Journal of Applied Psychology

1974, Vol. 59, No. 6, 717-721

EFFECTS OF THE SHORTER WORKWEEK ON SELECTED

SATISFACTION AND PERFORMANCE MEASURES
JOHN M. IVANCEVICH 1
Department of Business Administration, University of Kentucky

Effects of the 4-day, 40-hour workweek were examined by comparing experi-

mental (n = 104) and control (n — 106) subjects on measures of job satisfac-
tion, anxiety-stress, and performance over a 13-month period. The analysis
indicated that the workers in the 4-day, 40-hour division were more satisfied
with personal worth, social affiliation, job security, and pay; experienced less
anxiety-stress; and performed better with regard to productivity than their
control group (S-day, 40-hour) counterparts.

One of the more recent attempts to improve variables as satisfaction, anxiety-stress, ab-
the job satisfaction and performance of em- senteeism, and performance. The present arti-
ployees is the 4-day, 40-hour (4-40) work- cle describes such an experiment.
week (see Glickman & Brown, 1973). Basi- The specific hypotheses tested were that
cally, the shortened workweek involves more when 4-day, 40-hour (4-40) and S-day, 40-
hours of work per day for fewer days per hour (5-40) work groups are compared, the
week. It is estimated that between 60 and 70 4-40 subjects show significant improvements
companies per month are converting to the in (a) self-actualization, (b) autonomy, (c)
4-day workweek (Wheeler, Gurman, & Tarno- personal worth, (d) social affiliation, (e) job
wieski, 1972). The current estimate of the security, and (/) pay satisfaction (Hypothe-
number of firms using the shorter workweek sis 1); significant reduction in anxiety-
ranges from 700 to 1,000 (Chard & Caruth, stress (Hypothesis 2 ) ; significantly less
1973). unexcused absenteeism (Hypothesis 3); sig-
The literature suggests that firms are nificant improvements in performance ratings
adopting the 4-40 workweek because of po- (Hypothesis 4); and significant changes in
tential improvements in production and mo- the relationships of the job satisfaction facets,
rale, reduction in absenteeism rates (Steward anxiety-stress, and absenteeism with overall
& Larsen, 1972), and the convenience that it performance (Hypothesis 5).
provides employees by providing an extra day
METHOD
off (Poor, 1970). There are, however, some
inherent difficulties in the shortened work- Company Studied
week; for example, limitations on the hours The company studied has a total of 1,140 mana-
of female labor, possible overburdening of gerial and operating employees. These employees are
located in manufacturing facilities in four geographi-
supervisory personnel, overtime pay questions, cally dispersed operations in the Midwest. The four
and scheduling problems. More research is divisions were each engaged in work that could be
needed on the effects of the 4-40 workweek classified, from a technological standpoint, as inter-
and other workweek variations to clarify the mediately complex (Woodward, 196S). The company
possible benefits and costs of this change in designs, manufactures, and markets a packaging ma-
chine used in the food-processing industry.
working conditions.
A recent longitudinal study by Nord and Subjects
Costigan (1973) investigated the reactions of The company's executive committee, which wanted
employees over time. There is also a need for to convert all four divisions to the 4-40 workweek,
agreed to establish two divisions as experimental and
researchers to conduct experiments to study control 2 groups to be studied for 12 months. The
the shortened workweek as an independent 2
variable that could influence such dependent The term control division is used only to indicate
the division that was not converted to the 4-40
1
Requests for reprints should be sent to John M. workweek. The company rejected the request of the
Ivancevich who is now at the Department of Or- author to study the two divisions for at least two
ganizational Behavior and Management, University years so as to ascertain any "sleeper" effects in the
of Houston, Houston, Texas 77004. 4-40 division (Campbell & Stanley, 1963).
717
718 JOHN M. IVANCEVICH

