Académique Documents
Professionnel Documents
Culture Documents
This study analyzes the determinants of industrial accidents across 44 four-digit manufacturing
industries in Malaysia from 1993 to 2008 through the business cycle and structural approaches. The
results of pooled ordinary least square and fixed-effects estimations revealed that industrial accidents
in Malaysian manufacturing sector were negatively influenced by firm size and positively influenced by
business cycle. Consistent with the findings of previous studies in other countries, the empirical
evidence of this study supports the pro-cyclical behavior of injury rates in manufacturing industries
towards business cycle. The analysis demonstrates that both structural and cyclical variation effects
are important determinants of industrial accidents in Malaysia.
Key words: Industrial accident, cyclical variation, structural characteristics, occupational safety and health
act.
INTRODUCTION
Studies on the incidence of industrial accidents or injuries through tax and compensation benefit is inconclusive.
can be grouped into three approaches, namely: business The third approach focused on structural effect of the
cycle, labor market and structural approaches. The industry such as workers characteristics, firm size, and
business cycle approach to workplace injuries provides mechanization or capital intensity in the production
explanations as to how injury rates may be expected to process (Oi, 1974; Cooke and Gautschi, 1981; Viscusi,
vary over the course of the economic cycle (Kossoris, 1986; Currington, 1986). Demographic differentials in
1938; Leigh 1985; Robinson and Shor, 1989). These work injury rates could have been generated by several
studies support pro-cyclical relation, showing that the structural forces. If other things being equal, (such as
number of accident tends to increase during economic type of industry, occupation, firm size and safety of the
upswings and vice versa. Nevertheless, this approach work site), certain workers are innately more liable to be
does not explicitly consider the interaction of choices involved in industrial accidents (Oi, 1974). Despite the
made by employers between safety and profits and the three approaches, it is often assumed that the causes of
choices of employees between safety and wages in accidents vary across sectors (Coleman, 1981). A survey
determining the risk of an accident. This give rise to the report by Centers for Disease Control and Prevention
second approach, the market oriented approach to (1993) on fatal injuries in the United States (U.S.) during
industrial accident as proposed by Chelius (1974), Oi 1980 to 1989 shows that the largest number of fatalities
(1974), and Sider (1985). Their studies relate the injury occurred in the construction sector, followed by tran-
rate to the market factors, such as wage and government sportation, manufacturing, and primary economic sectors.
control. In general, their analysis shows that, under firms A large body of existing empirical analysis on industrial
optimization, occupational accident is determined by accidents focused on manufacturing and construction
wage level and employers incentives in accident pre- sectors. This is due to their natural hazard and both
vention. However, the impact of government intervention sectors are found to be highly responsive to the business
cycle, particularly in mature capitalist economies as
well as those in transition towards industrialized
economies
*Corresponding author. E-mail: (Robinson and Shor, 1989; Davies et al., 2009).
fatimahs@um.edu.my. With the vision of becoming an industrialized economy
2082 Afr. J. Bus. Manage.
0820
Said et al. 2092
0920
by the year 2020, higher risk of accident as factors that influence The remainder of this
Malaysia has started its they are not accustomed industrial accidents in paper is structured as
Industria- lization effort to the hazard of Malaysian manufacturing follows: an overview of
since 1960s. workplace environment. sector. Hence, the industrial accidents in
Industrialization has been Therefore, a study of objective of our study is Malaysia and followed by
an integral part in the workplace injuries in to examine the literature review.
Malaysian development Malaysian manufacturing determinants of Subsequently, this study
strategies and sector is particularly industrial accidents in the discusses the model
manufacturing sector relevant since it would Malaysian manufacturing specification and data,
has shown to be one contribute to a greater sector during which is followed by
of the important understanding of 1993 to 2008. We specify results and discussion.
backbones and a major factors that determine our empirical model Finally, this study
contributor to the industrial accidents in the based on two concludes and offers
Malaysian economy. The sector. approaches, namely: some policy implications.
