Volume 27, No.

11, 2006

Edited by ~ Édité par

Eddy S.W. Ng
Trent University

GENDER AND DIVERSITY IN ORGANIZATIONS

GENRE ET DIVERSITÉ DANS LES ORGANISATIONS

Proceedings of the Annual Actes de Congres annuel de la

Conference of the Administrative Section Genre et diversité dans les
Science Association of Canada, organisations de l’association
Gender and Diversity in Organizations des sciences administrative du Canada
Division

Banff, Alberta Banff, Alberta

June 3-6, 2006 3-6 Juin 2006

Copies of these proceedings are available Des exemplaires de ces Actes peuvent être
online. connectés.
 AACKNOWLEDGMENTS ~ REMERCIEMENTS
I would like to thank all the authors who
submitted their work to our division.
I am also indebted to everyone who
answered my call for reviewers. Your
contribution to the conference and division
is gratefully appreciated.
I would be remiss if I also did not thank
Jenni Hantula, my incredibly capable
research assistant. Finally, I would like to
acknowledge the generous resources
provided by the Office of Research at Trent
University.
Ed Ng
Trent University
Academic Reviewer
J'aimerais remercier tous les auteurs qui ont
soumis leur travail à notre division.
Je suis aussi redevable tout le monde qui a
répondu mon appelle des critiques. Votre
contribution à la conférence et à la division
est appréciée avec reconnaissance.
Je serais négligent si je n'ai pas remercié
aussi Jenni Hantula, mon aide de recherche
incroyablement capable. Finalement,
j'aimerais reconnaître les ressources
généreuses fournies par le Bureau de
Recherche à l'Université de Trent.
Ed Ng
L'Université de Trent
Responsable de l'évaluation
 REVIEWERS ~ ÉVALUATEURS

Kara Arnold Memorial University

Myrtle P. Bell University of Texas at Arlington
Ida Berger Ryerson University
John Douglas Bishop Trent University
Lynn Bradley Carleton University
Patricia Bradshaw York University
Merridee Bujaki University of Ottawa
Marie-Andrée Caron ESG-UQAM
Tania Casado University of Sao Paulo
Robin Church University of Winnipeg
Kelly Dye Acadia University
Lorraine Dyke Carleton University
Suzanne Gagnon McGill University
Gerald Hunt Ryerson University
Mark Julien Brock University
Rekha Karambayya York University
Waheeda Lillevik University of Michigan - Flint
Rosemary A. McGowan Wilfrid Laurier University, Brantford
Gloria Miller University of Regina
Albert J. Mills St Mary's University
Kim Morouney Wilfrid Laurier University
David Newhouse Trent University
Ranjna Patel Bethune-Cookman College
Kate Rowbotham University of Toronto
Julia Rowney University of Calgary
Mary Runte University of Lethbridge
Robert Runte University of Lethbridge
John Russell University of Saskatchewan
S.K. Saha Memorial University
Christine Silva Queen's University
Lisa Stickney Temple University
Peggy Wallace St Mary's University
Yang Yang University of Western Ontario
Margaret Yap Ryerson University
Asaf Zohar Trent University
 TABLE OF CONTENTS ~ TABLE DES MATIÈRES

BEST PAPER – PRIX DU MEILLEUR TEXTE

Financial Well Being and Women’s Retirement Satisfaction.......................................... 1
Gillian Leithman, Concordia University
Charles Draimin, Concordia University
Linda Dyer, Concordia University

BEST STUDENT PAPER – PRIX DU MEILLEUR TEXTE D'ÉTUDIANT

Do Expectations Concerning Absenteeism Differ Between Men and Women?
Indications from the Popular Press.................................................................................. 13
Eric Patton (Student), Concordia University
Gary Johns, Concordia University

Racial Inequality in Employment in Canada: Empirical Analysis and

Emerging Trends ............................................................................................................... 30
Mohammed Al-Waqfi, United Arab Emirates University
Harish C. Jain, McMaster University

Advancing Equity in Canadian Unions ........................................................................... 50

Janice Foley, University of Regina

Creating Diverse Non-Profit Boards: Engaging Multiple Dimensions of Power......... 64

Christopher Fredette (Student), York University
Patricia Bradshaw, York University
Sue Inglis, McMaster University

Human Resource Management Practices for Achieving Diversity Advantage:

Determinants and Outcomes of the Configuration of Diversity Management
Structures............................................................................................................................ 82
Alison M. Konrad, University Western Ontario
Yang Yang (Student), University Western Ontario
Cara Maurer (Student), University Western Ontario

What Women and Minorities Want: Work Values, Labour Market Confidence
and a Preference for Commitment to Diversity ............................................................ 101
Eddy S.W. Ng, Trent University
Greg J. Sears, Public Service Commission of Canada
Participation in Voluntary Training Activities in the Canadian Banking
Industry: Do Gender and Managerial Status Matter?................................................. 119
Stéphane Renaud, Université de Montréal
Lucie Morin, Université du Québec à Montréal
Julie Cloutier, Université du Québec à Montréal

Les Enjeux Organisationnels de la Diversité Culturelle: Portée et Limites des

Modèles d’Analyse Dominants........................................................................................ 130
Sid Ahmed Soussi, Université du Québec en Outaouais
Annie Côté, Université du Québec en Outaouais

More or Less: Hierarchical Differences in the Effects of Gender Composition

on the Hiring of Women Faculty .................................................................................... 144
Heather Geraci Perretta, Cornell University
Pamela S. Tolbert, Cornell University

ABSTRACTS ~ RÉSUMÉS

Emergence Of Human Diversity Strategies: The Role Of The Network As A

Mechanism......................................................................................................................... 163
Parshotam Dass, University of Manitoba

Do Women Managers Benefit From Women Only Management Training?

An Empirical Examination Of The Impacts Of One Program..................................... 164
Linda Scourtoudis, Carleton University
Lorraine Dyke, Carleton University

Strategic Responses To Complex Stakeholder Demands: A Case Study Of An

Aboriginal Nonprofit Organization................................................................................. 165
Bernie Williams, University of Lethbridge
Jill Finley, University of Lethbridge

Gender And Age Differences In Organizational Politics ..............................................166

Karen A. Sommerville, Carleton University
Lorrain S. Dyke, Carleton University

Gender As A Moderator Of Executive Career Success.................................................167

Barbara Orser, University of Ottawa
Catherine Elliott, University of Ottawa
Joanne Leck, University of Ottawa

Employment Equity And Backlash: An Examination Of The Perceptions Of

Employees Within The Federal Public Service: Preliminary Findings ....................... 169
Amy Warren, Memorial University
Travor Brown, Memorial University
ASAC 2006 Gillian Leithman
Banff, Alberta Charles Draimin
Linda Dyer
John Molson School of Business
Concordia University

FINANCIAL WELLBEING AND WOMEN’S RETIREMENT SATISFACTION

This study explored predictors of women’s retirement satisfaction. Survey

responses from 200 retirees revealed that the women were less financially
prepared and more financially anxious than men. Unlike their male counterparts,
their regrets over having delayed retirement planning affected their life
satisfaction. Recommendations for working women are proposed.

Introduction: Women and Retirement

Retirement from paid employment is a milestone event in a person’s life. Financial preparedness
for retirement means having an adequate income to sustain one’s current standard of living after one’s
working years are over. Unfortunately, a great number of people arrive at retirement financially ill
prepared for this stage of development. According to a Statistics Canada report (2001), 33% of families
whose main income earner is between 45-66 years of age will not have sufficient retirement income.
Forty three percent of these families have no private savings. Having an adequate income is particularly
problematic for female retirees – women are at higher risk than men of living an impoverished old age
(Perkins, 1995). The objective of this research is to study the factors that affect women’s financial
preparedness for retirement and their financial and psychological wellbeing in this stage of life.

Many of today’s female retirees belonged to a different generation, one that earned less money
than their male counterparts and placed their husband’s careers before their own. Women’s careers tended
to be shorter and more discontinuous than men’s, often interrupted by child bearing, child rearing and
caring for elderly family members (Timmerman, 2000; Townson, 2000). For a majority of present-day
female retirees, paid work was secondary to their primary role as wife and mother; in return for their
devotion, they expected financial support from their husbands for the duration of their lives (Perkins,
1992). These women’s life long dependence on their husbands for the management of all financial matters
placed them at a disadvantage in old age. Young working women today may face a similar future.
Women employees continue to have lower earnings than their male counterparts (Thornborrow &
Sheldon, 1995). Women still shoulder the primary responsibility for childcare, eldercare and housework,
and adjust their work schedules to accommodate these tasks by reducing the number of hours they work,
taking leaves of absence or leaving the workforce entirely (Statistics Canada, 2002). The consequences of
an interrupted work trajectory are greater than just forgone monetary gains, retirement benefits and
pension credits associated with employment. A woman’s absence from the workforce means that she also
loses seniority, impeding opportunities for promotions and pay raises (Hannon, 1995). In addition,
women live longer than men. Coupled with the increasing rate of later-life divorce, this figure translates
into women having an 85 percent chance of living out their golden years alone (Perkins, 1995). Thus even
today, young women face the grim possibility of living in poverty after retirement.

1
The Role of the Pension System

Public pension schemes were designed to ensure that most people would retire with some degree
of financial security. There are three tiers to the Canadian pension system: Old Age Security plan (OAS),
the Canada / Quebec Pension Plan (CPP/QPP) and private pension savings plans such as Registered
Retirement Savings Plans (RRSP) and employer pension plans. Indeed, the development of the Canadian
pension system over the last decades has generally improved financial preparedness for retirement.
According to Statistics Canada (1999), in 1980, 34% of Canadians over 65 years old lived below the low-
income cut-off. By 1996, this percentage had declined to 21%. Yet women have still been left behind. In
1997, Canadian women’s pension income was 61% that of men’s on average (Marshall, 2000).

The nature of the pension system helps explain why women continue to be disadvantaged
financially (McDonald, 2002; Street & Connidis, 2001). While the first tier of the system, the OAS, treats
Canadian men and women identically, it currently pays less than $600 a month, This, alone, is insufficient
to maintain a standard of living above the poverty level. Even augmented by the Guaranteed Income
Supplement (GIS), it provides a minimal level of retirement income. The second tier of the pension
system, C/QPP, is relatively disadvantageous for women because benefits are related to career earnings,
which reproduces, in retirement benefits, the lower career earnings of women. The third tier, private
pensions, is an arena in which income inequality is greatest between men and women. Employer pension
plans in the predominantly female service sector tend to be offered at unattractive terms or not offered at
all (Morrisette & Drolet, 1999). Men’s contributions to their RRSPs consistently exceed that of women’s.
Between the years of 1993 to 1999, men and women aged 35 to 44 contributed, on average,
approximately the same percentage of their earnings to their RRSPs, yet the average annual dollar value
of men’s contributions was $1,564 while women contributed only $848 (Régie des Rentes du Québec,
2004). The difference in average earnings, of course, explains the discrepancy. Forty-eight percent of
women have no private pension savings in contrast to only 24% of men (Statistics Canada, 2001).

Why Is Financial Preparedness So Difficult?

Traditional economic theories such as the life-cycle hypothesis are hard put to explain lack of
preparedness for retirement (Modigliani & Brumberg 1955). More recent work in behavioural economics
(Thaler, 1994) offers a more successful explanation. People’s rationality is constrained by time,
intelligence and the information at their disposal and thus they do not always make choices that are in
their long-term best interest as the life-cycle hypothesis suggests. Planning requires a great deal of work
(Rabin, 1998) and extensive knowledge of private and government pension plans (Lusardi, 1999). Lusardi
also points out that retirement is a once in a lifetime event, which does not afford people the opportunity
to learn from their mistakes. This is not surprising: financial planning, like any long-term decision-
making involves making simultaneous predictions of such personal imponderables as future earnings, age
of retirement, health in retirement, general retirement needs, and age of death. One must also predict stock
market returns, inflation and economic growth rates – areas where even the experts are often wrong
(Shulz, 1985).

Beyond the general difficulty of financial planning for retirement, Perkins (1995) suggests that
despite the advancements of the women’s movement of the 1960s, many women still believe the
“prevailing myth” that they will be cared for in their older age. It is this myth combined with women’s
fear of growing old that Perkins believes prevents women from asserting their independence and planning
for retirement. These social views are reflected in institutional structures. For example, in low paying,
non-unionized service jobs, pre-retirement education is not often available to women in the workplace
and the programs themselves are limited in scope when they are offered. In such circumstances, women’s
lack of financial preparedness is more acute than men’s.

2
 In sum, women’s lower lifetime earnings, interruptions from paid employment, and identification
with traditional gender roles will influence their ability to save for retirement. Thus we hypothesize: H1 -
Retired women will report a lower level of financial preparation for retirement than will men.

What is the effect of inadequate financial preparations for retirement? More specifically, do
women’s lower retirement incomes translate into a lower level of financial and psychological wellbeing in
retirement in comparison to male retirees? Are women less satisfied with their lives in retirement than
men? It might seem obvious that the decrease in income that comes with retirement would result in
anxiety about making ends meet and lower satisfaction (Choi, 2001). The relationship is not so simple,
however. Lowenstein, Prelec and Weber (1999) examined retired men’s perceptions of the sufficiency of
their income in comparison to the perceptions of working men. In contrast to working men, retired men
actually reported less anxiety about their finances than do those who are still employed. Financial anxiety
seemed to be greatest prior to retirement as people worry about whether or not they will have sufficient
income to fund all of retirement. Men with high-incomes and those who began saving for retirement at a
young age reported lower levels of financial anxiety than others. Once retired, however, the majority of
men quickly adapted to their new financial circumstances and reported relatively high levels of happiness.

Would the process be the same for retired women? Very few scholars have addressed this
question. Related research, though, suggests that there may be gender differences in the experience of
financial anxiety. In a comprehensive review of gender differences in well-being, Nolen-Hoeksema and
Rusting (1999) reported that women display higher rates of anxiety than do men across the life span.
Other research findings agree that women consistently report higher levels of anxiety than do men
(Feingold, 1994). In addition, despite Lowenstein et al’s (1999) findings that men do adjust to lower
retirement incomes, we have seen that women’s retirement incomes are lower yet. A woman’s longer life
expectancy and the increased risk of outliving her money—coupled with her greater propensity to
experience anxiety—are compelling factors that may influence her feelings of financial anxiety. Hence:
H2: Retired women will report greater levels of financial anxiety than will retired men.

Lowenstein et al (1999) also examined whether retired men display financial regret—reget about
past financial decisions regarding their retirement savings. Their findings indicated that the later an
individual begins to save for retirement, the greater the likelihood of regret, and that men who have
greater income from pension plans and savings display less financial regret. Interestingly, however, they
found that the financial regret expressed by retired men was surprisingly low on the whole. The
researchers noted that the men in their study did display some regret over how well they had implemented
a savings strategy, but they did not exhibit regret over how much they had in fact saved for retirement.

An examination of the financial literature regarding investor behaviour lends support to the
relatively low level of financial regret displayed by the men in the Lowenstein et al (1999) study.
According to Shefrin and Statman (1985) who studied gender differences in investing decisions, men who
have made poor financial decisions often wish to avoid a public admission that their judgment was
incorrect and will go to great lengths not to admit that they have made mistakes. Thaler (1980) echoes the
sentiment, maintaining that people will go to great lengths to avoid feelings of financial regret. Moreover,
he explains that having to admit that one has made mistakes is made more difficult by having to
acknowledge such mistakes to others. The idea of confronting the reality that one did not make an
appropriate financial plan for retirement is a regret that carries substantial consequences for one’s
household’s standard of living in retirement. Moreover, once the person has retired, there are few if any
opportunities to recover and regain one’s pre-retirement standard of living. On the other hand, if one does
not interpret one’s financial preparations for retirement as inadequate, then there is no need to feel regret
for missed opportunities.

3
 Today’s male retirees were socialized to identify with the role of economic provider. They were
taught that their primary duty in life was to bring home a paycheque and financially provide for their
families. Women, on the other hand, were socialized to embody the role of wife, mother and caregiver.
The women’s movement of the 1960s increased women’s choices, but the role of men never deviated
from that of breadwinner (Pinquart & Sorensen, 2000). Thus a man is more likely to feel accountable to
others for his failure to provide a comfortable retirement. Following Shefrin & Statman (1985), this may
mean that he will go out of his way to avoid feelings of regret so as not to have to admit to himself or
others that his financial plans were inadequate. A woman might be less driven to avoid financial regret.
Thus: H3 - Women will report higher levels of financial regret than men.

Life Satisfaction in Retirement

The majority of people lead fulfilling lives despite differences in material wealth (Diener &
Diener, 1996). The role of income on the wellbeing of retired men and women is not clear (Nolen-
Hocksema & Rusting, 1999). Calasanti (1996) argued that a retirement model that was initially designed
for white middle class men could not predict the wellbeing of women retirees. As such, one should expect
gender differences in the predictors of retirement wellbeing. A recent study by Bourque, Pushkar,
Bonneville & Béland (2002) lent support to Calasanti’s theory. The findings from the study revealed that
different variables influence the life satisfaction of men and women in old age. Although income was
found to predict the life satisfaction of both men and women, there was a stronger association for men.
Bourque et al (2002) point out that it is paradoxical that despite older women’s lower economic status,
widowhood and deteriorating health, they do not report that they are any less satisfied with their lives than
are men.

Following Calasanti (1996) and Bourque et al (2002) we believe that the interplay between
feelings of financial wellbeing and life satisfaction will differ for retired women and men, that is: H4 –
The relationships between financial anxiety, financial regret and life satisfaction will be moderated by
gender.

Method

Participants

Female and male retirees were recruited from community centers, retirement clubs, libraries and
seniors’ events in four Canadian provinces. Additional participants were recruited by means of snowball
sampling, that is, participants were asked for the names of other retirees who might be interested in
responding to the questionnaire. In the various locations, 550 paper questionnaires were distributed and
197 were returned, of which 9 could not be used because of incomplete data (for example, the respondent
did not indicate gender). The usable response rate to the paper questionnaires was 36%. In addition,
notices and flyers were placed in some locations inviting participation online. Twelve online responses
were received.

The sample was composed of 83 men and 117 women. Forty-six percent of the female retirees
were married; in contrast, the majority of male retirees, 85%, were married. With respect to education,
42% of the female sample and 56% of the male sample had received an undergraduate or post-graduate
university degree. The mean age of the women was 68.6 years old, and their average retirement age was
59.6. The men had a mean age of 70.5 years old, and the average age at which they had retired was 61.0.

4
Measures

All respondents were asked to complete a confidential self-report questionnaire. The distribution
of this questionnaire was effected in a variety of ways including individual administration in a face-to-
face context, by mail, and over the Internet. Questionnaires were printed in a large typeface to aid
responses among older respondents.

Financial Preparedness. Four tailor-made questions related to the age at which the respondents
began to prepare for retirement. The first form of preparation was saving for retirement, and response
categories were: “Under 30 years old,” “30-39 years old,” “40-59 years old,” “60 or over” and “Never.”
Three other forms of preparation were included, the age that the respondent first calculated the amount of
money needed to fund retirement activities, the benefits to be received from the Canada / Quebec Pension
Plan, and the amount of money needed to maintain his or her standard of living in retirement. Response
categories were coded as: “Under 40 years old,” “40-59 years old,” “60 or over” and “Never.” We also
combined all four questions into an index of financial preparedness. This index was found to be internally
consistent, with a reliability coefficient of .76.

Financial Anxiety. Four questions drawn from Lowenstein et al’s (1999) money anxiety index
were used to assess people’s subjective perceptions of the sufficiency of their retirement income.
Participants were asked to rate on a four-point scale their extent of agreement with each of these four
statements: “I feel anxious about money,” “With my current income I have no trouble making ends
meet,” “I have plenty of money to do the types of things that I enjoy,” “ I have sufficient income for my
wants and my needs.” The last three items were reverse scored. In this sample, the financial anxiety index
achieved an internal consistency reliability of .84.

Financial Regret. The financial regret scale also was drawn from an existing instrument
(Lowenstein et al, 1999), and assessed the extent to which people had experienced regret over the way
they had managed their income for retirement. The scale consisted of ten items and respondents were
asked to indicate on a four-point scale the extent to which they agreed with statements such as: “I should
have started saving earlier,” “I should have gotten more professional help in planning for retirement” and
“I should have spent more time with my family even if it meant earning less.” The average of the ten
items was the financial regret score, which was found to have acceptable reliability in this sample (alpha
= .86).

Life Satisfaction. The measure of life satisfaction was based on Floyd, Haynes, Doll,
Winemiller, Lemsky, Burgy, Werle and Heilman’s (1992) life satisfaction in retirement subscale, taken
from their Retirement Satisfaction Inventory. The nine-item scale is designed to measure aspects of a
retiree’s satisfaction with financial resources, interpersonal relationships, the adequacy of governmental
aid and community services and access to transportation. It should be noted that Floyd et al.’s (1992)
scale is composed of 10 items. One item in which respondents were asked to indicate how satisfied they
were with their marriage was, however, unintentionally left out of the current scale. Participants rated
their level of satisfaction on a six-point scale, and the internal consistency reliability of the index was .75
in this sample.

The questionnaire also asked participants for information about their gender, age, age of
retirement, income during their working years, income at retirement and other relevant descriptors.

5
 Results

Did the women in our sample start to prepare for retirement later than the men? We found that in
some respects, they did. We compared men and women on the index of financial preparedness that was an
average of all four preparedness items. We found that women, with a mean of 2.84 on the four-point
scale, were significantly less prepared than men, whose mean score was 2.49 (see Table 1). (Note that for
this measure, the higher the number, the less the financial preparedness). We concluded that there was
support for our first hypothesis about gender differences in financial preparations for retirement.

Follow-up chi square analyses revealed significant differences in the age at which women and
men first calculated the money they would need to fund retirement activities (p = .011), as well as the age
at which they first calculated the amount of money they would need to maintain their standard of living (p
= .008). Inspection of the data showed that the main differences occurred because 49% of the women had
never done calculations of funding retirement activities, and 39% had never done calculations related to
their standard of living. With regard to the other measures of financial preparedness, there were no male /
female differences in the data. Thus the distribution of ages at which women started to save for retirement
and first calculated the C/QPP was the same as men.

To examine our second hypothesis, we tested the extent to which women reported greater levels
of financial anxiety than men (see Table 1). The participants in the study revealed that, in general, they
were not particularly anxious about their finances. The average for women was 2.22, which was just
above the neutral point on the scale; men rated their anxiety as 2.03, also close to the neutral point. Even
so, there was a significant difference between these two means. The women were more anxious than the
men about their retirement finances (t = 2.03; p = .027, one tailed). Thus our second hypothesis was
supported. On the other hand, there was no difference in the level of financial regret displayed by men
and women in our sample (t = .069; n.s). Again, both means were close to the midpoint of the scale. We
rejected our third hypothesis.

Although it was not one of our hypotheses, we checked to see whether there were gender
differences in the reported life satisfaction of the men and women in our sample. We found that both
groups said that they were quite satisfied. The means were 4.55 (men) and 4.52 (women) on a six-point
scale, with no significant statistical difference between the two means, indicating that both groups were
satisfied with their lives in retirement to an equal extent.

Table 1

Gender differences in four major dependent variables

Construct Women Men p

Financial preparedness 2.84 2.49 .0005
(.78) (.73) one-tailed
Financial anxiety 2.22 2.03 .027
(.65) (.69) one-tailed
Financial regret 2.10 2.11 n.s.
(.57) (.45)
Life satisfaction 4.52 4.55 n.s.
(.81) (.73)

6
 For our final hypothesis, we were interested in exploring the associations between financial
anxiety, financial regret and the life satisfaction of men and women retirees. First, a multiple regression
was performed with life satisfaction as the dependent variable and financial anxiety and financial regret as
the predictors. We found that the overall equation was significant (p < .001), and that both predictors
contributed inversely to the variance in the dependent variable (financial anxiety: β = -.374; p < .001;
financial regret: β = -.204; p = .005). As regret and anxiety increased, life satisfaction decreased. When
we split the participants into two sub-samples, it was interesting to note that gender had an impact. The
pattern in the relationships differed for the men versus the women. For the women, both regret (p = .014)
and anxiety (p = .001) played a role in predicting life satisfaction. For the men, however, all the variance
in life satisfaction came from the impact of financial anxiety (p < .001). Financial regret among men was
unrelated to their satisfaction with retirement life (p = .263, n.s.).

In sum, significant gender differences were found between the age at which men and women
begin to calculate the money they would need for retirement activities and maintenance of their standard
of living, with women lagging behind men in their financial preparations. No gender differences were
observed on the financial regret measure, and female retirees were found to be just as satisfied with their
lives as were the male retirees. Although the majority of men and women in the sample did not exhibit
financial regret or financial anxiety, those who did also exhibited a decrease in life satisfaction. Moreover
women, but not men, who were regretful of their earlier retirement-preparation behaviour, displayed
decreased life satisfaction.

Discussion

Unlike anxiety, regret and satisfaction, which were are features of life in retirement, financial
preparedness for retirement requires planning—a combination of thinking and acting—over time.
Looking at the individual questions, women were as well-prepared as men in two respects, less well-
prepared in two others. Women tended to begin saving for retirement as early as men, and they also gave
some thought to one source of the retirement income (Canada/Quebec Pension Plan). However, in terms
of thinking about what they would actually need in retirement—to pay for retirement activities and, more
generally, to maintain their standard of living, on average women started much later than men. One is
struck, moreover, by the large number of women who never thought about these things at all. Averaging
these two measures, 44% of women never thought about what they would need. Only half as many men,
22%, were in that category. The index of financial preparedness better captures the multi-faceted nature
of the retirement planning process and it confirms that women are financially less well prepared than
men.

Women were more anxious about the financial aspects of their retirement. This is not surprising
given the discussion in the previous paragraph. Though, on average, women may have started saving as
early as men, they hadn’t thought as much about what they would need in retirement, so it is much more
likely they arrived at that stage of life with too little. And we should bear in mind that even if they saved
the same percentage of their income as men in their earning years, women would be in a far less
comfortable situation than would men in retirement. Their income was smaller and lives in retirement can
be expected to be longer. All other things equal, a woman would need to have accumulated more than a
man for the same level of financial security.

Insofar as experiencing regret, there is no difference between men and women. We had
hypothesized that the different social roles of men and women—provider versus caregiver—would lead to
different propensity to experience financial regret. It was presumed that the more public nature of the
breadwinner role would mean that there was a premium on saving face and thus not admitting failure to
provide. Regret would seem to be a well-managed emotion—regret was not much in evidence in our

7
sample—but the connection to one’s social role is not clear. One could build a different case: women are
not as well prepared financially for retirement as men, but because they give greater importance to their
role as caregiver, their failure to provide—either towards the household or, if they are alone, for
themselves alone—is less important than it would be for a man. They might feel regret for failed personal
relationships with children, friends or a divorced spouse, but the financial side of things is not a focus for
thought or feeling. Regret over non-financial issues is not measured in this study. It is particular striking
that women seem to have the same experience of financial regret as men given that, objectively, they had
much more cause to experience regret since the financial situation of women in retirement is, overall,
more precarious than men. This gives some support to the folk notion that you don’t have to be rich to be
happy.

On the other hand, as we saw in our final analysis relating anxiety, regret and gender to life
satisfaction, regret does have a differential impact on life satisfaction as between men and women. For
women—but not for men—regret has an inverse impact on life satisfaction in retirement. The anomalous
results for regret deserve more discussion and the empirical literature investigating regret provides a
useful framework. Gilovich & Medvec (1995) maintain that regrets of inaction, that is opportunities that
are forgone or things or situations that people have desired but have failed to act upon, tend to be more
burdensome than regrets of action, decisions that were taken that did not turn out as anticipated or that
resulted in failure. Regrettable actions are believed to be easier to rectify over time because one can make
necessary life changes so as to resolve the regrettable experience, find a lesson to be learned from a bad
decision and/or engage in dissonance reduction.

Regrettable inactions, however, are not as easy to resolve due a number of psychological
mechanisms that impede the reconciliation of the regretful inaction. With the passing of time regrettable
inactions tend to take on a powerful and intense momentum, as the more distance there is between us and
the missed opportunity, the better convinced we become that we would have been great at that job,
charming on that date, and at the top of our class if admitted to the prestigious school. Secondly, regrets
of inaction are difficult to move beyond because of the imagined consequences of what might have been.
Regrettable actions have known and finite consequences, regrettable inactions, however, are only
restricted to the contents and creativity of one’s own imagination (Gilovich & Medvec, 1995).

The value of understanding the consequences of regrets of inaction are of interest because it
serves to provide a comprehensive explanation as to why regret was related to life satisfaction for women
but not for men, Detailed analysis of the financial regret measure revealed that the female retirees were
most regretful of three of the nine items that composed the financial regret scale, all of which were regrets
of inaction. These women were primarily regretful of not having calculated the amount of money that
they would need in retirement, not starting to save at an earlier age and not being knowledgeable about
different savings and investment options. It is interesting to note that the former two findings (age that
people start to save for retirement and calculating the amount of money one needs to save in order to have
adequate retirement income) are the two items from the financial preparedness scale that indicated that
the men were better financially prepared for retirement than the women. It appears that adequate financial
preparation moderates financial regret and is what contributed to separating the financially satisfied men
from the financially dissatisfied women.

The findings from the present study suggest that men have more efficiently prepared for their
financial well-being in old-age, thereby shielding them from any regret that they might have otherwise
reported had they not been as prepared. Hayes (1995) maintains that inherent to the role of economic
provider is the responsibility of preparing for retirement. Thus, men have had more time and greater
opportunity to prepare for their future financial security. Men have therefore less reason to report
financial regret in comparison to women, seeing as they have been socialized to prepare for this later
stage of development.

8
 Additionally, the women in the study reported that one of their greatest financial regrets was not
being more knowledgeable about different savings and investments options. This concurs with the
findings of the National Center for Women and Retirement Research, the results of which indicate that
out of a national sample of 1,100 American women surveyed, 74% in contrast to 33% of men, reported
that their lack of knowledge regarding investment vehicles has been one of their greatest obstacles in
planning for their future financial security (cited in Hayes, 1995). In light of Gilovich & Medvec’s (1995)
association between regrets of inaction and their relationship to fear and lack of confidence, it is not
surprising that three-quarters of the women who were surveyed reported that fear of making a mistake
was their greatest impediment to planning for a financially secure retirement. Additionally, 54% of these
women have actually postponed planning for their financial well-being in old-age for fear of failing.

As in any scientific investigation, a number of limitations must be addressed. First, the financial
preparedness index is somewhat narrow; in particular, the existence of private pensions is a notable
omission. A second limitation of the current investigation is the manner in which gender differences were
measured. According to Hatch (2000), in order to establish gender differences in retirement, men and
women need to be matched on characteristics such as previous occupational status, age, length of time
since retirement and labour force history. Variation in any one of these dimensions, or a combination of
all of them, may therefore be the source that is contributing to the observed differences between men and
women. The current study did not match retired men and women on any of the aforementioned
characteristics. The greatest gender difference that existed between the men and women in the current
study was that 43.2% of the female retirees were married and 29.5% were widows in contrast to 87% of
the male retirees who were married and 3.7% were widowers. This may explain why the women in the
current study exhibited financial regret, whereas the men did not. The women have more reason to be
financially regretful than do the men, as being, single, divorced or widowed, for a woman, translates into
less retirement income in comparison to their married male counterparts.

Lastly, the generalizability of the findings of the current sample of primarily Quebec retirees is
questionable as these retirees are substantially more affluent than the average retired Quebecer. In 2001,
the average annual retirement income of a woman living in Quebec was $18,119 and the average annual
retirement income of her male counterpart was $28,654 (Régie des Rentes Quebec, 2003). In comparison,
the average annual retirement income of a female retiree in the current study fell between $40,000 to
$49,000 and the average annual retirement income of a male retiree in the current study fell between
$50,000 to $59,999. One of the reasons for this discrepancy, however, may be due to the wording of the
question, which asked respondents to indicate their approximate income or household’s income prior to
retirement. The wording of the question does not allow us to distinguish between the income of a single
individual and the income of his/her household. Thus, the true difference between a retired woman’s and
a retired man’s annual retirement income may actually be masked. Future research should consider an
individual’s income and that of the household’s separately. In addition, the sample was generally more
educated than norm.

Conclusion

The results of the present study indicate that women are less well-prepared for retirement than
men. They also tend to exhibit a greater degree of financial anxiety than men in retirement. Women do
not, on average, report more financial regret than men, nor is their level of life satisfaction in retirement
different than that of men. Further analysis indicates that to the extent that women—though not men—
experience financial regret, it will negatively affect the perception of the quality of their lives and their
life satisfaction in retirement, indicating that financial security is an integral component of a retiree’s
well-being, and suggesting that financial security is essential for successful aging. At the time that the
study was devised, it was, to our knowledge, one of the first investigations that attempted to assess the

9
financial regret and financial anxiety of female retirees. Therefore, our goal was to detect meaningful
patterns of association that would foster more specific causal hypotheses for future research. Although no
one particular investigation is sufficient to draw conclusions regarding the phenomena under study, the
present findings do suggest that future investigation is needed to assess the long-term effects of
inadequate financial preparation on the subjective well-being of men and women.

There are numerous reasons why women lag behind men in their financial preparations for
retirement. Financial security, however, is a fundamental component of retiree’s well-being. While the
current generation of female retirees is beginning to understand the necessity of adequate retirement
income, it is essential that future cohorts of young women be well informed of how to facilitate the
development of a financially secure old-age. A campaign to assist future generations of Canadian women
achieve adequate financial security should include the efforts of three parties: the federal and provincial
governments, business, and Canadian women themselves.

The Canadian governments and business can greatly benefit from instituting policies that take
into account women’s relationship to the workforce. Canadian women, however, must take responsibility
for their own future financial security. Whether or not Canadian women choose to marry and devote
themselves to their husbands and children is a matter of personal choice. They are, however, strongly
encouraged to inform themselves of their savings options regardless of whether or not they are in the
work force. Thus, whether a woman is a salaried worker, self-employed, a homemaker or a student, she is
strongly advised to familiarize herself with the regulations of public and private pension schemes. The
Quebec government, for example, has done an excellent job at compiling information specific to women
and retirement benefits. The government does emphasize that gender is an issue that will affect women’s
future financial security. As such, it promotes strategies to help women plan accordingly.

At present, women live longer, earn less and do not do not plan as efficiently for their financial
well-being in retirement as do men. Women must accumulate more to sustain themselves for a long
retirement. They have to start early. Even if a woman is married, she should still manage her finances as if
one day she will be on her own. Although a terribly daunting thought, it is a protective measure that will
safeguard women against divorce and widowhood should such an event occur. Divorce or widowhood
have left past generations of women financially ill-prepared for retirement and terrified of making any
financial decisions for fear of failure. Many of these women never made any financial decisions because
their husbands always managed the household’s finances. It may involve taking more risks with their
money than they do now. It certainly requires becoming more familiar with the workings of the financial
markets.

Traditional gender roles, institutional barriers and a lack of financial confidence have hindered
many Canadian women from planning for their financial well-being in old-age. Research has shown that
for those female retirees who have not adequately planned for their financial well-being in retirement, the
consequences are both psychologically and financially catastrophic. The experiences of these older
women can serve to enlighten younger generations of Canadian women and provide an example from
which to learn. As more and more women become aware of the psychological barriers that have impeded
them from taking control of their financial destinies, they are making saving for retirement a top priority.
What is most encouraging is that financial management has empowered these women and left them with a
sense that they are in control of their lives. Although money is no guarantee for a happy retirement,
financial security is a good place to start.

10
 References

Bourque, P., Pushkar, D., Bonneville, L., & Béland, F. (in press). Contextual effects on life satisfaction of
older men and women.
Calasanti, T. M. (1996). Gender and life satisfaction in retirement: An assessment of the male model.
Journal of Gerontology, 51B (1), S18-S29.
Choi, N. G.(2001). Relationship between life satisfaction and postretirement employment among older
women. International Journal of Aging and Human Development, 52 (1), 45-70.
Diener, E., & Diener, C. (1996). Most people are happy. Psychological Science 7 (3), 181-185.
Feingold, A. (1994). Gender differences in personality: A meta-analysis. Psychological Bulletin, 17, 532-
540.
Floyd, F. J., Haynes, S. N., Doll, E. R., Winemiller, D., Lemsky, C., Burgy, T. M., Werle, M., &
Heilman, N. (1992). Assessing retirement satisfaction and perceptions of retirement experiences.
Psychology and Aging, 7, 609-621.
Gilovich, T., & Medvec, V. H. (1995). The experience of regret: What, When, and Why. Psychological
Review, 102 (2), 379-395.
Hannon, K. (1995). Why the rules are different for women. Working Woman, September, 20-24.
Hatch, L. R. (2000). Adaptation to retirement. In J. Hendricks (ed.), Beyond Gender Differences:
Adaptation to Aging in Life Course Perspective (pp. 134- 174). Amityville, NY: Baywood Publishing
Company, Inc.
Hayes, C. (1995). Exclusive self-test helps women identify aspects of their personality that motivate and
hinder them in making smart money choices. Retrieved on October 2, 2004 from
http://www.worknmom.com/worknmom/finance_target.html
Lowenstein, G., Prelec, D., & Weber, R. (1999). What, Me Worry? A Psychological Perspective on
Economic Aspects of Retirement. With Comment by Matthew Rabin. In H. J. Aaron (Ed.),Behavioral
Dimensions of Retirement Economics (pp.215-251). Washington, D.C: Brookings Institution Press.
Lusardi, A. (1999). Information, Expectations and Savings for Retirement.
with comment by William G. Gale. In H. J. Aaron (ed.), Behavioral Dimensions of
Retirement Economics (pp. 81-115). Washington, D.C : Brookings Institution Press.
Marshall, K. (2000). Incomes of younger retired women: The past 30 years [Electronic version].
Perspectives on Labour and Income, 1 (3).
McDonald, L. (2002).The Invisible Retirement of Women. Research Paper, No. 69. SEDAP (McMaster
University).
Modigliani, F. & Brumberg, R. (1955). Utility Analysis and the Consumption Function: An Interpretation
of Cros-Section Data. In K. Kurihara (ed.) Post Keynesian Economics. London: Allen & Unwin. 1955
Morrissette , R. & Drolet, M. (1999). The evolution of pension coverage of young and prime aged
workers in Canada. Ottawa: Statistics Canada.
Nolen-Hoeksema, S., & Rusting, C. L. (1999). Gender differences in well-being. In Kahneman, D.,
Diener, E., & Schwarz, N. (Eds.). Well-being: The foundations of hedonic psychology (pp. 330-350). New
York: Russell Sage Foundation.
Perkins, K. (1992). Psychosocial implications of women and retirement. Social Work, 37, 526 – 532.

11
Perkins, K. (1995). Social (in)security: Retirement planning for women. Journal of Women and Aging, 7,
37 –15.
Pinquart, M., & Sörensen, S. (2000). Influences of Socioeconomic staus, social network, and competence
on subjective well-being in later life: A Meta analysis. Psychology and Aging, 15 (2), 187-224.
Rabin, M. (1998). Psychology and economics. Journal of Economic Literature, 36 (1) Research Library
Core, pg. 11.
Régie Des Rentes Québec (2003). Pour une retraite réussie - Planifiez et épargnez tôt et régulièrement.
Retrieved April 15, 2003, http://www.rrq.gouv.qc.ca/an/regie/16_06_02_30.htm
Régies des Rentes Québec (2004). Will Gender affect your retirement income? Prospects. Retrieved April
2004 http://www.rrq.gouv.qc.ca/an/services/Publications/prevoir_03_03_a.pdf
Schulz, J. H. (1985). The Economics of Aging (3rd ed.). Belmont, CA: Wadsworth.
Shefrin, H., & Statman, M. (1985). The disposition to sell winners too early and ride losers too long:
theory and evidence. The Journal of Finance, VOL XL, NO. 3, 777-791.
Statistics Canada (1999). A Portrait of Seniors in Canada, 3rd edn. Catagogue 89-519-XPE. Ottawa:
Statistics Canada.
Statistics Canada (2001). The Assets and Debts of Canadians: Focus on private pension savings.
Retrieved September 15, 2003, from http://www.statcan.ca/english/freepub/13-596-XIE/0100113-596-
XIE.pdf
Statistics Canada (2002). General social survey cycle 16. Caring for aging society, catalogue 89-582-
X1E. Ottawa: Statistics Canada. Retrieved on October 2, 2004 from
http://www.statcan.ca/english/freepub/89-582-XIE/pdf/89-582-XIE03001.pdf
Street, D., & Connidis, I. (2001). Creeping Selectivity in Canadian women’s pensions. In J. Ginn, D.
Street, & S. Arber (Eds.), Women, Work and Pensions: International Issues and Prospects (pp. 158- 178).
Philadelphia, PA: Open University Press.
Thaler, R. H. (1980). Toward a positive theory of consumer choice. Journal of Economic Behavior and
Organization, 1, 39-60.
Thaler, R.H. (1994). Psychology and savings policies. The American Economic Review 84 (2), 186-192.
Thornborrow, N. M., & Sheldon, M. B. (1995). Women in the labour force.
In J. Freeman (ed.), Women: A feminist perspective, 5th ed. (pp. 197-219).
Timmerman, S. (2000). Planning for the future: A focus on mature women. Journal of Financial Service
Professionals, 54 (1), 22- 24.
Townson, M. (2000). Reducing poverty among older women: the potential of retirement incomes policies.
Ottawa: Status of Women Canada. Retrieved June 16, 2004, http://www.swc-
cfc.gc.ca/pubs/0662659271/200008_0662659271_1_e.html.

12
ASAC 2006 Eric Patton (Doctoral Student)
Banff, Alberta Gary Johns
John Molson School of Business
Concordia University

DO EXPECTATIONS CONCERNING ABSENTEEISM DIFFER BETWEEN

MEN AND WOMEN? INDICATIONS FROM THE POPULAR PRESS.

Through a qualitative content analysis of newspaper articles that have appeared

in the New York Times over the last one hundred years, this study explores the
different expectations surrounding absenteeism for women as compared to men.
The paper suggests that a separate absence culture exists for women that may
legitimize their absenteeism, but may also perpetuate gender stereotypes and lead
to gender discrimination against women.

Introduction

A common finding in absenteeism research is that women are generally more absent from work
than men. In a research area that has occasionally been criticized for its emphasis on correlational designs
that have not generally provided strong results (Martocchio & Harrison, 1993; Nicholson, 1993;
Nicholson & Martocchio, 1995), the difference in absence along gender lines has been consistent.
Despite the regular findings, there has been an unfortunate lack of theorizing and research on why these
differences exist (Johns, 2003). While the temptation is often to ascribe these differences to simple
explanations involving issues such as childcare, studies that have addressed many reasons that would
appear to make intuitive sense have demonstrated inconsistent or contradictory results. Furthermore, few
if any studies have considered the consequences for women resulting from their greater absenteeism.

The purpose of this paper is twofold. First, the question of why women have more workplace
absence will be explored. Rather than suggesting that women’s higher absenteeism is due to childcare or
stress, the proposition that women are subject to different expectations and norms surrounding attendance
will be advanced. These different social expectations are likely based on widely held gender stereotypes.
Second, to the extent that expectations differ, and women are perceived to be more absent than men, this
paper will also consider what this means for women in the workplace. If the norms concerning
absenteeism of women are built on stereotypes, their perpetuation may be harming women in other areas
of employment.

In order to explore these issues, a content analysis of newspaper articles featuring women’s
absenteeism appearing in the New York Times from 1900 to 2004 will be undertaken. Although not a
traditional method in absenteeism research, a content analysis of widely diffused documents represents a
valuable pathway to understanding beliefs and norms surrounding absence from work, and can highlight
potential outcomes for absentees. As research areas, gender and absenteeism are distinguished by the fact
they are familiar concepts to the average person. Unlike many topic areas in the organizational sciences,
absenteeism is part of our everyday vocabulary and it is reasonable to suggest that each of us have had
close encounters with absenteeism in our lifetime. This familiarity and public nature of absenteeism is
reflected in the impressive amount of news coverage the concept has received over the years. In fact, over
the last 100 years, there have been over 3,000 articles in the New York Times featuring absence from
work. The prominence of absenteeism in the press represents an excellent opportunity for researchers to
study how the concept is represented in the public domain. Similar to absenteeism and to an even greater
extent, gender is a salient and familiar topic at the public level that most people feel they have expertise

13
on, even if the topic is not always well understood (Deaux & Lafrance, 1998). Given that gender is also
an issue regularly treated in the press (Fowler, 1991), analyzing how gender and absenteeism have been
combined provides a novel pathway toward uncovering social expectations and representations of
workplace absence by women.

From a management perspective, Abrahamson (1991, 1996) suggests that actual choices and policies
within organizations can be influenced by outside forces, including the popular press. It is quite clear that
business leaders read major newspapers more regularly than academic business research. In this sense, the
popular press, best sellers, and business gurus can have much more of an impact on what actually happens
in organizations than academic research (Abrahamson, 1991; 1996). Abrahamson advocates studying the
rhetorics disseminated through the mass media in order to get research into the mainstream, and notes that
the rhetorics present are a manifestation of the ideologies underpinning organizational topics of interest.
Given the regular coverage of absenteeism in the news, its presentation can influence and also reflect how
managers view it. As newspapers are shapers and mirrors of social values and beliefs on a wide range of
topics (Fowler, 1995; Krippendorff, 2004), a content analysis spanning a hundred years focusing on how
gender and absenteeism have been combined in the popular press can provide insight into stereotypes in
operation, the existence of a separate absence culture for women, and the consequences of such an
absence culture.

Prior to reviewing the articles featuring both gender and absenteeism, the next section will
describe the existing findings on women’s absence in organizational behaviour research, underline the
lack of clear explanation as to the results, and suggest mechanisms that may be driving expectations
surrounding absenteeism by women.

Absenteeism and Gender: Consistent Findings but Little Explanation

Absenteeism can be defined as "an individual’s lack of physical presence at a given location and
time when there is a social expectation for him or her to be there" (Martocchio & Harrison, 1993, p.263).
In breaking down this definition, two separate elements can be identified: 1) a lack of presence, and 2) a
social expectation to attend. As previously noted, from a “lack of presence” perspective, the fact that
women are more absent than men is a common finding in absence research (Johns, 2003). Numerous
studies have demonstrated the relationship, and the link has been verified meta-analytically (Côté &
Haccoun, 1991). Despite the fairly consistent findings, there has been a disappointing lack of theorizing
on why this relationship exists. There are, however, several presumptions and stereotypes that operate
when researchers try to explain women's absence. First, women's more fragile health has been pointed to.
Although research has shown that women have a better mortality rate than men, their morbidity rate,
defined as generalized poor health, is worse than men’s (Rodin & Ickovics, 1990). Women tend to suffer
from more symptoms, more insomnia, more migraines, more chronic conditions, more depression, and
have more doctor visits than men (Nelson & Burke, 2002; Rodin and Ickovics, 1990). Empirically,
however, results linking women's higher absence to poorer health have been inconsistent. Côté and
Haccoun (1991) did not find any evidence that ill health could explain their meta-analytic finding that
women are more absent than men. Furthermore, studies by Leigh (1983), Scott and McClellan (1990),
Vistnes (1997), and Mastekaasa (2000) present findings on illness absence and gender that are sometimes
contradictory. Scott and McClellan (1990) found that women report more illness absence than men,
while Leigh (1983) found that illness absence for men and women were the same. Mastekaasa (2000) also
found that illness absence for men and women were equal, except in the year of childbirth for women or
for single mothers in general. Vistnes (1997) found that women were more absent due to chronic illness
while men were more absent due to acute illnesses. The methodologies employed in many of these
studies also raise questions as to whether or not illness absence was accurately measured. Furthermore,
researchers have also found that men are certainly not immune to health problems, that differences in

14
morbidity rates between men and women are shrinking, and that working women are actually healthier
than women who do not work (Nelson & Burke, 2002; Rodin and Ickovics, 1990). Overall, the existing
research does not provide clear evidence or clear explanations as to women's purported higher rate of
illness absence.

A second widely held assumption used to explain women's higher rate of absence is that they bear
the brunt of family and childcare responsibilities. Several authors have suggested that women face a
double burden in terms of work and family duties that does not apply to men (Bratberg, Dahl, & Risa,
2002; Mastekaasa, 2000; Nelson & Burke, 2002; VanderHeuval & Wooden, 1995). The assumption is
that the double burden and role conflict faced by women will increase stress (Nelson & Burke, 2002) and
lead to increased absence. As such, women are deemed to have more family responsibilities and greater
stress because of it. While this argument has intuitive appeal, empirical findings do not consistently
support the notion that greater family responsibilities leads to higher absenteeism. Both Leigh (1983) and
Vistnes (1997) found a positive relationship between women's absence and number of children.
However, Scott and McClellan (1990), VanderHeuval and Wooden (1995), and Erickson, Nichols and
Ritter (2000) did not find any interaction between number of children and gender related to absence.
Furthermore, Bratberg et al. (2002), in a direct test of the double burden hypothesis, found that women
with more children actually had less absence, a result that was also found by Zaccaro, Craig, and Quinn
(1991). While such contradictions may be surprising, Mastekaasa (2000) offers some interesting insight
through his findings. Although predicting parenthood should lead to greater absenteeism for women due
to role conflict, he found that, overall, married men and women were very similar and that children only
affected married women when the kids were very young. In fact, he found that married women with older
children were less absent than married women with no children. The only group that had increased
absenteeism in relation to number of children was single mothers who had never been married; divorced
parents were similar to married parents. The author concluded that multiple roles have positive effects and
pointed to research indicating that married people and parents are less suicidal, have more social support,
and are healthier. The benefits of women's multiple roles in terms of health are also highlighted by Rodin
and Ickovics (1990). Again, the evidence that childcare and family stress is linked primarily to absence by
women is not clearly demonstrated from an empirical standpoint.

A third argument presented to explain women's greater absenteeism concerns the nature of their
jobs. According to this argument, women are more absent than men due to their less interesting jobs and
poorer working conditions. This is another argument that has intuitive appeal but little empirical support.
Specifically, studies by Mastekaasa and Olsen (1998) and Erickson et al. (2000) found that women's
absence was still higher than men's in identical jobs.

Overall, we are confronted with studies that suggest that women are more absent than men but
provide little explanation as to why. There is, however, a fourth reason that needs to be considered:
Perhaps women are more absent because it is more socially acceptable for them to be (Johns &
Nicholson, 1982). As such, the explanation for women’s lower attendance may concern the second
element of Martocchio and Harrison’s (1993) definition of absenteeism, i.e. there may exist different
social expectations to attend for women.

Studies focusing on social expectations surrounding absenteeism often employ Nicholson and
Johns's (1985) concept of absence cultures. An absence culture represents "a set of shared understandings
about absence legitimacy and the established customs and practice of employee absence and control”
(Johns & Nicholson, 1982 p. 136). As underlined by Harrison, Johns, and Martocchio (2000), absence
cultures involve awareness of the attendance behaviour of others, self-awareness of one's own
absenteeism, and some level of agreement about a proper level of absence. Through absence cultures,
absenteeism ceases to be an exclusively individual behaviour, but must also be viewed as a collective
behaviour (Nicholson, 1993; Nicholson & Martocchio, 1995). Empirically, this collective

15
conceptualization of absence behaviour has received support, primarily relating to workgroup norms and
supervisor attitudes on absenteeism (Gellatly & Luchak, 1998; Markham & McKee, 1995; Xie & Johns,
2000). However, absence culture theory as developed by Nicholson and Johns (1985) does not limit the
normative reach to proximal entities such as workgroups and immediate supervisors, but can be extended
to social groups. Absence cultures can exist for collectives such as members of an occupation, citizens of
a country, or, consistent with the focus of this paper, along gender lines.

Gender as a Social Category Subject to Stereotypes: Implications for Absence Cultures

Absence cultures are strongly linked to social expectations concerning attendance behaviour.
Social psychologists have regularly noted that expectations surrounding behaviours can be influenced by
social categorization and stereotypes about groups (Anderson, 1991; Anderson, Krull, & Weiner, 1996;
Deaux, 1984; Kelley & Michela, 1980). Anderson, Krull, and Weiner (1996) underline that individuals
will use knowledge structures, including stereotypes, to provide quick explanations for behaviour, and
that stereotypes can influence our expectations about appropriate behaviour in given situations and for
specific actors. Gender researchers are increasingly focusing on the group and categorization aspect of
gender to explain differences in how men and women are perceived and treated. While many studies in
organizational behaviour treat gender as an individual characteristic rather than a group identity, gender
researchers such as Deaux (1984) and Nelson and Burke (2002) assert that gender should be considered
primarily as a social category. Deaux (1984) places a strong emphasis on categorization and notes that
gender is subject to normative beliefs and leads to expectations beyond any individual-level trait of the
person. To the extent that gender is treated as a group category as opposed to an individual trait,
expectations of behaviour, including absenteeism, are likely to be influenced by widely held gender
stereotypes.

Apart from the obvious physical differences, the concept of gender can be considered a socially
constructed state (Nelson & Burke, 2002). Several authors (Burke, 2002; Durkin, 1987; Nelson & Burke,
2002) underline how gender roles are created through socialization from a very early age and lead to
stereotypical views of men and women. As underlined by Deaux and Lafrance (1998), gender stereotypes
cover a wide range of beliefs concerning physical characteristics, personality traits, role-related
behaviours, occupational preferences, specific competencies, and emotional disposition. The general
stereotypes are that men are technically competent, competitive, aggressive, rational, and more committed
to their careers, while women are emotional, nurturing, passive, relationship oriented, less committed to
work, and less motivated for success (Durkin, 1987; Nelson & Burke, 2002). Stereotypes are very
resistant to change and often operate on an almost unconscious level (Banaji & Hardin, 1996; Deaux &
Major, 1987). Halpern (2005) recently commented that, although women have become much more
present in the workplace over the last twenty-five years, most of the same stereotypes surrounding women
in terms of childcare duties, emotions, and so on have not evolved. Jost and Banaji (1994) suggest that
such a phenomenon can be conceived as system-justification, which refers to “the psychological process
whereby an individual perceives, understands, and explains an existing situation or arrangement with the
result that the situation or arrangement is maintained” (p.10). Since women have traditional roles and are
subject to long standing stereotypes, expectations and explanations will tend to reinforce these views.
The power of stereotypes and system-justification is underscored by the growing body of research
demonstrating that the differences between genders are actually quite small. Durkin (1987), Hyde (2005),
and Maccoby and Jacklin (1974) highlight that most psychological and behavioural differences are very
small, and that the widely held beliefs that men and women are different may be constructed more
through early childhood teaching and reinforced by the media than because of any real differences. The
important point is that, regardless of whether the differences are real or not, the stereotypes and
expectations are real and this is what will ultimately shape social reality and attributions surrounding
gender issues (Deaux, 1984; Deaux & Lafrance, 1998).

16
 What are the implications of commonly held gender stereotypes on expectations surrounding
absenteeism by women? The general stereotypes and beliefs about women such as their lower
commitment to work, their double-burden of work/childcare, their higher level of stress, and their more
fragile health can result in an absence culture whereby absence is more legitimate and acceptable for
women as compared to men. This is consistent with Jost and Banaji’s (1994) notion of system
justification and Deaux’s (1984) suggestion that gender stereotypes drive expectations of behaviour.
Based on the current “system”, women will be held to different rules in terms of illness, childcare duties,
and stress. First, due perceived fragile health, women will be judged less responsible for illness absence
than men. Illness will be viewed as a legitimate mitigating circumstance for women to be absent, but the
effect should be weaker for men. Second, due to the perceived emotional nature of women (Durkin,
1987; Nelson & Burke, 2002), women should be judged less responsible for stress absence than men.
Third, women's widely perceived role as primary caregiver to children will legitimize absence for family
reasons more than men. In fact, Harrison, Johns, and Martocchio (2000) have suggested that absence
culture norms surrounding child responsibilities may apply only for women and not for men. Nelson and
Burke (2002) stress that socially constructed gender roles can restrict men in their ability to fully enjoy
and participate in family life. As such, while men are being encouraged to be more involved in childcare,
it may not be as legitimate a reason to be absent as compared to women.

Overall, the different norms and expectations surrounding women’s absenteeism provide an
interesting explanation for the general finding that women are more absent than men. The stereotypes in
operation may be driving the differences. Expectations can strongly influence behaviour and gender
stereotypes can become self-fulfilling prophecies (Geis, 1992). The existence of an absence culture for
women can reconcile the fact that individual studies focusing on gender and absence have not found
consistent support for predictions concerning women and sickness absence, or women’s absence and
childcare duties, and so on. Actual health problems or family situations in relation to actual absence by
women may not be the proper level of analysis to study this issue. Beliefs and stereotypes concerning
women as a social category may be more important. As such, regardless of individual situation,
absenteeism by women may be higher than men in the aggregate due to differing social expectations
concerning attendance.

Consequences of a Separate Absence Culture for Women

A common conceptualization of absenteeism by managers is that it is a form of deviant

behaviour subject to disciplinary action (Edwards & Whitson, 1993; Johns, 1997; Johns & Nicholson,
1982). If the proposed absence culture for women exists, the deviant connotation of absence by women
should be much less as compared to men. As per the previous section, women should be viewed to have
many good, acceptable, or at least expected reasons to be absent from work. One would suspect that
family and childcare duties would be particularly expected for women. Nicholson and Goodge (1976)
suggested this and warned that, for women, domestic responsibilities, much like illness absence, may
represent a useful attribution for deflecting sanctions following an absence episode. Overall, generally
held stereotypes about women may create an absence culture in which absenteeism by women is more
legitimate and expected, and lead to actual higher absenteeism.

While the preceding suggests that women may benefit from absence norms that are somewhat
more lax and a perception that their absenteeism is less deviant, the absence culture outlined above
possibly leads to a second consequence that is extremely harmful to women: A perpetuation of biases
against women in the workplace. While stereotypes based on gender can lead to expectations, stereotypes
are also closely linked to discrimination against groups (Dipboye & Colella, 2005). The categorization
aspect of gender that Deaux (1984) and Durkin (1987) advocate when studying gender can be a strong
enabler of unfair biases against women. Dipboye and Colella (2005) suggest that discrimination is a
consequence of the short cuts taken in processing information on women or other historically

17
disadvantaged groups. In the workplace, it is widely acknowledged that women suffer from pay inequity
and still encounter glass-ceiling effects (Meyerson & Fletcher, 2000; Wilde, 2000). Research highlighted
by Cleveland, Vescio, and Barnes-Farrell (2005) indicate that women hold less prestigious and influential
jobs, advance more slowly in organizations, receive lower starting pay, and are less likely to benefit from
transfers and job changes than men. Deaux and Major (1987) underline that stereotypes about women can
create and perpetuate sex discrimination and segregation in employment. Women’s separate absence
culture driven by stereotypes may be a contributing factor to these barriers.

Even if absence by women is not viewed as deviant, absenteeism itself is still viewed very
negatively by managers. Absenteeism is associated with costs, lower performance, lower efficiency, and
disruptions to co-workers (Harrison & Martocchio, 1998; Nicholson & Martocchio, 1995). To the extent
that women are strongly associated with absenteeism, this can have negative effects for them in the
workplace. An absence culture that features a general expectation that women are more absent than men
may lead to women being discriminated against in hiring decisions, promotion decisions, and in terms of
pay equity. The same implicit stereotypes that justify a system (Jost & Banaji, 1994) in which women are
more absent can also lead system-justification for discriminating against women in the workplace. If, as
demonstrated by Durkin (1987), Maccoby and Jacklin (1974), and Hyde (2005), gender differences are in
reality very low, discrimination against women based on absenteeism is particularly unfair. Halpern
(2005) suggests that women are not living the typical stereotypes of spending more time on family duties,
but that implicit stereotypes along these lines hurt women in the workplace. It is interesting to note that,
in their review, Cleveland, Vescio, and Barnes-Farrell (2005) found that women face discrimination in
terms of hiring decisions, advancement, and pay, but less in performance appraisal. Perhaps the
explanation for this would be that hiring and promotion decisions rely on limited information, while
performance appraisal involves more individual and interpersonal consideration. In other words, group
categorization is more salient in hiring and promotion decisions than in performance appraisal where the
focus is more on the individual. As such, stereotypes surrounding women and absenteeism may prove to
be barriers for women in terms of access and advancement in organizations.

To recap, despite consistent findings that women are more absent than men, research focusing
specifically on health and family issues have not been able to explain women’s supposed higher
absenteeism. This section suggests that social expectations, in the form of a separate absence culture for
women based on widely held gender stereotypes, may be a factor in explaining higher absence by women.
This absence culture for women may legitimize absenteeism for this group and attenuate perceptions of
deviance surrounding women’s absence. At the same time, such an absence culture, regardless of
whether it leads to actual higher absenteeism for individual women, will also be a factor in perpetuating
biases and discrimination against women in the workplace. An important question that emerges is: How
can we study this or find evidence of such an absence culture? The traditional correlational designs
employed in absence research involving absence records (Johns, 2003) seem suboptimal for studying such
questions. As highlighted in the introduction, however, there exists a data source that would allow
researchers to link gender issues and absenteeism in terms of stereotypes, social representation, and
outcomes: the popular press.

