You are on page 1of 8

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/15833810

Safety Climate in Industrial Organizations: Theoretical and


Applied Implications

Article  in  Journal of Applied Psychology · March 1980


DOI: 10.1037/0021-9010.65.1.96 · Source: PubMed

CITATIONS READS
2,013 19,912

1 author:

Dov Zohar
Technion - Israel Institute of Technology
58 PUBLICATIONS   10,870 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Systematic review of The effect of Safety interventions on work injuries http://www.campbellcollaboration.org/lib/project/222/


View project

Safety Climate Lone Worker project View project

All content following this page was uploaded by Dov Zohar on 25 March 2014.

The user has requested enhancement of the downloaded file.


Journal of Applied Psychology
1980, Vol. 65, No. 1, 96-102

Safety Climate in Industrial Organizations:


Theoretical and Applied Implications
Dov Zohar
Technion—Israel Institute of Technology, Haifa, Israel

A 40-item measure of organizational climate for safety was constructed and


validated in a stratified sample of 20 industrial organizations in Israel. This
climate reflects employees' perceptions about the relative importance of safe
conduct in their occupational behavior. It can vary from highly positive to a
neutral level, and its average level reflects the safety climate in a given company.
It was shown that there is an agreement among employees' perceptions re-
garding safety climate in their company and that the level of this climate is
correlated with safety program effectiveness as judged by safety inspectors.
The two dimensions of highest importance in determining the level of this
climate were workers' perceptions of management attitudes about safety and
their perceptions regarding the relevance of safety in general production proc-
esses. It is proposed that organizational climate, when operationalized and
validated as demonstrated in this article, can serve as a useful tool in under-
standing occupational behavior.

The purpose of this article is to describe In their review article, James and Jones
a particular type of organizational climate (1974) distinguished between measures of
and to examine its implications. This climate organizational climate that are based on (a)
is a climate for safety in industrial organiza- structural properties of organizations such
tions. Writers of organizational climate dis- as size, structure, systems complexity, lead-
tinguish between holistic climate measures, ership style, and goal directions (e.g., Fore-
such as House and Rizzo's (1972) scale, and hand & Gilmer, 1964; Porter & Lawler, 1965)
specific climate measures. Examples for such and (b) perceptions held by employees about
specific climates are Litwin and Stringer's aspects of their organizational environment,
(1968) motivation climate, Schneider and summarized over individual employees (e.g.,
Bartlett's (1970) individual differences cli- Schneider, 1973; Sells, 1968; Tagiuri, 1968).
mate, or Taylor's (1972) creativity climate. In the present article we adopted this second
Obviously, then, any given organization interpretation of organizational climate.
creates a number of different climates, and Namely, climate was viewed as a summary
the term organizational climate has to be of molar perceptions that employees share
supplemented by an appropriate adjective about their work environments. Following
indicating which type of climate it is. To
follow Schneider's (1975) proposal, the term Schneider (1975), it is assumed that these
organizational climate should describe an perceptions have a psychological utility in
area of research rather than a specific or- serving as a frame of reference for guiding
ganizational measure. It is in this spirit that appropriate and adaptive task behaviors.
the concept of safety climate was developed. Based on a variety of cues present in their
work environment, employees develop co-
This study was supported by a grant from the Com- herent sets of perceptions and expectations
mittee for Preventive Action, Ministry of Labor, Israel. regarding behavior-outcome contingencies
The author wishes to thank Ezey Dar-El for his con- and behave accordingly (Dieterly & Schnei-
tinued encouragement through all phases of this work. der, 1974; Fleishman, 1953; Litwin &
Requests for reprints should be sent to Dov Zohar,
Faculty of Industrial and Management Engineering, Stringer, 1968). These coherent sets of or-
Technion, Haifa, Israel. ganizational perceptions, when shared and
Copyright 1980 by the American Psychological Association, Inc. 0021-9010/80/6501-00%$00.75

