Vous êtes sur la page 1sur 113

THREE ESSAYS ON THE QUANTIFICATION, VALIDATION, AND APPLICATION OF GRAYS ACCOUNTING VALUES

RAMON P. RODRIGUEZ, JR. International Business Doctoral Program

APPROVED:

Gary P. Braun, Ph.D., Chair

Stephen B. Salter, Ph.D.

Patricia Eason, Ph.D.

Sidney J. Gray, Ph.D.

Patricia D. Witherspoon, Ph.D. Dean of the Graduate School

Copyright

by Ramon P. Rodriguez, Jr. 2009

Dedication

To my wife and children. Thank you Rebecca, R.T., Robert, and Rex for your patience, love, and support during my doctoral program years. You have made the difficult times easier and the good times fantastic. To Brian, Kate, Manuel, and Wayne. Your support and successes have motivated me to continue my journey through tough times. To Fernie R. Aceves. I could not have accomplished this without all of your help. You have been a professional mentor and I have learned more about the practice of accounting during my time with you than during any other time in my career. You have been a wonderful father-in-law providing love, guidance, and support to us. You have also been a treasured friend.

THREE ESSAYS ON THE QUANTIFICATION, VALIDATION, AND APPLICATION OF GRAYS ACCOUNTING VALUES

by

RAMON P. RODRIGUEZ, JR., B.B.A, M.B.A.

DISSERTATION

Presented to the Faculty of the Graduate School of The University of Texas at El Paso in Partial Fulfillment of the Requirements for the Degree of

DOCTOR OF PHILOSOPHY

International Business Doctoral Program THE UNIVERSITY OF TEXAS AT EL PASO May 2009

UMI Number: 3358880

INFORMATION TO USERS

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

______________________________________________________________
UMI Microform 3358880 Copyright 2009 by ProQuest LLC All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.

_______________________________________________________________
ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, MI 48106-1346

Acknowledgements
I would like to acknowledge the support and guidance of my doctoral committee members Gary P. Braun, Ph.D., Stephen B. Salter, Ph.D., Patricia Eason, Ph.D., and Sidney J. Gray, Ph.D. I also gratefully acknowledge Lance Eliot Brouthers, Ph.D., for giving me the opportunity to show that I am worthy to be a member of this learned profession. I thank the PhD Project, the KPMG Foundation, the American Institute of Certified Public Accountants Fellowship for Minority Doctoral Students, the Deans Office, the Department of Accounting, the Department of Marketing and Management, and other members of the College of Business Administration at the University of Texas at El Paso as well as the School of Accountancy and other members of the College of Business at Southern Illinois University Carbondale for financial, intellectual, moral, and professional support during the completion of this dissertation. Most importantly, I am forever indebted to my wife and children, my parents, my sister and her family, my in-laws, and my friends for their constant support and love throughout my careers, my studies and my life.

GO MINERS!

Abstract
This doctoral dissertation creates, validates and uses a quantification of country-level accounting values based on Grays (1988) framework. This dissertation consists of three essays. The first essay reports the theoretical foundation and method for the quantification of the accounting values. The quantification results in reported scores for each of Grays (1988) four accounting values Professionalism versus Statutory Control, Uniformity versus Flexibility, Conservatism versus Optimism, and Secrecy versus Transparency. Accounting values scores are reported for 58 countries. The second essay reports the method and results of validation testing preformed on the accounting values scores. The primary validation method, Multi-Trait MultiMethod Analysis, failed to provide empirical validation to the accounting values scores. A secondary validation method, an ad-hoc research study replication, provided limited empirical validation to the accounting values scores. The third essay uses the accounting values scores as a primary variable in the explanation of differing levels of a countrys adoption of the International Financial Reporting Standards. The essay reports empirical evidence of the effect of accounting values on a countrys level of adoption of the International Financial Reporting Standards; however, this effect seems to be disappearing as more countries adopt the International Financial Reporting Standards.

vi

Table of Contents
Acknowledgements ..........................................................................................................................v Abstract .......................................................................................................................................... vi Table of Contents .......................................................................................................................... vii List of Tables ................................................................................................................................. ix List of Figures ..................................................................................................................................x List of Illustrations ......................................................................................................................... xi Overview ..........................................................................................................................................1 Essay 1 - The Quantification of Grays Accounting Values............................................................4 1.1 1.2 1.3 1.4 Introduction ....................................................................................................................4 Theoretical Framework of Accounting and Culture ......................................................5 Hofstedes Cultural Dimensions ..................................................................................10 Grays Accounting Values ...........................................................................................12 1.4.1 Professionalism versus Statutory Control ..........................................................14 1.4.2 Uniformity versus Flexibility .............................................................................11 1.4.3 Conservatism versus Optimism .........................................................................15 1.4.4 Secrecy versus Transparency .............................................................................15 Quantification of Grays Accounting Values ..............................................................16 1.5.1 Gray (1988) and GLOBE (House et al. 2004) ...................................................21 Mechanics of the Quantification ..................................................................................24 1.6.1 Professionalism versus Statutory Control ..........................................................26 1.6.2 Uniformity versus Flexibility .............................................................................26 1.6.3 Conservatism versus Optimism .........................................................................27 1.6.4 Secrecy versus Transparency .............................................................................27 1.6.5 Standardization of Quantified Scores ................................................................28 1.6.6 Characteristics of Quantified Scores ..................................................................29 1.6.7 Empirical Test of Face Validity .........................................................................32 Discussion and Conclusion ..........................................................................................34

1.5 1.6

1.7

Essay 2 - The Empirical Validation of the Accounting Values Scores .........................................37 2.1 2.2 Introduction ..................................................................................................................37 Multi-Trait Multi-Method Analysis .............................................................................38 2.2.1 Eddie (1990).......................................................................................................39 2.2.2 Salter and Niswander (1995) .............................................................................39 2.2.3 Gray and Vint (1995) .........................................................................................41 vii

2.2.4 2.2.5 2.2.6 2.2.7 2.3

Zarzeski (1996) ..................................................................................................41 La Porta et al. (1998) .........................................................................................43 Hope (2003) .......................................................................................................43 Analysis of Results ............................................................................................45

Replication of Jaggi and Low (2000)...........................................................................48 2.3.1 Variables and Regression Models ......................................................................49 2.3.2 Analysis of Results ............................................................................................54 Conclusion ...................................................................................................................61

2.4

Essay 3 - Using Grays Accounting Values to Explain Deviations from IFRS.............................63 3.1 3.2 Introduction ..................................................................................................................63 Literature Review and Hypothesis Development ........................................................64 3.2.1 Conservatism versus Optimism .........................................................................65 3.2.2 Secrecy versus Transparency .............................................................................67 Research Method .........................................................................................................68 3.3.1 Variables and Regression Models ......................................................................68 3.3.2 Analysis of Results ............................................................................................74 3.3.3 Additional Test of H1 and H2 ............................................................................76 Conclusion ...................................................................................................................81

3.3

3.4

Summary ........................................................................................................................................83 References ......................................................................................................................................87 Appendix 1 Composition of Eddies (1990) Accounting Values Indexes ..................................97 Appendix 2 Correlation of GLOBE and Hofstede Cultural Values Indices ...............................99 Vita.. .........................................................................................................................100

viii

List of Tables
Table 1.1: Matrix of Relationships of Accounting Values with Hofstede's Culture Values ........ 14 Table 1.2: Matrix of Relationships of Accounting Values with GLOBE Culture Values. ........... 22 Table 1.3: Accounting Practices Scores........................................................................................ 23 Table 1.4: Accounting Values Scores ........................................................................................... 25 Table 1.5: Descriptive Statistics, Test of Normality, and Pearson Correlations........................... 29 Table 1.6: Regression Results of Accounting Values Hypothesis Testing ................................... 33 Table 2.1: Multi-Trait Multi-Method Validity Results ................................................................. 46 Table 2.2: Descriptive Statistics on Regression Variables ........................................................... 52 Table 2.3: Jaggi and Low (2000) Replication Regression Results - Full Sample ........................ 55 Table 2.4: Jaggi and Low (2000) Replication Regression Results by Common Law and Code .. 58 Table 3.1: Dependent and Control Variables................................................................................ 70 Table 3.2: Pearson Correlations for Variables in Tests on Absence Measures ............................ 72 Table 3.3: OLS Regression Results .............................................................................................. 75 Table 3.4: Pearson Correlations for Variables in Tests on Use of IFRS ...................................... 79 Table 3.5: Ordinal Probit Regression Results ............................................................................... 80

ix

List of Figures
Figure 1.1: Gray's (1988) Framework of Accounting and Culture. ................................................ 5 Figure 1.2: Perera's (1989) Framework of Accounting and Culture............................................... 6 Figure 1.3: Doupnik and Tsakumis' (2004) Partial Refinement of Gray's (1988) Framework of Accounting and Culture ................................................................................................ 7 Figure 1.4: Synthesis of Previous Frameworks into an institutional Framework of Accounting and Culture .................................................................................................................. 10 Figure 1.5: Histograms of Accounting Values Scores .................................................................. 31

List of Exhibits
Exhibit 2.1: Composition of Salter & Niswanders (1995) Accounting Values Indexes ............. 40 Exhibit 2.2: Composition of Zarzeskis (1996) Disclosure Index ................................................ 42

xi

Overview
This doctoral dissertation creates, validates and uses a quantification of country-level accounting values based on Grays (1988) framework. Two decades after Gray (1988)

synthesized the extant literature to describe a framework of accounting values, a large scale country-level quantification of accounting values still does not exist. A large scale country-level quantification of accounting values would prove useful in a variety of studies. Many researchers have used Grays accounting values framework as theoretical support for their studies in various areas such as decision-making, firm valuation, materiality judgments, voluntary disclosure, and foreign listing requirements (Adams et al. 1993; Archambault and Archambault 2003; Arnold et al. 2001; Bao and Bao 2004; Buck and Shahrim 2005; Chow et al. 1995; Debreceny et al. 2002; Doupnik and Richter 2003; Doupnik and Salter 1995; Hussein 1996; King and Langli 1998; Lynn 1999; Meek et al. 1995; Radebaugh et al. 1995; Roberts and Salter 1999; Salter 1998; Salter et al. 2001; Schultz and Lopez 2001); however, the accounting values proxies used by these studies are problematic for one of two reasons either the study relies on the inclusion of three or more cultural variables as a proxy for a single accounting value or it relies on a newly created proxy that is not tested for validity nor reused in subsequent studies. The quantified accounting values created in this dissertation would mitigate both of these issues by (1) replacing multiple variables as a test of a single effect with one variable that tests a single effect, which should aid in the interpretation of empirical results, and by (2) providing a uniform measure of accounting values that can be used in all accounting values studies, which should aid in the generalizability of empirical results. This dissertation consists of three essays in which a large scale country-level quantification of Grays accounting values is created, tested, and applied.

The first essay reports the theoretical foundation and method for the quantification of the accounting values. Although the quantification of the accounting values is based on Grays (1988) framework, two other frameworks (Doupnik and Tsakumis 2004; Perera 1989) as well as an institutional theory (DiMaggio and Powell 1991, 1983; Oliver 1991; Selznick 1948; Selznick et al. 1969; Scott 1987, 2001) based synthesis of the three frameworks are presented and discussed. The quantification consists of a summative combination of the GLOBE (House et al. 2004) cultural values index in the manner described in Grays (1988) accounting values framework, which is then standardized to a range of 0 through 100 to aid with interpretation. This quantification results in reported scores for each of Grays (1988) four accounting values Professionalism versus Statutory Control, Uniformity versus Flexibility, Conservatism versus Optimism, and Secrecy versus Transparency. Accounting values scores are reported for 58 countries. The results of ordinary least squares multiple regression analyses based on Grays (1988) original four hypotheses provide empirical support for the face validity of the quantified accounting values. The second essay reports the method and results of validation testing preformed on the accounting values scores. The primary validation method, Multi-Trait Multi-Method Analysis, analyzes the correlations between the quantified accounting values scores and accounting values proxies created in six previous studies. This validation method assesses construct validity 1 by testing convergent 2 and discriminant 3 validity. The Multi-Trait Multi-Method Analysis provided only limited support for convergent validity and failed to provide empirical support for the

Construct validity is the totality of evidence about whether a particular operationalization of a construct adequately represents what is intended by theoretical account of the construct being measured (Cronbach and Meehl 1955). 2 Convergent validity is based on the extent to which a measure correlates with different measures with which it is theoretically predicted to correlate (Kotz et al. 1985). 3 Discriminant validity is based on the extent to which an operationalization does not correlate with operationalizations of different constructs (Kotz et al. 1985).

discriminant validity of the quantified accounting values scores. A secondary validation method, an ad-hoc research study replication, provided limited empirical validation to the accounting values scores. The third essay uses the accounting values scores as a primary variable in the explanation of differing levels of a countrys adoption of the International Financial Reporting Standards (IFRS). While dozens of countries now require or permit the use of IFRS or are converging with IFRS (International Accounting Standards Board 2009), the level of adoption varies between countries (Dean and Clarke 2005). Ding et al. (2005) created an index of a countrys Ding et al. (2005) reported the

nonconformity in its domestic GAAP with IFRS.

nonconformities as an absence of a specific rule from domestic GAAP or as an instance of a divergence of an existing domestic GAAP standard from the corresponding IFRS. I use Ding et al.s (2005) measures of the absence of IFRS pronouncements from domestic GAAP as the proxy for the differing levels of a countrys adoption of IFRS. I hypothesize and empirically show that countries that value conservatism have higher levels of deviation from IFRS and that countries that value secrecy have higher levels of deviation from IFRS. While this essay reports empirical evidence of the effect of accounting values on a countrys level of adoption of IFRS, an additional test of the hypotheses with newer data implies this effect may be disappearing as more countries adopt the International Financial Reporting Standards.

Essay 1 - The Quantification of Grays Accounting Values


1.1 Introduction Almost twenty years after Gray (1988) synthesized the extant literature to describe a framework of accounting values, despite several partial studies (Chanchani and Willett 2004; Eddie 1990; Salter and Niswander 1995), a large scale country-level quantification of accounting values still does not exist. This is unlike the underlying framework for Gray (1988), i.e., the cultural work of Hofstede (1980), which includes the creation of a country level quantification of cultural dimensions and has been widely tested and replicated (Kirkman et al. 2006; Sondergaard 1994). In an attempt to address this gap in the accounting values literature, I develop scores for each of the four accounting values described by Gray for approximately five dozen countries. The process of quantifying Grays (1988) accounting values will assist future researchers in examining the effect of accounting values on various accounting phenomena. For example, many researchers have used Grays 4 (1988; Radebaugh et al. 2006) accounting values framework as theoretical support for their studies in various areas such as decision-making, firm valuation, materiality judgments, voluntary disclosure, and foreign listing requirements (Adams et al. 1993; Archambault and Archambault 2003; Arnold et al. 2001; Bao and Bao 2004; Buck and Shahrim 2005; Chow et al. 1995; Debreceny et al. 2002; Doupnik and Richter 2003; Doupnik and Salter 1995; Hussein 1996; King and Langli 1998; Lynn 1999; Meek et al. 1995; Radebaugh et al. 1995; Roberts and Salter 1999; Salter 1998; Salter et al. 2001; Schultz and Lopez 2001). A quantified set of accounting values scores would greatly assist in the hypothesis testing of studies such as these.

In their textbook, Radebaugh, Gray, and Black (2006) extended the description of the framework presented by Gray (1988).

1.2

Theoretical Framework of Accounting and Culture Gray (1988) synthesized the work of others (Frank 1979; Mueller 1967, 1968; Nair and

Frank 1980; Nobes 1983, 1984) and presented a framework depicting the relationship between accounting and culture (Figure 1.1). Perera (1989) proposes a modification of the Gray (1988) framework (Figure 1.2). Additionally, in their review article, Doupnik and Tsakumis (2004) presented a partial refinement of Grays (1988) framework (Figure 1.3). Doupnik and Tsakumis (2004) specify Grays (1988) accounting systems construct as more than just country-level financial reporting rules, but also the individual- and firm-level application of these rules by accountants.

Figure 1.1: Gray's (1988) Framework of Accounting and Culture

Figure 1.2: Perera's (1989) Framework of Accounting and Culture

Figure 1.3: Doupnik and Tsakumis' (2004) Partial Refinement of Gray's (1988) Framework of Accounting and Culture The three frameworks (Doupnik and Tsakumis 2004; Perera 1989; Gray 1988) are similar to each other in most respects. Each framework incorporates the influence of external forces on the physical environment. Globalization and war, for example, affect societal-level physical or ecological environments. These external forces cannot be controlled by the typical individual, group, or society and can greatly influence the development of a society. Each framework also describes the effect of societal institutions on accounting, but have certain differences. Gray (1988) and Doupnik and Tsakumis (2004) show a complex relationship of (1) an indirect relationship as societal culture affects institutions and institutions influence accounting values and practices directly and (2) an additional direct relationship between accounting values and practices. Perera (1989) also acknowledges societal culture but posits that the effect of institutions on accounting values and practices will be simultaneous and separate.

The major difference between the three frameworks (Doupnik and Tsakumis 2004; Perera 1989; Gray 1988) is the relationship between societal values and institutional consequences. The Gray (1988) and Doupnik and Tsakumis (2004) frameworks illustrate a hierarchical relationship where (1) societal values influence accounting values and institutional consequences simultaneously and individually, and then (2) accounting values and institutional consequences influence accounting practices simultaneously and individually. Alternatively, Perera (1989) shows a direct effect of the physical environment on societal values and institutional consequences as well as an interrelatedness between societal values and institutional consequences. One possible solution to this difference is the application of institutional theory (DiMaggio and Powell 1991, 1983; Oliver 1991; Selznick 1948; Selznick et al. 1969; Scott 1987, 2001). Institutional theory describes the way in which organizations of all sizes and types influence the behaviors of members of a society. Institutional attributes such as legal system, political system, and religion affect not only the laws and norms of a society, but also the way its members think. These three pillars (Scott 2001) of institutional theory (regulatory, normative, and cognitive) encompass the institutional consequences construct of the Perera (1989) and Gray (1988) frameworks as well as their societal values construct. The regulatory pillar includes institutional consequences such as legal system, political system, capital markets, and corporate ownership. The normative pillar includes institutional consequences such as professional

associations, education, and religion. The cognitive pillar includes societal values. The different depictions of the antecedents and consequences of the institutional consequences and societal values constructs in the Perera (1989), Gray (1988), and Doupnik and Tsakumis (2004) frameworks are likely due to the complexities of the relationships and

processes described by institutional theory. Various institutional elements within a society can jointly or individually affect one or more of the three pillars of institutional theory. An example of the influence of societal values within the cognitive pillar of institutional theory is evident when Sharp and Salter (1997) note that it is interesting to speculate whether agency theory could have been developed to the extent it has in any culture other than an individualist one; the very idea of rigorously modeling self-interest may not have occurred to researchers in a collectivist culture. The complex interactions of the two constructs (institutional consequences and societal values) presented separately in the Perera (1989), Gray (1988), and Doupnik and Tsakumis (2004) frameworks is better depicted as a single construct under the umbrella of institutional theory. In other words, institutional theory reconciles the major difference in the frameworks. Doupnik and Salter (1995) performed a similar combination in their general model of accounting development. A combination of the Perera (1989), Gray (1988), and Doupnik and Tsakumis (2004) frameworks is presented as an institutional framework of accounting and culture in Figure 1.4. The institutional framework presented in Figure 1.4 depicts the influence of institutional elements on accounting values and accounting practices, with an additional influence of accounting values on accounting practices. These elements are diverse and broad. The list of institutional elements in Figure 1.4 is not exhaustive; however, culture has been shown to proxy for many of these. For example, Chui, Lloyd, and Kwok (2002) found a relationship between culture and corporate capital structure. Additionally, Stulz and Williamson (2003) linked culture to a countrys level of investor protection. I use culture as a proxy for the various institutional influences on accounting values.