Gender, Absenteeism, and the Popular Press: The Content Analysis

As emphasized in the introduction, absence from work has been featured prominently in the press
over the years. Based on the nature of the press, however, stories featuring absenteeism not only deal
with purported facts about absence, but should also provide insight into social norms and beliefs
surrounding the behaviour. Although the press is often presented as objective and unbiased, no text is ever
truly objective but always contains ideological underpinnings and reproduces social relations that exist
within society (Fairclough, 1995). Mass media have the power to influence knowledge, beliefs, values,

18
social relations, and social identities (Fairclough, 1995; Fowler, 1991) on issues ranging from gender,
patriotism, social class, money, leisure and family. The press is also a mirror for social norms and values
that already exist. The content of newspapers is not facts about the world but ideas articulated from a
particular ideological position (Fowler, 1991). A more accurate description is that news is the “end
product of a complex process which begins with a systematic sorting and selection of events and topics
according to a socially constructed set of categories” (Hall, Crichter, Jefferson, Clarke, & Roberts, 1978,
p.53). As previously mentioned, gender represents an important social category that is almost always
salient to some extent. Cleveland, Vescio, and Barnes-Farrell (2005) note that “understanding variations
in the experiences and outcomes of men and women at work requires that we look beyond the immediate
work environment to the societal and cultural context in which work takes place and to the developmental
history that people bring with them to work” (p. 149). The popular press represents an important societal
and cultural beacon that can provide clues as to the extent and consequences of gender categorization and
stereotypes.

There are reasons that absenteeism and gender will likely be treated in a way that will promote
stereotypical portrayals of the topics. For absenteeism, one would suspect that the portrayal would be
negative and simplified. Given that newspapers have a preference for negative stories (Fowler, 1991; Hall
et al., 1978), newspapers are much more likely to focus on the negative impact of absenteeism than any
positive features. Furthermore, as noted by Fowler (1991), newspapers are rarely interested in reporting
on complex processes, and prefer simplified versions of events highlighting outcomes. Consequently,
newspapers are less likely to report on the complex processes that cause absenteeism, and will focus
primarily on its negative impact. In terms of simplified explanations of absenteeism, links to gender
should prove a useful pathway with a strong emphasis on stereotypes. According to Fowler (1991) and
Hall et al. (1978), the social categories that newspapers use are very broad. The result is that news stories
often gloss over complexity in the pursuit of familiarity and often end up perpetuating stereotypes
(Fowler, 1991). Gender researchers such as Durkin (1987) have identified the popular media as an
important source and enabler of gender stereotypes. Media researchers such as Fowler (1991) note that
common stereotypes relating to women are often promulgated in the popular press with women often
defined almost exclusively through family roles, and as more emotional than men.

Overall, newspaper articles that combine absenteeism and gender may provide important insight
into how these topics are represented in the public domain and clarify the existence of a separate absence
culture for women. As underlined by Abrahamson (1996), far from being benign, newspaper stories on
workplace issues can be potentially influential on organizational policy and decision-making. As such an
important focus of study is whether or not the portrayal of absence by women may inadvertently send a
harmful message to managers about working women. To answer some of these questions, a content
analysis of articles featuring absenteeism and women appearing in the New York Times over the last 100
years was undertaken.

Methodology

Content analysis can be defined as any technique for making inferences by systematically and
objectively identifying specific characteristics within texts (Kabanoff, 1997; Stone, Dunphy, Smith &
Ogilvie, 1966). The emphasis on any technique indicates that content analysis should be considered a
series of techniques as opposed to a clear cut, rule-bound methodology. As noted by Weber (1990), there
is no single way of undertaking a content analysis and a great deal depends on the nature of the
substantive question. The common element in all content analyses is the creation of a coding scheme that
reduces a text into categories and extracts manageable data for analysis.

19
 For the present study, the analysis will be primarily qualitative in nature and, as recommended by
Miles and Huberman (1994), will focus on absenteeism as a naturally occurring, ordinary event in natural
settings, and will attempt to explicate how people come to understand, account for, take action, and
otherwise manage absence from work. This content analysis approach relies heavily on the work of
Krippendorff (2004) who advocates the use of both qualitative and quantitative analysis, and emphasizes
that content analysis is never just about words but is also concerned about where, when, and by whom a
text was written and read. From this perspective, it is important to justify the choice of texts.

The New York Times was chosen for several reasons. First, the availability of the database allows
for the extreme longitudinality that represents an important strength of the study. The availability of the
New York Times Historical Database spanning over 100 years will capture stories about women’s
absenteeism in different contexts and during periods when women were more or less present in the
workforce. Second, the New York Times covers a broad range of topics such as local, national and
international news, business news, and lifestyle trends. This variety also allows women’s absenteeism to
be reported on in different contexts. Finally, focusing on a single major newspaper avoids the need to sort
through identical stories that are picked up and run in smaller outlets. Overall, the New York Times is a
primary and well-recognized news source with original reporting on a host of issues.

Overall, 167 articles have appeared in the New York Times since 1900 focusing specifically on
women’s absenteeism. Each article was coded in a text-rich matrix (Krippendorff, 2004; Miles and
Huberman, 1994) for date, title of the article, job of workers, general treatment of absenteeism
(positive/negative), stakeholders identified in the articles, reported consequences of absenteeism, reported
causes of absenteeism, association between absenteeism and other organizational behaviour phenomena,
and type of article (news, feature, or editorial). This content analysis is part of a larger study reviewing
the total sample of over 3,000 articles featuring absenteeism in the New York Times. While the larger
study is still ongoing, gender and absence is one of the primary foci; the articles reviewed to date beyond
the 167 women/absenteeism articles detailed in the current study also provide an opportunity to compare
the differences in reporting between women and men in some specific contexts. Overall, as detailed next,
the review provided interesting evidence concerning the existence of an absence culture for women and
the perpetuation of gender stereotypes.

The Review

Grouping the articles by decade (Figure 1), two elements stand out. First, absenteeism by women
was an important story during World War II; this point will be focused upon shortly. Second, after the
war, absenteeism by women basically fell off the map only to resurface in an important way during the
1980s and 1990s. Acknowledging that the WWII articles reflect the extraordinary circumstances of the
times, the articles from the 1950s to the 1990s are consistent with the growing prominence of women in
the workplace that took place in the 1980s and 1990s.

In terms of content, an overall assessment is that the articles suggest that women are generally
more absent than men, and reflect a focus on women as a separate demographic group with specific
issues. Johns (1997) identified demographic models as one of several treatments of absenteeism in
academic research. Johns (1997) also identified several other models including withdrawal models
(absenteeism due to poor work attitudes), deviance models (absence as an indication of a deviant
personality or a faulty value system), conflict models (absenteeism as a labour tactic or a source of
disagreement between workers and managers), and medical models (absenteeism due to illness, pain,
smoking, and so on). Although all of these models are present in the full sample of 3,000 articles

20
 Figure 1

Articles by period

Articles by Period

90
80
70
60
50
Articles
40
30
20
10
0
Pre- 1940s 1950s 1960s 1970s 1980s 1990s 2000s
WWII

featuring absenteeism in the New York Times, it appears that the articles featuring absence by women
concern women’s traditional roles and some specific women’s issues. Out of the 167 articles, 101 focus
on domestic and family responsibilities for women. Other articles reference specific women’s issues such
as spousal abuse and specialized health problems. Articles also indicate that absenteeism by women, even
related to childcare, can lead to problems for women workers. While the following sections will expand
on the themes present in the reporting and explore trends over time, the oldest article in the sample,
1911’s “Bar Out Teachers With Small Babies” deserves special mention. While only a single data point,
this article is interesting as it reflects both women’s traditional roles as mothers and the barriers this can
create at work. The article describes a resolution passed by the New York City Board of Education
barring married women with young children from teaching positions. Board members explained that the
young mothers must resign or be fired, and justified the decision by suggesting that it is impossible to
care for a child and work at the same time. While this is perhaps the most blatant case of discrimination
against mothers and appeared in an era without legislative protection, other more subtle examples of this
line of thought appear elsewhere in the sample, including quite recently.

Women on the map: Absenteeism by women in the Second World War. The large number of
articles featuring absenteeism by women in the 1940s is a testimony to the important impact that women
had on war production during World War II. Given the almost total lack of articles about women and
absence prior to 1941, these articles also represent the first example of women being linked to
absenteeism in a significant way.

The most striking feature of the articles concerning absenteeism and gender is the difference in
how absence by men and women is presented. As part of the larger content analysis, all 510 articles
featuring absenteeism (both genders) of war plant workers were reviewed. Although men’s absenteeism is
often portrayed as deviant, unpatriotic, or caused by labour conflict, women’s absenteeism is presented as
being caused primarily by domestic responsibilities. While many articles during the Second World War
concerning men’s absenteeism present it as potentially leading to legislation that would see shirkers sent
immediately into the armed forces (e.g. 1943’s “House Body Votes Absentee Curb; Would Draft Navy
Yard Shirkers”), legislation discussed to curb women’s absenteeism was more supportive in nature and

21
was aimed at helping rather than punishing (e.g. 1943’s “President Speeds Women's Facilities” and
1944’s “More Services for Women Workers Asked As Essential to Keep Them on War Tasks”). While
absenteeism by men in wartime was equated with shirking, hangovers, horse racing and other
questionable reasons, such associations were not made for absence by women.

At the same time, stories featuring absenteeism by women were not unanimously supportive in
nature. Particularly in 1942 and early 1943, articles often portrayed women as a group of workers who
were too fragile, not committed, and not particularly suited for war work. December 1942’s “Women
Cautioned About War Jobs” and “Women Urged To Stick” suggest that women were actually retarding
the war effort due to their absenteeism. In these articles, women are portrayed as accident-prone (due to
their high heels, according to the U.S. Secretary of Labor who was a woman) and not serious about work
(e.g. apt to skip work in order to attend parties and get their hair done). Other articles in early 1943
suggested that women lacked the necessary toughness for war plant work and that they would be absent
“whenever they get a little scratch” (“Safety Pleaders Reach Los Angeles For Drive Against Plant
Accidents”–Feb. 1943). As 1943 advanced, however, there was increasing acknowledgement that women
were vital to war production, that their output was impressive, and that their absenteeism was due
primarily to childcare duties, transportation problems, and shopping needs. Articles such as 1943’s “It's A
Woman's War, Too", and “Their First 'Sub'”, and 1944’s “Women Have Increasing Role In Campaigns
To Speed War Plant Output, Cut Delays” attest to the valuable role that women were playing in the war
effort. In addition, articles such as 1943’s “Store Sets Up Unit In Jersey War Plant”, “Child Care Group
Formed In Buffalo”, and “Women In War Jobs Seen In Need Of Aid” indicate that women needed
assistance and services to combat their absenteeism. Overall, of the 81 articles featuring women’s
absenteeism during the Second World War, 46 linked absenteeism to domestic duties, and most of these
called for additional services for women.

In terms of the rest of the 81 articles, 7 emphasized women’s physical fragility, 6 stressed a
somewhat complacent attitude on the part of women, and other articles present women’s absenteeism as
linked to women's issues. For example, several articles, such as 1943’s “Beauty Shops Seen Speeding
The War”, suggest that making beauty shops and cosmetics available to women war plant workers would
help curb women’s absenteeism. Also, articles appeared describing the use of counsellors and social
workers to help women in the workplace. Articles such as 1943’s “Counsellors For Shop Morale” and
“Humanizing The Job” describe the need for women to have support dealing with issues, including shop
floor romances, fashion in the plant, and recreation. Finally, in conjunction with the articles highlighting
the importance of women to the war effort, wartime articles on women’s absenteeism also included calls
for recognition of women as equal, or better, than men and the suggestion that women were being unfairly
criticized for absenteeism. Articles in this vein included 1943’s “Women Defended On Absenteeism”,
“Women Workers Keeping At Posts”, and “Equal Pay Laws For Women Sought”. Before any permanent
gains or recognition could be obtained, however, the war ended and, as evidenced by the paucity of
articles during the 1950s and 1960s, women’s work disappeared from the scene.

1950s, 1960s, and 1970s: It’s a man’s world. The 21 articles featuring women’s absenteeism
over the 30-year period between 1950 and 1979 are an indication of how women were not a strong part of
the employment landscape over that time. Apart from one article in 1951 that advocated childcare (an
article that was written from the perspective of the U.S. returning to a war footing), the articles in the
1950s and 1960s commonly represented women’s absenteeism as an indication that they were not good
workers, and that men should be favoured. Short articles such as 1957’s “Office Absentees Cited” and
1969’s “Women Job Absenteeism Slightly Above Male Rate” noted that absenteeism by women was
higher than men’s, while articles such as 1956’s “A Secretary Is Set to Wed Or Get Ahead”, 1961’s
“Working Women Pose Problems, Employers Find”, and 1969’s “Employment Agency Backs Men As
Office Typists and Secretaries” portrayed women as unreliable, neurotic and inefficient from an
employment perspective.

22
 However, an interesting new angle that emerged during this period was the notion that women
were being discriminated against in the workplace and that exaggerated accounts about their absenteeism
were being used to create barriers. This idea had appeared once or twice in the articles from WWII but
became a larger focus over these years. A 1959 article on women stock brokers, “Women A Trial To
Some Brokers But Strong Force In Stock Field”, indicated that women brokers could be very successful
dealing with women or open-minded men clients, but that they faced barriers in being hired due to fears
of high absenteeism. As civil rights and the women’s liberation movement gained prominence in the
1960s and 1970s, articles such as “Women's Pay Gap Is Still Widening, U.S. Official Says” (1964),
“Women Are Breaking The Blue-Collar Barrier” (1973), and “To Women, Insurance Companies Are At
Fault On Many Things” (1974) suggested that women’s higher absenteeism was actually a myth being
used to discriminate against women in different business domains. Furthermore, two articles in the 1970s
“Women March Down Fifth In Equality Drive” (1970) and “Most Of The Nation's 'Alices' Stay On Job,
Ignoring NOW's Call For One-Day Strike” (1975) presented absenteeism as a form of organized protest
by the women’s movement to advance women’s rights. It should be noted, however, that both protests
were portrayed as unsuccessful. Still, the suggestion that higher absenteeism by women due to family
roles and lack of commitment is a myth being used to discriminate is interesting. Given the very few
articles that appeared during this period, however, this perspective seems to reflect more of a cry in the
wilderness as opposed to a strong trend.

The 1970s also witnessed the emergence of a trend that would dominate the reporting on
women’s absenteeism in the 1980s and 1990s: Corporate childcare and alternative work arrangements.
Articles such as “While Mother Works, The Company Takes Care Of The Children” (1970), “Woman
Economist Backs Part - Time Work” (1972) and “Workers Find 'Flextime 'Makes For Flexible Living”
(1979) revived a discourse that had been almost totally absent since the articles during World War II: the
need to provide women with assistance in order to help balance work and family.

1980s, 1990s, and 2000s: A new focus on women’s issues. The 1980s saw women stake a much
greater place in employment. This important presence of women in the workplace is underscored by the
number of articles featuring women’s absenteeism in the 1980s and 1990s. In all, 56 articles touching on
absenteeism by women appeared from 1980 to 1999. From a gender role perspective, 46 of these articles
featured women’s absenteeism as a result of domestic duties, most commonly involving childcare. Article
after article discusses the need for corporate childcare, the need for companies to fund daycares in their
communities, or the need for governments to get involved in childcare. Unlike the articles during World
War II, the childcare issue is overwhelmingly portrayed as a problem that private enterprise should deal
with as opposed to the government. Many articles have a strong advocacy aspect but also portray
corporate childcare as a win-win situation for mothers and companies that leads to corporate success.
Articles such as “Child Care And Business, Side By Side” (1986), “Child Care Connected With The
Workplace Makes Good Business Sense” (1987), “Wooing Workers In The 90's: New Role For Family
Benefits” (1988), and “Corporate Day Care Pays For Itself” (1989) strongly suggest that childcare is an
important recruitment and retention tool that reduces costly absenteeism. In addition to childcare, articles
in the 1980s and 1990s also advocated alternative work arrangements for women such as part-time work,
job sharing, flextime, and telecommuting to help women balance work and family life, and reduce
absenteeism. Examples of such articles include “A Fresh Look At Job Sharing” (1983), “Flexibility At
Work Helps Parents At Home” (1989), and “Working At Home, For Better For Worse” (1993). In the late
1990s and early 2000s, articles also appeared linking absenteeism to eldercare. It is interesting to note that
many of the family policy/absenteeism articles in the 1980s, 1990s and 2000s included fathers and sons as
also benefiting from family-friendly work policies; examples using women, however, were still
predominant.

23
 While many articles called for more family-friendly policies, it is important to note that the
benefits of family policies for companies are often presented as unsupported by hard data, and some
articles demonstrate resistance from managers and negative outcomes for women relating to childcare and
family policies. For example, the aforementioned “Wooing Workers In The 90's” (1988) article features
the comment from an executive that he has no intention of getting involved with daycares and would just
not hire women likely to become pregnant. Other articles suggest that many managers were taking a very
cautious approach to childcare support and that they were skeptical as to its cost efficiency. Furthermore,
there were indications in the 1980s and 1990s that women could still be fired due to childcare difficulties.
A particularly telling example was a three-story arc in 1991, including “Child Care In Conflict With A
Job”, describing how a Minnesota woman had been fired for excessive absenteeism caused by caring for
her very sick child. The articles focus on a lawsuit brought forth by the woman (which she won) asking
for unemployment insurance benefits that had originally been denied to her since she was fired. She did
not contest the firing per se, and the articles indicated that in the mind of the company, the judges, and
even the employee herself, the firing was justified. It is quite striking to juxtapose these 1991 articles with
the 1911 article highlighted at the beginning of the review. While the 1911 article suggested that caring
for children is incompatible with working and described policy to fire young women with children on this
basis, the 1991 articles indicate that this way of thinking still existed and was being enacted some 80
years later. In terms of discrimination based on women’s absenteeism, articles such as “Single Mothers
Join Suit To Enlist In The Military” (1984) and “On A Cultural Battlefield: Women In The Military”
(1997) described how young mothers were facing barriers enlisting and staying in the military due to
unsubstantiated fears about absenteeism, while 1995’s “For Many In Work Force, 'Glass Ceiling' Still
Exists” described the findings of a federal commission looking into discrimination that reported that
myths surrounding absenteeism by women was not allowing women to move beyond middle
management. The report suggested that purported higher absenteeism by women was based on
stereotypes and that, when maternity days are removed, women actually have less absenteeism than men.

Finally, from 1988 to 2004, there were 6 articles linking women’s absenteeism to
spousal/relationship abuse and sexual harassment. In fact, the final article in the sample from September
2004 “Victim Of Domestic Abuse Gets City Job Back” describes the case of a female parole officer who
had been fired for excessive absenteeism but was reinstated by order of the courts once it was revealed
that she was suffering from spousal abuse and was in hiding. There were also a few articles linking
absenteeism to women’s poverty and absenteeism as a labour tactic in nurses’ strikes. Overall, the
articles in this time period covered stories featuring women in many different jobs.

Discussion

The articles featuring absenteeism by women that have appeared in the New York Times over the
last one hundred years lend credence to the notion that social expectations concerning absence from work
are different between men and women. Articles focusing on women’s absenteeism most commonly
portray the behaviour as a work/family balance issue relating to childcare and other domestic duties.
While articles featuring absenteeism by men or gender-neutral articles tend to span many absence models
described by Johns (1997) such as withdrawal, deviance, or conflict, the articles on women’s absenteeism
primarily focus on a unique set of issues dominated by family roles.

24
 In terms of the common assumptions in academic research that absence by women is due to
poorer health, childcare, or lower status jobs, childcare emerges in the newspaper articles as the central
reason for women's absence. While the articles during World War II highlighted lack of physical stamina
as a reason for women’s absenteeism, and certain articles in the 1950s and 1960s indicated that women
were not serious about work, the childcare and domestic responsibility reasons were the most common
and appeared in each of the periods reviewed. Overall, the message conveyed in these articles is that
women can be strong contributors and are very present in the current workforce, but that they have other
priorities or special issues that tend to lead to higher absenteeism.

The ideas advanced by Nicholson and Goodge (1976) that women might be able to use their
family-based reasons for absence to ward off negative outcomes or punishment does not appear fully
supported by the articles. In fact, although not as high in terms of numbers, articles suggesting that
women are discriminated against or punished for their family/childcare absenteeism appeared in every
section of the review. In addition to the articles describing women actually being fired for family-related
absenteeism, it is interesting to speculate how this overall perception may impact women in general.
Indeed, several articles suggested that exaggerated fears of absence were resulting in discrimination
against women, and that women’s high rate of absenteeism is a myth. Halpern (2005) notes that
stereotypical role-based images of women have persisted over the years, even if they have not reflected
reality for quite a long time. The fact that men are also assuming family duties is emphasized by Halpern
(2005) and is demonstrated by a limited number of articles in the sample. Still, the message that work by
women will be disrupted by family duties is very strong. Another interesting element that emerges
through the review is that women have been on the cutting edge of many forms of alternative work
arrangements such as part-time work, flextime, job sharing, and telecommuting with one of the aims
being a reduction of absenteeism. Although these are positive changes that enhance the possibilities of
parents participating fully in the workforce, Cleveland, Vescio and Barnes-Farrell (2005) note that
employees in these alternative work arrangements (usually women) face serious barriers in terms of
career outcomes involving promotion and fringe benefits. Overall, advocates for women’s rights in the
workplace appear to be facing a difficult balancing act. On the one hand, they are at the forefront of
calling for more family-friendly policies, but in doing so they may be perpetuating stereotypes that
ultimately hold women back. From management’s perspective, the articles suggest that cost/benefit
analyses are the primary concern, and that women are still vulnerable to punishment when absenteeism
becomes excessive.

From an absence research perspective, the articles confirm some common assumptions but also
raise some questions for future study. First, consistent with general assumptions surrounding absenteeism
(Johns & Nicholson, 1982), the portrayal of absenteeism in all the articles is negative. While this notion
has an almost taken-for-granted quality to it, it is one that absence researchers such as Goodman and
Atkin (1984) have taken issue with, and is currently being questioned through the emergence of new
concepts such as presenteeism (Hemp, 2004). Second, the articles appearing in the New York Times
remind us that absenteeism does not only involve work factors, but outside forces as well. Johns (1997)
notes that withdrawal models, which link absenteeism to workplace factors such as job satisfaction or
organizational commitment, have historically been the most popular from an absenteeism research
perspective. In reality however, these factors represent only a small aspect of absenteeism. Outside forces
such as family responsibilities and contextual factors must be considered when studying absence. To limit
research to work and organizational factors and attitudes is only scratching the surface of the
phenomenon. At the same time, a third element that emerges from the articles is that, even if absenteeism
is caused by outside forces, managers still believe that it can be combated through management action.
This is consistent with Johns and Nicholson’s (1982) assertion that regardless of its cause, managers will
almost always believe that they can reduce absenteeism. The assumption is based on the belief that since
absenteeism is viewed as an organizational problem, the solutions must be available within the
organization (Johns & Nicholson, 1982). In this vein, companies and managers are portrayed as very

25
active in the newspaper articles. The articles indicate that organizations occasionally fired women, offered
services such as shopping facilities in WWII or childcare in the 1980s and 1990s, and implemented
alternative work techniques in order to combat absenteeism by women. Several articles in the sample
indicated that companies were resisting government involvement on many of these issues and preferred
solutions emerging from the private sector. Given the traditional male dominance in business, however, it
is questionable whether or not business leaders are truly the ones who should be developing policies on
these issues.

Finally, perhaps the most intriguing suggestion, which appeared only occasionally but with some
regularity, was the notion that women’s higher absenteeism is actually a myth. Although numerous
studies and a meta-analysis have demonstrated this, recent articles are less consistent on this point. Many
of the studies that this conclusion has been built upon were conducted prior to the mid-1980s. Given the
advances by women in the workplace over the last 25 years (Halpern, 2005), new studies, and perhaps a
meta-analysis covering studies from the late-1980s until today, are required. The belief that women are
more absent than men may reflect the absence cultures that exist along gender lines, but it is important
that such beliefs and expectations be revisited as times change. As Halpern (2005) underlined, women’s
participation in the workforce has undergone fundamental changes over the last 25 years but many of the
stereotypes from decades gone by still remain. As researchers, we must resist falling into this trap and
ensure that our studies evolve as times, and context, change.

Limitations and Conclusions

Limitations of this study include the single source and the nationality of the employees. All of the
articles featured American women. As absence cultures may vary across national cultures, future research
is needed to verify if expectations are the same for different peoples. One of the causes of this limitation
is the exclusive use of the New York Times as a data source. Focusing on the press in other countries
would lead to cross-cultural analysis in future studies. Expanding the source of texts in other ways would
also lead to interesting future studies. For examples, using newspapers with different editorial policy or
more business-focused publications such as the Wall Street Journal would perhaps offer another
perspective. Trade and practitioner journals would also be another interesting source for analysis. As
previously stated however, the New York Times is a rich source of stories on many different topic areas
that allows for a more varied analysis.

In conclusion, the 167 articles on women’s absence from work in the New York Times suggest
that there exists a particular absence culture for women based primarily on family roles. While this
absence culture has led to calls for more family-friendly policies in organizations, others have suggested
that women’s higher absenteeism is a myth that is leading to discrimination in the workplace. Future
research is needed to directly assess the existence of such an absence culture for women, to understand
how gender stereotypes influence social expectations to attend work, and to explore how such absence
cultures and stereotypes may perpetuate biases and discrimination against women workers.

Overall, this study combines research on absenteeism and gender using a methodology and data
source not regularly employed in studies on absence. The findings suggest that gender stereotypes are
very present in relation to absenteeism and, given the growing research demonstrating that actual
differences between men and women are very small in many respects, highlight the need for more
detailed research on gender and absence to determine if the differences are real or perceived, and what is
driving this differences or perceptions.

26
 References

Abrahamson, E. "Managerial fads and fashions: the diffusion and rejection of innovations." Academy of
Management Review, 16, (1991), 586-612.
Abrahamson, E. "Management fashion." Academy of Management Review, 21, (1996), 254-285.
Anderson, C.A. "How people think about causes: Examination of the typical phenomenal organization of
attributions for success and failure." Social Cognition, 9, (1991), 295-329.
Anderson, C.A., Krull,D.S., & Weiner, B. "Explanation: Processes and consequences." In Higgins, E.T.
& Kruglanski, A.W. (eds). Social Psychology: Handbook of Basic Principles, New York: Guilford Press,
(1996), 271-296.
Banaji, M R. & Hardin, C.D. "Automatic stereotyping." Psychological Science, 7, (1996), 136-141.
Bratberg, E., Dahl, S-A, & Risa, A.E. "“The double burden”: Do combinations of career and family
obligations increase sickness absence among women?" European Sociological Review, 18, (2002) 233-
249
Cleveland, J.N, Vescio, T.K., & Barnes-Farrell. "Gender discrimination in organizations". In Dipboye,
R.L. & Colella, A (eds.) Discrimination at work, New Jersey: Lawrence-Erlbaum, (2005), 149-176.
Côté, D., & Haccoun, R.R. "L’absentéisme des femmes et des homes: une méta -analyse." Canadian
Journal of Administrative Sciences, 8, (1991), 130-139.
Deaux K. "From individual differences to social categories. Analysis of a decades’s research on gender."
American Psychologist, 2, (1984),105-116.
Deaux, K. & Lafrance, M. "Gender". In Gilbert, D. T., Fiske, S. T., & Lindzey, G. (eds.)The handbook of
social psychology, Vol. 1 (4th ed.). New York, NY, US: McGraw-Hill, (1998), 788-827.
Deaux, K. & Major, B. "Putting gender into context: An interactive model of gender-related behavior."
Psychological Review, 94, (1987), 369-389.
Dipboye, R.L. & Colella, A. "An Introduction". In Dipboye, R.L. & Colella, A (eds.) Discrimination at
work, New Jersey: Lawrence-Erlbaum, (2005), 1-8.
Durkin, K. "Social cognition and social context in the construction of sex differences." In M.A. Baker
(ed) Sex differences in human performance, Wiley and Sons, (1987), 141-170.
Edwards, P. & Whitston, C. Attending to work: the management of attendance and shopfloor order.
Blackwell. 1993.
Erickson, R.J., Nichols, L., & Ritter, C. "Family influences on absenteeism: testing an expanded process
model." Journal of Vocational Behavior, 57, (2000), 246-272.
Fairclough, N. Media Discourse. Edward Arnold: New York, 1995.
Fowler, R. Language in the News: Discourse and ideology in the press. London: Routledge, 1991.
Gellatly, I.R. , & Luchak, A.A. "Personal and organizational determinants of perceived absence norms."
Human Relations, 51, (1998), 1085-1102
Geis, F.L. "Self-fulfilling prophecies: A social psychological view of gender." In Beall, A.E., &
Sternberg, R.J.(eds.) The psychology of gender. New York, NY, US: Guilford Press, (1993), 9-54.
Goodman, P.S. & Atkins, R.S. "Effects of absenteeism on individuals and organizations." In P.S.
Goodman, R.S. Atkins, and Assocs (eds), Absenteeism: New Approaches to Understanding, Measuring
and Managing Employee Attendance. Washington: Jossey Bass, (1984), 276-321

27
Hall, S., Crichter, C., Jefferson, T., Clarke, J., & Roberts, B. "The social production of the news". In
Policing the crisis: Mugging, the state, and law and order. London: Macmillan, (1978), 53-77
Halpern, D.F. "Psychology at the Intersection of Work and Family: Recommendations for Employers,
Working Families, and Policymakers." American Psychologist, 60, (2005), 397-409.
Harrison, D.A, Johns, G., & Martocchio, J.J "Changes in technology, teamwork, and diversity: New
directions for a new century of absenteeism research." Research in Personnel and Human Resources
Management, 18, (2000), 43-91.
Harrison, D.A , & Martocchio, J.J . "Time for absenteeism: A 20-year review of origins, offshoots, and
outcomes." Journal of Management, 24, (1998), 305-350.
Hemp, P. "Presenteeism: At Work - But Out of It." Harvard Business Review, 82(10), (2004, October),
49-58.
Hyde, J.S. "The Gender Similarities Hypothesis." American Psychologist, 60, (2005), 581-592.
Johns, G. "How methodological diversity has improved our understanding of absenteeism form work."
Human Resource Management Review, 13, (2003), 157-184.
Johns, G. "Contemporary research on absence from work: correlates, causes and consequences."
International Review of Industrial and Organizational Psychology, 12, (1997), 115-173.
Johns, G. & Nicholson, N. "The meanings of absence: New strategies for theory and research." In B. M.
Staw & L. L. Cummings (Eds.), Research in Organizational Behavior, CT: JAI Press., Vol. 4,
(1982),127-172,
Johns, G., & Xie, J.L. "Perceptions of absence from work: People’s Republic of China versus Canada."
Journal of Applied Psychology, 83, (1998) 515-530.
Jost, J.T., & Banaji, M.R. "The role of stereotyping in system-justification and the production of false
consciousness." British Journal of Social Psychology, 33, (1994), 1-27.
Kabanoff , B. "Computers can read as well as count: Computer-aided text analysis in organizational
research." Journal of Organizational Behavior, 18, (1997), 507-511
Kelley, H.H., & Michela, J.L. "Attribution theory and research." Annual Review of Psychology, 31,
(1980), 457-501
Krippendorff, K. Content Analysis. Sage. Thousand Oaks. 2004.
Leigh, J.P. "Sex differences in absenteeism." Industrial Relations, 22, (1983), 349-361.
Maccoby, E.E. & Jacklin, C.N. The psychology of sex differences. CA:Stanford University Press. 1974.
Markham, S.E. & McKee, G.H. "Group absence behavior and standards: A multilevel analysis." Academy
of Management Journal, 38, (1995), 1174-1190
Martocchio, J.J. & Harrsion, D.A. "To be there or not to be there?: Questions, theories and methods in
absenteeism research." Research in Personnel and Human Resources Management, 11, (1993), 259-328.
Mastekaasa, A. "Parenthood, gender and sickness absence." Social Science and Medicine, 50, (2000),
1827-1842.
Mastekaasa, A. & Olsen, K.M. "Gender, absenteeism, and job characteristics." Work and Occupations,
25, (1998), 195-228.
Meyerson, D.E., & Fletcher, J.K. "A modest manifesto for shattering the glass ceiling." Harvard Business
Review, 78 (1), (2000, January-February), 129-136.

28
Miles, M.B., & Huberman, A.M. Qualitative data analysis: A sourcebook of new methods. Beverly Hills,
CA: Sage. 1994.
Nelson, D.L, & Burke, R.J "A Framework for examining gender, work stress, and health." In Nelson,
D.L. & Burke, R.J. Gender, work stress, and health. Washington, DC: APA, (2002), 1-14
Nicholson, N. "Absence-there and back again." Journal of Organizational Behavior, 14, (1993), 288-290.
Nicholson, N., & Goodge, P.M. "The influence of social, organizational and biographical factors on
female absence." Journal of Management Studies, 13, (1976), 234-254.
Nicholson, N., & Johns, G. "The absence culture and the psychological contract: Who’s in control of
absence?" Academy of Management Review, 10, (1985), 397-407.
Nicholson, N., & Martocchio, J.J. "The management of absence: What do we know? What can we do?"
In G.R. Ferris, S.D. Rosen, & D.T. Barnum (Eds.) Handbook of Human Resources Management. Oxford:
Blackwell. (1995),567-614.
Rodin, J., & Ickovics, J.R. "Women’s Health: Review and research agenda as we approach the 21st
century." American Psychologist, 45, (1990), 1018-1034.
Scott, K.D. & McClelland, E.L. "Gender differences in absenteeism." Public Personnel Management, 19,
(1990), 229-253.
Stone, P.J., Dunphy, D.C., Smith, M.S., & Ogilvie, D.M. The General Inquirer: A Computer Approach to
Content Analysis. Cambridge: MIT Press. 1966.
VanderHeuval, A. & Wooded, M. "Do explanations of absenteeism differ for men and women?" Human
Relations, 48, (1995),1309-1330.
Vistnes, J.P. "Gender differences in days lost from work due to illness." Industrial and Labor Relations
Review, 50, (1997), 304-323.
Weber, R. Basic Content Analysis. 2nd ed. Beverly Hills, CA: Sage. 1990.
Wilde, C. "Women in IT strive for equal job compensation." Information Week, (2000, Jun. 12), 226-228.
Xie, J. L., & Johns, G. "Interactive effects of absence culture salience and group cohesiveness: A multi-
level and cross-level analysis of work absenteeism in Chinese context." Journal of Occupational and
Organizational Psychology, 73, (2000), 31-52.
Zaccaro, S.J., Craig, B., & Quinn, J. "Prior absenteeism, supervisory style, job satisfaction, and personal
characteristics : An investigation of some mediated and moderated linkages to work absenteeism."
Organizational Behavior and Human Decision Processes, 50, (1991), 24-44.

29
ASAC 2006 Mohammed Al-Waqfi
Banff, Alberta United Arab Emirates University

Harish C. Jain
McMaster University

RACIAL INEQUALITY IN EMPLOYMENT IN CANADA: EMPIRICAL ANALYSIS AND

EMERGING TRENDS

This paper examines the incidence of racial discrimination in employment in

Canada over the last two decades (1980-1999). Using data obtained from a
sample of 119 legal cases, the paper provides a quantitative and qualitative
analysis of the nature of racial discrimination legal decisions as well trends in
this area in Canada. Some policy recommendations to combat racial
discrimination in the workplace are suggested.

Canada’s population and workforce are becoming increasingly pluralistic. Forty-two percent of
Canadians report origins other than French or British, while sixteen percent of Canadians are foreign born
(Heritage Canada 2001). Successive census data show that Visible Minority (VM) population1 has almost
trebled over the last two decades from 4.7% in 1981 to 13.4% in 2001. Similarly the proportion of VMs
in the total labor force in Canada rose more than two-and-one-half times from 4.9 in 1981 to 12.6% in
2001 (Census Canada). With a highly diversified workforce, it is essential that equal opportunities be
available for all Canadians, including racial minorities in a non-discriminatory work environment.
Canada’s economic growth and prosperity in a highly competitive and global marketplace will depend on
full utilization of the talents, skills, knowledge, and energy of all Canadians.

Human rights laws in Canada at the federal and provincial levels prohibit discrimination on the
basis of race, colour, and ethnic origin and numerous other grounds. In addition, most human rights
statutes have provisions for voluntary special programs (employment equity programs) for disadvantaged
groups. Nonetheless, racial discrimination continues to exist in Canadian workplaces as this paper will
indicate. This phenomenon continues to challenge both researchers and policy makers who attempt to
understand its causes and investigate possible ways to eliminate it.

This article attempts to provide an assessment of the nature and recent trends in the incidence of
racial discrimination in employment in Canada. We used a sample of 119 published legal cases to provide
a quantitative analysis of racial discrimination in employment complaints in Canada over the last two
decades. We also examine the characteristics of racial discrimination incidents in some selected
workplaces in Canada.

The paper is divided into several sections. Section one discusses theories advanced for racial and
other types of employment discrimination. Section two briefly reviews previous studies that empirically
assessed various aspects of racial discrimination in employment contexts in Canada. Section three
provides an empirical assessment of the nature and trends of racial discrimination in employment in
Canada over the last two decades. Section four discusses the patterns of behavior and working conditions
leading to racial discrimination complaints. The last section summarizes the main conclusions of the study
and suggests some recommendations to combat racial discrimination in employment.

30
Theoretical Perspectives on Racial Discrimination

Racial discrimination is a multidimensional and complex phenomenon that is manifested in

various social and economic interactions within the society in which it is found. There are many theories
that try to explain the causes and effects of racial discrimination in employment by scholars from different
disciplines including economics, sociology, social psychology, and organizational behavior. This paper is
not intended to test the theories but to provide a brief background and some perspectives on the
theoretical reasons advanced for racial and other types of discrimination in the workplace.

One of the early economists who addressed racial discrimination in employment was Gary Becker
(1957). He argued that discrimination occurs as a result of prejudice or “taste for discrimination“ on the
part of the employer or his agents. Other economists including Arrow (1972 and 1998), Phelps (1972),
and Aigner and Cain (1977) have developed a model of discrimination that does not assume the employer
is prejudiced or forgoes profits to indulge in his taste to discriminate. The theory of statistical
discrimination is an information-based theory. Employers according to this theory rely on group
membership (such as race or gender) as a signal that allows them to improve their predictions of a
candidate’s potential to perform. For example, if blacks and whites happen to differ on average
productivity, then the experience of employers over time will cause them to use the observable
characteristic, such as race, as a surrogate for the unobservable characteristics which in fact cause the
productivity differences (Arrow 1998). Therefore, the employer will consistently hire and promote
whites over equally qualified members of the other groups.

Other theories such as the human capital theory as well as the social capital theory explain the
persistence of racial discrimination in terms of observed differences in the qualitative characteristics of
workers from different racial groups. According to the human capital approach (Becker, 1957), some
members of minority groups lack the necessary “human capital“ such as education, training and
experience. Thus members of minority groups may continue to be locked up in low-income and low–level
jobs, not only because of the discriminatory treatment against them, but also because of the lower quality
of their human capital (Jain, Sloane & Hurwitz, 2003; Jain 1982).

Many researchers have recently come up with alternative explanations for the racial
discrimination phenomenon that are not market based. Their explanations rely on social and
psychological interactions and their effects in shaping the work and life opportunities available for
members of racial minorities (see Arrow 1998; Loury 1998; Elsmile and sedo 1996). According to the
social capital theory (Loury 1977, 1987, and 1998), one’s investment in productive skills depends on
one’s position in the social structure. Research has shown that a very large fraction of jobs are filled
through referrals by current employees (Arrow 1998). Social segregation can lead to segregation in
employment opportunities through network referrals. Thus, preferences and investments in education and
training by individuals are influenced by social and psychological externalities.

Each of the theories discussed above took a different approach to explain the causes of racial
discrimination in employment. However, racial discrimination in employment contexts is a complex
phenomenon that is more likely to be determined by a number of factors including systemic2
discrimination.

Previous Evidence on Racial Discrimination in Employment in Canada

A common problem facing researchers attempting to measure racial discrimination is the fact that
in most cases discrimination is inferred from residuals not accounted for by other variables used to
explain disparities and lower achievements that may be experienced by various racial minority groups.

31
 This problem is difficult to overcome given the systemic nature of discrimination. This may
explain why researcher use a number of research methods in their attempt to empirically assess racial
discrimination. The different methods used complement rather than substitute one another in measuring
the levels and nature of racial discrimination in employment. For example, regression analyses are widely
used in assessing pay discrimination. Although these studies vary in their design and level of complexity,
the basic form involves regressing pay levels on a number of human capital variables (including
education and experience) and an additional variable representing race or ethnic group membership.
Audit studies (using actors or testers as job applicants) are often used to examine discrimination against
minority groups in access to various jobs in the labour market. Surveys, on the other hand, are used to
measure specific perceptions and attitudes towards racial minorities in the society (which are generally
perceived to result in discriminatory treatment of disadvantaged minorities). Finally, legal cases provide
a rich source of information about the experience of victims of racial discrimination in employment and
factors leading to this behaviour.

Studies using these methods have been conducted in Canada over the last few decades. A study
by Frances Henry (1978) surveyed a random sample of 617 white individuals and found that 16% of
those individuals could be considered extremely racist and an additional 35% inclined towards some
degree of racism. The report of the Royal Commission of Equality in Employment (Judge Abella 1984)
indicated that in meetings with non-whites all across Canada, they complained of facing discrimination
both overt and indirect. More recent commissioned studies on the status of visible minorities (VMs) in the
federal public service showed similar results (See Samuel 1997; Perinbam 2000). In one of the major
audit studies, over 400 jobs identified from the classified ads in the major newspapers in Toronto were
tested using pairs of white and visible minority job applicants with equal levels of qualification for the
jobs (Henry and Ginzberg 1985). When using direct in-person applications, the study found that offers to
whites outweighed offers to blacks by a ratio of three to one. Similarly, in phone inquiries, the percentage
of times that callers were told the job is open for them were significantly higher for white Canadian
callers compared to white immigrants, west Indian Blacks, and Indo-Pakistani callers due to latter’s
accent.

The lower rates of employment and pay for racial minorities are evidence of disparate conditions
between VMs and whites in the labour market. While part of this disparity can be attributed to differences
in human capital and productivity-related factors, there is a general belief that racial discrimination could
be one of the factors leading to the lower earnings and employment conditions experienced by VMs. For
example, a recent study by Pendakur and Pendakur (2000) showed that while there was an improvement
in relative earnings of both aboriginals and visible minority workers between 1971 and 1981, the earnings
gap was stable through 1991, and then there was a decline in relative earnings between 1991 and 1996.
Other studies of the earnings gap between whites and racial minorities provide, in general, an indirect
evidence of possible discrimination against racial minorities (see Howland and Sakellariou 1993; Baker
and Benjamin 1997; Boyd 1992; Bloom et. al. 1995; Meng, 1987).

There are a number of studies in the Canadian context that addressed the gap in representation
levels between whites and racial minorities. For example, Jain et al. (2000) found a significant under-
representation of racial minorities in selected police services across Canada and indicated that selection
and promotion policies that disadvantage minorities may be responsible for this under-representation.
Ornstein (2000), using the 1996 Canadian Census data, found that there is in general a pervasive disparity
between members of racial minorities and whites in the City of Toronto in pay, employment rates, and
other socio-economic indicators. At least two previous studies analyzed legal cases to examine racial
discrimination in employment in Canada ( Jain 1982 ; Andiapan et. al. 1989).

32
 The findings from previous studies indicate that racial discrimination is probably responsible for
part of the disparity in achievements between various VMs and whites in the Canadian labour market.
While it is important to conduct research examining whether racial discrimination exists, a more
important and compelling issue for research now is to understand the nature of systemic discrimination
against VMs and the causes and possible solutions for this problem. We need to understand the
characteristics of discriminatory behaviours and any specific patterns or trends in the development of this
phenomenon overtime. We also need to know more about the types of situations, behaviours, and
interactions that affect the employment relationship in a way that leads to discriminatory treatment. The
current paper attempts to contribute to achieving this objective by providing an analysis of published
legal cases of racial discrimination in employment in Canada over the last two decades. In the following
section we will examine the nature and trends of racial discrimination as represented in published legal
cases in employment over the past two decades.

Empirical Analysis

One of the most direct and objective ways to obtain data on racial discrimination, which we use in
this study, is the published reports of legal cases regarding complaints of racially discriminating
treatment. Our empirical analysis used 119 legal cases published in the Canadian Human Rights Reporter
(CHRR) between 1980-1999. Using data from these legal cases, we provide below a quantitative analysis
of the characteristics of racial discrimination in employment complaints in Canada over the last two
decades. We supplement these empirical findings with a qualitative analysis of selected legal cases in
various work contexts that illustrate the patterns of systemic discrimination against VMs in the workplace
in the next section.

Methodology

We identified a set of variables in our attempt to understand the nature of racial discrimination in
employment in Canada and the characteristics of individuals and organisations involved. Data on these
variables was obtained from a sample of 119 published legal cases involving racial discrimination
complaints. These reports provide a rich and detailed data that is difficult to obtain from other sources.
However, it must be noted here that the complaint driven legal cases convenience sampling approach used
in this study may introduce self-selection bias since the sample includes only those complaints that the
pertinent Human Rights Commission chose to take to tribunal/board of inquiry to seek redress for the
complainant/s; only a fraction of complaints filed before the Human rights Commissions across Canada
go to a tribunal3.

Both multinomial logistic regression and bivariate chi-square tests were used in the data analysis.
Multinomial logistic regression is used when an outcome variable has more than two categories. In this
case there is more than one logistic regression equation in the estimated model. The number of logistic
regression equations is equal to the number of categories minus one. Equations are built for all categories
except the last, which is used as a reference category. Each equation predicts the probability that a case is
(or is not) in a particular category (Tabachnick and Fiddell 1996, p. 599).

In order to examine if certain trends were developing in the incidence of racial discrimination
over the last two decades, we compared the characteristics of racial discrimination cases between the two
decades. This was done by applying chi-square tests on data from the two sub-periods (1980s and 1990s)
and by including time period as a binary categorical variable in the multinomial logistic models. We
present below the main results obtained.

33
 Table 1

Distribution of Legal Cases by Jurisdiction, Nature of Alleged Discrimination, Industrial Category,

Occupational Category of Complainants, Complainant’s Racial Group, Sex of Complainant,
Employment Stage, and Ownership of Organization*

1980-89 1990-99 Total Period (1980-1999) Chi-Square Tests

(%) (%) (%)
1- Jurisdiction (N=49) (N=67) (N=116)
B.C. Human Rights Council 14.3 23.9 19.8
Federal Human Rights Tribunal 10.2 29.9 21.6 Chi-Square = 21.65
Nova Scotia Board of Inquiry 6.1 6 6 D. of Freedom = 5
Ontario Board of Inquiry 65.3 25.4 42.2 Significant at P=.001
Quebec Provincial or District Court 0.0 9 5.2
Saskatchewan, Alberta, and Manitoba Boards of Inquiry 4.1 6 5.2

2- Nature of Alleged Discrimination (N=47) (N=66) (N=113)

Dismissal and Lay-off 36.2 28.8 31.9
Denied Promotion 10.6 9.1 9.7 Chi-Square = 4.61
Racial Harassment 8.5 16.7 13.3 D. of Freedom =4
Racial Harassment and Dismissal 14.9 25.8 21.2 Not Significant
Denied Access to Employment 29.8 19.7 23.9

3- Industrial Category (N=48) (N=70) (N=118)

Community, Business and Personal Services 52.1 25.7 36.4
Manufacturing, Mining, Quarries, and Construction 16.7 17.1 16.9 Chi-Square = 11.56
Public Administration 16.7 42.9 32.2 D. of Freedom = 3
Transportation and Communication and Other Utilities 10.4 8.6 9.3 Significant at P=0.01
Trade 4.2 5.7 5.1

4- Occupational Category of Complainant (N=47) (N=68) (N=115)

Blue Collar 48.9 38.2 42.6
Crafts and Formen/women 21.3 4.4 11.3 Chi-Square = 15.42
Labour/Genaral and Operatives 27.7 33.8 31.3 D. of Freedom = 3
White Collar 51.1 61.8 57.4 Significant at P=0.0025
Clerical and Services 29.8 14.7 20.9
Professional/Managerial/Technical/Sales 21.3 47.1 36.5
5- Complainant’s Racial Minority Group (N=45) (N=63) (N=108)
Aboriginal Indian 2.2 12.7 8.3
Black 51.1 28.6 38.0 Chi-Square = 12.61
Caucasian 6.1 14.3 11.1 D. of Freedom = 5
Far Eastern (Chinese, Phillipino, Fijian) 8.9 4.8 6.5 Significant at P=0.05
South Asians (Indian, Pakistani, etc) 31.1 31.7 31.5
Other Minority Groups (Middle Eastern, Latin American) 0.0 7.9 4.6

6- Sex of Complainant (N=46) (N=65) (N=111) Chi-Square = 0.081

Male 71.7 69.2 70.3 D. of Freedom = 1
Female 28.3 30.8 29.7 Not Significant

7- Ownership of Organization (N=49) (N=69) (N=118) Chi-Square = 0.996

Government, Quasi-Government, Crown Corporations, and D. of Freedom = 1
Publicly Funded 42.9 52.2 48.3 Not Significant
Private Company 57.1 47.8 51.7

8- Stage of Employment Relationship (N=47) (N=67) (N=114) Chi-Square = 1.648

Pre-Employment 29.8 19.4 23.7 D. of Freedom = 1
Post-Employment 70.2 80.6 76.3 Not Significant

9- Decision Ordered (N=48) (N=68) (N=116) Chi-Square = 0.232

Discrimination Found 54.2 57.4 56.0 D. of Freedom = 1
No Discrimination Found 45.8 42.6 44.0 Not Significant

10- Formal Relationship Discriminator (N=46) (N=71) (N=117)

Supervisor 63.0 50.7 55.6 Chi-Square = 8.458
Co-Worker 8.7 9.9 9.4 D. of Freedom = 3
Supervisor and Co-Worker 4.3 23.9 16.2 Significant at P=0.05
Selection Committee 23.9 15.5 18.8

* The total number of cases varies between the different variables due to lack of required data in some case reports.

34
Quantitative Analysis of Racial Discrimination Complaints

The quantitative analysis of the extent and trends in racial discrimination complaints driven legal
cases is presented in three main areas including general characteristics of cases, characteristics of
respondents, and characteristics of complainants. Table 1 shows the distribution of the 119 legal cases by
ten different factors including jurisdiction, industrial category of organizations involved, nature of the
discrimination complaint, stage of employment relationship, decision ordered, ownership of
organizations, occupational categories of complainants, visible minority group of complainants, sex of
complainants, and formal relationship with discriminator.

1. Characteristics of racial discrimination complaints. Our analysis shows that in 65 cases

(56% of total), tribunals found in favour of thecomplainants. The success rate of complaints did not
change significantly between the two sub-periods (1980s and 1990s). The chi-square test (Table1) shows
that the difference in success rate of complaints before human rights tribunals between the two periods is
not statistically significant. Table 1 shows that dismissal, refusal to hire, and harassment were the most
prevalent causes of complaints of racial discrimination in employment. It is interesting to note that the
percentage of cases involving racial harassment almost doubled between the 1980s and the 1990s from
23.4% to 42.4% of total cases. More than one third of all legal cases of racial discrimination alleged
harassment. This, in part, is what we explore next in our logistic regression analysis.

Table 2 below indicates the results of multinomial regressions with the nature of alleged
discrimination as the dependent variable. This variable has multiple categories. These are: dismissal,
harassment, and harassment with dismissal and denied access to employment. Dismissal: For dismissal
category of nature of alleged discrimination, positive and statistically significant coefficients were found
for ‘industrial category = community, business and personal services’ and for ‘industrial category =
manufacturing, mining, and construction.’ The odds ratios for the two variables were 11.39 and 11.70,
respectively. This means that the cases alleging dismissal due to racial discrimination are more likely to
be brought by the relevant Human Rights Commissions against organizations in these two industries than
others. The regression coefficients for time period and occupational category of complainant were not
statistically significant in this regression equation.

Racial Harassment: For legal cases involving racial harassment, (table 2) the regression
coefficients for ‘industrial category = public administration’ is positive and statistically significant
indicating that cases alleging harassment are more likely to be brought against organizations in the public
administration sector than other industrial groups. The odds ratio is 41.8 indicating that the odds of racial
harassment complaints in public administration are more than 41 times higher than those for other
industries. The regression coefficient for time period in table 2 was negative with marginal statistical
significance (p = 0.067). This indicates that the odds of racial harassment cases are lower in the 1980
compared to the 1990s. Also the regression coefficient for ‘industrial category = community, business,
and personal services’ is positive with an odds ratio of 11.8. However, this coefficient has only a marginal
statistical significance at p = 0.080.

35
 Table 2
1
Multinomial Logistic Regression for nature of alleged discrimination

Variables Dismissal Harassment Harassment with Dismissal

Coefficients B Chi- Sig. Exp(B) B Chi- Sig. Exp(B) B Chi- Sig. Exp(B)
Square Square Square
Intercept -0.903 0.577 0.448 2.418 2.278 0.131 0.253 0.50 0.823
Period=1980s -0.727 1.239 0.266 0.48 -1.693 3.349 0.067** 0.184 -2.160 6.833 0.009** 0.115
** ** **
Industrial Categ.=Community, Business, & Pers. Service 2.433 4.709 0.030 11.39 2.470 3.067 0.080 11.82 2.468 4.434 0.035 11.804
Industrial Categ.=Manufact., Mining, & Construction 2.460 4.67 0.031** 11.70 1.379 0.871 0.351 3.969 2.412 3.819 0.051** 11.161
Industrial Categ. = Public Administration 1.653 1.731 0.188 5.22 3.734 5.573 0.018** 41.86 1.182 0.727 0.394 3.260
Occup. Categ. Of Compln’nt= Blue Collar - Crafts 0.662 0.271 0.603 1.94 1.958 1.234 0.267 7.088 0.293 0.047 0.829 1.341
Occup. Categ. Of Compln’nt= Blue Collar – -0.223 0.056 0.82 0.80 1.187 0.934 0.334 3.278 -1.548 2.275 0.131 0.213
Labour/Gen’rl
Occup. Categ. Of Compln’nt= White Collar – Professional -0.654 0.547 0.460 0.520 -1.838 2.174 0.140 0.159 -2.714 6.761 0.009** 0.066
Model fit Chi-Square (21, N=101) = 34.89; P = 0.029; Cox and Snell = 0.29; NagelKerke = 0.31; McFadden = 0.13

(*) Reference categories for Period is 1990s; with reference to Industrial Categ, Transportation and Communications, and Occupational Categ. of Complainants, White Collar –
Clerical & Services , respectively
(**) Significant at the given P-value

1
Denied access to employment is used as reference category for the dependent variable. Denied promotion was dropped
due to the low number of effective cases in the sample.

36
 Racial Harassment with dismissal: For Racial harassment with dismissal, as the outcome
variable, the estimated coefficient for time period in this equation is negative and statistically significant
indicating that the odds of this type of complaints were less during the 1980s than the case during the
1990s. Also, the odds of these cases are lower in the case of white-collar professional complainants
compared to other occupational category groups. In terms of industrial category, the regression
coefficients were positive and statistically significant for ‘industrial category = community, business, and
personal services’ and for ‘industrial category = manufacturing, mining, and construction.’ The odds
ratios are 11.8 and 11.2 for these two industrial groups, respectively. Therefore, the odds are higher for
legal cases of harassment coupled with dismissal to be brought against organizations in these two groups
compared to other industrial groups.

About 76% of all cases are related to post-employment discrimination and only 23.7% to pre-
employment discrimination. Table 1 shows that the chi-square test found no statistically significant
difference in the distribution of cases between the two periods with respect to stage of employment
relationship.

In terms of remedies ordered in the cases where the complaints were successful (not shown in
table1), the most commonly ordered remedies are monetary compensation for pain and humiliation (N=39
or 59.1% of successful cases) and compensation for lost wages/salary (N=28, 42.4%). The percentage of
cases in which compensation for pain and humiliation, (sometimes called general damage payment,
punitive damages, or compensation for loss of health and welfare), remained almost the same over the
two sub-periods. The tendency of tribunals/courts to order punitive damages in compensation for
discrimination is an important development in the last two decades that is making discrimination in the
workplace somewhat more costly for employers than was the case before. Increasing punitive damage
payments and making the organisation responsible for the payment can be a very effective way of
encouraging organisations to apply policies to combat racial discrimination.

2. Characteristics of respondents. In terms of the breakdown of cases by the industrial category

of organisations, Table 1 shows that the highest percentage of complaints (N= 43, 36.4%) took place in
the community, business, and personal services followed by the public administration sector (N= 38,
32.2%) and manufacturing, mining, quarries, and construction (N=20, 16.9%). Our results show an
astonishingly high number of complaints brought against organisations in the public administration
sector. There seem to be a surge in such complaints in the 1990s. The percentage of complaints against
organisations in the public administration sector rose from 16.7% of the total number of complaints
during the 1980s to 42.9% during the 1990s. The difference in proportion of cases in the public
administration sector between the two decades is statistically significant (at P = 0.01) as shown by the
chi-square test. This indicates that there has been a trend of increasing cases in the public administration
over the study period.

In the majority of cases (N= 65, 55.6%) the person accused of discrimination was the supervisor
of the complainant. In other cases (N=19, 16.2%) both a supervisor and co-workers were involved. In 22
cases (18.8%) selection committees were named as respondents. The chi-square test indicates that there is
a statistically significant difference in the proportion of cases in the four categories between the 1980s
and the 1990s (chi-square = 8.5, P = 0.05) as shown in Table 1. There has been a relative decline in the
cases with supervisors and selection committees named as respondents and a relative increase in cases
where supervisors and co-workers and co-workers alone are accused of discrimination. The increase in
cases where co-workers are involved in discriminatory treatment of members of racial minority groups
can be partly explained by the increase in racial harassment cases which usually involve co-workers. In
58.4% of the cases, one person was involved in the alleged discriminatory behavior. The percentage of
cases decline as the number of individuals increase.

37
 3. Characteristics of Complainants. Our sample shows that the majority of racial discrimination
complaints came from white-collar employees (N= 66, 57.4%). Among white-collar complainants, the
majority of complaints (about 6 out of 10 or 63.6%) came from employees in the professional,
managerial, and technical group. The data for the two sub-periods indicate that the number of cases
involving blue-collar complainants have decreased from 48.9% to 38.2% between the 1980s and the
1990s. On the other hand there was a relative increase in cases brought by white-collar employees
especially those who are in managerial, professional and technical jobs. These results were also supported
by the chi-square test that was significant at the P = 0.05 level as shown in Table 1. This trend indicates
an increasing complexity in the racial discrimination problem as members of VMs are increasingly
competing on high level professional and managerial jobs that are still disproportionately dominated by
white males.

We found that cases filed by blue collar employees are more likely to be successful before the
human rights tribunal than those filed by while-collar employees. The success rates were 38.5% and
66.6% for white collar and blue-collar complainants, respectively. This can be explained by the high
levels of complexity of workplace relationships and intensity of competition on career opportunities at the
managerial, professional, and administrative levels compared to the case in production and shop floor
level jobs. The strong competition on limited career opportunities at the managerial levels may lead some
members of racial minority groups to misinterpret actions taken by others - that are perceived biased - as
racial discrimination due to their misunderstanding of the reality of their treatment. These
misinterpretations and misunderstandings are often uncovered in the course of court proceedings.

In terms of the breakdown of cases by complainants’ racial minority groups, the largest share of
complaints (N=41, 38%) came from Blacks followed by South Asians (N= 34, 31.5%) over the two
decades. The percentage of complaints from Blacks dropped from 51.1% to 28.6% between the 1980s and
the 1990s. The drop in complaints by Blacks was accompanied by an increase in complaints from other
racial groups (including Latin Americans, Aboriginal Indians, Caucasians, and others). This trend is
statistically significant as shown by the chi-square test in Table 1 (chi-square = 12.6, P = 0.05). This
increase in the diversity of complainants is expected to continue as more workers from diverse racial
minority groups continue to enter the labor market (Statistics Canada, 2000).

We also computed the distribution of complaints by the largest three minority groups (i.e; Blacks,
South Asians, and Aboriginals) across various types of industries (not shown in table1). For Blacks the
largest number of complaints were against organizations in the community and business services sector
(N=17, 41.5%) followed by manufacturing and public administration with each sector accounting for
21.9% of complaints (N=9 out of 41). For South Asians, the majority of complaints came from public
administration organizations (N=15, 42.9%) then community and business services (N=10, 28.6%), and
manufacturing (N=6, 17.1%). For Aboriginals 4 out of the 9 complaints (44.4%) came from public
administration organizations and 3 complaints (33.3%) from organizations in the community and business
services sector. We found that for Black complainants in 20 out of 41 cases (48.7%) human rights
tribunals judged in favor of the complainant. For South Asians the success rate was 62.8% (22 out of 35
cases) and for Aboriginals 7 out of 9 cases (77.7%) received judgments in favor of complainants. An
overwhelming majority of cases (N= 78, 70.3%) involved male complainants.