96
SAFETY CLIMATE 97

summarized for individual employees, are & Stahl, 1964; Planek et al., 1967). A third
defined here as organizational climates. characteristic was the existence of open
communication links and frequent contacts
between workers and management (Accident
Determining Safety Climate Dimensions
Prevention Advisory Unit in U.K., 1976;
To determine the various dimensions of Cohen et al., 1975). Another expression of
safety climate, a review of safety literature this free flow of information was found to
was undertaken. The purpose of this review be the carrying out of frequent safety in-
was to define organizational characteristics spections by appropriate personnel (Davis
that differentiate between high versus low & Stahl, 1964; Planek et al., 1967). General
accident-rate companies. It was assumed environmental control and good housekeep-
that such organizational features characterize ing was the fourth characteristic appearing
individual plants and the global perception consistently. Orderly plant operations, con-
of these by production workers, hence, form trolled environmental conditions, and high
the safety climate of that factory. usage of safety devices comprised this or-
One of the most consistent findings in the ganizational characteristic in low-accident
reviewed literature was that in factories hav- companies (Shafai-Sahrai, 1971; Smith,
ing successful safety programs, there was a Cohen, Cohen, & Cleveland, 1975).
strong management commitment to safety. A fifth characteristic was a stable work
This commitment was exhibited in a variety force with less turnover and older workers
of ways. Cohen, Smith, and Cohen (1975), (Cleveland et al., 1978; Cohen et al., 1975;
Mobley (Note 1), and Shafai-Sahrai (1971) Davis & Stahl, 1964). Although not specifi-
have all found that in low-accident com- cally studied, this factor probably reflected
panies, top management was personally in- better industrial relations and elaborate
volved in safety activities on a routine basis, personnel development practices in these
whereas such commitment was conspicu- factories. Finally, successful companies
ously absent in high-accident companies. had distinctive ways of promoting safety.
Cleveland, Cohen, Smith, and Cohen (1978) These included guidance and counseling,
and Shafai-Sahrai (1971) have reported that rather than enforcement and admonition. In
in low-accident companies safety matters addition, it included individual praise or
were given high priority in company meet- recognition for safe performance and enlist-
ings and production scheduling, based on ing workers' families in safety promotions
the conviction that safety is an integral part (Cleveland et al., 1978; Davis & Stahl, 1964;
of production systems and accidents are ac- National Safety Council, 1969).
tually symptoms of design faults in that When all these organizational charac-
system. teristics are integrated, it is possible to form
Another expression of management com- a coherent organizational pattern of a highly
mitment found to discriminate between safe company: Management is actively in-
companies was the rank and status of safety volved in safety management and creates a
officers; in the better companies they had general administrative control climate (Gri-
a higher status. This finding was reported maldi, 1970) in which work is to be performed.
by the Accident Prevention Advisory Unit This climate results in increased perform-
in the United Kingdom (1976), Cohen et al. ance reliability of workers, good housekeep-
(1975), Davis and Stahl (1964), and Planek, ing, and high design and maintenance stan-
Driessen, and Vilardo (1967). A second highly dards for work environments. There are
consistent organizational characteristic dis- well-developed personnel-selection training
criminating between companies was empha- and development programs in which safe
sis put on safety training. In better com- conduct is an integral part. Communication
panies it was designed as an integral part links between workers and management are
of new workers' training (Cohen et al., 1975; kept open, enabling a flow of information
National Safety Council, 1969; Mobley, regarding production as well as safety mat-
Note 1) or as a follow-up and periodic re- ters. Finally, general management philosophy
training for workers and supervisors (Davis is not strictly production oriented but also
98 DOV ZOHAR