External Influences Forces of nature Trade Investment Conquest

Ecological Influences Geographic Economic Demographic Genetic/hygienic Historical Technological Urbanization


Reinforcement

Institutionalism Regulatory, Normative, & Cognitive Legal system Political system Capital markets Corporate ownership Professional associations Education Religion Societal Values

Accounting Values Professionalism Uniformity Conservatism Secrecy

Accounting Practices Authority Application/Enforcement Measurement Disclosure

Figure 1.4: Synthesis of Previous Frameworks into an institutional Framework of Accounting and Culture 1.3 Hofstedes Cultural Dimensions Gray (1988) based his accounting values framework largely on the cultural work of Hofstede (1980). Hofstede (1980) described and quantified the structural elements of culture that affect business. He presented these structural elements as four underlying dimensions of societal culture. These four cultural dimensions are Large versus Small Power Distance, Strong

10

versus Weak Uncertainty Avoidance, Individualism versus Collectivism, and Masculinity versus Femininity. Hofstede (1984: 83-84) described these four cultural dimensions as: Large versus Small Power Distance Power Distance is the extent to which the members of a society accept that power in institutions and organizations is distributed unequally. This affects the behaviour of the less powerful as well as of the more powerful members of society. People in Large Power Distance societies accept a hierarchical order in which everybody has a place which needs no further justification. People in Small Power Distance societies strive for power equalization and demand justification for power inequalities. The fundamental issue addressed by this dimension is how a society handles inequalities among people when they occur. This has obvious consequence for the way people build their institutions and organizations. Strong versus Weak Uncertainty Avoidance Uncertainty Avoidance is the degree to which the members of a society feel uncomfortable with uncertainty and ambiguity. This feeling leads them to beliefs promising certainty and to maintaining institutions protecting conformity. Strong Uncertainty Avoidance societies maintain rigid codes of belief and behaviour and are intolerant towards deviant persons and ideas. Weak Uncertainty Avoidance societies maintain a more relaxed atmosphere in which practice counts more than principles and deviance is more easily tolerated. The fundamental issue addressed by this dimension is how a society reacts on the fact that time only runs one way and that the future is unknown: whether it tries to control the future or to let it happen. Like Power Distance, Uncertainty Avoidance has consequences for the way people build their institutions and organizations. Individualism versus Collectivism Individualism stands for a preference for a loosely knit social framework in society wherein individuals are supposed to take care of themselves and their immediate families only. Its opposite, Collectivism, stands for a preference for a tightly knit social framework in which individuals can expect their relatives, clan, or other in-group to look after them in exchange for unquestioning loyalty (the word "collectivism" is not used here to describe any particular political system). The fundamental issue addressed by this dimension is the degree of interdependence a society maintains among individuals. It relates to people's self-concept: I or we. Masculinity versus Femininity Masculinity stands for a preference in society for achievement, heroism, assertiveness, and material success. Its opposite, Femininity, stands for a preference for relationships, modesty, caring for the weak, and the quality of life. This fundamental issue addressed by this dimension is the way in which a society allocates social (as opposed to biological) roles to the sexes.

11

Some societies strive for maximum social differentiation between the sexes. The norm is then that men are given the more outgoing, assertive roles and women the caring, nurturing roles. As in all societies most institutions are populated by men. Such maximum-social-differentiation societies will permeate their institutions with an assertive mentality. Such societies become "performance societies" evident even from the values of their women. I have called these societies "masculine". (In the English language, "male" and "female" are used for the biological distinctions between the sexes; "masculine" and "feminine" for the social distinction. A man can be feminine, but he cannot be female.) Other societies strive for minimal social differentiation between the sexes. This means that some women can take assertive roles if they want to but especially that some men can take relationship-oriented, modest, caring roles if they want to. Even in these societies, most institutions are populated by men (maybe slightly less than in masculine societies). The minimum-social-differentiation societies in comparison with their opposite, the maximum-social-differentiation societies, will permeate their institutions with a caring, quality-of-life orientated mentality. Such societies become "welfare societies" in which caring for all members, even the weakest, is an important goal for men as well as women. I have called such societies "feminine". "Masculine" and "feminine" are relative qualifications: they express the relative frequency of values which in principle are present in both types of societies. The fact that even modern societies can be differentiated on the basis of the way they allocate their social sex role is not surprising in the light of anthropological research on non-literate, traditional societies in which the social sex role allocation is always one of the essential variables. Like the Individualism-Collectivism dimension, the MasculinityFemininity dimension relates to people's self-concept: who am l and what is my task in life? 1.4 Grays Accounting Values Gray (1988) described the four accounting values presented in Figure 1.4 in terms of their relationship to Hofstedes (1980) cultural dimensions. He summarized these relationships

through the presentation of four untested hypotheses. Additionally, he described the strength of the various relationships with the terms strong, less strong, and weak. I will use these

hypotheses and the three levels of relationship strengths to help create the accounting values scores. Each accounting value and related hypothesis is discussed below.

12

1.4.1

Professionalism versus Statutory Control Professionalism versus statutory control (referred to as professionalism) is defined as a

societys preference for the exercise of individual professional judgment and the maintenance of professional self-regulation as opposed to compliance with prescriptive legal requirements and statutory control (Gray 1988). This construct captures the importance of authority in a societys accounting rules and regulations (Perera 1989). Accountants must exercise professional

judgment in the course of their duties and this accounting value captures the variation in the extent of the use of professional judgment at the country level. Gray (1988) posited that professionalism has the strongest link with individualism and uncertainty avoidance. He also proposed a less strong link with power distance. Grays (1988) first hypothesis follows: H1: The higher a country ranks in terms of individualism and the lower it ranks in terms of uncertainty avoidance and power distance, the more likely it is to rank highly in terms of professionalism. Subsequently, Radebaugh, Gray, and Black (2006) expanded on professionalism by noting a weak relationship with masculinity and long term orientation. They state that

professionalism is negatively related with long term orientation and positively related to the assertiveness aspect of masculinity. The relationships between professionalism and all of the relevant cultural dimensions are presented in Table 1.1.

13

Table 1.1: Matrix of Relationships of Accounting Values with Hofstede's (1980) Culture Values
Relationship Uncertainty Strength Avoidance Strong Negative Professionalism vs. Less Strong Statutory Control Weak Strong Positive Uniformity vs. Less Strong Flexibility Weak Strong Positive Conservatism vs. Less Strong Optimism Weak Strong Positive Secrecy vs. Less Strong Transparency Weak Blank squares indicate no relationship. Accounting Value Long-Term Orientation Power Distance Negative Negative Negative Positive Positive Negative Positive Positive Negative Negative Negative Positive Individualism Masculinity Positive

1.4.2

Uniformity versus Flexibility Uniformity versus flexibility (referred to as uniformity) is defined as a societys

preference for the enforcement of uniform accounting practices between companies and for the consistent use of such practices over time as opposed to flexibility in accordance with the perceived circumstances of individual companies (Gray 1988). This construct captures the importance of enforcement in a societys accounting rules and regulations (Perera 1989). Accounting rules and regulations are, to some extent, applied differently between industries and over time. This accounting value captures these variations in the flexibility of accounting rules and regulations across reporting entities as well as across time at the country level. Gray (1988) posited that uniformity has the strongest link with uncertainty avoidance and individualism. He also proposed a less strong link with power distance. The relationships between uniformity and the cultural dimensions are presented in Table 1.1. Grays (1988) second hypothesis follows: H2: The higher a country ranks in terms of uncertainty avoidance and power distance and the lower it ranks in terms of individualism, the more likely it is to rank highly in terms of uniformity.

14

1.4.3

Conservatism versus Optimism Conservatism versus optimism (referred to as conservatism) is defined as a societys

preference for a cautious approach to measurement so as to cope with the uncertainty of future events as opposed to a more optimistic, laissez-faire, risk-taking approach (Gray 1988). This construct captures the importance of measurement in a societys accounting rules and regulations (Perera 1989). Accounting rules and regulations govern the acceptable methods for measuring and reporting the value of net assets and profits. The reported value of net assets and profits for the same underlying transactions can vary by country based on different levels of cautiousness. This accounting value captures variations in the level of cautiousness at the country level. Gray (1988) posited that conservatism has the strongest link with uncertainty avoidance. He also proposed a less strong link with individualism and masculinity. Grays (1988) third hypothesis follows: H3: The higher a country ranks in terms of uncertainty avoidance and the lower it ranks in terms of individualism and masculinity, the more likely it is to rank highly in terms of conservatism. Subsequently, Radebaugh, Gray, and Black (2006) expanded on conservatism by noting a strong relationship with long term orientation. They state that conservatism is positively related with long term orientation. The relationships between conservatism and all of the relevant cultural dimensions are presented in Table 1.1. 1.4.4 Secrecy versus Transparency Secrecy versus transparency (referred to as secrecy) is defined as a societys preference for confidentiality and the restriction of disclosure of information about the business only to those who are closely involved with its management and financing as opposed to a more transparent, open and publicly accountable approach (Gray 1988). This construct captures the

15

importance of disclosure in a societys accounting rules and regulations (Perera 1989). Financial statements are meant to depict the state of the firm in an accurate, yet concise, way. This balance between accuracy and conciseness results in variations in disclosure at the country level. Gray (1988) posited that secrecy has the strongest link with uncertainty avoidance, power distance, and individualism. He also proposed a less strong link with masculinity. Grays (1988) fourth hypothesis follows: H4: The higher a country ranks in terms of uncertainty avoidance and power distance and the lower it ranks in terms of individualism and masculinity, the more likely it is to rank highly in terms of secrecy. Subsequently, Radebaugh, Gray, and Black (2006) expanded on secrecy by noting a strong relationship with long term orientation. They state that secrecy is positively related with long term orientation. The relationships between secrecy and all of the relevant cultural

dimensions are presented in Table 1.1. 1.5 Quantification of Grays Accounting Values Although the quantification of accounting values on a scale as large (58 countries) as I am proposing has not been previously completed, previous researchers have attempted this on a smaller scale (one to 39 countries). Eddie (1990) and Salter and Niswander (1995) create proxies for the accounting values from measures such as the wording within audit opinions and the valuation of assets and liabilities. They then examine the association between these measures of accounting values and Hofstedes (1980) cultural dimensions as described in Grays (1988) four accounting value hypotheses. Eddie (1990) found statistically significant associations

between his measures of accounting values and Hofstedes (1980) cultural dimensions for all thirteen of Grays (1988) hypothesized relationships using data from thirteen countries 5 . Salter

Eddies (1990) results have been questioned (Chanchani and MacGregor 1999; Doupnik and Tsakumis 2004) due to the lack of independent validation of his accounting values measures.

16

and Niswander found statistically significant associations for six of Grays (1988) thirteen hypothesized relationships using data from twenty-seven countries. Others (Gray and Vint 1995; Hope 2003; Jaggi and Low 2000; Wingate 1997; Zarzeski 1996) have used a countrys disclosure index to focus specifically on the secrecy accounting value, finding full to partial support for Grays (1988) hypothesized relationships using six to thirty-nine different countries. These studies have quantified Grays (1988) accounting values in various ways, such as content analyses of audit opinion letters (Salter and Niswander 1995), the presence or absence of a professional exam (Salter and Niswander 1995), and content analyses of financial statements (Sudarwan and Fogarty 1996; Zarzeski 1996). A drawback to a majority of these studies is the subjectivity inherent in the method of quantification of the accounting values (Chanchani and MacGregor 1999; Doupnik and Tsakumis 2004). In addition to these smaller scale attempts at the quantification of accounting values, a large scale attempt has been initiated. Chanchani and Willett (2004) present an accounting values survey (AVS) meant to measure a countrys accounting values via financial statement preparers and users. While they state that the measurement of accounting values is in progress on a large scale, they have to date only presented results for India and New Zealand. One of their reasons for choosing these countries is their use of the English language. This avoids a very troublesome challenge of international business research survey translation. This problem must be addressed before this large scale quantification of accounting values can continue. I propose to create a large scale quantification of accounting values by treating each accounting value as a multi-item measure. The items comprising each accounting value measure are the cultural dimensions described by Grays (1988) four hypotheses. For example,

hypothesis 1 states that the higher a country ranks in terms of individualism and the lower it

17

ranks in terms of uncertainty avoidance and power distance, the more likely it is to rank highly in terms of professionalism. Therefore, to create the measure of a countrys professionalism, I propose to combine available measures of that countrys individualism, uncertainty avoidance, and power distance dimensions of culture in the directions and strengths posited by Gray (1988). This indirect method of quantification essentially creates a proxy for each accounting value using culture as a basis. Hofstedes (1980) cultural dimension scores would seem to be the likeliest candidate for the basis of the quantified accounting values because Gray (1988) used Hofstedes definitions of cultural dimensions to describe accounting values. However, a major limitation of Hofstedes scores is the age of the data. Hofstede collected his data in the 1960s and 1970s. Although Hofstede (2001) argues that his scores are valid because culture changes slowly over time, Shenkar (2001) posits that cultures can substantially change over time and cultural stability is an illusion, thereby eroding the utility of Hofstedes (1980) scores. This and other criticisms have led some researchers (Baskerville 2003; McSweeney 2002b, 2002a; Schwartz 1992, 1994) to argue that Hofstedes (1980) scores may no longer be completely valid. Schwartz (1994, 1999) presents cultural dimension scores based on data collected in the late 1980s and early 1990s; however, only 38 countries are represented. On the other hand, the GLOBE team present cultural dimension scores for 61 6 societies based on recently collected data (House et al. 2004). I use the GLOBE data as an alternative to Hofstedes (1980) cultural dimension scores for the basis of the accounting values scores due to the scope and age of the GLOBE data.

The GLOBE team administered their surveys in 62 societies, but scores for the Czech Republic are not presented due to response bias problems. Additionally, the GLOBE team captures different, yet similar, dimensions for two societies in each of three nations (Germany, South Africa, and Switzerland). Although culture can vary within the borders of a nation, accounting rules and regulations do not; therefore, I have averaged the GLOBE scores for the societies residing within one nation. This brings the total number of countries in my study to 58.

18

Initially, Hofstede (1980) described culture through the use of four dimensions uncertainty avoidance, power distance, individualism, and masculinity. Subsequently, he added a fifth dimension long term orientation (Hofstede 2001). As the state of the literature has progressed, the cultural dimensions have become more complex; for example, Schwartz (1994) described seven culture level value types. The GLOBE project (2004) followed Schwartz (1994) theory-driven approach to scale development (Hanges and Dickson 2004) and identified nine cultural dimensions. The GLOBE project is primarily based on the results of the survey of over 17,000 middle managers in three industries: banking, food processing, and telecommunications, as well as archival measures of country economic prosperity and the physical and psychological well-being of the cultures studied (House et al. 2004). The GLOBE cultural dimensions are uncertainty avoidance, power distance, future orientation, institutional collectivism, in-group collectivism, gender egalitarianism, performance orientation, assertiveness, and humane orientation. cultural dimensions are described as (House et al. 2004: 30, 239, 282, 513, 569): Uncertainty Avoidance is the extent to which a society, organization, or group relies on social norms, rules, and procedures to alleviate the unpredictability of future events. Power Distance is the extent to which a community accepts and endorses authority, power differences, and status privileges. Future Orientation is the degree to which a collectivity encourages and rewards future-oriented behaviors such as planning and delaying gratification. Institutional Collectivism is the degree to which organizational and societal institutional practices encourage and reward collective distribution of resources and collective action. In-Group Collectivism is the degree to which individuals express pride, loyalty, and cohesiveness in their organizations or families. The

19

Gender Egalitarianism is the degree to which a collective minimizes gender inequality. Performance Orientation reflects the extent to which a community encourages and rewards innovation, high standards, excellence, and performance improvement. Assertiveness is the degree to which individuals are assertive, confrontational, and aggressive in their relationships with others. Humane Orientation is the degree to which an organization or society encourages and rewards individuals for being fair, altruistic, friendly, generous, caring, and kind to others. The nine GLOBE cultural dimensions can be related back to Hofstedes (1980, 2001) five dimensions. Three of them can be equated on a one to one basis. Hofstede s (1980, 2001) uncertainty avoidance, power distance, and long term orientation are similar to uncertainty avoidance, power distance, and future orientation; however, as shown in Appendix 2, there are no statistically significant correlations between Hofstedes uncertainty avoidance, power distance, and long term orientation and GLOBEs uncertainty avoidance, power distance, and future orientation. Hofstedes (1980) individualism dimension has been split into institutional collectivism and in-group collectivism. As presented in Appendix 2, individualism is not

statistically associated with in-group collectivism and negatively correlated with institutional collectivism (R=-0.513; p=0.00). Hofstedes (1980) masculinity dimension has been

deconstructed into gender egalitarianism, performance orientation, assertiveness, and humane orientation. As shown in Appendix 2, masculinity is not statistically associated with any of these GLOBE variables. The GLOBE study decomposes the collectivism dimension of culture into institutional and in-group collectivism. The primary difference between the two is the level of analysis. The level of analysis concept portrays differences between the perspectives of individuals, groups,

20

organizations, societies, countries, and global regions. For example, an examination of the economy at a high level of analysis would yield data on trade imbalances and tariff impacts while an examination at a low level of analysis would yield data on consumer behavior and local demand. Institutional collectivism concerns a higher level of analysis, such as at the societal level. On the other hand, in-group collectivism focuses on a lower level of analysis, such as at the work group level of analysis. Accounting rules and regulations and, hence, accounting values, are applied at the country level. Therefore, institutional collectivism is the more

appropriate dimension for the quantification of accounting values. However, an exception to this is the professionalism accounting value. This accounting value is concerned with the level of decision-making authority granted at the individual accountant level. Consequently, the in-group collectivism dimension is more appropriate for the quantification of the professionalism accounting value. 1.5.1 Gray (1988) and GLOBE (House et al. 2004) Gray (1988) explained the influence of culture on accounting with an objective and comprehensive framework. Additionally, he equated the accounting values framework to

Hofstedes (1980) cultural dimensions in four testable hypotheses. However, in describing the relationships between the accounting values framework and Hofstedes cultural dimensions Gray (1988) left the door open for the use of a different measure of culture because he described the particular aspect of each dimension that related to the accounting values framework. I use Grays (1988) additional information to equate accounting values to the GLOBE cultural scores, which makes the following quantification of the accounting values possible. A matrix of these relationships is presented in Table 1.2.

21

Table 1.2: Matrix of Relationships of Accounting Values with GLOBE Culture Values
Relationship Strength Accounting Value Strong Professionalism vs. Less Strong Statutory Control Weak Strong Uniformity vs. Less Strong Flexibility Weak Strong Conservatism vs. Less Strong Optimism Weak Strong Secrecy vs. Less Strong Transparency Weak Accounting Value Uncertainty Future Avoidance Orientation Negative Negative Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Power Distance Negative Institutional Collectivism

Relationship Performance In-Group Humane Strength Orientation Collectivism Orientation Assertiveness Strong Negative Professionalism vs. Less Strong Statutory Control Weak Positive Strong Uniformity vs. Less Strong Flexibility Weak Strong Conservatism vs. Less Strong Negative Optimism Weak Strong Secrecy vs. Less Strong Negative Negative Transparency Weak Gender Egalitarianism is not associated with any Accounting Value. Blank squares indicate no relationship.

Additionally, the institutional framework presented in Figure 1.4 depicts accounting practices as well as accounting values. These practices can be equated to the accounting values on a one to one basis. Authority, enforcement, measurement, and disclosure are associated with professionalism, uniformity, conservatism, and secrecy, respectively (Gray 1988; Perera 1989; Radebaugh et al. 2006). In addition to the societal values scores in the GLOBE study (House et al. 2004), the authors present societal practices scores for all 61 societies. I have quantified the accounting practices scores using the same method as the accounting values quantification method described below. These accounting practices scores are presented in Table 1.3.

22

Descriptive and other statistics for these scores are presented later in Panel A of Table 1.5, and histograms of the distributions for these scores are presented later in Figure 1.5. Table 1.3: Accounting Practices Scores
Country Albania Argentina Australia Austria Bolivia Brazil Canada China Colombia Costa Rica Denmark Ecuador Egypt El Salvador England Finland France Georgia Germany Greece Guatemala Hong Kong Hungary India Indonesia Iran Ireland Israel Italy Japan Kazakhstan Kuwait Malaysia Mexico Morocco Namibia Netherlands New Zealand Nigeria Philippines Poland Portugal Qatar Russia Singapore Slovenia South Africa South Korea Spain Sweden Switzerland Taiwan Thailand Turkey USA Venezuela Zambia Zimbabwe Authority (Professionalism vs. Statutory Control) 46 61 87 40 84 71 72 14 56 70 90 44 48 63 71 62 74 50 51 81 64 55 91 39 37 24 50 85 75 67 72 38 18 42 48 74 100 86 33 20 70 45 91 79 0 58 71 57 53 61 62 35 38 48 92 70 37 35 Enforcement (Uniformity vs. Flexibility) 45 77 55 29 96 79 45 20 76 81 33 70 55 75 40 24 54 78 37 100 86 57 97 59 46 60 40 61 76 31 65 47 27 58 68 54 53 27 42 45 63 66 60 77 7 65 52 36 64 0 34 39 58 68 61 78 43 49 Measurement (Conservatism vs. Optimism) 57 92 53 27 80 80 41 42 92 79 15 81 61 76 38 21 68 86 40 90 98 65 100 64 61 77 52 60 84 39 70 67 27 62 90 63 30 56 51 59 82 67 52 98 8 70 44 64 77 0 31 55 76 74 60 81 62 67 Disclosure (Secrecy vs. Transparency) 79 88 68 29 100 76 57 48 84 100 49 87 62 66 50 25 60 91 30 84 93 71 88 67 71 79 66 72 78 40 68 76 39 64 83 62 53 65 40 69 78 63 68 89 0 66 55 45 63 3 36 56 82 68 72 84 77 65

23

1.6

Mechanics of the Quantification Each accounting value is a summative combination of the GLOBE cultural dimensions

listed in Table 1.2. More specifically, the scores for the cultural dimensions in Table 1.2 that comprise each of the accounting values are added together to create a single score for each of the four accounting values. In addition, I have applied a weighting system to the quantification process. This weighting system is based on the accounting values literature. Gray (1988; Radebaugh et al. 2006) used the terms strong, less strong, and weak to describe the relationships between the cultural dimensions and the accounting values. In order to facilitate a weighted combination of the multiple items comprising an accounting value, I have translated this into a weight of three, two, and one for strong, less strong, and weak, respectively. In other words, a relationship described as less strong carries twice the effect in my weighting method as a relationship described as weak. Similarly, a relationship described as strong carries three times the effect of a relationship described as weak. As an alternative, I used weights of four, two, and one for strong, less strong, and weak (i.e., strong carried twice the weight of less strong and less strong carried twice the weight of weak); however, the difference between the two weighting systems was small or nonexistent 7 and I employ the three, two, one weighting method here. This weighting method is straightforward and considers the varying strengths of the relationships described by Gray (1988; Radebaugh et al. 2006). The weighting is explained in greater detail for each accounting value below.