Patterns of Behavior Leading to Discrimination Complaints

As indicated above, there are several potential causes for racially discriminating behavior.
However, unlike other types of offenses and illegal behaviors, individuals committing racial
discrimination almost never openly admit of their wrong doing even after they are legally charged with
discrimination. This may be due to the fact that offenders in these cases usually have their own

38
interpretations of the situation that may make them view their actions as legitimate and justifiable.
Another reason is the social stigma associated with such attitudes and behaviors despite its widespread
existence in the society. Therefore, courts and human rights tribunals rely on careful analysis of all the
circumstantial evidence in order to make their judgments.

In order to understand the specific interactions and workplace relationships that lead to
discriminatory treatment and discrimination complaints we provide below a qualitative analysis of
selected incidents of racial discrimination in employment in light of the theoretical perspectives presented
above in this paper.

Systemic Discrimination. In Najar v. Cook (1992), a case of direct access discrimination based
on accent or ethnic origin, the complainant (a young man who came to Canada as refugee from Iran)
telephoned to apply for a job of traffic controller early during the time specified by the company for
applicant to call. He was told that all jobs were already filled. When Mr. Najar’s neighbor (a white
Canadian lady) called later, that same day, she was told that the position was still available. This is a clear
case of intentional denial of access to employment based on ethnic origin. The cause for this behavior
could be simply prejudiced attitude or possibly some generalized belief about inability of members of
specific minority groups to perform the job duties effectively as indicated by the theory of statistical
discrimination. Other cases of access discrimination are consistent with the latter interpretation. For
example, in the case of Daisly v. Fantasy cuts (1988), the two owners of a beauty salon, who were both
Chinese women, decided to terminate the employment of a Black lady shortly after they hired her because
they believed that “she was scaring away customers” and that if they hired a “nice White girl” business
would improve. Another example of denied access to employment that supports the notion of statistical
discrimination is found in Scott v. Wheeler (1986). The company in this case which had problems with a
Black employee working on a project decided to stop hiring any Blacks in this project taking race as an
indicator of potential work performance problems.

As indicated earlier, discriminatory behavior in employment contexts may be motivated in some

cases by a state of perceived conflict in the mind of specific individuals in specific work situations
leading them to treat others with disfavor. We described several factors leading to conflict and
subsequently discrimination including prejudiced attitudes (taste for discrimination), statistical
discrimination, competition for job opportunities, and differences in cultural values, language, and
interpersonal communication styles.

The published legal cases provide a variety of examples of situations involving conflicts that
could be caused by various factors as suggested above. For example, in Jahal and Grewal v. Cowichan
(1987) a couple, of East Indian origin, were hired as laborers in a twenty week project and were
terminated after only four days on the job based on the recommendation of their foreman. Several
colleagues of the complainants testified that they were subjected to differential treatment and racial slurs
by the foreman since their first day at work despite the fact that they had “worked hard.” The chair of the
board of inquiry found no reasonable ground for the termination and ordered the respondent to pay each
complainant the complete amount of lost wages due to the discriminatory action.

Other cases are far more complicated and involve more subtle motives for discrimination. The
case of Grover v. National Research Council of Canada (NRC) (1992) is a telling example of such cases.
Dr. Grover, a research physicist who specialized in optics, was hired by NRC in 1981. Dr. Grover was
described by the tribunal as “an outstanding scientist with high peer recognition internationally and a
scientist who was conducting, on behalf of NRC, important work in several fields of optics.” Dr. Grover’s
problems started in 1986 when he applied for promotion. The tribunal in this case found that there was a
concerted effort by two directors at NRC to thwart Dr. Grover’s career progression by denying him the

39
promotion that he obviously deserved. His testimony was accepted by the tribunal in which Dr. Grover
testified that his director told him that “he was ambitious and a troublemaker.” He further said that he
knew that Dr. Grover wanted to be a section head or a director and that he would see to it that Dr. Grover
would never become either. This is a clear example of a deep conflict that could be caused by a complex
set of job related factors such as professional jealousy and fear of competition for positions of power and
authority. Also, there is an indication of personal/emotional conflict that could be caused by prejudiced
attitudes as well as cultural differences and differences in values and interpersonal communication styles.
A personal and emotional conflict situation is also shown in the case of Almeida v. Chubb Fire Security
Division (1984), where one of two reasons given by the respondent for denying Mr. Almeida promotion
was that “he didn’t get along with others.” Again the tribunal found no facts to substantiate this claim and
ruled that the complainant was discriminated against based on his race and colour.

A highly publicized case involving several complainants brought by the National Capital Alliance
on Race Relations (NCARR), on behalf of a group of Health Canada employees, reveals a similar pattern
of discrimination by top management at Health Canada against highly qualified employees who were
members of visible minority groups. The significance of this case stems from the numerous incidences of
discrimination that took a systemic nature, even at the highest levels in the organization, and from the fact
that Health Canada is a major employer in the federal public sector (Health Canada employs 7306
workers and has an estimated budget of close to $3 billion in the fiscal year 2001-2002).(4) In NCARR v.
Canada (Health and Welfare) (1997), a tribunal found that Health Canada discriminated on a systemic
basis against VMs employees by failing to promote them to senior managerial positions. Testimonies in
this case show a concerted effort by top officials at Health Canada directed at preventing members of
VMs from achieving promotion to managerial positions. The tribunal found that some senior managers at
Health Canada perceived members of VMs as unfit for management positions and that “cultural
differences” represented their main reason for having such perception.

Dr. Liston, the Assistant Deputy Minister, states in his testimony:

“Employees who are being considered solely for ‘technical’ positions seem to fare better than
when being considered for ‘management’ positions. The cultural differences are minimized when we are
only looking for the scientific approach. However, when we start looking for the ‘soft skills’ such as
communicating, influencing, negotiating…quite often their cultural heritage has not emphasized these
areas and they are at a disadvantage.”

Obviously, such stereotyping is dangerous and can lead to discrimination even without the
intention to discriminate, as it clouds the ability of people holding this view to fairly assess members of
racial minorities based on their individual merits.

Racial harassment. We indicated in the previous section that complaints involving racial
harassment in the workplace constituted more than one third of the total complaints filed over the last two
decades and that the number of these cases have almost doubled between the 1980s and the 1990s. Racial
harassment deserves special attention in our attempt to understand the causes of racial discrimination in
employment, since this behavior is directed at a more personal level and therefore is probably more
damaging emotionally, as it attacks the dignity and self esteem of individuals subjected to racial
harassment.

Racial harassment cases do not usually involve making decisions to choose between members of
minority and majority groups as is the case in other situations that involve discrimination in decisions
regarding hiring and promotion or other aspects of the employment conditions of members of racial
minorities. Also, racial harassment is most commonly practiced by co-workers who have no official
authority over the recipients of this treatment. It is mainly motivated by the sheer discomfort of

40
association with members of racial minorities because they are perceived as incompatible and inferior.
This behavior can be explained as resulting from a perception of personal as well as cultural and socio-
economic superiority in the mind of specific individuals towards members of VMs that justifies in the
mind of perpetrators treating those individuals as inferiors. Since individuals with such prejudiced views
find themselves having to associate with members of VMs in the workplace, they tend to use racial slurs
and derogatory treatment as a way of expressing and communicating their feeling of their own superiority
and their resentment to treating members of VMs as equal co-workers and equal individuals.

Respondents in racial harassment cases usually admit to the racial slurs and name-calling, but
they often contend that it is not racially motivated, therefore not meant to discriminate against members
of VMs. They usually argue that racial jokes and name-calling is just done ‘in fun’, ‘between friends’,
and ‘not meant to give offence’. For example, this was the justification given by respondents in the case
of Naraine v. Ford Motor Co. of Canada (1996). However, the board of inquiry in this case found that the
workplace was characterized by an “extremely racially hostile environment.” The evidence presented in
this case indicated that employees at Ford Canada plant referred to their African-Canadian co-workers as
‘nigger’, ‘little nigger’, ‘lazy nigger’, and ‘boy’. Such derogatory terms cannot be taken as ‘friendly’ and
‘unhurtful’ but are more likely to be an expression of intolerance and prejudiced attitudes. Mekenzie
(1994) indicates that there are two related effects of racial verbal abuse. The first is felt at the level of the
individual victim and attacks his or her dignity and self-worth. The second is felt at the level of the
workplace and involves creating a “climate of oppression” in which true equality of employment becomes
difficult.

The role of recipients of discriminatory treatment. Most of the theoretical perspectives on

racial discrimination in employment that we discussed earlier in this paper viewed victims of racially
discriminating treatment largely as passive recipients of such treatment. However, it is important to
highlight the role of recipients of perceived or potential discriminatory treatment in shaping the course of
subsequent interactions in the specific situation and the outcome of such interactions. Complaints of racial
discrimination are, of course, a result of individuals or groups deciding to challenge the perceived
discrimination to the extent of filing a legal complaint against the alleged discriminator. However,
individuals subjected to discriminatory treatment do not always file complaints. In many cases they do
not even resist or fight the discrimination openly and resort to accommodation instead as a defense
mechanism. The reason is that such confrontation may involve high costs for members of minority
groups, in terms of emotional stress, as well as potential further deterioration of employment conditions
due to retaliation. This explains why it is usually common for individuals, who are subjected to
discriminatory treatment, to maneuver their relationships with perpetrators of such offences in a way that
attempts to reduce the conflict and minimize their losses due to discrimination.

In situations where strong organizational commitment to stop discrimination is perceived to be

lacking, or policies to penalize discriminators are ineffective or nonexistent, it may be seen a wiser tactic
from a personal utility perspective, for individuals subject to discrimination to avoid open confrontation.
They may rely instead on avoidance and other tactics, that mitigate the degree of conflict with the
discriminating party, in order to reduce the effects of discrimination on their career achievements. An
example of this non-confrontational behavior is demonstrated by Dr. Dhillon’s testimony in the case of
NCARR v. Health Canada (1997). Dr. Dhillon testified that he was continuously denied promotion and
he believed that this was caused by racial discrimination. Although he had appealed the decision not to
promote him, on a number of occasions, he never alleged discrimination in any of his appeals and had
never filed a legal complaint. Dr. Dhillon’s explanation for his decision not to openly defy discrimination
is that “he thought that such course of action would cause bad feelings and cause unnecessary antagonism
in the workplace.” He believed that “he had to work within Health Canada and, therefore, had to live
with the situation” (NCARR v. Health Canada 1997).

41
 On the other hand, the legal cases reveal many situations where individuals who have not
actually been subjected to discrimination filed complaints of racial discrimination due to misconceptions
or misinterpretations of the reality of their treatment. For example, in the case of Persard v. Sudbury
Regional Police Force (1995), the complainant alleged that he was discriminated against when he was
refused employment as a police officer, despite the fact that he was given a special treatment by the police
service, in processing his application for employment. The special treatment was due to a concern of the
chief of the police force about the lack of visible minority officers in the police service and also due to an
intervention in support of Mr. Persard’s application by a member of parliament (M.P) in the region.

In some situations, members of racial minority groups receive a specific unfavorable treatment
from an employer (or his agents) because of legitimate reasons, such as not meeting the requirements of
the job or violations of the employer’s rules or policies. However, because the complainant does not agree
with the justification given by the employer, he/she may attribute the perceived unfair treatment to racial
discrimination. Examples of such situations are found in numerous cases where boards of inquiry found
no evidence of racial discrimination (see for example Daniels v. Hamilton-Wentworth Police Services
Board, 1996; Khalil v. Ontario College of Art, 1997). This type of misperception can be exacerbated by
lack of trust and rapport between the two parties in situations where relationships are marred with
personal conflicts and interpersonal communication difficulties. Complainants in these cases tend to feel
that somebody is conspiring against them and consequently find their racial group membership as an
available explanation for their perceived mistreatment. It may be beneficial in order to reduce the
incidence of this type of behavior, where possible, to encourage better communication with members of
racial minority groups and to provide them with clear explanation for decisions that they may perceive as
unfair or discriminatory.

Conclusions And Recommendations

Our review of previous studies and analysis of legal cases indicate that racial discrimination in
employment is still a problem in Canada. Incidents of racial discrimination take place in various types of
organizations and industries in both the public and private sectors. It also affects individuals from various
racial groups including both white-collar and blue-collar workers. This phenomenon is complex and
multidimensional. There are various theoretical perspectives that examine the causes and consequences of
racial discrimination in the workplace.
Analysis of 119 legal cases adjudicated by various boards of inquiry across Canada produced several
important findings and revealed a number of trends in the incidence of racial discrimination over the last
two decades. Among the most significant conclusions are the following:

Conclusions

1- Most of the legal cases of racial discrimination over the last two decades (76%) are related to treatment
discrimination (after hiring). At lease three factors likely account for this result. First, human rights
legislation and employment equity policies have been in place in Canada for a relatively long period of
time. These laws and policies are believed to be relatively more effective in eliminating “access
discrimination” than the more subtle and harder to combat “treatment discrimination.” Second, people
usually have more to lose by accepting discriminatory treatment without complaint when they are
employed than when they are applying for employment. Third, it is more likely for a person to
recognize if he or she is being treated unequally while at the post-hiring stage rather than at the pre-
employment stage.
2- Majority of the complaints (57.4%) were brought by white-collar employees. There was a relative
increase in the percentage of complaints brought by white-collar (relative to blue-collar) employees
between the 1980s and the 1990s from 51.1% to 61.8%. This may indicate that the racial

42
 discrimination problem is moving to a new level. As members of racial minorities succeed in moving
to higher level managerial and professional positions they start to face discrimination at this level. The
higher number of complaints by white-collar employees may be caused by strong competition on job
opportunities among these employees. This could lead to more friction among employees, which
could motivate discriminatory behaviour. Policies to combat racial discrimination should consider this
change in the nature of discrimination today.
3- Complaints involving racial harassment coupled with dismissal increased over the decade of the
1990s. The increasing trend in racial harassment cases is indicative of increasing levels of workplace
conflict as racial and ethnic diversity continues to increase in Canadian workplaces. Such cases were
considered a rarity during the 1970s (Mackenzie 1994). Racial harassment complaints were more
successful before human rights tribunals than other types of racial discrimination complaints.
4- Analysis of legal complaints over the past two decades reveals that the nature of alleged
discriminatory behaviours tends to vary between organizations in different industry groups.
Complaints of dismissal due to racial discrimination are more likely to be brought against
organizations in the ‘community, business, and personal services’ and manufacturing, mining, and
construction’ sectors compared to other industrial groups. Complaints of racial harassment are more
likely to be brought against organizations in the public administration sector.
5- The percentage of complaints involving public administration organizations rose from 16.7% during
the 1980s to 42.9% during the 1990s. It is interesting to note here that in 1998 the public
administration sector employed only 5.5% of the total labour force in Canada compared to 38.5%
employed by the services sector (including health, educational, business, and other service industries)
and 15.7% employed by the manufacturing industries (Statistics Canada 2000).
6- An overwhelming majority of cases (70.3%) involved male complainants. Many explanations can be
given for this observation. It could be due to the lower labor force participation levels for women than
men especially among racial minority groups. Furthermore, the fact that women are generally
concentrated at lower level jobs where there is less competition on job opportunities could offer
another explanation for the low level of complaints by females.
7- In the majority of cases (55.6%), the complaints were against supervisors. In 16.2% of cases
supervisor and co-workers were named as respondents.
.
Policy recommendations: To help combat racial discrimination in employment, we make the
following policy recommendations:

1- The fact that systemic discrimination (that is, both intentional and unintentional discrimination) exists
means that combating racial discrimination should focus on creating a workplace environment with
zero tolerance policies for systemic discrimination. These policies should be enforced consistently
and should be built into the reward/penalty system within any organization. The most recent report
that reviewed the Canadian Human Rights Act recommended treating issues of discrimination in the
workplace the same way that workplace health and safety policies are treated (La Forest, 2000). A
person subjected to racial discrimination or harassment in the workplace ought to be legally allowed
to refuse to work in such an environment on the basis that this treatment represents a risk to his/her
psychological/emotional health and well-being. However, in the long run education and training are
maybe the best means to eliminate prejudiced attitudes and misconceptions that lead to racial
intolerance and discrimination.
2- Combating the racial discrimination in employment problem requires bold and strong measures by the
government to enforce anti-discrimination legislation. A recent task force report (La Forest, 2000)
indicated that the human rights laws are not being enforced as vigorously as necessary due to lack of
resources. The current complaint-driven approach to enforcing human rights law may be inadequate.
The subtle and often unintentional nature of discrimination may make it difficult for victims to realize
that they have been discriminated against. Also, the lengthy complaint processing procedure may

43
 cause victims to fear retaliation and further deterioration in their employment conditions before their
problems are properly addressed and remedial measures are ordered. An alternative approach may
involve speeding up the processing time for human rights complaints, establishing a follow-up
mechanism to ensure against retaliation, and requiring that organizations be more directly involved in
ensuring fair treatment of their visible minority employees.
3- The fact that treatment discrimination persists and is more prevalent than access discrimination
suggests that current employment equity programs that focus on removing barriers to employment
and require evidence on fair representation may not be sufficient. Other measures to ensure fair
treatment of members of racial minority groups such as mandatory surveys to assess organizational
culture and gauge the degree to which employees experience a discrimination free work environment
may be required. Empowering racial minority individuals and helping them resist and/or cope with
discrimination may be a useful policy direction as well.
4- Discriminatory behaviours vary in nature and contexts. Policies to combat racial discrimination
should take these variations into consideration. For example, combating racial harassment may
require certain interventions involving education and sensitivity training. On the other hand
discrimination by denying certain individuals promotion to managerial positions due to intentional or
unintentional discrimination may require different types of interventions. In order to aid policy
makers in this regard, we recommend that future research should focus on understanding the dynamic
processes and workplace contexts that may motivate different types of discriminatory behaviours.
5- Legislation should strongly emphasize the liability of employers for any acts of racial discrimination
by their employees.
6- The cost of racial discrimination should be high enough to encourage employers to adopt staffing and
managerial policies to prevent racial discrimination. We suggest increasing punitive damage
payments for victims of racial discrimination and holding employers liable for discrimination.
7- Various organizational policies and procedures including reward systems as well as other
interventions such as diversity training and effective communication should be designed to help
achieve equality.

Notes:
(1) The term visible minorities is used in Canada to refer to all persons who are of non-white or non-caucasian
origins including Blacks, Asians, and others. Visible minority groups include persons who were born in either
Canada or other countries. Aboriginal people, on the other hand, include three groups; First Nations (once called
Indians), the Inuits (who were the first Aboriginal group in Canada), and the Metis (descendants of early French and
English fur traders who married First Nations women).

(2) Systemic discrimination refers to direct or intentional discrimination and adverse effect discrimination that may
not be intentional but has disproportionate effect on women and minorities relative to non-minorities such as the
hiring requirement of height or weight (La Forest, 2000).

(3) Two points have to be stated here regarding these cases; first, the cases published in the CHRR do not
necessarily include all the cases adjudicated by Tribunals or boards of Inquiry across Canada since some of these
cases are not published. Second, the cases adjudicated by these tribunals represent only the tip of the iceberg in
terms of the total number of complaints brought to the various Human Rights Commissions across Canada. Some of
these complaints are resolved at early stages by conciliation and mediation or dismissed for being frivolous and
vexatious and other reasons. For instance, the Canadian Human Rights Commission (CHRC) received 7,450 signed
complaints between 1988 and 1997, excluding pay equity complaints. The CHRC dismissed or did not proceed with
about two-thirds of these complaints. It sent only 6 percent of the complaints to the Canadian Human Rights
Tribunal ((Report of the Auditor General of Canada, Sept. 1998, page 10-5). Thus, only a fraction of complaints go
to tribunals or boards of inquiries. As noted earlier, CHRR publishes most but not all of these cases.

44
4. Personal communication by telephone with an official at Health Canada.

5. We wish to acknowledge the partial assistance of the Social Sciences and Humanities Research Council (SSHRC)
and the Canadian Race Relations Foundation (CRRF). As well, we wish to thank Professors Rick Hackett and
Aaron Schat and Mr. Jiao, a doctoral student, for their helpful comments on an earlier draft of the paper.

45
 References

Abella, R. Silberman, Equality in Employment: The report of the Commission on Equality in

Employment. Ottawa: Supply and Services Canada, 1984.
Aggarwal, A. P, Sexual Harassment in the Workplace, Toronto: Butterworths, 1987.
Aigner, D.J. and Glen G. Cain, “Statistical Theories of Discrimination in Labor Markets.” Industrial
Labor Relations review, 3 (2), (1977).
Andiappan, Palaniappan, Crestohl Mark, and Jang Singh, “Racial Discrimination in Employment in
Canada.” Relations Industrielles 44(4), (1989), 827-849.
Arrow, Kenneth J., “What Has economics to Say about Racial Discrimination?” Journal of Economic
Perspectives 12(2), (1998), 91-100.
---------- “Models of Job Discrimination.” In Anthony H. Pascal, (ed.) Racial Discrimination in Economic
Life, Lexington, Mass.: Heath, (1972).
Auditor General of Canada, Report of the Auditor General of Canada
To the House of Commons. Ottawa, (September 1998).
Baker, Michael and Dwayne Benjamin, “Ethnicity, Foreign Birth and Earnings: A Canada/U.S.
Comparison.” Transition and Structural Change in the North American Labor Market. Michael G. Abott,
Charles M. Beach, and Richard P. Chaykowski, eds. IRC Press, Queens University at Kingston, (1997).
Becker, Gary S., The Economics of Discrimination. Chicago: The University of Chicago Press, (1957).
--------------, Human Capital: A Theoretical and empirical Analysis With Special Reference to Education.
3rd ed. Chicago: The University of Chicago Press, (1993.
Billingsley, Brenda and Leon Muszynski, No Discrimination Here? Toronto Employers and the Multi-
Racial workforce. Toronto: Social Planning Council of Metropolitan Toronto and The Urban alliance on
Race Relations, (1985).
Bloom, D. E., G. Grenier and M. Gunderson, "The Changing Labor market
Position of Canadian Immigrants," Canadian Journal of Economics 28(4b), (1995), 987-1005.
Boyd, M., "Gender, Visible Minority and Immigrant Earnings Inequality:
Reassessing an Employment Equity Premise," in V. Satzwick (ed).
Deconstructing a nation: Immigration, Multiculturalism and Racism in '90s
Canada, Halifax: Fernwood Publishing. pp. 279-321, (1992).
Canadian Human rights Commission, The Canadian Human Rights act: A Guide. Internet: www.chrc-
ccdp.ca, (1999).
“The Constitution Act.” 1982. as enacted by Canada act 1982 ( U.K., c.11)
Darity, Williams A. Jr and Patrick L. Mason, “Evidence on Discrimination in employment: Codes of
Color, Codes of Gender.” Journal of Economic Perspectives 12(2), (1998), 63-90.
Elmslie, Bruce and Stanley Sedo. 1996. “Discrimination, Social Psychology, and Hysteresis in Labor
Markets.” Journal of Economic Psychology 17:465-478.
Fernandez, John P. 1981. Racism and sexism in Corporate Life: Changing Values in American Business.
Lexington Books.
Heckman, James J., “Detecting Discrimination.” Journal of economic Perspectives 12(2), (1998), 101-
116.

46
Henry, Frances, “The Dynamics of Racism in Toronto.” Department of Anthropology. York
University.Toronto, (1978).
--------------- and Effie Ginsberg, Who Gets the work? A Test of Racial Discrimination in Employment.
Toronto: The Urban Alliance on Race Relations and The Social Planning Council of Metropolitan
Toronto, (1985).
Heritage Canada, “Annual Report on the Operation of Canadian Multiculturalism Act,” 1999-2000.
Internet: www.pch.gc.ca., (2001).
Howland, Juliet and Christos Sakellariou, “Wage Discrimination, Occupational Segregation and Visible
Minorities in Canada.” Applied Economics 25, (1993), 1413-1422.
Hum, D. and W. Simpson, "Wage Opportunities for Visible Minorities in
Canada," The Income and Labor Dynamics Working Paper Series. Ottawa:
Statistics Canada, (1998).
Ilgen, D.R., and M.A. Youtz, “Factors Affecting the evaluation and Development of Minorities in
Organizations.” Research in Personnel and Human Resource Management: A Research Annual. K.
Rowland and G. Ferris eds. Greenwich, Conn.: JAI Press, pp. 307-337, (1986).
Jain, Harish C., Sloane, Peter J. & Hurwitz, Frank M., Employment Equity and Affirmative Action: An
International Comparison. M. E. Sharpe. Armonk: New York, (2003).
Jain, Harish C., “Race and Sex discrimination in employment in Canada: Theories, evidence, and
Policies.” Relations Industrialles 37(2), (1982), 342-366.
Jain, Harish, Parbudyal Singh, and Carol Agcos, “Recruitment, Selection and Promotion of Visible
Minority and Aboriginal Police Officers in Selected Canadian Police Services.” Canadian Public
Administration 42(3), (2000), 46-74.
Kramer, Roderick M., “Intergroup Relations and Organizational Dilemmas: The Role of Categorization
Processes.” Research in Organizational Behavior. L.L. Cummings and B.M. Staw eds. Vol. 13, pp.191-
228. Greenwich, CT: JAI Press, (1991).
La Forest, G.V., Promoting Equality: A New Vision. Report of the Canadian Human Rights Act Review
panel. Ottawa: Ministry of Justice, (2000).
Levitin, T., R.R. Quinn, and G.L. Staines, “Sex Discrimination Against The American working Woman.”
American Behavioral Scientist 15, (1971), 238-254.
Loury, Glenn C., “A Dynamic Theory of Racial Income Differences.” In P.A. Wallace and A. Lamond,
eds. Women, Minorities and Employment Discrimination. Lexington, MA: Lexington Books, (1977).
----------, “Why Should We Care About Group Inequality?” Social Philosophy and Policy 5, (1987), 249-
71.
----------, “Discrimination in the Post-Civil Rights Era: Beyond Market Interactions,” Journal of
Economic Perspectives 12(2), (1998),117-126.
Mackenzie, Ian R., “Racial Harassment in the Workplace: Evolving Approaches.” Canadian Labor &
Employment Law Journal 3, (1994), 287-311.

Meng, R., “The Earnings of Canadian Immigrant and Native Born Males.” Applied Economics 19,
(1987), 1107-1119.
Muszynski, Leon and Jeffrey Reitz, Racial and Ethnic Discrimination in Employment. Social Planning
Council of Metropolitan Toronto, Working Paper #5, (1982).

47
Ontario Human Rights Commission, The Ontario human rights code, Commission’s website:
www.ohrc.on.ca., (2000)
Ornstein, Michael, Ethno-Racial Inequality in the City of Toronto: An Analysis of the 1996 Census.
Institute for Social Research. Toronto: York University, (2000).
Pelled, Lisa H., Kathleen M. Eisenhardt, and Katherine R. Xin, “Exploring the Black Box: An Analysis of
Work Groups Diversity, Conflict, and Performance.” Administrative Science Quarterly 44, (1999), 1-28.
Pendakur, K. and R. Pendakur, "The Colour of Money: Earnings
Differentials Among Ethnic Groups in Canada," Canadian Journal of Economics 31(1), (1998), 518-548.
Perinbam, Lewis, “Embracing Change in the Federal Public Service.” A Report by the Task Force on the
Participation of Visible Minorities in the Federal Public Service. Ottawa: Treasury Board of Canada,
(2000).
Phelps, Edmond S. 1972. “The Statistical Theory of Racism and Sexism.” American Economic Review
62, (1972), 659-61.
Samuel, John, Visible Minorities and the Public Service of Canada. A Report Submitted to The
Canadian Human Rights Commission, Ottawa, (1997).
Statistics Canada, “Labor Force Survey”. Catalogue no. 71004-XCB, Statistics Canada Website
www.statcan.ca, (2000)
Tabachnick, Barbara G and Fidell, Linda S., Using Multivariate Statistics, 3rd edition. New York: Harper
Collins, (1996).

48
Legal Cases Cited

Action travail des femmes v. Canadian National Railway, (1987) 8 C.H.R.R. D/4210

Daniels v. Hamilton-Wenworth (Regional Municipality) Police Services Board), (1996) 31 C.H.R.R. D/18

Daisley v. Fantasy Cut, (1987) 9 C.H.R.R. D/4447.

Ganganelli Almeida v. Chubb Fire Security Division, (1984) 5 C.H.R.R. D/2104.

Gladstone Leslie Scott v. Foster Wheeler Limited, (1985) 6 C.H.R.R. D/2885.

Griggs V. Duke Power Co., (1971) 401 U.S. 424

Grover v. National Research Council of Canada, (1992) 18 C.H.R.R. D/1.

Johal and Grewal v. Lake Cowichan (Village), (1984) 8 C.H.R.R. D/3643.

Khalil V. Ontario College of Art, (191997) 31 C.H.R.R. D/121.

Najar v. Cook, (1992) 18 C.H.R.R. D/232.

Naraine v. Ford Motor Co. of Canada, (1996) 27 C.H.R.R. D/230.

National Capital Alliance on Race Relations v. Canada (Health and Welfare), (1997) 28 C.H.R.R. D/179.

Persard v. Sudbury Regional Police Force, (1993) 19 C.H.R.R. D/336.

Ontario Human Rights Commission and O’Malley v. Simpson Sears Ltd. (1985) 7 C.H.R.R. D/3102
(Ont.).

49
ASAC 2006 Janice Foley
Banff, Alberta Faculty of Business Administration
University of Regina

ADVANCING EQUITY IN CANADIAN UNIONS

This study reports equity activists’ perceptions of what factors currently hold
back progress on equity within Canadian unions. The main solution identified is
increased representation of equity groups on union decision making bodies which
will require structural changes and reallocation of resources to equity initiatives
and to member education and communication.

Introduction

There has been much discussion in recent years about whether or not Canadian unions will be able
to survive in the increasingly hostile environment they face. As a result of increasing global
competitiveness, employers have demanded greater flexibility, labour unfriendly legislation has been
passed, and jobs have been lost in the high-paid manufacturing sector where unions have historically been
strong. In addition, unions have yet to make major inroads into the service sector in Canada where jobs are
increasingly to be found. While union density has stayed fairly steady in Canada since the 1970s and
presently sits at around 30% (OECD, 2004), the most recent research indicates it has been trending
downward (Riddell and Riddell, 2003). There are fears that union density in Canada will fall to U.S.
levels, currently around 13% (OECD, 2004).

Current thinking on union renewal suggests that a many-pronged attack will be necessary to
assure the future of organized labour (Waddington, 1999). But among the recommendations made, priority
is given to new organizing (Fairbrother and Yates, 2003) because the strength of the labour movement lies
in its numbers. Increasing union density in the service sector is therefore a necessity, but presents a
number of challenges with which unions have little experience (Kumar and Murray, 2003; Murray, 2002;
2004). A major issue is that many of those who work in the service sector are youth, minority group
members or women, whose needs typically differ to some extent from those of the prototypical union
member, a middle-aged white male. To increase their appeal to these groups, unions must find a way to
address this need without alienating the traditional membership.

A Brief History of Equity Initiatives in Canadian Unions

Addressing the needs of non-traditional members has been an issue facing unions since the 1960s,
when women first began entering the labour force in large numbers (Cobble, 1993). Female union activists
began agitating for more inclusive unions when the Report of the Royal Commission on the Status of
Women was tabled in 1970. Early studies revealed that at that time, one of the more pervasive problems
women faced in unions was a lack of input into decision-making (Briskin and McDermott, 1993; White,
1993). Few women held elected positions, there were few female staff representatives and almost no
females on bargaining committees. Union policies designed to facilitate female participation in unions
were rare. For instance, union meetings were held in the evenings and childcare was rarely provided for
union events. The harassment of female union members by their male ‘brothers’ at union events and in the
workplace was commonplace (Foley, 1995). Structures established to address equity concerns, such as
women’s committees, were rare and whether or not they were established was determined by male-

50
dominated union executive committees. In the broader society, legislation regarding human rights,
employment equity and pay equity had not yet been passed. Occupational segregation was the norm, with
most women performing low-paid, administrative roles in the workplace.

This state of affairs provided much impetus for the nascent women’s movement in unions that
emerged in Canada in the early 1970s. Its immediate objectives were to get women’s committees
established; to ensure female representation on union decision making bodies was proportionate to the size
of the female membership; to amend union constitutions to afford women better protection against
harassment and discrimination, and to make their wording more inclusive; and to negotiate collective
agreement language favourable to women. As the face of the union movement became more diverse, the
objectives broadened, whereby it was deemed necessary for all equity groups to be appropriately
represented on union executives, and to have their interests actively pursued.

Looking back thirty years, it is apparent that progress has been made. Equity structures are now in
place in most unions, and new equity legislation has also been passed which rectifies, at least on paper,
many of the past injustices women and equity group members faced. However it is still the case that
women and minority group members are under-represented in union leadership positions, which suggests
that the changes to date have not been sufficient. Why this is the case, and what additional changes would
be needed to make equity a reality, is presented below, based on comments collected from a sample of
union equity advocates.

Methodology

Interested participants were solicited at a labour federation-sponsored educational event for

women activists that was held in the spring of 2004. Face-to-face interviews were conducted that summer
with twenty-one of these activists. Their years of experience in the equity field ranged from 1 to 20 years,
while years of active involvement in unions ranged from 2 to 30 years. They were asked to discuss their
views on the level of equity that had been achieved in their unions to date, and what factors were
advancing and holding back the equity agenda. The interview guideline can be found in Appendix 1. The
interviews, which generally lasted 60 to 90 minutes, were tape-recorded and transcribed. Thirteen of the
transcripts were selected for analysis, provided by activists who had at least 5 years of experience as
equity advocates. The transcripts were coded for emergent themes and patterns, and predominant themes
were identified.

Findings

The findings fell into two categories, outcomes, i.e. successes and failures, and explanations for
these outcomes.

Successes and Failures

The major success reported in the study related to the equity structures now in place in many
unions that made them more welcoming for equity group members. These included equity committees,
designated leadership and training seats for equity group members, and equity group conferences. In
addition, inclusive language and prohibitions against harassment and discrimination had been written into
union constitutions and collective agreements. Non-progressive unions were increasingly seen as
anomalies because it was now politically unacceptable in union forums to not support equity initiatives.
Remarks which in past years would have been applauded were no longer being well-received.

51
“…when these guys get up and make some really anti-equity remark, instead of being cheered they’re kind
of booed.”

It was no longer unheard of for equity group members to win leadership races. “It is so cool to see
someone … who wouldn’t have been … truly welcome ten years ago suddenly holding office.” In some
unions equity had become a priority, with activists reporting, “…equity is now a consideration in pretty
much everything we do.” In most cases, equity initiatives were funded, perhaps not generously, but
consistently. “There’s usually money.” “I think if we asked for it, it would get done.” It was also felt that
members in general were more aware of the issues facing equity group members. “It’s not just the people
at the top level that realize you need to be inclusive of equity issues.” They were also less fearful of equity
initiatives. “…they’re now realizing that this person has to write the same test to get that job, that they
have to put in years of service [to get promotions.]”

Nonetheless, informants felt there was a long way to go. Support for equity initiatives was still at
the conceptual stage and concrete action to make it a reality, such as setting goals and deadlines, was
minimal. “They don’t do anything that they’ve got [in the collective agreement] … to make it work.”
There was a sense that leaders were reluctant to pursue equity issues too aggressively if they were facing
backlash from the membership. One activist reported that the women’s committee in her union had been
required to hide its survey on barriers to equity within a broader union survey because, in the local
president’s opinion, “we have to be careful not to offend anyone.” Another informant mentioned that her
local leadership had wanted to refuse to allow her to carry a banner in the gay pride parade because, “…we
got a lot of negative backlash from members last year when that was carried in the parade.”

Even in the absence of backlash, informants suspected that support for equity initiatives might be
less than 100 percent. “There are still leaders that no matter how much information you gave them or
education, they’re just not going to accept it…There’s still that personality and … religious beliefs too.”
Not all leaders supported social unionism, with its attendant concerns about equity. “Someone … faxed…
saying, why are we getting involved in this, this isn’t a union issue, this is a social issue and it’s none of
our business and I’m a Catholic and I don’t believe in this …” Some informants felt that leader self-
interest also reduced support for equity initiatives. “…[T]here’s underlying fears predominantly by union
male leadership…they’re afraid of losing their seats, losing what they deem to be power and control.”

Activists also felt that equity issues were given low priority during bargaining. One informant
reported that her local president had told her, “… these equity committees can tell us what they want… I
don’t think the chairs at the bargaining table are going to listen to what that committee says.” It was quite
obvious as well that the membership, both male and female, continued to view female leaders with
suspicion, regardless of their competence. Some of the issues were, “Will men trust a woman? Will they
think she is too soft? Will she be able to be strong at the bargaining table?” Fears existed that female
leaders would “make changes that nobody can live with.” One informant mentioned that when an
extremely competent female was elected president in her workplace after the then-current male president
went out-of-scope, negative comments were made about “look what you left us with, a woman in control.”

At a more basic level, activists reported that it was difficult to track what kind of progress was
being made because records were incomplete. Equity group members could be hard to identify visually,
and might be reluctant to identify themselves. Nor were they necessarily convinced that separate equity
structures were needed. Informants reported being asked,”why are you on this [women’s] committee, we
don’t need that, we have equality.” But all the informants reported that women remained seriously
underrepresented on union governing bodies. “We have three staff reps, president, vice- president, second
vice-president, treasurer and secretary, there’s just myself and the secretary who are women.” “There’s
26 people around the [decision-making] table, five are women.” Few non-traditional members were

52
present at union conventions, “There are very few visible minorities, there are very few non-visible
minorities, there are very few equity groups represented.”

Although these problems were less apparent in some unions than in others, overall the
interviewees felt that equity group members still lacked voice and decision-making power in unions.
While at times progress was made, it seemed to erode in subsequent periods so there was little net
advance. Activists pointed out, for example, that abortion was once again on the table. “That’s a fight that
we fought in the seventies … why is it back?”

Explanatory Factors

The main explanatory factors identified as affecting progress on equity were equity and union
structures, resources, member education, and communication. Each is discussed below.

Shortcomings of equity structures. As noted, the establishment of equity structures was seen as
a major triumph in the eyes of the activists interviewed. Still, some unions had lagged well behind others
in terms of introducing these structures, and some of the activists noted that their unions had only recently
introduced them while other mentioned that they still did not exist. Where they did exist, respondents
identified a number of “holes” that effectively limited their utility for advancing equity.

One of the more commonly cited problems was the fact that designated seats on national
executives were not necessarily held by those most qualified to fill them. The worst scenario, according to
activists, was where incumbents were merely using equity positions to gain advancement in the union. “…
They’re doing it because … they’re tired of doing what they’re doing… so let’s try this.” Sometimes they
lacked the experience or links to their constituencies required to do a good job. “… [W]e have had a gay
man in our diversity seat for the last year and a half, who has not functioned at all… [He] was … not in
any way, shape or form ready.”

According to activists, the wrong people often got into these seats because, by constitution, they
were elected positions that had to be filled at national conventions. Only equity group members present at
convention were eligible to run for these positions, and their chances of success were greatest if their
supporters were in attendance. But there were no guarantees that the best candidates, let alone their
supporters, even got to convention, for reasons that will be discussed later. Another issue was that since
equity group representation on decision making bodies was low, seat-holders could be reluctant to push
their issues too vigorously for fear of ostracism or other negative repercussions.

Another problem with equity structures was that equity committees were only advisory to the
executive committee. While none of the activists reported that their local executives had imposed
mandates, in some cases guidelines for the committees, linked to the union’s strategic plan, had been
provided. All initiatives then had to relate to that plan. Committee initiatives could be blackballed, such as
in the case of resolutions to be forwarded to convention. “We had so many difficulties, barriers, fights,
struggles, over several resolutions that dealt with constitutional changes that would bring about, or at
least encourage, a bit more opportunity for equity seeking groups.”

The sentiment was also expressed several times that setting up committees and designating
positions was a highly effective way to marginalize equity issues and equity group members. “It’s always
been a great way to dump it on someone so they don’t have to deal with it anymore.” By putting these
structures in place, a union could show that it had fulfilled its responsibilities while avoiding making
equity a union-wide concern or assuming responsibility for outcomes. One informant commented, “…
committees are good but they also can marginalize the issues and if you’re not bringing them up in the
main discussion areas, you’re starting to lose a battle…”

53
 Also, while committee membership was often quite circumscribed by union constitutions in an
effort to ensure that all constituencies were represented on local and national committees, union executives
usually appointed some of the members. That could mean that committee members served out of a sense
of duty rather than a sense of commitment to the issues, about which they might be fairly uninformed.
Sometimes committee seats were reserved for regional or national representatives or members of other
equity committees. These linkages helped in one sense, in that the committees’ issues could be conveyed
to one another and to the regional or national executive, and vice versa, which allowed the various equity
groups to work collaboratively toward common goals. But on the other hand, the time pressures the equity
group representatives faced were severe, to the point that several of the informants felt that these
individuals were being abused. “We’re also responsible for…abusing people’s time, their energy… you
must be here and here before you can be there.”

They pointed out that this integrative structure could be self-defeating because “… it dictates that
they should be there, but they may have no interest…no ability …” Another problem was that it could
make members more hesitant to get involved. “The next person that may have had an interest is a little
hesitant about becoming involved because they don’t want to be used and abused.” Within the union
culture it was not considered acceptable for elected position-holders to delegate any of their
responsibilities to others, regardless of the others’ qualifications, willingness or availability, which
removed a possible solution to the time-crunch. Also, since they spent much of their time attending
meetings, they had little time to actually spend working on equity initiatives. “[The human rights
committee has] a chairperson who has no time to chair the committee.”

The time commitment required to carry out the duties attached to leadership positions was perhaps
the most frequently mentioned problem with union structures. But eliminating the formal linkages
between committees or to the national union executive committees where the bulk of the decision making
occurred was not an acceptable alternative because then, whether or not equity group interests would be
represented would solely depend on whether or not equity group members ran for, and won, local
elections.

Resource issues. The union executive committees determined budget allocations to the
committees. Budget levels were a source of concern because they rarely permitted more than 3 to 4
meetings per year. Since the members came from different geographical regions there were few
opportunities for additional face-to-face contact. Also, with a diversity of views on the committee, it could
often be difficult for the committee members to agree on a course of action. Several activists took issue
with the type of initiatives the women’s committees adopted, such as raising awareness of breast cancer
prevention techniques or the incidence of domestic violence, because they did not appreciably advance
equity. But there was a sense that the harder topics might be avoided in part due to underfunding. “I think
[the budget] should be more to substantially do anything different.”

A related problem was that there were frequently a number of equity committees established, and
they were all treated the same in terms of the number of meetings funded. They competed with one
another for extra resources and attention from the executive committee. “… [E]ach one is trying to
advance their own agenda and running over somebody else in the meantime.” So if one committee asked
for extra funding, the other committees would demand the same. This effectively meant that none of the
committees were given much leeway once the budget was set. At the time of data collection, one of the
respondents reported with great frustration that her women’s committee could not afford to provide
childcare as well as a banquet at an upcoming women’s conference. The union was reluctant to help out
because the constitution specified that all committee events were to be self-funding. Enrolment was
restricted to 50 attendees for the same reason, which again emphasizes the importance of the funding
allocated to the committee.

54
 Training issues. The lack of resources allocated to membership equity training was also seen as
problematic. “We’re not doing a lot of cultural sensitivity issue educating … we aren’t educating enough
on pay equity and women … we aren’t educating at all on solidarity and pride issues, there’s never been
any education on youth.” Informants believed that initial involvement in the union would lead to future
participation. “You get involved with one thing in the union and all of a sudden you get all the material, so
then you go ‘Oh gee, that might be interesting’ so then you … get involved with something else.” The lack
of general training afforded to union members was also an issue. “There’s no union education for people
who just want to participate on some level but not in a formal capacity.”

Training seats were not routinely reserved for equity group members either, despite the expressed
need to get more equity group representation on decision-making bodies. Most seats were filled on a first
come, first served basis which was a problem for anyone not monitoring the union events calendar, or who
did not know how to go about accessing the training they wanted. Lack of profile in the local could also
restrict training opportunities where there was a competition for training seats. “It’s a popularity contest
because there might be so many positions designated… and then you have to be voted to go… you would
have to have your support in order to get in.” It was also noted that male members continued to be more
interested in taking training, and that locals did not always make it easy for members with family
responsibilities to attend educational sessions, for example by providing childcare. Locals that were
sensitive to the problem were usually able to arrive at times and places that would enable the bulk of the
interested membership to attend.

Informants noted that the absence of, or inability to access, training was hugely detrimental to
anyone seeking to get into a leadership position. People with the training were “going to be the ones that
end up taking the positions within the local executive.” Informants believed that advancing equity could
only effectively be done from a seat on the union executive, but it appeared to many of them that “the
same damn bunch” got to the educational events, conferences, and conventions. This conferred an
advantage on them because to get elected to an officer position or to a designated seat, members needed to
be extremely well-informed about how the union functioned, and to be in attendance at convention.

While formally everyone had the same chance to reach high level positions, the actual barrier was
seen to be “the non-inclusion … the lack of encouragement…the [not] going out and seeking people of
visible minority status …” A related issue was that since members did not get much education about
equity issues, they were not predisposed to support equity candidates over others. Informants also felt that
equity group members perhaps needed more encouragement and support to get them to run for office and
to persist until they succeeded. As one informant confided, “it’s terribly frightening, you know, you put
your best foot forward and it gets shot off.” Another mentioned, “Men approach things thinking, well of
course people are going to vote for me, whereas women don’t.”

Communication issues. All the activists felt that communication with members was lacking, even
to the point where one reported having encountered members in the past who had been unaware they were
unionized. As a result members tended to be apathetic, disinterested in union activities, or worse, “they
just don’t even see [the union] as relevant …” Maintaining communication with the membership was not
always attempted. “We had this executive that didn’t tell us anything.” Even if communication was
attempted, it was rarely directed to individual members, in some cases due to cost, in others because
membership lists were not available, or simply because it was not practical. “There’s no way to send out
300 resolutions to every member and say, I’m taking a poll, what do you think?” Communicating
electronically via web-sites and e-mail was difficult because not all members had computers.

Therefore information was sent to local officers and sometimes to stewards, who were supposed to
convey the information to the membership via such means as bulletin boards or face-to-face conversations.
But there was no official way to keep track of whether this occurred, and informants thought “shuffle and

55
file under G” was common. A few of the activists could attest to the fact that stewards did not always
convey information to the membership. Members were unlikely to be aware of when this happened unless
they were plugged into the union’s informal communication network. Even then there could be an
impenetrable layer of secrecy. One well-connected informant mentioned that she had only become aware
that her union was in severe financial trouble after the problem had become acute.

A more specific communication issue as it pertained to equity issues was that informants reported
that equity representatives, themselves included, sometimes tried to soften their message because they did
not want to rock the boat too much. “I think we’re almost afraid to bring up the equity issue in front of all
these guys… it’s like the rolling of the eyes, the ‘oh we’re not going to that are we?’…” At the other
extreme, comments were made about ineffectual presentation of equity ideas by committee members. “…
it’s like they want to get a certain number of resolutions forward and they don’t really think them out, and
they almost open themselves up to ridicule.” Especially strident and emotional communications from
equity representatives were also seen to detract from the importance of their message. “… they’re shrill
and they’re goofy and they’re combative… the second somebody says something they take instant
offense.”

Other problems with union structures. As indicated previously, union structures tended to be
highly bureaucratic and difficult to understand, and membership communication was limited.
Opportunities to get involved in the union tended to be made available only to members who had
displayed an active interest in the union by attending local meetings. Members who could not attend due
to time constraints, meeting times or locations, shift work or the like were at a severe disadvantage. Not
only did they not receive information about local affairs and upcoming union events, but they could be
barred from formal position-holding and denied opportunities to attend training sessions, conferences and
conventions. One informant who had been a shop steward for years but who could never get to meetings
because she worked evenings noted that it had been “raised to me as a reason why I shouldn’t be privy to
going to this conference or that convention.” Uninvolved members were also less likely to know how to
access opportunities for training and knowledge acquisition that could eventually open doors to leadership.

One important avenue to leadership was finding a mentor, which was particularly important
because of the general lack of information-sharing and training available to ordinary members. Mentors
could encourage members to run for office, teach them how to run successfully, get them entrée to
different committees, and familiarize them with executive level planning and decision making procedures
and priorities. Mentors could also help members learn the political ropes, critical to advancing in the
union, such as how to use intimidation or alliance-formation to get their own way, or how to use obscure
rules of order to limit discussion at union meetings or control the floor at conventions.

Activists pointed out that these political tactics could be used to either support or hinder equity
initiatives. The problem was that mentoring opportunities tended to be restricted to those who had
demonstrated a willingness to make the union their top priority, which was not feasible for many women.
Consequently women were less likely to develop the level of political acumen required to run successfully
for union leadership positions. Also, few role models existed of leaders who lived balanced lives, or who
had significant family responsibilities, which therefore acted as a deterrent for many potential female
leaders. “She was the only person at that point that any woman in the group if they were involved in the
union could really look up to, to see that oh, you know, maybe there is a light to look at.”

Another problem with union structures was that the supposedly democratic structures, intended to
give voice to members, in reality afforded leaders a great deal of autonomy. For example, although local
leaders were supposed to send out notices about all equity events, if they chose not to do so there were no
consequences. Similarly, although theoretically the membership could vote leaders out of office for
malfeasance, they had to know and be sufficiently concerned about what their leaders were doing to take

56
action. Since apathy was high, few members came out to local meetings, and membership communication
was limited, it was unlikely that either of those conditions would be met. Representation structures
themselves were faulty for despite the diversity within their locals, most convention delegates were white
males. Although they were supposed to be acting in the interests of the membership, poor meeting
attendance meant that delegates rarely got much guidance regarding what stance to take on equity issues.

Consequently, although union leaders were somewhat constrained by union constitutions that set
out the responsibilities of their offices, and by the need to at least appear to uphold the national union
position in regard to equity issues, on the whole leaders’ personal belief systems could dictate how or
whether to push for equity at convention. Although they tended to be more liberal than their members,
there were exceptions. As one activist reported, “I think a lot of them really believe that women should be
wherever their wives are, which is at home looking after children and grandchildren … making sure their
supper is on the table.”

Some activists nevertheless believed there was enough membership control and enough ethics to
ensure leaders “at least make some effort to do what they know is right.” Others were less sure. One
informant, noting that there was quite a split within her own local leadership in terms of who believed in
equity issues and who did not, felt certain that, ”When it comes to their secret ballots they are not voting
for [equity initiatives.]” Activists like her believed that“… a whole lot of men in power positions have to
retire and be replaced by people from equity seeking groups….” Others noted that their leadership was
rock solid on equity, and suffered for it. “They’ve been criticized for going too far because they’ll stick
with it.”

Discussion

What the data make clear is that significant obstacles to equity still exist within some Canadian
unions. Figure 1 presents a conceptual model of the path leading to progress on equity, incorporating the
comments of the study participants. It indicates that the path to union leadership proceeds from
involvement to knowledge, to local position holding, to convention attendance, and then to union
leadership. Initial involvement leads to the development of preliminary understanding of union operation
and goals. It inspires further involvement in union educational or political activities, and brings about
exposure to potential mentors. Knowledge and experience may result in formal position-holding within the
local, perhaps as a steward, a committee member, or even a local officer, which increases the probability
of attending conventions as a voting delegate.

Convention attendance brings with it the possibility of exposure to higher-level mentors and
appointments to important union committees. Convention delegates also elect union officers, equity seat
incumbents, debate equity resolutions, and establish union priorities, policies and practices. It is very
important, according to respondents, that equity group members should be well-represented at convention
because equity group support increases the likelihood that good incumbents will be elected to equity seats,
equity resolutions will be passed, and equity group members elected to union executive positions. It is at
the executive level that major decisions are made about how to implement the resolutions passed at
convention. That is also where decisions are made regarding resource allocations to equity structures and
to membership education on equity, and on the degree and type of communications members will receive.
These were all important factors identified in this study as affecting progress with respect to equity.

57
 Figure 1

Factors Affecting Progress on Equity

Member
Participation
Structures
Union Resource allocation
Further Local position Convention position Training
education holding attendance holding Communication
Decision making

Knowledge &
exposure to
mentors Progress on
equity

58
 Perhaps the major insight arising from this study is that whether or not equity is achieved can
largely be attributed to the leadership. It was evident that activists believed that equity-conscious leaders
could make a real difference. “It takes a really effective leader in the top position to help push the agenda.
They certainly can control some of the issues and how they’re brought up and brought forward.”
Although they also admitted that membership recalcitrance had some impact, as did societal values,
resource issues, and problems with accountability structures, they believed that leaders could overcome
these stumbling blocks. They could address the first two issues through membership education, and
implement changes to their communication and decision-making processes to give members more
information and more voice. A better informed and empowered membership would then be able to initiate
structural changes via resolutions to convention to ensure leaders acted in the members’ interests. The
resource issues could also be addressed with membership support. So a key question is how union leaders
come to be in a position where they can have such a powerful impact on outcomes, particularly given the
supposedly democratic nature of unions.

In answering this question, it is useful to refer to Bradley’s resource-based power model (1999)
which identifies 9 different power resources that can be drawn upon to give “one sex the capacity to
control the behaviour of the other”(Bradley, 1999: 33), which she refers to as gendered power. Given that
unions and other organizations have been recognized to be gendered (Acker, 1990), this theory has proven
useful in analyzing the power dynamics operating in a number of UK workplaces and unions (Bradley,
1999; Healy and Kirton, 2000) and may be useful for generating insights into how leaders affect
outcomes, including progress toward equity.

The nine power resources Bradley identifies are economic power, positional power, technical
power, physical power, symbolic power, collective power, personal power, sexual power and domestic
power. All are defined in Bradley (1999: 34-35), but several have proven to be particularly useful in
understanding the workings of power within unions and are defined here. Positional power is “power and
authority gained by virtue of holding positions.” Technical power arises from technical expertise, and in a
union context, one example of this would be knowledge of the rules of order governing meetings.
Symbolic power is “the ability to impose one’s own definitions, meanings, values and rules on a situation,
to give one’s own experience primacy” and derives in part from control over communication mechanisms.
Collective power is “the ability to organize groups of people to pursue common goals,” which is
associated with positional power. Personal power arises from knowledge, character, social skills,
experience, political skills, and the like.

It is fairly clear, based on informant comments, that union leaders tend to have these power
resources in more abundance than does the rest of the membership. Informant comments also offer
insights into how this occurs, which are captured in Figure 2. Here explicit recognition is given to how
factors like exposure to educational opportunities, mentoring and local position holding give rise to not
only personal and technical power, but also positional, symbolic and collective power. Possession of these
power resources increases the likelihood of convention attendance and union position holding. What is
particularly captured in Figure 2 is how local and union position holding give rise to Michels’ Iron Law of
Oligarchy (Michels, 1949), which ensures that existing leaders will be unseated only with the greatest of
difficulty because they control the power and material resources of the organization, including the
communication and decision making mechanisms.

59
 Figure 2

The Role of Power in Advancing Equity

Positional/sym-
bolic/collective
Member power
Participation

Structures
Union Resource allocation
Further Local position Convention position Training
education holding attendance holding Communication
Decision making
Knowledge &
Progress on
exposure to
equity
mentors
Personal/tech--
nical/ power

60
 Whether or not union leaders will choose to utilize their power to advance the equity agenda or
hold it back is largely a matter of individual conscience, given the problems with representation
structures, already discussed. Union executives can mobilize collective power to ensure they maintain
their privileged position, for instance by ensuring that equity committees are under-funded, or refusing to
establish adequate numbers of equity seats on the executive committee. As informants noted, they can and
have employed sexual power to intimidate and silence female leaders, preventing them from vigorously
pursuing equity issues. On the other hand, they can use their positional power to ensure that equity issues
become totally integrated into union affairs, and use their symbolic power to frame equity issues in ways
that generate membership support. They can use their positional power to increase membership education
regarding equity issues and to establish better communication structures.

Figures 1 and 2 identify some logical intervention points that would help to advance the equity
agenda. Helping women and equity group members to acquire the personal and technical power that
would lead to union leadership positions that carry with them high levels of positional, symbolic and
collective power, would be the ultimate way to advance the equity agenda. But structural changes and
increased resource allocations to equity initiatives might have to occur before this could be achieved. For
instance, special efforts should be made initially to make sure equity group members can participate in
union affairs. Changing union policies and practices to ensure that child care is provided at all union
events and meetings are held at appropriate times and locations, would be obvious starting points. Seats
should be reserved for equity group members at all educational events, and potential mentors should be
encouraged to keep an eye out for potential leaders within these groups. Equity group members should
also have access to conventions where they should be paired up with experienced activists who can help
them make sense of the proceedings and the issues raised.

To increase the collective power of those equity group members already holding local and union
positions, their constituency groups should be empowered within the union. This could be accomplished,
for instance, by ensuring that all equity committees and all equity seat holders had representation at
convention. Resolutions generated at equity conferences and by equity committees could be automatically
forwarded directly to conventions, with serious discussion of them mandated. Equity group budgets could
be increased, and specialized training provided for equity group leaders and committee members
regarding how to increase and use their own power resources.

Many of these proposed changes would require constitutional amendments, which could only be
passed with high levels of support from voting delegates at union conventions. Since this support might
not be forthcoming without extensive membership and leadership education about why an increased
emphasis on equity was necessary, resources would have to be allocated to that end. If the leadership
made a strong, visible commitment to equity, and the membership developed a greater acceptance of the
need for it, equity group members might become more willing to assume the risks associated with running
for leadership positions. There would also be more support for them if they ran unsuccessfully, which
would possibly encourage them to try again. In the end there would be greater equity group representation
on the main union decision making bodies, and a greater possibility of achieving equity.

After several decades of efforts to make unions more welcoming to non-traditional members,
some progress has been made. I have argued that further progress lies largely with getting more equity
group members into leadership, and have suggested why that is the case. Greater efforts must be made to
ensure proper representation of non-traditional groups on union executives because the democratic deficit
within unions is rapidly assuming crisis proportions. The union renewal literature unequivocally
maintains that if unions are to remain powerful, they must develop more inclusive ways of operating. The
activists interviewed for this study made a number of suggestions in that regard. All that is needed now is
the will and the commitment to rectify the current situation by making equity a priority.

61
 Appendix 1
Interview Guideline
.
Personal Background:
1. Please tell me what union you belong to, how big it is, what its gender breakdown is, and whether
the gender breakdown or the number of non-traditional members has changed in the past 5 years.

2. What is your current position or title in the union, and what other positions or titles have you held
in the past in this or another union.

3. How long have you been involved in efforts to promote equity?

4. To what extent do you feel that equity needs in unions today are different than they were 10 years
ago? Please explain.

5. Which equity initiatives are you, or have you been, particularly involved in i.e. women,
minorities, gays/lesbians?

6. Why is advancing these equity initiatives important to you?

7. To what extent do you feel that equity for non-traditional members has been achieved in your
union?

Union Information:
8. What kinds of equity initiatives, if any, are currently underway in your union, i.e. women,
minorities, gays/lesbians? Have they changed over time? If so, how?

9. What formal structures have been put in place to promote equity within your union? When were
these structures put in place? Please describe them.

10. Are there any informal structures or mechanisms in place? When and how did they emerge?
Please describe them.

11. Have there been any changes to these formal structures or informal mechanisms in the past few
years? If not, go on to the next question. Otherwise, please describe any changes made,
explaining why they were made, who was involved, and how they were introduced.

12. To what extent do you believe that existing equity mechanisms within your union are effective?
What formal or informal structural changes would be desirable? Would it be better to have only
one, rather than multiple, formal equity structures?

13. What do you see in the future for your union in respect to achieving greater equity for non-
traditional members: greater success, less success, little change? Why?

14. What would need to change to ensure that greater equity was achieved? [Prompt with: external
environment, union leadership, existing formal and informal structures, membership interests.]

62
 References

Acker, J., “Hierarchies, Jobs, Bodies: A Theory of Gendered Organizations,” Gender and Society, 4 ,
(1990), 139-58.
Bradley, H, Gender and Power in the Workplace. Analysing the Impact of Economic Change,
Basingstoke: Macmillan, 1999.
Briskin, L. & McDermott, P. (eds.), Women Challenging Unions: Feminism, Democracy and Militancy,
Toronto: University of Toronto Press, 1993.
Cobble, D. (ed.), Women andUnions. Forging a Partnership, Ithaca: ILR Press, 1993.
Fairbrother, P. & Yates, C. (eds), Unions in Renewal. A Comparative Study. London: Continuum, 2003.
Foley, J., Redistributing Union Power to Women: The Experiences of Two Women’s Committees.
Unpublished doctoral dissertation, University of British Columbia, 1995.
Healy, G. & Kirton, G., “Women, Power and Trade Union Government in the UK,” British Journal of
Industrial Relations, 38(3), (2000), 343-360.
Kumar, P. & Murray, G., “Strategic Dilemma: The State of Union Renewal in Canada,” in P. Fairbrother
& C. Yates (eds.), Unions in Renewal. A Comparative Study, London: Continuum, (2003), 200-220.
Michels, R., Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern
Democracies, New York: Free Press, 1949.
Murray, G., “Unions in Canada: Strategic renewal, strategic conundrums,” in P. Fairbrother & G. Griffin
(eds.) Changing Prospects for Trade Unionism. Comparison Between Six Countries. London: Continuum,
(2002), 93-136.
Murray, G., “Union Myths, Enigmas , and Other Tales: Five Challenges for Union Renewal,” Studies in
Political Economy, 74, (fall 2004), 157-169.
Organization for Economic Co-operation and Development, OECD Outlook. Paris: OECD, 2004.
Riddell, C. & Riddell, W. C., “Changing Patterns of Unionization: The North American Experience,
1984-1998.” Dept. of Economics, University of British Columbia, Discussion Paper No. 01-23 (2003).
Waddington, J., “Situating Labour Within the Globalization Debate,” in J. Waddington (ed),
Globalization and Patterns of Labour Resistance, London: Mansell (1999), 1-28.
White, J., Sisters and Solidarity. Women and Unions in Canada, Toronto: Thompson Educational
Publishing, 1993.

