You are on page 1of 21

PERSONNEL PSYCHOLOGY

1994,41

ANTECEDENTS OF INVOLUNTARY TURNOVER DUE


TO A REDUCTION IN FORCE
MURRAY R. BARRICK, MICHAEL K. MOUNT
Department of Management and Organizations
University of Iowa
J. PERKINS STRAUSS
Augustana College

This study examined antecedents to involuntary turnover due to a re-


duction in force. The authors used structural equation modeling and
logit regression analysis with a sample of 194 salespeople to test an ex-
ploratory process model of involuntary turnover. Results showed that
general mental ability and conscientiousness were indirectly correlated
with involuntary turnover through job performance, whereas the rela-
tionship of tenure, gender, job involvement, sales volume, and super-
visory ratings of job performance to turnover was direct. The results
demonstrate that involuntary turnover decisions are significantly re-
lated to performance; however, other variables also influenced these
decisions. Implications and future research needs are discussed.

Early research on turnover in organizations focused on understand-


ing and predicting why people voluntarily leave (for earlier reviews see
Brayfield & Crockett, 1955; Herzberg, Mausner, Peterson, & Capwell,
1957; March & Simon, 1958). Recent research has focused on refining
models that explain characteristics of voluntary leavers (e.g., Dalessio,
Silverman, & Schuck, 1986; Gerhart, 1990; Mobley, Griffeth, Hand, &
Meglino, 1979; Steers & Mowday, 1981). Researchers have also recently
begun to investigate the turnover construct itself (e.g., Campion, 1991),
recognizing that turnover can be functional or dysfunctional to the firm’s
performance (e.g., Dalton, Krackhardt, & Porter, 1981; McEvoy & Cas-
cio, 1987), avoidable or unavoidable from the organization’sperspective
(e.g., Abelson, 1987), or voluntary or involuntary from the employee’s
perspective (e.g., Dalton et al., 1981).
Very little research has examined the characteristics of those who
leave for involuntary reasons, particularly when the firm decides who
should leave (e.g., termination for poor performance or dismissal for re-
duction in force). This is somewhat surprising since the firm’s choice of
who should leave may have as large an impact on organizational effec-
tiveness as the employee’s choice to leave. In light of the large number

Correspondence and requests for reprints should be sent to Murray R. Barrick, De-
partment of Management and Organizations, Universityof Iowa, Iowa City, IA 52242
COPYRIGHT 0 1994 PERSONNEL PSYCHOLOGY. INC.

515
516 PERSONNELPSYCHOLOGY

of downsizings and layoffs that have recently occurred, it is as impor-


tant to understand organizational choices as it is to understand individ-
ual choices in turnover behavior. The purpose of the present paper is
to examine antecedents of the involuntary turnover process when the
turnover reflects the organization’schoice of who leaves during a reduc-
tion in force. In the following pages we first briefly discuss involuntary
turnover as it is operationalized in the current study and review the lit-
erature on antecedents of involuntary turnover. We then formulate an
initial model of involuntary turnover which will be tested using data col-
lected in a large firm that underwent a reduction in force.
When studying involuntary turnover, it is important to know the rea-
sons for leaving. In conducting this study, it became apparent that some
of the turnover decisions associated with the reduction in force were
based on mutual agreement rather than solely due to the firm’s choice.
That is, there were a few cases where older employees voluntarily ac-
cepted an early retirement incentive package offered by the firm as part
of the reduction in force. Although the voluntary-involuntary turnover
distinction has been recognized to be on a continuum (Campion, 1991),
in the current study, early retirement was considered primarily a vol-
untary turnover decision rather than involuntary turnover and, conse-
quently, was excluded. Also excluded were unavoidable turnover situ-
ations when individuals left for reasons beyond their (or the organiza-
tion’s) control, such as death, illness, or spouse moving (Abelson, 1987).
Thus, according to Abelson’s expanded avoidability taxonomy (p. 393),
the type of turnover considered in the current study is both involuntary
(from the individual’s perspective) and avoidable (from the organiza-
tion’s perspective). Further, one could consider this type of turnover
functional for the organization using Dalton et al.’s (1981) definition of
functionality (i.e., the individuals terminated would most likely be poor
performers). Therefore, when referring to involuntary turnover in this
study, we restrict our analysis to involuntary/avoidable turnover (those
employees who were dismissed by the organization during the reduction
in force).
We were unable to locate any studies investigating antecedents to
involuntary turnover in organizations undergoing a reduction in force.
However, four studies investigated antecedents of involuntary turnover
due to the organization’s decision to terminate the employment relation-
ship. Wanous, Stumpf, and Bedrosian (1979) investigatedjob survival of
1,736 newly hired workers across a variety of organizations. They found
that demographic variables (e.g., sex, race, age) had little or no impact
on involuntary turnover (i.e., discharged for cause or not recommended
for reenlistment). However, for those who left involuntarily, mean job
performance levels were significantlylower than the mean performance
MURRAY R. BARRICK ET AL. 517