people oriented, as evidenced by various Method


supportive policies described above. All the
organizational characteristics described Questionnaire Development
above were corroborated in a comprehen- Based on the industrial safety literature described
sive review article published by Cohen above, seven organizational dimensions were included
in the initial version of the safety climate questionnaire.
(1977). (Although his article was published Each of these climate dimensions was represented by
after the present study was completed, it seven items that were short statements with 5-point
was used in preparing this article.) scales for evaluating subjects' agreement with them.
Based on the reviewed literature, it was All items were phrased positively so that full agreement
decided that the safety climate questionnaire resulted in a higher score in this dimension. This pro-
cedure resulted in a questionnaire of 49 items. The
would include the following dimensions: (a) questionnaire was then given to a pilot sample of 120
perceived management attitudes towards production workers in four factories. Workers were
safety, (b) perceived effects of safe conduct interviewed by a team of three interviewers who read
on promotion, (c) perceived effects of safe each item aloud and recorded subjects' agreement to
it on the 5-point scale. These data were then factor
conduct on social status, (d) perceived or- analyzed using a principal-components factor analysis
ganizational status of safety officer, (e) per- with varimax rotation.
ceived importance and effectiveness of This procedure resulted in eight factors that largely
safety training, (f) perceived risk level at overlapped the original ones, thus confirming the valid-
work place, and (g) perceived effectiveness ity of the theoretical considerations for developing
these questionnaire items. Nine items found to be un-
of enforcement versus guidance in promot- related to any specific factor were pulled out, resulting
ing safety. in a 40-item questionnaire. Items assigned to each fac-
These dimensions, therefore, included tor had a loading greater than .49 on that factor. Table 1
those organizational characteristics found to lists these factors with their respective eigenvalues and
the number of questionnaire items representing them.
discriminate between high versus low ac- It should be noted that Factor 8 was retained, despite
cident-rate companies. the fact that its eigenvalue was less than 1, which is
Based on the literature of organizational the lowest recommended value for factor retention
climate and the literature of organizational (Guttman, 1954). This was done because in a dis-
criminant analysis that will be referred to below (Table
safety practices, two hypotheses were for- 4), it proved to be of a high discriminant value.
mulated: (a) Workers in different companies
share common perceptions regarding safety
in their organization. The sum of these per- Questionnaire Administration
ceptions is the safety climate in each or- Twenty factories were selected for questionnaire ad-
ganization, (b) Safety climate can vary from ministration. Factory selection was done in a quasi-
a less favorable to a more favorable one. random manner. Using a national listing of large in-
dustrial organizations (i.e., those having more than 500
Its level in each company is expected to workers), 5 factories were randomly chosen from each
be correlated with that company's safety of four production categories: metal fabrication, food
record. processing, chemical industry, and textile industry. Out

Table 1
Principal-Components Factor Analysis of the Safety Climate Questionnaire
No. of
%of questionnaire
Factor Eigenvalue variance items

Perceived importance of safety training programs 9.84 40.9 6


Perceived management attitudes toward safety 4.63 19.3 9
Perceived effects of safe conduct on promotion 2.53 10.6 7
Perceived level of risk at work place 2.34 9.7 5
Perceived effects of required work pace on safety 1.66 6.9 3
Perceived status of safety officer 1.17 4.8 5
Perceived effects of safe conduct on social status 1.07 4.4 2
Perceived status of safety committee .84 3.4 3
SAFETY CLIMATE 99