I tested the two weighting methods using paired t tests of means analyses. The 4-2-1 and 3-2-1 weighting methods produced no statistically significant difference in means for the Uniformity versus Flexibility (t= 0.490, p=0.626) and Secrecy versus Transparency (t=-0.991, p=0.326) accounting values. The 4-2-1 method produced a Professionalism versus Statutory Control accounting value with a mean of 45.32 and the 3-2-1 method produced a mean of 45.81. This small difference (0.49) is statistically significant (t=2.137, p=0.037). The 4-2-1 method produced a Conservatism versus Optimism accounting value with a mean of 52.39 and the 3-2-1 method produced a mean of 51.61. This small difference (0.78) is statistically significant (t=-3.902, p=0.000).The four accounting values produced with the 4-2-1 method were virtually perfectly correlated with the four accounting values produced with the 3-2-1 method (Pearson R0.998).

24

Table 1.4: Accounting Values Scores


Country Albania Argentina Australia Austria Bolivia Brazil Canada China Colombia Costa Rica Denmark Ecuador Egypt El Salvador England Finland France Georgia Germany Greece Guatemala Hong Kong Hungary India Indonesia Iran Ireland Israel Italy Japan Kazakhstan Kuwait Malaysia Mexico Morocco Namibia Netherlands New Zealand Nigeria Philippines Poland Portugal Qatar Russia Singapore Slovenia South Africa South Korea Spain Sweden Switzerland Taiwan Thailand Turkey USA Venezuela Zambia Zimbabwe Professionalism vs. Statutory Control 31 33 65 84 22 48 68 60 27 45 80 21 13 0 66 87 64 33 87 41 29 62 48 63 38 28 66 54 51 82 56 39 36 17 15 15 100 29 16 28 39 50 31 25 62 47 43 54 49 69 97 26 6 36 65 17 50 42 Uniformity vs. Flexibility 82 64 26 20 80 83 11 75 69 62 15 81 88 100 28 2 47 47 23 80 68 55 43 54 84 100 31 33 56 26 36 79 70 80 89 61 2 38 90 70 49 58 81 40 43 51 53 27 62 0 12 93 96 64 21 83 49 60 Conservatism vs. Optimism 67 61 22 16 60 75 14 49 70 50 0 79 78 83 22 9 36 56 14 67 71 51 52 53 80 87 27 34 61 38 32 65 67 74 83 65 11 31 86 68 34 46 71 54 42 44 51 54 67 1 4 70 100 77 19 77 57 61 Secrecy vs. Transparency 78 54 26 14 78 66 16 58 56 56 9 72 86 81 27 0 38 50 21 64 66 60 46 54 80 84 31 33 55 39 38 78 68 77 84 63 17 59 72 66 44 46 81 44 45 43 58 43 55 4 9 78 100 68 24 69 49 62

25

1.6.1

Professionalism versus Statutory Control As shown in Table 1.2, the professionalism accounting value is strongly and negatively

associated with uncertainty avoidance, weakly and negatively associated with future orientation, less strongly and negatively associated with power distance, strongly and negatively associated with in-group collectivism, and weakly and positively associated with assertiveness. To

illustrate the quantification, the accounting values scores for Australia will be calculated. Using the weights described earlier and the GLOBE study (House et al. 2004) cultural dimension scores, the professionalism score for Australia is the weighted sum of strong and negative uncertainty avoidance (3 X -3.98), weak and negative future orientation (1 X -5.15), less strong and negative power distance (2 X -2.78), strong and negative in-group collectivism (3 X -5.75), and weak assertiveness (1 X 3.81). This score is standardized and reported in Table 1.4 with the professionalism scores of the other 57 countries. Descriptive and other statistics for this score are presented later in Panel B of Table 1.5. 1.6.2 Uniformity versus Flexibility As shown in Table 1.2, the uniformity accounting value is strongly and positively associated with uncertainty avoidance, less strongly and positively associated with power distance, and strongly and positively associated with institutional collectivism. Using the

weights described earlier and the GLOBE study (House et al. 2004) cultural dimension scores, the uniformity score for Australia is the weighted sum of strong uncertainty avoidance (3 X 3.98), less strong power distance (2 X 2.78), and strong institutional collectivism (3 X 4.40). This score is standardized and reported in Table 1.4 with the uniformity scores of the other 57 countries. Descriptive and other statistics for this score are presented later in Panel B of Table 1.5.

26

1.6.3

Conservatism versus Optimism As shown in Table 1.2, the conservatism accounting value is strongly and positively

associated with uncertainty avoidance, strongly and positively associated with future orientation, less strongly and positively associated with institutional collectivism, and less strongly and negatively associated with performance orientation. Using the weights described earlier and the GLOBE study (House et al. 2004) cultural dimension scores, the conservatism score for Australia is the weighted average of strong uncertainty avoidance (3 X 3.98), strong future orientation (3 X 5.15), less strong institutional collectivism (2 X 4.40), less strong and negative performance orientation (2 X -5.89). This score is standardized and reported in Table 1.4 with the conservatism scores of the other 57 countries. Descriptive and other statistics for this score are presented later in Panel B of Table 1.5. 1.6.4 Secrecy versus Transparency As shown in Table 1.2, the secrecy accounting value is strongly and positively associated with uncertainty avoidance, strongly and positively associated with future orientation, strongly and positively associated with power distance, strongly and positively associated with institutional collectivism, less strongly and negatively associated with performance orientation, and less strongly and negatively associated with humane orientation. Using the weights

described earlier and the GLOBE study (House et al. 2004) cultural dimension scores, the secrecy score for Australia is the weighted sum of strong uncertainty avoidance (3 X 3.98), strong future orientation (3 X 5.15), strong power distance (3 X 2.78), strong institutional collectivism (3 X 4.40), less strong and negative performance orientation (2 X -5.89), and less strong and negative humane orientation (2 X 5.58). This score is standardized and reported in

27

Table 1.4 with the secrecy scores of the other 57 countries. Descriptive and other statistics for this score are presented later in Panel B of Table 1.5. 1.6.5 Standardization of Quantified Scores The GLOBE cultural dimension scores were gathered with a survey using a seven-point scale (House et al. 2004). This characteristic was transferred to the unstandardized accounting values scores 8 . For increased clarity of understanding and the intuitiveness of the accounting values scores, the scores presented are standardized to create a range between zero and 100. Standardization converts variables to a common scale and is generally performed to create data with a mean of zero and standard deviation of one (Hair et al. 2006). For the accounting values scores, I standardized to a given range rather than a given mean and standard deviation. This creates scores that look similar to one another, but allows for different, albeit similar, means and standard deviations. This is important because readers will likely be more comfortable reading a secrecy score for Australia of 26 rather than 4.465. Furthermore, comparing the secrecy scores of Australia and China may be easier for some if the scores are 26 and 58 rather than 4.465 and 4.780. The unstandardized and standardized scores are perfectly correlated, have the same distribution, and are statistically identical in all other respects.

The unstandardized accounting values scores are not presented.

28

Table 1.5: Descriptive Statistics, Test of Normality, and Pearson Correlations


Panel A: Accounting Practices Scores Std. Dev. 21.86 KolmogorovSmirnov Z (Asymp. Sig.) 0.641 (0.805) 0.416 (0.995) 0.753 (0.621) 1.091 (0.184) Correl. Correl. Correl. 1. 2. 3. (Sig.) (Sig.) (Sig.)

Accounting Value 1. Professionalism vs. Statutory Control (Authority) 2. Uniformity vs. Flexibility (Enforcement) 3. Conservatism vs. Optimism (Measurement) 4. Secrecy vs. Transparency (Disclosure) n = 58

Mean 57.89

Min. 0.00

Max. 100.00

54.94

21.18

0.00

100.00

0.404 (0.002) 0.162 (0.224) 0.277 (0.036) 0.883 (0.000) 0.857 (0.000) 0.887 (0.000)

61.45

22.94

0.00

100.00

64.55

21.16

0.00

100.00

Panel B: Accounting Values Scores Std. Dev. 23.33 27.15 24.85 23.56 KolmogorovSmirnov Z (Asymp. Sig.) 0.527 (0.944) 0.770 (0.593) 0.785 (0.568) 0.663 (0.772) Correl. Correl. Correl. 1. 2. 3. (Sig.) (Sig.) (Sig.)

Accounting Value 1. Professionalism vs. Statutory Control 2. Uniformity vs. Flexibility 3. Conservatism vs. Optimism 4. Secrecy vs. Transparency n = 58

Mean 45.81 54.96 51.61 52.47

Min. 0.00 0.00 0.00 0.00

Max. 100.00 100.00 100.00 100.00

-0.842 (0.000) -0.859 (0.000) -0.865 (0.000) 0.935 (0.000) 0.955 (0.000) 0.946 (0.000)

Panel C: Correlations between Accounting Practices Scores and Accounting Values Scores Professionalism vs. Statutory Control (Sig.) 0.203 (0.126) n = 58 Uniformity vs. Flexibility (Sig.) 0.425 (0.001) Conservatism vs. Optimism (Sig.) 0.623 (0.000) Secrecy vs. Transparency (Sig.) 0.519 (0.000)

1.6.6

Characteristics of Quantified Scores Table 1.5 presents some descriptive information about the quantified scores. The

accounting practices scores in Table 1.3 do not have a theoretical foundation; instead they are the

29

result of accounting values theory applied to cultural practices data. They are presented for informational purposes and the associated statistics presented in Panel A of Table 1.5 and later in Figure 1.5 are not discussed. The accounting values scores have means of 46 to 55 and standard deviations of 23 to 27, as presented in Panel B of Table 1.5. Due to the standardization step of the quantification process, all of the scores range from zero to 100. A normal distribution is an important characteristic in data such as these because of the assumptions of ordinary least squares regression. Table 1.5 Panel B shows the Kolmogorov-Smirnov Z scores for all four accounting values scores. All four of the accounting values scores have Kolmogorov-Smirnov Z scores that are not statistically significant. This indicates normal distributions for all four scores (Massey 1951). Hair et al. (2006) suggest the use of graphical plots in conjunction with

statistical tests. Accordingly, Figure 1.5 presents histograms of all four accounting values scores with normal distribution curves. normally distributed. These also appear to be acceptably, albeit not perfectly,

30

Figure 1.5: Histograms of Accounting Values Scores 31

Table 1.5 Panel B also presents the correlations between the four accounting values scores. The absolute values of the correlations are in the range of 0.842 to 0.955. These high correlations are caused by using many of the same eight cultural dimensions to create each of the accounting values scores in a summative manner. For example, uncertainty avoidance is

strongly associated with all four accounting values. This one cultural dimension makes up at least 19% (3/16 of secrecy) and as much as 38% (3/8 of uniformity) of each of the accounting values scores. These high correlations could cause multicollinearity in regression models where two or more of the quantified accounting values are in the same model. However, the scores still have utility when used separately. Many studies focus on a single accounting value at a time (e.g., Debreceny et al. 2002; Gray and Vint 1995; Hope 2003; Jaggi and Low 2000; King and Langli 1998; Meek et al. 1995) and the accounting values scores would work well in these situations. 1.6.7 Empirical Test of Face Validity Grays (1988) accounting value hypothesis were stated in terms of Hofstedes (1980) cultural measures and the GLOBE (House et al. 2004) study provides alternate measures of culture; therefore, an empirical test of the relationship between these accounting values scores and Hofstedes cultural dimension scores using Grays (1988) four hypotheses may appear to simply be a test of the similarity of the GLOBE and Hofstede measures of cultural dimensions. However, I believe it is a necessary initial test of the accounting values scores in order to provide face validity. I test these hypotheses using regression analysis on four models where each of the four accounting value scores is explained by the cultural dimensions in each of Grays (1988) four hypotheses. For example, hypothesis 1 is tested by examining the relationship between the professionalism score and Hofstede s (1980) individualism, uncertainty avoidance, and power

32

distance scores. There are 49 countries 9 with both GLOBE and Hofstede values reported. These 49 countries comprise the sample for the regression analysis. Table 1.6: Regression Results of Accounting Values Hypothesis Testing
Hypothesis 1 Professionalism vs. Statutory Control + 0.317 (0.02) Hypothesis 2 Uniformity vs. Flexibility -0.449 (0.00) 0.124 (0.11) 0.345 (0.01) Hypothesis 3 Conservatism vs. Optimism -0.717 (0.00) 0.166 (0.05) Hypothesis 4 Secrecy vs. Transparency -0.424 (0.00) 0.086 (0.20) 0.377 (0.00) 0.117 (0.12) 0.581 0.543 15.245 0.000

Pred. Individualism Uncertainty Avoidance Power Distance Masculinity

Pred.

Pred.

Pred.

-0.208 (0.03) -0.351 (0.01)

0.143 (0.07) 0.600 0.573 22.470 0.000

R-Square 0.472 0.586 Adj. R-Sqr. 0.437 0.558 F-Statistic 13.400 21.240 F-Stat. Sig. 0.000 0.000 n = 49 Standardized coefficients are presented. Significance levels are presented in parentheses below coefficients. All hypothesis tests are one-tailed. F tests of the models are two-tailed.

Table 1.6 presents the tests of the thirteen relationships posited in Grays (1988) four hypotheses. These tests statistically support nine of the thirteen relationships. The tests of the relationships comprising the professionalism and uniformity accounting values are all statistically significant at the p=0.05 level or better, with the exception of the relationship between uniformity and uncertainty avoidance, which was in the predicted direction but has a pvalue of 0.11. This provides an initial level of validity to these two accounting values scores.
The 49 common countries are Argentina, Australia, Austria, Brazil, Canada, China, Colombia, Costa Rica, Denmark, Ecuador, Egypt, El Salvador, England, Finland, France, Germany, Greece, Guatemala, Hong Kong, Hungary, India, Indonesia, Iran, Ireland, Israel, Italy, Japan, Kuwait, Malaysia, Mexico, Netherlands, New Zealand, Nigeria, Philippines, Poland, Portugal, Russia, Singapore, South Africa, South Korea, Spain, Sweden, Switzerland, Taiwan, Thailand, Turkey, USA, Venezuela, and Zambia.
9

33

The relationship between secrecy and uncertainty avoidance is in the predicted direction; however, it failed to receive statistical support (p=0.20). The masculinity dimension is not statistically related to either the conservatism or secrecy accounting values. This is likely due to the use of the decomposed aspects of the masculinity dimension, specifically assertiveness, performance orientation, and humane orientation, instead of Hofstedes (1980) masculinity score. However, the other relationships comprising the conservatism and secrecy accounting values are all in the predicted direction and, with one exception, statistically significant at the p=0.05 level or better. Overall, these tests provide, at the very least, face validity to the accounting values scores. 1.7 Discussion and Conclusion There are several limitations in my quantification of the accounting values. Among these is the assumption of the validity of Grays (1988) accounting values. Doupnik and Tsakumis (2004) review numerous empirical tests of Grays (1988) framework and conclude that the framework has not been subjected to adequate empirical inquiry to positively support it empirically. These empirical tests of Grays (1988) framework include Salter and Niswander (1995), who found empirical support for only six of the thirteen relationships posited in Grays (1988) four hypotheses. Sudarwan and Fogarty (1996) found empirical support for only four of the relationships. While there are other studies that support Grays (1988) framework, Eddie (1990) found empirical support for all thirteen hypothesized relationships, there is still enough ambiguity to withhold a judgment of final empirical support for the Gray (1988) framework. A second limitation is the appearance of circular logic in the quantification and validation of the accounting values. Grays (1988) hypotheses are used to quantify the accounting values and the same relationships are used to provide initial validation to the quantified accounting

34

values. A solution to this apparent circular logic would be quantifying the accounting values using primary data gathered from hundreds of subjects from each of dozens of countries. This solution would be an ambitious and worthwhile undertaking, but in the interim my quantification using secondary data provides a necessary second-best solution. Another limitation is the coverage of countries. An advantage to using secondary data to quantify the accounting values is the large number of countries included in the quantification. However, this number is limited by the secondary data used in the quantification. In this case, the GLOBE data covers only 58 of the worlds 194 countries. Furthermore, the initial validity testing of the quantified accounting values is done on only 49 countries due to the limited overlap of the GLOBE and Hofstede data. This leaves three-quarters of the worlds countries out of the process. The future availability of a more inclusive cultural dataset would be the best solution, but a valid combination the GLOBE (House et al. 2004), Hofstede (1980), and Schwartz (1994, 1999) datasets could substantially increase the coverage of countries. This is left for future endeavors. In this essay I have presented an institutional framework of the relationship between accounting and culture. This framework integrates institutional theory with prior accounting values frameworks (Doupnik and Tsakumis 2004; Gray 1988; Perera 1989). quantified Grays (1988) accounting values for 58 countries. I have also

While attempts have been

previously made to quantify Grays (1988) accounting values with survey instruments (e.g., Chanchani and Willett 2004), this work has not yet been completed. As an alternative to collecting primary data on accounting values, I have used theory formulated by Gray (1988) from the work of others (Frank 1979; Mueller 1967, 1968; Nair and Frank 1980; Nobes 1983, 1984) and cultural dimensions quantified by the GLOBE project (House et al. 2004) to enrich the

35

accounting values research stream by presenting quantified accounting values scores for the purpose of expanding and testing this theoretical realm by future researchers. The next step in this process is to validate these quantified accounting values scores. I do this in Essay 2 with standard validation techniques and through the use of multiple replications of previous studies. Essay 3 applies the quantified accounting values scores to an examination of deviations of a countrys domestic GAAP from International Financial Reporting Standards.

36

Essay 2 - The Empirical Validation of the Accounting Values Scores 2.1 Introduction With the creation of the accounting values scores in Essay 1 comes the need to test the validity of these scores in a variety of ways. The accounting values scores will provide only limited help to international accounting researchers unless they are shown to be a reliable proxy for the accounting values construct. A traditional validity test as well as a replication of a previous study using Grays (1988) framework will be undertaken in an attempt to validate the accounting values scores. The accounting values scores created in Essay 1 will be tested using two different methods. First, previous research that has attempted to test Grays (1988) framework (e.g., Eddie 1990; Gray and Vint 1995; Hope 2003; Salter and Niswander 1995) will be used as various methods of construct measurement in a Multi-Trait Multi-Method validation analysis. The accounting values scores created in Essay 1 are tested against various proxies for accounting values created in these previous studies. The expectation is that quantified accounting values created in Essay 1 will be properly related to the various other proxies for accounting values. For example, Salter and Niswander (1995) created multiple measures of each of the four accounting values and tested these measures against Hofstedes (1980) cultural dimension scores using Grays (1988) hypothesized relationships. The Multi-Trait Multi-Method analysis will use the Salter and Niswander (1995) measures, along with others, to provide validity to the accounting values scores created in Essay 1. Second, other researchers have used Grays (1988) framework as theoretical support for their studies (e.g., Arnold et al. 2001; Doupnik and Riccio 2006; Jaggi and Low 2000; Meek et al. 1995; Schultz and Lopez 2001). The second method for validating the accounting values

37

scores involves replicating one of these studied by replacing their proxy for culture with the appropriate quantified accounting values score. The expectation is that quantified accounting values created in Essay 1 will improve the explanatory value of each model and, therefore, improve the hypothesis testing of the studies. The results of these two validation methods are mixed. The Multi-Trait Multi-Method analysis shows limited evidence of convergent validity and no evidence of discriminant validity. Conversely, the replication method shows evidence of the usefulness of the quantified accounting values scores. Taken together, I reach the conclusion that, although the accounting values scores are not as statistically sound, they do provide help in addressing research questions dealing with culture and international accounting practices and systems. 2.2 Multi-Trait Multi-Method Analysis Multi-Trait Multi-Method analysis was introduced by Campbell and Fiske (1959) and tests construct validity using correlation analysis between measurements obtained when each of a number of traits is measured by each of a number of methods (Kotz et al. 1985). Cronbach and Meehl (1955) define construct validity as the totality of evidence about whether a particular operationalization of a construct adequately represents what is intended by theoretical account of the construct being measured. Construct validity is primarily comprised of convergent and discriminant validity. Convergent validity is based on the extent to which a measure correlates with different measures with which it is theoretically predicted to correlate and discriminant validity is based on the extent to which an operationalization does not correlate with operationalizations of different constructs (Kotz et al. 1985). As a test of the construct validity of the accounting values created in Essay 1, I use Multi-Trait Multi-Method analysis to compare similarities and differences with accounting value proxies created in the following six studies.