experimental division (n = 104) was converted to the ranging from .70 to .85 were generated from the 19-
4-40 system, while the control group (n = 106) con- item job satisfaction measures used.
tinued to function on the 5-40 schedule. The anxiety-stress measure consisted of nine ques-
The experimental and control subjects were similar tions in which employees were asked how frequently
on such characteristics as mean age, mean seniority, they "felt bothered" about a number of job-related
mean time in present grade, unexcused absenteeism matters (see Kahn, Wolfe, Quinn, Snoek, & Rosen-
rate, education level, and average annual wage. The thai, 1964). Examples of questions are the following:
experimental and control divisions are both located (a) not knowing just what the people you work
in suburbs of large cities, but are 120 miles apart. with expect of you; and (b) feeling that you have
The subjects participating in the study were all too heavy a workload, one that you can't possibly
members of the United Steelworkers of America finish during an ordinary workday. Employees re-
Union. Neither division has had a strike for the sponded to each item on a 5-point scale (never to
past 10 years. nearly all the time).
The unexcused absence rate was calculated on the
basis of hours per month missed on the job, where
Measures the reason was not known by the supervisor.
A questionnaire was administered during working Four performance facets examined were quality,
hours to the total operating employee population in level, effectiveness as a team member, and knowledge
both divisions. The anonymous questionnaire was of job. The performance index was a multitrait scale,
completed at three different time periods over a 13- developed by the company, which was completed by
month period. These three time periods are desig- the supervisor of each employee. The index was the
nated (a) Time Before, one- month before conversion average of 10 (6-point) ratings evaluating the em-
to the 4-40 workweek; (b) Time After 1, three ployee's performance on the four facets relative to
months after conversion; and (c) Time After 2, 12 "other employees who are now doing the same work"
months after the conversion. (Spearman-Brown split-half reliability = .81).
Job satisfaction was denned as the extent to which
the job rewards meet or exceed the employees' per- Change t tests were used to study the net differ-
ception of an equitable level of rewards on six job ences between the changes in the mean scores for
facets—self-actualization, autonomy, personal worth, both the experimental and the control groups
social affiliation, job security, and pay (Porter & (McNemar, 1962). Although changes in Time After 1
Lawler, 1968). and Time Before, and Time After 2 and Time After
The job satisfaction facet measures were generated 1 scores were analyzed, only the Time After 2 —
from a 19-item questionnaire, previously pretested Time Before net difference changes, the most rele-
for clarity and reliability in a third division of the vant, are reported in this study.
company. Examples of items are the following: (a)
"There is the opportunity to develop friendships in
the company." (b) "The compensation program for RESULTS
employees in jobs similar to mine is fair." The re-
Table 1 presents data for the evaluation of
spondents answered each of the 19 items on a 1 to
6-point Likert scale (strongly disagree to strongly changes in job satisfaction and anxiety-stress.
agree). Acceptable coefficient alpha reliability levels There was a small, but significant improve-

TABLE 1
MEAN SCORES, NET CHANGES, AND t VALUES FOR SATISFACTION
FACETS AND ANXIETY STRESS VARIABLES

Experimental division (« = 104,) Control division (n = 106)

Net dif-
changesd
TB» TAib TA 2 « TA 2 -TB TB TA, TA 2 TAz-TB
Satisfaction facets
Self-actualization 3.51 3.S8 3.57 .06 3.47 3.46 3.53 .06 .00 0
Autonomy 4.47 4.39 4.48 .01 4.34 4.39 4.38 .04 -.03 .38
Personal worth 4.41 4.49 5.48 1.07 4.42 4.61 4.62 .20 .87 6.79*
Social affiliation 4.18 4.30 4.97 .79 4.40 4.48 4.41 .01 .78 7.01*
Job security 3.61 3.95 4.38 .77 3.72 3.68 3.79 .09 .70 6.99*
Pay 3.04 3.46 4.21 1.17 3.01 3.09 3.10 .09 1.08 10.59*
Overall satisfaction 3.83 4.04 4.75 .92 3.76 3.79 3.81 .05 .87 10.02*
Anxiety-stress 2.61 2.42 1.86 -.75 2.30 2.24 2.25 -.05 -.70 9.11*

Note. The higher the mean score for the first seven variables indicates a larger amount. TB = Time Before; TAi = Time After 1;
TAa 2 = Time After 2.
b
Mean score of data collected one month before conversion to 4-day, 40-hour workweek,
c
Mean score of data collected three months after conversion.
d
Mean score of data collected 12 months after conversion.
The differences in changes for TAj — TB period in the experimental and control divisions.
 SHORTER WORKWEEK EFFECTS ON SATISFACTION AND PERFORMANCE 719
TABLE 2
MEAN SCORES, NET CHANGES, AND t VALUES FOR UNEXCUSED
ABSENCE AND PERFORMANCE FACTORS