63
ASAC 2006 Christopher Fredette (student)
Banff, Alberta Patricia Bradshaw
Schulich School of Business
York University

Sue Inglis
McMaster University

CREATING DIVERSE NON-PROFIT BOARDS:

ENGAGING MULTIPLE DIMENSIONS OF POWER1

This paper presents the results of interviews with key informants who are
engaged with efforts to diversify non-profit boards of directors. The strategies
and approaches they use are reflected in a multi-dimensional model of power.
Implications for on-going efforts to enhance diversity are suggested.

Introduction

The importance of diversity has been recognized as an essential component of organizational

survival, whether from the perspective of performance, creativity or stakeholder satisfaction. To this end,
organizations – especially those operating in the voluntary sector – are starting to take action to enhance
diversity at all levels, including board levels. Researchers are tracking these changes and a growing
number of studies have been conducted examining the composition of boards and the conclusion seems to
be that “the vast majority of trustees are white, more trustees are male than female, and boards draw their
members disproportionately from members of the upper and middle classes (Ostrower and Stone, 2005).
While boards are becoming more diverse it is at an uneven pace (Abzug, 1996; Abrug and Galaskiewicz,
2001; Kang and Cnaan, 1995; Ostrower, 2002).

The advantages of diversity on nonprofit boards are increasingly well articulated in the literature
(Fletcher, 1999; Berthoud & Greene, 2001; Daley & Angulo, 1994; Daley & Marsiglia, 2001; Fink,
Pastore & Riemer, 2001; Mukherjee, 2001; Miller, 1999). Factors such as being more responsive to the
community and clients, better able to respond to priorities of funders and the improved ability to make
creative decisions have all been identified as driving factors for change. While increasingly boards
understand the value of diversity, in reality, they continue to struggle with how to become more reflective
of their communities. A number of more practitioner focused articles have been written to help boards
move in the direction of more diversity (Berthroud & Green, 2001; Buchanan & O’Neill, 2001; Devine,
2005; Mukherjee, 2001).

Missing from most discussions of diversity in the boardroom is an in-depth and comprehensive
understanding of the critical role of power in moving the diversity agenda ahead. This paper draws on
interviews with key informants from the nonprofit sector to explore the role of multiple dimensions of
power in creating more diverse boards. Examinations of power in the boardroom to date have tended to
examine the way power is manifested in the relationship between the board and the staff (e.g. Golensky,
1993; Kramer, 1965; Murray, Bradshaw and Wolpin, 1992). This paper will extend this focus on power to
the topic of creating diverse boards by drawing on a model described by Bradshaw (1998) and Bradshaw
and Boonstra (2004). The paper explores the multifaceted nature of power within the context of diversity
as experienced and understood by community leaders currently serving in the non-profit sector. Through

1 We gratefully acknowledge the support of SSHRC Grant in the Non-profit Sector in Canada Joint Initiatives

64
the use of qualitative research, relying specifically upon in-depth interview techniques, this paper
illustrates both the barriers faced by leaders as well as the change approaches afforded to them through
the use of power based strategies in the context of their roles on boards of directors. In undertaking this
work we aim to contribute to the current state of understanding regarding the importance of the personal
and structural power in board diversity efforts as well as the impact of surface and deep facets of power
on organizational diversity efforts.

Research Methodology

Semi-structured, in-depth, exploratory interviews were conducted with eighteen community

leaders serving on non-profit boards of directors in two large, culturally diverse cities in Ontario, Canada.
Using a structured interview protocol, based on open-ended or ‘grand-tour’ questions combined with
more exploratory questions we dialogued with the informants about both their personal experiences on
nonprofit boards and their views of the more organizationally focussed approached to creating diverse and
inclusive boards. The initial set of informants was selected using a purposeful sampling approach (Patton,
1990). This approach was supplemented with a snow-ball sampling technique in which informants were
asked to recommend other potential interviewees to complete the sample. Informants included those with
leadership experience with boards of directors in the non-profit sector as well as being members of under
represented groups (e.g. the sample included seven youth (ages 18- 29), 9 females, several members of
the LGBT communities and many visible and ethno-racial minorities including people from Asia, Eastern
and Western Europe, Sri Lanka, Africa, Iran and Indigenous backgrounds). In all cases, informed consent
was obtained from research participants in advance of any data collection. The interviews ranged in
length from forty-five minutes to two hours and were electronically recorded. Verbatim transcriptions of
each interview allowed for systematic thematic analysis (Kvale, 1996). The collection and analysis of
research data for this paper provided a structured multiple informant design which accentuate the central
tenets of verification and generalization of a previously established theoretical model (Eisenhardt, 1989).
In contrast to traditional methods for the development of grounded theory, the implementation of a
multiple case format of evaluation facilitated the integration of conceptual frameworks from previous
research rather than requiring strict reliance on data generated grounded theory (Eisenhardt, 1989). In this
regard, Yin (2003) in agreement with Eisenhardt (1989), emphasized the importance of balancing a pre-
established conceptual framework or set of theoretical assumptions, while simultaneously remaining
receptive to new directions resulting from investigation. Our multiple interviews offered the potential to
explore the phenomenon of diversity in terms of an established theoretical model on power, while also
capturing the insight and understanding of confidential respondents in an innovation manner.

Revisiting the Power Model Through the Lens of Diversity

The Transformational Model of Power, first introduced by Bradshaw (1998), proposes the
conceptualization of power in four distinct ways, based on an awareness of two competing polarities or
tensions. Power in this model can be conceptualized as manifesting itself at either an overt or a latent
level of consciousness and held by individual actors as well as within the systems and edifices of the
organization. Pursuing a multi-dimensional approach to the conceptualization of power results in a
dynamic and interactive model based on four unique elements: Surface-Personal Power; Surface-
Structural Power; Deep-Personal Power; and Deep-Structural Power. In the next sections of the paper we
will explore each perspective on power, the underlying power sources of each dimension and the
strategies that boards are using to respond to organizational challenges and shortcomings related to
increasing diversity. In each case we provide illustrative quotes to illustrate how the power sources are
being used to enhance diversity on the board.

65
Surface-Personal Power

This dimension of individual power emphasizes the knowable, manageable and transferable
aspects of influence as a resource to be accumulated and maintained. In early conceptual writing, power
was characterized as a measure of the ability of an actor to exert influence, such that person A has more
power than B to the extent that A can get B to do something they would not otherwise choose to do (Dahl,
1975; Emerson, 1962; Wrong, 1965). As early as 1959, French and Raven created a typology of power
which includes five sources of personal power all of which are both manifest and identifiable. From this
perspective, power derives from expertise, legitimate authority, referent power, rewards and coercion and
association with other powerful people (Kotter, 1985; Mechanic, 1962). From our perspective, this
traditional view of power exemplifies a goal-oriented approach best characterized by personal action,
where power rests with the behaviours, skills, possessions and qualities of an active-knowing agent.
Consistent with this reasoning, much of the popular literature has focused on potential strategies for
enhancing personal power and success. These strategies are seen as knowable and relatively easy to teach,
thus individuals even at the lower end of the organizational hierarchy can work to increase their personal
power through education, improved interpersonal skills or network to develop associations with powerful
others (Mechanic, 1962).

Although this view has conventionally focused on the actions and capability-building of the
individual, many non-profit boards are actively undertaking initiatives which serve to instil and enhance
the bases of power held by members from traditionally disadvantaged groups. Some of these are
described in the next sections. (See Table 1).

Building Expert Power. Stemming from the assumption that an important source of personal
power originates from the specific acquisition of knowledge and expertise by the individual, many board
initiatives have centred on improving the knowledge base of new and traditionally marginalized board
members. Strategies adopted by the boards of directors in this respect addressed organization-driven
training and education intended to familiarize and orient members with organizational practices and
activities. Strategic activities undertaken at the board level included explicit commitment to: targeted
recruitment initiatives focused at incorporating diverse members with technical and personal expertise;
socialization practices in which new members undergo organization orientation; and, providing technical
and process training for new board members. Similarly, respondents strongly emphasized the benefits of
simplifying board practices through the use of plain language and flexible debate procedures which
facilitate newcomer input during the orientation period.

In addition, a number of participants introduced the idea of knowledge-transfer initiatives aimed

at improving the relevance of participant’s previous experiences and mitigating the expertise gap by
simplifying the ways in which the board conducts itself. The focus on integrating previous experience was
strongly communicated by in two ways: first, the emphasis on recruiting members with underlying
expertise, “we tried to think together of women that we were working with in the community who were
from you know more marginalized communities or traditionally marginalized communities, and decided
to actually target them”; and second, in valuing individuals for their personal lived-expertise, “this
collective said, 'no, we actually value those experiences and components of your identity. We think they're
important and we think that you have unique perspective and we'd like you to come and share that
perspective with us’.”

Enhancing Associative Power. Recognizing that many of the activities undertaken within the
board context build on associative power as the basis of successful abilities to communicate, make
decisions and politic, many boards have endeavoured to encourage stronger associative bonds between
board members and the community. Focusing specifically on improving the relational ties between
dominant members or powerful coalitions and organizational newcomers as well as members of

66
traditionally marginalized groups, boards have adopted strategies consistent with the central tenets of
association and mentoring, thereby managing the institutionalization of networking activities. Although
simplistic sounding, the benefits derived from mentoring relationships can be incredibly powerful as
noted by one informant, “it's like cultural translation for any new member, but particularly for new
members who either don't have a lot of experience or don't you know if there's sort of a dominant culture
at the board.” In addition, the benefits of mentoring relationships extend beyond simply providing
guidance in times of uncertainty; these social connections form the foundation for developing the
information networks necessary to gain access to powerful others. Comprehending the potential benefits
of encouraging relationships between powerful board members and new or marginalized members, one
informant recounted undertaking a particularly insightful approach to enhancing access to those board
members she believed to be among the strongest and most powerful, “Well recently what we started to do
is have a sort of a mentorship happening, where we bugged older senior board members, and I targeted
basically the women who are very strong. Women of colour who are active in the community and active
on the board, to beg them to put something else on their plate which is to mentor the younger and less
experienced board members.” Reflecting on this statement, this is perhaps one of the best
characterizations of power as a manageable phenomenon in which personal power may be redistributed of
through a series of critical associative linkages. The importance of building bonds through association is
more obviously apparent when it is absent, as another informant noted with respect to his organization,
“women have been asked to come on the board and have not been provided with the kind of supports to
ensure their real value added inclusion in the decision making, planning you know those kinds of things.”

Increasing Legitimate Authority. In the power literature, legitimate authority has been
attributable to the position power, or that which results from the position held by an individual relative to
others in the organizational hierarchy. From the perspective of non-profit boards of directors, legitimate
authority speaks to the roles, positions and committees memberships which consistently reflect leadership
and decision-making power among board members. In this regard, thematic analysis revealed two
dominant strategic approaches for reallocating personal power among diverse board members, especially
those of visible minorities and marginalized communities. First, respondents described the importance of
the transferring the role of board chairperson from powerful traditional group members to members of
underrepresented groups. Although potentially controversial, this strategy provides less powerful
members with the opportunity to exercise voice and shape the discourse while limiting potential
repercussions as illustrated, “I chaired a committee for almost ten years. Worked with the government
everyday and I know that I antagonized some people because they made some comments, they backed off
eventually.”

The second strategy related to legitimate authority identified by respondents centred on the need
for inclusion and representation within the decision-making hub. Whether through the creation of a seat at
the boardroom table or through the assumption of a leadership role in advisory committees, informants
frequently stressed the need to ensure and encourage access to venues in which members could exercise
influence when expressing their views. Consistent with the theme of legitimate authority, one informant
discussed the degree to which winning an elected position provided legitimacy for issues and concerns
that were central to her, “I think that in some ways, staff felt that I could be a good representative, that I
had a good understanding of the issues and the staff and the relationships, and so I ran for election and
was nominated and elected and was made the vice-chair of the chapter in Toronto that year.” One
prominent way to decrease a traditional power differential is to increase the degree to which the roles that
diverse and marginalized members play in the board are legitimate and also visible, central and critical.
By improving the visibility, centrality and criticality of previously marginalized individuals, organizations
are able to provide the opportunity to demonstrate their relevance within the context of the boardroom.

67
Surface-Structural Power

Whereas the Surface-Personal perspective emphasizes power as manifest – knowable,

manageable, and transferable - within and among individuals, the Surface-Structural dimension assumes
that power is visible but resides instead within the structure and configuration of an organization itself. As
a result, power may be seen as belonging to any dominant collective operating within a defined structure
rather than stemming from individuals based on their traits, skills or behaviours; thus structural factors act
as the overriding influence in understanding power relations. Implicit in this view of power is that a
sophisticated and aware subunit or interest group can learn to become more politically astute and enhance
their interests by increasing their ability to exert influence over sources of power (Pfeffer, 1994).

At first blush, organizational structure may appear rigid, acting as an impediment to individual
action and agency, however it is perhaps more constructive to consider the underlying dynamics of intra-
group negotiation upon which many of the visible structural dimensions have been anchored and founded.
From this perspective, by consciously and pragmatically manoeuvring into the best position within the
structure, traditionally marginalized groups and coalitions may be able to leverage and thereby reorient
structural power to better suit their collective interests. Clearly, by altering the structural parameters of a
given context, the previously dominant power relations can be surfaced, addressed and reconfigured. To
this end, Kanter (1977) has argued that if the numbers of women or other minorities are increased in
managerial positions the organizational dynamics affecting them can also be changed. Similarly, others
have stressed the importance of coalitions, arguing that formal stratification of the organization constrains
and facilitates the creation and collapse of coalitions between interest groups (Bacharach & Lawler,
1980). Further, some suggest that the creation of structured power games allows organizational members
deal to with the dynamics of uncertainty (Crozier & Friedberg, 1980).

Boards of directors, especially those active in the nonprofit sector, have acknowledged the need
to strike a balance with respect to power differentials between groups in traditionally central positions and
those that have historically been marginalized. The following sections present the thematic analysis of the
interviews grouped into board initiated strategies which support the dimension of a Surface Structural
power, or power that rests in organizational structures which are both identifiable and manifest.(See Table
2)

Increasing Access and Control of Scarce Resources. The control of scarce resources, including
the control of uncertainty, has frequently been considered a source of power and case studies of
organizations reveal many resources and uncertainties which can be managed to enhance power (Crozier,
1964; Hickson, Hinings, Lee, Schneck & Pennings, 1971; Salancik & Pfeffer, 1974). Scarcity, regardless
of the specific commodity – money, information, talent or similar resource – is both an extremely
common and an extremely powerful motivator in organizational contexts, as it suggests that only a limited
number of needs will be satisfied. As a result, those in possession of the ability to limit or control the
access of others are able to derive immense power from their relative standing. This phenomenon is very
prominent in the non-profit sector and many respondents reflected on past events in which information
was withheld during decision-making activities or the pet-projects of dominant coalitions were advanced
through the allocation of critical resources, to the detriment of other priorities. As a result, much of the
discussion offered by informants centred upon the notions of encouraging accountability and monitoring
of activities and the reform of existing board-level practices. In recognition of the critical nature of
relationship between resources and change, with respect to the impact of scarcity one informant
suggested, “The boards will wake up if the funders ask for it. … I mean look at the university, the only
way this university even dealt with pay equity and employment equity was because we are funded by the
federal government.” From this perspective, strategies adopted by the boards of directors focused on the
need to mitigate the power derived from scarcity in two distinct ways. The first strategy centres on
facilitating a climate which systematically encourages greater access to information for members of

68
diverse and marginalized constituencies. As one informant noted, “It's working because there's been a
systematic effort to create a culture on the board, to promote safety, to promote conditions under which
people can participate fully. To promote conditions under which people can share diverse experiences
and ideas and be heard. ”

The second prominent strategy that emerged from discussions with informants focused on
mandating a representative composition for positions which exercise decision making power including
both committees and advisory groups. Paradoxically, decision-making power often extends beyond the
grasp of the voting board members in that the alternatives from which decision-makers often choose are
reflective of the opinions and choices of appointed advisory committees and other groups. From this
perspective, boards are recognizing and attempting to address the need to have members of marginalized
groups present throughout the organization and board’s decision-making structure. As one informant
noted, “we've had people on the board, I've met who could tell you very little of what was happening
because they come just to the meetings, they haven't participated in any standing committees and there is
no rule that says that they have to. So, they come and they rubber stamp.”

Assuring Flexibility. Organizational rigidity commonly referred to as bureaucracy within

practitioner communities, has traditionally been associated with formal organizational structures wherein
reliance upon rules and routines serve to ensure stability, thus inhibiting change and adaptation. In many
instances, this form of structural constraint has been leveraged against individuals as a means through
which dominant groups can maintain the appearance of a stable and legitimate base of control, while
avoiding direct challenges to authority. In contrast, many innovative boards of directors have recognized
the enormous benefit of addressing structural dimensions of the organization which have traditionally
limited the influence of marginalized individuals and groups, thus altering and in some cases removing
unnecessary formality to create more open fluid dialogue and systemic flexibility. To this end, two
prominent sets of strategies emerged during our thematic analysis, including: the need to address the
impact of formality and formalized structure on flexibility; and, the ability to fluidly alter board structure
or enact and disband advisory committees to enhance a board’s ability to address complex or emerging
challenges.

The first strategy centred on the need for boards to encourage structural flexibility through an
active process of systematically reviewing and challenging the existing policies, bylaws and constitution
of the organization. Recalling one such board meeting, one informant reflected on her board’s resolution
to remove structural barriers and increase flexibility, “we talked about that actually and we have made a
commitment to minimize our bylaws.” This board’s ability to identify and address the ways in which its
rules and routines were impeding its members’ ability to contribute to their full potential is consistent
with the systematic approaches introduced by other respondents. Consistent with the notion of adapting
structure to suit an organization’s diversity initiatives, one informant discussed her board’s desire to enact
new policy to reach diversity-related objectives, “The board commits to a diversity plan and builds in
different policies and give direction to the paid staff around policies related to diversity… I think that that
is absolutely essential. One of the things that I pushed for was having policies and having ways of
indicating that we had progress...”

The second strategy that emerged from our conversations with active board members focused on
the board’s ability to flexibly approach difficult issues by altering its own structure or through the fluid
use of provisional advisory committees. With respect to the alteration of board structure, one respondent
discussed how his board was able to improve flexibility stating, “Like if we just lay out the three, you
know the basic responsibilities of a board you can be very creative with how you structure your board in
order to ensure that those responsibilities and sort of legal responsibilities, organizational and legal
responsibilities are covered. You can structure a board to cover all of that and do whatever else you want
to do, you can be very creative.” This perspective suggests changing board structure as one possible way

69
of unlocking flexibility in the boardroom. Consistent with the underlying rationale behind changing board
structure, some respondents discussed the benefits derived from the use of temporary advisory
committees. In this regard, one informant shared how her board had instituted a provisional advisory
committee, composed of previously marginalized members, with a specific hiring focus, “An advisory
committee that may just want to do HR stuff…So if this committee of youth just wants to work on that area
that's a great way, that might be a great way to, to affect diversity at the board level you, that's a little bit
more creative than just putting people in chairs of the board. So allowing a kind of board structure that's
a little bit more creative and a little bit more interesting so that you can facilitate your goals, you know
and see your motives through.”

Demonstrating Relevance. Strategic contingencies theory (Hinings, Hickson, Pennings &

Schneck, 1974; Perrow, 1970) suggests that those who control tasks which are critical and of relevance to
the strategic direction of the organization will also have more power. Not surprisingly, a number of
prominent board driven initiatives emerged from our thematic analysis which centred on the notion of
building a business case for diversity. From this perspective, two distinct yet related strategies dominated
much of the discourse offered by our respondents. The first strategy focused on the need to establish a
socialized culture of understanding to facilitate a business-like climate, focusing members on importance
of similarity and shared interest. Many of the most successful organizations, regardless of profit
orientation, have undertaken efforts to establish consistent internal cultures which serve to unite
stakeholders with a salient dominant mindset. In this regard, the board members involved in this study
recognized the need to establish strong social norms consistent with their organizations’ missions. As one
informant succinctly testified, “that's part of building the organizational culture, because then it becomes
very clear to new people coming in that this is how things work here, and you know is there a goodness of
fit. It becomes very clear if it isn't, and people self select out or you select them out. So yeah, but that's
part of all the mechanisms you build in at every step throughout the organization to address that.”
Similarly, others elaborated on the guiding role that culture takes affords when facing challenges to the
relevance of diversity-related initiatives, comparing their current board experience to that of past
situations in the absence of strong culture “The board has created a culture that values diversity when
that happens, and is vigilant to that and is very cognizant when that process is falling down to bring it
back on track. I've certainly been on boards where that has not occurred and it's not a good experience.”

In contrast to the strategy of institutionalizing cultural openness, an alternate strategy offered by

respondents stressed the need to approach diversity issues with the same methodology used to tackle other
complex business decisions. The second strategy to arise from an analysis centred on making the business
case using a two-phase approach: first, implementing demographic audits to identify diversity gaps,
making them visible; and second, undertaking specific actions to address the critical gaps where inclusion
of members from marginalized communities could benefit from improved representation. The underlying
impetus behind demographic auditing stems from principles of employment equity legislation, which
mandated organizations to focus on internal composition. The mindset pervasive in this type of approach
is one of establishing representative balance as proposed by one respondent, “Gender diversity was very
con - at least starting in the 80s for - it was very consciously planned to make sure to maintain a
balance.” and in undertaking ratio or cost-benefit analysis, “And ah, one of the comments that ah, from
the person that they'd sent down to us, to ah, talk to us about it was, was that you know, we had overall a
good ratio of different ethnic backgrounds.” provided by another. With respect to the methodology
employed to perform a demographic audit, a great deal of consistency emerged from the participant
transcripts. In the vast majority of cases, respondents elaborated upon a process in which their board
agreed to identify demographic gaps, undertake a plan to correct the under-representation and establish
some form of monitoring procedure for the future. While the exemplar offered in table 2 provides an
excellent example of the processes embarked on at the board level, another informant offered the
following, “if they do an environmental scan within that service, find out the gaps and way and then look
at ways in which they could fill those gaps, they could fill those needs, that's the place to start. And then, I

70
think that they have to set up rigorous goals and, and also indicators that will indicate whether they
achieved those goals or not. I think that just to talk about it, doesn't make any points without the answers,
because I think you've got to monitor it because the situations we get into, or could get into is that ah
situation we have at the hospital.”

Deep-Personal Power

The Deep-Personal perspective on power has been largely ignored in the field of organizational
behaviour, having its roots in the psychoanalytic, post-modern and feminist theories. This power
dimension differs from the other areas in several critical ways, beginning with the assumption that power
is inherently diffuse amongst individuals located throughout a social system. This diffusion allows
individuals to become active agents who can deploy their power even if they are at the bottom of the
hierarchy or relatively powerless (Whittington, 1992; Winter, 1996). Implicit in this approach is the
recognition that power relations are often latent or even unconscious and are thereby embedded in the
actual psyche of the individual (Bolen, 1992; Brown, 1994; Starhawk, 1990). From this perspective,
members of oppressed groups often fail to understand how they collude in maintaining the very systems
which oppress them (Burrell & Morgan, 1979).

In the face of latent control mechanisms there are a number of sources of power which boards of
directors can mobilize on behalf of the individual. McIntosh (1989) defines privilege as a knapsack of
taken-for-granted and unearned assets which provide special provisions which help certain individuals
advance. Often, unconscious over-privilege can be found in members of the upper class, heterosexual,
male, white and/or able bodied communities. Knowledge of the advantages which result from these
privileges is usually denied by those who gain at the disadvantage of others. Authentic existence requires
those with power and privilege to make this explicit and then act on the consequences of this unearned
advantage. Boards that enact provisions to facilitate the naming of inequality and attempt to generate
authentic opportunities to act in congruence with an individual’s own values and beliefs, undertake a
process to identify powerlessness, alienation and complicity resultant from the dominant systems. In
addition, facilitating the exercise of individual self-knowledge dictates that board members recognize and
thereby acknowledge their assumed personal power and privilege. (See Table 3)

Acknowledging Privilege and Tokenism. Within the frame of privilege, it is perhaps most
difficult to recognize one’s own sources of privilege than to acknowledge the privilege afforded others. In
this regard, boards of directors demonstrated resolve to address issues of latent privilege by focusing on
two distinct strategies: first, undertaking discussions to challenge individual self-awareness and systemic
sources of privilege; and second, systematically recognizing and labelling tokenism and apathy within the
scope of board activities.

In one participant’s description of the pervasive nature of privilege, she addressed two crucial
issues regarding the inequalities derived from implicit sources of power. First, relating to her self-
awareness of economic privilege, our informant noted, “…these kinds of issues can happen even when
you're not looking at diversity, big D, that you know it could be six white middle class women sitting on
the board together, and there might be difficulties in terms of you know saying or doing things that would
cause other people pain without thinking.” In addition, the further dynamic of diversity only serves to
cloud the implications of shared realities, “The politics of interpersonal relations period, and then where
you have the additional flavour of not recognizing a shared base of experience. It can always happen, it
happens a lot.” Moving the discussion beyond a one-dimensional framework based on colour or gender,
one key respondent demonstrated an enriched awareness of the multiple bases of privilege by sharing his
insights relating to the experiences of others, “Black women have found a lot of stress. Because one, they
are facing two forms of discrimination. Being a woman and being Black. So, it'd be interesting, I'd kind of
separate that.”

71
 Consistent with the approaches used to challenge sources of privilege; non-profit boards of
directors face similar battles when confronting tokenism and apathy within the scope of their activities.
Reflecting on a recent board meeting, an informant recounted the challenge she posed to others regarding
the superficial claims of inclusiveness being offered by apathetic others “…they were serving white
women. And I remember saying to the ED in the organization, why don't you just say that that's the kind
of organization you are. Like don't pretend to be something that you're not.” Alternatively, others
illustrated ways in which their boards were challenging the conventional understanding of diversity
within the board itself, “The people they select to have the gender balance, really don't balance the
gender…. most often there's very little difference between the female corporate president and the male
corporate president…The women who become, say, head of a corporation, does not change the culture of
the corporation to become women-friendly. The culture of the corporation is still male - do, do you
address the issues of childbirth? You know, all the - even the working conditions - no, nothing, nothing
has been addressed. They, they only replace a man, you, you dress differently. You talk differently but
everything else is the same.” Facilitating this type of challenge to the dominant meaning systems is
indicative of how boards are challenging convention and recognizing opportunities for improvement in
services currently offered and the ways in which they represent and include the interests of those from
traditionally marginalized communities. As one informant succinctly noted “I guess in the context of
violence against the women and the work that we do, the diversity and coming from a point of view where
diversity is important is around issues of representation and inclusion. So it's really important in this
agency that we have a diverse representation of women at all levels, and from my perspective the
challenge is how to do that that's not in a tokenistic way.”

Increasing Self-Knowledge. Dysfunctional boardrooms fail to recognize the important role that
boards of directors plays in expanding individuals’ knowledge of self. From this perspective, the failure to
recognize the importance of facilitating an individual’s acts of defiance and resistance to current
organizational dogma or the ability for individuals to challenge the current dominant logic hampers
organizational efforts to remain flexible and adaptive. Attempts to increase board member self-knowledge
have been undertaken by focusing on two distinctive strategies, both aimed at smoothing the process of
individual expression. One strategy focuses on the creating ideological space for individuals to resist
pressures of assimilation into the mainstream doctrine. The underlying notion of personal resistance
encapsulates an interesting paradox as an individual must not only be a part of the board, he or she must
also remain able to exclude themselves from its dogma. An ideal illustration of the paradoxical
relationship between membership and resistance is provided by one informant discussing the challenges
faced by another member of his organization, “I think the organization generally wants to become a more
open and responsive to change, but this trans-sexual activist is going to have a really hard time being a
board member, because his role has traditionally been to challenge, to confront, to illuminate, to protest,
and that does not make a great board member. In terms of because really for a board member you need
somebody who’s willing to compromise and accommodate, and push the board members to be patient and
be aware of their power.” One half of this paradox – resisting the conventional doctrine – is illustrated by
the member’s willingness to confront and challenge the status quo, which appears to stand in
contradiction to being a great board member. While the organization wants to be responsive to change, it
struggles to include the some of the voices that exercise the greatest personal agency.

Building on the theme of self-knowledge, the second board-level strategy to emerge from our
discussion with non-profit board members is that of accepting acts of individual challenge and defiance.
Recognizing the strength to challenge and defy the status quo in a fellow board member, one informant
revelled when recounting, “I think it's just incredible how well things work in this particular organization
and this particular board. And we've been lucky enough to have some extremely powerful women of
colour who have sort of I don't know set a tone around, not a tone, they've laid the ground work I guess
that people are challenged in their racist assumptions. Like I think that's what happens right when you
have a disability and you challenge the stereotypes and whatever.”

72
 From our interviews with active board members it appears clear that participants expect to face
mainstream opposition to their unconventional thinking, however, it appears equally evident that
members seek the opportunity to be heard. The risks associated with challenging conventional wisdom
appear somewhat less salient than would otherwise be expected, as outlined by one respondent discussing
her experience with other board members, “there's going to be a certain amount of resistance or to, to
down-grade, but you need to keep on hammering them with, ah, you know, the, the research if you do it
with the kinds of things like the statistically or even to bring out and say, well, these are the kinds of -
there is a ...” Similarly, the ability to constructively contribute new insight from within the framework of
the board of directors is appears to be worth the risk as captured by a younger female respondent, “she's
been on the board before and the only piece of advice that she gave me was, ‘Even if you think what
you're going to say is stupid and they're no going to go for it, say it anyway cause that's what they need.’
And so essentially I've been doing that and any opportunity I have to say ‘Well why don't we try this
instead, it's new, it's fresh, it's action.’ I'll say it even though I know most people will go ‘Oh no don't do
that’.” Finding ways to accommodate these diverse voices in the boardroom is both difficult and often
divisive, yet is a critical factor in generating creative insight and encouraging organizational growth.

Confronting Disciplinary Micro-Practices. Whereas focusing on improving self-knowledge

centred on the board’s ability to create room for marginalized members to challenge and exercise
resistance, confronting disciplinary micro-practices focuses instead on strategies which sanction the
individual board members’ ability to challenge dominant meaning systems and push the resistance of
others confronting issues of diversity. Perhaps the most innovative – and seemingly least threatening –
strategy to emerge from our thematic analysis centred on the instrumental use of humour. As evident in
the following passage, many of our respondents spoke frankly about using humour to engage others in
difficult dialogue, “I, I find that, that, you know humorous, in a humorous way, you can really hit very
hard so that you can, you know, the people can really really get it.” Similarly, another respondent spoke
of using humour to launch a serious challenge during a tenuous discussion of diversity and
marginalization, “And so I just said to them, I said, ‘is there anyway it would be my colour, or is it
because I am too outspoken?’ I don't bow down to people just because they think I should. I guess it's a
little of both.”

In contrast to the humour-based strategies used as an instrument for change, boards of directors
are endeavouring to implement a second strategy to address diversity in terms of deep personal power
which stresses each individual member’s capability to confront dominant meaning systems and promote
resistance of others. One informant relayed her struggle with marginalization and the challenges posed by
the board’s reinforcing systems of power, “I haven't talked about diversity per say. I have struggled and
when I related it to power issues, I, ahh, dominance and oppression, that become more easier for me to
focus and ask questions because I think that the end of the structure, regarding the power, and the
reinforcement of that power. Unless we can break that continual reinforcement, we will never be
successful.”

Deep-Structural Power

Through a process of careful reflection, we come to understand an organization’s meaning system

as representative of the latent or deep cultural nature of power relations and the source of its endurance.
Applying tenets of the radical humanist paradigm (Burrell & Morgan, 1979), we come to understand how,
cultural artefacts, language, rituals and values construct meaning for organizational members and how
they simultaneously work to suppress conflict, prevent issues from being identified and control the actual
agenda for decision making and non-decision making. In these ways power relations becomes entrenched
in the organization and those who can set the agenda, manage the meaning systems and have others
believe their definitions of reality have more power than those who do not (Alvesson, 1992; Hardy, 1985;
Murray & Bradshaw-Camball, 1993; Rosen, 1985; Smircich & Morgan, 1982;). Thus power is increased

73
to the extent that the group which defines reality has others accepting their definition in unquestioning
and taken-for-granted ways.

From the work of more critical theorists we can further understand the deeper aspects of power by
reflecting on how groups come to consent to their own domination and subjugation in a passive mode
characterized by lack of resistance. Lukes (1974) argues that people accept the status quo and their role in
it because they view the current systems as natural and unchangeable. The role of ideological hegemony
(Clegg, 1990) is important to understand as we see that existing organizational and societal structures are
supported by inherently classed, gendered and raced assumptions and values (Mills, 1992). The social
construction of normal then becomes objectified and is seen as being a given which is unchangeable.
Other ways of constructing meaning are suppressed and silenced with resulting inequities of power. We
see the interests of the dominant group as “objective” and as legitimate knowledge claims instead of
seeing all interests as subjectively created realities embedded in multiple truth claims.

The Surface-Structural perspective looks at power from a more positivistic perspective, with
power originating in structures and objective resources, in contrast, a deep-cultural view of power focuses
on the interpretive and post-modern perspectives for dominant lenses and underlying assumptions. Power
is a cultural artefact which becomes entrenched in the hands of certain dominant and privileged groups.
This dynamic exists to the extent that the meaning systems in which the relations of power are embedded
are shared collectively by various interest groups and are reproduced through discourses, practices and
routines within organizations (Townley, 1994). Undercurrents of critical deconstructionism were
pervasive throughout much of the discourse with respondents, many echoing the need to reach beyond
superficial change and resolve the frustration and alienation felt by the marginalized. (See Table 4)

Emancipating the Agenda. The notion of emancipating the agenda focuses critically on
improving the board’s ability to attack latent structural barriers and leap beyond taken-for-granted
conventional obstacles. In this regard, respondents frequently spoke in terms of advancing a social justice
perspective rather than traditional notions of representative diversity. One respondent spoke to the need to
expand the definition of inclusion to integrate mission and diversity in ways that contribute to the
organization’s social focus, transferring power to otherwise marginalized individuals and groups, “We've
not tried to get you know an abused women on the board. I think a responsible way of doing something
like that would be to have something like what the women abuse council does and the counselling
cooperative do, where they have an accountability committee, where everybody knows that these are the
women who are consumers of the service or have been consumers of the service. Have been abused, they
are identified, and they have power within the structure or outside the structure to influence the board.”

In addition to the approaches already discussed, other respondents spoke specifically of the need
for definitive formal redress to tackle the inseparable relationship between power and diversity. To this
end, commentary focused on consciously enacting change through the board’s own policy and process,
placing items on the agenda “so right from the very beginning the issue of diversity and intersection and
power is right there, right on the table usually right at the beginning. I mean in our five year strategic
direction I think, I think that improving gender and increasing gender, female involvement, increasing
ethno-cultural involvement in the organization is like the top two in terms of our five year strategic
direction plan.” or putting forward motions to affect change “Having the opportunity to participate in the
debate um and discussion of motions in my mind, it's almost as effective as the voting usually is. For us,
most of our voting is consistent by the time we vote, everyone votes in the same way anyway.” In other
discussions, dialogue provided by informants made it equally evident that agenda could be provide the
power to avoid future traps as well as escape current ones, “And all of them had a commitment to
diversity, but some of them had been more effective on creating a diverse and inclusive environment. And
a lot of them had philosophical arguments against policies that entrench you know, negatively used
quotas you know, or policies that entrench positions.” Despite variance in the multitude of tactics offered

74
by respondents, all share an underlying motivation to change from within; to operate on the agenda from
within the organization and address the misalignments between the claimed ideology and the multiple
realities that result.

Deconstructing Meaning Systems. In many cases, opportunities for change are less available
within the parameters of current board activities; agendas for change are lost in the dominant
superstructure or suppressed, falling instead on deaf ears. Increasingly, board members and collectively
boards, recognize the need to truly question the beliefs and assumptions upon which they base their world
views. This is perhaps nowhere more pressing than in the non-profit sector as organizations struggle to
better satisfy each of the organization’s critical stakeholders. To this end, respondents discussed highly
critical strategies in which their boards undertook debates about the nature, meaning and composition of
diversity, asking what diversity means to each member, to the organization and to the society it serves.
Challenging the meaning of diversity creates expansive space for inclusive thinking as it attacks the very
core of categorical division, an intuitively central precursor to marginalization. The message voiced by
one informant, “Like we are all of every other community and we often have multiple identities that we're
balancing. So when we define ourselves as a community for purposes like ours, it immediately makes you,
you immediately begin to deconstruct that. Like what community? Whose community? Whose we?” is
highly indicative of the dialogue offered by others, which deconstructs the conventional notions of static
diversity. Building on the theme of deconstruction, another respondent discussed challenges she had made
to the very understanding that others had of the board’s role and purpose in the organization, “And the
board is supposed to be, they're supposed to be the stewards, the leaders, the visionaries of the
organization and how can they can be care taking their organization; whether it's the constituents they
serve but also their staff, their volunteers their board members. How can they be creating an environment
that is comfortable if these issues are not on their radar?” These critical attempts to deconstruct dominant
assumptions and meaning structures which underpin the board’s foundation challenge the status quo in
ways not yet discussed. They demonstrate focus of purpose and an inherent desire to surface latent facets
of the organization that impede its path to excellence.

Challenging the Established Definitions of “Normal”. Unlike many of the previous challenges
to conventional meaning systems or dominant ideologies, boards of directors that actively encourage
members to confront normalized and taken-for-granted governing routines through which assumptions
may be surfaced, thereby revealing the latent power structures. Questioning power-providing structures
and revealing underlying structural power imbalances involve exploration of the rules, routines and
culture unconsciously held within both the board and the organization. One striking example of the need
for boards to challenge hypocrisy and the existing states of “normal” was offered by a respondent
engaged in an organization working to improve socio-economic conditions of impoverished people
abroad that he perceived to have lost its’ focus and priorities, “…when these people came for board
meetings it was more socializing than things being done at the board because, you know, the dinner, the
cocktails and all the other - I mean, you know, the idea of going across from Halifax to Victoria and
places like that when the board meetings are held. Everything is paid for. We even had board meetings
overseas.” It would be naïve to imply that all boards and board members openly accept attempts to refute
the existing states of normal, however, the very act of challenging the dominant assumptions of power
that underpin the board’s activities is often enough to surface mechanisms of marginalization. From this
perspective, an informant’s recollection of her boardroom experience illustrates how those within the
margins of power can pose a challenge to the existing normal, “Before I came on the executive, what was
happening at the board meeting was the executive would decide things - come to the board and present to
the board and the board says yes. .. the chair, thought that I would be the nice person to be appointed to
the executive. After about the third executive meeting, of course, she told me she said, ‘I am very
disappointed in you because, you know, we want executive solidarity’. Ah, so I said, you'll never get that
as long as I'm on the executive. You can be sure because I came on the board to represent certain views
and you will hear about those things’.”

75
 In terms of practical application, those who pushback against the predominant cultural swell
strive to reduce stakeholder alienation, which our respondents so often depicted as resulting from an
inability to recognize the reality endured by marginalized constituents. Three respondents provided
strikingly similar examples of this theme: first, “And the centre and they put those ah, kiosk there and the
initial designation was for drivers' licenses and you just fill these forms up and just apply. A lot of people,
you know, can't read, you know, so it's ludicrous.” second, “I said, ‘There is no point in me sitting here if
I'm gonna have so and so on the Health Council because they are from a different strata of society.’ They
can't even fathom the problems that these people have.” and finally, “I think that the best thing that has
ever come of it is the biased comments, is that people before thought that they knew what was good for
everybody else, started realizing that they don't.” Challenges of this nature serve to shift the thought
paradigm within the boardroom, where boards surface contradictions based on unchecked assumptions
which impair their organization’s ability to reach its aspirations.

Discussion and Conclusion

In this paper we have applied Bradshaw’s (1998) multi-dimensional model of power in order to
thematically present the ideas of informants from the voluntary sector regarding how they work to create
more diverse and inclusive boards. In presenting this paper we have used the words of the informants as
often as possible in order to communicate the depth of their insights and the complexity and magnitude of
the challenges that they face. Creating diverse boards is not a simple or straight forward endeavour. Each
informant had stories to tell, strategies to describe and success and frustrations to share. What emerged is
a complex array of possible approaches to creating more inclusive and diverse boards.

Prominent board driven initiatives emerged from our thematic analysis, several focusing
specifically on improving surface personal power bases of traditionally marginalized board members
including: enhancing an individual’s level of expertise; increasing association and relational bonds among
board members with specific focus on strengthening the social linkages between dominant and
marginalized board members; and, ensuring the inclusion of members from underrepresented groups in
positions characterized with legitimate authority. From a surface structural perspective some board driven
initiatives that emerged from our thematic analysis included increasing access to and control of critical
and scarce organizational resources by marginalized groups, ensuring structural flexibility, as well as
increasing the visibility of diverse groups thereby increasing their power within the boardroom. Also
emergent from our thematic analysis, were three broad approaches taken by boards that sought to surface
and address the underlying dynamics of the marginalized individuals’ struggles to gain more latent
sources of power. Primary among these approaches was the challenge of addressing privilege and the
need to endow opportunities to stand outside the dominant discourse, seeing how boards have
unconsciously conferred power to some and how this unearned right can be challenged and redistributed.
Similarly, through a process of questioning and de-familiarization individual board members described
how they replace an attitude of deference to power with an ability to make autonomous choices by
enhancing self-knowledge. Similarly, encouraging individuals to not only challenge internalized
disciplinary micro-practices, but also surface such unconscious disciplinary practices of others, permits
boards to change consciousness in ways which are often subversive and which create the possibility of
new realities. Finally, we found example of the deep structural power and thematic analysis revealed three
prevalent board-level strategies intended to deconstruct conventional limitations to change. These themes
included: emancipating the agenda; deconstructing current meaning systems; and challenging the
established definitions of what is “normal”.

The themes that emerged suggest the possibility that we can confront the inequalities of power
and the dynamics of exclusion and marginalization in the boardroom from different directions at the same
time or over time. Individuals can have their power basis enhanced and the structures, both
manifest/conscious and latent/unconscious/deep, can be adjusted and deconstructed. In these ways we can

76
 target change efforts and hopefully create sustained and profound change in the boardroom. Previous
exploration of organizational change from a power perspective (Bradshaw and Boonstra, 2004) suggests
that profound change will only happened if the multifaceted nature of power is understood and more than
one approach engaged.
Table 1 - Dimensions of Surface-Personal Power
Emergent Theme
Building Expert Power -
Orienting and Socializing
New Members
Enhancing Associative
Power
Creating Links to Powerful
Others
Increasing Legitimate
Authority
Putting Leadership in the
Hands of Marginalized
Communities
Table 2 - Dimensions of Surface-Structural Power
Emergent Theme
Increasing Access and
Control of Scarce
Resources
Shared Access to Sources
of Power
-
Assuring Flexibility
Leveraging Diversity to
Improve Responsiveness
and Dynamism
Demonstrating Relevance-
Making the Business Case
Quote
We had a full afternoon of orientation and there's a board package that they gave
to me as well as a renewed version of my board binder with all the policies and
everything in it cause apparently the one they originally gave me was wrong. At
that point in time it was a really worth while afternoon just kind of touring both
facilities and getting to know what the term policy governance actually means.
Um, however up until that point I've been paired up, I have a mentor and she's
been really helpful.
Women of colour who are active in the community active on the board, to beg
them to put something else on their place which is to mentor the younger and less
experienced board members.
So I thought because I got the names of these women who might have been
wonderful women, but it was just the same old pattern of grabbing from the
community and plucking them in so we decided not to actually do that. Instead
what we did was we tried to think together of women that we were working with
in the community who were from you know more marginalized communities or
traditionally marginalized communities, and decided to actually target them and
make phone calls and talk to them, and also be very open about that we were
looking for women of colour so they knew that this is what we're doing.
Quote
I'll tell you something, a long time ago it was written into the policy of the
organization … All of those committees have two coordinators, and it was
written that one must be female and one must be a male coordinator, and the
board must have a female and a male co-chair. We don't have one president, we
have a female and male co-chair. And I can tell you for sure we would not have,
we still have work to do on our gender equity, but we wouldn't have the level of
gender equity that we have now if we didn't have those policies written in.
So yeah I think a lot of things drive the push for diversity, and last of all but not
the least, organizational culture, in terms of recognizing that diversity brings
richness. Diversity brings new ideas. Diversity brings growth. Diversity brings
dynamism. Diversity brings energy. And lack of diversity means sameness,
dullness, lack of growth, unresponsive, so I think organizational culture which
includes willingness, political will, you know commitment, values, so yeah so all
of the above.
How you can make a diversity plan. The first thing that we did since I was a part
of this was we did an environmental scan for what the gaps were. Not only for
the board, but for the staff, and also for the clientele that we serve, and then we
took steps to try on how we were going to fill those gaps. And then we do up
those plans. Not only did we do a plan, but we monitor the plan and it starts
every six months. And we found that if we stuck to the plan, that you need
someone as well, kind of like a sounding board if you should have harassment
77
problems.

Emergent Theme
Acknowledging Privilege
and Tokenism
Naming Apathy
Increasing Self-Knowledge
-
Generating Insight from
Others' Lived Experience
Confronting Disciplinary
Micro-Practices
Exercising Resistance on
Behalf Others' Interest
Table 4 - Dimensions of Deep-Structural Power
Emergent Theme
Emancipating the Agenda -
Putting Diversity on the
Record
Deconstructing Meaning
Systems
Debating the Meaning of
Diversity
Challenging the Established
-
Definitions of "Normal"
Surfacing Contradictions
Between Assumptions and
Aspirations
Table 3 - Dimensions of Deep-Personal Power
Quote
So I started doing work in terms of inclusion, and one of the things that I faced
almost immediately was a challenge from a very powerful person on the board
who said to me, "Well isn't our focus in XXX women and refugee women of
colour?" And I almost was blown out of my seat. This was the strategic planning
day and I though oh my god, and I didn't know what to say, I was like floored, so
I said, "well okay, maybe I missed something." Okay then just tell me about the
work in the immigrant refugee with the colored communities that you do with
women with disabilities in those communities, given that women with
disabilities are 20% of those population. Silence, no discussion right, so really in
fact that the work that we were reporting to do with those communities weren't in
fact for all women in those communities. So it sort of made the point, and you
know those are the kinds of challenges that I think we face.
So from an individual perspective it also creates conditions, which make it very
hard for people to offer their best because they bring a lot of diversity in addition
to their minority status or being a person of colour or a woman, whatever. But I
think it's important to name that, to be constantly vigilant to it, to have good team
process, to recognize that you can build in ways of speaking to community and in
integrating voices directly. Using the person's own experiences and expertise.
Consulting others when you don't have that expertise, recognizing limits and
boundaries.
We need to talk to the organizational committee level, which is this little sub
community that is more issues and more conservative about how we need to be
aware of the language. So if somebody is talking at a very - talking in social
work language or talking in board language, that we rather then you know that
we be sensitive to the body saying and stop the person, and say could you just
explain what that means, or would you mind just elaborating a little bit for new
members, or that we will take on that onus.
Quote
So that's what's driving it is our desire as a board to be - there was a great phrase
that we stole from somebody else, "reflective of the community but not
representative of the community". So in other words what we're attempting to do
is we want to have people on the board who know communities, and bring that
that can inform the boards deliberation.
I think for some of the women on our board they would never be exposed to the
levels of diversity in there own either workplace or that kind of thing that they
are here. So, and I would say the same is true for the women when they're
coloured about the women who wouldn't ever come in contact with people with
disabilities or deaf people or anything like that. So it enriches their view of the
world. But I think it has influenced the people who are in the dominant culture,
who need to be the ones who share power and give up power. They need to be
able to see that people can take that power and take responsibility right.
The best way to serve, to surface and create space for those people is to have out
people in leadership positions in your organizations. So even if you think it's not
a problem, I'd like to see what happens if you did get an out person on your
board, I'd be curious to see what would start happening at your organization.
78
 References

Alvesson, M. “Leadership as social integrative action: A study of a computer consultancy company”.

Organization Studies, 13(2), 185-209, 1992.
Abzug, R. “The Evolution of Trusteeship in the Unitied States: A Roundup of Findings From Six Cities”,
Nonprofit Management and Leadership, 7: 101- 111, 1996.
Abrug, R. & J. Galaskiewicz, “Nonprofit Boards: Crucibles of Expertise or Symbols of Local Identities?”
Nonprofit and Voluntary Sector Quarterly, 30: 51-73, 2001
Bacharach, S. B., & Lawler, E. J. Power and politics in organizations. San Francisco, CA: Jossey Bass,
1980.
Berthoud, H., and Greene, R.D. “A Multi-Faceted Look at Diversity: Why Outreach is Not Enough.”
The Journal of Volunteer Administration, 2-10, 2001.
Bolen, J. S. Ring of power: The abandoned child, the authoritarian father, and the disempowered
feminine. New York: Harper Collins, 1992.
Bradshaw, P., "Power as Dynamic Tension and Its Implication for Radical Organizational Change",
European Journal of Work and Organizational Psychology, 7(2), pages 121-143, 1998.
Bradshaw, P. and J. Boonstra, “Power Dynamics and Organizational Change”, Organizational
Development and Change, ed.by J. Boonstra., Wiley, p. 279 – 300, 2004.
Brown, L. S. Subversive dialogues: Theory in feminist therapy. New York: Basic Books, 1994.
Buchanan, A., and O’Neill, M. “Inclusion and Diversity: Finding Common Ground for Organizational
Action [online]. A Deliberative Dialogue Guide.” Canandian Council for International Co-operation,
2001. Available from World Wide Web: http:// www.web.net/ccic-cci.
Burrell, G., & Morgan, G. Sociological paradigms and organisational analysis: Elements of the
Sociology of Corporate Life. London: Heinemann, 1979.
Clegg, S. R. Frameworks of power. London: Sage,1990.
Crozier, M. The bureaucratic phenomenon. Chicago: University of Chicago Press, 1964.
Crozier, M., & Friedberg, E. Actors and systems: The politics of collective action (A.Goldhammer,
Trans.). Chicago, The University of Chicago Press, 1980.
Daley, J.M., & Angulo, J. “Understanding the Dynamics of Diversity within Nonprofit Boards.” Journal
of Community Development Society, 25 (2), 172-188. 1994.
Daley, J.M., & Marsiglia, F.F. "Social Diversity within Nonprofit Boards: Members' Views on Status and
Issues." Journal of the Community Development Society, 32 (2), 290-309, 2001.
Dahl, R. A. “The concept of power.” Behavioral Science, 2, 201-215, 1975.
Devine, M.A. “Inclusion of Youth with and without Disabilities: More than Just Sharing the Same
Space.” In P.A. Witt, & L.L. Caldwell (eds.). Recreation and Youth Development. Pennsylvania: Venture
Publishing, 2005.
Eisenhardt, K. “Building theories from case study research”. Academy of Management Review, 14: 532-
550, 1989.
Emerson, R. M. “Power-dependence relations.” American Sociological Review, 27, 31-41, (1962).
Fink, J. S., Pastore, D. L., & Riemer, R. A. “Do Differences Make a Difference? Managing Diversity in
Division IA Intercollegiate Athletics.” Journal of Sport Management, 15 (1), 10-50, 2001.

79
Fletcher, K. “Building Diverse Boards: Lessons from a Case Study of Planned Parenthood Affiliates.”
In J. Miller, K. Fletcher, & R. Abzug (eds.). Perspectives on Nonprofit Board Diversity. Washington,
D.C.: National Centre for Nonprofit Boards, 1999.
Golensky, M. “The Board-executive relationship in nonprofit organizations: Partnership or power
struggle?” Nonprofit Management and Leadership, 4: 177-1919, 1993.
Hardy, C. The nature of unobtrusive power. Journal of Management Studies, 22(4), July,
384-399, 1985.
Hickson, D. J., Hinings, C. R., Lee, C. A., Schneck, R. E., & Pennings, J. M. “A strategic
contingencies' theory of intraorganizational power.” Administrative Science Quarterly, 16, 216-229, 1971.
Kang, C. & R. Cnaan, “New Findings on Large Social Services Organization Boards of Trustees”,
Administration in Social Work, 19: 17-44, 1995.
Kanter, R. M. Men and women of the corporation. New York: Basic Books, 1977.
Kotter, J. P. Power and influence: Beyond formal authority. New York: The Free Press, 1985.
Kramer, R. “Ideology, states and power in board-executive relationships” Social Work, 10: 107-114.
1965.
Kvale, S. Interviews: An Introduction to Qualitative Research Interviewing. Thousand Oaks,
California: Sage Publications, 1996.
Lukes, S. Power: A radical view. London: Macmillan Education Ltd., 1974.
McIntosh, P. “White privilege: Unpacking the invisible knapsack.” Peace and Freedom, July
August, 1989.
Mechanic, D. “Sources of power of lower participants in complex organizations.” Administrative
Science Quarterly, 7, December, 349-364, 1962.
Miller, J. “Managing Board Diversity: The Importance of Differentiating Between Diversity and
Representativeness. In J. Miller, K. Fletcher, and R. Abzug (eds.). Perspectives on Nonprofit Board
Diversity. Washington, D.C.: National Center for Nonprofit Boards, 1999.
Mills, A. J. “Organization, gender and culture.” In A. J. Mills & P. Tancred (Eds). Gendering
organizational analysis (pp. 93-111). Newbury Park, CA: Sage, 1992.
Mukherjee, A. The Kaleidoscopic Organization. Diversity and Inclusivity in Organizations: A
Framework for Change. Developed for the United Way of York Region, 2001.
Murray, V., Bradshaw, P. and Wolpin, J. “Power in and around nonprofit board: A neglected dimension
of governance” Nonprofit Management and Leadership 3: 165-182, 1992.
Murray, V.V., & Bradshaw-Camball, P. “Temptations and dilemmas in the interpretive perspective on
organizational politics.” In. G. Dlugos, W. Dorow, & D. Farrell (Eds.). Organizational politics: From
conflict suppression to rational conflict-management (pp. 51-70), Gabler, 1993.
Ostrower, F. Trustees of Culture: Power, Wealth and Status on Elite Arts Boards, Chicago: University of
Chicago Press. 2002.
Ostrower, F. & M. Stone, “Governance: Research Trends, Gaps, and Future Prospects”, In The Nonprofit
Sector: A Research Handbook, 2nd ed. W. Powell and R. Steinberg. New Haven: Yale University Press,
2005.

80
Patton, M. Q. Qualitative Evaluation and Research Methods. (2nd ed.) Newbury Park, California: Sage
Publications, 1990.
Perrow, C. “Departmental power and perspective in industrial firms.” In. M. N. Zaid (Ed.). Power in
organizations. Nashville: Vanderbilt University Press, 1970.
Pfeffer, J. Managing with power: Politics and influence in organizations. Boston, MA: Harvard
Business School Press, 1994.
Rosen, M. “Breakfast at spiro's: Dramaturgy and dominance.” Journal of Management, 11(2),
31-48, 1985.
Salancik, G.R., & Pfeffer, J. “The bases and use of power in organizational decision making: The
case of a university.” Administrative Science Quarterly, 19, 453-473, 1974.
Smircich, L., & Morgan, G. “Leadership as the management of meaning.” Journal of Applied
Behavioural Science, 18, 257-273, 1982.
Starhawk. Truth or dare: Encounters with power, authority, and mystery. Harper San Francisco, 1990.
Townley, B. “Reframing human resource management: Power, ethics and the subject at work.” London:
Sage Publications, 1994.
Whittington, R. “Putting Giddens into action: Social systems and managerial agency.” Journal of
management studies, 29 (6), 693-712, 1992.
Winter, S. “The "power" thing.” Virginia law review. 82 (5), 721- 835, 1996.
Wrong, D. H. “Some problems in defining social power.” The American Journal of Sociology, 673 -681,
1965.
Yin, R.K.. Case study research: Design and methods, 3rd edition. Thousands Oaks, CA: Sage
Publications, Inc., 2003.

81
ASAC 2006 Alison M. Konrad
Banff, Alberta Yang Yang (Ph.D. student)
Cara Maurer (Ph.D. student)
Richard Ivey School of Business
University of Western Ontario

HUMAN RESOURCE MANAGEMENT PRACTICES FOR ACHIEVING DIVERSITY

ADVANTAGE: DETERMINANTS AND OUTCOMES OF THE CONFIGURATION OF
DIVERSITY MANAGEMENT STRUCTURES

Adopting the resource based view and institutional theory, we classified

diversity initiatives into three categories labeled institutional, strategic, and
configurational diversity structures, respectively. We suggest that the structures
will have different effects on the two components of diversity advantage:
capital and process advantage.

A number of studies in Australia, Canada, and the United States have shown that managerial
practices intended to increase the demographic diversity of the workforce are associated with greater
representation of women and minority groups (French, 2001; Holzer & Neumark, 2000; Konrad &
Linnehan, 1995; Leck & Saunders, 1992; Leck, St. Onge & LaLancette, 1995; Moore, Parkhouse &
Konrad, 2001). Each of the studies has examined a somewhat different set of practices. Holzer and
Neumark’s study was relatively narrow, examining only recruiting activities. The other sets of authors
examined a comprehensive set of human resource management practices (HRM), including recruitment,
selection, training, promotion, and conditions of service as well as managerial support for diversity-
related programs.

Only two papers have theorized about the configuration of diversity-related management
practices. In their qualitative study, Ely and Thomas (2001) identified three different managerial
perspectives on diversity: the “discrimination and fairness” perspective, focusing on creating fair
processes in the organization in order to comply with legal mandates, “access and legitimacy”
perspective, focusing on access to a more diverse customer base, and “integration-and-learning”
perspective, focused on integrating the knowledge, skills and work methods of a diverse workforce into
organizational processes. French (2001) theorized that diversity-related HRM practices reflect four
configurations. In the classical disparity structure, there is no acknowledgement of structural barriers for
members of historically excluded groups and no diversity-related HRM practices are put into place. An
anti-discrimination structure, strives for equal treatment of all identity groups, similar to Ely & Thomas’
(2001) discrimination and fairness perspective. The affirmative action approach takes the additional step
of acting to redress past discrimination. In the fourth structure, which French (2001) labels gender equity,
reflecting Australia’s legislative focus on eradicating gender discrimination in employment, “Issues such
[as] gender, culture and cultural requirements and the impact of family commitments on work
requirements, their effect on career choices, as well as access to education are to be recognized and
addressed in the workplace” (p. 271). French’s factor analytic findings supported her four-category
model.

82
 The current paper contributes to the advances made by previous authors by developing and
testing a theory regarding the determinants of the configuration of diversity-related structures in
organizations. We argue that the implementation of diversity-related structures reflects management’s
motivation to engage a diverse workforce for strategic reasons, for legitimacy reasons or both. In the
following sections, we develop a theoretical view of the determinants and outcomes of organizational
configurations of diversity management structures based on these motivations. We test our ideas on
survey and archival data collected from a sample of 155 Canadian organizations.

Theoretical Background

The Resource Based View

The resource based view (RBV; Barney, 1991;2002) argues that an organization can acquire and
develop resources and capabilities to create a competitive advantage. Organizational resources include
both tangible (e.g., plant, equipment, and raw materials) and intangible (e.g., culture, human capital)
assets. Resources that are valuable, rare and inimitable are sources of competitive advantage when
combined with supportive organizational factors, such as organizational structure and culture (Barney,
2002). Capabilities are organizational processes that “integrate, reconfigure, gain and release resources”
(Eisenhardt & Martin, 2000, p.1107). Some researchers suggest that capabilities are also a source of
competitive advantages if they have the three characteristics of high value, rarity, and inimitability (e.g.,
Teece, Pisano, & Shuen, 1997).

The application of the RBV to human resource management argues that employees are important
resources and can create a competitive advantage (e.g., Schuler & MacMillan, 1987; Ulrich, 1991)
because human resources can be value-adding, unique, and inimitable (Wright & McMahan, 1992).
Furthermore, human resource practices such as employee selection, training, appraisal, and rewards
determine the organization’s capabilities in terms of the ability to leverage the value of its human assets
(cf. Wright & McMahan, 1992), and as such, can serve as a means for creating competitive advantage.
Furthermore, the strategic human resource management perspective argues that human resource practices
should be aligned with strategy, and the degree of congruence between the strategies and practices
determine whether an organization can achieve superior performance (Delery & Doty, 1996).