38

2.2.1

Eddie (1990) Eddie (1990) constructed indexes of accounting values and used them to test the link

between a countrys societal values and characteristics of its accounting system. He selected the factors comprising the indexes on the basis of theoretical reasoning (Eddie 1990). These 40 factors, ten factors per accounting value, are listed in the Appendix of this dissertation. Similar to Nobes (1984), information for scoring the factors for each country was discerned from works of reference, study of financial reports and general business literature (Eddie 1990). Each factor was then scored judgmentally on a six point scale (Eddie 1990). Eddie (1990) used correlation analysis to empirically support all of Grays (1988) hypothesized relationships. Eddies (1990) accounting values scores are presented in his paper. The data covers thirteen countries, twelve of which overlap with the accounting values scores created in Essay 1. This data is method 2 in the Multi-Trait Multi-Method analysis. 2.2.2 Salter and Niswander (1995) Salter and Niswander (1995) also created scores for Grays (1988) accounting values in order to test the link between accounting values and systems and Hofstedes (1980) cultural constructs. They also selected the items comprising the indexes on the basis of theoretical reasoning. These items and their sources are listed in Exhibit 2.1 of this Essay. The items comprising the Professionalism versus Statutory Control and Secrecy versus Transparency accounting values are from secondary sources (e.g., Bavishi 1991), while the items comprising the Uniformity versus Flexibility and Conservatism versus Optimism accounting values are from primary sources (i.e., Doupnik and Salter 1993). Salter and Niswander (1995) used multiple regression analysis to empirically support six of Grays (1988) thirteen hypothesized relationships.

39

Exhibit 2.1: Composition of Salter & Niswanders (1995) Accounting Values Indexes Salter and Niswanders (1995) accounting values scores are not presented in their paper. I have used a newer edition (Bavishi 1995) of their secondary sources to recreate their accounting values scores for Professionalism versus Statutory Control and Secrecy versus Transparency. This yielded Professionalism versus Statutory Control scores for 43 countries and Secrecy versus Transparency scores for 41 countries, of which 35 and 36, respectively, overlap with the accounting values scores created in Essay 1. The primary data collected in Doupnik and Salter (1993) was not available; therefore, I was not able to recreate their accounting values

40

scores for Uniformity versus Flexibility and Conservatism versus Optimism. The recreated data is method 3 in the Multi-Trait Multi-Method analysis. 2.2.3 Gray and Vint (1995) Gray and Vint (1995) tested the link between the Secrecy versus Transparency accounting value and system and Hofstedes (1980) cultural constructs using a comprehensive database of disclosure practices of large companies in 27 countries. The database used was the result of a joint project by the University of Glasgow and Deloitte Haskins and Sells International and published by Gray, Campbell, and Shaw (1984) (Gray and Vint 1995). The joint project gathered the data via a survey instrument containing 512 questions relating to disclosure requirements and practices (Gray et al. 1984). Gray and Vint (1995) used univariate regression analysis to empirically support all four of Grays (1988) hypothesized relationships concerning the Secrecy versus Transparency accounting value. They used multiple regression analysis to empirically support two of Grays (1988) four hypothesized relationships concerning the Secrecy versus Transparency accounting value. Gray, Campbell, and Shaws (1984) disclosure scores are presented in Gray and Vints (1995) paper. The data covers 27 countries, 23 of which overlap with the accounting values scores created in Essay 1. This data is method 4 in the Multi-Trait Multi-Method analysis. 2.2.4 Zarzeski (1996) Zarzeski (1996) calculated a disclosure index for firms in France, Germany, Hong Kong, Japan, Norway, the United Kingdom, and the United States in order to test the link between accounting practices and Hofstedes (1980) cultural constructs. She examined 256 corporate annual reports to create firm-level disclosure scores (Zarzeski 1996). She used prior accounting disclosure studies (Adhikari and Tondkar 1992; Barrett 1976; Choi 1973; Singhvi and Desai

41

1971) to choose the disclosure items (Zarzeski 1996). These disclosure items and associated weights are presented in Exhibit 2.2. Each firms disclosure score is the ratio of sum of the weights of the disclosure items appearing in the corporate annual report over the maximum score of 52 (Zarzeski 1996). She uses firm-level multiple regression analysis with repeated measures for the culture variables to provide empirical support for three of Grays (1988) four hypothesized relationships concerning the Secrecy versus Transparency accounting value.

Exhibit 2.2: Composition of Zarzeskis (1996) Disclosure Index

42

Zarzeski (1996) presents the firm-level scores averaged by country. This data covers seven countries, six of which overlap with the accounting values scores created in Essay 1. Zarzeskis (1996) scores are measures of transparency and should be negatively related to the Secrecy versus Transparency accounting value created in Essay 1. This data is method 5 in the Multi-Trait Multi-Method analysis. 2.2.5 La Porta et al. (1998) Although they do not discuss culture or accounting values, La Porta et al. (1998) investigate the legal rules protecting corporate shareholders and creditors and the quality of enforcement of these rules in 49 countries. As part of their study, they use an index created by examining and rating companies' 1990 annual reports on their inclusion or omission of 90 items falling in the categories of general information, income statements, balance sheets, funds flow statement, accounting standards, stock data, and special items (La Porta et al. 1998; La Porta et al. 2000). La Porta et al.s (1998) disclosure index is presented in their paper. The data covers 41 countries, 36 of which overlap with the accounting values scores created in Essay 1. La Porta et al.s (1998) scores are measures of transparency and should be negatively related to the Secrecy versus Transparency accounting value created in Essay 1. This data is method 6 in the MultiTrait Multi-Method analysis. 2.2.6 Hope (2003) Hope (2003) examines the relative roles of legal origin and national culture in explaining firm-level disclosure levels. He creates a dataset that includes 1851 disclosure scores from 1583 firms reported in two separate CIFAR editions (Bavishi 1991; 1995) covering 39 countries (Hope 2003). The data is a measure of the extent of information disclosure in annual financial

43

statements and reported by the Center for International Financial Analysis and Research (CIFAR) for leading non-financial companies in 42 countries in the first half of the 1990s (Bavishi 1991; Hope 2003; 1995). Hope (2003) uses univariate analysis to provide empirical support for three of Grays (1988) four hypothesized relationships concerning the Secrecy versus Transparency accounting value. When he separates the sample by legal origin, Hope (2003) provides univariate empirical support for all four of Grays (1988) hypothesized relationships in common law countries and univariate empirical support for three of Grays (1988) four hypothesized relationships in code law countries. Hope (2003) next uses multiple regression analysis to provide consistent empirical support for one of Grays (1988) four hypothesized relationships concerning the Secrecy versus Transparency accounting value. When he separates the sample by legal origin, Hope (2003) uses multiple regression to provide empirical support for three of Grays (1988) four hypothesized relationships in common law countries and empirical support for one of Grays (1988) four hypothesized relationships in code law countries. Interestingly, Hopes (2003) multivariate analysis shows relationships between disclosure practices and the Secrecy versus Transparency accounting value that is one way for common law countries and opposite for code law countries, with the common law countries relationship more closely matching Grays (1988) hypothesis. Hope (2003) presents the firm-level scores averaged by country. This data covers 42 countries, 36 of which overlap with the accounting values scores created in Essay 1. Hopes (2003) scores are measures of transparency and should be negatively related to the Secrecy versus Transparency accounting value created in Essay 1. This data is method 7 in the MultiTrait Multi-Method analysis.

44

2.2.7 Analysis of Results Campbell and Fiske (1959) describe a subjective method of assessing convergent and discriminant validity. This method requires the creation of a correlation table, which is divided into blocks. The symmetrical blocks on the diagonal of the correlation table are the heterotraitmonomethod blocks (Campbell and Fiske 1959; Kotz et al. 1985). These are the correlation coefficients of the traits within a single method. The nonsymmetrical blocks in the correlation table are the heterotrait-heteromethod blocks (Campbell and Fiske 1959; Kotz et al. 1985). The heterotrait-heteromethod blocks for the method of interest contain the validity diagonals. Table 2.1 presents the results of the Multi-Trait Multi-Method analysis for the accounting value scores created in Essay 1. The coefficients on the validity diagonals in Table 2.1 have been outlined in bold rectangles.

45

Table 2.1: Multi-Trait Multi-Method Validity Results y


T1M1 T2M1 T3M1 T4M1 T1M2 T2M2 T3M2 T4M2 T1M3 T4M3 T4M4 T4M5 T4M6 T4M7 1 -.842** -.859** -.865** T1M1 58 58 58 58 ** 1 .935** .955** -.842 T2M1 58 58 58 58 ** ** 1 .946** .935 -.859 T3M1 58 58 58 58 ** ** ** 1 .955 .946 -.865 T4M1 58 58 58 58 .374 -.655* -.823** -.618* 1 -.907** -.920** -.965** T1M2 12 12 12 12 13 13 13 13 .371 -.907** 1 .965** .848** -.168 .377 .640* T2M2 12 12 12 12 13 13 13 13 .443 -.920** .965** 1 .905** -.182 .441 .702* T3M2 12 12 12 12 13 13 13 13 -.448 .702* .859** .668* -.965** .848** .905** 1 T4M2 12 12 12 12 13 13 13 13 .195 -.077 -.150 -.068 .216 -.420 -.290 -.105 T1M3 35 35 35 35 12 12 12 12 * ** ** ** * * * .537 .625 .583 .499 -.652 .667 -.382 .586 T4M3 36 36 36 36 12 12 12 12 .306 -.405 -.360 -.264 .504 -.196 -.229 -.470 T4M4 23 23 23 23 7 7 7 7 -.444 -.442 -.495 -.513 -1.000** 1.000** 1.000** 1.000** T4M5 6 6 6 6 2 2 2 2 ** ** ** ** .493 -.458 -.554 -.526 -.525 -.494 -.449 .431 T4M6 36 36 36 36 11 11 11 11 * ** ** ** * -.505 -.621* -.696* .340 -.545 -.548 -.463 .663 T4M7 36 36 36 36 12 12 12 12 Pearson coefficients and country counts are presented. **Correlation is significant at the 0.01 level (2-tailed). *Correlation is significant at the 0.05 level (2-tailed). Traits: T1 - Professionalism vs. Statutory Control T2 - Uniformity vs. Flexibility T3 - Conservatism vs. Optimism T4 - Secrecy vs. Transparency Methods: M1 - Accounting Values Scores Created in Essay 1 M2 - Eddie (1990) M3 - Salter and Niswander (1995) M4 - Gray and Vint (1995) M5 - Zarzeski (1996) M6 - LaPorta et al. (1998) M7 - Hope (2003)

1 -.135 43 37 -.135 1 37 41 .245 -.638** 23 23 -.241 -.657 7 7 -.109 -.778** 35 37 .062 -.963** 38 41

1 27 .901 6 .737 22
*

**

.686 23

**

1 7 .550 7 .575 7

1 41 .819 37
**

1 42

Campbell and Fiske (1959) specify four requirements for the statistical support of construct validity (Kotz et al. 1985). First, the elements of the validity diagonal should be substantial (Campbell and Fiske 1959; Kotz et al. 1985). Using statistical significance as a measure for this requirement, only five of the ten coefficients pass this test. This first

46

requirement is a test of convergent validity (Campbell and Fiske 1959; Kotz et al. 1985). The accounting value scores created in Essay 1 exhibit limited convergent validity. Campbell and Fiskes (1959) second requirement is any element on the validity diagonal should be larger than all other elements in the corresponding row and column of its heterotraitheteromethod block (Kotz et al. 1985). For example, 0.374 should be larger than the absolute value of -0.655, -0.823, -0.618, -0.168, -0.182, and -0.448. Only one of the ten coefficients pass this test, as only Zarzeskis (1996) measure of disclosure is more closely related to the Secrecy versus Transparency accounting value than it is to the other three accounting values created in Essay 1. Campbell and Fiskes (1959) third requirement is any element on the validity diagonal should be larger than all the nondiagonal elements in the corresponding row of the heterotraitmonomethod block to its side and all nondiagonal elements in the corresponding column of the heterotrait-monomethod above it (Kotz et al. 1985). For example, 0.374 should be larger than the absolute value of -0.907, -0.920, -0.965, -0.842, -0.859, and -0.865. coefficients pass this test. Campbell and Fiskes (1959) fourth requirement is the same pattern should be exhibited by the nondiagonal elements of all heterotrait-monomethod blocks as well as all heterotraitheteromethod blocks (Kotz et al. 1985). For example, the pattern of correlation coefficients for the traits within method one should be similar to the pattern of correlation coefficients for the traits within method two. This test is the most subjective of the four, but the pattern of None of the ten

correlation coefficients for the traits within method one seems to be similar to the pattern of correlation coefficients for the traits within method two; however, that does not seem to be the case for the heterotrait-heteromethod blocks.

47

Campbell and Fiskes (1959) final three requirements are tests of discriminant validity (Kotz et al. 1985). The accounting values created in Essay 1 appear to show limited evidence of convergent validity and no evidence of discriminant validity. I replicate a study by Jaggi and Low (2000) as a second attempt to provide some evidence of construct validity of the accounting values created in Essay 1. 2.3 Replication of Jaggi and Low (2000) Jaggi and Low (2000) examine the impact of legal origin (common law versus code law) and national culture on the extent of firm-level disclosures in six countries. Based on previous theoretical and empirical studies (La Porta et al. 1998; Ball et al. 1999; Meek and Saudagaran 1990; Salter and Doupnik 1992), Jaggi and Low (2000) argue that legal origin has a significant influence on financial statement disclosures. They tested the following hypothesis: H1: The level of financial disclosures by firms in common law countries is higher than that in code law countries. Jaggi and Low (2000) also argue that due to the influence of cultural values on the development of a countrys legal origin, legal origin as a construct contains all of the explanatory variance necessary to describe the relationship between financial statement disclosures and cultural values. Additionally, legal origin as a construct contains explanatory power beyond that of cultural value; due to differences in the development of capital markets as well as in the ownership and capital structure at the firm level, the information needs of financial statement users differs in common law and code law countries (Jaggi and Low 2000; La Porta and Lopezde-Silanes 1998). Based on this, they tested the following hypothesis: H2: Influences of cultural values on financial disclosures by firms will be significantly less in common law countries compared to code law countries. Jaggi and Lows (2000) results show that firms in common law countries have higher levels of financial statement disclosures than firms from code law countries. They also show that 48

the impact of cultural values on financial disclosures in common law countries is insignificant and the results for code law countries were mixed and inconsistent with Grays (1988) hypotheses (Jaggi and Low 2000). Overall, they show no evidence of a consistent and direct impact of cultural values on financial statement disclosures when legal origin is included in the model (Jaggi and Low 2000). In an attempt to provide some support for the construct validity of the accounting values created in Essay 1, I replicate Jaggi and Lows (2000) data and retest their hypotheses. However, instead of using Hofstedes (1980) four cultural dimension scores to proxy for the cultural values construct, I use the Secrecy versus Transparency accounting value score quantified in Essay 1. 2.3.1 Variables and Regression Models For their dependent variable, Jaggi and Low (2000) used the firm-level disclosure scores from the third edition of the Center for International Financial Analysis and Researchs (CIFAR) International Financial Reporting Index for Industrial Companies (Bavishi 1993). The CIFARs disclosure scores are presented for the top 1000 global companies. This number was reduced in Jaggi and Lows (2000) sample to 505 firms from 28 countries due to the non-availability of Hofstedes (1980) country data and Global Vantages financial data. This was further reduced to the final sample of 401 firms from 6 countries by Jaggi and Lows (2000) screening requirement of at least 20 firms from each country. In replicating the data for the dependent variable, I use the fourth edition of the CIFAR firm-level disclosure scores (Bavishi 1995), due to the limited availability of the third edition. Similar to Jaggi and Low (2000), I begin with the CIFARs disclosure scores for the top 1000 global companies. This number is reduced in my sample to 541 firms from 31 countries due to

49

the non-availability of country data corresponding to the Secrecy versus Transparency accounting value index created in Essay 1 and COMPUSTAT Globals financial data. This is further reduced to the final sample of 382 firms from 6 countries by Jaggi and Lows (2000) screening requirement of at least 20 firms from each country. The data in Jaggi and Lows (2000) study are from fiscal year 1991 financial statements while my data represents fiscal year 1993; however, the six countries in my replication are the same countries represented in Jaggi and Lows (2000) study Canada, France, Germany, Japan, the United Kingdom, and the United States. The variables of interest in Jaggi and Lows (2000) study are legal origin and cultural values. Legal origin is a dichotomous variable with values of common law or code law and is taken from La Porta et al. (1998). I use the exact same data in the replication. As the measure of cultural values, Jaggi and Low (2000) use Hofstedes (1980) cultural dimension scores and relate them to the cultural preference for secrecy or transparency by using the relationships hypothesized by Gray (1988). In the replication, I use the Secrecy versus Transparency

accounting value score quantified in Essay 1. The control variables in Jaggi and Lows (2000) study are country-level market capitalization and firm-level measures of multi-nationality, size, and debt ratio. In the Jaggi and Low (2000) study, as well as in my replication, market capitalization is represented by the countrys average market capitalization in US$ divided by GDP in US$. Jaggi and Low (2000) collected their market capitalization data from the Emerging Stock Markets Factbook (1996) and the World Development Report (Mundial 1990, 1991, 1992) and my replication data is from the World Banks World Development Indicators database via the NationMaster website (http://www.nationmaster.com).

50

Jaggi and Lows (2000) measure of a firms multi-nationality is a dichotomous variable with firms categorized as multi-national or not and is taken from CIFARS Global Company Handbook (Bavishi 1992). Due to the limited availability of this publication, I use an alternate method to categorize the firms and create my dichotomous variable. I label a firm as multinational if they report any foreign income and/or foreign income tax expense. In my replication and Jaggi and Lows (2000) study, firm size is measured as total assets in US$ and debt ratio is calculated by dividing total debt by total assets. All of my financial data comes from fiscal year 1993 and is pulled from COMPUSTAT Global and Jaggi and Lows (2000) data is from fiscal year 1991 and is pulled from Global Vantage. Descriptive statistics of both sets of data are presented in Table 2.2.

51

Table 2.2: Descriptive Statistics on Regression Variables


CIFAR MKTC MNC SIZE DEBT Panel A: Jaggi & Low (2000) Full Sample (N=401) Mean Standard Deviation Minimum Maximum 73.09 6.03 47.00 88.00 0.74 0.34 0.25 1.31 0.78 0.41 0.00 1.00 10,395.68 18,393.97 297.42 184,325.50 0.28 0.15 0.00 0.89

Panel B: Replication Full Sample (N=382) Mean Standard Deviation Minimum Maximum 76.08 7.33 44.00 92.00 0.73 0.22 0.23 1.20 0.52 0.50 0.00 1.00 14,416.57 24,029.60 236.02 251,506.00 0.30 0.17 0.00 1.41

Panel C: Jaggi & Low (2000) Common Law Countries (N=263) Mean Standard Deviation Minimum Maximum 74.21 5.52 47.00 88.00 0.68 0.18 0.53 1.04 0.78 0.42 0.00 1.00 8,751.11 19,300.68 297.42 184,325.50 0.26 0.14 0.00 0.89

Panel D: Replication Common Law Countries (N=260) Mean Standard Deviation Minimum Maximum 77.43 6.41 50.00 92.00 0.82 0.16 0.59 1.20 0.56 0.50 0.00 1.00 11,904.09 25,101.40 339.30 251,506.00 0.28 0.16 0.00 1.41

Panel E: Jaggi & Low (2000) Code Law Countries (N=138) Mean Standard Deviation Minimum Maximum 70.96 6.40 52.00 88.00 0.85 0.51 0.25 1.31 0.79 0.41 0.00 1.00 13,529.89 16,135.61 569.14 84,825.50 0.31 0.16 0.04 0.80

Panel F: Replication Code Law Countries (N=122) Mean 73.19 0.54 0.42 19,771.04 0.35 Standard Deviation 8.29 0.20 0.50 20,664.14 0.19 Minimum 44.00 0.23 0.00 236.02 0.00 Maximum 90.00 0.69 1.00 119,031.79 0.76 Definition of Variables: CIFAR = Disclosure scores from Center for International Financial Analysis and Research (Bavishi 1995); MKTC = Mean Market Capitalization in US$ as a percentage of GDP in US$; MNC = 1 if the firm is a multinational, 0 otherwise; SIZE = Total Assets in US$ (millions); DEBT = Total Debt divided by Total Assets.