Experimental division (n = 104) Control division (n = 106)

Net dif-
changesd
TB» TA.b TA,« TA2-TB TB TAi TA2 TAi-TB
Unexcused Absences 337.0 321.3 331.4 -5.6 349.0 357.2 341.3 -7.7 2.1 .27
Performance facets
Quality of production 3.83 3.71 4.01 .18 3.75 3.79 3.81 .06 .12 .19
Level of production 3.81 3.99 4.24 .43 3.84 3.81 3.96 .12 .31 5.03*
Effectiveness as team
member 3.71 3.94 4.54 .83 3.41 3.62 3.59 .18 .65 9.27*
Knowledge of job 3.31 3.39 3.36 .05 3.93 4.04 3.99 .06 -.01 .09
Overall performance 3.67 3.96 4.04 .37 3.73 3.81 3.84 .11 .26 4.72*

Note. The higher the mean for all of the variables, except unexcused absences, suggests better performance. TB = Time Before;
TAa = Time After; TAl = Time After 1; TA 2 = Time After 2.
b
Mean score of data collected one month before conversion to 4-day, 40-hour workweek.
0
Mean score of data collected three months after conversion.
d
Mean score of data collected 12 months after conversion.
The differences in changes for TA2 — TB period in the experimental and control divisions.
*P < .01.

ment in overall satisfaction in the experi- perceived anxiety-stress index in the experi-
mental division which had converted to the mental division.
4-40 work schedule. Four of the job satis- Table 2 presents data relevant to the un-
faction facets (personal worth, social affilia- excused absences and performance facets hy-
tion, job security, and pay) appeared to potheses. With respect to absences, the results
account for this general increase in satisfac- showed no significant changes in either the
tion. There was no discernible increase in the 4-40 or the 5-40 divisions. Thus, Hypothesis
self-actualization and autonomy facets. In 3 was not supported. The supervisory per-
spite of this latter finding, Hypothesis 1 was formance ratings of the employees showed
generally supported. significant improvement for the experimental
As predicted in Hypothesis 2, there wa.s a group in the rate of production, effectiveness
significant (but small) improvement in the as a team member, and overall performance,

TABLE 3
CHANGES IN CORRELATION COEFFICIENTS OF BEHAVIORAL AND ABSENCE VARIABLES WITH
PERFORMANCE BETWEEN TIME AFTER 2 AND TIME BEFORE DATA

Overall performance versus

experimental division Overall performance versus control division

Variables
Net change
TBa
Difference Difference
TA»b TB TA» in correla-
TA»-TB TA2-TB
tion"

Self-actualization .17 .09 -.08 .15 .14 -.01 -.07

Autonomy .13 .07 -.06 .11 .13 .02 -.08
Personal worth .19 .09 -.10 .13 .04 -.09 -.01
Social affiliation .18 .46 .28 .12 .19 -.07 .21*
Job security .01 .54 .53 .09 .01 -.08 .61*
Pay -.02 .48 .50 -.12 -.07 .05 .45*
Anxiety-stress -.11 -.19 -.08 -.07 -.02 .05 -.13
Unexcused absences -.06 -.09 -.03 -.03 -.12 -.09 -.06

Note.
a
Tests of significance were computed after r to z transformation of correlation coefficients.
b
Mean score of data collected one month before conversion to 4-day, 40-hour workweek.
0
Mean score of data collected 12 months after conversion.
The differences in changes for TA2 — TB period in the experimental and control divisions.
** < .01.
720 JOHN M. IVANCEVICH