of the 20 selected factories, 4 declined to cooperate Table 2


and were, therefore, substituted by others selected in Multiple-Range Test of Factories Based on
the same manner. All factories had a worker popula- Respective Safety Climate Scores
tion of 500-1,000 workers and exhibited a wide range
of technologies and safety records. Questionnaire ad- Production categories
ministration was limited to production workers only, M n of
and in each plant a stratified random sample of 20 work- Group score plants Chem Met Text Food
ers was selected. Sample stratification was based on the
relative size of the various production departments in 1 186.6 5 4 1
the factory, resulting in a random sample in which the 2 153.3 4 1 1 1 1
various departments were proportionally represented. 3 141.1 4 2 1 I
Workers were interviewed by one member of the three- 4 120.8 7 1 2 4
member interviewing team. During the interview, ques-
tionnaire items were read aloud and interpreted if Note. Chem = chemical, Met = metal, Text = textile.
necessary. Workers' responses were then recorded on
a 5-point scale for each item.
plants in each group. One evident charac-
Questionnaire Interpretation teristic of this table is that chemical plants
To obtain safety climate scores, responses for each have the highest safety climate scores,
item in the questionnaire were given values from 1 to 5. whereas food processing plants have the
The value 5 was given for high agreement with a state- lowest scores. Metal processing and textile
ment, and a value of 1 was given for disagreement factories fall in between. These data could
with a statement. Each questionnaire could thus be
assigned a single score indicating the safety climate be expected based on the technologies and
level for that individual worker. This score was the risk levels involved. Chemical plants have
sum of values for all items in the questionnaire. The the highest risks in their production proc-
safety climate level for a given factory was determined esses, followed by metal fabrication and
by the average score of all 20 workers interviewed in
that plant. This procedure of representing the climate
textile factories. It is interesting, therefore,
by a single score was based on theoretical considera- although expected based on our second hy-
tions whereby all climate dimensions described condi- pothesis, that the resulting safety program
tions and procedures affecting safety programs' effec- practices were reflected in the safety climate
tiveness. Since these wer,e considered additive in nature, levels of these companies.
a high score would indicate more favorable conditions
and procedures. An attempt to test the second hypothesis
directly by correlating safety climate scores
Results with standard safety measures such as ac-
cident-frequency rate and accident-severity
To test the hypothesis stating that work- rate was terminated due to the apparent lack
ers' perceptions of their work place safety of reliability of these measures. This lack of
were relatively homogeneous, the variance reliability resulted from the fact that these
of safety climate scores within factories was measures were based on reports used for
compared with the variance between fac- workers' compensation purposes. Because
tories using a one-way analysis of variance. of different insurance policies and a system
The resulting F ratio was highly significant, of penalties, some factories had highly in-
F(19, 380) = 52.4,p < .001, hence support- flated figures, whereas others had a bias in
ing the notion of a definable safety climate the opposite direction. As an alternative,
in industrial organizations. therefore, four experienced safety inspectors
Using a multiple-range test (Nie, Hull, working at Israel's Institute of Safety and
Jenkins, Steinbrenner, & Bent, 1975, pp. Hygiene were asked to rank order the se-
427-428), the 20 plants were divided into lected factories according to safety practices
four groups based on differences between and accident-prevention programs. Ranking
their respective group mean climate scores. was done separately in each production
Each group included all factory pairs having category, since judges otherwise had great
scores that differed by a range smaller than difficulty in comparing various factories.
the shortest significant range for the .05 level The textile factories and three others were
of significance. Table 2 lists group means not rank ordered because of judges' insuf-
and respective production categories of ficient familiarity with their functioning.
100 DOV ZOHAR

Table 3
Ranking of Factories Within Each Production Category
Metal Chemical Food

Judge Judge Judge

Rank A B C D score A B C D score A B C D score

1 a a a a 156.3 f f f f 156.9 j J 1 j 135.8


2 b b b b 116.2 g g g g 158.8 k k k k 103.2
3 c c c c 124.4 h h h h 155.3 1 1 J 1 108.5
4 d d d d 115.4 i i i : 148.1
5 e e e e 110.4

Note. When different factories share the same rank, the climate score is the one given to the modal factory
in that rank. Factories are represented by lowercase letters.

Due to their geographical dispersion, it was pair of ranks. When fewer factories were
impossible to get the inspectors to visit these ranked, there was a stronger effect of low-
factories, and they had to be omitted. Re- ering the resulting correlation coefficient.
sults of this procedure are given in Table 3. This relatively high agreement between
The agreement between judges' ranking judges' ranking and safety-climate scores
and respective safety climate scores of fac- therefore supported the second hypothesis
tories was tested using Spearman rank cor- and the validity of the safety climate ques-
relation coefficients in each production tionnaire. Agreement among judges was
category. These were rs = .90 (metal), rs also high, as can be seen in Table 3. In
= .80 (chemical), and rs = .50 (food). These the metal category there was complete
correlations are based on small ns ranging agreement between judges, whereas in the
between three and five, hence they should chemical and food categories, disagreements
be interpreted cautiously. Inspection of resulted from interchanges between single
Table 3 indicates that all disagreements pairs of factories, resulting in high overall
between judges' ranking and safety-climate agreement.
ranks resulted from an interchange in a single Finally, a stepwise discriminant analysis
was used to find the smallest number of cli-
Table 4 mate dimensions that are sufficient to dis-
Stepwise Discriminant Analysis of the criminate between different factories based
Safety Climate Questionnaire on their safety climate levels. Results of
this analysis are given in Table 4. Based on
Fto Wilks' lambda criterion (Nie et al., 1975),
enter or Wilks'
climate dimensions accounted for most of
Climate dimension remove lambda
the separation between factories. These di-
Perceived importance of mensions, listed in decreasing discriminant
safety training 141.12 .12 power, are (a) perceived importance of safety
Perceived effects of required training programs, (b) perceived effects of
work pace on safety 81.44 .02
Perceived status of safety required work pace on safety, (c) perceived
committee 58.92 .006 status of safety committee, and (d) perceived
Perceived status of safety status of safety officer.
officer 48.81 .002 The data in Table 4 can be interpreted
Perceived effects of safe as indicating that two climate dimensions
conduct on promotion 16.85 .0009
Perceived level of risk at are most influential in determining safety
work place 7.96 .0006 climate levels. The first dimension is the
Perceived management perceived relevance of safety to job behavior.
attitudes toward safety 6.74 .0005 This relevance is reflected by regarding
Perceived effect of safe
conduct on social status 1.88 .0004 safety training as an important prerequisite
for successful performance and by viewing
SAFETY CLIMATE 101