52

The full regression model tested by Jaggi and Lows (2000) is as follows: DISCi = i + i1UAi + i2PDi + i3INDi + i4MASi + i5LSi + i6DEBTi + i7SIZEi + i8MKTCi + i9MNCi + i, where: DISCi = disclosure index of company i; UAi = uncertainty avoidance value of company i; PDi = power distance value of company i; INDi = individualism value of company i; MASi = masculinity value of company i; LSi = 1 if company i belongs to common law countries, and 0 if company i belongs to code law countries; DEBTi = debt-to-asset ratio for company i; SIZEi = firm size, proxied by log of total assets of company i; MKTCi = country's market capitalization divided by GDP for company i; MNCi = 1, if company i is multinational, otherwise zero; i1-9 = coefficients of variables from 1 to 9; i = residual term. The full regression model tested in the replication is as follows: DISCi = i + i1SecrecyVsTransparencyi + i2LSi + i3DEBTi + i4SIZEi + i5MKTCi + i6MNCi + i , where: DISCi = disclosure index of company i; SecrecyVsTransparencyi = Secrecy versus Transparency accounting value score quantified in Essay 1 of company i; LSi = 1 if company i belongs to common law countries, and 0 if company i belongs to code law countries; DEBTi = debt-to-asset ratio for company i; SIZEi = firm size, proxied by log of total assets of company i; MKTCi = country's market capitalization divided by GDP for company i; MNCi = 1, if company i is multinational, otherwise zero; i1-6 = coefficients of variables from 1 to 6; i = residual term. The sub-sample regression model tested by Jaggi and Lows (2000) is as follows: DISCi = i + i1UAi + i2PDi + i3INDi + i4MASi + i5DEBTi + i6SIZEi + i7MKTCi + i8MNCi + i,

53

where: DISCi = disclosure index of company i; UAi = uncertainty avoidance value of company i; PDi = power distance value of company i; INDi = individualism value of company i; MASi = masculinity value of company i; DEBTi = debt-to-asset ratio for company i; SIZEi = firm size, proxied by log of total assets of company i; MKTCi = country's market capitalization divided by GDP for company i; MNCi = 1, if company i is multinational, otherwise zero; i1-8 = coefficients of variables from 1 to 8; i = residual term. The sub-sample regression model tested in the replication is as follows: DISCi = i + i1SecrecyVsTransparencyi + i2DEBTi + i3SIZEi + i4MKTCi + i5MNCi + i, where: DISCi = disclosure index of company i; SecrecyVsTransparencyi = Secrecy versus Transparency accounting value score quantified in Essay 1 of company i; DEBTi = debt-to-asset ratio for company i; SIZEi = firm size, proxied by log of total assets of company i; MKTCi = country's market capitalization divided by GDP for company i; MNCi = 1, if company i is multinational, otherwise zero; i1-5 = coefficients of variables from 1 to 5; i = residual term. 2.3.2 Analysis of Results As shown above in Table 2.2, Jaggi and Lows (2000) data provided empirical support for hypothesis 1. The mean of the disclosure scores for the common law country firms was 3.25 greater than those of the code country firms (t-value=5.07, p<0.001). The replication data also supports hypothesis 1. The mean of the disclosure scores for the common law country firms was 4.24 greater than those of the code country firms (t-value=5.47, p<0.001). Jaggi and Low (2000) test hypothesis 2 with hierarchical regression analysis. They first report the results of a partial model and then add variables until they finally report the results of the full model. Further, they avoid the negative effects of multicollinearity by using four separate final models instead of one. The Hofstede (1980) cultural dimension scores for the six

54

countries in their sample are correlated at a high level; therefore, Jaggi and Low (2000) must run four final models each containing the control variables, the legal origin variable, and one of the Hofstede (1980) cultural dimensions. I report their results in two of the columns within Table 2.3. The first column of Table 2.3 shows their partial model without the cultural proxy and the third column shows an aggregation of the results from their four final models, which included the cultural proxy variables. Table 2.3: Jaggi and Low (2000) Replication Regression Results - Full Sample
Jaggi & Low (2000) Table 5 Model 1 60.27 (0.000) + 1.64 (0.023) 5.52 (0.000) 0.51 (0.016) 1.15 (0.274) 3.96 (0.000) Full Sample Replication Model 1 60.67 (0.000) 11.52 (0.000) 1.61 (0.012) 2.05 (0.004) -7.69 (0.000) 0.86 (0.188) Jaggi & Low (2000) Table 5 Models 3 - 6 + Full Sample Replication Model 2 60.61 (0.000) 11.47 (0.000) 1.60 (0.015) 2.05 (0.004) -7.69 (0.000) 0.90 (0.337) 0.00 (0.491) Pred.

Intercept

MKTC

MNC

SIZE-log

DEBT

n.s.

LS

mixed results

Secrecy vs. Transparency

mixed results

R-Square 0.195 0.195 Adj. R-Sqr. 0.239 0.184 0.241 - 0.342 0.182 F-Statistic 26.07 18.21 22.15 - 35.70 15.14 F-Stat. Sig. 0.000 0.000 0.000 - 0.000 0.000 N 401 382 401 382 Unstandardized coefficients are presented. Significance levels are presented in parentheses below coefficients. All hypothesis tests are one-tailed. F tests of the models are two-tailed. Definition of Variables: SIZE-log = Logarithm of Total Assets in US$; LS = 1 if the firm belongs to a common law country, 0 if code law country firm; Secrecy vs. Transparency = Accounting Value Scores developed in Essay 1 (represented by Hofstede (1980) scores in Jaggi & Low (2000)); See Table 2.2 for other variables.

55

The results of Jaggi and Lows (2000) partial model are mostly as expected. Firms in countries with a high average market capitalization have higher disclosure scores than firms in countries with a lower average market capitalization. Multi-national firms have higher

disclosure scores than domestic firms. Larger firms have higher disclosure scores than smaller firms and firms from common law countries have higher disclosure scores than firms from code law countries. They did not find a statistically significant effect between debt ratio and financial statement disclosures. The results of the partial model in my replication are very similar to Jaggi and Lows (2000) results. Larger firms, multi-national firms, and firms from countries with a high average market capitalization all have higher disclosure scores than their opposites. However, contrary to expectations, my replication data shows that firms with higher debt ratios have lower disclosure scores than firms with lower debt ratios. Additionally, my replication data shows no relationship between legal origin and disclosure scores. The results of Jaggi and Lows (2000) full models show that larger firms, multi-national firms, and firms from countries with a high average market capitalization all have higher disclosure scores than their opposites. Their full model also shows no statistically significant relationship between debt ratio and financial statement disclosures. With regard to the legal origin and culture variables, their full models produce mixed results (i.e., regression coefficients with predicted relationships, no relationships, and relationships contrary to the prediction). The results of the full model in my replication are once again very similar to Jaggi and Lows (2000) results. Larger firms, multi-national firms, and firms from countries with a high average market capitalization all have higher disclosure scores than their opposites. Again, contrary to expectations, my replication data shows that firms with higher debt ratios have lower

56

disclosure scores than firms with lower debt ratios. As would be expected from the results of the aggregation of Jaggi and Lows (2000) four full models, my replication shows no statistically significant relationship between disclosure scores and either legal origin or culture. Next, Jaggi and Low (2000) split the sample into a sub-sample of firms from common law countries and a sub-sample of firms from code law countries. They then run their full model on these two sub-samples. Their results are reported in the second column of Table 2.4. For firms in common law countries, they found the expected results for the control variables. Specifically, larger firms, multi-national firms, firms with higher debt ratios, and firms from countries with a high average market capitalization all have higher disclosure scores than their opposites. With regard to the culture variable, they found mixed results (i.e., regression

coefficients with predicted relationships, no relationships, and relationships contrary to the prediction). For firms in code law countries, they found that larger firms and multi-national firms have higher disclosure scores than their opposites. They found mixed results for firms from countries with a high average market capitalization and no statistically significant relationship between debt ratio and financial statement disclosures. With regard to the culture variable, they found mixed results. These results led them to the conclusion that culture has no effect on the level of financial statement disclosures in the presence of a measure of legal origin. They generalized this to the larger conclusion that culture does not matter in the presence of a measure of legal origin.

57

Table 2.4: Jaggi and Low (2000) Replication Regression Results by Common Law and Code Law Countries
Divided Sample Replication Model 1 Panel A: Common Law Countries Intercept 54.85 (0.000) Pred. MKTC MNC SIZE-log DEBT Secrecy vs. Transparency R-Square Adj. R-Sqr. F-Statistic F-Stat. Sig. N Panel B: Code Law Countries Intercept MKTC MNC SIZE-log DEBT Secrecy vs. Transparency + + + + + + + + 22.75 (0.000) 2.76 (0.000) 0.90 (0.108) -3.40 (0.052) Jaggi & Low (2000) Table 6 Models 1 - 4 + + + + + mixed results 0.372 0.362 37.76 0.000 260 Divided Sample Replication Model 2 63.59 (0.000) 35.90 (0.000) 3.63 (0.000) 1.72 (0.008) -3.07 (0.062) -0.98 (0.000) 0.436 0.424 39.21 0.000 260

0.332 - 0.332 27.01 - 27.01 0.000 - 0.000 263

56.14 (0.000) -10.46 (0.012) 2.14 (0.087) 5.36 (0.000) -0.06 (0.495)

mixed results mixed results + + n.s. mixed results

34.96 (0.000) -34.46 (0.000) 0.22 (0.437) 6.36 (0.000) -1.46 (0.367) 0.89 (0.000) 0.338 0.309 11.83 0.000 122

R-Square 0.140 Adj. R-Sqr. 0.111 0.296 - 0.296 F-Statistic 4.77 12.49 - 12.49 F-Stat. Sig. 0.001 0.000 - 0.000 N 122 138 Unstandardized coefficients are presented. Significance levels are presented in parentheses below coefficients. All hypothesis tests are one-tailed. F tests of the models are two-tailed. Definition of Variables: See Tables 2.2 and 2.3

58

The replication results of the partial regression model on the two sub-samples are presented in the first column of Table 2.4. For the firms in common law countries, the results are very similar to the results of the partial model on the full sample. Specifically, larger firms, multi-national firms, and firms from countries with a high average market capitalization all have higher disclosure scores than their opposites. Additionally and contrary to Jaggi and Lows (2000) expectations and results, my replication data shows that firms with higher debt ratios have lower disclosure scores than firms with lower debt ratios. For the firms in code law countries, larger firms and multi-national firms have higher disclosure scores than their opposites. Contrary to Jaggi and Lows (2000) expectations, firms from countries with a high average market capitalization have lower disclosure scores than their opposites. These replication results also show no statistically significant relationship between debt ratio and financial statement disclosures. The replication results of the full regression model on the two sub-samples are presented in the third column of Table 2.4. For the firms in common law countries, the results are very similar to the results of the partial model on the full sample and the partial model on the common law country firm sub-sample. Specifically, larger firms, multi-national firms, and firms from countries with a high average market capitalization all have higher disclosure scores than their opposites. Additionally and contrary to expectations, my replication data shows that firms with higher debt ratios have lower disclosure scores than firms with lower debt ratios. With regard to the culture variable, the results show that, for firms in common law countries, firms in countries with a preference for transparency have higher disclosure scores. For the firms in code law countries, larger firms have higher disclosure scores than smaller firms. Contrary to

expectations, firms from countries with a high average market capitalization have lower

59

disclosure scores than their opposites.

These replication results also show no statistically

significant relationship between financial statement disclosures and debt ratio or multinationality. The most interesting finding of this replication is the statistically significant relationships between the Secrecy versus Transparency accounting value and the disclosure index with legal origin in the regression model. This is contrary to Jaggi and Lows (2000) findings and

conclusions. Similar to this are the results for average market capitalization. Jaggi and Low (2000) found the sign of the relationship was opposite between the two sub-samples (although with some mixed results in the code law sub-sample). This relationship is also seen in their interaction model (results not presented in this essay) with a significant interaction between legal origin and average market capitalization. They did not run an interaction between legal origin and their culture variables, but if they had they likely would have found a significant interaction between the variables. This may have led them to see that there is a relationship between culture and financial statement disclosure that varies with legal origin. Hope (2003) was motivated to reexamine this issue in part because he believed that limitations in Jaggi and Lows (2000) sample size led them to the conclusion that culture had no effect on financial statement disclosure levels. As noted earlier, with a larger sample size and more diverse collection of countries, Hopes (2003) multivariate analysis showed relationships between disclosure practices and the Secrecy versus Transparency accounting value that was one way for common law countries and opposite for code law countries, with the common law countries relationship more closely matching Grays (1988) hypothesis. This is similar to my finding that for firms in common law countries higher disclosure scores are associated with a countrys preference for transparency, but for firms in code law countries lower disclosure scores

60

are associated with a countrys preference for transparency. It is possible that, in addition to the limitations pointed out by Hope (2003) of a small sample of firms and low level of country diversity, Jaggi and Lows (2000) study suffered an increased level of statistical noise because of the high correlations between the measures of culture and subsequent multiple full models needed to address the potential for multicollinearity. These limitations led Jaggi and Low (2000) to the likely false conclusion that culture does not matter in the presence of a measure of legal origin. My quantified accounting values may help to prevent these false conclusions in future studies. 2.4 Conclusion A test of construct validity failed to provide empirical support for the validity of the accounting values quantified in Essay 1. Multi-Trait Multi-Method analysis results demonstrated that the quantified accounting values appear to show limited evidence of convergent validity and no evidence of discriminant validity. This is a major shortcoming in the usefulness of the accounting values scores. As a second attempt to provide some evidence of the usefulness of the accounting values scores, I replicated a study by Jaggi and Low (2000), which examined the relationship between financial statement disclosures and legal origin and culture. Their results led them to the conclusion that culture has no effect on the level of financial statement disclosures in the presence of a measure of legal origin. They generalized this to the larger conclusion that culture does not matter in the presence of a measure of legal origin. My replication provides evidence that they may have arrived at these conclusions in error. This is also supported by a reexamination of the issue by Hope (2003). Although the accounting values scores created in Essay 1 are not as statistically sound as I would like, they do provide help in addressing research questions dealing with culture and

61

international accounting practices and systems. In Essay 3, I use the quantified accounting values to help address one of these questions.

(Nobes and Parker 1991; David and Brierley 1985; Chow and Wong-Boren 1987; Choi 1991, 1993; Waterhouse various; Lybrand 1988)

62

Essay 3 - Using Grays Accounting Values to Explain Deviations from IFRS 3.1 Introduction With the advances in the scope and global acceptance of the International Accounting Standards Boards (IASB) International Financial Reporting Standards (IFRS), international accounting standards have received increasing attention in the academic literature (e.g., 2006; Haverals 2007; Nobes 2006; Suzuki 2007; Tyrrall et al. 2007). Many of these studies are documenting the effect of IFRS on the accounting and business practices of a particular country. Domestic accounting regulatory bodies are considering these effects when deciding on the level of adoption or adaptation of IFRS (Zeghal and Mhedhbi 2006). Gray (1988) has presented a framework for describing the accounting culture of a country and, in Essay 1, I have provided a quantified proxy for his four accounting values. In this essay, I provide empirical evidence concerning the effect of accounting values, specifically conservatism and secrecy, on a countrys level of adoption of international standards. This work provides an important starting point for future international accounting research. In general, the inclusion of a measure of the intrinsic preferences of a country

concerning accounting measurement and disclosure will be a critical variable in all types of international research. More specifically, the study of the deviations of domestic GAAP from IFRS is ongoing (Beckman et al. 2007; Daske 2006; Ding et al. 2007; Horton and Serafeim 2006; Van der Meulen et al. 2007; van Tendeloo and Vanstraelen 2005). For example, in the case of the U.S., domestic GAAP and IFRS have been converging over time (Haverty 2006); however, there is still a substantial lack of comparability between the two sets of standards (Haverty 2006). This study would provide empirical support for the notion that, despite the global push for convergence in accounting standards, country-level preferences for accounting

63

measurement and disclosure provides important explanatory power in differences between domestic GAAP and IFRS. 3.2 Literature Review and Hypothesis Development More than 100 countries now require or permit the use of IFRS or are converging with IFRS (International Accounting Standards Board 2009); however, the level of adoption can vary greatly between countries (Dean and Clarke 2005). Some countries are accepting IFRS

completely and even considering disbanding their local accounting standard settings boards (e.g., Canada 10 ) while others are adopting IFRS, but keeping their standards setters in order to adapt and modify the international standards to meet local needs, thereby producing various forms of IFRS (Dean and Clarke 2005). For example, on January 1, 2005, Australia adopted IFRS;

however, the Australian Accounting Standards Board (the AASB) continues to issue Australian equivalents to International Financial Reporting Standards (A-IFRS). The costs and benefits of IFRS adoption weigh heavily in the decisions concerning a countrys level of adoption. One theory contends that accounting standards that produce high quality financial statements, such as IFRS, lead to a lower cost of capital for firms (Levitt 1998; Dargenidou et al. 2006; Leuz and Verrecchia 2000; International Accounting Standards Board 2002; Sengupta 1998; Bailey et al. 2006; Easley and O'Hara 2004). However, others have found that the convergence of IFRS and local GAAP does not automatically produce positive outcomes such as lower cost of capital (Barth et al. 1999; Cuijpers and Buijink 2005; Cohen 2006; Leuz 2003; Daske 2006; Botosan 1997). Jermakowicz and Gornik-Tomaszewski (2006) surveyed 112 European Union (EU) companies required to use both EU GAAP and IFRS for financial statement reporting purposes.
Canada considered disbanding their Accounting Standards Board (AcSB), but has since decided against it. However, after January 1, 2011, accounting standards for Canadian public companies will be set by the IASB and the AcSB will simply attempt to influence future standards.
10

64

Among other things, they reported that the respondents found the process of IFRS compliance is costly, complex, and burdensome. They also reported that these companies do not expect to lower their cost of capital by implementing IFRS. Other studies have reached similar

conclusions (Cuijpers and Buijink 2005; Cohen 2006; Daske 2006; Botosan 1997). These costs and benefits (or lack thereof) are important considerations when deciding the level of convergence between domestic GAAP and IFRS. It is important to consider that these costs and benefits are examined by individual members of a decision-making group (e.g., the US FASB) through a lens shaped by an individuals socialization into a societys cultural norms. For accounting situations, this socialization has been described in terms of accounting values (Gray 1988; Perera 1989) and can be represented by the index of quantified accounting values created in Essay 1. These accounting values are Professionalism versus Statutory Control

(professionalism) which influences authority, Uniformity versus Flexibility (uniformity) which influences enforcement, Conservatism versus Optimism (conservatism) which influences measurement, and Secrecy versus Transparency (secrecy) which influences disclosure; however, as the IFRS are primarily addressing measurement and disclosure, the conservatism and secrecy accounting values are the focus of this essay. 3.2.1 Conservatism versus Optimism Conservatism is defined as a societys preference for a cautious approach to measurement so as to cope with the uncertainty of future events as opposed to a more optimistic, laissez-faire, risk-taking approach (Gray 1988). The IFRS has been theoretically argued and empirically shown to be less conservative than the standards of many countries. Ball (2006) points out numerous instances where measurement under IFRS is closer to fair value than historical cost as well as his expectation that the standards will continue to move in that

65

direction. Although Cairns (2006) argues against the notion that IFRS are fair value based standards, he does point out the numerous contexts in which the less conservative fair value measurement is required instead of the more conservative historical cost. Ding and Su (2008) describe the results of a study commissioned by the Chinese Ministry of Finance on the impact of a switch from Chinese GAAP to IFRS. The study of 1348

companies reports that about 1000 companies would increase their beginning balances of net assets by an average of 9 percent, while 259 companies would record an average decrease of 4 percent (Ding and Su 2008). Bertoni and De Rosa (2006) use the restatements of the 2004 annual reports of 42 companies to empirically show that Italian GAAP is more conservative than IFRS. Separate from the issue of IFRS conservatism, but relevant to the discussion, Wstemann and Kierzek (2005) identify inconsistencies in IFRS revenue recognition. For these reasons, countries that prefer a cautious approach to measurement would likely be wary of adopting IFRS. Although it encompasses more than just measurement, it is pertinent in the discussion of measurement IFRSs true and fair view principle is the source of much heated discussion and uncertainty (e.g., Alexander 2006; Dean and Clarke 2005; Nobes 2006; Wstemann and Kierzek 2005, 2006). Until the meaning of true and fair, with respect to IFRS, becomes more stable, clear, and less uncertain, countries that value conservatism will be reluctant to embrace IFRS. Therefore, countries that value conservatism should have higher levels of deviation from IFRS, which leads to the following hypothesis: H1: There is a positive association between conservatism and the size of differences between IFRS and domestic GAAP.