Some researchers have applied the rationales of the RBV to diversity and suggest that diversity
can provide a competitive advantage (e.g., Richard, 2000; Richard & Murthi, 2004). For instance, it is
suggested that racial diversity is valuable since it can increase an organization’s sensitivity to customers’
needs and improve the organization’s response to environmental change; it is rare due to its limited
supply; and it is inimitable since the socially complex dynamics of a diverse workforce are difficult to be
copied (Richard & Murthi, 2004). Hence conceptually, diversity has the potential to be an important
resource. However, human resources do not “produce on their own” (Richard & Johnson, 2004, p.135),
especially when the resources consist of members of different demographic groups. The suggestion that a
diverse workforce is naturally a valuable resource ignores the complex and often negative processes
associated with diversity at the individual, inter-group, and organizational levels. For instance, individual-
level stereotyping and categorization can increase negative prejudice and discrimination (e.g., Fiske,
1998). Power differences and inequality between demographic groups can increase inter-group conflict
(e.g., Linnehan & Konrad, 1999). And lastly, the natural tendency for organizations to become
homogeneous is inconsistent with the goal of multiculturalism, and can mitigate the proposed positive
effects of diversity (Richard & Johnson, 2001). For these reasons, diversity must be managed in order to
generate value-adding capabilities for the organization. Moreover, Oliver (1997) suggests that resource
acquisition and deployment cannot avoid the impact of institutional factors.

83
Institutional Theory

Institutional theory addresses both external and internal factors that influence organizational
structures. Research focusing on external factors suggests that external environments control important
resources, and organizations have to conform to environmental pressures in order to acquire legitimacy
that is necessary for survival (DiMaggio & Powell, 1983; Meyer & Rowan, 1977). Institutional theorists
have identified three dimensions of institutional environments: constituents, forces, and logic. For a
specific organizational field, the institutional constituents include “regulatory structures, governmental
agencies, laws, courts, and professions” (Scott, 1987, p.498). And institutional forces include coercive,
normative and mimetic forces (DiMaggio & Powell, 1983; Scott, 1995). The logic of coercive forces is
instrumental and the conformity pressure derives from fear of losing benefits. The logic of normative
forces is “appropriateness” and the pressure of conformity comes from “expectations” or “social
obligations”. And the logic of cognitive or mimetic force is “sensemaking” and the pressure of conformity
comes from social identity (Scott, 1995).

For research focusing on internal institutionalization, it is suggested that adopted practices or

structures will experience pre-, semi-, and full stages of institutionalization (cf. Barringer & Milkovich,
1998). And structures that do not reach full institutionalization will eventually be abandoned or
substituted (Tolbert & Zucker,1996). The relationship between internal and external institutional forces is
not independent, however. For instance, based on a power perspective, Lawrence and Jennings (2001)
suggest that external institutional forces can facilitate internal institutionalization processes.

Institutional theory has been adopted to explain organizations’ structural characteristics (Meyer,
Scott, & Deal, 1983; Meyer, Scott, & Strang, 1987; Scott, 1987; Scott & Meyer, 1987; Singh, Tucker, &
House, 1986) and organizational change (Hinings & Greenwood, 1988; Tolbert & Zucker, 1983). Recent
empirical research demonstrates that institutional theory has strong explanatory power regarding the
adoption of novel government structures (e.g., Lee & Pennings, 2002), family initiatives (Goodstein,
1994), flexible benefits plans (Barringer & Milkovich, 1998), accommodations for workers with
disabilities (Balser, 1999) and diversity management structures (Konrad & Linnehan, 1995).

Conceptually, institutional theory is appropriate for research on diversity. For instance, external
institutional forces exist that pressure organizations to become more diverse. Coercive institutional
pressures come from legislation, such as equal employment opportunity and affirmative action in the
United States, and the Equity Employment Act in Canada. Normative pressure derives from influences of
community and professional associations, such as the Society for Human Resource Management in the
United States, which conducts surveys on organizational diversity practices and regularly publishes
articles and guidelines for how to manage diversity effectively. Mimetic forces mainly arise from an
organization’s competitors, and markets and media can convey information regarding organizations’
diversity practices and facilitate the imitative processes. These external forces come up against
institutionalized structures internal to organizations that support individual similarity, reward inequality,
and emphasize homogeneity as multiculturalism is often contrary to organizations’ principles of
efficiency (e.g., Hoobler, 2005). As a result of the staying power of institutionalized internal structures,
organizations vary in the extent to which the adopt diversity management practices promulgated in the
institutional environment.

Diversity Management Structures

We suggest that resource considerations and institutional forces jointly motivate organizations to
adopt different diversity structures. When organizations pursue workplace diversity for strategic reasons,
the motivation for diversity is to achieve success through the successful execution of strategy. When

84
organizations pursue diversity for legitimacy reasons, the motivation for diversity is to avoid the threat of
regulation or damage to their reputation. In this section, we classify diversity structures into three
categories based on the strength of these two forces (see Figure 1).

Figure 1

Relationships between Motives for Diversity and Diversity Structures

Success Achievement

Weak Strong

Threat Weak Classical Disparity Structures Strategic Diversity Structures

Avoidance Strong Institutional Diversity Structures Configurational Diversity Structures

Strategic Diversity Structures

We define strategic diversity structures as diversity management structures adopted due to an

organization’s need to achieve success through a diverse workforce. The rationale underlying these
structures is that a diverse workforce is a valuable resource which can facilitate strategy formulation and
implementation. Hence, the goal of strategic diversity structures is to create and exploit heterogeneous
human resources through effective management. These structures are shown in the upper right-hand
quadrant of Figure 1 as indicating strong motivation for success achievement and weak motivation for
threat avoidance.

To realize the goal, strategic diversity structures will emphasize retaining and deploying
demographically heterogeneous resources to realize economic goals. First, the structures will be
associated with organizations’ business strategies. This prediction conforms to the argument of the RBV
that resources can create a competitive advantage only when they are used for formulating and
implementing business strategies. Second, strategic diversity structures need to ensure that differences
across different demographic groups are maintained, and employees’ different needs for individual
development are addressed, such as using individualized socialization and internal labor markets.
Research on human resource management has suggested that individualized socialization processes, such
as individualized coaching or mentoring, are effective in terms of accommodating individual differences
(McMillan-Capehart, 2004). And internal labor markets can increase values of human resources because
promotion opportunities generate motivation to perform well, learn new skills, and stay with the firm
(Lepak & Snell, 1999). Third, strategic diversity structures will reward diversity to emphasize its
importance for an organization. Finally, the structures will exploit diversity by integrating diversity into
organizational processes that can facilitate an organization’s value creation or adaptation. For instance,
organizations may increase diversity at boundary spanning positions in cases where adaptation to
demographic changes of customers is valuable. Or, organizations may increase diversity in product
development positions if organizational strategy is innovation-oriented.

Institutional Diversity Structures

Institutional diversity structures are diversity management structures adopted due to an

organization’s need to avoid the negative effects of violating the requirements of the external

85
environment. The rationale for adopting institutional diversity structures is to facilitate the acquisition of
legitimacy. Hence, the goal of institutional diversity structures is to conform to institutional forces
through increasing the representation of traditionally underrepresented demographic groups. These
structures are shown in the lower left-hand quadrant of Figure 1, indicating weak motivation for success
achievement and strong motivation for threat avoidance.

To realize this goal, the structures will first exhibit commonality across different organizations.
Organizations’ adoption of the same set of diversity structures can be attributed to their conforming to
legislative mandates or their mimicking the “best practices” of the industry in which they operate,
consistent with the isomorphic effects proposed by institutional theory (DiMaggio & Powell, 1983).
Second, the adopted structures will not necessarily fit with the organization’s business strategy and may
be independent from other human resources practices. Third, as a reflection of the external environments,
institutional diversity structures will exhibit a focus on external recruitment as external recruitment has a
quick and obvious effect on increasing diversity. Finally, similar to the diversity perspective of “fairness-
and-justice” (Ely & Thomas, 2001), the structures will link diversity to visible positions in order to show
an organization’s conformity to institutional pressures.

Configurational Diversity Structures

The sources of the different structural characteristics of strategic and institutional diversity
structures are organizations’ different motives for pursuing diversity. Strategic diversity structures reflect
an organization’s need to retain and exploit a diverse workforce, a need to outperform competitors in
generating economic gains, and a need to adapt to changing demographics and globalization. Institutional
diversity structures reflect an organization’s need to avoid the negative effects of violating the
requirements of external environments. The organization is driven by a need to avoid uncertainty, a need
to maintain stability, and a need to reduce environmental influences.

It is possible, however, for an organization to be driven by both success achievement and threat
avoidance at the same time. Under such circumstances, we argue that configurational diversity structures
will be developed to address the two types of needs. The goal of configurational diversity structures is
both to create and maintain organizational heterogeneity to simultaneously achieve success through a
unique and inimitable competitive advantage and to be viewed as legitimate in order to avoid the negative
effects imposed by external and internal institutional forces.

To realize this goal, configurational diversity structures will first relate diversity to corporate
strategies. In addition to guiding business development, corporate strategies also consider organizational
adaptation to external and internal environments, such as changes of institutional factors (Oliver, 1997).
Second, the structures will integrate the mandated diversity structures or “best practices” with
idiosyncratic structures to accommodate the two different types of needs. The mandated structures or
“best practices” have symbolic effects for organizations by showing conformity, whereas idiosyncratic
practices reflect organizations’ distinctive approaches to managing diversity in the best way given their
unique internal and external situation. Third, the structures will include the practices addressing both
legitimacy and performance needs, such as the external recruitment and internal motivation, development
and retention, and both individualized and institutionalized socialization tactics. These seemly
contradictory practices will be adopted by the structures in a balanced way in order to address both
motives. Finally, in the sense that multi-faceted efforts are needed to meet strategic and legitimacy needs,
configurational diversity structures will integrate the notion of diversity into the whole human resource
management process from selection, training, appraising to rewarding, and evenly distributing diversity
across a variety of jobs.

86
 One could argue that some organizations may have neither a need for achievement nor a need for
threat avoidance in the area of diversity. Examples might be small businesses that are not subject to as
many legal requirements for equal employment operating in niches that do not exhibit much demographic
diversity. We label those human resources structures without a motive to pursue diversity “classical
disparity management structures” following French (2001).

Hypotheses

Following Boxall’s (1998) view that human resources provide two forms of competitive
advantage, specifically, capital advantage and process advantage, we suggest that the organizational
benefits of diversity include both a capital and a process dimension. Diversity capital advantage refers to
an organization’s superior composition of different demographic and cultural groups that can ensure that
the organization outperforms its competitors in terms of the availability of a variety of knowledge, skills
and perspectives. Diversity process advantage consists of superior problem solving and superior social
integration and communication processes deriving from different demographic groups that can help an
organization to execute operations more quickly, efficiently and effectively, and thus outperform its
competitors.

In this study, we measure diversity capital advantage as the percentages of women, visible
minorities, and persons with disabilities employed by the organization at various hierarchical levels. To
capture diversity process advantage, we examine the relationship between diversity structures and firm
financial performance. Below, we discuss the relationship between the four types of diversity structures
and diversity advantage to derive a set of hypotheses about the antecedents of diversity capital advantage
and diversity process advantage.

Strategic Diversity Structures and Diversity Advantage

Strategic diversity structures are a reflection of an organization’s desire for diversity in order to
achieve performance goals. The structures can first lead to a diversity capital advantage. Specifically,
strategic diversity structures will increase the demographic and cultural diversity of organizational
members to match changing customer demographics, and therefore to exploit market opportunities (Cox,
1994). This outcome derives from the close association between the diversity management structures and
the business strategy. Conceptually, one of the basic pillars of the business case for diversity is the
assertion that a diverse employee base is valuable for bringing intelligence concerning increasingly
diverse sets of customers inside the organization (Cox & Blake, 1991). Additionally, because people with
different demographic characteristics can bring different styles, insights, and perspectives into an
organization (Ely & Thomas, 2001), increasing cultural and demographic diversity will also enhance the
variety of knowledge and skills. Realizing both of these factors requires that the organization diversify its
human capital base, and organizations that seek diversity for strategic reasons must bring this variety of
knowledge inside their boundaries by using recruitment and selection practices that are likely to generate
a diverse set of highly qualified employees.

Beyond bringing more diversity into the organization, strategic diversity structures can enhance
diversity capital by investing in internal training and development. Organizations with a strategic view of
diversity are likely to invest in their diverse workforces in order to achieve superior motivation, skill
building, and retention outcomes. Internal training and development increases the value of human
resources because they constitute “isolation systems” (Rumelt, 1984). Isolation systems are characterized
by practices emphasizing developing knowledge and skills that are specific to an organization, and
therefore are a necessary condition for creating inimitable resources (Barney, 2002; 1991). Therefore,
training and development investments in a diverse workforce can result in the creation of a rich human

87
capital base comprised of people with a unique blend of varied experiences, perceptions, values, and
beliefs who are educated in the specific knowledge and skill sets necessary to understand and execute the
particular mission and strategy of the business.

Finally, strategic diversity structures may focus on ways to enhance managers’ commitment to
diversity as well as their skills for managing a diverse workforce. Managers’ commitment to diversity is
a prerequisite for the effectiveness of diversity management (e.g., Robinson & Dechant, 1997).
Managers’ commitment can also signal the value of diversity to outsiders, which in turn will increase an
organization’s attractiveness to members of a variety of demographic groups due to the rationale of
person-organization fit (Richard & Murchi, 2004). The attractiveness will make an organization
outperform its competitors in terms of selecting and recruiting talented and knowledgeable people from
diversely demographic backgrounds, the group of people whose supply is suggested to be limited
(Richard & Murchi, 2004). One method for increasing managers’ commitment to diversity is to create a
link between effective diversity management and the managerial performance appraisal and reward
system. Hence, we expect that organizations with a strategic view of diversity will hold managers
accountable for effective diversity management and that these organizations will be more attractive to a
diverse set of demographic and cultural groups. Thus, we propose:

H1a: Strategic diversity structures will be positively associated with an organization’s diversity
capital advantage, as assessed by organizational employment statistics.

Strategic diversity structures can also lead to diversity process advantage. First, linking diversity
to business strategies, such as marketing and innovation, can increase demographic group members’
positive attitudes toward diversity (Richard & Kirby, 1999), hence increasing goal consensus. Goal
consensus can increase the congruency of employees’ value systems, which has been described as
“diversity absorptive capability” (Ofori-Dankwa & Julian, 2002). Value congruence combined with
demographic diversity has been posited to result in creative decision making processes (Ofori-Dankwa &
Julian, 2002). Value congruence around the organization’s diversity goals is likely to be especially
important for realizing the strategic value of diversity because employees who do not accept the
organization’s diversity goals are likely to cause many process difficulties. In order for diversity process
advantage to be achieved, all employees must be willing to be respectful of differences, to accommodate
diverse needs, and to be willing to learn from coworkers who are different from themselves. As such,
diversity process advantage requires that all employees embrace the organization’s diversity goals.

Second, the internal diversity training, individual development, and retention emphases
characterized by strategic diversity structures will increase people’s awareness and acceptance of
individual differences. An understanding and openness to difference is likely to result in increased
communication and improved rapport across different demographic and cultural groups.

Third, applying diversity to meet business needs and create value can improve organizational
processes. For instance, emphasizing the implication of diversity for marketing and linking diversity to
boundary positions will make an organization’s workforce match the demographic characteristics of the
organization’s customers, and therefore speeding up the organization’s responses to customer and markets
changes, and facilitating problem-solving processes. Also, emphasizing the implication of diversity for
innovation can provide demographic group members with opportunities to participate in an organization’s
decision making processes, and therefore opportunities to air their views. These conditions are proposed
to be related to higher quality of decision making (Nemeth, 1992).

Finally, practices that enhance management commitment to diversity are likely to result in
diversity process advantages. Without management commitment, diversity becomes a set of policies on
paper without any substantive impact on organizational processes (Cox, 1994). Because organizations

88
that adopt diversity for strategic reasons intend for diversity to influence the way the organization does
business, practices must have substantive impact for the organization to achieve its goals. Holding
management accountable for the effective management of diversity requires managers to achieve the
potential problem-solving and creativity benefits of having diverse knowledge and viewpoints inside the
organization.

In summary, strategic diversity structures are designed to influence the organization’s internal
processes, with the result that members of a diverse set of demographic and cultural groups feel fully
included in both formal decision-making processes and informal social relationships. Thus, we propose:

H1b: Strategic diversity structures will be positively associated with an organization’s diversity
process advantage, as assessed by return on assets.

Institutional Diversity Structures and Diversity Advantage

Similar to strategic diversity structures, institutional diversity structures can also lead to diversity
capital advantage. Institutional diversity structures will increase demographic and cultural diversity of
organizational members in order for an organization to conform to legislative demands, as well as to
respond to the normative and mimetic pressures. The institutions that exert pressures on organizations to
increase representation of different demographic groups usually emphasize quantitative outcomes of
diversity. Quantitative measures are easier to be acquired and compared. Also, quantitative outcomes are
more observable and will have a significant symbolic effect to outsiders, which in turn can increase an
organization’s attractiveness to different demographic groups and facilitate the organization’s selecting
and recruiting a diverse talent pool. One could argue that the structures will not be able to bring a superior
outcome to an organization due to their homogeneity across a variety of organizations. But based on the
rationale of commonality of organizations’ capabilities (Eisenhardt & Martin, 2000), we suggest that the
structures still can create a distinctive composition of different demographic groups and the composition
cannot not be easily imitated by its competitors. Oliver (1996) suggests that organizational responses to
institutional pressures can result in competitive advantage due to the fact that managers in different
organizations respond show different reactions to such pressures. In addition, the structures will make an
organization outperform those competitors who are less motivated to avoid the negative effects of non-
compliance, for instance, because they are less likely to come under scrutiny and hence, less subject to
institutional pressures. Thus, we propose:

H2a: Institutional diversity structures will be positively associated with an organizations’

diversity capital advantage, as assessed by organizational employment statistics.

However, unlike strategic diversity structures, institutional diversity structures will not relate to
diversity process advantage. First, the independence of institutional diversity structures from other human
resource management structures might lead to inconsistency between different practices. The positive
effects of each “best” diversity practice might be offset due to the conflict between the practice and an
organization’s other human resource management structures. Without the support from other human
resource practices, the structures cannot maintain their effectiveness. Second, the structures’ weak linkage
to an organization’s strategies makes them difficult to convey messages likely to generate consensus
among organizational members about valuing diversity. The lack of consensus will hinder employees’
sense making processes and is unable to result in a shared value system. As argued above, consensus-
building around the value of diversity may be particularly important for diversity process advantage in
order to motivate all employees to respect differences, accommodate differences, and show a willingness
to learn across difference. Unless employees are willing to engage in these behaviors, the processes of
communication, relationship-building, and decision-making are likely to be hindered by diversity
(Williams & O’Reilly, 1998).

89
 Third, the structures’ external recruitment focus and neglect of motivation, development, and
retention will have a negative effect on employees’ commitment. Although an organization is able to copy
all the “best” practices, it cannot easily imitate other organizations’ diversity related processes, especially
the individually-focused processes that are customized for a particular configuration of demographics and
cultures.

The symbolic efforts of “valuing diversity” may not be enough to convince employees that the
structures have met the promise. To the extent that institutional diversity structures cannot create an
environment whereby employees will share similar values and at the same time maintain distinctively
demographic characteristics, we propose:

H2b: Institutional diversity structures will not be associated with an organization’s diversity
process advantage, as assessed by return on assets.

Configurational Diversity Structures and Diversity Advantage

In order to simultaneously address a need to achieve success and a need to avoid threat,
configurational diversity structures intend to make substantial efforts in obtaining a diverse talent pool,
and therefore will lead to diversity capital advantage. First, as with strategic diversity structures, the
configurational structures will be integrated with the business strategy as well as the other human
resource practices in the organization, including the diversity-related “best practices” adopted for
legitimacy concerns. The configuration combining the “best practices” with other diversity practices in a
synergistic way will increase an organization’s effectiveness in selecting and recruiting highly educated
and skilled employees from a variety of demographic groups. Second, like strategic practices,
configurational practices will include internal training, development, and retention systems to improve
employees’ skills, increase their commitment, and therefore their intention to stay. The combination of
external recruitment and internal development will facilitate updating and maintaining an appropriate
diversity composition that keeps pace with changes in the diversity of the organization’s environment.

Third, the configurational approach considers the appropriate distribution of demographic and
cultural groups for achieving both legitimacy and business goals. As a result, the configurational
structures will aim to distribute the diverse organizational talent pool among decision-making groups
where diversity is important to achieving business strategy as well as among positions that are highly
visible for symbolizing legitimacy to internal and external constituencies. This combination of results is
not generated by either strategic or institutional structures alone, and consistent with Oliver’s (1996)
argument, achieving both strategic and legitimacy goals is likely to be associated with superior
performance. The combination of institutional and strategic structures is also likely to be more difficult
for competitors to imitate, as an organization with a configurational diversity structure has achieved
integration between strategic diversity practices and institutionalized diversity practices as well as
integrating both with business strategy and the total human resource management system. Hence, the
structures will not only relate to diversity capital advantage, but also may have a stronger effect than that
of the simple sum of the other two structures. Thus, we propose:

H3a: Configurational diversity structures will be positively associated with an organization’s

diversity capital advantage, and the positive effect will be stronger than that of the sum of strategic and
institutional diversity structures.

Configurational structures will also be more effective for creating diversity process advantage.
On one hand, the structures’ association with corporate strategies ensures employees’ goal consensus and
common understanding of the value of diversity, and therefore will increase an organization’s “diversity
absorptive capability” and improve its decision making and problem-solving processes (Ofori-Dankwa &

90
Julian, 2002). On the other hand, the association of diversity structures with other human resource
management structures will convey consensus information about diversity across different functional
areas, thereby reducing dysfunctional conflicts among practices. The horizontal and vertical fit combined
together will increase employees’ shared understanding of equality and fairness of the diversity
management structures. Research has suggested that perceived fairness of human resource management
structures are positively associated with their ease of being accepted, a characteristic labeled
“acceptability” of human resource management (Bretz, Milkovich, & Read, 1992; Waldman & Bowen,
1998). And acceptability is proposed as a determinant of the extent to which human resource management
structures can influence employees’ attitudes and behaviors (Bowen & Ostroff, 2004). We suggest that
the high acceptability of configurational diversity structures will also influence employees’ attitudes and
behaviors toward diversity, such as their willingness to comply with diversity policies and readiness for
change. Due to the deep roots of organizational tendency to become homogeneous, an organization’s
wide acceptance of diversity will facilitate other types of organizational change (Iles & Hayers, 1997).
The readiness for change in conjunction with the “even distribution” of diversity across an organization
will increase mutual learning, a process which is proposed to be related to the quality of decision making
(Ely & Thomas, 2001).

In addition to influencing people’s attitudes and behaviors, acceptable human resource

management structures foster a positive organizational climate (Bowen & Ostroff, 2004). A positive
organizational climate for diversity has important implications for the effectiveness of diversity since it
ensures individual differences in values, opinions and beliefs are encouraged and accepted (McMillan-
Capehart, 2004). Because they are acceptable, configurational diversity structures can improve the
consensus of employees’ value systems on the one hand, and maintain distinctive demographic
differences on the other hand. The appropriate tension between differences and similarity will improve
employees’ social interaction and communication, and facilitate cooperation and problem solving
processes. In the sense that socially complex dynamics among demographically diverse groups are
difficult to imitate (Richard & Murthi, 2004), the dynamics created through patterns of complex social
relationships will have a stronger effect on diversity process advantage. Thus, we propose:

H3b: Configurational diversity structures will be positively associated with an organization’s

diversity process advantage, and the effect will be stronger than that of the sum of the strategic and
institutional diversity structures.

Method

We developed a survey to measure diversity-related HRM structures in Canadian organizations.

The survey was carefully designed with input from Canadian Human Resource Managers and Canadian
line managers in the banking, telecommunications, petroleum products, engineering, power generation,
and consulting industries. In-depth interviews were conducted with 7 Canadian managers and consultants
specializing in managing diversity. In these interviews, human resource management practices for
managing diversity were identified. The first draft of the survey was distributed to a pilot sample of 6
diversity management specialists, who provided comments on the survey form. The final four-page
survey form focused on six types of diversity-related structures: (1) linking diversity to strategic business
goals and human resource planning (α = .85), (2) recruiting a diverse workforce (α = .71), (3) selecting a
diverse workforce (α = .70), (4) training and developing a diverse staff (α = .51), (5) monitoring the
effectiveness of staffing for diversity (α = .78), and (6) providing work-life flexibility (α = .89).

In Fall, 2004 and Spring, 2005, the survey was mailed to the HR managers in 935 firms. The
sample of firms included those organizations with at least 100 employees that were either listed in the
Globe and Mail 1000 largest companies or had submitted an employment equity report in 2002 (the most

91
recent set of reports available on-line to the public at the time). HR managers Human Resource managers
from 155 medium to large-sized Canadian organizations responded to the survey (17.6% response rate
after 4 mailings and at least one telephone contact). Respondents received a summary report of the
results.

Measures

Diversity structures. Following Arthur (1992) and French (2001) we created indices reflecting
the 6 types of structures included in the survey and conducted a cluster analysis on the basis of the
organizational scores on the 6 indices (see Table 1). Findings from the cluster analysis indicated three
patterns of diversity-related structures among the Canadian organizations in our sample. The first pattern,
representing 43 organizations in our sample (31%) showed low levels of diversity-related structures in all
seven areas, indicating low motivation to engage in workplace diversity issues. We labeled this pattern
the “classical disparity” approach, following French (2001) and Figure 1.

The second pattern, representing 50 organizations in our sample (36%) showed moderate to high
levels of diversity-related structures in the areas of selection, training and development, and monitoring
the effectiveness of staffing for diversity and relatively low levels of all other diversity-related structures.
We labeled this pattern “institutionalized diversity structures” because Canadian case law has legitimated
certain selection practices and de-legitimated others (e.g., a number of interview questions are illegal in
Canada, such as questions about gender, ethnicity, family structure, or disability status). Also, Canadian
employment equity legislation requires the development of monitoring systems in order for firms to
provide the numbers needed for their annual employment equity reports (required only of certain
industries in Canada).

The third pattern, representing 45 organizations in our sample (33%) showed the highest levels of
diversity structures in all 7 areas, and particularly in the areas of linking diversity to strategic business
goals and human resource planning, recruiting a diverse workforce, providing work-life flexibility and
creating an inclusive working environment. We labeled this pattern the “configurational” approach
because these firms showed high levels of the institutionalized structures and also had high levels of
strategic structures linking diversity to the business strategy, actively recruiting to bring in a more diverse
workforce, and developing substantive work-life and inclusiveness programs in order to retain a diverse
workforce.

Antecedents of Diversity Structures. To test our theory regarding the impact of motivations for
diversity on diversity-related HRM structures, we examined whether several factors predicted the cluster
categorization for the organizations in our sample. Institutional pressures for diversity were measured
with survey responses assessing organizational size (n of employees), whether the organization was
subject to Canadian employment equity legislation, whether the organization was a signatory to the
Canadian Federal Contractors Program (which has diversity requirements), whether the organization had
an HRM department (reflecting normative pressures resulting from the HRM profession’s attention to
workplace diversity issues), and whether the organization had a diversity expert on staff (also reflecting
normative institutional pressures).

Strategic motivation for diversity was measured by assessing whether the HRM manager was
included in the development of the organization’s strategic direction, whether the firm used high
involvement management practices (a four-item index asking what percentage of employees work in self-
managing teams, problem-solving or quality groups, project teams or task forces, and employee policy or
strategy committees α = .72), and whether the firm was internationalized (had any employees in foreign
operations).

92
 Outcomes of Diversity Structures. Outcome measures were taken from archival sources. Of
the 155 respondents to our survey, 81 had filed employment equity reports, and we examined the
percentage of women, visible minorities, and persons with disabilities employed in 2002 and 2003 to
assess diversity capital advantage. Because firms operated in a variety of industries with different human
resource bases, some would be expected to have greater participation of various constituencies due to the
nature of the work as well as the geographic location of the organization. Hence, rather than predicting
the base percentages of women, visible minorities, and persons with disabilities, we calculated the growth
in representation of all of these groups between 2002 and 2003 (the latest employment equity statistics
available on-line to the public). Growth statistics were calculated by taking the number employed for
each constituency in 2003 (e.g., the number of women), subtracting the number in 2002, and dividing the
difference by the total in 2002 (i.e., growth in terms of the % of the total in 2002).

We assessed diversity process advantage by examining the association between diversity

structures and return on assets in 2003. Of the 155 respondents to our survey, 87 were public companies
with data on firm financial performance publicly available.

Results

Nominal regression analysis was used to predict whether the diversity structures in each
organization belonged to the classical disparity, institutional, or configurational clusters. Findings
indicated that organizational size and the presence of an HRM department distinguished between the
configurational cluster and the classical disparity cluster (the reference category in the nominal regression
analysis). The presence of a diversity expert on staff, being subjected to employment equity legislation,
and being a signatory to the Federal Contractors Program distinguished between the configurational
cluster and the classical disparity cluster as well as between the institutional cluster and the classical
disparity cluster. These findings support the notion that institutional pressures result in the development
of institutional and configurational diversity-related structures in organizations.

The inclusion of HRM in developing business strategy distinguished the configurational cluster
from the classical disparity cluster but not between the institutional cluster and the classical disparity
cluster. This finding is consistent with the idea that only organizations with a strategic interest in
developing diversity develop diversity-related structures linked to the business strategy, actively reach out
to bring more diversity into the workplace, and build substantive programs aimed at retaining a diverse
workforce. The utilization of high involvement management practices and organizational
internationalization were not related to the development of diversity-related structures in our sample.

H1a and H1b predicted that strategic diversity structures would be positively associated with
diversity capital advantage and diversity process advantage, respectively. Because we did not observe a
purely strategic cluster of diversity structures in our sample, these predictions could not be tested
separately from the impact of the configurational structures.

H2a and H2b predicted that institutional diversity structures would be positively associated with
diversity capital advantage but not with diversity process advantage. Regressions analyses controlling for
firm size and industry sector (manufacturing or service) indicated that firms in the institutional cluster
showed significantly greater growth in the representation of women than firms in the classical disparity
cluster (the comparison category in the analysis), supporting the positive impact on diversity capital
advantage. Institutional diversity structures were unrelated to the growth of visible minorities or persons
with disabilities employed, however. Regression also indicated that firms in the classical disparity and
institutional clusters did not differ in return on assets, supporting the prediction of no impact on diversity
process advantage.

93
 Table 1

Regressions Predicting Diversity Capital Advantage

DV = Growth Rate of Females DV = Growth Rate of Persons with

Disabilities
Step Predictor First β Second β Final β First β Second β Final β
1 Governed by the EEA .05 .09 .07 .08 .03 .03
Employee Number -.06 -.01 .00 -.06 -.13 -.13
Industry .02 .01 -.01 .04 .06 .06

2 Strategic/ Diversity Structures (Dummy 1) -.18 .06 .27* .29†

3 Institutional Diversity Structures (Dummy 2) .38** .03

Change in R2 at step 1 .01 .01

Change in R2 at step 2 .03 .07*
Change in R2 at step 3 .09** .00
Overall model F 2.10† 1.11
Overall adjusted R2 .06 .01

N 81 71
†p<.10.
*p < .05. **p < .01. ***p < .001.

H3a and H3b predicted that configurational diversity structures would be positively associated
with both diversity capital advantage and diversity process advantage and that the benefits of
configurational structures would be greater than the benefits of either institutional or strategic structures
alone. Regression analyses controlling for firm size and industry sector indicated that firms in the
configurational cluster showed significantly higher growth in the employment of persons with disabilities
than firms in the classical disparity cluster, which supported H3a. The configurational cluster did not
differ from the classical disparity cluster in growth of women or visible minorities, however. Regressions
also indicated that firms in the configurational cluster showed significantly higher return on assets than
firms in the classical disparity cluster, supporting H3b.

Table 2

Regressions Predicting Diversity Process Advantage

DV = ROA 2003
First β Second β Final β
Step Predictor
1 Governed by the EEA .16 .09 .11
Employee Number .03 -.04 -.01
Industry .10 .06 .06

2 Strategic/ Diversity Structures (Dummy 1) .24† .18

3 Institutional Diversity Structures (Dummy 2) -.12

Change in R2 at step 1 .04

Change in R2 at step 2 .04†
Change in R2 at step 3 .01
Overall model F 1.60
Overall adjusted R2 .04

N 83

†p<.10. *p < .05. **p < .01. ***p < .001

94
 Discussion

The current paper introduces the concepts of diversity advantage and diversity management
structures, and examines how the three types of structures are associated with the two types of diversity
advantage. Due to a possibly detrimental effect of diversity, we suggest that diversity advantage can serve
as an intermediate outcome in terms of bridging diversity and organizational performance. The diversity
advantage, consisting of capital and process advantages, therefore can be created through effective
management. As such, the paper contributes to conceptual understanding of the determinants of diversity-
related structures in organizations and how those structures are configured to achieve management’s
motivations for diversity.

We suggest that organizations’ adoption of diversity management structures will reflect their
different motives to diversity. The adoption of strategic diversity structures derives from a motive of
success whereby diversity is treated as a strategic resource. Strategic diversity structures are characterized
as linking diversity to business strategies, to internal training, development and retention, to reward
systems and to an organization’s value creation and adaptation processes.

By comparison, the adoption of institutional diversity structures derives from a motive of threat
avoidance. Organizations adopt institutional diversity structures to avoid negative consequences.
Legislation, professional associations and competitors are all sources of conformity pressures. The
diversity structures influenced by these forces have such characteristics as linking diversity to the “best
practices”, to “institutionalized socialization tactics”, to visible positions, and to a focus on external
recruitment.

Finally, configurational diversity structures are developed to simultaneously address both

motives. Configurational diversity structures show such characteristics as linking diversity to existing
human resource management structures, integrating the “best practices” with other diversity practices,
associating diversity with corporate strategies, balancing organizational similarity and differences of
demographic groups, and distributing diversity across different jobs.

Our findings provide evidence of the existence of both institutional and strategic diversity-related
structures in Canadian firms and linked the presence of these structures to institutional and strategic
motivations to engage with workplace diversity issues. Canadian firms fell into three clusters. The first
cluster, which we labeled the “classical disparity” group following French (2001), showed low levels of
diversity management structures across the 7 dimensions included in our survey. The second cluster,
which we labeled the “institutional” group, showed moderate to high levels of diversity-related structures
in the areas of selection, training and development, and monitoring the effectiveness of staffing for
diversity, activities which are required by case law and employment equity legislation. We labeled the
third cluster “configurational,” because these firms showed the highest levels of diversity structures in all
7 areas, and particularly in the areas of linking diversity to strategic business goals and human resource
planning, recruiting a diverse workforce, providing work-life flexibility and creating an inclusive working
environment. Validating our view that diversity structures reflect either a strategic or an institutional
motivation toward diversity, we found that factors associated with institutional pressures differentiated
the institutional and configurational clusters from the classical disparity cluster, while the inclusion of
HRM in developing business strategy differentiated the configurational cluster but not the institutional
cluster from the classical disparity cluster.

Adopting the RBV, institutional theory and strategic human resource management, we suggested
that the strategic, institutional and configurational diversity structures have different implications for
creating a diversity capital advantage and a diversity process advantage. Specifically, the former refers to
an organization’s superior composition of groups with different demographic and cultural characteristics.

95
 The latter refers to superior problem solving processes in conjunction with superior social
integration and communication processes between diverse groups.

Our findings indicated that as predicted, institutional diversity structures were associated with
diversity capital advantage, specifically, growth in the employment of women, but not with diversity
process advantage, specifically, return on assets. Additionally, configurational diversity structures were
associated with diversity capital advantage in terms of growth in the employment of persons with
disabilities, and with diversity process advantage in terms of return on assets. These findings partially
confirm our prediction that only configurational diversity structures that combine strategic and legitimacy
goals for diversity result in both diversity capital advantage and diversity process advantage. The
findings also suggest that the increase in the employment of women in Canadian firms appears to be the
result of institutional pressures rather than a belief on the part of firms that including more women will
result in improved firm financial performance.

The current study has both conceptual and empirical limitations. Empirically, despite several
contacts by telephone and mail, we were only able to achieve a 17.6% response rate to our survey. The
fact that the survey was not anonymous probably reduced response rate, given that we were asking a set
of detailed questions about the sensitive topic of diversity. The benefit of using a confidential rather than
an anonymous survey format was the ability to match the survey data to archival sources of information
on employment statistics and firm financial performance. Hence, we traded away some generalizability
for improved internal validity because the analyses examining the outcomes of diversity management
structures are not subject to common methods bias. The fact that our outcome measures are assessed at a
later point in time than our independent measures presents an interpretive difficulty, and we will have to
wait until the 2004 - 2005 employment statistics and firm financial performance numbers are available to
conduct more definitive analyses.

Conceptually, the issue of change over time is left unaddressed. The three types of diversity
structures can develop and change over time, and the role that each of the three diversity structures plays
during different diversity development stages is left unaddressed. For instance, institutional diversity
structures might have a stronger positive effect in the earlier diversity development stage, whereas
strategic and configurational diversity structures might have a positive effect in the later stage in the
development of an organization’s diversity initiative. Future conceptual work should adopt a dynamic
view to examine different stages of diversity development, and how the match between diversity
development stages and structures influence organizational performance. Also, future research could
examine the institutionalization process for different types of diversity structures, and how the processes
ultimately influence organizational performance.

96
 References

Arthur, Jeffrey B., “The Link Between Business Strategy and Industrial Relations Systems in American
Steel Minimills,” Industrial and Labor Relations Review, 45(3), (April, 1992), 488-506.
Balser, D. B. Implementing New Employment Law: A Contested Terrain, Dissertation, Cornell
University, available from Dissertation Abstracts International, 1999.
Barney, Jay B., “Firm Resources and Sustained Competitive Advantage,” Journal of Management, 17,
(March, 1991), 99-120.
Barney, Jay B., “Evaluating Firm Strengths and Weaknesses: The Resource-Based View,” Gaining and
Sustaining Competitive Advantage, Boston, MA: Addison-Wesley, J. Barney (Ed.), (2002), 149-192.
Barringer, Melissa W. & Milkovich, George T., “A Theoretical Exploration of the Adoption and Design
of Flexible Benefit Plans: A Case of Human Resource Innovation,” Academy of Management Review,
23(2), (April, 1998), 305-324.
Bowen, David E. & Ostroff, Cheri, “Understanding HRM-Firm Performance Linkages: The Role of the
‘Strength’ of the HRM System,” Academy of Management Review, 29(2), (April, 2004), 203-221.
Boxall, Peter, “Achieving Competitive Advantage through Human Resource Strategy: Towards A Theory
of Industry Dynamics,” Human Resource Management Review, 8(3), (1998), 265-288.
Bretz, Jr., Robert D., Milkovich, George T. & Read, Walter, “The Current State of Performance Appraisal
Research and Practice: Concerns, Directions, and Implications,” Journal of Management, 18(2),
(June,1992), 321-352.
Cox, Taylor H., Cultural Diversity in Organizations: Theory, Research and Practices, San Francisco:
Berrett-Koehler, 1994.
Cox, Taylor H. & Blake, Stacy, “Managing Cultural Diversity: Implications for Organizational
Competitiveness”, Academy of Management Executives, 5(3), (August, 1991), 45-56.
Delery, John E. & Doty, Harold D., “Modes of Theorizing in Strategic Human Resource Management:
Tests of Universalistic, Contingency, and Configurational Performance Predictions,” Academy of
Management Journal, 39(4), (August, 1996), 802-835.
DiMaggio, Paul J. & Powell, Walter W., “The Iron Cage Revisited: Institutional Isomorphism and
Collective Rationality in Organizational Fields,” American Sociological Review, 48(2), (April, 1983),
147-160.
Edelman, Lauren B. “Legal Ambiguity and Symbolic Structures: Organizational Mediation of Civil
Rights Law,” American Journal of Sociology, 97(6), (May 1992), 1531-1576.
Ely, Robin J. & Thomas, David A., “Cultural Diversity at Work: The Effects of Diversity Perspectives on
Work Group Processes and Outcomes,” Administrative Science Quarterly, 46(2), (June, 2001), 229-273.
Eisenhardt, Kathleen M. & Martin, Jeffrey A., “Dynamic Capabilities: What Are They?” Strategic
Management Journal, 21(10/11), (2000), 1105-1121.
Fiske, Susan T., “Stereotyping, Prejudice, and Discrimination,” The handbook of social psychology,
Boston: McGraw-Hill, D. T. Gilbert, Susan T. Fiske, & Gardner Lindzey (Ed.), (1998), 357-411.
French, Erica, “Approaches to Equity Management and Their Relationship to Women in Management,”
British Journal of Management,12, (2001), 267-285.
Goodstein, Jerry D., “Institutional Pressures and Strategic Responsiveness: Employer Involvement in
Work-Family Issues,” Academy of Management Journal, 37(2), (April, 1994), 350-382.

97
Hinings, B. & Greenwood, R., “The Normative Prescription of Organizations,” Institutional Patterns and
Organizations: Culture and Environment, Cambridge, MA: Ballinger, L. G. Zucker (Ed.), (1988), 53-70.
Hoobler Jenny M., “Lip Service to Multiculturalism: Docile Bodies of the Modern Organization,”
Journal of Management Inquiry, 14(1), (March, 2005), 49-56.
Holzer, Harry J. & Neumark, David, “What Does Affirmative Action Do?” Industrial and Labor
Relations Review, 53(2), (Jan., 2000), 240-271.
Iles, Paul & Hayers, Paromjit K., “Managing Diversity in Transnational Project Teams: A Tentative
Model and Case Study,” Journal of Managerial Psychology, 12(2), (1997), 95-117.
Konrad, Alison M. & Linnehan, Frank, “Formalized HRM Structures: Coordinating Equal Employment
Opportunity or Concealing Organizational Practices?” Academy of Management Journal, 38(3), (June,
1995), 787-820.
Lawrence, Thomas B., Winn, Monika I. & Jennings, P. Devereaux, “The Temporal Dynamics of
Institutionalization,” Academy of Management Review, 26(4), (2001), 624-644.
Leck, Joanne D. & Saunders, David M, “Hiring Women: The Effects of Canada’s Employment Equity
Act,” Canadian Public Policy, 18(2), (June, 1992), 203-220.
Leck, Joanne. D., St. Onge, Sylvie, & LaLancette, Isabelle, “Wage Gap Changes among Organizations
Subject to the Employment Equity Act,” Canadian Public Policy, 21(4), (December, 1995), 387-400.
Lee, Kyungmook, & Pennings, Johannes M, “Mimicry and the Market: Adoption of A New
Organizational Form,” Academy of Management Journal, 45(1), (Feb.2002), 144-162.
Lepak, David P., & Snell, Scott A., “The Human Resource Architecture: Toward A Theory of Human
Capital Allocation and Development,” Academy of Management Review, 24(1), (Jan. 1999), 31-48.
Linnehan, Frank, & Konrad, Alison M., “Diluting Diversity: Implications for Intergroup Inequality in
Organizations,” Journal of Management Inquiry, 8(4), (Dec.1999), 399-414.
McMillan-Capehart, Amy, “Heterogeneity or Homogeneity: Socialization Makes the Difference When
Diversity Is at Stake,” Paper presented at the annual meeting of the Academy of Management, New
Orleans, 2004.
Meyer, John. W., & Rowan, B., “Institutionalized Organizations: Formal Structure as Myth and
Ceremony”, American Journal of Sociology, 83, (1977), 340-363.
Meyer, John. W., Scott, W.R., & Deal, T.R. “Institutional and Technical Sources of Organizational
Structure: Explaining the Structure of Educational Organizations”, Organizational environments: Ritual
and rationality, Beverly Hills, CA: Sage, J. W. Meyer, & W.R. Scott (Eds.), 1983, 45-70.
Meyer, John. W., Scott, W. Richard, & Strang, David, “Centralization, Fragmentation, and School
District Complexity,” Administrative Science Quarterly, 32(2), (June, 1987), 186-201.
Moore, Mark E., Parkhouse, Bonnie L., & Konrad, Alison M. “Women in Sport Management: Advancing
the Representation through HRM Structures”, Women in Management Review, (2001), 16(2), 51-61.
Nemeth, C.J., “Minority Dissent as A Stimulant to Group Performance,” Group Process and Productivity,
Newbury Park, CA: Sage, S. Worchel, W.Wood, & J.A. Simpson (Eds.), (1992), 95-111.
Ofori-Dankwa, Joseph C. & Julian Scott D., “Toward Diversity and Similarity Curves: Implications for
Theory, Research and Practice,” Human Relations, 55(2), (2002), 199-224.
Oliver, Christine, “The Institutional Embeddedness of Economic Activity,” Advances in Strategic
Management, Greenwich, Conn.: JAI Press, J. Baum & J. Dutton (Eds.), Vol. 13, (1996), 163-186.

98
Oliver, Christine, “Sustainable Competitive Advantage: Combining Institutional and Resource Based
Views,” Strategic Management Journal, 18(9), (Oct. 1997), 697-713.
Richard, Orlando C., “Racial Diversity, Business Strategy, and Firm Performance: A Resource Based
View,” Academy of Management Journal, 43(2), (Apr. 2000), 164-177.
Richard, Orlando. C., Barnett, Tim, Dwyer, Sean, & Chadwick, Ken, “Cultural Diversity in Management,
Firm Performance, and the Moderating Role of Entrepreneurial Orientation Dimensions,” Academy of
Management Journal, 47(2), (Apr. 2004), 255-266.
Richard, Orlando, C., & Johnson, Nancy B., “Understanding the Impact of Human Resource Diversity
Practices on Firm Performance,” Journal of Managerial Issues, XIII(2), (Summer 2001), 177-195.
Richard, Orlando, C., & Johnson, Nancy B., “High Performance Work Practices and Human Resource
Management Effectiveness: Substitutes or Complements?” Journal of Business Strategies, 21(2): (Fall
2004), 133-148.
Richard, Orlando. C., & Kirby, Susan L., “Organizational Justice and the Justification of Work Force
Diversity Programs”, Journal of Business and Psychology, 14(1), (Fall 1999), 109-118.
Richard, Orlando C., McMillan, Amy, Chadwick, Ken, & Dwyer, Sean, “Employing An Innovation
Strategy in Racially Diverse Workforces: Effects on Firm Performance,” Group & Organization
Management, 28(1), (2003), 107-126.
Richard, Orlando. C., & Murthi, B.P.S., “Does Race Matter within A Multicultural Context: Alternate
Modes of Theorizing and Theory Testing”, Paper presented at the annual meeting of the Academy of
Management New Orleans. 2004.
Robinson, Gail, & Dechant, Kathleen, “Building A Business Case for Diversity”, Academy of
Management Executive, 11(3), (Aug. 1997), 21-31.
Rumelt, R.P., “Toward A Strategic Theory of the Firm”, Competitive strategic management, Englewood
Cliffs: NJ: Prentice-Hall, R.Lamb (Ed.), (1984), 556-570.
Schuler, Randall S. & Jackson, Susan E., “Linking Competitive Strategies with Human Resource
Management Practices”, Academy of Management Executive, 1(3), (Aug. 1987), 207-219.
Scott, W. Richard, “The Adolescence of Institutional Theory,” Administrative Science Quarterly, 32(4),
(Dec.1987), 493-511.
Scott, W. Richard, Institutions and Organizations, CA, Thousand Oaks: Sage, 1995.
Scott, W. Richard, & Meyer, John W., “Environmental Linkages and Organizational Complexity: Public
and Private Schools,” Comparing public and private schools, New York: Fulmer Press, H. M. Levin, &
James, T. (Ed.), (1987), 128-160.
Singh, Jitendra, V., Tucker, David J. & House, Robert J., “Organizational Legitimacy and the Liability of
Newness,” Administrative Science Quarterly, 31(2), (June 1986), 171-193.
Teece, David J., Pisano, Gary, & Shuen, Amy, “Dynamic Capabilities and Strategic Management”,
Strategic Management Journal, 18(7), (1997), 509-533.
Thomas, David A. & Ely, Robin J, “Making Differences Matter: A New Paradigm for Managing
Diversity,” Harvard Business Review, 74(5), (1996), 79-90.
Tolbert, Pamela S. & Zucker, Lynne G., “Institutional Sources of Change in the Formal Structure of
Organizations: The Diffusion of Civil Service Reforms, 1880-1935”, Administrative Science Quarterly,
28(1), (Mar. 1983), 22-39.

99
Tolbert, Pamela S., & Zucker, Lynne G., “Institutionalization of Institutional Theory”, Handbook of
Organizational Studies, London: Sage, S. Clegg, C.Hardy, & W. Nord (Eds.), (1996),175-190.
Ulrich, D. “Using Human Resources for Competitive Advantage,” Making organizations competitive,
Sam Francisco, CA: Jossey-Bass, R. Kilmann, I. Kilmann, & Associates (Eds.), (1991), 129-155.
Waldman, David A. & Bowen, David E., “The Acceptability of 360-degree Appraisals: A Customer-
Supplier Relationship Perspective,” Human Resource Management, 37(2), (Summer 1998), 117-129.
Williams, K., & O'Reilly, C. “The Complexity of Diversity: A Review of Forty Years of Research”,
Research on Managing in Groups and Teams, Greenwich, CT: JAI Press, D. Gruenfeld, & M. Neale,
(Eds.), Vol. 20, (1998), 77-140.
Wright, Patrick M. & McMahan, Gary C., “Theoretical Perspectives for Strategic Human
Resource Management,” Journal of Management,18(2), (June 1992), 295-320.

100
ASAC 2006 Eddy S.W. Ng
Banff, Alberta Business Administration Program
Trent University

Greg J. Sears
Public Service Commission of Canada

WHAT WOMEN AND MINORITIES WANT: WORK VALUES, LABOUR MARKET

CONFIDENCE, AND A PREFERENCE FOR COMMITMENT TO DIVERSITY1

This study explored potential differences in the work values and labour market
perceptions of ethnic minorities and women relative to the majority group. Ethnic
minorities and women reported higher extrinsic, social, and altruistic work values
than the majority group, but lower confidence in their labour market prospects.

Introduction

Many industrialized countries are addressing a shortage of labour force skills by encouraging
immigration of workers from other countries (Weinstein, 2002). This trend is particularly evident in
North America. In Canada, for example, immigrants accounted for 70 percent of workforce growth over
the past decade and are expected to account for all workforce growth by 2011 (Statistics Canada, 2003a).
Nearly one-half (46 percent) of recent immigrants to Canada report non-European descent (Statistics
Canada, 2003b). Among these individuals, immigrants from China represent the most prevalent group,
followed by East Indians, Filipino, and Vietnamese (Statistics Canada, 2003a). This pattern of
immigration is expected to continue, with even greater numbers of ethnic minorities immigrating into
Canada in future. Complementing this immigration trend, a significant proportion of ethnic minorities
(approximately two-thirds) currently residing in Canada are under the age of 35 (c.f. Dheer & Barnard,
1999).

A major challenge facing employers is understanding and connecting with younger-generation

ethnic minorities. If organizations are to recruit effectively and retain individuals from this emerging
segment of the workforce, they must understand the fundamental values and the challenges that ethnic
minorities face in the labour market (McKay & Avery, 2005; Weinstein, 2002). Younger generation
ethnic minorities, and children of ethnic minority immigrants in particular, often have distinct values and
experiences that fit neither the lives of their parents nor those of the mainstream society. As Dheer and
Barnard (1999: 16) state, these individuals often have “… two sets of values, languages, often two
wardrobes, two sets of movie stars and music videos, two kinds of humour, and two lives – one at home,
and one out there.”

One of the myths that persists regarding ethnic minority children is they are highly assimilated
into Western society (Darity, 2005; Levin & Leong, 1973). Acculturation, however, does not imply
assimilation (Hood & Koberg, 1994; Sayegh & Lasry, 1993; Teske & Nelson, 1974). Ethnic minority
children may alternate between the values of mainstream society and those of their parents’ culture. They
may, for example, adhere to the culture learned at home, which in turn guides who they marry, how they
raise their kids, and the languages they speak, while at the same time conforming to mainstream values
when it comes to socializing, education, and work (Dheer & Barnard, 1999). Given the unique challenges
1
We thank D-Code Consulting for making the data available to us.

101
younger-generation ethnic minorities face in society, coupled with our limited understanding of the work
and labour market perceptions of ethnic minorities, in this study we investigate the work values and job
search self-efficacy of ethnic minorities about to enter the labour market and how they compare to those
reported by the majority group.

Drawing on self-determination theory (Deci & Ryan, 1985, 1987; Deci, Connell, & Ryan, 1989;
Ryan & Deci, 2000) and the “model minority” hypothesis (Bell, Harrison, & McLaughlin, 1997; Cheng,
1997; Petersen, 1966; Taylor & Stern, 1997), we propose that individuals holding minority status (i.e.,
ethnic minorities and women) may possess a different pattern of work values than majority group
members. Lyons, Duxbury, & Higgins (2005) defined work values as generalized beliefs about the
desirability of various aspects of work (e.g., pay, autonomy, working conditions) and work-related
outcomes (e.g., accomplishment, fulfillment, prestige). We compare ethnic minorities and majority group
members on four work values, including extrinsic, instrinsic, social, and altruistic work values (Lyons et
al. 2005; Pryor, 1987; Ros, Schwartz, & Surkiss, 1999, Super, 1970). Extrinsic values reflect material
aspects of work, such as pay, benefits, and job security. Intrinsic values pertain to the inherent
psychological and cognitive satisfactions of working, such as interesting work, challenge, variety, and
intellectual stimulation. Social values concern relations with supervisors, coworkers, and others in the
organization. Altruistic values reflect one’s desire to help others a make a contribution to society.

Although research has examined the differences in work values among different age generations
(e.g., Loughlin & Barling, 2001; Lyons et al., 2005; Smola & Sutton, 2002), no research has
systematically explored the patterns of differences likely to exist between ethnic minorities and
mainstream society. Likewise, although the focus on theory development in this study centers on ethnic
minorities, we also deemed it important to examine the work values of women. Very few comparative
studies have been conducted in this domain. Moreover, Hood & Koberg (1994) have argued that women
may share a common experience with other minority groups (e.g., ethnic minorities) in the organizational
acculturation and socialization process.

In addition to potential differences relating to work values, labour market data have consistently
indicated that ethnic minorities and women are less successful in obtaining desired labour market
outcomes (e.g., Hum & Simpson, 2003). What is less studied in this area, however, are the perceptions of
self-efficacy of ethnic minorities and women regarding their labour market performance. In this vein, we
explore the labour market confidence of minority groups and their interests in employers’ commitment to
promoting workforce diversity and social responsibility. While research has established that ethnic
minorities and tend to be less successful in the labour market, it is important to determine whether these
minority groups perceive an inequity in the labour market that may require corrective measures. Given
the increasing numbers of “model minorities” and women in the Canadian workforce (Agocs, 2002;
Conference Board of Canada, 2000a; Statistics Canada, 2003a), recent gains in the labour market
outcomes of ethnic minorities and women (Jain & Lawler, 2004; Simpson, Sturges, Woods, & Altman,
2004), and the relationship likely to exist between labour market self-efficacy and labour market
outcomes, it is critical that we assess the self-efficacy of minority groups and how they compare to the
majority group. In view of these factors and ongoing debate surrounding minority attitudes toward
affirmative action and other formalized organizational diversity policies (Bell et al., 1997; Tougas,
Beaton, & Veilleux, 1991; Tougas & Veilleux, 1989), the link between minority status and preference for
organizations committed to employee diversity and social responsibility was also explored.

102
 Conceptual Background and Theory

Minority Group Disadvantage in the Work Force

Many people choose to immigrate to seek out better opportunities, either for themselves or for
their children. Earlier immigrants to North America (e.g., 1890s to 1920s) tended to be unskilled and
poor, and were forced to start out from the very bottom (Thompson & Weinfield, 1995). Contemporary
immigrants, in contrast, tend to have stronger skill sets and higher levels of education. In 2001, over 80
percent of skilled immigrants to Canada held university degrees, as did 20 percent of family-class
immigrants, and 13 percent of refugees (Conference Board of Canada, 2004a). Despite impressive
educational levels, many immigrants, particularly those of ethnic descent, are underemployed and are not
working in their trained professions. Several examples of minority underemployment abound. Zhou and
Gatewood (2004) cite the example of Filipino immigrants with a college education and transferable job
skills who are limited to work as nannies and caregivers. Similarly, many East Indian immigrants trained
as surgeons and computer programmers can only find work as taxicab drivers (c.f. Zhou & Gatewood,
2000).

Perhaps surprisingly, despite experiencing substantial skill shortages, organizations have not
actively invested in recruiting from the vast pool of skilled immigrant workers. The most frequently cited
reasons for why organizations may be overlooking immigrant workers include: language barriers,
unfamiliarity with “the Canadian way”, a lack of recognition of foreign credentials and work experience,
and disparate performance standards (Conference Board of Canada, 2004b). Furthermore, ethnic
minorities report barriers to employment relating to a number of cultural differences, including their
ethnicity, race, language, accent, or religion (Statistics Canada, 2003b). Due to these employment
barriers and their experience of both direct and indirect forms of discrimination (Harvey, Siu, & Reil,
1999; Hum & Simpson, 2003), many ethnic minorities are struggling from the bottom of the
socioeconomic ladder and are facing difficulty securing acceptable employment.

Cultural Assimilation

Historically, the most common approach for immigrants to gain access to the opportunity
structure of society is through assimilation (Alba & Nee, 1997). Assimilation refers to a process in which
immigrants become full participants in the institutions of the host society and identify completely with
that society (Eisenhardt, 1954). Gordon (1964) conceptualized the classical assimilation process as
consisting of several stages, ranging from cultural, structural, and marital assimilation, to identificational,
attitude-receptional, behavioral-receptional, and civic assimilation. For many economic immigrants, the
process of assimilation becomes a top priority as they strive to become more and more similar to the host
society in life skills, manner, and outlook (Rumbaut, 1997).

Research has shown that assimilation into the dominant or host society facilitates the upward
mobility of immigrants. Daneshvary, Herzog, Hofler, & Schlottman (1992), for example, report that how
well ethnic minorities do in the U.S. labour market hinges on the process and speed of assimilation, and
the “Americanization” of immigrant workers. Likewise, Bratsberg, Ragan, & Nasir (2002) found young,
male immigrants who are naturalized (a process in which an immigrant obtains citizenship in the U.S.)
assimilated better into American society, and attained better jobs and higher wages. This finding may be
explained by the fact that those who are naturalized may demonstrate a greater commitment and ability to
learn the host country’s language, history, and culture, thereby facilitating assimilation. Kossoudji (1989)
also suggested that assimilation results in a rise in occupational status and earnings, stemming from a
greater awareness and knowledge about the local labour market, acquisition of language capabilities, and
acquisition of host country credentials. From the point of view of assimilation theorists, becoming
similar to the dominant group can bring about upward social and economic mobility (Daneshvary et al.,

103
1992; Rumbaut, 1997) while maintaining strong attachments to one’s native culture and language will
negatively affect the assimilation process (Child, 1943; Warner & Srole, 1945). Consistent with this
perspective, ethnic minority immigrants often feel they are expected to abandon their old ways of life and
assimilate into the mainstream through educational attainment, the development of job skills, inter-racial
marriages, and high levels of occupational achievement over successive generations (Zhou, 1997).

The pressure to assimilate and the conditional acceptance of ethnically distinctive groups by the
host society have taken a heavy toll on many children of immigrant parents. Many immigrant parents
place higher expectations on their children to work hard and perform well in school. Kim (1999, c.f.
Zhou & Gatewood, 1997), for example, reported that Korean immigrants have pushed their children
toward prestigious universities to regain the social status the parents had lost in the host society. As a
function of these pressures and witnessing their parents’ struggle to build from modest beginnings,
children of immigrants have tended to become more education and success-driven, demonstrating strong
needs for career achievement (Darity, 1997; Gans, 1992).

The “Model Minority” Hypothesis

Many ethnic minorities, particularly those of Asian descent, have been labeled “model
minorities” (Cheng, 1997). The model minority hypothesis was first advanced in the U.S. by Petersen
(1966) who described Asian Americans2 as success stories relative to other minority groups (e.g.,
African-Americans and Hispanic-Americans). According to the “model minority” hypothesis, Asian
immigrants tend to be too successful to be considered disadvantaged. They typically have high levels of
schooling, high occupational achievement, and high median family income (Cheng, 1997; Kim & Lewis,
1994; Taylor & Stern, 1997). These outcomes may be attributed, in large part, to a “model minority”
value system that emphasizes hard-work and education (Darity, 2005; Rumbaut, 1997).

While the model minority hypothesis has largely been applied in the context of the U.S., we
suggest in this study that a similar “model minority” phenomenon exists in Canada which may influence
the work values and labour market perceptions of younger generation ethnic minorities. Based on the
model minority hypothesis and self-determination theory, we postulate that ethnic minorities in Canada
may place a different emphasis on various work values relative to the majority group.

Self-Determination Theory

Self-determination theory (Deci & Ryan, 1985; 1987; Deci, et al., 1989; Ryan & Deci, 2000)
aims to explain the specific driving forces behind behavioral self-regulation and motivation. The theory
posits that an individual’s motivation will depend upon his or her sense of self-competence and belief that
they have a deliberate choice over their behavior. According to this perspective, the functioning input
affecting the initiation and regulation of intentional behavior can be classified as either autonomous (i.e.,
providing choice), or controlling (i.e., pressuring one to think, feel, or behave in specified ways; Deci et
al., 1989). Autonomy involves acting with a sense of volition and choice, while controlled behaviour
involves acting with a sense of pressure and coercion. Autonomous motivation occurs when individuals
are engaged in the activity because of intrinsic rewards (e.g., I work because it is fun). Controlled
motivation entails expectations of receiving a reward (e.g., I work because I value the reward outcome)
(Gagne & Deci, 2005; Deci, 1971).