share of manufacturing Industrial accidents the business cycle
sector to Gross Domestic have been the subject of and structural
Product (GDP) increased growing number of approaches. Difficulties OVERVIEW OF
significantly from only academic research since of obtaining data on INDUSTRIAL
12.2% in 1970 to 30.1% the last three decades. wage premium and on ACCIDENT IN
in 2010. Apparently, this However, large body of employees protection MALAYSIA
sector has been the research focusing on the measures for each
major sector in creating causes of industrial industry prevent us from Tables 1 and 2,
employ- ment accidents is dominated incorporating the labor respectively present the
opportunities. In 1970, by empirical studies in market oriented approach number of industrial
employment in the industrialized countries, in our model. accidents by sectors and
manufacturing sector such as European by types of accident in
represented only 9.4% of countries and the U.S. In Malaysia during 1994 to
total employment Malaysia, existing studies 2008. There was
(Malaysia, 1976). In line on industrial accident significant decline in
with the industrialization were mainly focused on the total number of
process, the share of the issues of the industrial accidents
employment in the establishment and reported for all sectors,
manufacturing sector enforcement of the a decrease of 55.30%
increased over the occupational safety and from
years. As at health act (OSHA) and 125,506 in 1994 to
2010, the share of the evolution of safety 56,095 in 2008. Among
employment in the related regulations all sectors, the number of
sector has (Jamaluddin, 1994; accidents reported for the
incre Rahmah and Sum, 2000; manufacturing sector has
ased Mansur et al., 2003; been the highest
to Ariffin et al., 2006; throughout the period.
27.8 Rampal and Nizam, 2006; This reflects workers in
% Lugah et al., 2010; the manufacturing sector
Surienty et al., 2011). are exposed to higher
(Mala
Empirical study on accidental risks.
ysia,
industrial accidents in It can be observed
2010)
Malaysia, however are that the pattern of
.
still lacking and mostly accidents reported varies
It is often the case that
concentrated on the from one sector to
rapid expansion of
construction sector another, reflecting the
manufac- turing
(AbdulHamid et al., 2008; difference of hazard
industries during
Ali et al., 2010; Zakaria et across sectors. As shown
economic expansion is
al., 2010). Apart from in Table 2,
associated with large
20.60% of total fatal
employment of new these studies, Mansor et
accidents and 37.91% of
workers and new al. (2011) examine the
total disablement
technologies, influence of individual
accidents in 2008
machineries and factors and nature of job
involved workers from the
equipments. While the on accident among
manufacturing sector.
application of new workers at port sites.
Although there has been
technologies would However, to the best of
significant reduction of
expose new hazards to our knowledge, no
total accident in the
the workers, hiring new attempt has been made
manufacturing sector,
worker might as well pose to specifically investigate accident cases which
2102 Afr. J. Bus. Manage.
1021
caused fatality and meant hiring new
disablement shown an workers who are not
increment. Between 1998 accustomed to the
and 2008, fatal accidents hazards of their new
increased from 256 to jobs and hence
268 cases, while increased the number of
disablement increased accidents reported. After
from 5,823 to 9,701 2000, total accidents
cases.
Figure 1 illustrates the
trend of industrial
accidents in the
manufacturing sector
reported from 1993 to
2008. Overall, total
industrial accidents in
the manufacturing sector
were on a declining
trend, except from 1998
to
2000 which shows an
upward trend. An upward
trend of industrial
accidents during this
period was attributable to
Malaysian economic
recovery from the Asian
financial crisis which hit
Malaysia in the middle
of 1997. The upward
and downward trend in
total accidents during
economic crisis and its
recovery partly explain
the influence of business
cycle over industrial
accidents. During
economic crisis in 1997,
firms tended to reduce
both the volume and cost
of production in response
to decrease in aggregate
demand. Reducing
production involves the
lay-off of newly hired,
less experienced and
unskilled workers who
are normally more
vulnerable to accident at
the workplace. Hence by
running the plants with
the experienced and
skilled workers during
economic recession helps
to reduce the number of
accident cases reported.
As the economy began
to recover in 1998, there
was an increase in
employment due to
increase in production.
Increase in employment
during economic upswing
Said et al. 2112
1121
i
C
n
(1) a
t
p
e
where Yit is the dependent i
r
variable, i is entity, t is time, t
c
i (i = 1 . . . a
e
.n) is the n entity-specific l
p
intercept, and i is the t
coefficient for independent i s
variable, Xit and it is the n =
error term. Based on the t Coeffici
general fixed-effects model, e ent for
we rewrite Equation 1 into the n each
following s indepe
s i ndent
p t variable
e y ;
c PW =
i Percentage of =
f production
i workers in the
E
c industry
r
a F=
r
t Percentag
o
i e of
r
o female
n workers in
the t
:
industry e
C r
V m
(2) Where: = = Industry
I =
R C
y Y
= c
e
l
a
i
T c r
h a
e l
i v
n a
j r
Table 3. Descriptive statistics.