higher work pace as potentially hazardous. mental control, maintenance and house-
The second climate dimension is the per- keeping, worker selection and training,
ceived management attitude toward safety, information flow, and so forth. Indeed, ac-
which can be readily exhibited in workers' cident analyses in most companies reveal
eyes by the organizational status of both these relationships and the literature re-
the safety officer and safety committee. The viewed points in this direction.
status of the safety committee can be as- The concept of safety climate implies that
sessed by the level of those managers who production workers indeed have a unified
actively participate in it and by the actual set of cognitions regarding the safety aspects
implementation of its decisions, whereas the of their organization. As proposed by
status of the safety officer can be assessed Schneider (1975), such perceptions and be-
by executive authority relegated to him (e.g., havior-outcome expectations can guide and
authority to remove workers from produc- direct job behaviors accordingly. Further-
tion hall or to stop production processes more, these cognitions are largely related
when safety regulations are not followed). to perceptions of management attitudes
To summarize, then, the data in this study about safety and its relevance to general
supported both hypotheses, namely: (a) production processes. It can therefore be
Safety climate can be regarded as a char- concluded that a genuine change in manage-
acteristic of industrial organizations, and ment attitudes and increased commitment
(b) safety climate is related to the general are prerequisites for any successful attempt
safety level in these organizations. at improving the safety level in industrial
organizations. Attempting to improve safety
Discussion levels, as we often see, with new safety reg-
ulations, poster campaigns, and depart-
Organizational safety climate, as defined mental safety contests without first securing
in this article, has both theoretical and ap- sincere management commitment might be
plied significance. The main implication is missing the forest for the trees. This is the
that management commitment to safety, basic applied implication of the concept of
with its multitude of expressions, is a major safety climate in industrial organizations.
factor affecting the success of safety pro- Finally, the apparent difficulties the safety
grams in industry. Such expressions might inspectors had in comparing factories from
be the establishment of job-training pro- different production categories points toward
grams, relegation of executive authority to the potential use of safety-climate scores
safety officials, participation of high-level as a common denominator for comparing
managers in safety committees, and taking different factories. A major source of dif-
safety into consideration in job design, in- ficulty has been the different technologies
cluding required work pace. Often, manage- and risk levels involved in the various fac-
ment views safety as a technical and inde- tories. Using a measure such as the one pro-
pendent aspect of the production process, posed here, which is independent of these
detached from other management operations. factors, can therefore enable such com-
Yet, not willing to ignore its responsibility parisons when studying industrial safety
in this regard (and complying with govern- programs.
ment regulations), management assigns all
responsibility to specified safety personnel Reference Note
without relegating to them any executive 1. Mobley, W. H. Managerial evaluations of safety
power. The view proposed by this author motivation and behavioral hypothesis (Contract
as well as others (Cohen, 1977; Grimaldi, Rep. CDC-74-1815). Cincinnati, Ohio: National In-
1970) is quite to the contrary. Safety should stitute for Occupational Safety and Health, 1974.
be regarded as an integral part of the pro- References
duction system closely related to the overall
degree of control management has over pro- Accident Prevention Advisory Unit. Success and fail-
ure in accident prevention. London: Her Majesty's
duction processes (Grimaldi, 1970). Hence, Stationery Office, 1976.
it should not be detached from general man- Cleveland, R. J., Cohen, H. H., Smith, M. J., & Cohen,
agement responsibilities such as environ- A. Safety program practices in recordholding plants.
102 DOV ZOHAR