2
The “model minority” descriptor commonly applies to immigrants and their descendants from East Asia (e.g.,
China, Japan, Korea) and South Asia (e.g., India, Pakistan, Sri Lanka) (Cheng, 1997)

104
Hypotheses

Drawing on self-determination theory and the model minority hypothesis, we posit that minorities
may display different extrinsic and intrinsic work values. Having directly or vicariously experienced
discrimination (Harvey et al., 1999) and feeling a strong external pressure to succeed (Cheng, 1997;
Kennedy & Schumacher, 2005; Staniec, 2004), minorities (i.e., ethnic minorities and women) may be
expected to place a strong emphasis on extrinsic rewards such as pay, benefits, and job security. Thus,
minorities may be more inclined to regulate their behavior through controlled motivation. High internal
and external expectations to succeed serve to reinforce the control element of the functioning input. In
contrast, having always enjoyed dominant status in society, the majority group (i.e., Whites, males) may
have less salient extrinsic (or “survival”) needs. As a result, they may derive their motivation through a
greater sense of autonomy and will seek to fulfill their intrinsic work needs. Based on this rationale, we
hypothesize:

H1a: Individuals holding minority status in society (i.e., ethnic minorities,

women) will report higher extrinsic work values than those with a more
dominant status (i.e., White, males).

H1b: Individuals holding a more dominant status in society (i.e., White, males)
will report higher intrinsic work values than those with a minority status
(i.e., ethnic minorities, women).

Social capital theory suggests that individuals in structurally advantageous positions can leverage
their social networks to achieve higher performance, faster promotions, and greater career success (Burt
1992; Podolny & Barron, 1997). Research has indicated that women and ethnic minorities consistently
have weaker social networks compared to White males (Ibarra, 1993, 1995). This may be partly due to
ethnic minorities not being seen as knowing “the right types of people” in the organization to be included
in one’s social networks (Brief & Hayes, 1997). Although minorities appear to be less successful in
establishing social networks, a self-determination perspective would suggest that model minorities may
actively seek to improve their social networks to increase their perceptions of autonomy as well as
opportunities for forging successful extrinsic outcomes (e.g., career advancement, higher salaries). Thus,
given the greater challenges minorities may face in establishing social networks in the organization and
the importance of social connections to yielding other desired work outcomes (e.g., career advancement),
we hypothesize that minorities will report a stronger desire to work with good people and develop strong
social connections in the organization.

H1c: Individuals holding a minority status in society (i.e., ethnic minorities,

women) will report higher social work values than those with a more
dominant status (i.e., Whites, males).

We also explore how ethnic minorities and women differ in their altruistic values when compared
to the majority group. Altruistic values are conceived here as one’s identification with employer
commitments to diversity, social responsibility, and work-life balance, and a desire to make a personal
impact in one’s job. In line with self-determination theory, we expect that those who have historically
been in a disadvantaged position will have stronger views regarding the importance of social justice and
equal treatment in organizations. Minorities may feel they have fewer opportunities (i.e., less choice or
autonomy) in organizations, and therefore, favour jobs in which the employer recognizes the importance
of providing better opportunities to disadvantaged groups. Accordingly, we hypothesize:

105
 H1d: Individuals holding a minority status in society (i.e., ethnic minorities,
women) will report higher altruistic work values than those with a more
dominant status (i.e., Whites, males).

Lastly, we are interested in exploring the levels of labour market confidence among women and
ethnic minorities. Examining the confidence or self-efficacy of minority groups regarding their
employment prospects is of particular importance as it helps shed light on whether women and minorities
perceive themselves to be disadvantaged. In light of prior research indicating that minorities have
traditionally been less successful in the labour market (e.g., Hum & Simpson, 2003), are more likely to be
underemployed (e.g., Zhou & Gatewood, 2000), and even “model minorities” have tended to perceive
discrimination and inequity in the workplace (Bell et al., 1997), we posit that minorities will report less
labour market confidence than those in the majority group. Specifically, we hypothesize:

H2: Compared to individuals holding more dominant status in society (i.e.,

Whites, males), individuals with a minority status (i.e., ethnic minorities,
women) will report: less confidence in their ability to find a job they desire,
a greater willingness to accept a less than ideal job, and longer time
expectations for securing full-time employment.

Finally, prior research has suggested that individuals with different levels of ability may possess
different work perceptions (e.g., Ng & Burke, 2005; Trank, Rynes, & Bretz, 2002). Although we offer no
specific hypotheses regarding the role of ability in work values and labour market perceptions, we
investigated the effects of this factor on an exploratory basis. Self-determination theory suggests that
competence (ability) contributes to individuals’ perceptions of control and autonomy, therefore we
deemed it important to include this variable in this study. In particular, we are interested in how this
variable may combine (interact) with minority status in influencing individuals’ work values and
perceptions of labour market confidence.

Method

Sample and Methodology

This study is part of a broader research project commissioned by a consortium of large Canadian
companies interested in better understanding the job, organizational, and career preferences of post-
secondary students. The original data were collected from 20,771 college and university students across
Canada using an internet-based survey. Given our interest in students closest to entering the labour
market, only students who were expecting to graduate and seek their first full-time employment within the
next year were included in the study. Overall, this sample consisted of 5,738 complete cases.

Table 1 shows the demographic characteristics of the sample in this study. A majority of the
students (75.8 percent) were between 20-25 years of age. Thirty-six percent were enrolled in arts and
humanities, 39.1 percent were enrolled in science and engineering, and 25.1 percent were enrolled in
commerce/business studies. Approximately 63 percent of the respondents in the sample were female; 33
percent were ethnic minorities.

Measures

Work Values. Four well-established work values were assessed in this study: extrinsic, intrinsic,
social, and altruistic work values (e.g., Lyons et al., 2005; Pryor, 1987; Ros, Schwartz, & Surkiss, 1999,
Super, 1970). In reference to their upcoming search for full-time employment, respondents were asked to

106
indicate the level of importance of fourteen items designed to measure these four work values. Students
responded to each item using a five-point scale (1 = not at all important, 5 = essential). Extrinsic values
were measured with three items (α = .58): “good initial salary,” “good health and benefits plan,” and “job
security.” Intrinsic values were assessed with four items (α = .64): “opportunities for advancement,”
“challenging work,” “good training opportunities/developing new skills,” and “organization is a leader in
its field.” Social values were measured with two items (α = .85): “good people to work with” and “good
people to report to.” Altruistic values were assessed with four items (α = .70): “strong commitment to
employee diversity,” “commitment to social responsibility,” “work-life balance,” and “opportunities to
make a personal impact.”

Ability. In line with prior work using grade point average (GPA) as a proxy measure for ability
and competence (e.g., Ng & Burke, 2005; Trank, Rynes, & Bretz, 2002), ability was assessed using self-
reported GPA. Respondents were asked to indicate whether their GPA was in the upper first-class (A to
A+), lower first-class (B+ to A-), second-class (D to B), or third-class (F) range. Research has shown that
grades are highly correlated with abilities (c.f. Afarian & Kleiner, 2003). Higher grades represent not
only intellectual ability but also motivation to do high quality work; thus, our ability measure has direct
links to work performance (Howard, 1986). Due to a very limited number of students reporting failing
(F) grades (less than 1%), these cases were excluded from the analysis.

Labour Market Confidence. Three items were developed for this study to measure students’
overall confidence in the labour market for their current/upcoming job search. The three items were: “how
confident are you in finding a job you really want” (“job confidence”;1 = not at all confident to 5 = very
confident), “would you be willing to accept a job that is not ideal but may be a good starting point for
your career” (“2nd Best”; 0 = no; 1 = maybe; 2 = yes), and “when do you expect to find full-time
employment” (“Time to Employment”; 1 = within one year of graduation to 6 = I already have).

Results

The means, standard deviations, and correlations for all the variables in this study are reported in
Table 1. Results for the Multivariate Analyses of Covariance (MANCOVAs) comparing gender, ethnic,
and ability (GPA) groups in terms of work values and labour market confidence are reported in Tables 2
and 3. Students’ program of study was entered as control variable in each of the MANCOVAs due to
significant differences in student work values and labour market confidence as a function of this variable.
For example, arts and humanities students rated altruism more highly than science or business students (F
(2, 4695) = 57.46, p< .001). Arts and humanities students also indicated a greater preference for
organizations with a strong commitment to diversity than science or business students, (F (2, 4715) =
23.83, p< .001), and reported a longer expected duration before obtaining full-time employment (F (2,
4130) = 93.60, p< .001).

107
 Table 1

Means, standard deviations, and zero-order correlations for all variables

M s.d. 1 2 3 4 5 6 7 8 9 10 11 12 13
1 Gender 1.37 .48 -
2 MinorityStat 1.67 .47 -.02 -
3 GPA 2.29 .62 .04** -.09** -
4 Age 4.03 1.21 .11** -.06** -.17** -
5 Program 1.89 .77 .16** -.11** .01 .04** -
6 Extrinsic 11.63 2.07 -.12** -.07** .02 -.03* -.49** -
7 Intrinsic 16.23 1.38 .00 -.09** -.06** .04** .14** .33** -
8 Altruistic 14.92 3.05 -.20** -.14** -.03 .02 -.13** .33** .42** -
9 Social 8.50 1.53 -.10** -.10** -.04** .02 .02 .26** .41** .49** -
10 Diversity 3.43 1.20 -.19** -.26** .04** .02 -.08** .31** .34** .78** .35** -
11 Social Resp 3.49 1.08 -.15** -.04* -.04** .01 -.15** .18** .25** .78** .38** .46** -
12 Confidence 3.44 1.15 .02 .12** -.13** .04** .06** .03 .18** .07** .05** .02 .05** -
13 2nd Best 1.79 .46 -.07** -.03 .03* -.04** -.05** -.01 -.05** -.02 .00 .02 -.02 -.12** -
14 Time to FT 3.43 1.68 .04** -.04* -.07** .14** .21** .02 .15** -.07** .03* -.03* -.08** .28** -.06**
** p < .01 (two-tailed), * p < .05 (two-tailed)

Coding (for dummy/categorical variables):

Gender: 1 = Female, 2 = Male

MinorityStat: 1 = Ethnic minority, 2 = Whites
GPA: 1= Upper first class (A+/A), 2 = Lower first class (A-), 3 = Second class (B/C/D)
Age: 1 = Under 16, 2 = 17-19, 3 = 20-22, 4 = 23-25, 5 = 26-28, 6 = 29-30, 7 = Over 30
Program: 1 = Arts & humanities, 2 = Science & engineering, 3 = Commerce/Business
Confidence (in looking for FT job): 1 = Not at all confident, 5 = Very confident
2nd Best (willing to accept less ideal job): 1 = Yes, 2 = No
Time to FT (employment): 1 = Already have, 6 = Within one year of graduation

108
Work Values

Table 2 reports results of the MANCOVAs relating to the four types of work values: extrinsic,
intrinsic, social, and altruistic.

Table 2

Work values by gender, minority status, GPA; gender x minority status; gender x GPA; and
minority status x GPA

Extrinsic Intrinsic Social Altruism

Mean s.d. Mean s.d. Mean s.d. Mean s.d.
Program
Arts 11.74** 1.95 8.52*** 1.59 15.88*** 2.39 15.45*** 3.11
Science 11.49** 1.92 8.38*** 1.47 16.18*** 2.31 14.43*** 2.98
Commerce 11.51** 1.91 8.60*** 1.62 16.75*** 2.41 14.47*** 2.94

Gender
Female 11.89* 2.00 8.64* 1.50 16.29* 2.29 15.32*** 2.90
Male 11.60* 2.14 8.45* 1.68 16.58* 2.47 14.56*** 3.28

Minority Status
V.M. 12.08*** 2.15 8.76*** 1.54 16.75*** 2.38 15.74*** 5.04
White 11.41*** 2.03 8.32*** 1.48 16.12*** 2.34 14.15*** 3.08

GPA
Upper 1st cl 12.12** 2.08 8.79** 1.54 16.85*** 2.36 15.46** 3.26
Lower 1st cl 11.48** 2.06 8.48** 1.47 16.45*** 2.27 14.66** 3.07
2nd class 11.63** 2.10 8.35** 1.67 16.01*** 2.50 14.71** 3.10

Gender x Minority
Minority Female 12.13 2.12 8.90 1.47 16.53 2.26 15.95 2.88
White Female 11.64 1.94 8.38 1.50 16.04 2.30 14.70 2.89
Minority Male 12.02 2.19 8.63 1.57 16.97 2.54 15.53 3.21
White Male 11.18 2.09 8.26 1.68 16.19 2.40 13.60 3.21

Gender x GPA
Female Upper 12.04 2.00 8.74 1.41 16.36* 2.26 15.76 2.89
Female Lower 11.76 1.96 8.67 1.41 16.41* 2.26 15.14 2.86
Female 2nd Cl 11.87 2.10 8.51 1.55 16.09* 2.36 15.07 2.86
Male Upper 12.20 2.22 8.85 1.69 17.34* 2.59 15.16 3.69
Male Lower 11.20 2.09 8.30 1.55 16.49* 2.30 14.18 3.12
Male 2nd Class 11.40 2.14 8.19 1.68 15.92* 2.58 14.34 3.23

Minority x GPA
Minority Upper 12.81** 2.35 9.17a 1.11 17.35 1.98 16.83** 2.86
Minority Lower 11.61** 2.11 8.59a 1.46 16.68 2.26 15.12** 3.08
Minority 2nd Cl 11.81** 2.15 8.53a 1.60 16.22 2.50 15.25** 2.97
White Upper 11.43** 1.89 8.42a 1.61 16.34 2.42 14.09** 3.14
White Lower 11.34** 2.02 8.38a 1.47 16.22 2.25 14.20** 3.03
White 2nd Class 11.46** 2.07 8.16a 1.69 15.79 2.41 14.16** 3.13
***p<.001, **p<.01, *p<.05, ap<.10

109
 Extrinsic Values. As predicted in Hypothesis 1a, respondents in minority status positions placed
more importance on extrinsic work outcomes (e.g., salaries, benefits, security) relative to minority group
members. Ethnic minorities (F (1, 2685) = 29.01, p< .001) and women (F (1, 2685) = 5.39, p < .05) both
rated extrinsic rewards more highly than Whites and men. Complementing these findings, a significant
interaction emerged (F (2, 2685) = 5.19, p < .01) indicating that minority status interacts with GPA in
influencing extrinsic values. Specifically, ethnic minorities with the highest grades provided significantly
higher ratings (M = 12.81) compared to each of the other groups (M = 11.31 Whites with second-class
grades to M = 11.81 minorities with lower first-class grades). This finding is consistent with both self-
determination theory and the model minority hypothesis, and supports the notion that holding a minority
status in society (mitigated perceptions of control), particularly when coupled with higher levels of ability
(competence), will tend to result in a higher value placed on extrinsic work outcomes.

Intrinsic Values. Ethnic minorities (F (1, 2685) = 20.15, p < .001) and men (F (1, 2685) = 4.37,
p < .05) placed the greatest weight on intrinsic outcomes (opportunities, challenging work, new skills,
leader in its field) relative to Whites and women. A significant interaction also surfaced between gender
and GPA (F (2, 2685) = 5.28, p < .05). Men with the highest grades (M = 17.34) valued intrinsic
outcomes significantly more than others (M = 15.92 males with second-class grades to M = 16.49 males
with lower first-class grades). Overall then, this finding supports Hypothesis 1b relating to women, but
not ethnic minorities. Results lend credence to the notion that holding a dominant position in society with
respect to gender, particularly when coupled with higher ability, will tend to result in a higher weight
placed on intrinsic work values. Contrary to expectations, however, ethnic minorities were more likely
than Whites to value intrinsic outcomes in their work.

Social Values. Consistent with Hypothesis 1c, ethnic minorities (F (1, 2685) = 22.37, p < .001)
and women (F (1, 2685) = 4.26, p < .05) rated social values (e.g., good people to work with and to report
to) more highly than others. Although marginally significant (F (2, 2685) = 2.45, p < .10), there was also
evidence of an interaction between minority status and GPA. Ethnic minorities with the highest GPAs (M
= 9.17) rated social values more highly than others (M = 8.16 Whites with second-class grades; M = 8.59
minorities with lower first-class grades). This may suggest that, despite the majority group’s ability to
capitalize on social networks, minority groups may have a stronger desire to build strong social networks
at work to enhance their social capital.

Altruistic Values. In line with Hypothesis 1d, ethnic minorities (F (2, 2685) = 22.37, p < .001)
and women (F (2, 2685) = 17.54, p < .001) provided higher ratings on altruistic values (i.e., commitment
to diversity and social responsibility, work-life balance, opportunity to make a personal impact) than
others. An interaction was evidenced (F (2, 2685) = 6.38, p < .01), such that ethnic minorities with the
highest grades (M = 16.83) reported the highest ratings on altruism while Whites with higher grades
reported the lowest ratings on this dimension (M = 14.09). Indeed, analyses of two central items of
interest comprising the altruism scale -- commitment to diversity and social responsibility -- mirror this
pattern. Ethnic minorities (F (1, 2685) = 174.70, p < .001 diversity; F (1, 2685) = 21.12, p < .001 social
responsibility) and women (F (2, 2685) = 23.31, p < .001 diversity; F (1, 2685) = 17.35, p < .001 social
responsibility) placed a stronger emphasis on these items than did others. Ethnic minorities with the
highest grades (M = 4.14 diversity; M = 3.86 social responsibility) placed the most weight on diversity
and social responsibility, while Whites with the highest grades (M = 2.79 diversity; M = 3.16 social
responsibility) provided among the lowest ratings on these items. Overall, these findings are consistent
with self-determination theory and suggest that an element of self-interest (e.g., Bell et al., 1997) may
play a key role in individual’s attitudes regarding the importance of an organization’s commitment to
diversity and social responsibility.

110
Labour Market Confidence

Table 3 reports results for the MANCOVAs on the three indicators of labour market confidence:
confidence in finding a job (“confidence”), willing to accept a job that is not ideal (“2nd Best”), and time
expectation for employment (“Time to Empl.”).

Table 3

Labour market confidence by gender, minority status, GPA; gender x minority status;
gender x GPA; minority status x GPA

Job Confidence 2nd Best Time to Empl.

Mean s.d. Mean s.d. Mean s.d.
Program
Arts 3.40 1.11 1.82 .43 3.07*** 1.42
Science 3.43 1.11 1.80 .40 3.40*** 1.53
Commerce 3.58 1.12 1.76 .40 3.96*** 1.67

Gender
Female 3.46 1.13 1.81** .43 3.60 1.64
Male 3.49 1.10 1.73** .50 3.55 1.71

Minority Status
V.M. 3.28*** 1.12 1.76 .46 3.58 1.57
White 3.67*** 1.10 1.78 .46 3.56 1.72

GPA
Upper 1st cl 3.63*** 1.13 1.71* .53 3.83*** 1.80
Lower 1st cl 3.53*** 1.10 1.79* .47 3.59*** 1.69
2nd class 3.27*** 1.12 1.81* .43 3.30*** 1.60

Gender x Minority
Minority Female 3.29 1.16 1.84** .40 3.65 1.52
White Female 3.63 1.11 1.78** .44 3.55 1.69
Minority Male 3.28 1.08 1.68** .52 3.52 1.63
White Male 3.70 1.10 1.77** .46 3.58 1.72

Gender x GPA
Female Upper 3.63 1.15 1.78 .40 3.87 1.74
Female Lower 3.48 1.13 1.83 .42 3.56 1.68
Female 2nd Cl 3.27 1.14 1.83 .39 3.36 1.59
Male Upper 3.64 1.12 1.63 .55 3.79 1.83
Male Lower 3.58 1.11 1.75 .54 3.62 1.75
Male 2nd Class 3.27 1.12 1.79 .48 3.24 1.63

Minority x GPA
Minority Upper 3.35 1.22 1.65 .63 3.81 1.67
Minority Lower 3.36 1.14 1.81 .46 3.64 1.56
Minority 2nd Cl 3.15 1.09 1.82 .42 3.30 1.54
White Upper 3.92 1.06 1.77 .49 3.85 1.84
White Lower 3.69 1.07 1.77 .47 3.54 1.75
White 2nd Class 3.39 1.13 1.81 .43 3.29 1.64
***p<.001, **p<.01, *p<.05

111
 Supporting Hypothesis 2, White and male students reported higher overall labour market
confidence. Whites reported significantly greater confidence in finding employment than ethnic
minorities (F (1, 2384) = 28.96, p < .001). Women expressed a greater willingness than men to accept a
less than ideal job (F (1, 2384) = 8.77, p < .01). Indeed, a significant interaction emerged (F (2, 2384) =
6.98, p < .01) indicating that ethnic minority women (M = 1.84) were most likely to accept a less than
ideal position while ethnic minority men (M = 1.68) were least likely to do the same. Results also
indicated that students with higher grades were more confident in finding employment than those with
lower GPAs (F (1, 2384) = 19.68, p < .001) and expected to find employment faster than those with lower
grades (F (2, 2384) = 19.93, p < .001).

Taken together, our findings regarding the three perceptual measures of labour market confidence
are in line with economic data indicating that ethnic minorities and women experience poorer labour
market outcomes. Consistent with self-determination theory, the results signal that holding a dominant
status in society (e.g., being White, male, and in possession of good grades) tends to result in higher
levels of labour market confidence, while being in a minority position (e.g., being an ethnic minority,
female, possessing lower grades) will tend to be associated with less confidence in finding suitable
employment, and indeed a greater willingness to settle for less than ideal job opportunities.

Discussion

We applied the theory of self-determination and the “model minority” hypothesis in examining
whether differences may exist between minorities (i.e., ethnic minorities and women) and the majority
group (i.e., Whites and men) with respect to their work values and labour market confidence. Consistent
with the self-determination perspective, ethnic minorities and women placed a stronger emphasis on
extrinsic work values (e.g., pay, benefits, job security) than Whites and males, while women appear to
value intrinsic outcomes less than men. We also found, however, that ethnic minorities value intrinsic
outcomes (e.g., greater job challenge, development of new skills) discernibly more than Whites. This
finding is consistent with recent descriptive research indicating that minority employees, as with majority
group members, report strong expectations in terms of receiving fulfilling work, opportunities for
advancement, and assurances that their efforts will pay off (Fortune, 1999). Our result regarding the
comparatively high intrinsic values of ethnic minorities also, contrary to self-determination theory, may
suggest that a positive or reciprocal relationship exists between extrinsic and intrinsic motivational
patterns and work values. Indeed, Alderfer (1972) has suggested that individuals may be motivated by
more than one need at a time (i.e., valuing both extrinsic and intrinsic rewards at the same time). If this is
the case, it is conceivable that one of the means by which minorities attain success in their jobs (extrinsic
outcomes) is via a higher level of intrinsic motivation, and not strictly extrinsic motivation as suggested
by the model minority hypothesis. Future longitudinal research is needed examining this question.

Consistent with self-determination theory, minority status was linked to a greater emphasis on
social values. Although perhaps disadvantaged in their opportunities to build social networks, there is
evidence that ethnic minorities and women are increasingly recognizing the inherent value in networking
and relationship building and the benefits they provide in terms of access to information and resources,
and potential brokerage opportunities. Tung (2002), for example, reported new forms of networks arising
out of ethnic professionals and women’s groups such as “Chinese Professionals Association,” and “new
girls network.” This greater interest in networking, coupled with the tendency for minorities to have a
stronger group (collectivistic) orientation in general (e.g., Bell et al., 1997) may explain their stronger
social values.

112
 Our results also indicated that women and ethnic minorities were more likely to value altruism as
a key work outcome, signalling that those with minority status may be more inclined to empathize with
those in disadvantaged positions. What is particularly noteworthy is the fact that ethnic minorities with
higher levels of ability felt factors such as an organization’s commitment to diversity and social
responsibility are especially desirable while Whites with higher levels of ability found these to be
significantly less attractive work outcomes. This finding suggests that a strong element of self-interest
may be driving both majority and minority group perceptions. The self-interest hypothesis (Bell et al.,
1997) predicts that people’s attitudes will be shaped largely by positive expected outcomes that satisfy
one’s own self-interests. Thus, it appears that minorities feel a personal interest in levelling the playing
field with respect to their societal opportunities. Commensurate with earlier research suggesting minority
group members feel positive actions toward minorities may come at the expense of the majority group
(e.g., Falkenberg & Boland, 1997; Seijts & Jackson, 2001; Son Hing, Bobocel, & Zanna, 2002), findings
from this study suggest that majority group members with higher ability levels are less inclined to favour
altruistic values particularly relating to diversity and social responsibility. We call for more research in
this area. Ng & Burke (2005) have previously reported higher achievers (including men) who were
exposed to diversity messages tend to be more supportive of such programs. More research is needed
investigating contingency variables that influence the relationship between ethnicity and attitudes toward
diversity programs.

With respect to our results regarding labour market perceptions, ethnic minorities and women
reported less confidence regarding their labour market prospects than those with a more dominant societal
status. This finding confirms that the labour market confidence reported by minorities and women largely
mirrors the discrimination they face in the labour market. Interestingly, we found evidence that ethnic
minority women are most likely to accept a job that is not ideal. This finding reinforces the notion that
ethnic minority women may face a particular disadvantage with respect to job opportunities offered in
Canadian society. Barnum, Liden, and DiTomaso (1995), for example, found minority women
experienced a “double jeopardy” owing to their membership in two disadvantaged groups. Likewise,
Elliott and Smith (2004) reported that African-American women to be the most disadvantaged among
women and racial minorities in the U.S. In contrast to ethnic minority women, ethnic minority men were
most reluctant to accept a non-ideal job. This may be explained by the fact that ethnic minority men, who
often see themselves as the family breadwinners, may feel a stronger desire to succeed (i.e., reach a
higher societal status) to “save face” in society (e.g., Yang, 2002). It is this “do or die” attitude that has
led many immigrants to be labelled “model minorities.”

Limitations

One of the limitations of this study is more specific information regarding the ethnic background
of the minority group in the sample was not available. Given our sample is comprised exclusively of
final-year post-secondary students, and ethnic minorities in our sample reported significantly higher
GPAs than the majority group (r = -.08, p < .01), we suggest that the bulk of the minorities in our data set
embody the prototypical model minority (i.e., minorities with higher education levels and skill).
Nevertheless, although a majority of ethnic minorities in Canada are from the Asia Pacific region, many
have diverse backgrounds, education, and socioeconomic status. For instance, immigrants from East Asia
(e.g. Hong Kong, Taiwan) are more likely to immigrate as skilled workers and business immigrants,
while those from Southeast Asia (e.g., Vietnam) are more likely to immigrate as refugees (Conference
Board of Canada, 2004a; Zhou & Gatewood, 2000). Future studies should seek to examine the different
values and levels of labour market self-efficacy of different ethnic subgroups.

In a similar vein, we were unable to ascertain the generation or specific level of acculturation of
ethnic minorities in this sample. Consistent with the classical assimilation process, it is expected that
second and subsequent generations are more likely to have values and attitudes more similar to the

113
dominant group. Ng & Burke (2005), for example, found differences among first generation immigrants
and their preferences for affirmative action. Future research should compare the work values and labour
market confidence of minorities with different levels of acculturation.

Another potential limitation is that GPAs were self-reported by the students and may suffer from
some degree of social desirability bias. While prior research indicates that self-reported GPA closely
approximates objective accounts of GPA (Afarian & Kleiner, 2003; Trank, Rynes, & Bretz, 2002), we
recommend that future studies assessing students’ ability and achievement incorporate multiple measures.

Conclusion

Consistent with self-determination theory and the model minority hypothesis, this study indicates
that ethnic minorities and women report higher extrinsic, social, and altruistic work values than the
majority group. Ethnic minorities and men report higher intrinsic work values than Whites than Whites
and women. This study extends the application of self-determination theory to the study of diversity
issues and demonstrates the utility of this theory in explaining minority – majority group work value
differences. Despite strong work values, findings from this study indicate that ethnic minorities and
women report lower confidence in their labour market prospects relative to the majority group. Both
ethnic minorities and women also rate an organization’s commitment to diversity and social responsibility
as highly important values. While more research is needed probing the effectiveness of diversity
initiatives in organizations and how these initiative may be improved (McKay & Avery, 2005), results
from this research suggest organizations should adopt a more proactive stance in their efforts to recruit
and retain minorities. As part of their recruitment efforts, organizations should appeal to the work values
of minorities in both the design of work in the organization and in their recruitment strategy. Given the
untapped human capital that exists among the minority population, coupled with organizations’ limited
efforts to embrace and leverage these capabilities, recruitment and retention efforts targeted toward
minorities provide a valuable means of forging a distinct competitive advantage in the war for talent.

114
 References

Alba, R. & Nee, V. (1997). Rethinking assimilation theory for a new era of immigration. International
Migration Review, 31(4): 826-874.
Alderfer, C.P. (1972). Existence, Relatedness, and Growth; Human Needs in Organizational Settings.
NY: Free Press.
Afarian, R. & Kleiner, B.H. (2003). The relationship between grades and career success. Management
Research News, 26(2/3/4): 42-51.
Agocs, C. (2002). Canada’s employment equity legislation and policy, 1987-2000. International Journal
of Manpower, 23(3): 256-276.
Barnum, P., Liden, R.C., & DiTomaso, N. (1995). Double jeopardy for women and minorities: Pay
differences with age. Academy of Management Journal, 38(3): 863-901.
Bell, M.P., Harrison, D.A., & McLaughlin, M.E. (1997). Asian American attitudes toward affirmative
action in employment: Implications for the model minority myth. Journal of Applied Behavioral Science,
33(3): 356-377.
Bratsberg, B., Ragan, J.F. Jr., & Nasir, Z.M. (2002). The effect of naturalization on wage growth: A
panel study of young male immigrants. Journal of Labor Economics, 20(3): 568-597.
Brief, A.P. & Hayes, E.L. (1997). The continuing “American Dilemma”: Studying racism in
organizations. Journal of Organizational Behavior, 4: 89-105.
Burt, R.S. (1992). Structural Holes: The Social Structure of Competition. Cambridge, MA: Harvard
University Press.
Cheng, C. (1997). Are Asian American employees a model minority or just a minority? The Journal of
Applied Behavioral Science, 33(3): 277-290.
Child, I.L. (1943). Italian or American? The Second Generation in Conflict. New Haven: Yale
University Press.
Conference Board of Canada (2004a). Performance and Potential 2004-05: How Can Canada Prosper in
Tomorrow’s World? Ottawa.
Conference Board of Canada (2004b). The voices of visible minorities: Speaking out on breaking down
barriers. Ottawa.
Daneshvary, N., Herzog, H.W., Hofler, R.A., & Schlottmann, A.M. (1992). Job search and immigrant
assimilation: An earnings frontier approach. The Review of Economics and Statistics, 74(3): 482-492.
Darity, W. Jr. (2005). Stratification economics: The role of intergroup inequality. Journal of Economics
and Finance, 29(2): 144.153.
Deci, E.L. (1971). Effects of externally mediated rewards on intrinsic motivation. Journal of Personality
and Social Psychology, 18: 105-115.
Deci, E.L., Connell, J.P., & Ryan, R.M. (1989). Self-determination in a work organization. Journal of
Applied Psychology, 74(4): 580-590.
Deci, E.L. & Ryan, R.M. (1985). Intrinsic Motivation and Self-Determination in Human Behavior. NY:
Plenum.
Deci, E.L. & Ryan, R.M. (1987). The support of autonomy and the control of behaviour. Journal of
Personality and Social Psychology, 53: 1024-1037.

115
Dheer, B. & Barnard, R. (1999). Canada’s first generation of visible minorities comes of age. Horizons,
2(3): 16, 23. Ottawa: Policy Research Secretariat.
Eisenhardt, S.N. (1954). The Absorption of Immigrants. London: Routledge & Kegan Paul.
Elliott, J.R. & Smith, R.A. (2004). Race, gender, and workplace power. American Sociological Review,
69(3): 365-386.
Falkenberg, L. E., & Boland, L. (1997). Eliminating barriers to employment equity in the Canadian
workplace. Journal of Business Ethics, 16, 963-975.
Fortune (1999). The 50 Best Companies for Asians, Blacks, and Hispanics, July 19.
Frohlich, N. (1974) Self-interest and altruism, what difference? The Journal of Conflict Resolution,
18(1): 55-73.
Gagne, M. & Deci, E.L. (2005). Self-determination theory and work motivation. Journal of
Organizational Behavior, 26: 331-362.
Gans, H.J. (1992). Second-generation decline: Scenarios for the economic and ethnic futures of the post-
1965 American immigrants. Ethnic and Racial Studies, 15(2): 173-192.
Gordon, M.M. (1964). Assimilation in American: Life the Role of Race Religion and National Origins.
NY: Oxford University Press.
Hall, E.T. (1976). Beyond Culture. Garden City, NY: Anchor Press/Doubleday.
Harvey, E.B., Siu, B., & Reil K.D.V. (1999). Ethnocultural groups, period of immigration and
socioeconomic situation. Canadian Ethnic Studies, XXXI (3): 95-103.
Hood, J.N. & Koberg, C.S. (1994). Patterns of differential assimilation and acculturation for women in
business organizations. Human Relations, 47(2): 159-181.
Howard, A. (1986). College experiences and managerial performance. Journal of Applied Psychology,
71(3): 530-552.
Hum, D. & Simpson, W. (2003). Labour market training of new Canadians and limitations to the
intersectionality framework. Canadian Ethic Studies, XXXV (3): 56-69.
Ibarra, H. (1993). Personal networks of women and minorities in management: conceptual framework.
Academy of Management Review, 18: 422-447.
Ibarra, H. (1995). Race, opportunity, and diversity of social circles in managerial networks. Academy of
Management Journal, 38: 673-703
Jain, H.C. & Lawler, J.L. (2004). Visible minorities under the Canadian Employment Equity Act., 1987-
1999. Relations Industrielles, 59(3): 585-609.
Kennedy, K. & Schumacher, P. (2005). A collaborative project to increase the participation of women
and minorities in higher level mathematical courses. Journal of Education for Business, 80(4): 189-193.
Kim, P.S. & Lewis, G.B. (1994). Asian Americans in the public sector: Success, diversity, and
discrimination. Public Administration Review, 54(3): 285-290.
Kossoudji, S.A. (1989). Immigrant worker assimilation: Is it a labor market phenomenon? The Journal
of Human Resources, 24(3): 494-527.
Levin, J. & Leong, W. (1973). Comparative reference group behavior and assimilation. Phylon, 34(3):
289-294.
Loughlin, C. & Barling, J. (2001). Young workers’ values, attitudes, and behaviours. Journal of
Occupational and Organizational Psychology, 74(4): 543-558.

116
Lyons, S., Duxbury, L. & Higgins, C. (2005). An empirical assessment of generational differences in
work-related values. Paper presented at the 2005 Administrative Sciences Association of Canada
Conference, Toronto, ON.
McKay, P.F. & Avery, D.R. (2005). Warning! Diversity recruitment could backfire. Journal of
Management Inquiry, 14(4): 330-336.
Ng, E.S.W. & Burke, R.J. (2005). Person-organization fit and the war for talent: Does diversity
management make a difference? International Journal of Human Resource Management, 16(7): 1195-
1210.
Perlmann, J. & Waldinger, R. (1997). Second generation decline? Children of immigrants, past and
present – a reconsideration. International Migration Review, 31(4): 893-922.
Perlmann, J. & Waldinger, R. (1998). Are the children of today’s immigrants making it? Public Interest,
132: 73-96.
Petersen, W. (1966). “Success story, Japanese American style.” New York Times, January 6.
Podolny, J.M. & Baron, J.N. (1997). Relationships and resources: Social networks and mobility in the
workplace. American Sociological Review, 62: 673-693.
Pryor, R.G.L. (1987). Differences among differences: In search of general work preference dimensions.
Journal of Applied Psychology, 72: 426.433.
Ragins, B.R. (1997). Diversified mentoring relationships in organizations: A power perspective.
Academy of Management Review, 22(2): 482-521.
Ros, M., Schwartz, S.H., & Surkiss, S. (1999). Basic individual values, work values, and the meaning of
work. Applied Psychology – An International Review, 48: 49-71.
Ryan, R.M. & Deci, E.L. (2000). Self-determination theory and the facilitation of intrinsic motivation,
social development, and well-being. American Psychologist, 55(1): 68-78.
Rumbaut, R.G. (1997). Assimilation and its discontents: Between rhetoric and reality. International
Migration Review, 31(4): 923-960.
Smola, K.W. & Sutton, C.D. (2002). Generational differences: Revisiting generational work values for
the new millennium. Journal of Organizational Behavior, 23(4): 363-382.
Sayegh, L. & Lasry, J.-C. (1993). Immigrants’ adaptation in Canada: Assimilation, acculturation, and
orthogonal identification. Canadian Psychology, 34(1): 98-
Seijts, G.H. & Jackson, S.E. (2001). Reactions to employment equity programs and situational
interviews: A laboratory study. Human Performance, 14(3): 247-265.
Simpson, R., Sturges, J., Woods, A., & Altman, Y. (2004). Career progress and career barriers: Women
MBA graduates in Canada and the UK. Career Development International, 9(4/5): 459-477.
Son Hing, L.S. Bobocel, R.D., & Zanna, M.P. (2002). Meritocracy and opposition to affirmative action:
Making concessions in the face of discrimination. Journal of Social Personality and Social Psychology,
83(3): 493-509.
Statistics Canada (2003a). Canada’s multicultural portrait: The changing mosaic. Minister of Industry,
Ottawa.
Statistics Canada (2003b). Ethnic Diversity Survey: portrait of a multicultural society. Minister of
Industry, Ottawa.
Staniec, J.F.O. (2004). The effects of race, sex, and expected returns on the choice of college major.
Eastern Economic Journal, 30(4): 549-562.

117
Super, D.E. (1970). Work Values Inventory. Boston: Houghton Mifflin.
Taylor, C.R. & Stern, B.B. (1997). Asian-Americans: Television advertising and the “model minority”
stereotype. Journal of Advertising, 26(2): 47-61.
Teske, R.H.C., & Nelson, B.H. (1974). Acculturation and assimilation: A clarification. American
Ethnologist, 1(2): 351-367.
Thompson, J.H. & Weinfeld, M. (1995). Entry and Exit: Canadian Immigration Policy in Context. In
Charles F. Doran and Ellen Reisman Babby (Eds.), Being and Becoming Canada, Special Edition of the
Annals of the American Academy of Political and Social Science. 538: 185-98.
Tougas, F., Beaton, A.M., & Veilleux, F. (1991). Why women approve of affirmative action: The study
of a predictive model. International Journal of Psychology, 26(6): 761-776.
Tougas, F., & Veilleux, F. (1989). Who likes affirmative action: Attitudinal processes among men and
women. In E.A. Blanchard and F. J Crosby (Eds.), Affirmative Action in Perspective. New York:
Spinger Verlag.
Trank, C.Q., Rynes, S.L., & Bretz, R.D. Jr. (2002). Attracting applicants in the war for talent:
Differences in work preferences among high achievers, Journal of Business & Psychology, 16(3): 331-
345.
Tung, R.L. (2002). Building effective networks. Journal of Management Inquiry, 11(2): 94-101.
Warner, W.L. & Srole, L. (1945). The Social Systems of American Ethnic Groups. New Haven: Yale
University Press.
Weinstein, E. (2002). Migration for the benefit of all: Towards a new paradigm for economic
immigration. International Labour Review, 141(3): 225-252.
Yang, S. (2002). Chaeyoun-saving (face saving) due to Korean job loss: Listening to men’s voices.
Journal of Comparative Family Studies, 33(1): 73-95.
Zhou, M. (1997). Segmented assimilation: Issues, controversies, and recent research on the new second
generation. International Migration Review, 31(4): 975-1008.
Zhou, M. & Gatewood, J.V. (2000). Mapping the terrain: Asian American diversity and the challenges of
the twenty-first century. Asian American Policy Review, X: 5-29.

118
ASAC 2006 Stéphane Renaud
Banff, Alberta École de relations industrielles
Université de Montréal

Lucie Morin
Julie Cloutier
École des sciences de la gestion
Université du Québec à Montréal

PARTICIPATION IN VOLUNTARY TRAINING ACTIVITIES

IN THE CANADIAN BANKING INDUSTRY:
DO GENDER AND MANAGERIAL STATUS MATTER?

This study investigates whether gender and managerial status act as significant
correlates of participation in voluntary training. Our theoretical foundation rests
on human capital and systemic discrimination theories. Data come from the
computerized records of a bank’s employees. In short, results show that both
gender and managerial status have a differential impact on participation in
voluntary training: women participate more than men and manager’s
participation is higher than non-managers’ participation. Also, individual
characteristics and productivity-related variables impact differently on
participation by gender and managerial status.

Introduction

In the last decade, due to deregulation, banking in numerous countries has evolved into a
customer-focused industry (Karr, 1994). In the early 1990s, in Canada and elsewhere, governments
reformed all laws and regulations pertaining to financial institutions (Harris and Pigott, 1997). This
deregulation has created more competition from national markets as well as international markets. For
instance, Canadian banks now have to compete with the ING Bank, an international bank that offers
insurance, banking and asset management services in Europe, Asia-Pacific, North and South America.
Furthermore, unlike traditional banks, ING does not have physical branches; it provides its services via
the Internet and other office technologies. Thus, compared to other industries, banking is increasingly
international.

The increased competition, in turn, has forced banks to change from being low-cost producers, to
developing and offering value-added services in various financial areas. Quality of service has become
the ultimate factor differentiating one organization from another and, indeed, determines whether or not it
will survive (Combs and Bourne, 1995). This focus on customer services has had, among other things, the
effect of making the various roles and responsibilities of banking sector workers more complex. This has
created a need for new skills and knowledge or competencies.

In organizations, competencies can be developed through mandatory and voluntary training

activities (Goldstein and Ford, 2002; Noe, 2004; Wexley and Latham, 2002). Mandatory or internal
training refers to employer-provided training or the supply of training whereas voluntary or external
training refers to the demand for training or training that an employee can undertake without the
employer’s approval. Voluntary training is growing rapidly across the globe mainly because the
responsibility for the learning process is increasingly being placed on the individual. In Canada, in

119
response to the new deregulated environment, many banks have opted for a blended training approach.
They developed their internal supply of training and facilitated access to voluntary training by forming
partnerships with colleges and universities as well as the Canadian Bankers Association. However, little is
known about those who participate in these voluntary training activities.

Several economic investigations based on national surveys have focused on the correlates of
participation in mandatory training, including Miller (1994) and Wooden and VandenHeuvel (1997) for
Australia; Booth (1991), Green (1993) and Greenhalgh and Stewart (1987) for Great Britain; Oosterbeek
(1998) for the Netherlands; and Altonji and Spletzer (1991) and Veum (1996) for the USA. In short,
results have consistently shown that younger and more highly educated workers are more likely to receive
training from their employers, as are workers in larger firms. Furthermore, training probabilities differ by
gender, with male workers receiving, in general, more training from their employer than female workers.
With regard to voluntary training, very little empirical literature exists. Findings from a study conducted
by Miller (1994) have indicated that, all things being equal, women undertake more voluntary training
than men. One possible explanation of this effect is that firms have a discriminatory attitude that forces
women to gain competencies through voluntary training.

The aim of this study is to investigate the correlates of participation in voluntary training in the
Canadian banking industry. In our analyses, we choose to differentiate female workers from male workers
because this industry is a major employer of women and new opportunities have developed for female
professionals with the deregulation of financial services. We also choose to differentiate managers from
non-managers because, as Noe (1996, p.123) noted, “in most companies, managers have more
opportunities for salary progression and career advancement than employees in other positions.”
Furthermore, compared to non-managers, managers receive more of the mandatory training expenditures
in both U.S. (Dolezalek, 2004; Galvin, 2002, 2003) and Canadian organizations (Harris-Lalonde, 2001 in
Saks and Haccoun 2004).

Our theoretical foundation rests on human capital theory (Becker, 1962) and systemic
discrimination theory (see Chicha, 1997; Doeringer and Piore, 1971; England, Farkas, Kilbourne and
Dou, 1988). Human capital theory is “a ready-made model of who is likely to demand and to be provided
with training” (Green, 1993, p. 103). It postulates that workers invest in training to increase their
productivity, and thus to enhance their benefits, in particular salaries and promotions. A worker’s decision
to participate in training is therefore based on an analysis of the expected return on investment. And
correlatively, the factors which contribute to increasing training costs or reducing benefits have a negative
impact on participation in training. In short, human capital theory suggests that, like other investments,
training decisions are influenced by economic incentives. Most empirical studies investigating the supply
and demand of training have used this theory.

Our reasoning is also based on systemic discrimination theory which provides an explanation for
employment inequalities between men and women. From this theoretical perspective, the gender gaps
observed are due to discrimination against women in terms of organizational behaviour, practices and
decision-making processes (Doeringer and Piore, 1971). This discrimination is based on the fact that
employers hold stereotypes regarding the gender-based roles of workers. For example, some employers
tend to consider women’s salary to be secondary income since men are perceived to be the main providers
for their family, which gives their work greater prestige and greater value compared to that of women
(Chicha, 1997). Also according to Chicha, employers, in general, tend to perceive women as less
motivated and less attached to their organization because of the time they devote to family
responsibilities, which justifies lower salaries. Similarly, Cobb-Clark and Dunlop (1999) suggest that
employers are generally less inclined to offer promotions to married women with children due to career
interruptions caused by family obligations while married men with children deserve to obtain greater total
compensation (e.g., salaries, training, promotion opportunities) because they are eager to invest more in

120
their career since they have to support their family. In short, the existence of organizational stereotypes
regarding women’s social role accounts for female workers’ low return on investments in schooling, work
experience, organizational tenure and training. Empirical results have shown that, after having controlled
for the variables of human capital, women are promoted less often, earn less than men and the return on
their investments is lower (see Renaud and Vallée, 2005; Christofides and Swidinsky, 1994).

Empirical Strategy

According to human capital theory, a first general equation of voluntary training participation can
be specified as follows:

Ti = ∀0 + ∀1Xi + ∀2Si + ∀3Mi + 0Ti (1)

where Ti represents the training participation for individual i, and is a function of Xi which is a vector of
frequently used control variables that include individual characteristics and productivity-related
explanatory variables (age, schooling, organizational tenure, part-time status and hourly wage rate) for
individual i, Si is gender (1=female), Mi is managerial status, ∀0 is a constant and 0Ti is a random
component of the equation. Age, schooling, organizational tenure, part-time status and hourly wage rate
are all expected to be negatively related to training participation because as they increase, an individual’s
perception of the marginal utility of training should decrease. For gender, women in general receive less
mandatory training from their employer than do men (Booth, 1991; Green, 1993; Miller, 1994).
Researchers have suggested that this difference can be explained by systemic discrimination. As a result,
in a context of mandatory training only, women’s human capital, all things being equal, is less than men’s
human capital. In line with human capital theory, voluntary training can then be seen as a necessary
strategy for women to catch up. Hence, women are expected to participate more in voluntary training than
men and ∀2 should be positive. As for managerial status, managers are expected to have a greater desire
to engage in voluntary training than non-managers because they expect that training increases their
chance of getting promotions, which are more lucrative at the management level than at the non-
management level. Thus, ∀3 is expected to be positive.

Based on systemic discrimination theory, Equation 1 is broken down by gender and takes the
following forms:

TFi = β0 + β1XFi + β3MFi + 0TFi (2)

TMi = γ0 + γ1XMi + γ3MMi + 0TMi (3)

where TFi and TMi represent the training participation for individual female and male i respectively and are
a function of XFi and XMi which are vectors of individual characteristics for individual female and male i,
MFi and MMi are managerial status, β0 and γ0 are constants and 0TFi and 0TMi are the error term of each
equation. Following Equations 2 and 3, managerial status is expected to have a differential impact on
participation in voluntary training based on gender. The difference in training participation between
female managers and female non-managers is expected to be greater than the difference between male
managers and male non-managers. This hypothesis can be explained by the fact that, although women
have made significant progress in the banking industry, it appears that they are still confined to lower
managerial levels (Culpan, Akdag and Cindogvlu, 1992). Furthermore, at a certain level of the
hierarchical structure, women no longer have access to promotions and mandatory training (Renaud and
Vallée, 2005). They thus make up for this disadvantage by participating in voluntary training. Hence, the
category of female managers is the one with the most incentive to engage in voluntary training.
Accordingly, β3 should be greater than γ3.

121
 Equation 1 can also be rewritten by managerial status and then takes the following forms:

TMAi = δ0 + δ1XMAi + δ2SMAi + 0TMAi (4)

TNMAi = ζ0 + ζ1XNMAi + ζ2SNMAi + 0TNMAi (5)

where TMAi and TNMAi represent the training participation for managers and non-managers i respectively
and are a function of XMAi and XNMAi which are vectors of individual characteristics for managers and
non-managers i, SMAi and SNMAi are gender, δ0 and ζ0 are constants and 0TMAi and 0TNMAi are the error term
of each equation. Following Equations 4 and 5, gender is expected to have a differential impact on
participation in voluntary training based on managerial status. The difference in training participation
between female managers and male managers is expected to be smaller than the difference between
female non-managers and male non-managers. As an explanation, one could argue that compared with
male managers, female managers are more likely to use voluntary training in order to enhance their
human capital because of the glass ceiling and its negative impact on access to mandatory training.
Moreover, male non-managers can be divided into two groups who perceive low utility in voluntary
training: part-timers with low attachment to their employers (for example students) and young
professionals on the “management career track” who receive, in general, extensive mandatory training.
On the contrary, female non-managers in banking usually stay longer in non-management positions and
receive very little career-development mandatory training. Thus, if they want to progress, they
necessarily have to engage in voluntary training. In addition, in the banking industry in general, there are
very few male non-managers compared to female non-managers (Renaud and Vallée, 2005). Hence, δ2
should be smaller than ζ2.

To sum up, based on both human capital theory and systemic discrimination theory, it is expected
that female managers should be the ones who have the most interest in voluntary training and male non-
managers the ones who have the least. However, one question still remains: within each category, what
are the productivity-related and individual characteristics that are associated with participation in
voluntary training? To answer this question, Equations 2 to 5 have to be rewritten by gender and
managerial status and take the following form:

TFMAi = η0 + η1XFMAi + 0TFMAi (6)

TFNMAi = θ0 + θ1XFNMAi + 0TFNMAi (7)
TMMAi = ι0 + ι1XMMAi + 0TMMAi (8)
TMNMAi = κ0 + κ1XMNMAi + 0TMNMAi (9)

where TFMAi , TFNMAi , TMMAi and TMNMAi are, respectively, training participation for a female manager,
training participation for a female non-manager, training participation for a male manager and training
participation for a male non-manager.

Data

The data used to test our equations empirically come from a large Canadian bank that has
numerous branches across Canada as well as in several countries. In order to face the challenges
associated with the deregulation of the Canadian financial markets and increasing international
competition, in 1996, the bank established a partnership with Canadian universities to facilitate access to
voluntary training opportunities taken by employees on their personal time. The bank’s involvement was
limited to reimbursing tuition fees and instructional material. This is a clear case of voluntary training
which is on the demand side of training. Qualitative information collected from the employer confirmed
that participation in the program was completely voluntary as the organization’s first objective was to

122
develop employees’ responsibility for their own human capital. Since the beginning of the program, some
employees have actively participated while others have not. Data were retrieved from the computerized
records of the employees for 1998, that is, three years (January 1996 to December 1998) after the
voluntary training partnership was established. Data provide information on 1923 employees,
representing roughly 20 percent of the bank’s total workforce. A random probability sampling procedure
was used by the financial institution to make up this sample.

Training participation is measured using a dummy variable coded 1 if an employee had completed
one voluntary training activity or more from January 1, 1996 to December 31, 1998, and otherwise is
coded 0. This coding implies the use of logistic regression analysis. Gender is measured by a dummy
variable coded 1 for women, and 0 for men. Managerial status is measured by a dummy variable coded 1
for manager, and 0 for non-manager. Individual characteristics and productivity-related explanatory
variables used in this study include age, schooling, organizational tenure, part-time status and hourly
wage rate. Age is measured by three age dummy-variables. These are coded 1 if the employee is aged
between 30-39, 40-49 or over 50, and otherwise are coded 0. The reference group is made up of
employees aged under 30. Schooling is assessed by a set of four educational binary variables. They are
coded 1 if the employee has a community college degree, a university certificate, an undergraduate
degree, a master’s and/or a Ph.D. degree, and otherwise are coded 0. The reference group is composed of
the employees who have a high school education or less. Organizational tenure is measured using a
continuous variable expressed in years. Part-time status is measured by a dummy variable coded 1 for
part-timers, and otherwise is coded 0. Finally, hourly wage rate is measured using a continuous variable
expressed in dollars.

Table 1

Mean sample characteristics of employees

Full sample Women Men Managers Non-managers

N=1923 N=1519 N=404 N=869 N=1054
Training participation 0.20 0.22 0.14 0.25 0.16
Employee is a woman 0.79 0.61 0.94
Employee is a manager 0.45 0.35 0.84
Age 30-39 0.35 0.36 0.32 0.35 0.34
Age 40-49 0.39 0.41 0.33 0.41 0.38
Age 50 and over 0.13 0.13 0.14 0.13 0.13
College 0.17 0.17 0.18 0.15 0.19
University certificate 0.07 0.06 0.13 0.11 0.04
Undergraduate degree 0.16 0.09 0.40 0.30 0.04
Master or more 0.03 0.01 0.08 0.05 0.00
Organizational tenure 13.41 13.93 11.46 14.60 12.43
Part-time status 0.36 0.43 0.07 0.04 0.62
Hourly wage rate 17.55 15.57 24.99 23.43 12.70

Table 1 reports mean sample characteristics of all employees used in subsequent analyses. It also
provides mean by gender and managerial status (women, men, managers, non-managers). This table
reveals that 20 percent of employees participated in voluntary training during the first three years of the
voluntary training program. Our sample includes 79 percent of women and 45 percent of managers. An
examination of the distributions for voluntary training participation reveals differences between women

123
and men and between managers and non-managers: 22 percent participation rate for women compared to
14 percent for men, 25 percent participation rate for managers compared to 16 percent for non-managers.

Empirical Results

Table 2 reports training-equation estimates for the full sample and separate estimates for women
and men, managers and non-managers and related sub-categories. Results from Equation 1 first indicate
that, as predicted, women participate more than men (∀2 = 0.471, p < 0.01). A possible explanation is that
women, because of systemic discrimination in mandatory training, have to engage in voluntary training to
compensate for their lack of competencies. Second, findings show that managers’ participation is higher
than non-managers’ participation (∀3 = 1.017, p < 0.01). This result is in line with the view that returns
on participation are higher for managers than they are for non-managers.

The hypothesis that the difference in training participation between female managers and female
non-managers is greater than the difference between male managers and male non-managers finds
empirical support from Equation 2 and 3 estimates: β3 = 1.108 and is greater than γ3, which equals 0.851.
These results suggest the presence of a glass ceiling in the organization under study. One could argue that
women’s limited access to mandatory training leads them to adopt voluntary training as a parallel
strategy.

Results from Equation 4 and Equation 5 show that gender has a differential impact on
participation in voluntary training based on managerial status: The difference in training participation
between female managers and male managers (δ2 = 0.265) is smaller than the difference between female
non-managers and male non-managers (ζ2 = 0.978). These results suggest that systemic discrimination
against women regarding promotion and training begins to manifest itself at the lower levels of the
organizational structure. These results also support the hypothesis that women use voluntary training to
compensate for the lack of mandatory training to the extent that they believe that they have a chance to
get a promotion.

The purpose of Equations 6 to 9 was to answer the following question: In general and within each
category, what are the individual and productivity-related characteristics associated with participation in
voluntary training? In short, the results indicate different patterns of determinants according to managerial
status and gender1. Age and hourly wage rate are significantly related to participation in voluntary
training and produce more constant effects across the sub-categories. On the other hand, the other
variables under study are related quite differently to participation in voluntary training by sub-category.
Our results are explained in detail below, one variable at a time.

In regard to age, in Equation 1, all things being equal, individuals aged 50 and over participate
significantly less than individuals in all other age categories. This result is in line with human capital
theory. Estimates vary slightly across the four sub-categories under study (Equations 6 to 9). An
examination of significant estimates only shows that age is a stronger predictor of male managers’
participation in voluntary training than of employees in all other sub-categories. Estimates suggest that
male managers aged 50 and over are the ones who see the least benefits in participating in voluntary
training. Also, the difference between age 50 and over estimates from Equation 6 and Equation 8 gives
empirical credence to the presence of systemic discrimination. Up until age 50, women’s participation
rate is higher than that of men. These results support the hypothesis that women view voluntary training
as a strategy that allows them to get around the limited access to mandatory training.

1
The results from Equation 9 should be looked at with caution given the small number of observations.

124
 Table 2

Voluntary training estimates from logistic regressions

Equation 1 Equation 2 Equation 3 Equation 4 Equation 5 Equation 6 Equation 7 Equation 8 Equation 9

Full sample Women Men Managers Non- Female Female non- Male Male non-
managers managers managers managers managers

Age 30-39 0.140 0.110 0.018 0.473 0.025 0.732 - 0.041 - 0.031 2.816
Age 40-49 - 0.015 - 0.041 - 0.498 0.408 - 0.354 0.775 - 0.425 - 0.723 1.093
Age 50 and over - 1.390 ** - 1.269 ** - 3.061 ** - 1.807 ** - 0.972 ** - 1.340 * - 1.026 ** - 3.297 ** - 16.069
College - 0.179 - 0.219 0.134 - 0.340 - 0.040 - 0.431 - 0.071 0.034 1.921
University certificate 0.209 0.295 0.225 0.431 - 0.370 0.534 * - 0.213 0.461 - 21.000
Undergraduate degree - 0.639 ** - 1.079 ** 0.208 - 0.353 - 0.843 - 0.830 ** - 0.746 0.280 - 19.869
Master or more - 2.318 ** - 20.123 - 0.615 - 1.853 * - 19.365 - 19.721 - 19.819 - 0.512 - 14.362
Organizational tenure 0.046 ** 0.031 ** 0.126 ** 0.075 ** 0.007 0.050 ** 0.005 0.134 ** 0.238
Part-time status - 0.173 - 0.168 - 1.108 - 0.871 * 0.114 - 0.920 * 0.172 - 19.039 - 4.374 *
Hourly wage rate - 0.029 * - 0.032 - 0.054 * - 0.055 ** 0.119 * - 0.057 ** 0.141 * - 0.054 * - 1.852 *
Employee is a woman 0.471 ** 0.265 0.978 *
Employee is a manager 1.017 ** 1.108 ** 0.851

Constant - 2.147 ** - 1.391 ** - 2.537 ** - 1.224 ** - 4.018 ** - 0.723 - 3.278 ** - 1.728 ** 17.682 *

NagerKerke R2 0.136 ** 0.127 ** 0.210 ** 0.245 ** 0.036 ** 0.227 ** 0.028 * 0.212 ** 0.483

N 1918 1516 402 868 1050 531 985 337 65

* : p < 0.05; ** : p < 0.01 (one-tailed tests)

125
 As regards schooling, results from Equation 1 offer moderate support for our general hypothesis:
Only the undergraduate degree level and higher are significant and negative. Moreover, the results of
Equations 6 to 9 indicate that this variable is significantly associated with female managers only. In other
words, female managers who hold a master’s degree or higher are significantly less inclined to participate
in voluntary training than female managers with an undergraduate degree, and those who hold an
undergraduate degree are significantly less inclined to undertake training than those who hold a university
certificate. These results are in line with human capital theory which predicts that interest in training
decreases as individuals increase their capital through education. Moreover, systemic discrimination
theory is also a reasonable explanation because the link is not found in men. Therefore, it is plausible that,
regardless of male managers’ level of schooling, the firm offers training to them so that they will have the
required competencies to advance in the hierarchical structure. On the other hand, women’s limited access
to mandatory training encourages those who hold a university certificate or an undergraduate degree to
continue their training on a voluntary basis in order to compensate for the lack of mandatory training.

Contrary to the general hypothesis put forward, organizational tenure is positively associated with
the probability of participating in voluntary training, except for non-managers. This result could be
explained by the compensatory nature of voluntary training. According to human capital theory, the
greater organizational tenure individuals have, the lower the marginal utility of training activities given
the remaining number of years they have to work. Thus, individuals will be less and less inclined to
undertake training in order to enhance their human capital. Similarly, beyond a certain number of years,
the firm is increasingly reluctant to offer training to them. Since training demand follows the same curve
as training supply, the demand for compensatory training (voluntary training) remains steady regardless
of the number of years of service in the firm.

Organizational tenure produces also a stronger effect on male managers than on female managers.
A plausible explanation is that, through the career of managers, the firm’s training supply (mandatory
training) decreases more rapidly than the managers’ training demand (for a lateral movement, promotions,
or a career change, for example). Therefore, the greater the managers’ organizational tenure in the firm,
the less their training needs are met by the firm, and the more likely they are to resort to voluntary
training. However, because of women’s limited access to mandatory training, the gap between training
supply and training demand is greater for women. Given the joint effect of systemic discrimination and
gradual decrease in training supply throughout the years of service, the probability of resorting to
voluntary training increases more considerably for men since women already use voluntary training as a
parallel strategy from the beginning of their career.