66

This relationship should be clearer when focusing on individual accounting standards that concern asset and liability valuation as well as revenue recognition, which leads to the following sub-hypothesis: H1a: There is a positive association between conservatism and the size of differences between IFRS measurement-specific standards and domestic GAAP measurement-specific standards. 3.2.2 Secrecy versus Transparency Secrecy is defined as a societys preference for confidentiality and the restriction of disclosure of information about the business only to those who are closely involved with its management and financing as opposed to a more transparent, open and publicly accountable approach (Gray 1988). The IFRS has been theoretically argued and empirically shown to be more transparent than the standards of many countries. The U.S. Securities and Exchange Commission (SEC) sees IFRS as a more transparent global financial reporting environment (Rosen 2007). For example, Weienberger et al. (2004) point out that the extensive reporting obligations of US GAAP and IFRS are due to a focus on capital market oriented users and German GAAP, with its focus on creditor protection, limitations on profit distribution, and linkage with tax reporting requirements, will produce a balance sheet and income statement that is less informative for investment decisions than IFRS or US GAAP. Daske and Gebhardt (2006) present empirical evidence that disclosure quality increases significantly in Austrian, German, and Swiss firms that report in IFRS rather than their domestic GAAP. Also, Hodgdon et al. (2008) find that firm compliance with the disclosure requirements of IFRS reduces information asymmetry and enhances the ability of financial analysts to provide more accurate forecasts. However, the increased transparency required by IFRS comes with increased costs of compliance for firms (Ball 2006; Tarca 2004; Jermakowicz and Gornik-

67

Tomaszewski 2006). Additionally, Jeanjean and Stolowy (2008) present empirical evidence that a change from local GAAP to IFRS does not increase earnings quality as measured by earnings management. Therefore, firms in countries that value secrecy will find it hard to justify these costs with the supposed benefits and will pressure their standard setters to resist convergence with IFRS. Consequently, countries that value secrecy should have higher levels of deviation from IFRS, which leads to the following hypothesis: H2: There is a positive association between secrecy and the size of differences between IFRS and domestic GAAP. This relationship should be clearer when focusing on accounting standards that attempt to increase transparency (i.e., increased disclosure), which leads to the following sub-hypothesis: H2a: There is a positive association between secrecy and the size of differences between IFRS disclosure-specific standards and domestic GAAP disclosure -specific standards. 3.3 Research Method Ding et al. (2005) examined the relationship between the deviation of domestic GAAP from IFRS and societal culture. They used Hofstedes (1980) cultural dimensions as one of their measures of societal culture and presented their own absence and divergence scores as the measure of the deviation of domestic GAAP from IFRS, which are further discussed by Ding et al. (2007). As a test of my hypotheses I use Ding et al.s (2005) measure of the level of IFRS pronouncements missing from domestic GAAP along with the accounting scores created in Essay 1 instead of Hofstedes (1980) cultural dimensions. 3.3.1 Variables and Regression Models Ding et al. (2005) created an index of a countrys nonconformity in its domestic GAAP with IFRS. They obtained their data from Nobes (2001) report of comparisons of domestic GAAP to IFRS by audit firm partners located in 62 counties all over the world for the fiscal year

68

ending December 31, 2001. These partners were asked to compare their countrys domestic GAAP to benchmarks based on international accounting standards. Ding et al. (2005) quantified the differences between the various domestic GAAPs and IFRS reported by Nobes (2001). They identified 111 potential nonconforming items between domestic GAAP and IFRS and then counted the actual differences in the items for all 62 countries. Ding et al. (2005) reported the nonconformities as an absence of a specific rule from domestic GAAP or as an instance of a divergence of an existing domestic GAAP standard from the corresponding IFRS. The range of the nonconformity score, which is the sum of the absence score and the divergence score, is 0 to 111 with 0 representing a perfect harmony with IFRS and 111 a complete deviation from IFRS. Ding et al. (2005) further divide their absence score into a score for the absence of a specific rule on recognition and measurement from domestic GAAP and a score for the absence of a specific rule requiring disclosure from domestic GAAP. I use Ding et al.s (2005) measures of the absence of IFRS pronouncements from domestic GAAP as the dependent variable in my hypothesis tests. I choose absence over the divergence score and the nonconformity score for two reasons. First, with only 42 observations in my hypothesis testing models, I would like to use a dependent variable with the greatest potential for impact in order to overcome any potential lack of statistical power. A complete absence of an IFRS from domestic GAAP is easier to objectively measure and more telling about that country than a divergence of domestic GAAP from IFRS. Additionally, the qualitative nature and extent of the divergence is important, but cannot be objectively captured by the divergence score. Ding et al.s (2005) nonconformity score is a sum of the absence score and the divergence score and, therefore, carries the same concerns as the divergence score alone. The second reason that I use the absence score as my dependent variable is the availability of

69

individual scores concerning measurement and disclosure, which are required to test my subhypotheses. Only absence is reported in terms of measurement and disclosure scores as well as an aggregate score, the measures for divergence and nonconformity are only available as aggregate scores. Ding et al.s (2005) variables of interest in their hypothesis testing models are Hofstedes (1980) four cultural dimension scores and a measure of legal origin. For my hypothesis testing models, I replace Hofstedes (1980) four cultural dimension scores with the Conservatism versus Optimism and Secrecy versus Transparency scores quantified in Essay 1. I use the same

measure of legal origin as Ding et al. (2005). They use the same common law or code law classification as La Porta et al. (1997) and use the classification method devised by the University of Ottawa for missing countries (Ding et al. 2005). The complete dataset is presented in Table 3.1 and correlations of the data are presented in Table 3.2. Table 3.1: Dependent and Control Variables
Panel A: Data Country Argentina Australia Austria Brazil Canada China Colombia Denmark Ecuador Egypt Finland France Germany Greece Hong Kong Hungary India Indonesia Ireland Israel Italy Japan Malaysia Mexico Morocco Netherlands New Zealand Philippines Poland Portugal Russia Singapore Slovenia South Africa Absence 47 22 34 36 4 23 31 22 22 21 18 40 14 40 18 12 0 15 27 18 30 0 56 10 23 24 23 29 38 4 40 7 Absence (Measurement) 34 19 17 24 0 17 18 12 11 13 9 22 12 24 14 7 0 6 18 13 24 0 43 8 17 22 18 21 30 3 25 6 Absence (Disclosure) 13 3 17 12 4 6 13 10 11 8 9 18 2 16 4 5 0 9 9 5 6 0 13 2 6 2 5 8 8 1 15 1 Use of Common IFRS PDI IDV MAS UAI Law -1 49 46 56 86 0 1 36 90 61 51 1 1 11 55 79 70 0 1 69 38 49 76 0 -1 39 80 52 48 1 80 20 66 30 0 -1 0 1 18 74 16 23 0 1 0 1 80 38 52 68 0 1 33 63 26 59 0 1 68 71 43 86 0 1 35 67 66 65 0 1 60 35 57 112 0 1 68 25 57 29 1 46 80 88 82 0 -1 77 48 56 40 1 -1 78 14 46 48 0 1 28 70 68 35 1 0 13 54 47 81 1 1 50 76 70 75 0 -1 54 46 95 92 0 -1 104 26 50 36 1 0 81 30 69 82 0 70 46 53 68 0 1 38 80 14 53 0 1 22 79 58 49 1 -1 94 32 64 44 0 68 60 64 93 0 1 63 27 31 104 0 93 39 36 95 0 1 74 20 48 8 1 0 1 49 65 63 49 1 Disclosure Requirement 0.50 0.75 0.25 0.25 0.92 0.42 0.58 0.00 0.50 0.50 0.75 0.42 0.33 0.92 0.92 0.50 0.67 0.67 0.67 0.75 0.92 0.58 0.50 0.67 0.83 0.42 1.00 0.83 Anti-Director Rights 4 4 2 3 4 1 3 2 2 2 2 1 1 4 2 2 3 3 0 3 3 0 2 4 4 2 3 4 Access to Equity 4.1 6.4 6.1 5.0 6.0 5.0 5.7 3.2 4.9 6.0 6.0 5.8 5.2 6.4 6.5 5.2 5.9 5.8 5.4 6.4 6.3 4.5 5.8 6.5 5.8 5.5 6.1 6.3 Market Sophistication 3.9 6.0 5.5 5.3 6.0 4.3 5.8 3.2 3.2 5.9 5.7 6.1 4.5 6.0 5.0 3.6 5.7 5.3 4.3 5.4 5.4 4.4 5.8 5.7 4.0 5.2 5.9 5.8 Logarithm of GNI 3.71 4.56 4.60 3.67 4.56 3.44 4.71 3.45 3.13 4.61 4.56 4.56 4.34 4.45 2.91 3.15 4.66 4.27 4.51 4.58 3.74 3.90 4.63 4.44 3.15 4.26 4.47 3.73

70

Table 3.1: Dependent and Control Variables (continued)


Panel A: Data (continued) Absence Country Absence (Measurement) South Korea 15 10 Spain 28 10 Sweden 10 5 Switzerland 42 27 Taiwan 19 16 Thailand 29 23 Turkey 47 35 United Kingdom 0 0 United States 6 2 Venezuela 41 35 Zimbabwe Mean 23.45 15.95 Std. Dev. 14.10 10.58 Min. 0 0 Max. 56 43 Panel B: Description of the Variables Absence Absence (Measurement) Absence (Disclosure) Use of IFRS PDI IDV MAS UAI Common Law Absence (Disclosure) 5 18 5 15 3 6 12 0 4 6 7.50 5.22 0 18 Use of IFRS -1 1 1 0 -1 -1 1 1 -1 1 0 0.28 0.92 -1 1 PDI IDV MAS UAI 60 18 39 85 57 51 42 86 31 71 5 29 34 68 70 58 58 17 45 69 64 20 34 64 66 37 45 85 35 89 66 35 40 91 62 46 81 12 73 76 55.5 50.4 53.2 23.0 23.9 18.7 11 12 5 104 91 95 Common Law 0 0 0 0 0 1 0 1 1 0 1 62.7 0.31 24.3 0.47 8 0 112 1 Disclosure Requirement 0.75 0.50 0.58 0.67 0.75 0.92 0.50 0.83 1.00 0.17 0.50 0.62 0.24 0.00 1.00 Anti-Director Rights 2 2 2 1 3 3 2 4 5 1 3 2.51 1.19 0 5 Access to Equity 5.3 5.2 5.9 6.1 6.2 6.1 5.8 6.5 6.4 3.4 5.9 5.66 0.80 3.2 6.5 Market Sophistication 4.6 5.2 5.8 6.4 5.0 4.3 4.1 6.7 6.7 3.9 4.0 5.12 0.94 3.2 6.7 Logarithm of GNI 4.25 4.44 4.64 4.76 4.13 3.48 3.73 4.60 4.65 3.78 2.53 4.10 0.59 2.53 4.76

Absence of specific rules on recognition/measurement or disclosure based on 2000 survey of partners in the large accountancy firms in more than 60 countries (Nobes 2001). Source: Ding et al. (2005). Absence of specific rules on recognition and measurement based on 2000 survey of partners in the large accountancy firms in more than 60 countries (Nobes 2001). Source: Ding et al. (2005). Absence of specific rules requiring disclosures based on 2000 survey of partners in the large accountancy firms in more than 60 countries (Nobes 2001). Source: Ding et al. (2005). A trichotomous variable based on data as of February 2008 with the value of negative one for countries that prohibit the use of IFRS, zero for countries that permit IFRS, and one for countries that require IFRS. Source: IAS Plus website (http://www.iasplus.com/country/useias.htm). Power Distance Index is the extent to which the less powerful members of organizations and institutions (like the family) accept and expect that power is distributed unequally. Source: Hofstede (1980). Individualism on the one side versus its opposite, collectivism, is the degree to which individuals are integrated into groups. Source: Hofstede (1980). Masculinity versus its opposite, femininity, refers to the distribution of roles between the genders. Source: Hofstede (1980). Uncertainty Avoidance Index deals with a society's tolerance for uncertainty and ambiguity. Source: Hofstede (1980). An indicator variable with the value of one for common law countries and zero otherwise.

Disclosure Requirement

An index aggregating investor protection. One is added for each of the following six proxies for the strength of specific disclosure requirements pertaining to the promoter's: (1) prospectus, (2) insiders' compensation, (3) ownership by large shareholders, (4) insider ownership, (5) contracts outside the normal course of business, and (6) transactions with related parties. The index is the average of the six proxies. Source: La Porta et al. (2006). An index aggregating shareholder rights. The index is formed by adding 1 when: (1) the country allows shareholders to mail their proxy vote, (2) shareholders are not required to deposit teir shares prior to the General Shareholders' Meeting, (3) cumulative voting or proportional representation of minorities on the board of directors is allowed, (4) an oppressed minorities mechanism is in place, (5) the minimum percentage of share capital that entitles a shareholder to call for an Extraordinary Shareholder's Meeting is less than or equal to 10 percent (the sample median), and (6) when shareholders have preemptive rights that can only be waived by a shareholders meeting. Source: La Porta et al. (1998). Average country response from the World Economic Forum's 2005 Executive Opinion Survey to the following statement: Raising money by issuing shares on the local stock market is (1 = nearly impossible, 7 = quite possible for a good company). Source: Lopez-Carlos et al. (2005). Average country response from the World Economic Forum's 2005 Executive Opinion Survey to the following statement: The level of sophistication of financial markets in your country is (1 = lower than international norms, 7 = higher than international norms). Source: Lopez-Carlos et al. (2005). The logarithm of 2006 GNI per capita (formerly GNP per capita), which is the gross national income, converted to U.S. dollars using the World Bank Atlas method, divided by the midyear population. Source: World Bank website (http://www.worldbank.org).

Anti-Director Rights

Access to Equity Market Sophistication Logarithm of GNI

71

Table 3.2: Pearson Correlations for Variables in Tests on Absence Measures


1. 2. 3. 4. 5. 6. 1. Absence Sig. 2. Absence (Measurement) 0.950 Sig. (0.00) 3. Absence (Disclosure) 0.766 0.527 Sig. (0.00) (0.00) 4. Conservatism 0.351 0.433 0.064 Sig. (0.02) (0.00) (0.69) 5. Secrecy 0.260 0.369 -0.053 0.958 Sig. (0.10) (0.02) (0.74) (0.00) 6. PDI 0.176 0.322 -0.185 0.724 0.693 Sig. (0.27) (0.04) (0.25) (0.00) (0.00) 7. IDV -0.205 -0.290 0.039 -0.712 -0.672 -0.695 Sig. (0.20) (0.07) (0.81) (0.00) (0.00) (0.00) 8. MAS 0.054 0.069 0.005 0.132 0.161 0.047 Sig. (0.74) (0.67) (0.97) (0.41) (0.31) (0.77) 9. UAI 0.474 0.391 0.486 0.343 0.226 0.118 Sig. (0.00) (0.01) (0.00) (0.03) (0.16) (0.46) 10. Common Law -0.483 -0.393 -0.507 -0.134 -0.028 -0.166 Sig. (0.00) (0.01) (0.00) (0.40) (0.86) (0.30) Statistically significant correlations are bolded (2-tailed tests). 7. 8. 9.

0.053 (0.74) -0.158 (0.32) 0.225 (0.16)

0.150 (0.35) 0.087 -0.532 (0.59) (0.00)

Univariate analysis, as presented in Table 3.2, provides initial empirical support for three of the four hypotheses. The Pearson correlation coefficient for the relationship between Absence and Conservatism shows a statistically significant positive association (p = 0.02), which supports hypothesis H1. The Pearson correlation coefficient for the relationship between Absence

(Measurement) and Conservatism shows a statistically significant positive association (p < 0.01), which supports hypothesis H1a. The Pearson correlation coefficient for the relationship between Absence and Secrecy shows a marginally statistically significant positive association (p = 0.10), which supports hypothesis H2. The Pearson correlation coefficient for the relationship between Absence (Disclosure) and Secrecy shows no statistically significant association, which fails to support hypothesis H2a. Additionally, Table 3.2 reports a high correlation between the

72

accounting values Conservatism and Secrecy. This is due to a limitation in the method of quantification and is discussed in Essay 1. The full regression model tested by Ding et al. (2005) is as follows: ABSi = i + i1PDIi + i2IDVi + i3MASi + i4UAIi + i5Common lawi + i, where: ABSi = Absence index of company i; PDIi = power distance value of company i; IDVi = individualism value of company i; MASi = masculinity value of company i; UAIi = uncertainty avoidance value of company i; Common lawi = 1 if company i belongs to common law countries, and 0 if company i belongs to code law countries; i1-5 = coefficients of variables from 1 to 5; i = residual term. The full regression model used for testing hypothesis 1 is as follows: ABSi = i + i1Conservatismi + i2Common lawi + i, where: ABSi = Absence index of company i; Conservatismi = Conservatism versus Optimism accounting value score quantified in Essay 1 of company i; Common lawi = 1 if company i belongs to common law countries, and 0 if company i belongs to code law countries; i1-2 = coefficients of variables from 1 to 2; i = residual term. The full regression model used for testing hypothesis 2 is as follows: ABSi = i + i1Secrecyi + i2Common lawi + i, where: ABSi = Absence index of company i; Secrecyi = Secrecy versus Transparency accounting value score quantified in Essay 1 of company i; Common lawi = 1 if company i belongs to common law countries, and 0 if company i belongs to code law countries; i1-2 = coefficients of variables from 1 to 2; i = residual term. The full regression model used for testing hypothesis 1a is as follows: ABS-Mi = i + i1Conservatismi + i2Common lawi + i,

73

where: ABS-Mi = Measurement-specific absence index of company i; Conservatismi = Conservatism versus Optimism accounting value score quantified in Essay 1 of company i; Common lawi = 1 if company i belongs to common law countries, and 0 if company i belongs to code law countries; i1-2 = coefficients of variables from 1 to 2; i = residual term. The full regression model used for testing hypothesis 2a is as follows: ABS-Di = i + i1Secrecyi + i2Common lawi + i, where: ABS-Di = Disclosure-specific absence index of company i; Secrecyi = Secrecy versus Transparency accounting value score quantified in Essay 1 of company i; Common lawi = 1 if company i belongs to common law countries, and 0 if company i belongs to code law countries; i1-2 = coefficients of variables from 1 to 2; i = residual term. 3.3.2 Analysis of Results Ding et al. (2005) hypothesized a non-directional relationship between their absence measure and culture (Perera 1994), as well as a negative relationship between their absence measure and common law countries (Hope 2003; Jaggi and Low 2000). They were unable to empirically support either relationship due to the failure of their hypothesis testing model to reach statistical significance at the 0.05 level. In Table 3.3 Panel A Models 1 and 2, I have replicated Ding et al.s (2005) tests (model significances of p < 0.05). Model 2 produced statistically significant support for their hypothesis of a negative relationship between their absence measure and common law countries (p < 0.05). Model 1 produced a statistically significant relationship between absence and one of Hofstedes (1980) cultural dimensions uncertainty avoidance (p < 0.01); however, this relationship becomes only marginally significant (p = 0.10) when the common law variable is added in Model 2.

74

Table 3.3: OLS Regression Results


PDI Sig. IDV Sig. MAS Sig. UAI Sig. Common law Sig. Conservatism Sig. Secrecy Sig. Observations 2 R 2 Adj R F Sig. (F ) 2 R Change F Change Sig. (F Change) Pred. Sign Model 1 (?) 0.059 (0.77) (?) -0.092 (0.66) (?) -0.012 (0.94) (?) 0.454 (0.00) () H1 ( + ) H2 ( + ) 41 0.24 0.16 2.905 (0.04) 41 0.31 0.21 3.140 (0.02) 0.07 3.327 (0.08) 42 0.24 0.22 12.734 (0.00) Panel A: Absence Model 2 Model 3 0.048 (0.81) -0.059 (0.77) 0.039 (0.79) 0.287 (0.10) -0.313 -0.491 (0.04) (0.00) Model 4 Model 5

-0.454 (0.00) 0.283 (0.02)

-0.485 (0.00)

42 0.32 0.29 9.190 (0.00) 0.08 4.524 (0.04) Panel B: Absence Panel C: Absence (Measurement) (Disclosure) Pred. Sign Model 1 Model 2 Pred. Sign Model 1 Model 2 Common law () -0.401 -0.350 () -0.514 -0.516 Sig. (0.00) (0.01) (0.00) (0.00) Conservatism H1a ( + ) 0.381 Sig. (0.00) Secrecy H2a ( + ) -0.070 Sig. (0.31) Observations 42 42 42 42 2 R 0.16 0.30 0.27 0.27 2 Adj R 0.14 0.27 0.25 0.23 F 7.662 8.506 14.395 7.196 Sig. (F ) (0.01) (0.00) (0.00) (0.00) 2 R Change 0.14 0.01 F Change 8.008 0.262 Sig. (F Change) (0.01) (0.61) Tests of significance on variables with a predicted sign are one-tailed. Tests of significance on variables with no predicted sign are two-tailed. All model tests of significance are two-tailed. Statistically significant tests are bolded.

0.238 (0.04) 42 0.30 0.26 8.288 (0.00) 0.06 3.156 (0.08)

75

Table 3.3 Panel A Models 3, 4, and 5 show tests of the relationship between the deviation of domestic GAAP from IFRS and two of Grays (1988) accounting values conservatism and secrecy. In the base model (Model 3), common law countries are shown to be negatively related (p < 0.01) to Ding et al.s (2005) absence measure. Model 4 adds conservatism to Model 3 and Model 5 adds secrecy to Model 3. Both of the accounting values are positively and significantly (p < 0.05) related to the absence measures. This provides empirical support for hypotheses H1 and H2 as countries that value conservatism and secrecy are shown to have higher deviations from IFRS when compared to domestic GAAP. Table 3.3 Panel B presents tests of the two sub-hypotheses. Ding et al. (2005) also deconstruct their absence measure into a measure of the absence from domestic GAAP of measurement-specific IFRSs and a measure of the absence from domestic GAAP of disclosurespecific IFRSs. These two measures are the dependent variables in the four Models reported in Panel B. Common law country classification is the control variable in the two base models (Models 1 and 3) and continues to be statistically significant (p < 0.01). When added to Model 1, conservatism is positively and significantly (p < 0.01) related to the measurement-specific absence measure. This provides empirical support for hypothesis H1a as countries that value conservatism are shown to have higher deviations from IFRS measurement-specific standards when compared to domestic GAAP. When added to Model 3.3, secrecy has no significant explanatory power beyond the variability explained by the common law variable; therefore, hypothesis H2a is not empirically supported. 3.3.3 Additional Test of H1 and H2 Despite the ingenuity of Ding et al.s (2005) absence measures, a potential drawback of their work is the age of the data. It is based on data collected in 2000 (Nobes 2001) and the

76

years since then have seen increased convergence between domestic GAAP and IFRS. As an alternative measure for additional tests, I have created a Use of IFRS measure based on data reported by IAS Plus (www.iasplus.com) as of February 2008. This measure categorizes

countries as prohibiting (value of -1), permitting (0), or requiring (1) the use of IFRS. The measure is similar to one used by Hope et al. (2006); they categorized countries as adopting IFRS (value of 1) or not (0) as of 2005. Hope et al. (2006) investigated the factors that influence a countrys decision to adopt IFRS. They hypothesized that, due to bonding (Coffee 2002) and signaling theories (Spence 1973; Tarca 2004), countries with less stringent domestic investor protection would be more likely to adopt IFRS to attract foreign equity investment and as a signal to the rest of the world. As measures of domestic investor protection, they used the countrys disclosure requirements (La Porta et al. 2006), anti-director rights (La Porta et al. 1998), and capital market access (Schwab et al. 1999). For control variables, they used the logarithm of per capita 1996 gross national product (Mundial 1996) and a proxy for market development the ratio of stock market capitalization held by small shareholders to gross domestic product (La Porta et al. 1999). Hope et al. (2006) reported (Table 5 Model 5) that a countrys disclosure requirements (negative relationship), anti-director rights (negative relationship), and capital market access (positive relationship), as well as the control variable market capitalization (positive relationship), all had significant effects (p 0.05) on the countrys decision to adopt IFRS. Gross national product, their other control variable, had at least a marginally significant (p 0.10) negative relationship with a countrys decision to adopt IFRS by 2004 (Table 4 Models 5 and 6), but no significant relationship with a countrys decision to adopt IFRS by 2005 (Table 5 Models 5 and 6) (Hope et al. 2006).