Variables IR S KI PW F CV
Mean 3.90 134.21 120.63 57.86 21.33 5.60
Median 1.99 75.80 73.76 65.06 17.40 3.06
Maximum 54.86 1399.63 1367.41 92.17 75.20 136.61
Minimum 0.00 1.51 1.76 0.13 0.00 -83.87
Std. Dev. 5.50 202.49 174.46 21.26 18.20 23.65
The injury rate, as a proxy for is the fixed effects model. which is Model 1. As fixed effects method. One
industrial accidents, is Table 3 presents a shown in Model 1, firm possible explanation for
measured by the percentage summary of the descriptive
statistics of the variables size, capital intensity the difference is that the
of accidents reported per
worker employed. Firm size used in this study. During and cyclical variation variation among
is measured by employees 1993 to 2008, the average are the significant industries in terms of
per establishment and injury rate among the sample variables under pooled capital intensity is not an
capital intensity is measured manufacturing industries was OLS. R-squared value for important factor that
by the value of fixed assets per approximately 3.90% ranging causes the injury rates to
from a minimum of 0 to a
pooled regression is
worker where these two
maximum of lower than that of Models vary across the sample
independent variables take
the natural logarithm form. 54.86%. The lowest and the 2 and 3 whereby only industry.
Production workers and highest injury rate came from 25% of variation in Production workers
female workers are tanneries dependent variable can (KI) and female workers
respectively measured as a and leather finishing be explained by the (F) are found to have
percentage of total industries and metal and positive influence over
wood working machinery explanatory variables
employment. The cyclical
variation variable is measured manufacturing, respectively. included in the model. injury rates (IR) and only
by the percentage change of The average for firm size and Models 2 and 3 are the significant under fixed-
total employment in the capital intensity was 134.21 fixed-effects estimation effects estimation. Our
manufacturing industries. and 120.63%, respectively. results. Under this finding with respect to
Most studies on Based on the cut-off point of production workers is
specification, we treat
occupational accidents 120, on average, the
and business cycle Model 2 as the reference consistent with Viscusi
Malaysian manufac- turing
support the existence of pro- industries can be classified model. In Model 3, (1986). The result is
cyclical relationship where the as capital intensive or production workers (PW) justifiable as production
number of accidents tends heavy industry (UNIDO, 1985; is excluded to isolate the workers are those who
to increase during economic Fatimah and Saad, 2004). are directly involve in
possible influence of this
upswings and reduce during The mean for total production
variable over female firms operation and
economic recession (Kossoris, worker and female worker
1938; Leigh, 1985; Robinson were 57.86 and 21.33%, workers (F) resulting from having direct contact with
and Shor, 1989). Thus, we respectively. The average our measurement machineries and
expect a positive relationship cyclical variation was 5.60% method. equipments. Hence,
between injury rates and throughout the period of The results of this study increase in the fraction of
cyclical variation. Similarly, study. production workers in
capital inten- sity, production
reveal a strong negative
worker, and female worker relationship between firm manufacturing plants
are expected to have positive size (S) and injury rate would increase the injury
influence over the injury rates. R (IR) as the sign of the rates. In contrast, Viscusi
As for firm size, a negative E coefficient and its level of (1986) found a negative
relationship with injury rate is S significance are rela- tionship between
expected in the sense that U consistent under the female workers and
larger firms are better in L
controlling accidents among three models. This finding industrial accidents where
workers as compared to
T is consistent with the he expected that higher
smaller firms (Cooke and S theory (Oi, 1974; Smith, fraction of female workers
Gautschi, 1981; McVittie et al., 1979) and supports the involve less physical
1997) A effort and pose lower risk.
Three types of industrial
empirical findings of
accidents are included in
N previous studies (Cooke A positive sign of
the study, viz. D and Gautschi, 1981; female workers in this
fatal accidents, permanent McVittie et al., 1997). It study leads us to draw a
disability and t emporary D turns out that the larger number of inferences. A
disability. Data on industrial common explanation is
accidents were obtained from
I the firm, the lower the
annual report published by S injury rates. This could
social security organization C be attributed to a
(SOCSO). The annual survey U proper safety
of manufacturing industry, S precaution practiced by
published by the Department
of Statistics, provides data on
S larger firm or adoption of
total employees, fixed assets I safety machinery and
and total establish- ments for O equipments.
each industry. Unpublished N The coefficient for
data of production workers in capital intensity (KI) is
manufacturing plants were
The regression results for positive and highly
obtained from the
Department of Statistics. W e three different model significant under pooled
precede the analysis of this speci- fications are OLS (Model 1) and
study by estimating our reported in Table 4. The consistent with the
balanced panel data using second column of the findings of Currington
pooled ordinary least square (1986). However, this
table shows the results
(OLS) method and compare
from pooled OLS variable appears to be
the results with those under
Equation 2 specification which estimates, insignificant under the
Table 4. Determinants of industrial accidents in Malaysian manufacturing industries.