The variable of part-time status is only significantly associated with participation in Equations 6
and 9, that is, for female managers and male non-managers. The negative link is in line with our
prediction based on human capital theory. A negative effect was also found for male non-managers.
However, the results show that part-time status is not linked with the participation of female non-
managers. The explanation could be that women do not work part-time out of preference and therefore
part-time status cannot be viewed as an indicator of women’s attachment to the firm.

Lastly, for the variable hourly wage rate, the results of Equation 1 indicate a negative relation
with participation in voluntary training, as suggested by human capital theory. Moreover, all things being
equal, the higher the salary the individuals earn, the less they participate in voluntary training. This
negative effect is more marked for male non-managers than for male and female managers. Thus, it seems
that the more individuals increase their salaries – for example, through promotions-, the higher their
marginal productivity for the firm, which in return probably encourages the firm to invest in training these
individuals. Consequently, individuals are less interested in pursuing voluntary training. On the other
hand, the hourly wage rate exerts a positive influence on female non-managers’ participation in voluntary
training. These results suggest that as female non-managers are promoted to higher-level positions, they

126
gradually see the opportunity to gain access to a managerial position. To this end, they tend to enhance
their human capital through voluntary training since they are generally disadvantaged in the context of
mandatory training. From a different perspective, it is also plausible that, given the decreasing
opportunities to get a promotion because of discrimination, female non-managers decide to invest in
voluntary training in order to enhance their value on the market and their inter-organizational mobility.

Conclusion

A review of the literature on voluntary training in organizational settings reveals that very little
knowledge has been acquired on the subject, in particular as regards the characteristics of the participants
in this type of training. Our study aimed at identifying the correlates of participation in voluntary training
based on gender and managerial status, and therefore we examined four sub-categories: female managers,
male managers, female non-managers and male non-managers. Our goal was to better understand the
motivations of these different categories of employees to participate in this type of training.

The results showed that the probability of participating in voluntary training varies according to
gender and managerial status. This probability is explained in particular by the differential effect
produced by the individuals’ productivity-related characteristics (age, schooling, organizational status,
part-time status, and hourly wage rate) according to gender and managerial status. In line with human
capital theory, it can be suggested that male managers, female managers, male non-managers and female
non-managers participate in voluntary training in different proportions not only because of their different
characteristics but also because of their different assessments of the marginal utility of voluntary training.
These results point to the importance of pursuing research on the determinants of voluntary training by
examining separately the four sub-categories of employees.

The positive effect of managerial status as well as the negative effect of age, hourly wage rate and
part-time status are in line with the predictions of human capital theory. It thus seems that individuals
participate in voluntary training in order to increase their productivity, and consequently, their benefits
(e.g., promotions, salaries). Similarly, the significant effect generally shown by the level of schooling
suggests that the marginal utility of voluntary training decreases with the level of schooling. In other
words, it seems that the higher the workers’ level of schooling, the smaller the gap between the
individuals’ training needs (i.e. total training demand) and training supply (mandatory training).

Moreover, the results highlight the differential gender-based effect on voluntary participation
produced by the variables, i.e. managerial status, age, hourly wage rate, schooling and organizational
tenure. These differential effects support systemic discrimination theory. Although the banking industry is
predominantly female, the results suggest that systemic discrimination against women regarding
promotion exists at the lower levels of the hierarchical structure, coupled with a glass ceiling phenomenon
at the managerial levels. This discrimination has the effect of limiting women’s access to mandatory
training. They thus consider voluntary training as a parallel strategy to compensate for the lack of
mandatory training, as long as they believe that voluntary training has a marginal utility. Therefore,
women tend to undertake voluntary training when they believe that this training will help them to get a
promotion, or when they think that it will help them to enhance their inter-organizational mobility and
thus increase their salaries.

As regards the validity of our model, the variables selected explain, in proportions of 13 percent
to 25 percent, the individuals’ decision whether or not to participate in voluntary training. These variables
are more useful in explaining the participation of managers (25 percent) than of non-managers (4
percent). Our model’s variables also help to better account for men’s decisions to participate (21 percent)
than women’s decision to participate (13 percent). This suggests that future research should refine the

127
explanatory model and include other variables such as participation in mandatory training, hierarchical
level, and perceived opportunities for promotion. Future research could also seek to answer the following
questions: To what extent does voluntary training compensate for the lack of mandatory training? To
what extent do women perceive that they are being discriminated against as regards promotion? To what
extent is voluntary training used to get a promotion or to enhance organizational mobility? What are the
constraints related to voluntary training? Are these constraints more prevalent among women or men,
among managers or non-managers, and so on?

128
 References

Altonji, J.G., and J.R. Spletzer. 1991. Worker characteristics, job characteristics, and the receipt of on-the
job training. Industrial and Labor Relations Review, 45(1), 58-80.
Becker, G. 1962. Investment in human capital: a theoretical analysis. Journal of Political Economy, 70, 9-
49.
Booth, A.L. 1991. Job-related formal training: who receives it and what is it worth?. Oxford Bulletin of
Economics and statistics, 53(3) 281-295.
Chicha, M.T. (1997). L’équité salariale: Mise en oeuvre et enjeux. Cowansville, Québec : Éditions Yvon
Blais Inc.
Christofides, L.N.and Swidinsky, R. (1994). Wage determination by gender and visible minority status:
Evidence from the 1989 LMAS. Canadian Public Policy – Analyse de politiques, vol. 20(1), 34-51.
Cobb-Clark, D.and Dunlop, Y. (1999). The role of gender in job promotions. Monthly Labor Review, vol.
122, 32-38.
Combs, H.W., and Bourne, S. G. (1995). Quality in Business : Preparing Retail Banking for a
Competitive Environment. Review of Business, 17 (1), 3-6.
Culpan, O., Akdag, F., & Cindogvlu, D. (1992). Women in Banking: A Comparative Perspective on the
Integration Myth. International Journal of Manpower, vol. 13 (1), 33-40.
Doeringer, P. and Piore, M.J. (1971). Internal labor markets and manpower analysis. Lexington, D.C:
Heath.
Dolezalek, H. (2004). Training magazine’s annual comprehensive analysis of employer-sponsored
training in the United States. Training, vol. 41 (10), 20-36
England, P., Farkas, G., Kilbourne, B.S. and Dou, T. (1988). Explaining occupational sex segregation and
wages: Findings from a model with fix effects. American Sociological Review, vol. 53, 544-558.
Galvin, T. 2002. 2002 Industry Report: Training magazine’s 21rd annual comprehensive analysis of
employer-sponsored training in the United States. Training, 39 (10), 24-73.
Galvin, T. 2003. 2003 Industry Report: Training magazine’s 22rd annual comprehensive analysis of
employer-sponsored training in the United States. Training, 40(10), 21-36.
Goldstein, I.L., and Ford, K.J. (2002). Training in organizations (4th ed). Belmont, CA: Wadsworth-
Thomson Learning.
Green, F. 1991. Sex discrimination in job-related training. British Journal of Industrial Relations, 29(2),
295-305.
Green, F. 1993. The determinants of training of male and female employees in Britain. Oxford Bulletin of
Economics and Statistics, 55(1) 103-123.
Greenhalgh C. et M. Stewart. 1987. The effects and determinants of training. Oxford Bulletin of
Economics and Statistics, 49(2), 171-191.
Harris, S.L. and C.A. Pigott. 1997. “A Changed Landscape for Financial Services.” The OECD Observer.
No. 206, June/July, www1.oecd.org/publications/observer/206/028-031a.pdf .
Karr, J. (1994). The customer is at the Center of Retail Revolution. American Banker, 159 (October
issue), p. 16.

129
Miller. P.W. 1994. Gender discrimination in training: an Australian perspectiv. British Journal of
Industrial Relations, 32(4), 539-565.
Noe, R.A. 1996. Is career management related to employee development and performance? Journal of
Organizational Behavior, 17, 119-133.
Oosterbeek, H. 1998. Unravelling supply and demand factors in work-related training. Oxford Economic
Papers, 50(2), 266-284.
Renaud, S. and Vallée, M-C. (2005). Les facteurs associés à la mobilité ascendante selon le sexe dans le
secteur bancaire au Canada. Regards sur le travail, vol.1(3), 15-35.
Saks, A.M., and Haccoun, R.R. (2004). Managing performance through training and development (3ème
ed.). Scarborough, Ontario: Nelson Thomson Learning.
Veum, J. R. 1996. Gender and race differences in company trainging. Industrial Relations, 35(1), 13-45.
Wexley, K. N. et Latham, G. P. 2002. Developing and training human resources in Organizations 3rd ed..
Upper Saddle River, NJ: Prentice Hall.
Wooden, M. A. et VandenHeuvel. 1997. Gender discrimination in training: a note. British Journal of
Industrial Relations, 35(4), 627-633.

130
ASAC 2006 Sid Ahmed Soussi
Banff, Alberta Annie Côté
Université du Québec en Outaouais
Gatineau (Qc), Canada

LA DIVERSITÉ CULTURELLE DANS LES ORGANISATIONS : ANALYSE CRITIQUE DES

FONDEMENTS THÉORIQUES DU MANAGEMENT INTERCULTUREL

Cette communication pose la question du recours aux pratiques de

management interculturel par les organisations dans un contexte de
diversité culturelle. Elle rend compte de trois grands modèles
d'intégration de la diversité culturelle issus du management interculturel.
N’étant pas institutionnalisées par les politiques RH, les pratiques de
«gestion interculturelle» se présentent comme de véritables défis au sein
des organisations.

La diversité culturelle : un enjeu récurrent

La question de la diversité culturelle dans les organisations se pose de façon récurrente en raison
de la diversification croissante, sur les plans identitaire et ethnoculturel, des milieux de travail (Chevrier,
2003; d'Iribarne et al., 2002). Les transformations récentes du travail ont fait l'objet de nombreux travaux
convergeant vers le constat d'une profonde réorganisation de la société industrielle (Boltanski et
Chiapello, 2004; Castells, 2001), notamment par la contraction irréversible de son secteur secondaire au
profit d'une «économie nouvelle» (Sériyex, 2000) et d'une «économie du savoir» (Castells, 2001). Au-
delà de leurs dynamiques combinées, les flux migratoires (Helly, 2000), l'émergence de l'économie du
savoir (Viprey, 2003) et le recours croissant à une main-d'œuvre hautement qualifiée ont fortement
contribué à ces transformations. Les frontières culturelles autrefois bien délimitées deviennent de plus en
plus floues (Steinmueller, 2002). On assiste à une transformation des structures organisationnelles avec,
entre autres, l'émergence du fonctionnement en réseau dans lequel intervient davantage une composition
diversifiée de travailleurs sur le plan culturel. L'objectif fondamental de ces communautés de
connaissance «est la production et la circulation de savoirs nouveaux et qui interconnectent des personnes
appartenant à des entités différentes, voire rivales» (David et Foray, 2002:13).

Pour les organisations, c'est un défi de taille qui s'impose en matière de gestion et d'intégration de
personnels culturellement hétérogènes. Le défi se fait d'autant plus grand que les structures mêmes du
système de relations industrielles post-fordiste subissent une mutation des acteurs en relation avec la
montée des enjeux identitaires. Les besoins de main d'œuvre qualifiée forcent les entreprises à repenser
leurs stratégies pour attirer et conserver les travailleurs issus de la diversité culturelle.

Ce travail propose une synthèse critique du corpus d’études produit par la recherche en
management interculturel et présente une revue des principaux modèles d’analyse.

À l'inverse de la main-d'œuvre qui fait un déplacement surtout vers l'intérieur, les flux de capitaux
et de marchandises ont une croissance essentiellement tournée vers l'extérieur. La mobilité des entreprises
à l'échelle internationale suppose la gestion d'un personnel réparti dans plusieurs pays, ce qui impose à
l'organisation une relation exogène avec la diversité culturelle, d'où la nécessité de nouvelles pratiques de
gestion et d'intégration de la main-d'œuvre (Lainé, 2004).

130
 La conjugaison des phénomènes des flux migratoires et de la mondialisation sous-tend
diversification de la main-d'œuvre (Helly, 2000). Le premier renvoie surtout à un mouvement endogène
alors que le second a une portée exogène. Les flux migratoires amènent des travailleurs issus de cultures
différentes au sein des entreprises locales et la mondialisation tend à déplacer les entreprises vers de
nouveaux espaces culturellement distincts. Ce dernier cas a été largement documenté par les diverses
études du management interculturel qui se sont intéressées au caractère exogène de la diversité culturelle
(Trompenaars et Hampden-Turner, 2004; Lainé, 2004; Chevrier, 2003; Adler, 1991; Hofstede 1980).
Toutefois, le corpus théorique du management interculturel a nettement écarté l’élément endogène alors
même que la question culturelle et identitaire se pose à travers les deux problématiques. Par ailleurs, ces
études sont fondamentalement tournées vers des modèles instrumentaux visant la performance
organisationnelle.

Sur le plan empirique, la gestion de la diversité culturelle échappe à toute formalisation dans les
organisations. Ces modèles (du management interculturel) sont rarement intégrés dans les fonctions
ressources humaines des entreprises. Les organisations procèdent le plus souvent par improvisation et au
cas par cas (Lainé, 2004).

Au-delà de l'intégration dans le milieu de travail se pose aussi le problème de l'intégration sociale
des individus (Drudi, 2003). Ils sont confrontés dans leur espace d'accueil à des transformations radicale
au niveau identitaire (Cardu et Sanschagrin, 2002) conjuguées à un univers de références inhabituel
(Cardu et Sanschagrin, 2002; Drudi, 2003).

Sur le plan méthodologique, la question de la gestion de la diversité culturelle a surtout fait l'objet
d'études quantitatives (Trompenaars et Hampdon-Turner, 2004; Hofstede, 1980). Peu d'études qualitatives
ont été consacrées à un regard critique dans la perspective des transformations récentes et de la croissance
de la diversité au travail. Par ailleurs, l'intérêt des études du management interculturel privilégie
résolument une approche instrumentale souvent associée à des styles de gestion traitant de l’implantation
de cadres occidentaux à l’étranger (aspect exogène) (Trompenaars et Hampdon-Turner, 2004; Lainé,
2004; Adler, 1991; Hofstede, 1980). Ces études construisent des repères pragmatiques permettant
d'intégrer la dimension culturelle et d'éviter les impairs reliés aux incompréhensions et aux conflits
potentiels dans les organisations. Le défi de ces études est de sensibiliser de manière concrète les
managers aux difficultés que dresse l’interculturel en vue d'une meilleure efficacité dans leurs fonctions.
En revanche, ces études n'offrent pas de modèle de gestion ancré dans les rapports au travail entre acteurs
porteurs d'identités et de cultures différentes. Bien que les enquêtes aient donné lieu à un contenu riche en
données quantitatives (Tromprenaars et Hampdon-Turner, 2004, Hofstede, 1980), une lacune demeure en
termes de résultats qualitatifs opérants dans un environnement endogène.

Ce travail propose une synthèse critique du corpus d’études produit par la recherche en
management interculturel et présente une revue des principaux modèles d’analyse. Nous présentons ici les
différentes écoles de pensée qui caractérisent ce corpus à travers l’analyse de trois grands modèles. Un
modèle des références nationales dominant tourné vers le déterminisme des cultures nationales sur les
stratégies des organisations et fondé sur des critères nationaux (Trompenaars et Hampdon-Turner, 2004;
Lainé, 2004; Hofstede, 1994, 1980; Adler, 1991; Hall, 1984, 1979, 1971). Un deuxième modèle,
mobilisant des critères sociétaux et construit autour des identités locales et de leurs impacts sur certaines
fonctions des organisations (organisation et gestion des relations du travail) dans des sociétés
culturellement distinctes (d'Iribarne et al., 2002). Quelques études critiques récentes de ces deux modèles
tendent à former ce qui apparaît comme un troisième modèle en émergence qui souligne la nécessité d'une
approche anthropologique de la diversité culturelle au travail. Ce modèle se pose en alternative aux
courants théoriques nationaux et sociétaux.

131
 Le modèle des références nationales

Le courant regroupant les modèles nationaux d'analyse s’inscrit dans une perspective
épistémologique fonctionnaliste traditionnelle et domine à ce titre la littérature sur le management
interculturel. Avec leurs travaux fondateurs, Hall (1979) et Hofstede (2002, 1991, 1980) en apparaissent
comme les précurseurs. Plus récemment, les travaux de transfert de Trompenaars et Hampden-Turner
(2004) ont renforcé la visibilité de ce courant. À ceux-ci, s'ajoutent plusieurs autres travaux fondés sur les
mêmes postulats (Lainé, 2004; Adler, 1991; Usunier, 1991, 1985).

Postulats et définition de la culture comme un système de communication

À partir de travaux réalisés sur les relations entre les indiens Hopis et Navajos et leurs
environnements institutionnels et culturels dans les années 1930, l'anthropologue américain Hall (1971)
propose une définition de la culture qui renvoie à l'analyse des idéaux-types «profondément enracinés
dans le biologique et le physiologique» (Hall, 1971: 16). À la façon de l'ethnologie classique, c'est par
l'observation des comportements qu'il édifie une conception de la culture qui se veut déterministe.
L'homme ne peut se soustraire à sa propre culture puisque celle-ci s'inscrit jusque dans son système
nerveux et influence la perception de son environnement (Hall, 1971). C'est à travers la culture que sont
dictés les comportements des individus. C'est par l'étude de la communication1, c'est-à-dire du «système
de création, d'émission, de rétention et de traitement de l'information» (Chevrier 2003: 42), que Hall
élabore sa définition de la culture. Il divise la communication en deux parties : les mots (verbal) et le
comportement (non-verbal) (Hall, 1984). Pour décoder les messages exprimés par ces deux formes de
communication, il propose trois dimensions "cachées" : le contexte, le temps et l'espace.

Contexte riche versus contexte pauvre. Le contexte donne un sens à la communication et agit
sur celle-ci de manière inconsciente et invisible (Hall, 1984). Pour évaluer le contexte d'une
communication, il faut déterminer la quantité d'informations échangées par les interlocuteurs. Lorsque
l'information transmise est implicite et exige une compréhension approfondie du contexte pour saisir le
sens de la communication, le contexte est "riche"2. À l'inverse, un contexte est "pauvre" lorsqu'il y a une
transmission détaillée de l'information qui demande peu de connaissance du contexte3. La signification, le
contexte et l'information sont donc reliés à travers un équilibre dynamique. Ainsi, «un apport
d'information doit compenser une perte de contexte pour que la signification reste constante» (Hall, 1984:
75-76).

Pour Hall (1984), les stratégies de communication des Américains et des Japonais se réfèrent à
des contextes différents. Les Américains (pauvres en contexte) sont dérangés par la manière subtile des
Japonais (riches en contexte) d'en venir là où ils veulent alors que ces derniers s'imaginent à tort que leurs
interlocuteurs devinent leur raisonnement. D'un autre côté, les Japonais voient les explications logiques et
précises des Américains comme une façon «de pénétrer leur esprit et de penser à leur place» (Hall, 1984:
78).

1
Voir Usunier (1985) qui établit un parallèle entre la communication interculturelle et le management international.

2
Par exemple, un groupe d'ingénieurs d'une même entreprise peuvent utiliser un bon nombre de sous-entendus au cours d'une
discussion sans créer d'ambiguïté car ceux-ci n'ont besoin que de peu d'explications pour comprendre ce dont il est question.

3
C'est le cas de la cour de justice où chacun des faits doit être clairement précisé et méticuleusement organisé.

132
 Temps monochrone versus temps polychrone. La culture repose sur une conception temporelle
de l'activité et de l'expérience. Celle-ci se reflète par la manière dont les individus les planifient, les
organisent et en parlent. Selon Hall (1984), les sociétés sont orientées vers deux types de conception du
temps qui constituent chacune l'extrémité d'un continuum sur lequel il est possible de situer les cultures:
le temps monochrone et le temps polychrone. Le temps monochrone est représenté comme une ligne
droite sur laquelle s'inscrivent le passé, le présent et le futur. Les individus exécutent leurs tâches
méthodiquement comme s'il y avait un découpage entre chacune d'elles (Hall, 1984, 1979). Dans
l'organisation, cela peut se traduire par une rigidité des horaires, une division du travail et des tâches. La
perception polychrone du temps peut s'illustrer symboliquement par une forme circulaire. Les individus
sont engagés dans plusieurs relations et activités sans programme formellement établi4.

Proxémique culturelle. La distance physique entre les individus apparaît comme un mode de
communication forgé par la culture. La distance interpersonnelle "convenable" est différente selon les
sociétés (Hall, 1971). Tout comme la segmentation des activités, l'espace physique est compartimenté.
Tout le modèle de Hall est fondé sur les marges de manœuvre reliant les trois dimensions duelles
précédentes.

La culture: une programmation mentale

Suite aux travaux de Hall (1979, 1971), Hofstede (1980) s'engage dans une voie alternative
fondée sur une convergence supposée des cultures. À partir d'une recherche quantitative sur les valeurs
personnelles en lien avec le contexte du travail, et prenant comme échantillons les personnels des
nombreuses filiales IBM dans le monde, il élabore un modèle d'analyse construit à partir de sept
dimensions duelles à portée universelle. Pour lui la culture est une sorte de programmation mentale de
l'ensemble des façons de penser, de réagir ou d'agir qu'intériorise un individu par l'apprentissage continu
et le plus souvent inconscient. La culture ainsi acquise se projette dans le milieu de travail. Hofstede
(1980) retient quatre façons par lesquelles la culture se cristallise. Il illustre les niveaux de manifestation
de la culture par une pelure d'oignon avec au centre les valeurs, puis respectivement les rituels, les héros
et les symboles. Les trois dernières couches correspondent aux pratiques identifiées par l'observation
externe, même si le sens culturel qui leur est accordé est invisible. Ces programmes mentaux
correspondent à des niveaux de culture qui sont: le niveau national, le niveau relatif à l’appartenance à un
groupe régional et/ou ethnique et/ou religieux et/ou linguistique, le niveau relatif au sexe de l’individu, le
niveau relatif à la génération, le niveau relatif à l’origine sociale, à l’instruction et à la profession exercée
et le niveau relié à l’organisation ou l’entreprise à laquelle appartient l’individu.

Dimensions universelles. Les systèmes de valeurs auxquels fait référence Hofstede (1980) se
retrouvent également au cœur des organisations. Il a donc dressé l'inventaire de leurs particularités qui a
donné lieu à la construction d'un modèle dimensionnel des cultures nationales à partir de l'expression de
quatre valeurs ou dimensions prédominantes qui caractérisent les écarts culturels : la distance
hiérarchique, le contrôle de l'incertitude, le degré d'individualisme ou de collectivisme et le degré de
masculinité ou de féminité. Plus tard, Hofstede (1991) ajoutera une cinquième dimension : l'orientation
vers le court ou le long terme d'une société. Les travaux de Hofstede (1980, 1991) ont animé ceux de
plusieurs chercheurs qui lui ont succédé. Les dimensions fondamentales proposées par ce dernier ont par
ailleurs été maintes fois reprises et modifiées (Lainé, 2004; Mutabazi et al., 1994; Adler, 1991, Usunier,
1985) afin de leur offrir une portée plus pratique que théorique.

4
Les individus issus de l'Amérique Latine et du Moyen-Orient donnent parfois l'impression aux Américains d'aborder le temps
de manière chaotique, où tout peut changer selon les aléas de la vie. Pour cette conception, il n'est pas rare de voir des rendez-
vous annulés ou reportés à la dernière minute. Par ailleurs, le contact humain et les relations interpersonnelles sont privilégiés, ce
qui explique souvent les fluctuations dans ce qui a été prévu au départ.

133
La culture comme un ensemble d'universaux

Trompenaars et Hampdon-Turner (2004), dont les travaux sont des plus récents, proposent un
modèle à vocation instrumentale et consultative pour les organisations. Partant du principe que chaque
culture élabore ses propres choix de résolution de problèmes, ils dégagent sept dimensions duelles
fondamentales de la culture. Leur conceptualisation repose sur un modèle polarisé : dans chaque
dimension, c'est au point médian que se situe l'efficience visée par l'organisation5.

La culture: une définition plurielle. D'autres études ont poursuivi la même démarche, mais dans
d'autres contextes (Lainé, 2004; Mutabazi et al., 1994; Adler, 1991; Usunier, 1985). La voie adoptée
conserve une vocation instrumentale visant la gestion interculturelle, comme la synergie culturelle6
d'Adler (1991) et l'ATLAS7 de Lainé (2004). Pour Symington (1983), la culture est «un ensemble
complexe qui comprend le savoir, les croyances, l'art, la morale, les coutumes, ainsi que toutes les
aptitudes et les habitudes qu'un être humain acquiert en tant que membre d'une société» (Symington, 1983
: 1). Pour Barnouw (1963), c'est «le mode de vie d'un groupe d'individus, la combinaison de tous les
modèles plus ou moins stéréotypés de comportement appris, que chaque génération transmet à la suivante
au moyen du langage et de l'imitation» (Barnouw, 1963: 4).

Adler (1991) a développé six dimensions empruntées à Kluckhohn et Strodtbeck (1961) qui sont
les qualités individuelles des membres, leur relation avec la nature et avec le monde, leur relation avec les
autres membres, leur activité principale, leur orientation spatiale et leur orientation temporelle. Comme
pour celles de Hofstede (1980), ces dimensions servent à cerner les différents comportements au travail.
Adler (1991) en propose deux de plus et ne reprend que le couple individualisme/collectivisme.

Le modèle des références sociétales

Ce modèle vise, comme le précédent, à approfondir l'impact de la culture sur les pratiques de
management afin de mieux harmoniser ces pratiques avec les situations locales. Toutefois, il se démarque
par la définition de la culture qu'il propose et par sa manière de rendre compte des spécificités culturelles.

Les postulats de la culture comme un système de sens

C'est principalement à travers l'approche interprétative de d'Iribarne (2002, 1989) et de son équipe
qu'elle a été médiatisée. Ce dernier construit un corpus théorique hybride sur le plan disciplinaire:
jumelage de l'ethnographie avec une grille d'interprétation qui repose sur l'histoire des cultures politiques
et des systèmes d'organisation sociale. La démarche met en œuvre un inventaire des cultures politiques et
en examine les ancrages historiques respectifs. À partir du postulat que les modes de régulation sont

5
Par exemple, pour l'une de leurs dimensions qui est l'universalisme versus le particularisme, les auteurs démontrent qu'une
société universaliste accueille la diversité (dans une optique positive) mais manque de discernement dans l'appréciation des
singularités individuelles (dans une optique négative).

6
Adler (1991) propose la notion de synergie culturelle comme une approche intégrée de la gestion interculturelle. La synergie
culturelle est une stratégie pour gérer la diversité culturelle et ainsi transcender les conséquences qu'elle produit au sein des
organisations. Cette méthode serait tout indiquée pour traiter des conflits dans une situation de multiculturalisme.

7
ATLAS est un outil pratique qui permet de déterminer la dimension culturelle à travers une méthode d'identification des
variables culturelles. Voir à ce sujet l'ouvrage de Lainé, S. (2004), Management de la différence. Apprivoiser l'interculturel.

134
récurrents aux différents niveaux d'un espace national8, il dresse une typologie des cultures. Ce modèle
reconnaît au courant de l'anthropologie symbolique sa définition de la culture comme étant un processus
d'interprétation : les individus attribuent des significations à leur réalité à partir d'un référentiel de sens.
Cette unité de système de sens renvoie à un processus tant individuel que social: chacun interprète de
manière personnelle ce qu'il vit en se rapportant aux normes institutionnelles et formelles partagées et
c'est par ces normes que se construit le contexte d'interprétation de la culture politique. Dès lors les
comportements individuels et organisationnels s’inscrivent résolument dans une optique stratégique. Il
importe alors «de mettre en rapport les stratégies des acteurs avec le sens qu'ils accordent aux situations
où ils se trouvent et aux actes qu'ils sont susceptibles d'accomplir» (d'Irbarne et al., 2002: 343). On peut
alors cesser «de concevoir une culture comme un univers de pratiques uniformes pour l'appréhender
comme un ensemble de codes au sein duquel la diversité des pratiques prend sens (que ces pratiques
soient majoritaires ou déviantes, marquées par l'immobilisme ou porteuse de nouveauté)» (Iribarne (d') et
al., 2002: 347). Dans ce modèle, les référents expliquant des conduites sont permanents et signent le
rattachement à une culture mais l'expression de ces conduites peut varier. Par exemple, les rapports
d'autorité dans les organisations en France seraient tributaires des modèles aristocratiques -sous-officier et
soldat-, clérical -maître et apprenti- puis servile -maître et laquais- (Chevrier, 2003). La culture politique
provient ici du fait que «la révolution a aboli les ordres dans leur forme ancienne, mais n'a pas balayé la
conception fondamentale de la société organisée en groupes sociaux hiérarchisés» (Chevrier, 2003: 76).
Cette idée est au cœur des modèles sociétaux. C'est le passage du clan à l'individu qui, par le travail,
trouve son identité. Ainsi, avec la Révolution Industrielle, l'individu naît avec le travail, ce qui désormais
donne lieu à une nouvelle hiérarchisation par le travail. Cette hiérarchisation renvoie à une codification
selon le mérite de la formation des individus. Les fonctions qui se trouvent au sommet de l'entreprise
reviennent à ceux qui sont passés par les grandes écoles.

Les référents qui organisent les formes de liens sociaux sont tout aussi présents dans les situations
de travail. L'entreprise déploierait des modalités de fonctionnement qui se conforment aux codes
d'interprétation enracinés dans le contexte social plus large. La culture politique transportée à travers le
parcours historique s'intègre au sein des pratiques de gestion de l'organisation.

Portée et limites

Modèles des références nationales. Hall, compte tenu du contexte des États-Unis des années
1960 jusqu'aux années 1980, s'est opposé à l'idée dominante de l'universalité des façons de communiquer.
Il démontre l'importance de la culture dans les conduites individuelles. Néanmoins, son corpus théorique
comporte des propositions contradictoires notamment au niveau de sa définition déterministe de la
culture. Alors qu'il maintient l'hypothèse que l'individu est si emprisonné par sa propre culture qu'il ne
peut être contrôlé de l'extérieur, il soutient que les Japonais sont polychrones lorsqu'ils travaillent entre
eux mais monochrones dans leurs relations avec les Occidentaux (1984). Comment pouvons-nous être
conditionnés par notre culture, qui constitue l'essence de ce que nous sommes, et en même temps
s'approprier aisément un autre système temporel opposé dans certaines situations?

L'idée d'une unité culturelle pour exprimer le caractère commun semble avoir été écartée dans les
travaux de Hall (Chevrier, 2003). Cette omission soulève des interrogations, entre autres, lorsqu'il affirme
qu'«à un niveau préconscient, la monochronie est masculine, et la polychronie, féminine» (Hall, 1984:
66). Devons-nous supposer qu'il y a une culture différente pour les hommes et les femmes d'une même
nation?

8
Voir les travaux de Crozier (1963) qui a cherché à comprendre l'influence de la façon de vivre en société des Français dans le
fonctionnement des entreprises françaises.

135
 Pour sa part, Hofstede (1980) cherche à mettre en évidence une relativité des pratiques
managériales selon les contextes culturels. Ces recherches ont eu un impact majeur sur le management
interculturel. En dépit du travail colossal de Hofstede (1980), les résultats obtenus sont des données
quantitatives très spécifiques d’une multinationale tout aussi spécifique et d’un seul type de travailleurs
lié au domaine de l'informatique. Cela a-t-il une incidence sur la portée des résultats? Ces derniers
peuvent-ils être projetés sur d'autres milieux? La question demeure entière.

Un autre problème méthodologique important dans la catégorisation des cultures nationales de

Hofstede (1980) consiste à départager ce qui se rapporte aux aspects individuels qui sous-tendent des
éléments pouvant appartenir à d'autres composantes de la culture comme la "culture professionnelle" ou la
"culture organisationnelle". Ces composantes ne s'insèrent qu'en partie dans la culture nationale.

Les travaux de Trompenaars et Hampdon-Turner (2004) postulent l'existence de six dimensions

duelles universelles. C'est leur dualité qui est porteuse des réponses traitant les problèmes de
communication interculturelle et c'est dans ces réponses que se cristallise le caractère spécifique d'une
culture. Cette conception universelle de la culture semble peu nuancée et occulte substantiellement
l'aspect anthropologique de la diversité culturelle, nonobstant le fait que ce modèle d'analyse est certes
facile à opérationnaliser pour les entreprises. Il est également culturellement marqué en raison de son
choix de prôner le one best way à travers des formules destinées directement aux managers.

Le modèle des références nationales se caractérise finalement par son traitement instrumental de
la diversité culturelle en proposant des repères pragmatiques pour la gestion interculturelle. Cela alors
même que cadre théorique reste marqué par des incohérences conceptuelles et sémantiques.

Modèles des références sociétales. Le modèle des références sociétales met en lumière la
récurrence des attitudes et des comportements individuels et organisationnels dans les cultures. Il y a une
volonté de normaliser et d'uniformiser les pratiques de management en termes de performance
organisationnelle. Il se démarque sur le plan méthodologique par une démarche qualitative qui vise des
échantillons plus restreints dans des entreprises différentes. Cette méthode d'étude de cas permet une
justification plus pointue des mécanismes organisationnels. En outre, ce n'est pas la logique universelle
comme dans le premier modèle qui intéresse d'Iribarne (2002, 1989) mais bien une logique locale
permanente qui s'exprime par les pratiques de gestion et le fonctionnement global d'une société. Son cadre
théorique contient des incohérences et une hétérogénéité dans l'analyse qui font obstacle à sa
généralisation.

Un modèle en émergence : vers une typologie de la culture?

C’est à partir de la remise en question de la portée et des limites des deux précédents modèles
qu’un troisième modèle semble prendre corps. L'approche de d'Iribarne (2002, 1989) suscite plusieurs
interrogations d'ordre théorique et au niveau de sa définition de la culture. Dupuis (2004) souligne le
manque de cohérence et les contradictions qui ne permettraient pas la construction d'un modèle empirique
à portée explicative générale. C'est par les réponses alternatives à certaines critiques de fond que se
construit progressivement ce nouveau modèle d'analyse dont l'articulation principale est la construction
d'une typologie anthropologique des cultures.

136
Une critique du modèle des références sociétales

Une instabilité terminologique. D'Iribarne emploie des expressions comme les "traits
fondamentaux" (1989 : 12), les "principes fondamentaux" (1989 : 20), les "grands principes" (1989 : 201),
le "principe d'ordre" (1993: XIII), les "principes communs" (1993: XIX), les "valeurs" (1989: 14, 1998:
258), les grands repères (1998: 259). Bien qu'il reconnaisse que le recours à ces différents concepts traduit
probablement une même idée chez d'Iribarne, Dupuis (2004) soutient que l'instabilité terminologique de
ce dernier n'est pas sans conséquence et peut mener à la confusion car elle renvoie à des paradigmes et
des perspectives disciplinaires différents.

Une incohérence théorique. Dupuis (2004) reproche à d'Iribarne certaines disparités qui mènent
tout droit vers une incohérence de son corpus et une impasse théorique : «nous passons ainsi de référents
ultimes propres à une culture à des universaux vécus différemment selon les cultures» (Dupuis, 2004: 25).
En effet, d'un côté, il lui est impossible de rendre compte d'une unité de la culture compte tenu que les
catégories universelles éclipsent désormais les référents ultimes qui permettraient de le faire. D'un autre
côté, étant donné que ce sont les associations communes qui donnent lieu au sens, qui lui est propre à
chaque culture, l'élaboration d'une typologie des cultures ne peut être conçue. Cela supposerait la
présence d'éléments communs aux différentes cultures. En revanche, d'Iribarne ne peut catégoriser les
cultures car elles renvoient à des spécificités locales. Or un choix s'impose entre les deux, qui passerait
selon nous par une définition anthropologique contemporaine de la culture articulée autour d'une
conception plus dynamique. L'idée de logique ou de référent ultime, entre autres, s'inscrivent dans un
cadre prédéterminé désormais entaché de caducité face aux transformations récentes du travail et des
organisations.

Synthèse et propositions critiques

Selon la définition de Cuche (1996), la culture serait un ensemble d'éléments en interaction les
uns avec les autres. Le choix des oppositions fondamentales extraites des études anthropologiques de
Lévi-Strauss (1958) -sur lesquelles s'appuie d'Iribarne- confirme la logique déterministe des deux
premiers modèles.

La notion de référent ultime, telle que proposée par d'Iribarne, laisse sous-entendre une
interprétation différente de ce que veut l'auteur. En attribuant aux "référents ultimes" un déterminisme
intrinsèque à chaque culture, d'Iribarne subordonne les comportements des individus et des organisations
à une pré-notion fortement approximative : l'inconscient collectif attaché à toute culture. La question se
pose alors de pouvoir repérer la référence ultime propre à chaque culture. Nous suggérons avec Dupuis
(2004) un retour à l'idée de "valeur", définie comme un référent commun. Ce dernier agit comme point de
repère pour les individus «ne serait-ce que pour mieux s'y opposer» (Dupuis, 2004: 27). Il y a donc
présence d'une interaction «qui consiste pour un individu pourvu de certains attributs à établir des
préférences entre des états différents d'un système social» (Boudon et Bourricaud, 1990: 665). Un
système de valeurs n'est jamais parfaitement stable mais il y a persistance d'un noyau dur de la culture qui
évolue avec le temps. À cet effet, la définition de valeur de Rocher (1969: 56) comme «une manière d'être
ou d'agir qu'une personne ou une collectivité reconnaissent comme idéale et qui rend désirables ou
estimables les êtres ou les conduites auxquels elle est attribuée» semble répondre aux risques
d'approximation évoqués plus haut. En raison de la notion de "dynamique de valeur" qui en découle et qui
porte la représentation symbolique des pratiques associées à une culture, la logique fondée sur les
référents ultimes devient désormais sans objet. Cette proposition porte potentiellement les composantes
d'une typologique des cultures dans la mesure où chaque culture se définit par une combinaison
d'interactions unique de ses valeurs. Une telle proposition présente tous les éléments de fécondité
nécessaire au développement de ce troisième modèle en émergence.

137
Vers une typologie de la culture?

Le modèle des références nationales contient une variété de définitions de la culture; tantôt un
système de communication (Hall, 1984, 1979, 1971), tantôt une programmation mentale (Hofstede, 1980)
ou une solution à des situations de crise (Trompenaars et Hampdon-Turner, 2004). Ces définitions ont en
commun de reposer sur des dimensions qui "cataloguent" les cultures nationales sur une base d'indicateurs
quantitatifs encadrant une analyse comparative des valeurs, croyances, attitudes et comportements.
Partageant cette même tendance relativiste, l'ambition du modèle des références sociétales est de créer un
modèle d'analyse appliqué à la gestion interculturelle des organisations. À la différence du premier
modèle, il fait une discrimination des cultures politiques historiquement déterminées et conçoit la culture
comme une logique héritée et collectivement reconduite par la reproduction des statuts sociaux et
professionnels et, finalement, des comportements des individus et des organisations.

Le modèle en construction écarte les valeurs comme concept fondateur en raison de sa portée
universelle et globalisante. Il conduit la définition de la culture vers une dynamique de valeurs
caractérisée par une combinaison unique d'interactions propre à chaque culture. La culture apparaît ainsi
non seulement comme une dynamique interne mais également comme un processus en perpétuelle
mutation, pourvu d'une capacité évolutive en fonction du temps et de l'espace. Ce modèle est en mesure
d'intégrer les études culturelles et ethnographiques ayant pour objet les comportements des et dans les
organisations. Certaines interrogations interpellent cependant ce modèle. À l'heure de la mondialisation et
du métissage, peut-on encore penser à une typologie reposant uniquement sur une délimitation des
frontières géographiques? Comment décrire une culture? Quelles en sont les caractéristiques
significatives?

Enjeux de performance ou innovation organisationnelle?

Le corpus d'études du management interculturel apparaît caractérisé d'abord par le rôle central
qu'il attribue à la gestion de la diversité culturelle en tant qu'enjeu de performance organisationnelle (Côté
et Soussi, 2006). De la théorie des universaux (Hall, 1979) à celle de la schématisation dichotomique de
Trompenaars et Hampdon-Turner (2004), en passant par les dimensions duelles (Hofstede, 1980, 1994),
ces travaux mettent en avant des résultats quantitatifs visant avant tout le "prêt-à-appliquer"
organisationnel dans la gestion des différences culturelles. Au-delà de ces modes d'emploi, subsiste une
zone d'imprévisibilité qui oblige les gestionnaires à implanter des pratiques nouvelles face à des situations
inédites en matière de diversité culturelle. De ce fait, l'on voit émerger un réel corpus empirique de
pratiques managériales apparaissant comme d'authentiques innovations organisationnelles (Côté et
Soussi, 2006).

Bien que les modèles présentés soient dominants, ils n'ont guère insisté sur la nécessité
d'introduire des nouvelles pratiques d'intégration de la diversité culturelle formalisées au niveau des
fonctions de l'organisation (RH ou autre). La gestion interculturelle reste une préoccupation définie en
termes de gains de productivité. Dans le modèle des références nationales ces pratiques relèvent, de façon
diffuse et transversale aux différentes fonctions de l'organisation, sans volonté récursive de les intégrer
comme prérogative ad hoc de la fonction RH. Le modèle des références sociétales renvoie, quant à lui, à
l'adéquation des référentiels culturels avec l'organisation mais seulement au niveau de certains secteurs de
l'organisation du travail, comme les programmes de gestion de la qualité notamment, et non dans un
projet RH intégrateur sur les plans professionnel et organisationnel. Jusqu'ici, l'intérêt porté à cette
question par le management interculturel, s'inscrit surtout dans la perspective fonctionnaliste des théories
des organisations dont l'objectif est «d'apprivoiser l'interculturel» pour une meilleure efficacité. De
surcroît, c'est l'aspect exogène (des entreprises à l'étranger) de la diversité qui y est traité, sans aborder le
volet endogène, celui de l'insertion des individus dans un milieu de travail local commun, culturellement
marqué.

138
 Enfin, le modèle critique, en émergence, porte un intérêt aux pratiques tenant compte des identités
individuelles mais sans manifester une préoccupation socialement innovante en matière de diversité
culturelle. L'interrogation demeure dans la mesure où l'on ne dispose que de peu d'études appropriées et
encore moins d'une définition de la culture couvrant les principaux indicateurs d'analyse de ce modèle.
Dupuis (2004) souligne l'importance de tenir compte de l'individu dans l'espace de travail mais l'idée n'est
pas poussée davantage. En somme, la question de la prise en charge de la diversité culturelle reste entière
et peut susciter des tentatives d'innovations organisationnelles dans les entreprises qui vivent cette réalité
au quotidien et à laquelle font défaut un appareillage théorique adéquat et des données empiriques fertiles.

Par ailleurs, en termes de GRH, deux principaux défis interpellent la fonction RH: «celui de
développer des processus viables par lesquels la gestion des ressources humaines peut contribuer à la
planification stratégique dans les organisations, ainsi que d'adopter et de promouvoir la pensée stratégique
au sein de la fonction elle-même» (Saba et Pucik, 1997: 7). Il s'agit d'aligner les composantes
organisationnelles à celles de l'environnement externe (Saba et Pucik, 1997). Cet environnement est
touché par les facteurs du flux migratoire, de l'émergence d'une économie du savoir et de la
mondialisation. Le besoin se fait de plus en plus pressant d'introduire dans les orientations de la fonction
RH une gestion qui tient compte de l'intégration sociale et professionnelle des individus dans un milieu
interculturel.

Conclusion. La diversité culturelle : un cheval de Troie pour la recherche?

Une leçon peut être tirée de cette synthèse critique. C'est la nécessité d'une passerelle
épistémologique de l'anthropologie vers les sciences de gestion. L'approche anthropologique étudie
l'organisation en tant que dynamique d'un système de relations (Lévis-Strauss, 1958) et non comme une
entité isolée. Sur le plan anthropologique, les comportements organisationnels apparaissent moins fondés
sur une rationalité singulière. Ils renvoient davantage aux représentations symboliques qui leur sont
intrinsèques qu'aux considérations stratégiques que leur attribuent les modèles d'analyse dominants du
management interculturel. Les sciences de la gestion partagent peu cette conception de l'organisation en
raison de leur intérêt pour l'organisation en tant qu'objet empirique observable, sur lequel il est possible
d'agir concrètement.

L'abstraction des enjeux attachés à la diversité culturelle s'impose comme un défi majeur face
auquel les sciences de gestion semblent a priori démunies en termes de modèles d'analyse à portée
systémique. Devant cet écueil, l'appareillage conceptuel anthropologique propose des éléments d'analyse
que les théories de l'organisation pourraient mettre en valeur. Les deux disciplines gagneraient à élaborer
une approche méthodologique complémentaire.

Il s'agit d'exploiter les conditions favorables à la création d'une épistémologie commune aux
sciences de la gestion et de l'anthropologie. Cela suppose pour les premières qu'elles transcendent leur
conception traditionnelle de l'entreprise. De plus, la recherche en gestion s'inscrit encore dans un
paradigme positiviste des plus orthodoxes alors qu'elle pourrait s'enrichir d'un questionnement qui prenne
en compte les incidences organisationnelles des représentations symboliques. En cela, les recherches en
anthropologie des organisations sont fertiles. Cela suppose aussi l'abandon partiel, par le management
interculturel, d'un postulat devenu dominant depuis quelques décennies : son projet d'une science de
l'action et de la décision, et en retour qu'il s'approprie un corpus théorique qui ne lui demeure accessible
qu'au prix d'une volonté avérée de transdisciplinarité.

Ce rapprochement interdisciplinaire autour des sciences de l'organisation –sur la base des enjeux
de la diversité culturelle- ouvrirait la voie à une exploitation potentiellement riche de résultats de
recherche qui sont restés jusque-là confinés dans des disciplines cloisonnées comme l'ethnologie et

139
l'anthropologie des organisations. Il s'agit pour l'anthropologie d'examiner une situation spécifique avec
un fort contrôle des variables plutôt que de chercher à comprendre la globalité d'une culture (Bonnafous-
Boucher, 2005). Toutefois, ni le relativisme contextuel propre à l'anthropologie, ni l'universalisme associé
au management ne permettent d'entrevoir l'intégration d'une typologie des cultures qui tienne compte des
acteurs de terrain alors qu'un investissement théorique en ce sens constituerait une solide posture pour
appréhender l'irruption récurrente et inexorable de la diversité culturelle dans le contexte de
l'organisation.

Une difficulté méthodologique apparaît cependant avec cette voie interdisciplinaire. Les
organisations constituent des espaces clos qui, de façon toute légitime, n'ouvrent leurs portes que
lorsqu'elles y trouvent un intérêt en matière d'efficacité et de performance. Or, la question de la diversité
culturelle semble justement un objet de recherche qui converge vers un même centre d'intérêt pour les
organisations et cette nouvelle voie de recherche exigeant la contribution des études anthropologiques. La
problématique de la diversité culturelle pourrait être ce cheval de Troie salutaire qui permettrait
d'entrouvrir à la recherche le champ souvent hermétique des organisations.

140
 Synthèse des modèles d’analyse

Modèle des références nationales Modèle des références sociétales Modèle émergent

Variable Universaux de la culture Noyau dur de la culture Valeurs : combinaison unique

Référents ultimes d'interactions

Méthodologie Approche anthropologique Approche culturaliste Approche ethno-organisationnelle

Méthodologie quantitative Méthodologie qualitative Méthodologie qualitative
Perspectives Anthropologie Sciences administratives Sciences de gestion
disciplinaires Sociologie des organisations Ethnologie Anthropologie
des organisations
Hofstede (1980) d'Iribarne (2002, 1989) Dupuis (2004)
Trompenaars et Hampdon-Turner Chevrier (2003) Côté et Soussi (2006)
Travaux dominants (2004) Usunier (1991, 1985)
Hall (1971) Lainé (2004)
Adler (1991)
Saba et Pucik (1997)
Hall (1971) : d'Iribarne (2002, 1989): En construction :
- système de communication - système de sens - dynamique de valeurs
Hofstede (1980) : - noyau dur - référents anthropologiques
Définitions de la - programmation mentale - référents ultimes
culture Trompenaars, Hampdon-Turner - repères principaux
(2004)
- solution aux dilemmes universaux
Enjeux Vocation exogène Vocation exogène Vocation endogène et exogène
organisationnels Pas de formalisation en GRH Intégration collective Intégration socio-
Intégration individuelle dans le Intégration de l'organisation dans le organisationnelle
milieu milieu socio-institutionnelle

141
 Références

Adler, N., Comportement organisationnel. Une approche multiculturelle, 2ème Édition, Les Éditions
Reynald Goulet Inc, 1991.
Barnouw, V., Culture and Personality, Homewood, Ill.: Dorsey Press, 1963.
Boltanski, L., Chiapello, E., Le nouvel esprit du capitalisme, Paris, Gallimard, 2004.
Bonnafous-Boucher, M., Anthropologie et gestion, Paris, Économica, 2005.
Boudon, R., Bourricaud, F., «Valeurs», Dictionnaire critique de la sociologie, Paris, PUF, 1990, 663-670.
Cardu, H., Sanschagrin, M., «Les femmes et la migration: les représentations identitaires et les stratégies
devant les obstacles à l'insertion socioprofessionnelle à Québec», Recherches féministes, 15(2), 2002, 87-
122.
Castells, M., L’ère de l’information, Paris, Fayard, 2001.
Chevrier, S., Le management interculturel, Paris, PUF, 2003.
Côté, A., Soussi, A. S., «La diversité culturelle au travail : portée et limites des modèles théoriques
dominants», Actes de Colloque du CRISES, no: HS0601. coll. Hors-Série, 2006, 112-119.
Crozier, M., Le phénomène bureaucratique, Paris, Éditions du Seuil, 1963.
Cuche, D., La notion de culture dans les sciences sociales, Paris, Éditions La découverte, coll. Repères,
1996.
David, P. A., Foray, D., «Une introduction à l'économie et à la société du savoir», Revue Internationale
des Sciences Sociales, (Mars 2002), 13-27.
Dupuis, J.-P., «Problème de cohérence théorique chez Philippe d'Iribarne: une voie de sortie»,
Management international, 8(3), 2004, 21-29.
Drudi, G., «Impact de la non-reconnaissance professionnelle sur l'identité des personnes immigrantes.
Situation des personnes issues des minorités noires», Vivre ensemble, 11(38), (Automne 2002/ Hiver
2003), 13-16.
Hall, E.T., La dimension cachée, Paris, Éditions du Seuil, 1971.
Hall, E. T., Au-delà de la culture, Paris, Éditions du Seuil, 1979.
Hall, E.T., La danse de la vie. Temps culturel, temps vécu, Paris, Éditions du Seuil, 1984.
Helly, D., «Le multiculturalisme canadien: de la promotion des cultures immigrées à la cohésion sociale
1971-1999», Cahiers de l'URMIS, no. 6, 2000.
Hofstede, G., Culture’s consequences: International differences in work related values, Beverly Hills,
CA: Sage Publications, 1980.
Hofstede, G., Culture and Organizations: Sofware of the Mind, Londre, MacGraw Hill, 1991.
Hofstde, G., Vivre dans un monde multiculturel. Comprendre nos programmations mentales, Paris, Les
Éditions d'organisation, 1994.
Hofstede, G., Culture’s consequences: Comparing Values, Behaviors, Institutions and Organizations
Across Nations, Sages Publications, 2002.
Iribarne, P. (d’), La logique de l’honneur, Paris, Éditions du Seuil, 1989.
Iribarne, P. (d’), La logique de l’honneur, Paris, Éditions du Seuil, 1993.

142
Iribarne, P. (d’), Henry A., Segal, J.-P., Chevrier, S., Globokar, T., Cultures et mondialisation. Gérer
par-delà les frontières, Paris, Éditions du Seuil, 1998.
Iribarne, P. (d’), Henry, A., Segal, J.-P., Chevrier, S., Globokar, T., Cultures et mondialisation. Gérer
par-delà les frontières, Paris, Éditions du Seuil, 2002.
Kluckhohn, F., Strodtbeck, F.L., Variations in Values Orientations, Evanston, Ill.: Row, Peterson, 1961.
Lainé, S., Management de la différence. Apprivoiser l'interculturel, Afnor, 2004.
Lévis-Strauss, C., Anthropologie structurale, Paris, Plon, 1958.
Mutabazi, E., Altman, Y., Klesta, A., Poirson, P., Management des ressources humaines à l'international.
Filiales-fusions-acquisitions-alliances-coopérations, Paris, Éditions Eyrolles, 1994.
Rocher, G., «Introduction à la sociologie générale», Tome 1: L'action sociale, Montréal, Hurtubise HMH,
1969.
Saba, T., Pucik, V., Le rôle stratégique de la fonction ressources humaines au sein des entreprises
internationales: élaboration d'un cadre conceptuel, Fonds de recherche de l'École de relations
industrielles, (Février 1997).
Sériyex, H., La nouvelle excellence. Réussir dans l'économie nouvelle, Paris, Maxima Laurent Du Mesnil,
2000.
Steinmueller, W. E., «Les économies fondées sur le savoir. Leur liens avec les technologies de
l'information et de la communication», Revue Internationale des Sciences Sociales, (Mars 2002), 159-
173.
Symington, J. W., «Learn Latin America's Culture», New York Times, (23 septembre 1983).
Trompenars, F., Hampdon-Turner C., Au-delà du choc des cultures. Dépasser les oppositions pour mieux
travailler ensemble, Paris, Éditions d’Organisations, 2004.
Usunier, J.C., Management international, Paris, PUF, 1985.
Usunier, J.C., «Perceptions du temps des affaires et cultures nationales: une comparaison internationale»,
Économies et sociétés, Séries sciences de gestion, no. 17,(Avril 1991), 81-114.
Viprey, M., «Un modèle d'immigration de main-d'œuvre ouvert, fondé sur la sélectivité», Chronique
internationale de l'IRES, no. 84, (Septembre 2003).

143
ASAC 2006 Heather G. Perretta (Student)
Banff, Alberta Pamela S. Tolbert
Faculty of Organizational Behavior
Cornell University

MORE OR LESS: HIERARCHICAL DIFFERENT IN THE EFFECTS OF

GENDER COMPOSITION ON THE HIRING OF WOMEN FACULTY

Using data drawn from a random sample of baccalaureate-granting colleges and

universities, this research examined predictions derived from both demand-side
and supply-side theoretical frameworks about the impact of organizational sex
composition on the hiring of women faculty. Our results indicate that the effects
of women’s proportional representation on hiring patterns varied markedly across
hierarchical levels.

Introduction

In the nearly 25 years since Pfeffer (1983) identified organizational demography as a promising
area of study, researchers have explored the impact of the demographic composition of work groups and
organizations on a variety of outcomes (see Reskin, McBrier, and Kmec [1999] for an overview). In
conjunction with Baron and Bielby’s (1980) call to “bring the firms back in” to stratification research,
many empirical investigations have focused specifically on the relationship between the proportional
representation of women and minorities in organizations and various aspects of inequality (Konrad and
Pfeffer 1991; Shenhav and Haberfeld 1992; Tolbert 1986). While this research has provided ample
documentation of the influence of demography on organizational dynamics and stratification outcomes,
the specific processes that produce variations in demographic composition – hiring, promotion, and
turnover – have been relatively neglected (Bygren and Kumlin 2005; Mittman 1992; Tolbert et al. 1995).
In addition, much of the extant research has been based on cross-sectional rather than longitudinal data,
often making questions of causal influence difficult to address. Finally, demographic studies have
typically focused on relations among peers – individuals at the same organizational level – and neglected
questions of whether or how the impact of group proportions varies at different hierarchical levels (for
exceptions, see Cohen, Broschak, and Haveman 1998; Elvira & Cohen 2001; Ely 1994; 1995; Gorman
2005). Little is known about the relative effects of having a high proportion of minorities (i.e., members
of socially subordinate groups) in more powerful positions compared to the effects produced by having a
high proportion of minorities in lower, less powerful positions.

One notable effort to address these issues is represented by the work of Cohen et al. (1998). In a
study using longitudinal data from savings and loan institutions, the authors examine the influence of
women’s representation at different levels of an organization at one point in time on the subsequent
movement of women into management positions through hiring or promotion. The findings of their study
suggest that the distributions of women at different levels have varying effects on women’s movement
into mid-level and lower-level management positions. Their results also suggest that demographic
composition may affect promotion and hiring processes in different ways, but limitations of the data
prevent the authors from examining these two processes separately for appointments to mid- and high-
level organizational positions.

Intuitively, it seems reasonable to expect that the effect of an organization’s gender composition
on women’s promotion from within will differ from the effect on the hiring of women from outside the
firm, particularly for positions with more power. The proportion of women in an organization at a given

144
level should be positively related to the likelihood that an internally-filled position at the same or a higher
level will be given to a woman, simply because there are apt to be more women among potential
candidates (Huffman 1999). Moreover, a high rate of promotion of men from ranks containing a large
proportion of women may be perceived as overt discrimination. Under such circumstances, demographic
composition is apt to exert a more passive influence on women’s subsequent internal mobility. On the
other hand, it is not clear that organizations with a large proportion of women will necessarily be more
likely to hire a woman when making an external hire. In fact, one theoretical approach to the impact of
group proportions proposed by Blalock (1967), stressing intergroup competition for resources, leads to
the opposite prediction: as the proportion of women in a group increases, resistance to hiring more
women should also increase. In general, we argue that decisions about external hires represent a purer
measure of the active influence of women’s proportional representation on group relations and personnel
outcomes.

Hence, in this study, we focus specifically on the impact of the gender composition of the faculty
on the hiring of new faculty members by higher education institutions. We incorporate measures of
women’s proportional representation among the faculty at different hierarchical ranks in an examination
of the hiring of women both for untenured and tenured positions. Our analyses are based on data
collected for a random sample of baccalaureate-granting institutions, from 1988 through 1997.

Theories of Demographic Effects: Evidence and Hypotheses

Both demand-side approaches to labor market processes, focusing on preferences and choices of
employers, and supply-side approaches, focusing on preferences and choices of job applicants, suggest
that the proportion of individuals with a given social characteristic in an organization may affect hiring
outcomes. Below, we describe the general theoretical logic and supporting empirical evidence for these
approaches, and use them to derive predictions about the relation between the demographic composition
of an organization and hiring outcomes.

Demand Side Approaches

One demand-side approach, articulated by Blau (1977) and Kanter (1977; 1993), draws on social
contact theory to argue that increases in the proportional size of a minority group will lead to a more
favorable social environment for that group. According to this perspective, social prejudice flourishes in
situations where members of a majority group have limited opportunities to interact with minority group
members. Based on this assumption, Blau argued that an increase in the proportional size of a minority
group, by increasing the likelihood of intergroup interactions, will reduce the majority’s propensity to
engage in stereotyping of and discrimination against minority members. These positive effects of
intergroup interactions will gradually decline as the minority group approaches parity with the majority.

Similarly, Kanter argued that women in skewed groups, or ‘tokens,’ suffer from heightened
visibility, a sense of isolation, and pressure to conform to gender-based stereotypes – in short, a relatively
hostile work environment. As the proportions of men and women become more balanced, increases in
male-female interaction are expected to occur, resulting in the elimination of the negative consequences
of tokenism for women. Like Blau, Kanter argued that frequent interaction should disconfirm men’s
propensity to subscribe to gender stereotypes, and dissipate tendencies to define in-group/out-group
boundaries along gender lines. Thus, her analysis also suggests that increases in the proportion of women
in a group will result in a more favorable environment for women, at a declining rate. As proportions
become more equivalent, gender will become less salient to interacting parties, and gender composition
should cease to influence organizational dynamics.

145
 In line with Kanter’s and Blau’s arguments, a number of empirical studies have found a positive
relationship between the proportion of women in a unit and various indicators of a favorable social
environment for women. For example, Konrad and Pfeffer’s (1991) study of women and minorities
within higher education institutions showed that having a higher proportion of women in administrative
positions significantly increased the likelihood that a vacant administrative position would be filled by a
woman. Likewise, research by Blum, Fields, and Goodman (1994), based on a sample medium and large
firms, and by Huffman (1999), using a set of service and manufacturing firms, indicated that
organizations with a higher proportion of women overall had a higher proportion of women in
management positions; similar results were found by Kulis (1997) in a study of colleges and universities.
Baron, Mittman and Newman’s (1991) study of the California civil service from 1979-1985 indicated that
gender integration of departments (requiring increases in the rate at which women were hired) occurred to
a greater extent during this period in departments with a higher proportion of women.

Findings from the study by Cohen et al. (1998), examining the effects of variations in the
proportion of women at different hierarchical levels on hiring and promotion of women managers at low
and mid-level ranks in savings and loan organizations, can also be taken as providing partial support for a
social contact approach. Their analyses indicated that women were more likely to be hired or promoted
into vacant positions when they were already proportionally well represented at a given job level, and that
for mid-level positions, such movement was also positively related to the proportion of women in higher-
level positions. Other findings from their research are less compatible with a social contact perspective.
For example, the proportion of women at mid-level positions had no effect on hirings and promotions of
women at lower-levels; likewise, the proportion at lower management ranks had no effect on promotions
or hires for mid-level positions.