77

As an additional test of hypotheses H1 and H2, I replicate the tests in Hope et al. (2006) with updated data and the conservatism and secrecy accounting values variables. The dependent variable for this additional test is created from data from the IAS Plus website (http://www.iasplus.com) operated by the Deloitte Touche Tohmatsu CPA firm. The use of IFRS by country as of February 2008 is reported in terms of required, permitted, and prohibited. I have quantified this variable by coding countries that require their public companies to use IFRS as 1, countries that permit the use of IFRS as 0, and those that prohibit the use of IFRS as 1. The disclosure requirements variable (La Porta et al. 2006) is recent and not modified for the additional test of hypotheses H1 and H2. The anti-director rights variable (La Porta et al. 1998) is also not modified because neither updated data nor an acceptable substitute was available. An updated version of the capital market access variable (Lopez-Claros et al. 2005) was available and is used for this additional test. Hope et al. (2006) used two control variables. The first was a proxy for market development the ratio of stock market capitalization held by small shareholders to gross domestic product (La Porta et al. 1999). I was not able to update this proxy; however, I found financial market sophistication (Lopez-Claros et al. 2005) in the World Economic Forums Global Competitiveness Report 2005-2006. The other control variable used by Hope et al. (2006) was the 1996 per capita GNP by country. This has been updated to the 2006 per capita GNI for the additional test of hypotheses H1 and H2. Finally, although Hope et al. (2006) included a common law variable in their sensitivity analysis and found no significant relationship between it and the adoption of IFRS, I have included it in this additional test because of the associations found in my earlier tests and reported in Table 3.3 Panels A and B.

78

Table 3.4: Pearson Correlations for Variables in Tests on Use of IFRS


1. 2. 3. 4. 1. Use of IFRS Sig. 2. Conservatism -0.267 Sig. (0.10) 3. Secrecy -0.239 0.958 Sig. (0.14) (0.00) 4. Disclosure Requirement -0.412 -0.184 -0.076 Sig. (0.01) (0.26) (0.64) 5. Anti-Director Rights -0.209 -0.188 -0.085 0.520 Sig. (0.20) (0.25) (0.61) (0.00) 6. Access to Equity -0.091 -0.488 -0.367 0.727 Sig. (0.58) (0.00) (0.02) (0.00) 7. Market Sophistication 0.207 -0.803 -0.732 0.465 Sig. (0.21) (0.00) (0.00) (0.00) 8. Logarithm of GNI 0.374 -0.737 -0.716 0.158 Sig. (0.02) (0.00) (0.00) (0.34) 9. Common Law -0.115 -0.124 -0.002 0.645 Sig. (0.49) (0.45) (0.99) (0.00) Statistically significant correlations are bolded (2-tailed tests). 5. 6. 7. 8.

0.450 (0.00) 0.362 (0.02) 0.038 (0.82) 0.629 (0.00)

0.697 (0.00) 0.334 (0.04) 0.538 (0.00)

0.755 (0.00) 0.393 (0.01)

-0.027 (0.87)

79

Table 3.5: Ordinal Probit Regression Results

Common Law Sig. Disclosure Requirement Sig. Anti-Director Rights Sig. Access to Equity Sig. Market Sophistication Sig. Logarithm of GNI Sig. Conservatism Sig. Secrecy Sig. Observations Goodness of Fit Tests Log Likelihood Model Chi-square Sig. (Model Chi-square)

Use of IFRS IFRS Not Permitted = -1 IFRS Permitted = 0 IFRS Required = 1 Pred. Sign Model 1 Model 2 (+) 3.360 3.517 (0.03) (0.03) () -12.881 -15.163 (0.00) (0.00) () -0.400 -0.358 (0.20) (0.23) (+) 1.067 1.227 (0.16) (0.14) (?) -0.493 0.192 (0.67) (0.89) (?) 3.173 3.598 (0.06) (0.05) H1 ( ) 0.033 (0.16) H2 ( ) 39 46.275 24.522 (0.00) 39 45.130 25.667 (0.00)

Model 3 3.918 (0.03) -18.427 (0.00) -0.318 (0.27) 1.318 (0.14) 0.535 (0.70) 4.297 (0.04)

0.060 (0.05) 39 42.885 27.912 (0.00) 0.51 0.61 0.39

Pseudo R2 Cox and Snell 0.47 0.48 Nagelkerke 0.56 0.58 McFadden 0.35 0.36 Tests of significance on variables with a predicted sign are one-tailed. Tests of significance on variables with no predicted sign are two-tailed. All model tests of significance are two-tailed. Statistically significant tests are bolded.

The dependent variable, Use of IFRS, is a trichotomous variable with possible values of 1 (IFRS prohibited), 0 (IFRS permitted), or 1 (IFRS required), which warrants the use of ordinal probit regression analysis (McCullagh 1980) 11 . The results of this analysis are reported in Table

As an additional test, I transformed the trichotomous dependent variable into a dichotomous variable with a value of 1 for IFRS permitted or required and 0 for IFRS prohibited and tested the hypotheses with bivariate logistic

11

80

3.5. Model 1 presents the results of the relationship between the control variables and the use of IFRS. As found in the primary hypothesis testing reported in Table 3.3, common law countries are less likely to have differences between their domestic GAAP and IFRS (p < 0.05). Of the three hypotheses in Hope et al. (2006), only the negative relationship between disclosure requirement and the required use of IFRS is statistically significant (p < 0.01). Interestingly, the marginal significance (p < 0.10) of the negative relationship between gross national product and the use of IFRS reported in Model 5 of Table 3.4 by Hope et al. (2006) is a positive relationship (p < 0.10) in my analysis (Table 3.5 Model 1). Table 3.5 Models 2 and 3 present the results of additional tests of hypotheses H1 and H2. Although the significant variables in Model 1 retain their significant relationships to the dependent variable, neither of the variables of interest (conservatism in Model 2 and secrecy in Model 3) is significantly related to the use of IFRS. 3.4 Conclusion This essay examines the effect of accounting values on the level of use of IFRS in approximately 40 countries. The decision to replace current domestic GAAP with a set of foreign constructed accounting standards must be justified with many benefits. These benefits, such as high quality financial statements and increased transparency, are valued differently by countries with differing accounting values. I hypothesize and find empirical support for the hesitancy of countries that value conservatism and secrecy to fully implement IFRS. A limitation of this study is the disappearance of the relationship between accounting values and the level of adoption of IFRS for later time periods (the analysis presented in Table 3.5 with dependent variable data from 2008 versus the analysis presented in Table 3.3 with dependent variable data from 2001). The changing role of the control variable Gross National

regression analysis. The results were similar to the results of the ordinal probit regression analysis, including no empirical support for the hypotheses.

81

Product (Income) can possibly shed some light on this limitation. Hope et al. (2006) posited that firms in countries with smaller economies would prefer the bonding effects associated with IFRS and press their regulators to adopt the international standards. They found empirical evidence for this relationship using dependent variable data of adoption of IFRS by 2005. In Table 3.5, I present tests of this relationship using dependent variable data of adoption of IFRS by 2008 and find that countries with larger economies are more likely to require IFRS. This could simply be the effects of increase convergence between local GAAP and IFRS, which would erode the explanatory power of accounting values and other predictor variables as more and more countries are adopting some form of IFRS. A future study could examine the explanatory power of accounting values at predicting the actual level of implementation or compliance with IFRS of firms operating in countries that have adopted IFRS.

82

Summary
This dissertation reports the creation, validation, and application of the quantification of Grays (1988) accounting values framework. Essay 1 reported the creation of scores for each of Grays (1988) four accounting values Professionalism versus Statutory Control, Uniformity versus Flexibility, Conservatism versus Optimism, and Secrecy versus Transparency. These scores cover 58 countries and are a combination of the GLOBE (House et al. 2004) cultural values scores in the manner hypothesized in Grays (1988) accounting values framework. The resulting accounting values scores are normally distributed and standardized to a range of 0 through 100. The scores exhibit means that range from 45.81 to 54.96 and standard deviations that range from 23.33 to 27.15 meaning that about 68 percent of the scores have an approximate range of 26 through 77, which also suggests a normal distribution. An empirical test of face validity using multiple regression analysis shows that the quantified accounting values scores conform to the theoretical predictions of Grays (1988) original four hypotheses. Essay 2 reported the validation of the quantified accounting values scores. The primary validation method, Multi-Trait Multi-Method Analysis, did not empirically support the construct validity of the quantified accounting values scores as it provided only limited support for the convergent validity and failed to provide empirical support for the discriminant validity of the quantified accounting values scores. A secondary validation method, an ad-hoc research study replication, provided limited empirical validation to the accounting values scores. Essay 3 used the scores that were quantified in Essay 1 and validity tested in Essay 2 as the primary variables in the explanation of differing levels of a countrys adoption of the International Financial Reporting Standards (IFRS). The essay examined the effect of

measurement and disclosure related accounting values on the level of use of IFRS in 42

83

countries. The decision to undertake the cost associated with replacing current domestic GAAP with a set of foreign constructed accounting standards must be justified with many benefits. These benefits, such as high quality financial statements and increased transparency, are valued differently by countries with differing accounting values. I hypothesized and found empirical support for the hesitancy of countries that value conservatism and secrecy to fully implement IFRS. As an additional test of the hypotheses I used recent data gathered from the IAS Plus website categorizing the use of IFRS by country in terms of required, permitted, and prohibited. This more recent test of the effect of measurement and disclosure related accounting values on the level of use of IFRS in these 39 countries failed to empirically support the hesitancy of countries that value conservatism and secrecy to require or allow the use of IFRS. The three essays of this dissertation provide the starting point for a new method of analyzing the relationship between accounting values and various accounting phenomena. The accounting values proxies used by previous studies were problematic for one of two reasons either the study relied on the inclusion of three or more cultural variables as a proxy for a single accounting value or it relied on a newly created proxy that had not been tested for validity nor reused in subsequent studies. The quantified accounting values created in this dissertation would mitigate both of these issues by (1) replacing multiple variables as a test of a single effect with one variable that tests a single effect, which should aid in the interpretation of empirical results, and by (2) providing a uniform measure of accounting values that can be used in future accounting values studies, which should aid in the generalizability of empirical results. Although the tests in Essay 2 failed to conclusively validate the quantified accounting values scores, their usefulness was shown in Essay 3 as an explanatory factor in a countrys decision on how fully to implement the International Financial Reporting Standards.

84

A major limitation in the quantification of Grays (1988) accounting values framework is the high correlation between the four accounting values. The absolute values of the correlations are in the range of 0.842 to 0.955. These high correlations are caused by using many of the same eight cultural dimensions to create each of the accounting values scores in a summative manner. These high correlations could cause multicollinearity in regression models where two or more of the quantified accounting values are in the same model. The high correlations also suggest a larger problem of a lack of discriminant validity, which was shown to be the case in Essay 2. The lack of discriminant validity suggests that four separate and distinct accounting values were not captured in the quantification of the accounting values framework, but rather four slightly different views of a single aggregated accounting values construct was captured. This limitation arose due to the theoretical summative combination of the cultural values rather than a factor analysis, which would have produced four distinct and orthogonally-related accounting values. Factor analysis was not available as a method of quantification because of the non-availability of the required data. A factor analysis method of quantification would have required multiple observations per country and the GLOBE data provides only one observation per country 12 . Additionally, a factor analysis method of quantification would have required multiple and different components for each accounting value. While multiple components were available for each accounting value, these multiple components were the same components used in the creation of all four accounting values, albeit in different quantities. This limitation can only be overcome through the use of a different set of source data for the quantification. A second limitation in this dissertation arises from the conclusion in Essay 3 that a more recent test of the effect of measurement and disclosure related accounting values on the level of
12

In fact, the GLOBE team captures different, yet similar, dimensions for two societies in each of three nations (Germany, South Africa, and Switzerland); however, this would not be sufficient for the factor analysis requirements.

85

use of IFRS has disappeared as more countries adopt IFRS. The conclusion was reached because the primary hypothesis testing model used data collected in 2001 and a secondary hypothesis testing model used data collected in 2008. The result of the first test showed a statistically significant accounting value effect on the level of IFRS adoption, while the result of the second test showed no statistically significant accounting value effect on the level of IFRS adoption. However, this conclusion is based on a rudimentary measure of the level of use of IFRS. The measure in the primary hypothesis testing model is the product of hand collecting data on 111 potential nonconforming items between domestic GAAP and IFRS for each of 62 countries. By contrast, the measure in the secondary hypothesis testing model simply categorizes countries as prohibiting, permitting, or requiring the use of IFRS. This dramatic drop in variance within the measure could solely account for the finding of no statistically significant accounting value effect on the level of IFRS adoption leading to the inference of an incorrect conclusion. This limitation can be overcome through the collection of more accurate data as a recent measure of the level of IFRS adoption. Future research based on the work in this dissertation could include addressing both limitations above as well as investigating the effect accounting values have on the relationship between actual IFRS implementation in practice and the level of IFRS adoption within a countrys accounting rules and regulations. For example, a countrys generally accepted

accounting practices could differ from the letter of the law as laid out in IFRS. Could accounting values help explain any differences by country of the actual practice of accounting from the formalized rule of law set forth in the national accounting rules and regulations?

86

References
Adams, C. A., P. Weetman, and S. J. Gray. 1993. Reconciling national with international accounting standards. European Accounting Review 2 (3):471-494. Adhikari, A., and R. H. Tondkar. 1992. Environmental Factors Influencing Accounting Disclosure Requirements of Global Stock Exchanges. Journal of International Financial Management & Accounting 4 (2):75-105. Alexander, D. 2006. Legal Certainty, European-ness and Realpolitik. European Accounting Review 15:65-80. Archambault, J. J., and M. E. Archambault. 2003. A multinational test of determinants of corporate disclosure. The International Journal of Accounting 38 (2):173-194. Arnold, D. F., Sr., R. A. Bernardi, and P. E. Neidermeyer. 2001. The association between European materiality estimates and client integrity, national culture, and litigation. International Journal of Accounting 36 (4):459-483. Bailey, W., G. Andrew Karolyi, and C. Salva. 2006. The economic consequences of increased disclosure: Evidence from international cross-listings. Journal of Financial Economics 81 (1):175-213. Ball, R. 2006. International Financial Reporting Standards (IFRS): pros and cons for investors. Accounting & Business Research 36:5-27. Ball, R., S. P. Kothari, and A. Robin. 1999. The Effect of International Institutional Factors on Properties of Accounting Earnings. SSRN eLibrary. Barrett, M. E. 1976. Financial Reporting Practices: Disclosure and Comprehensiveness in an International Setting. Journal of Accounting Research 14 (1):10-26. Barth, M. E., G. Clinch, and T. Shibano. 1999. International accounting harmonization and global equity markets. Journal of Accounting and Economics 26 (1-3):201-235. Bao, B. H., and D. H. Bao. 2004. Income Smoothing, Earnings Quality and Firm Valuation. Journal of Business Finance & Accounting 31 (9-10):1525-1557. Baskerville, R. F. 2003. Hofstede never studied culture. Accounting, Organizations & Society 28 (1):1-14. Bavishi, V. B. 1991. International Accounting and Auditing Trends. 2nd ed. Princeton, N.J.: Center for International Financial Analysis and Research. . 1992. CIFAR's Global Company Handbook. Princeton, N.J.: Center for International Financial Analysis and Research.

87

. 1993. International Accounting and Auditing Trends. 3rd ed. Princeton, N.J.: Center for International Financial Analysis and Research. . 1995. International accounting and auditing trends. 4th ed. Princeton, N.J.: Center for International Financial Analysis & Research. Beckman, J., C. Brandes, and B. Eierle. 2007. German Reporting Practices: An Analysis of Reconciliations from German Commercial Code to IFRS or US GAAP. ADVANCES IN INTERNATIONAL ACCOUNTING 20:253. Bertoni, M., and B. De Rosa. 2006. Measuring balance sheet conservatism: empirical evidence from Italian first time adopters of IFRS. Paper read at Emerging issues in international accounting & business conference, at Padova, Italy. Botosan, C. A. 1997. Disclosure Level and the Cost of Equity Capital. The Accounting Review 72 (3):323-349. Buck, T., and A. Shahrim. 2005. The translation of corporate governance changes across national cultures: the case of Germany. Journal of International Business Studies 36 (1):42-61. Cairns, D. 2006. The Use of Fair Value in IFRS. European Accounting Review 15:5-22. Campbell, D. T., and D. W. Fiske. 1959. Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin 56 (2):81-105. Chanchani, S., and A. MacGregor. 1999. A synthesis of cultural studies in accounting. Journal of Accounting Literature 18:1-30. Chanchani, S., and R. Willett. 2004. An empirical assessment of Gray's accounting value constructs. International Journal of Accounting 39 (2):125-154. Choi, F. D. S. 1973. Financial disclosure and entry to the European capital market. Journal of Accounting Research 11 (2):159-175. . 1991. Handbook of International Accounting. New York: Wiley. . 1993. Handbook of International Accounting: Supplement. New York: Wiley. Chow, C. W., and A. Wong-Boren. 1987. Voluntary financial disclosure by Mexican corporations. The Accounting Review 62 (3):533-541. Chow, L. M. Y., G. K. K. Chau, and S. J. Gray. 1995. Accounting Reforms in China: Cultural Constraints on Implementation and Development. Accounting & Business Research 26 (1):29-49. Chui, A. C. W., A. E. Lloyd, and C. C. Y. Kwok. 2002. The Determination of Capital Structure: Is National Culture a Missing Piece to the Puzzle? Journal of International Business Studies 33 (1):99-127.

88

Coffee, J. C. 2002. Racing towards the Top?: The Impact of Cross-Listings and Stock Market Competition on International Corporate Governance. Columbia Law Review 102 (7):1757-1831. Cohen, D. A. 2006. Does Information Risk Really Matter? An Analysis of the Determinants and Economic Consequences of Financial Reporting Quality. SSRN eLibrary. Coopers Lybrand. 1988. International Financial Reporting and Auditing - A Guide to Regulatory Requirements. New York: Coopers & Lybrand. Cronbach, L., and P. Meehl. 1955. Construct validity in psychological tests. Psychological Bulletin 52 (28):1-302. Cuijpers, R., and W. Buijink. 2005. Voluntary adoption of non-local GAAP in the European Union: A study of determinants and consequences. European Accounting Review 14 (3):487-524. Dargenidou, C., S. McLeay, and I. Raonic. 2006. Expected earnings growth and the cost of capital: an analysis of accounting regime change in the European financial market. Abacus 42 (3-4):388-414. Daske, H. 2006. Economic Benefits of Adopting IFRS or US-GAAP: Have the Expected Costs of Equity Capital Really Decreased? Journal of Business Finance & Accounting 33 (34):329373. Daske, H., and G. Gebhardt. 2006. International financial reporting standards and experts' perceptions of disclosure quality. Abacus 42 (3/4):461-498. David, R., and J. E. C. Brierley. 1985. Major Legal Systems in the World Today. 3rd ed. London: Stevens & Sons. Dean, G., and F. Clarke. 2005. 'True and fair' and 'fair value' accounting and legal will-o'-thewisps. Abacus 41 (2):1-8. Debreceny, R., G. L. Gray, and A. Rahman. 2002. The determinants of Internet financial reporting. Journal of Accounting and Public Policy 21 (4):371-394. DiMaggio, P. J., and W. W. Powell. 1983. The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields. American Sociological Review 48 (2):147-160. . 1991. The New Institutionalism in Organizational Analysis. Chicago: University of Chicago Press. Ding, Y., O.-K. Hope, T. Jeanjean, and H. Stolowy. 2007. Differences between domestic accounting standards and IAS: Measurement, determinants and implications. Journal of Accounting & Public Policy 26 (1):1-38.