when compared with the control group. of 4-40 advocates, who assumed positive
Quality and job knowledge ratings showed no effects of these facets (Wheeler, Gurman, &
differences. Consequently, Hypothesis 4 was Tarnowieski, 1972). Similarly, the lack of a
only partially supported. reduction in absenteeism for the 4-40 group
To test Hypothesis 5, which predicted sig- does not substantiate previous claims (Nanus
nificant changes in the relationships between & Adelman, 1971; Poor, 1970; Wheeler,
overall performance and other indices of satis- 1970).
faction and anxiety, the net correlation These results were obtained in two divisions
changes between the Time After 2 and Time in a single company, and generalizations must
Before were examined for the experimental be made with caution. Also, the study was
and control divisions (see Table 3). The re- limited to a 13-month period, and it is pos-
sults for the experimental groups showed some sible that some effects may show up or di-
significantly greater changes than those oc- minish over a longer time. Nord & Costigan's
curring in the control group. The results sug- (1973) results suggest this possibility, and
gest that stronger positive relationships be- we cannot rule out partial "Hawthorne
tween social affiliation, job security, and pay effects."
with overall performance occurred in the ex- The more positive relationships in the ex-
perimental division after the conversion to the perimental division between performance and
4-40 workweek. However, there was no sig- social affiliation, job security and pay might
nificant improvement indicated for the self- be partially the result of such an effect. The
actualization, autonomy, personal worth, employees may have considered themselves
anxiety-stress, and unexcused absenteeism re- more secure because they were the only plant
lationships with overall performance over in the organization that converted to a rela-
time. Thus, only partial support of Hypothesis tively modern 4-40 schedule. There are, of
5 is shown. course, possible intervening variables, such as
physical fatigue, attitudes toward leisure time,
DISCUSSION job pace, etc. which could be examined as
moderators of these effects. Future studies
Overall, the results are generally supportive might well examine these issues.
with regard to positive effects of the 4-40
workweek on a number of the indices investi- REFERENCES
gated. There appeared to be no negative
effects on these indices. However, not all in- Alderfer, C. P. A new theory of human needs. Organi-
zational Behavior and Human Performance, 1969,
dices showed significant effects. The results 4, 142-1IS.
emphasize the need to evaluate possible effects Campbell, D. T., & Stanley, J. C. Experimental and
on a variety of different criteria. Thus, the quasi-experimental designs for research. Chicago:
findings with certain performance facets for Rand McNally, 1963.
Chard, C. A., & Caruth, D. L. A look at the four-day
the 4-40 division seemed to support the con- workweek. Business Studies, 1973, 12, 40-48.
tention of advocates of this schedule that the Glickman, A. S., & Brown, Z. B. Changing schedules
shorter workweek may result in improved per- of work: Patterns and implications. Washington,
formance (Hall & Nougaim, 1968; Poor, D. C.: American Institutes For Research, 1973.
1970), although not all aspects of performance Hall, D. T., & Nougaim, K. E. An examination of
Maslow's need hierarchy in an organizational set-
showed significant change. ting. Organizational Behavior and Human Perform-
Furthermore, the present study indicated ance, 1968, 3, 12-35.
that after conversion to the 4-40 workweek, Kahn, R. L., Wolfe, D. M., Quinn, R. P., Snoek,
there were positive effects on anxiety levels J. D., & Rosenthal, R. A. Organizational stress.
New York: Wiley, 1964.
and on some facets of job satisfaction. How- McNemar, Q. Psychological statistics. New York:
ever, there was no discernible improvement in Wiley, 1962.
satisfaction in autonomy and self-actualiza- Nanus, B., & Adelman, H. M. Work and leisure,
tion facets. These latter facets are those which 1980. Business Horizons, 1971, 14, 5-10.
Nord, W. R., & Costigan, R. Worker adjustment to
Alderfer (1969) labels as "growth" facets. the four-day week: A longitudinal study. Journal
This finding is not consistent with the claims of Applied Psychology, 1973, 58, 60-66.
 SHORTER WORKWEEK EFFECTS ON SATISFACTION AND PERFORMANCE 721
Poor, R. 4 Days, 40 Hours. Cambridge, Mass.: Wheeler, K. E. Small business eyes the four-day
Bursk & Poor Publishing, 1970. workweek. Harvard Business Review, 1970, 43,
Porter, L. W., & Lawler, E. E., III. Managerial atti- 142-147.
tudes and performance. Horaewood, 111.: Irwin, Wheeler, K. E., Gurman, R., & Tarnowieski, D. The
1968. four-day week; An AM A research report. New
Steward, G. V., & Larsen, J. M. A Four-day—Three- York: American Management Association, 1972.
day per week application to a continuous produc- Woodward, J. Industrial organization. London: Ox-
tion operation. Management of Personnel Quarterly, ford University Press, 1965.
1971, 10, 13-20. (Received September 24, 1973)