In general, this approach suggests that increases in the representation of women in an

organization should lead to increases in the likelihood of hiring women, albeit at a declining rate. Thus,
we hypothesize that:

Hypothesis 1a: Women’s proportional representation among the faculty of an institution

will have an inverted U-shaped relation to the rate of hiring women at tenured and
untenured levels.

Work within this theoretical framework has not specifically focused on the ways in which
hierarchical differentiation (i.e., the representation of a minority group in higher level, more powerful
positions versus lower-level positions) may affect intergroup relations. One plausible extension rests on
the assumption that the distribution of minorities at higher organizational levels will simply amplify
predicted effects. Thus, having a greater representation of women among higher ranks might be expected
to be especially effective in dispelling negative stereotypes of women and creating more support for
hiring additional women faculty. Again, following the logic of this approach, a curvilinear effect is
predicted: as women’s representation among the tenured faculty approaches parity with men, further
changes in the proportion of women will have less impact on interaction probabilities, and thus will have
less impact on hiring decisions. Based on this distinction, we hypothesize that:

Hypothesis 1b: Women’s proportional representation among the tenured faculty of an

institution will have an independent, inverted U-shaped relation to the rate of hiring
women at tenured and untenured levels.

In contrast to social contact arguments, an intergroup competition perspective gives much more
weight to material interests as a driving force in group relations. The general tenets of this perspective
were articulated in an early analysis by Blalock (1967), in an explanation of findings from research on
residential segregation. He posited competition for economic and political resources as the driving force

146
behind intergroup relations, conceiving of such competition in terms of a zero sum contest, such that “the
success of one [group] implies a reduced probability that others will also attain their goals” (Blalock
1967: 73). Blalock argued that when a minority group is relatively small, majority group members are
unlikely to perceive it as representing a competitive threat to the majority’s control of key resources.
However, as the minority group grows in size, it poses an increasing threat to the majority, who respond
with discriminatory behaviors designed to keep the minority group from gaining a greater share of
community resources. At some threshold level of proportional size, the minority group is expected to
achieve enough power to counteract discriminatory actions by majority members, thus, reversing the
relationship between proportional size and discriminatory outcomes.

While Blalock’s (1967) original work was based on studies of community race relations, Reskin
et al. (1999) and Cassirer (1996) argue that these hypothesized dynamics are likely to hold for gender
relations within organizations as well. There are several studies that provide a reason to expect threat-
responses by men to the increased representation of women in an organization. In particular, research has
indicated that the negative consequences associated with the feminization of occupations (Reskin and
Roos 1990) also often appear at the organizational level. For example, Pfeffer and Davis-Blake (1987), in
a study of college administrators, found that increases in women’s representation in the administration of
an institution were negatively associated with the compensation levels of both male and female
administrators. Likewise, Martin and Harkreader (1993), studying the distribution of women among
different job ladders at a military base, found that both men and women received lower salaries when they
were on ladders containing a higher proportion of women.

A study by Shenhav and Haberfeld (1992) of compensation practices in privately owned firms
found that the pay of men and women was more equal in more demographically balanced firms than in
firms where group proportions were skewed; further analysis indicated that these results were due largely
to the fact that men in firms with a higher proportion of women had relatively low levels of compensation
compared to men in firms with proportionately fewer women. Similarly, Tolbert (1986) found that the
average salary differential between male and female faculty was lower in institutions with a higher
proportion of female faculty, primarily as a result of men’s comparatively low salary levels in such
institutions.

Such findings provide a rationale for expecting a relation between the proportion of women and
levels of discrimination or resistance to further increases in the number of women in a department or
organization, as suggested by Blalock’s (1967) theory (see also Jacobs 1992; Kraus and Yonay 2000;
Pazy and Oron 2001). In line with this approach, a study by Tolbert et al. (1995) of faculty turnover in
departments of sociology, found an inverted U-shaped association between women’s representation in a
department and women’s rates of turnover. Their interpretation suggested that male faculty’s propensity
to discriminate against female faculty increased as the proportion of women in a department increased,
thus leading to higher rates of turnover until women faculty attained a critical mass and were able to exert
countervailing power.

In sum, growth in the proportion of women in an organization is expected to pose an increasing

threat to men in the organization, who are likely to respond with discriminatory actions, including
resistance to hiring additional women faculty members. However, once women achieve some threshold
level of representation within the organization, they can be expected to exert countervailing power against
the dominant male group; this should lead to a greater likelihood of hiring women faculty. Hence, we
hypothesize:

Hypothesis 2a: Women’s proportional representation among the faculty of an institution

will have a U-shaped relation to the rate hiring of women at tenured and untenured levels.

147
 Again, work within this tradition has not focused on the potential impact of the distribution of
minorities at different levels. One plausible assumption is that increases in the proportion of women at
higher ranks within an organization are especially likely to pose a threat to men’s perceived control of
organizational resources; hence, such increases could be likely to have a particularly negative impact on
the hiring of women for both higher and lower-ranking positions, until some threshold level of
representation among higher ranks is attained. Following this logic, we postulate:

Hypothesis 2b: Women’s proportional representation among the tenured faculty of an

institution will have an independent, U-shaped relation to the rate of hiring women at
tenured and untenured levels.

Supply-Side Approaches

Two major supply-side approaches have often guided studies of organizational demography,
similarity-attraction theory and social identity theory. Although these differ in the psychological
mechanisms that are emphasized, they generally converge in their predictions of behavior. According to
the similarity-attraction paradigm, individuals are attracted to and prefer to interact with others who hold
similar attitudes and beliefs. Thus, individuals are more likely to be comfortable in and join groups with
a larger proportion of similar-minded persons in them. When direct information on others’ attitudes is not
available, individuals are assumed to rely on other characteristics, including ascriptive characteristics such
as race and gender, to draw inferences about beliefs and attitudes; physical resemblance is expected to
translate into resemblance in beliefs (Byrne 1971). Hence, individuals are expected to be attracted to
groups with a larger proportion of individuals who have demographic characteristics that are similar to
their own.

In line with the notion that individuals use ascriptive characteristics to make judgments of
similarity, social identity theorists argue that such characteristics are a common basis for the formation of
social categories that individuals use to classify themselves and others (Tajfel 1978; 1982). A key
premise of this perspective is that social evaluations attached to such categories affect individuals’ self-
esteem and, as a consequence, people seek to ensure that the groups to which they belong are positively
evaluated. Various aspects of organizations can signal the status or valuation assigned to different social
groups (Ashforth and Mael 1989; Chatman et al. 1998; Chattopadhyay, Tluchowska, and George 2004;
Ely 1994; 1995; Turner 1987; Wharton 1992). Demographic composition readily serves as one such
signal: the presence of a relatively large number of organizational members with a particular
characteristic is likely to be taken as an indication of the acceptance of individuals who share that
characteristic, affirming that social identity. Thus, both similarity-attraction and social identity theories
suggest a positive, linear relation between the proportion of women in an organization and women’s
propensity to seek and accept employment in that organization.

Although empirical research based on these theoretical approaches has not specifically examined
individuals’ employment decisions, several studies based on this approach have investigated related
phenomena, such as organizational attachment and commitment (Bacharach and Bamberger 2004; Sacco
and Schmitt 2005; Wharton and Baron 1987; Williams and O’Reilly 1998). Overall, this research
provides mixed evidence on the impact of gender relations in organizations. For example, using survey
data from lower-level employees in 151 work units of three large organizations, Tsui et al. (1992) found
that, in general, individuals who differed from others in their unit in terms of gender were less
psychologically committed to the organization, and more likely to report plans to leave. However,
separate analyses for men and women indicated that this result primarily reflected men’s dissatisfaction
with work groups in which they were dissimilar to others; the effects of dissimilarity on women’s
reactions were not significant. In contrast, research conducted by Chatman and O’Reilly (2004), based on
a sample of professional employees working in project teams, found that women in groups with more

148
women expressed a significantly higher level of normative commitment to the organization and generally
higher levels of positive affect than groups with a smaller proportion of women. However, women
indicated the highest propensity to change groups when they worked in entirely female teams. A study by
Graves and Elsass (2005) using teams of undergraduate students, found that sex dissimilarity had no
significant effect on individuals’ expressed attachment to their team, nor on their contributions to the
team’s tasks.

Some of the behaviorally-oriented research that draws on social contact theory as a framework,
discussed above (e.g., Baron, Mittman and Newman, 1991; Konrad and Pfeffer, 1991), is also compatible
with these supply-side theories, since these studies only examined the linear impact of increases in the
proportion of women in an organization: they indicated a positive relation between the proportion of
women and the likelihood of hiring more women. These analyses could thus be interpreted from a
supply-side perspective as well.

Because the logic of both similarity-attraction and social identity theories suggests only a positive
linear relation between the proportion of women in an organization and its attractiveness to potential
hires, we hypothesize that:

Hypothesis 3a: Increases in the proportional representation of women among the faculty of an
institution will have a positive effect on the rate of hiring of women at tenured and untenured
levels.

Much of the empirical work based on these traditions has neglected the potential effects of
differences in the hierarchical distribution of women in the organization. Extending the general logic of
social identity theory, however, it might be expected that a greater representation of women at higher,
more senior levels serves as an especially strong validation of women’s social identity, and makes such
organizations more attractive to female applicants.

Results from Ely’s (1995) research are compatible with this argument. In a study of junior and
senior-level women in two types of law firms, sex-integrated and male-dominated, she found that women
in male-dominated firms were more critical of women’s attributes in relation to their firms’ requirements
for success, and that junior-level women had more negative attitudes towards senior-level women in such
firms than women in sex-integrated firms. She interprets these findings as responses to the devaluation of
women’s social identity in firms with relatively few high-level women. Extending this logic to hiring
decisions, it could be expected that women would find the prospect of employment in organizations with
a comparatively large number of women in higher-level positions affirming of their social identity and
thus be more likely to seek or accept offers in such organizations than in those with fewer women in such
positions. This reasoning is consistent with a number of studies that have shown that having a woman in
a leadership role increases the likelihood of other female hires (Konrad and Pfeffer 1991; Pfeffer, Davis-
Blake, and Julius 1995). Following these arguments, we hypothesize:

Hypothesis 3b: Increases in the proportional representation of women among the

tenured faculty of an institution will have a positive effect on the rate of hiring women
for tenured and untenured positions.

Data, Measures and Analyses

The data used to test our hypotheses are taken from the Integrated Post-Secondary Education
Data System (IPEDS), a battery of surveys sent by the U.S. Department of Education to all institutions

149
whose primary purpose is to provide post-secondary education. Most IPEDS surveys are conducted
annually, but data on faculty hires and composition are collected every other year, as part of an EEOC
survey that began in 1987. We used data from this survey, along with data from other IPEDS files, to
create a five-panel study design, predicting hiring patterns for the period 1988-1997. A random sample of
colleges and universities was drawn from the set of all four-year institutions that offered baccalaureate
and higher-level degrees in 1987, yielding an initial set of 590 institutions for which we collected data.
Because we were interested in hiring processes for tenure-track positions, we dropped institutions that
reported having no tenure-track faculty (tenured or untenured) for the entire period under consideration,
reducing our sample to 510 institutions. We analyzed an average of five observations per institution,
yielding a sample size of 2,550 observations.

We examined three hiring outcome measures in each panel year: the number of women hired by
a university into any tenure-track faculty position, the number hired into untenured positions, and the
number hired into tenured positions. On average, universities hired approximately 10 new faculty
members in each year of our study, of which 3-4 were women, though this varied considerably across
universities. Approximately 90 percent of all hires were for untenured positions.

One of our key predictors was women’s overall representation in the organization; to measure
this, we included the ratio of women faculty to total faculty. To account for the curvilinear effects posited
by Kanter (1977) and Blalock’s (1967) theories, we also included the squared term of this measure in our
models. Women’s proportional representation in higher-status, more powerful positions was measured by
the ratio of tenured women faculty to total tenured faculty. Once again, we included the square of this
variable to account for possible curvilinear effects. Because most independent variables are lagged by
one year, and the data on staffing were collected by IPEDS only in odd-numbered years, we took the
average of two consecutive odd years to create the lagged proportional measures.

In addition to the compositional measures, we included a number of variables that are likely to
influence hiring patterns or the demographic composition of an organization. Since organizations hiring
larger numbers of faculty at a given level are more likely to hire more women at that level, ceteris
paribus, we included the total number of faculty hired overall in predicting the number of total women
hired, the number of hires at untenured levels in models predicting the number of untenured women hired,
and the number of faculty hired at tenured levels in models predicting the hiring of tenured women.
Unlike other independent variables, which were lagged by a year, these measures were contemporaneous
with the dependent variables.

Since not every institution made hires each year (on average, about 80 percent of the schools
made one or more hires annually), we addressed the problem of sampling bias by using a version of
Heckman’s (1979) correction for sample selectivity (see also Berk 1983). We specified a probit model
predicting the probability of a school hiring any faculty in a given year, which included measures of
percent tenured faculty, classification as a research institution by the Carnegie classification system, the
size of the student body, the age of the school, and whether or not the school had experienced a decline in
enrollment from 1988 through 1997 of 10% or more. The probit coefficients generated by this analysis
(not shown here, but available on request from the authors) were used to construct a selection bias
estimator, which was included in all models.

A number of other organizational characteristics were also included in the models as control
variables. Larger organizations are apt to be more bureaucratized, and bureaucratization has been shown
to have a positive effect on women’s employment opportunities (Kulis and Miller-Loessi 1992; Reskin et
al. 1999). We used the log of the total number of full-time and part-time graduate and undergraduate
students as our measure of size. Organizations founded at a time when women were largely excluded
from the workplace may continue to be less likely to provide opportunities for women than organizations

150
that were founded when women’s labor force participation had become more institutionalized (Huffman
1999; Reskin et al. 1999). Hence, we included age, measured as the difference between year of founding
and 1987, to control for such imprinting effects (Stinchcombe 1965).

A large proportion of female students may create a constituent-based demand for female faculty
(Reskin et al. 1999; Tolbert and Oberfield 1991). Alternatively (or perhaps in addition), a large
proportion of female students may serve as a proxy for other factors that influence both the gender
composition of an institution and its hiring outcomes, such as a curricular emphasis on fields of study that
are female-dominated. For these reasons, we controlled for the size of the female student body by
including the ratio of the female students to the total enrollment. The dependence of public organizations
on state and federal funding has been argued to reduce discriminatory propensities in such organizations.
Therefore, a dummy variable for public control was included in our models. Studies of the impact of
organizational growth/decline on women’s demographic representation yield contradictory results, with
Szafran (1984) reporting that women bear a disproportionate share of layoffs during periods of decline,
and Stover (1994) reporting an association between departmental decline and the likelihood of having a
female department manager. Thus, tendencies toward gender discrimination may be heightened in times
of constrained resources and/or declining resources may lead to a decreased likelihood that male
applicants will accept employment with such organizations. To address either possibility, we controlled
for organizational decline by including a dummy variable to identify organizations that experienced a
10% or greater decline in student enrollment between 1988 and 1996.

We also controlled for two financial characteristics that may influence both women’s proportional
representation and rates of hiring. It has been argued that organizations with more slack resources allow
members to indulge preferences for employing select groups of workers (Becker 1971; Tolbert and
Oberfield 1991). We controlled for this by including a measure of the market value of an institution’s
endowment. The research emphasis of academic institutions has been postulated to negatively affect
women’s proportional representation among the faculty, because women are presumed to prefer
employment in less-demanding teaching institutions (Kulis and Miller-Loessi 1992; Stafford and Johnson
1974). We controlled for the research emphasis of academic institutions by including a measure of the
proportion of an institution’s total expenses that are devoted to research.

Our dependent variables represent count data; as is common with such data, their distribution is
characterized by a high proportion of zeroes and low numbers and a long right-hand tail. The common
solution to skewed distributions, taking the log of the variables, was not appropriate since some of the
cases have a legitimate value of zero. Analyses also indicated that the variables were characterized by
“overdispersion,” having a conditional variance greater than the conditional mean. Given the
characteristics of the data, negative binomial regression models were used for our analyses (Long 1997);
we used the nbreg procedure from Stata Version 9.0.

Results

Table 1 presents the correlations among the variables used in our analysis, along with their means
and standard deviations. Not surprisingly, the correlations between the total number of new hires and the
number of women hired, overall and at each level, are very high. Likewise, the measures of proportion of
women faculty in an institution overall, and the proportion of women within the ranks of tenured faculty
are also highly correlated (r=.86). The relation between the number tenured women faculty hired and the
number of untenured women is only moderate, however (r=.11).

Because the majority of women were hired for untenured positions, models predicting total
number of women hired (not shown here, but available on request from the authors) and total number of

151
untenured women hired yielded virtually the same results. Hence, we focus on the comparison between
hiring for untenured and tenured positions. Table 2 presents the results of the analyses examining the
impact of the proportional measures of women faculty in an institution on patterns of hiring for untenured
positions.

The first model includes only the control variables as predictors; the second model adds the
measure of the overall proportion of women faculty; the third model includes a squared term for this
proportional measure; the fourth model adds the measure of the proportion of women among the tenured
faculty, and the fifth includes a squared term for the latter, along with the other proportional measures.

Not surprisingly, in the first model, total number of untenured hires is a strong predictor of the
number of women hired into untenured positions; based on the z-score for this coefficient, this is by far
and away the best predictor. Net of this factor, a number of other organizational characteristics still have
significant effects on the rates of hiring women; most of these effects are maintained across different
models.

The size of the endowment has a negative effect on the hiring of women at this level, consistent
with the argument that the existence of slack resources in an organization enables the exercise of
preferences for associating with members of higher status groups (Becker 1972). The proportion of
expenses devoted to research also exerts a negative effect on the hiring of women for untenured positions.
This may be taken as providing some support for the argument that, because of gender and family roles,
women prefer employment in less research-oriented institutions, although it is worth noting that the
effects of this variable are reversed for hires at the tenured level, as will be shown. The proportion of
female students has a very strong positive effect on hiring women for untenured positions in the first
model, though this variable declines in impact considerably once measures of the proportion of women
faculty are introduced. This suggests that this variable is a useful proxy for other factors, such as
disciplinary emphasis, that may increase an institution’s propensity to hire women. Size has a strong
positive effect: larger institutions are more likely to hire women for untenured positions, all else being
equal. Finally, the coefficient for the measure of public control is significant, but contrary to expectations,
negative: private institutions hire women at a higher rate than public institutions, once other factors (such
as size of endowment) are controlled.

In Model 2, the addition of the measure of the total proportion of women on the faculty is
strongly positive, consistent with supply-side predictions (hypothesis 3a). The addition of a squared term
for this measure in Model 3, however, is also significant, suggesting a curvilinear relation that is more
consistent with a social contact approach (hypothesis 2a), which posits a declining impact of group
proportions on cross-group interactions as groups become more equal in size. Since there is no
theoretical reason to expect a declining effect of women’s proportional representation on the
attractiveness of an organization to job candidates, the curvilinearity of this relation suggests supply-side
approaches may be less relevant to explaining hiring outcomes than demand-side.

152
 Table 1

Descriptive Statistics and Correlations

Variables Mean S.D. X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 X11 X12 X13 X14 X15 X16 X17
X1 No. hires,
women 3.99 6.42 1.00
X2 No. hires, total
9.73 16.25 .95 1.00
X3 No. hires,
untenured women 3.67 5.73 .93 .89 1.00
X4 No. hires,
total untenured 8.67 14.17 .88 .94 .94 1.00
X5 No. hires,
Tenured women .33 2.34 .46 .42 .11 .11 1.00
X6 No. hires,
total tenured 1.06 5.50 .53 .52 .20 .21 .95 1.00
X7 Decline
.15 .36 -.08 -.09 -.07 -.08 -.04 -.05 1.00
X8 Endowment
6.08e-07 3.43e-8 .20 .27 .17 .22 .12 .22 -.04 1.00
X9 Research
expenditures .03 .06 .34 .41 .34 .40 .09 .17 -.07 .25 1.00
X10 Pct. women
students .55 .14 -.04 -.09 -.05 -.09 -.00 -.03 -.12 -.09 -.25 1.00
X11 Size
3.48 .50 .52 .53 .54 .54 .10 .17 -.06 .18 .44 -.06 1.00
X12 Public
.41 .49 .24 .24 .26 .26 .02 .05 -.08 -.07 .29 -.05 .57 1.00
X13 Age
111.32 46.75 .13 .16 .14 .16 .02 .06 .01 .30 .12 -.05 .08 -.13 1.00
X14 Pct. women
faculty 28.96 14.02 -.04 -.10 -.05 -.10 .09 .03 -.12 -.10 -.22 .59 -.06 -.10 -.04 1.00
X15 Pct. women
faculty2 1035.13 1014.27 -.09 -.14 -.10 -.14 .00 -.04 -.10 -.09 -.20 .58 -.16 -.15 -.08 .92 1.00
X16 Pct tenured
women faculty 24.04 14.43 -.08 -.13 -.09 -.13 -.01 -.05 -.08 -.09 -.21 .55 -.08 -.06 -.08 .86 .86 1.00
X17 Pct. tenured
women faculty2 786.30 1067.85 -.11 -.14 -.12 -.14 -.02 -.05 -.07 -.07 -.17 .49 -.16 -.13 -.10 .76 .87 .93 1.00

153
 Table 2

Negative Binomial Regression Models Predicting

Untenured Female Faculty Hires

Variable Model 1 Model 2 Model 3 Model 4 Model 5

Constant -2.410*** -2.550*** -3.006*** -3.058*** -3.067***
(.255) (.255) (.264) (.264) (.265)
Total tenured hires .053*** .053*** .052*** .051*** .051***
(.002) (.002) (.002) (.002) (.002)
Decline .018 -.009 -.011 -.002 -.002
(.053) (.053) (.052) (.052) (.052)
Endowment -1.50e-10** -1.43e-10** -1.22e-10** -1.14e-10* -1.13e-10***
(.53)e-10 (.53)e-10 (.52)e-10 (.52)e-10 (.52)e-10
Research expenditures -1.201*** -2.083** -.560*** -.656* -.664*
(.332) (.331) (.335) (.334) (.335)
% Female students .906*** .231 .385+ .425* .429*
(.170) (.209) (.213) (.212) (.213)
Size (log) .556*** .614*** .598*** .624*** .626***
(.073) (.073) (.073) (.073) (.073)
Public -.099* -.080+ -.102* -.078+ -.077+
(.047) (.047) (.046) (.047) (.047)
Age .000 .001 .001 .0001 .001
(.000) (.000) (.000) (.000) (.000)
% Women faculty .011*** .043*** .050*** .051***
(.002) (.005) (.005) (.008)
% Women faculty2 -.046e-2*** -.035e-2*** -.04e-2***
(.007)e-2 (.006)e-2 (.01)e-2
% Tenured women -.040*** -.015*
(.007) (.007)
% Tenured women2 .049e-3
(.097)e-3
Selection bias estimator -2.409*** .177*** .152*** .149*** .149***
(.255) (.025) (.025) (.025) (.025)
LR chi sq. 2266 2295*** 2346*** 2364*** 2364***
Standard errors shown in parentheses
***p<.0001; **p<.01; *p<.05; +p<.10

Along the same lines, in Model 4, the coefficient of the measure of women’s proportional
representation at tenured levels is significantly negative, contrary to hypothesis 3b, which is premised on
the argument that having a relatively large number of women in higher level positions would make an
organization more attractive to female job applicants. This negative effect is more consistent with group
competition arguments, insofar as women’s increasing representation in higher level positions may
provoke concerns about the distribution of power among men and women in the organization, and thus
result in more negative reactions to women. Interestingly, this result holds only when measures of the
total representation of women are included in the model. In other analyses (not shown here, but available
on request from the authors) which included the proportion of women among the tenured faculty as the
only proportional measure in the model, the coefficient for this measure was positive but non-significant.
This suggests that any negative reaction to women’s representation at senior levels depends partly on their
overall representation in the organization.

154
 All proportional measures are contained in Model 5, including a squared term for the proportion
of tenured level women. The latter measure is non-significant, but the coefficient for the percent tenured
women remains significantly negative, while the measures for the overall representation of women among
the faculty continue to indicate a curvilinear relation. Following the logic of competition theory, it may
be that the representation of women among tenured faculty is not generally not large enough to allow the
exercise of countervailing pressure. An alternative interpretation of this result invokes the “queen bee”
syndrome, described by Kanter (1977), which suggests that senior women may play the role of
gatekeepers, actively resisting the addition of more women to the organization. The results of the analysis
of the hiring of tenured women, presented next, make this interpretation less plausible, however.

Table 3 presents counterparts of the models in Table 2, predicting the hiring of women for
tenured positions. Again, the total number of tenured faculty hires has a very strong, significant relation
to the number of female hires at this level. In contrast to the analyses for untenured women, the
coefficient for the measure of decline is significantly negative in these models: institutions with declining
enrollments hire fewer tenured women (net of the effects of total tenured hires). As in the hiring of junior
level women, endowment value also has a significant negative relation to the hiring of women, suggesting
that wealthy institutions appear to be less likely to hire women faculty at either level.

In contrast to the models for hiring untenured women, though, research emphasis has a significant
positive impact on the hiring of senior-level women. The proportion of female students has no effect on
the hiring of tenured women, even when there are no other proportional measures included in the model.
This is in contrast to Table 2, where the size of the female student body was very significant in the first
model. The measure of organizational size also has a significant, positive effect on the hiring of women
faculty at the tenured level. Thus, all else equal, larger schools tend to hire relatively more women for
both untenured and tenured positions. And finally, the significant negative coefficient for public control
again suggests that once other factors are taken into account, the hiring rates of women are lower at public
institutions than private.

In sharp contrast to the pattern of hiring for untenured women, the second and third models
indicate that the overall proportion of women faculty has neither a linear nor curvilinear effect on the rate
at which women are hired for senior-level positions. Only in Model 4 does a small positive effect obtain
for any of the proportional measures: as the proportion of women among the tenured faculty increases,
institutions are more likely to hire women for senior-level positions. One interpretation of this is that as
organizational members become more accustomed to seeing women at higher organizational levels, they
are more willing to hire women at these levels; this is consistent with a social contact perspective. This
effect disappears, however, in Model 5, including all proportional measures, perhaps as a result of
multicollinearity in these measures. Overall, these results suggest that the hiring of women at tenured
levels is much less influenced by the sex composition of organizations than hiring at lower-level,
untenured positions.

Summary and Conclusions

Knowledge of the forces that shape organizational hiring practices is a key component of
understanding gender segregation and desegregation at occupational levels, since women’s entry into
most occupations rests on organizations’ hiring processes. However, while workplace gender integration
at a higher level of analysis is a function of processes operating at lower levels, as others have noted
(Baron and Bielby 1985), it does not require corresponding integration at lower levels. That is,
occupations can become desegregated, while members of those occupations continue to be relatively
segregated within individual organizations. Thus, analyses at the organizational and within-
organizational levels are crucial to fully understanding larger patterns of workplace change.

155
 Table 3

Negative Binomial Regression Models Predicting Tenured Female Faculty Hires

Variable Model 1 Model 2 Model 3 Model 4 Model 5

Constant -7.054*** -7.113*** -7.008*** -7.218*** -7.213***
(.952) (.957) (.985) (.985) (.987)
Total untenured hires .330*** .328*** .328*** .330*** .330***
(.025) (.025) (.025) (.026) (.025)
Decline -.584* -.579* -.594* -.594* -.594*
(.264) (.264) (.264) (.265) (.265)
Endowment -5.82e-10** -5.74e-10** -5.75e-10** -592e-10** -5.93e-10**
(1.96)e-10 (1.96)e-10 (1.97)e-10 (1.97)e-10 (1.97)e-10
Research expenditures 2.614** 2.655** 2.650* 2.815** 2.828**
(1.118) (1.118) (1.137) (1.139) (1.149)
% Female students .514 .249 .250 .114 .099
(.699) (.831) (.831) (.839) (.856)
Size (log) 1.212*** 1.235*** 1.234*** 1.242*** 1.242***
(.260) (.264) (.264) (.265) (.265)
Public -.561** -.553** -.553** -.590** -.592**
(.195) (.195) (.195) (.197) (.198)
Age .001 -.001 -.001 .000 .000
(.002) (.002) (.002) (.002) (.002)
% Women faculty .004 .004 -.005 .006
(.007) (.018) (.019) (.028)
% Women faculty2 .006e-3 -.017e-2 -.01e-2
(.226)e-3 (.025)e-2 (.04)e-2
% Tenured women .023+ .026
(.013) (.027)
% Tenured women 2 -.031e-3
(.036)e-3
Selection bias estimator -.017 -.024 -.023 -.021 -.022
(.087) (.088) (.089) (.089) (.089)
LR chi sq. 757 757*** 757*** 761*** 761***
Standard errors shown in parentheses
***p<.0001; **p<.01; *p<.05; +p<.10

The academic profession, which has witnessed a marked decrease in sex segregation over the last
several decades, provides a good illustration of this. Women have entered academia at a growing rate
over the last 40 years, increasing their representation from less than 20% of all full-time college and
university faculty in 1960 to nearly a third by 1997 (U.S. Department of Education 2000). But the rate of
desegregation of the professoriate has been far from even across colleges and universities. To date,
women are much better represented in the faculty of smaller, less prestigious, less research-oriented, and
lower-paying institutions than in more elite institutions that often have much higher levels of faculty
compensation (Jacobs 1996). For example, recent data indicate that women constitute, on average, nearly
40 percent of faculty members in liberal arts institutions, about 35 percent in masters-granting
institutions, but only a little more than a quarter of those at elite, research universities (U.S. Department
of Education 2000). Our research seeks to explain such patterns by examining the organizational factors
that affect rates of hiring women for faculty positions, focusing in particular on the impact of the existing

156
representation of women in the faculty on hiring outcomes. A key general finding is that this impact
varies by the hierarchical level at which hires are made.

Partly in line with supply-side theories, our study showed that organizations with a greater
proportion of women are more likely to hire more women for untenured positions. Our analyses indicated
that this relation is curvilinear, however, as theories based on social contact precepts suggest. However,
we also found that increases in the proportion of women among the tenured faculty were associated with
decreases in the rate of hiring untenured women, which is at odds with both supply-side and social contact
theories. This result is more compatible with a group competition approach, stressing economic and
political concerns that can be triggered among majority group members by the growth of a minority
group. In the case of organizations, it may be that such concerns are less likely to surface when increases
in the proportion of women are mostly confined to lower-level positions, and likely to be activated only
when women move into higher-level positions.

The results of the analyses of hiring for tenured positions, on the other hand, indicated that
demographic composition had very little impact on the rates of hiring women into such positions. Only
increases in the proportion of tenured women had a marginally positive effect on hiring women for
tenured positions. One interpretation of this effect is that hiring women for tenured positions is affected
by the extent to which organizational members are accustomed to having women in such positions; this is
consistent with the general logic of social contact theories. Overall, though, demographic composition
appears to play only a small role in the hiring of women into more senior positions. In part, the lack of
effect may reflect the fact that tenured-level hires are relatively rare events (on average, universities made
one tenured hire per year, and only about one in four of these was a woman). It may also be that such
hires are shaped more by year-to-year variations in the pool of senior faculty who are known to be
available and by organizational resources that enable such hires, than by demography-related group
dynamics.

Our results do indicate that the hiring of women faculty is affected by a number of other
organizational characteristics, in addition to demographic composition. Schools with larger endowments
were significantly less likely to hire women, for untenured as well as for tenured positions. Becker’s
(1972) notion of “tastes for discrimination” provides one framework for interpreting this result – insofar
as endowments provide organizations with more slack resources, highly endowed institutions are more
likely to have the ability to pay the premiums required for indulging in preferences for higher status
members. Controlling for the size of the endowment, private schools are somewhat more likely to hire
women, at both ranks. Again interpreting this from a resource perspective, more uncertain revenue flows
at private institutions may make them more inclined to hire women, partly because their average
compensation levels are less than men (Jacobs 1996).

On the other hand, the impact of research emphasis on institutions’ propensity to hire women
appears to vary by level of hire: institutions that are more research-oriented are significantly less likely to
hire women for untenured positions, but more likely to hire them for tenured positions. Supply-side
approaches suggest that the demands of research institutions make them less attractive to women who
must cope with greater domestic responsibilities (Kulis and Miller-Loessi 1992). In this context, the
gender differences in hiring by rank could reflect the fact that junior-level women are apt to be at a life
stage where family responsibilities are greatest; hence they prefer not to accept employment in research-
intensive universities. Since research universities have a relatively low proportion of women among their
faculty overall, they may devote more effort to hiring senior-level women.

Differences also appeared in the impact of the proportion of female students in an institution on
the rates of hiring women for untenured and tenured positions: it had no effect on hiring at the senior-
level, but generally a significant positive effect on hiring at the junior-level. In general, it appears that

157
hiring women for senior-level positions is less driven by concern with filling specific positions (and thus
affected less by the disciplinary composition of the institution, which is likely to be linked to the
proportion of female students) and more by discretionary considerations – e.g., whether desired senior
faculty are available for hiring, and whether the institution has the resources to make such hires.

The findings from our study suggest that future work on demographic composition needs to pay
closer attention to the way in which the distribution of women at different hierarchical levels in an
organization affects patterns of hiring and promotion of other women. In addition, more research on the
relationship between changes in the distribution of women across ranks and relevant individual attitudes
and perceptions of both men and women would enhance both understanding and predicting the impact of
demographic changes. More direct exploration of the micro-level mechanisms implied by the theories
discussed here could lead development of these perspectives as well as to better empirical tests (Reskin
2003).

The patterns found in our research suggest a number of potentially fruitful avenues for further
research. For example, while our results failed to provide support for social identity theory’s predictions,
detailed information on individual’s perceptions of opportunity across demographically-diverse
organizations might suggest refinements to the theory. Both similarity-attraction theory and social
identity theory suggest that women applicants to an organization are encouraged by the presence of other
women, with social identity theory specifying the particular influence of women in higher-level ranks.
But it may be that at some threshold level of representation, potential women applicants come to view an
organization as being a “female-ghetto” (O’Farrell and Harlan 1984), a situation that might counteract
similarity-attraction processes and discourage women from applying. Evidence of the existence of this
would suggest a curvilinear pattern to social identity theory.

A key assumption of a social contact perspective is that the rate of cross-group interactions
increases as a minority group increases in size, but there may be other social forces that counteract the
increased opportunities for interaction. Empirical research aimed at documenting this assumption would
provide important support for this perspective. As a numerical minority group increases in size, the
statistical probabilities of intra-group interaction also rise. Whether this affects relations within gender
groups, and whether such effects are similar for men and women also merits consideration (Riordan 2000;
Tolbert, Graham, and Andrews 1999).

Finally, how the proportion of women at higher levels of an organization affects men’s and
women’s perceptions of opportunities for advancement, attachment to the organization, overall job
satisfaction and other such outcomes deserves more attention. The limited number of women at higher
ranks in most organizations has constrained such research to date, but there is some evidence that the
‘glass ceiling’ is slowly thinning; thus, increasing numbers of women at higher levels of organizations
should facilitate our ability to explore such questions in more depth.

158
 References

Ashforth, Blake E. & Mael, Fred, “Social Identity Theory and the Organization,” Academy of
Management Review, 14(1), (January 1989), 20-39.
Bacharach, Samuel B. & Bamberger Peter A., “Diversity and the Union: The Effect of Demographic
Dissimilarity on Members’ Union Attachment,” Group and Organization Management, 29 (3), (June
2004), 385-410.
Baron, James N. & Bielby, William T., “Bringing the Firms Back In: Stratification, Segmentation and the
Organization of Work,” American Sociological Review, 45(5), (October 1980), 737-765.
Baron, James N. & Bielby, William T., “Organizational Barriers to Gender Equality: Sex Segregation of
Jobs and Opportunities,” Pp. 233-251 in Gender and the Life Course, edited by Alice S. Rossi, New
York: Aldine, 1989.
Baron, James N., Mittman, Brian S. & Newman, Andrew E., “Targets of Opportunity: Organizational and
Environmental Determinants of Gender Integration with the California Civil Service, 1979-1985,”
American Journal of Sociology 96(6), (May 1991), 1362-1401.
Becker, Gary S., The Economics of Discrimination, Second Edition, Chicago IL: University of Chicago
Press, 1971.
Berk, Richard A., “An Introduction to Sample Selection Bias in Sociological Data,” American
Sociological Review, 48(3), (June 1983), 386-398.
Blalock, Hubert M. Jr., Toward a Theory of Minority-Group Relations, New York: Wiley, 1967
Blau, Peter M., Inequality and Heterogeneity, New York: Free Press, 1977.
Blum, Terry C., Fields, Dail L., & Goodman, Jodi S., “Organization-Level Determinants of Women in
Management,” Academy of Management Journal, 37(2), (April 1994), 241-268.
Bygren, Magnus & Kumlin, Johanna, “Mechanisms of Organizational Sex Segregation,” Work and
Occupations, 32(1), (February 2005), 39-65.
Byrne, Donn, The Attraction Paradigm, New York: Academic Press, 1971.
Cassirer, Naomi R., “Race Composition and Earnings: Effects by Race, Region and Gender,” Social
Science Research, 25(4), (December 1996), 375-399.
Chattopadhyay, Prithviraj, Tluchowska, Malgorzata, & George, Elizabeth, “Identifying the Ingroup: A
Closer Look at the Influence of Demographic Dissimilarity on Employee Social Identity,” Academy of
Management Review, 29(2), (April 2004), 180-202.
Chatman, Jennifer A., Polzer, Jeffrey T., Barsade, Sigal G., & Neale, Margaret A., “Being Different Yet
Feeling Similar: The Influence of Demographic Composition and Organizational Culture on Work
Processes and Outcomes,” Administrative Science Quarterly, 43(4), (December 1998), 749-780.
Chatman, Jennifer A. & O’Reilly, Charles A., “Asymmetric Reactions to Work Group Sex Diversity
Among Men and Women,” Academy of Management Journal, 47(2), (April 2004), 193-208.
Cohen, Lisa. E., Broschak, Joseph P., & Haveman, Heather A., “And Then There Were More? The
Effect of Organizational Sex Composition on the Hiring and Promotion of Managers,” American
Sociological Review, 63(5), (October 1998), 711-727.
Elvira, Marta M. & Cohen, Lisa E., “Location Matters: A Cross-Level Analysis of the Effects of
Organizational Sex Composition on Turnover,” Academy of Management Journal, 44(3), (June 2001),
591-605.

159
Ely, Robin J., “The Effects of Organizational Demographics and Social Identity on Relationships Among
Professional Women,” Administrative Science Quarterly, 39(2), (June 1994), 203-238.
Ely, Robin J., “The Power in Demography: Women’s Social Constructions of Gender Identity at Work,”
Academy of Management Journal, 38(3), (June 1995), 589-634.
Gappa, Judith M. & Leslie, David W., The Invisible Faculty, California: Jossey-Bass Publishers, 1993.
Gorman, Elizabeth H., “Gender Stereotypes, Same-Gender Preferences, and Organizational Variation in
the Hiring of Women: Evidence from Law Firms,” American Sociological Review, 70(4), (August 2005),
702-728.
Graves, Laura M. & Elsass, Priscilla M., “Sex and Sex Dissimilarity Effects in Ongoing Teams: Some
Surprising Findings,” Human Relations, 58(2), (February 2005), 191-221.
Heckman, James J., “Sample Selection Bias as a Specification Error,” Econometrica, 45(1), (January
1979), 153-161.
Huffman, Matt L., “Who’s in Charge? Organizational Influences on Women’s Representation in
Managerial Positions,” Social Science Quarterly, 80(4), (December 1999), 738-756.
Jacobs, Jerry A., “Women’s Entry into Management: Trends in Earnings, Authority, and Values Among
Salaried Managers,” Administrative Science Quarterly, 37(2), (June 1992), 282-301.
Jacobs, Jerry A., “Gender Inequality in Higher Education,” Annual Review of Sociology, 22, (1996), 153-
185.
Johnson, George E. & Stafford, Frank P., “The Earnings and Promotion of Women Faculty,” American
Economic Review, 64(6), (December 1974), 888-903.
Kanter, Rosabeth Moss, “Some Effects of Proportions on Group Life: Skewed Sex Ratios and Responses
to Token Women,” American Journal of Sociology, 82(5), (March 1977), 965-990.
Kanter, Rosabeth Moss, Men and Women of the Corporation, New York: Basic Books, 1993.
Konrad, Allison M. & Pfeffer, Jeffrey, “Understanding the Hiring of Women and Minorities in
Educational Institutions,” Sociology of Education, 64(3), (July 1991), 141-157.
Kraus, Vered & Yonay, Yuval P., “The Effect of Occupational Sex Composition on the Gender Gap in
Workplace Authority,” Social Science Research, 29(4), (December 2000), 583-605.
Kulis, Stephen S., “Gender Segregation Among College and University Employees,” Sociology of
Education, 70(2), (April 1997), 151-173.
Kulis, Stephen S. & Miller-Loessi, Karen, “Organizations, Labor Markets, and Gender Integration in
Academic Sociology,” Sociological Perspectives, 35(1), (Spring 1992), 93-117.
Long, J. Scott, Regression Models for Categorical and Limited Dependent Variables, Thousand Oaks,
CA: Sage Publications, 1997.
Martin, Patricia Y. & Harkreader, Shelia, “Multiple Gender Contexts and Employee Rewards,” Work and
Occupations, 20(3), (August 1993), 296-336.
Mittman, Brian S., “Theoretical and Methodological Issues in the Study of Organizational Demography
and Demographic Change,” Research in the Sociology of Organizations, 10 (1992): 3-53.
O’Farrell, Brigid & Harlan, Sharon L., “Job Integration Strategies,” Pp. 267-291 in Sex Segregation in the
Workplace: Trends, Explanations, Remedies, edited by Barbara F. Reskin, Washington, DC: National
Academy Press, 1984.

160
Pazy, Asya & Oron, Israela, “Sex Proportion and Performance Evaluation Among High-Ranking Military
Officers,” Journal of Organizational Behavior, 22(6), (September 2001), 689-702.
Pfeffer, Jeffrey, “Organizational Demography,” Research in Organizational Behavior, 5 (1983): 299-357.
Pfeffer, Jeffrey & Davis-Blake, Alison, “The Effects of the Proportion of Women on Salaries: The Case
of College Administrators,” Administrative Science Quarterly, 32(1), (March 1987): 1-24.
Pfeffer, Jeffrey, Davis-Blake, Alison & Julius, Daniel J., “AA Officer Salaries and Managerial Diversity:
Efficiency Wages or Status?” Industrial Relations, 34(1), (January 1995), 73-93.
Reskin, Barbara F., “Including Mechanisms in Our Models of Ascriptive Inequality,” American
Sociological Review, 68(1), (February 2003), 1-21.
Reskin, Barbara F., McBrier, Debra B. & Kmec, Julie A., “The Determinants and Consequences of
Workplace Sex and Race Composition,” Annual Review of Sociology, 25 (1999): 335-361.
Reskin, Barbara F. & Roos, Patricia A., Job Queues, Gender Queues: Explaining Women’s Inroads into
Male Occupations, Philadelphia: Temple University, 1990.
Riordan, Christine M., “Relational Demography Within Groups: Past Developments, Contradictions, and
New Directions,” Research in Personnel and Human Resources Management, 19, (2000), 131-173.
Sacco, Joshua M. & Neal Schmitt, “A Dynamic Multilevel Model of Demographic Diversity and Misfit
Effects,” Journal of Applied Psychology, 90(2), (March 2005), 203-231.
Shenhav, Yehouda & Haberfeld, Yitchak, “Organizational Demography and Inequality,” Social Forces,
71(1), (September 1992), 123-143.
StataCorp, Stata Statistical Software: Release 9.0. College Station, Texas: Stata Corporation, 2005.
Stinchcombe, Arthur L., “Social Structure and Organizations,” Pp. 142-193 in Handbook of
Organizations, edited by James G. March, Chicago: Rand McNally, 1965.
Stover, Dana L., “The Horizontal Distribution of Female Managers Within Organizations,” Work and
Occupations, 21(3), (August 1994), 385-402.
Szafran, Robert F., “Female and Minority Employment Patterns in Banks,” Work and Occupations, 11(1),
(February 1984), 55-76.
Tajfel, Henri, Differentiation Between Social Groups, New York: Academic Press, 1978.
Tajfel, Henri, Social Identity and Intergroup Relations, Cambridge: Cambridge University Press, 1982.
Tolbert, Pamela S., “Organizations and Inequality: Sources of Earnings Differences Between Male and
Female Faculty,” Sociology of Education, 59(3), (July 1986), 227-35.
Tolbert, Pamela S., Graham, Mary E. & Andrews, Alice O., “Group Gender Composition and Work
Group Relations,” Pp. 179-202 in Handbook of Gender and Work, edited by Gary N. Powell, Thousand
Oaks, CA: Sage, 1999.
Tolbert, Pamela S. & Oberfield, Alice A. “Sources of Organizational Demography: Faculty Sex Ratios in
Colleges and Universities,” Sociology of Education, 64(4), (October 1991), 305-315.
Tolbert, Pamela S., Simon, Tal , Andrews, Alice O. & Rhee, Jaehoon, “The Effects of Gender
Composition in Academic Departments on Faculty Turnover,” Industrial and Labor Relations Review,
48(3), (April 1995), 562-579.
Tsui, Anne S., Egan, Terri D. & O’Reilly, Charles A., “Being Different: Relational Demography and
Organizational Attachment,” Administrative Science Quarterly 37(4), (December 1992), 549-579.

161
Turner, John C., Rediscovering the Social Group: A Self-Categorization Theory, Oxford: Basil Blackwell,
1987.
U.S. Department of Education, National Center for Education Statistics. Digest of Education Statistics,
2000. Washington, DC: 2000.
U.S. Department of Education, National Center for Education Statistics. Integrated Post-Secondary
Education Data System. Washington DC: 1988-1997.
Williams, Katherine Y. & O’Reilly, Charles A., “Demography and Diversity in Organizations: A Review
of 40 Years of Research,” Research in Organizational Behavior, 20, (1998), 77-140.
Wharton, Amy S., “The Social Construction of Gender and Race in Organizations: A Social Identity and
Group Mobilization Perspective,” Research in the Sociology of Organizations 10, (1992), 55-84.
Wharton, Amy S. & Baron, James N., “So Happy Together? The Impact of Gender Segregation on Men
at Work,” American Sociological Review, 52(5), (October 1987), 574-87.

162
ASAC 2006 Parshotam Dass
Banff, Alberta Faculty of Management
I. H. Asper School of Business
University of Manitoba

EMERGENCE OF HUMAN DIVERSITY STRATEGIES:

THE ROLE OF THE NETWORK AS A MECHANISM

The objective of this paper is to explore an organic view of human resource

diversity strategies in organizations. We propose a theoretical model that uses the
network as a mechanism to facilitate the interaction of cognitive perspectives of
diversity with external and internal pressures for diversity leading to the emergence
of human diversity strategies.

A cantaloupe changes its color by watching another cantaloupe.-An Indian Proverb

Relevant to the theme of the conference, there is a dire need to reach new heights in the area of
diversity practically, conceptually, ethically, and empirically. Our knowledge and actions regarding diversity
have tremendous implications for the quality of life for the people and the society at large. Diversity
researchers have made tremendous progress over the years (e.g., Ely & Thomas, 2001; Ibarra, 1992, 1993,
1995; Ibarra, Kilduff, & Tsai, 2005). In this paper, we intend to build on this progress by taking a strategic
perspective to diversity. We propose the question: How do human diversity strategies emerge in
organizations? Three trends in the broad literature on management may facilitate in the process of building a
conceptual model and developing a better understanding to answer this question. First, within the field of
strategic management, a modern concept of strategy has emerged that takes an organic perspective (Farjoun,
2002), which goes beyond classical, mechanistic models of designing strategies to adaptive, learning models
of “growing” strategies. Second, there is a surge in the recent literature on social mechanisms that represent a
mid-range theoretical focus and may explain social phenomena (Bunge, 2004; Davis & Marquis, 2005;
Hedstroem & Swedberg, 1998). Third, during the 1990s and beyond, research on personal, organizational
networks and social capital has accumulated. These networks provide opportunity, motivation and ability
(Adler & Kwon, 2002) and may enable us to explicate their use as a mechanism. Networks are recognized
because of their use in collecting and disseminating information; learning, creating and communicating
knowledge and other resources; opportunity identification, evaluation and capture; and decision-making and
strategic choice, among others (Adler & Kwon, 2002; Gulati, 1998).

The objective of this paper is to explore an organic view of human resource diversity to answer the
question posed above. In this paper, we offer a theoretical model that uses the network as a mechanism, which
facilitates the interaction of individual-level cognitive perspectives of diversity and external and internal
pressures for diversity and homogeneity leading to emergence of human diversity strategies in organizations.
We use the term “growing strategies” to represent an organic perspective, consistent with the metaphor of
organism, as compared to designing strategies, which reflects a machine metaphor. The term mechanism
though derived from mechanistic roots is now used and accepted in an organic sense as well (Bunge, 2004).

The paper is organized as follows. First, we review the relevant background literature that reflects a
gap as well as points to how to fill the gap. Second, we propose a model of human diversity strategy
emergence and change in organizations and suggest propositions and develop their rationale. Finally, we
conclude by drawing out implications for future theory, research and practice.

163
ASAC 2006 Linda Scourtoudis (Student)
Banff, Alberta Lorraine Dyke
Eric Sprott School of Business
Carleton University

DO WOMEN MANAGERS BENEFIT FROM WOMEN ONLY MANAGEMENT TRAINING?

AN EMPIRICAL EXAMINATION OF THE IMPACTS OF ONE PROGRAM

Research shows that women continue to face barriers to reaching senior management
(Catalyst, 2004). Women often receive less career encouragement, have more difficulty
obtaining mentors and developing networks, and receive fewer training opportunities than
their male counterparts (Lyness and Thompson, 1997; Tharenou, Latimer and Conroy, 1994).
Management training programs for women have grown in popularity in recent years as a
means to help women overcome these barriers. One specific program of this type is the
Management Development Program for Women (MDPW) at Carleton University in which
participants are able to develop management skills in a supportive all-female environment.

The primary focus of this research was to determine if the MDPW offered at Carleton
University has been able to promote positive change for women with respect to certain
variables that are associated with enhanced career success. This research also examined
whether the MDPW has been able to promote positive career outcomes for women.

Both longitudinal and retrospective data were collected. The first survey was a follow-up
post-test for the twenty-five MDPW alumnae who participated in a 1999 study of their
experiences. The second questionnaire sought comparative judgments from the remaining
two hundred and forty-five alumnae (a retrospective study). Both surveys used previously
validated measures. The response rate was 50%.

This study found that for the women who participated in the MDPW, the program had a
significant positive impact in their lives. The program played a role in helping these women
raise their sense of self-worth and increase their belief in their ability to manage their own
careers. These women have also taken a more active approach to career planning. Having
participated in the MDPW seems to have also shaped their self-perceptions of job
competence and career satisfaction. Respondents have also advanced notably in their careers.
Eighty-three percent of respondents in this study had received at least one promotion since
completion of the MDPW and the majority had experienced increases in responsibility and
challenge in their jobs.

Training is one means to achieve workplace equity. The results of this research suggest that
the MDPW is helping women to believe in themselves by giving them the necessary tools to
overcome some of the external barriers they face. The findings from this research should
encourage other women to participate in similar programs and encourage managers to support
women employees’ participation in similar programs. Educators should give serious
consideration to the development and refinement of women only management training
programs.

164
ASAC 2006 B.E. (Bernie) Williams
Banff, Alberta Jill Finley (Student)
Faculty of Management
University of Lethbridge

STRATEGIC RESPONSES TO COMPLEX STAKEHOLDER DEMANDS: A CASE STUDY OF

AN ABORIGINAL NONPROFIT ORGANIZATION

Many nonprofit organizations today are facing increasingly complex demands,

as well as financial inadequacies and resource constraints. This case study
uncovered unique leadership characteristics as the key success factor that
contributed to the development of strategic responses designed to mediate
changing environmental conditions and respond to the demands of multiple, and
often competing, stakeholders.

Canada’s Nonprofit Sector

Nonprofit organizations are a key component of Canadian society that serve a variety of needs
and provide a wide range of services. Change has been constant throughout the history of Canada’s
nonprofit sector. The nonprofit sector has been continuously moving from its backstage place among the
institutional components of Canadian society towards a more high profile, mid-stage position (Reed &
Howe, 1998). This shift has been fueled by government decisions to downsize and delegate some of their
social responsibilities. The nonprofit sector not only commences their duties at the fringes of government
programs, but it plays an important role in communicating the government’s Voluntary Sector Task
Force’s overall goal of ‘enhancing the quality of life for Canadians’ (Government of Canada, 1999).
Understandably then, the nonprofit sector is seen as the sector that fills in where other systems have fallen
short (Salamon, 1987). In an epoch characterized by significant societal changes that include economic
swings, sustainable development challenges, struggle for, and maintenance of a national identity and
reduced legitimacy to Canada’s democratic institutions, studying nonprofit organizations can offer public
policy implementers an understanding of the forces that influence the strategy and structure of nonprofit
organizations.

The nonprofit sector has been facing increasing complexity and responding to change for many
years. The introduction of the neo-conservative political ideology in the 1980’s provided a focal point to
the commencement of a new reality--a reduction of federal funding designated for social program
development and enhancement. In subsequent years, this philosophy filtered down to the provincial and
municipal levels, leaving nonprofit organizations to deal with the resulting social and economic
restructuring now occurring at all three levels of government. Nonprofit organizations are facing a new
climate characterized by unprecedented competition for donations, combined with shrinking government
support (Bush, 1992; Weisbrod, 1988).

165
ASAC 2006 Karen A. Somerville (Student)
Banff, Alberta Lorraine S. Dyke
Eric Sprott School of Business
Carleton University

GENDER AND AGE DIFFERENCES IN ORGANIZATIONAL POLITICS1

Results from a sample of 81 executives indicate that women are not averse to
political behaviour and behave just as politically as men. There are, however,
some gender differences in the content of respondents’ political behaviour. Older
executives were less likely to regard politics as harmful but no more likely to
engage in political behaviour.

Organizational politics are a reality for all organizations (Ferris, King, Judge & Kacmar, 1991).
In some cases, politics may get more attention than other important aspects of organizational life, such as
clients or leadership (Zaleznik, 1989). Although politics are often regarded as a negative organizational
dynamic (Mainiero, 1994), politics can be positive for both organizations and for individuals (Ferris &
Kacmar, 1992).

The necessity of political behaviour appears to increase at higher organizational levels (Gandz &
Murray, 1980), as usually the organizational climate becomes more political at these levels (Pfeffer,
1992). Further, some research has found differences in how men and women view organizational power
and politics (e.g., Hardesty & Jacobs, 1986; Helgeson, 1990; Legare, 1995).

However, to date, there has been relatively little empirical research relating to political behaviour.
In addition, despite the aging North American workforce (Kanfer & Ackerman, 2004), very little research
has considered the relationship of age to political attitudes and behaviours (Treadway, Ferris, Hochwarter,
Perrewe, Witt, & Goodman, 2005).

The purpose of this paper is to examine group differences in individual political behaviour and
attitudes amongst a group of employees who are expected to be politically active and aware – senior
managers. Specifically, we examined whether executives felt that politics were necessary or harmful to
organizations and which political strategies they employed. We also explored gender and age differences
in political attitudes and behaviour.

1
The authors wish to acknowledge the research assistance of Shirley McKey.

166
ASAC 2006 Barbara Orser
Banff, Alberta School of Management
Catherine Elliott
Faculty of Education
Joanne Leck
School of Management
University of Ottawa

GENDER AS A MODERATOR OF EXECUTIVE CAREER SUCCESS 1

Previous modeling of objective and subjective executive career success specify gender as one covariate,
among many. Theoretical reasons why gender is better specified as a moderator variable are advanced
and confirmed through empirical testing. Findings reveal that sector, age, and experience are factors that
are moderated by gender in the determination of executive career success.

The objective of this research program is to examine how gender influences executive career success. As
a theoretical framework, this paper builds on the conceptual model of career success advanced by Judge,
Cable, Boureau and Bretz (1995). Where Judge and his colleagues recognized gender among
determinants of career success, they did not treat gender as a moderating variable nor did they examine
interaction effects of gender with other determinants of career success. This paper addresses this potential
shortcoming with a view to gaining a yet better understanding of ways in which managers achieve success
and of the role of gender.

There is considerable research about barriers to career advancement that are specific to women. For
example, Gerkovich Griffith et al., (1989) and Ragins, et al., (1998) document perceived barriers to
success specific to senior women such that they leave organizations citing lack of opportunity and
compensation. These have been recognized by some employers that have undetakern remedial action.
Examples have been cited by Lobel (1991), Black (2001), Ellis and Sonnenfeld (1994), and Brady and
McLean (2002). The model postulated by Judge et al. was a substantive improvement over earlier
attempts to model career success, attempts that have been described as “piecemeal” and atheoretical
(Gattiker & Larwood, 1989). Nonetheless the model derived by Judge et al. still does not account for
documented gender differences in managers’ values, aspirations and perceptions of success (Gilligan,
1982; Gallos, 1989; Duxbury, Dyke & Lam, 1999, 2000; Parker & Chusmir, 1992; Sturges, 1999; Dyke
and Orser, 2005).

To the extent that the role of gender in the model is not properly specified, gender influences or biases
may obscure, or be obscured by, seemingly neutral organizational processes such as selection criteria and
job prerequisites (Levey et al., 1999; Egan and Benedick, 1994). Likewise, failure to account properly for
gender effects may support gender stereotypes about women’s ambition and workplace commitment
(Bank of Montreal, 1991) and legitimatize wage disparity.

This research is of interest to employees, employers and academics who seek to understand factors that
influence career success. The information will assist in the development of training aimed at those
making career choices and to women who are weighing the pros and cons of a career or job change.

1
The authors wish to acknowledge the financial support of the Social Sciences and Humanities Research Council of
Canada (Grant No. 410-99-0824; (PI) XX, (CI) XX, and the research assistance of Kelly Love, Alisa Bhegani,
Courtney Kada and Bernie Rawlings.

167
Finally, the study may also be of interest to policy makers and feminist organizations that seek to ensure
equitable employment opportunities. This research therefore builds on Canadian2 and international
research3 that seeks to explain why women are significantly less likely to attain advancement. To do so,
the paper opens with a brief description of the model of executive career success advanced by Judge,
Cable, Boudreau and Bretz (1995). Three theoretical perspectives that have been advanced to help explain
gender differences in career success are then presented. The research methodology and summary findings
follow. The paper concludes with a discussion of findings and study limitations.

2
For example see Bank of Montreal, 1991; Women’s Executive Network, 1999; Duxbury, L., Dyke, L., and Lam, N., 1999 and
2000; Orser, 1997, 2000 and 2001.
3
For example, see Egan and Bendick, 1994; Fine and Bolman, 1998; Ragins et al., 1998; Ragins and Cotton, 1999; Stroh et al.,
1992; Schaubroeck and Lam, 2002; Sach et al, 1992.

168
ASAC 2006 Amy M. Warren
Banff, Alberta Travor C. Brown
Faculty of Business
Memorial University

EMPLOYMENT EQUITY AND BACKLASH: AN EXAMINATION OF THE PERCEPTIONS

OF EMPLOYEES WITHIN THE FEDERAL PUBLIC SERVICE: PRELIMINARY FINDINGS

Employment Equity programs within the Federal government have been put in place to
ensure that higher numbers of traditionally underrepresented groups, such as women, are
employed in the Federal public service. Under the Employment Equity Act, Federal
government departments are to develop and implement Employment Equity programs as well
as submit data yearly illustrating that they have hired employees from traditionally
underrepresented groups. The evidence suggests that there are more women in the Federal
workplace (based on statistics provided by the Treasury Board Secretariat (TBS). The
positions women occupy still tend to be in the administrative category, with only a small
number of women being promoted to management and executive positions (TBS, 2001).

Unfortunately, to date much of the research on such programs has focused on representation
statistics (i.e., number of target group members hired, promoted, etc.) To our knowledge, no
study has examined how Employment Equity programs are perceived by public sector
employees and if these perceptions lead to instances of backlash towards the groups covered
by Employment Equity.

In the present study, we surveyed 217 staff members employed within a Federal public
service department branch, with particular focus on questions concerning employee
perceptions towards Employment Equity programs and backlash. A total of 108 people
responded to the survey (49.8% response rate). Overall the study found that perceptions of
Employment Equity programs are slightly negative and instances of backlash do occur.
Specific findings include that; (1) men who are not members of one of the four designated
employment equity groups are more likely to feel that Employment Equity programs are
unnecessary and more likely to believe that such programs promote reverse discrimination;
(2) older employees are more likely view Employment Equity programs are necessary; (3)
managerial staff are more likely than non-managerial staff to believe that hiring and
promotion practices are merit-based and that people hired under Employment Equity are
qualified; and (4) employees in non-managerial, traditionally male-dominated occupations
were more likely than management to believe that Employment Equity promotes reverse
discrimination when compared to the management group.

The Federal public service is large and studies on other branches could add value to
Employment Equity research. Also a study which focuses more specifically on members of
the three other Employment Equity groups may help gain further insight into the
effectiveness of Employment Equity programs and the perceptions surrounding Employment
Equity. To further investigate why this may be so, focus groups with employees may be
necessary.

169