89

Ding, Y., T. Jeanjean, and H. Stolowy. 2005. Why do national GAAP differ from IAS? The role of culture. International Journal of Accounting 40 (4):325-350. Ding, Y., and X. Su. 2008. Implementation of IFRS in a regulated market. Journal of Accounting and Public Policy 27 (6):474-479. Doupnik, T. S., and E. L. Riccio. 2006. The influence of conservatism and secrecy on the interpretation of verbal probability expressions in the Anglo and Latin cultural areas. International Journal of Accounting 41 (3):237-261. Doupnik, T. S., and M. Richter. 2003. Interpretation of uncertainty expressions-a cross-national study. Accounting, Organizations and Society 28 (1):15-35. Doupnik, T. S., and S. B. Salter. 1993. An Empirical test of a Judgemental International Classification of Financial Reporting Practices. Contact 24 (1):41-60. . 1995. External Environment, Culture, and Accounting Practice: A Preliminary Test of A General Model of International Accounting Development. International Journal of Accounting 30 (3):189-207. Doupnik, T. S., and G. T. Tsakumis. 2004. A Critical Review of Tests of Gray's Theory of Cultural Relevance and Suggestions for Future Research. Journal of Accounting Literature 23:1-48. Easley, D., and M. O'Hara. 2004. Information and the Cost of Capital. The Journal of Finance 59 (4):1553-1583. Eddie, I. A. 1990. Asia Pacific cultural values and accounting systems. Asia Pacific International Management Forum 16:22-30. Frank, W. G. 1979. An Empirical Analysis of International Accounting Principles. Journal of Accounting Research 17 (2):593-605. Gray, S. J. 1988. Towards a Theory of Cultural Influence on the Development of Accounting Systems Internationally. Abacus 24:1-15. Gray, S. J., L. G. Campbell, and J. C. Shaw. 1984. International financial reporting: A comparative international survey of accounting requirements and practices in 30 countries. New York: St. Martin's Press. Gray, S. J., and H. M. Vint. 1995. The impact of culture on accounting disclosures: some international evidence. Asia-Pacific Journal of Accounting 21:33-43. Hair, J. F., W. C. Black, B. J. Babin, R. E. Anderson, and R. L. Tatham. 2006. Multivariate Data Analysis. Sixth ed. Upper Saddle River, NJ: Pearson Prentice-Hall.

90

Hanges, P. J., and M. W. Dickson. 2004. The development and validation of the GLOBE culture and leadership scales. In Culture, leadership, and organizations: The GLOBE study of 62 societies, edited by R. J. House, P. J. Hanges, M. Javidan, P. W. Dorfman and V. Gupta. Thousand Oaks, Calif.: Sage Publications, 122151. Haverals, J. 2007. IAS/IFRS in Belgium: Quantitative analysis of the impact on the tax burden of companies. Journal of International Accounting, Auditing & Taxation 16 (1):69-89. Haverty, J. L. 2006. Are IFRS and U.S. GAAP converging?: Some evidence from People's Republic of China companies listed on the New York Stock Exchange. Journal of International Accounting, Auditing & Taxation 15 (1):48-71. Hodgdon, C., R. H. Tondkar, D. W. Harless, and A. Adhikari. 2008. Compliance with IFRS disclosure requirements and individual analysts' forecast errors. Journal of International Accounting, Auditing and Taxation 17 (1):1-13. Hofstede, G. H. 1980. Culture's consequences, international differences in work-related values. Beverly Hills: Sage Publications. . 1984. Cultural dimensions in management and planning. Asia Pacific Journal of Management 1 (2):81-99. . 2001. Culture's consequences : comparing values, behaviors, institutions, and organizations across nations. 2nd ed. Thousand Oaks, Calif.: Sage Publications. Hope, O.-K. 2003. Firm-level Disclosures and the Relative Roles of Culture and Legal Origin. Journal of International Financial Management & Accounting 14 (3):218-248. Hope, O.-K., J. Jin, and T. Kang. 2006. Empirical evidence on jurisdictions that adopt IFRS. Journal of International Accounting Research 5 (2):1-20. Horton, J., and G. Serafeim. 2006. Market Response to and the Value Relevance of Reconciliation Adjustments from UK GAAP to IFRS GAAP: First Evidence from the UK. Available at SSRN: http://ssrn.com/abstract=923582. House, R. J., P. J. Hanges, M. Javidan, P. W. Dorfman, and V. Gupta. 2004. Culture, leadership, and organizations: The GLOBE study of 62 societies. Thousand Oaks, Calif.: Sage Publications. Hussein, M. E. 1996. A comparative study of cultural influences on financial reporting in the US and The Netherlands. The International Journal of Accounting 31 (1):95-120. International Accounting Standards Board. 2002. International Accounting Standards Committee Foundation, Annual Report 2002. London. . IFRSs around the world. International Accounting Standards Committee Foundation 2009 [cited. Available from http://www.iasb.org/About+Us/About+the+IASB/IFRSs+around+the+world.htm.

91

International Finance Corporation. 1996. Emerging Stock Markets Factbook. Washington, D.C.: International Finance Corporation. Jaggi, B., and P. Y. Low. 2000. Impact of Culture, Market Forces and Legal System on Financial Disclosures. International Journal of Accounting 35 (4):495-519. Jeanjean, T., and H. Stolowy. 2008. Do accounting standards matter? An exploratory analysis of earnings management before and after IFRS adoption. Journal of Accounting and Public Policy 27 (6):480-494. Jermakowicz, E. K., and S. Gornik-Tomaszewski. 2006. Implementing IFRS from the perspective of EU publicly traded companies. Journal of International Accounting, Auditing & Taxation 15 (2):170-196. King, R. D., and J. C. Langli. 1998. Accounting Diversity and Firm Valuation. International Journal of Accounting 33 (5):529567. Kirkman, B. L., K. B. Lowe, and C. B. Gibson. 2006. A quarter century of Culture's Consequences: a review of empirical research incorporating Hofstede's cultural values framework. Journal of International Business Studies 37 (3):285-320. Kotz, S., N. L. Johnson, and C. B. Read 1985. Encyclopedia of statistical sciences. Vol. 5, Lindeberg condition to multitrait-multimethod matrices. Vol. 5. New York, N.Y.: Wiley. La Porta, R., and F. Lopez-de-Silanes. 1998. Capital Markets and Legal Institutions. In Beyond the Washington Consensus: Institutions Matter, edited by S. J. y. P. Burki, Guillermo. Washington, D.C.: The World Bank, 67-86. La Porta, R., F. Lopez-de-Silanes, and A. Shleifer. 1999. Corporate Ownership Around the World. Journal of Finance 54 (2):471-517. . 2006. What Works in Securities Laws? Journal of Finance 61 (1):1-32. La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R. W. Vishny. 1997. Legal Determinants of External Finance. Journal of Finance 52 (3):11311150. . 1998. Law and Finance. Journal of Political Economy 106 (6):1113-1155. . 2000. Investor protection and corporate governance. Journal of Financial Economics 58 (1-2):3-27. Leuz, C. 2003. IAS Versus U.S. GAAP: Information Asymmetry-Based Evidence from Germany's New Market. Journal of Accounting Research 41 (3):445-472. Leuz, C., and R. E. Verrecchia. 2000. The Economic Consequences of Increased Disclosure. Journal of Accounting Research 38:91-124.

92

Levitt, A. 1998. The Importance of High Quality Accounting Standards. Accounting Horizons 12 (1):79-82. Lopez-Claros, A., K. Schwab, and M. E. Porter. 2005. The Global Competitiveness Report 20052006: Palgrave Macmillan. Lynn, B. E. 1999. Culture and intellectual capital management: a key factor in successful ICM implementation. International Journal of Technology Management 18 (5):590-603. Massey, F. J. J. 1951. The Kolmogorov-Smirnov test of goodness of fit. Journal of the American Statistical Association 46 (253):68-78. McCullagh, P. 1980. Regression Models for Ordinal Data. Journal of the Royal Statistical Society. Series B (Methodological) 42 (2):109-142. McSweeney, B. 2002a. Hofstede's model of national cultural differences and their consequences: A triumph of faith a failure of analysis. Human Relations 55 (1):89-118. . 2002b. The essentials of scholarship: A reply to Geert Hofstede. Human Relations 55 (11):1363-1372. Meek, G. K., C. B. Roberts, and S. J. Gray. 1995. Factors Influencing Voluntary Annual Report Disclosures By U.S., U.K. And Continental European Multinational Corporations. Journal of International Business Studies 26 (3):555-572. Meek, G. K., and S. M. Saudagaran. 1990. A survey of research on financial reporting in a transnational context. Journal of Accounting Literature 9 (145):82. Mueller, G. G. 1967. International accounting. New York: Macmillan. . 1968. Accounting Principles Generally Accepted in the United States Versus those Generally Accepted Elsewhere. International Journal of Accounting of Accounting Education and Research 3 (2):93-102. Mundial, B. 1990. World Development Report. New York: Oxford University Press. . 1991. World Development Report. New York: Oxford University Press. . 1992. World Development Report. New York: Oxford University Press. . 1996. World Development Report 1996: From Plan to Market. Washington/New York: World Bank/Oxford University Press. Nair, R. D., and W. G. Frank. 1980. The Impact of Disclosure and Measurement Practices on International Accounting Classifications. The Accounting Review 55 (3):426-450. Nobes, C. W. 1983. A Judgemental International Classification of Financial Reporting Practices. Journal of Business Finance and Accounting 10 (1):1-19.

93

. 1984. International classification of financial reporting. London: Croom Helm. , ed. 2001. GAAP 2001 - a survey of national accounting rules benchmarked against international accounting standards: IFAD. . 2006. Revenue Recognition and EU Endorsement of IFRS. European Accounting Review 15 (3):81-89. Nobes, C. W., and R. H. Parker. 1991. Comparative international accounting: Prentice Hall New York. Oliver, C. 1991. Strategic Responses to Institutional Processes. The Academy of Management Review 16 (1):145-179. Perera, M. H. B. 1989. Towards a framework to analyse the impact of culture on accounting. International Journal of Accounting 24:42-56. . 1994. Culture and international accounting: Some thoughts on research issues and prospects. ADVANCES IN INTERNATIONAL ACCOUNTING 7:267285. Price Waterhouse. various. Doing Business in ... New York: Price Waterhouse. Radebaugh, L. H., G. Gebhardt, and S. J. Gray. 1995. Foreign Stock Exchange Listings: A Case Study of Daimler-Benz. Journal of International Financial Management & Accounting 6 (2):158-192. Radebaugh, L. H., S. J. Gray, and E. L. Black. 2006. International accounting and multinational enterprises. 6th ed. New York: J. Wiley. Roberts, C. B., and S. B. Salter. 1999. Attitudes towards Uniform Accounting: Cultural or Economic Phenomena? Journal of International Financial Management & Accounting 10 (2):121-142. Rosen, C. 2007. SEC, Users Voice Support for IFRS at Roundtable. Journal of Accountancy 203 (6):73-73. Salter, S. B. 1998. Corporate financial disclosure in emerging markets: Does economic development matter? The International Journal of Accounting 33 (2):211-234. Salter, S. B., and T. S. Doupnik. 1992. The relationship between legal systems and accounting practices: a classification exercise. Advances in International Accounting 5 (1):3-22. Salter, S. B., D. M. Guffey, and J. J. McMillan. 2001. Truth, Consequences and Culture: A Comparative Examination of Cheating and Attitudes about Cheating among U.S. and U.K. Students. Journal of Business Ethics 31 (1):37-50.

94

Salter, S. B., and F. Niswander. 1995. Cultural Influence On the Development of Accounting Systems Internationally: A Test of Gray's [1988] Theory. Journal of International Business Studies 26 (2):379-397. Schultz, J. J., Jr., and T. J. Lopez. 2001. The impact of national influence on accounting estimates: Implications for international accounting standard-setters. International Journal of Accounting 36 (3):271-290. Schwab, K., M. E. Porter, and J. D. Sachs. 1999. The Global Competitiveness Report 1999. Geneva, Switzerland: World Economic Forum. Schwartz, S. H. 1992. Universals in the content and structure of values: Theoretical advances and empirical tests in 20 countries. In Advances in experimental social psychology, edited by M. Zanna. New York: Academic, 1-65. . 1994. Beyond individualism/collectivism: New cultural dimensions of values. In Individualism and collectivism: Theory, method, and applications, edited by U. Kim, H. C. Triandis, C. Kagitcibasi, S. C. Choi and G. Yoon. London: Sage, 85119. . 1999. A Theory of Cultural Values and Some Implications for Work. Applied Psychology 48 (1):23-47. Scott, W. R. 1987. The Adolescence of Institutional Theory. Administrative Science Quarterly 32 (4):493-511. . 2001. Institutions and Organizations. 2nd ed. Thousand Oaks, CA: Sage Publications. Selznick, P. 1948. Foundations of the Theory of Organization. American Sociological Review 13 (1):25-35. Selznick, P., P. Nonet, and H. M. Vollmer. 1969. Law, Society, and Industrial Justice. New York: Russell Sage Foundation. Sengupta, P. 1998. Corporate Disclosure Quality and the Cost of Debt. The Accounting Review 73 (4):459-474. Sharp, D. J., and S. B. Salter. 1997. Project Escalation and Sunk Costs: A Test of the International Generalizability of Agency and Prospect Theories. Journal of International Business Studies 28 (1):101-121. Shenkar, O. 2001. Cultural Distance Revisited: Towards a More Rigorous Conceptualization and Measurement of Cultural Differences. Journal of International Business Studies 32 (3). Singhvi, S. S., and H. B. Desai. 1971. An empirical analysis of the quality of corporate financial disclosure. The Accounting Review 46 (1):129-138.

95

Sondergaard, M. 1994. Research Note: Hofstede's Consequences: A Study of Reviews, Citations and Replications. Organization Studies (Walter de Gruyter GmbH & Co. KG.) 15 (3):447. Spence, A. M. 1973. Job Market Signaling. Quarterly Journal of Economics 87 (3):355-374. Stulz, R. M., and R. Williamson. 2003. Culture, openness, and finance. Journal of Financial Economics 70 (3):313-349. Sudarwan, M., and T. J. Fogarty. 1996. Culture and accounting in Indonesia: An empirical examination. The International Journal of Accounting 31 (4):463-481. Suzuki, T. 2007. Accountics: Impacts of internationally standardized accounting on the Japanese socio-economy. Accounting, Organizations & Society 32 (3):263-301. Tarca, A. 2004. International Convergence of Accounting Practices: Choosing between IAS and US GAAP. Journal of International Financial Management & Accounting 15 (1):60-91. Tyrrall, D., D. Woodward, and A. Rakhimbekova. 2007. The relevance of International Financial Reporting Standards to a developing country: Evidence from Kazakhstan. International Journal of Accounting 42 (1):82-110. Van der Meulen, S., A. Gaeremynck, and M. Willekens. 2007. Attribute differences between US GAAP and IFRS earnings: An exploratory study. International Journal of Accounting 42 (2):123-142. van Tendeloo, B., and A. Vanstraelen. 2005. Earnings management under German GAAP versus IFRS. European Accounting Review 14 (1):155-180. Weienberger, B. E., A. B. Stahl, and S. Vorstius. 2004. Changing from German GAAP to IFRS or US GAAP: A Survey of German Companies. Accounting in Europe 1:169-189. Wingate, M. L. 1997. An examination of cultural influence on audit environments. Research in Accounting Regulation:129-148. Wstemann, J., and S. Kierzek. 2005. Revenue Recognition under IFRS Revisited: Conceptual Models, Current Proposals and Practical Consequences. Accounting in Europe 2 (1):69 106. . 2006. True and Fair View Revisited A Reply to Alexander and Nobes. European Accounting Review 15 (3):91-116. Zarzeski, M. T. 1996. Spontaneous harmonization effects of culture and market forces on accounting disclosure practices. Accounting Horizons 10:18-37. Zeghal, D., and K. Mhedhbi. 2006. An analysis of the factors affecting the adoption of international accounting standards by developing countries. The International Journal of Accounting 41 (4):373-386.

96

Appendix 1 Composition of Eddies (1990) Accounting Values Indexes


Factors Comprising Professionalism Index Nature of Professional Organization 1. 2. 3. 4. 5. 6. 2. 3. 4. 5. Independence of the accounting profession from government organizational structure Activity in conducting research and setting standards Age of accounting profession Membership employed in the private sector versus government sector Range of services offered by the profession Strength of disciplinary rules and enforcement Extent of legal liability of accountants for negligence or malpractice Extent of continuing professional education requirements Reliance on judgment, true and fair view as opposed to statutory compliance Extent of qualified audit reports

Nature of Membership

Factors Comprising Uniformity Index Accounting Policy 1. 2. 3. 4. 5. Accounting disclosure is regulated by a uniformity plan as opposed to individual corporate selection Accounting measurement and valuation rules are specified by statute Accounting rules for measurement and valuation are required to be the same as taxation rules Stock exchange listing rules require uniform disclosure, measurement and valuation Accounting information is primarily used by government for national planning and macroeconomic purposes Companies change accounting policies over time Companies change measurement and valuation practices over time Companies within the same industry have the same accounting policies, measurement and valuation practices Companies within the same group have the same accounting policies, measurement and valuation practices A change in accounting practices is reported and includes a restatement for prior periods

Nature of Membership 1. 2. 3. 4. 5.

97

Factors Comprising Conservatism Index Valuation of Assets and Liabilities 1. 2. 3. 4. 5. 1. 2. 3. 4. 5. Fixed assets are revalued to amounts above historical cost The net realizable value of marketable securities is reported if above historical cost Goodwill is amortized and not written off immediately The equity method is used to value long-term investments in associated companies Contingent liabilities are reported in the financial statements The realization principle is followed in reporting income Income from abnormal and extraordinary items is included on the income statement and not taken directly to reserves All gains or losses on foreign currency transactions are included in net income The use of the equity method for reporting income from associated companies Tax-effect accounting is used to calculate the taxation expense

Recognition of Income

Factors Comprising Secrecy Index Dissemination of Corporate Information 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Extent of availability of annual reports for public corporations, enterprises and large private companies Extent of additional timely disclosures to securities markets Extent of use of employee reporting Extent of social reporting disclosures Extent of segment (geographic and line of business) disclosures Extent of non-statutory required disclosures Extent of use of audit committees within corporations Extent of full group disclosures (all subsidiaries and associated companies) Extent of disclosure of interests in joint ventures Extent of disclosure of accounting policies

98

Appendix 2 Correlation of GLOBE and Hofstede Cultural Values Indices


1. 1. 1 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

41 .279 1 (0.08) 41 41 .131 -.097 1 3. (0.42) (0.55) 41 41 41 -.229 1 .453 .354 4. (0.00) (0.02) (0.15) 41 41 41 41 .016 .186 .465 .523 5. (0.00) (0.00) (0.92) (0.25) 41 41 41 41 -.092 -.286 .463 -.414 6. (0.57) (0.07) (0.00) (0.01) 41 41 41 41 -.155 .024 -.054 -.040 7. (0.33) (0.88) (0.74) (0.80) 41 41 41 41 .139 .290 .650 .542 8. (0.00) (0.39) (0.07) (0.00) 41 41 41 41 .111 .556 .412 .421 9. (0.00) (0.49) (0.01) (0.01) 41 41 41 41 .016 -.209 -.455 -.513 10. (0.00) (0.92) (0.19) (0.00) 41 41 41 41 .190 .197 .178 .001 11. (0.23) (0.22) (0.26) (1.00) 41 41 41 41 .205 .060 12. -.371 .415 (0.20) (0.71) (0.02) (0.01) 41 41 41 41 .081 -.386 .287 -.193 13. (0.76) (0.13) (0.26) (0.46) 17 17 17 17 GLOBE Cultural Values Variables: 1. Future Orientation 2. Performance Orientation 3. Assertiveness 4. Institutional Collectivism 5. In-Group Collectivism 6. Power Distance 7. Humane Orientation 8. Uncertainty Avoidance 2.

1 41 -.052 1 (0.75) 41 41 -.232 1 -.448 (0.14) (0.00) 41 41 41 .294 .094 -.272 (0.06) (0.56) (0.09) 41 41 41 .092 .081 -.083 (0.57) (0.61) (0.61) 41 41 41 -.098 .059 .145 (0.54) (0.71) (0.37) 41 41 41 -.101 .077 -.005 (0.53) (0.63) (0.97) 41 41 41 .108 -.034 -.452 (0.50) (0.00) (0.83) 41 41 41 -.231 .185 .065 (0.37) (0.48) (0.81) 17 17 17 Hofstede Cultural Variables: 9. Power Distance 10. Individualism 11. Masculinity 12. Uncertainty Avoidance 13. Long Term Orientation

1 41 .707 (0.00) 41 -.731 (0.00) 41 .127 (0.43) 41 .271 (0.09) 41 .155 (0.55) 17

1 41 -.653 (0.00) 41 .131 (0.42) 41 .056 (0.73) 41 .128 (0.63) 17

1 41 .036 (0.82) 41 -.230 (0.15) 41 -.348 (0.17) 17

1 41 .157 (0.33) 41 .054 (0.84) 17

1 41 .169 (0.52) 17

1 17

Statistically significant correlations at the 0.05 level are bolded (2-tailed tests). The following information is reported for each variable per row: Pearson Correlation Sig. (2-tailed) N

99

Vita
Ramon P. Rodriguez, Jr., the first child of Ramon P. Rodriguez, Sr. and Betty Huitron and brother of Kim Esquivel, was raised in Frankfurt, West Germany, San Antonio, Texas, and Augsburg, West Germany. He graduated from Augsburg American High School in Augsburg, West Germany. He spent four years serving in the United States Air Force and six years in the restaurant business performing jobs as varied as disc jockey and restaurant manager. He

received a Bachelors of Business Administration in accounting from The University of Texas at San Antonio. He also earned a Masters of Business Administration from UTSA. In the fall of 2005, he entered the Graduate School at the University of Texas at El Paso in pursuit of a Doctor of Philosophy in International Business. He is a Certified Public Accountant and has worked at two public accounting firms as well as in the corporate sector (oil and gas industry). Additionally, he has taught accounting courses at UTSA, UTEP, and Southern Illinois University Carbondale. His work has been published in the Journal of International Accounting Research.

Permanent address:

3 Morgan Drive Murphysboro, Illinois 62966

This dissertation was typed by the author.

100