Cincinnati, Ohio: National Institute for Occupational Nie, N. H., Hull, C. H., Jenkins, J. G., Steinbrenner,
Safety and Health, 1978. K., & Bent, D. H. Statistical package for the social
Cohen, A. Factors in successful occupational safety sciences (2nd ed.). New York: McGraw-Hill, 1975.
programs. Journal of Safety Research, 1977,9, 168- Planek, T., Driessen, G., & Vilardo, F. J. Evaluating
178. the elements of an industrial safety program. Na-
Cohen, A., Smith, M., & Cohen, H. H. Safety pro- tional Safety News, August 1967, 60-63.
gram practices in high vs. low accident rate com- Porter, L. W., & Lawler, E. E., III. Properties of
panies—An interim report (U.S. Department of organizational structure in relation to job attitudes
Health, Education and Welfare Publication No. 75- and job behavior. Psychological Bulletin, 1965, 64,
185). Cincinnati, Ohio: National Institute for Occu- 23-51.
pational Safety and Health, 1975. Schneider, B. The perception of organizational climate:
Davis, R. T., & Stahl, R. W. Safety organization and The customer's view. Journal of Applied Psychol-
activities of award-winning companies in the coal ogy , 1973,57,248-256.
mining industry (Information Circular No. 8224.) Schneider, B. Organizational climates: An essay.
Pittsburgh, Pa.: U.S. Department of Interior, Bureau Personnel Psychology, 1975, 28, 447-479.
of Mines, 1964. Schneider, B., & Bartlett, C. J. Individual differences
Dieterly, D., & Schneider, B. The effect of organiza- and organizational climate. II: Measurement of or-
tional environment on perceived power and climate: ganizational climate by the multitrait-multirater
A laboratory study. Organizational Behavior and matrix. Personnel Psychology, 1970, 23, 493-512.
Human Performance, 1974, / / , 316-337. Sells, S. B. An approach to the nature of organiza-
Fleishman, E. E. Leadership climate, human relations tional climate. In R. Tagiuri & G. Litwin (Eds.),
training, and supervisory behavior. Personnel Psy- Organizational climate: Explorations of a concept.
chology, 1953,6,205-222. Boston: Division of Research, Harvard Business
Forehand, G. A., & Oilmer, B. H. Environmental vari- School Press, 1968.
ation in studies of organizational behavior. Psy- Shafai-Sahrai, Y. An inquiry into factors that might
chological Bulletin, 1964, 62, 361-382. explain differences in occupational accident experi-
Grimaldi, J. V. The measurement of safety engineering ence of similar size firms in the same industry (Tech.
performance. Journal of Safety Research, 1970, 2, rep.). East Lansing, Mich.: Division of Research,
137-159. Graduate School of Business Administration, Michi-
Guttman, L. Some necessary conditions for common gan State University Press, 1971.
factor analysis. Psychometrika, 1954, 19, 149-161. Smith, M. J., Cohen, H. H., Cohen, A., & Cleveland,
House, R. J., and Rizzo, J. R. Toward the measure of R. J. On-site observations of safety practices in plants
organizational practices: Scale development and with differential safety performance. National Safety
validation. Journal of Applied Psychology, 1972,56, Congress Transactions (Vol. 12). Chicago: National
388-396. Safety Council, 1975.
James, L. R., & Jones, A. P. Organizational climate: Tagiuri, R. The concepts of organizational climate. In
A review of theory and research. Psychological R. Tagiuri & G. Litwin (Eds.), Organizational cli-
Bulletin, 1974, 81, 1096-1112. mate: Explorations of a concept. Boston: Division
Litwin, G. H., & Stringer, R. A., Jr. Motivation and
organizational climate. Boston: Division of Re- of Research, Harvard Business School Press, 1968.
Taylor, C. W. (Ed.). Climate for creativity. Elmsford,
search, Harvard Business School, 1968.
National Safety Council. Award-winning programs. In, N.Y.: Pergamon Press, 1972.
Accident prevention manual for industrial operations
(6th ed.). Chicago: Author, 1969. Received January 8, 1979 •

View publication stats

You might also like