for those who stayed or left voluntarily (the performance differences,


expressed in standard score units is d = -.57). Additionally, job atti-
tudes (e.g., how favorably the job was perceived by the employee) were
significantlyrelated (d = -.24) to involuntary turnover (more favorable
attitudes were associated with less turnover).
La Rocco, Puch, and Gunderson (1977) investigated retention deci-
sions for 642 naval personnel. In general, those who were discharged or
not recommended for reenlistment (involuntaryturnover) were younger
(d = -.58), less likely to be married (d = -.95), had lower performance
ratings (d = -.62), and were less satisfied ( d = -.12) than those who
were eligible and did reenlist (stayers) or those who were eligible and
did not reenlist (voluntary turnover). Furthermore, those with longer
job tenure (d = -.47) and greater job involvement (d = -3) were less
likely to be involuntary leavers.
Stumpf and Dawley (1981) found that sex (d= -.35; females with-
drew less often), age (d = -.48), tenure (d = -.72), and performance
(d = -34) were all negatively and significantly related to involuntary
(dismissal) and voluntary turnover of bank tellers. However, three in-
dices of performance (cash balance records, promotional increases, and
overall performance ratings) were more strongly correlated with invol-
untary turnover (d = -1.17, -.86, -.78, respectively) than with volun-
tary turnover (d = -.91, -.71, -.64, respectively).
Wells and Muchinsky (1985) considered three types of turnover in
their study of 420 credit managers from a large finance company: quit-
ters (voluntary turnover), those fired (involuntary turnover), and those
promoted. They found that promoted employees performed signifi-
cantly better than those who voluntarily quit (d = .65), and that those
who voluntarily quit performed significantlybetter than employees who
were involuntary leavers (d = SO). Significantlymore males were in the
involuntary turnover group than were in the voluntary turnover group ( d
= -.37), and more minority employees were in the voluntary turnover
group than in either the promoted or involuntary turnover groups ( d =
.31). However, age, education, and class rank did not vary significantly
across the three groups.
Although Wanous et al. (1979), La Rocco et al. (1977), Stumpf and
Dawley (1981), and Wells and Muchinsky (1985) all considered invol-
untary turnover in their analyses, the reason for the terminations was
for poor performance. Although this is conceptually similar to a reduc-
tion in force, there is a difference. For example, it is possible that the
relationship between performance and dismissal is stronger when there
is no organizational goal to restructure the organization. In contrast,
when the termination is due to a reduction in force (i.e., restructuring),
it becomes imperative to layoff employees who would otherwise not be
518 PERSONNEL PSYCHOLOGY

terminated. Thus, when an organization decides to downsize, although


one would expect performance to be a strong correlate with turnover,
there may be other factors involved in the termination decision. The
primary strength of the current study, then, is that it is the first attempt
to investigate antecedents of turnover in an organization going through
a reduction in force.
Our purpose in this study was to formulate and test a model assess-
ing antecedents to involuntary turnover. In developing this model, we
sought to enhance our understanding of the relative importance of vari-
ous individual differences in attitudinal, dispositional, demographic, and
behavioral variables as antecedents of involuntary turnover, and to use
such knowledge to enable decision makers facing reductions in force in
the future to better consider the best interests of the organization and
the individual.
As indicated above, there is a paucity of research pertaining to in-
voluntary turnover. Considering this when developing our model, we
included several key antecedent variables from the voluntary literature
that were particularly applicable to involuntary turnover. In doing so,
we recognized that the antecedents of voluntary turnover are different
due to the different nature of involuntary turnover (e.g., La Rocco et
al., 1977; McEvoy & Cascio, 1987; Stumpf & Dawley, 1981; Wanous et
al., 1979; Wells & Muchinsky, 1985). Therefore, the model tested in
this study should be viewed as an exploratory model of the involuntary
turnover process.

A Model of Involuntary Turnover

Most models of antecedents of voluntary turnover propose that indi-


vidual difference characteristics (e.g., demographics) and the environ-
ment (e.g., job characteristics) influence the employees’ affective re-
sponses (e.g., job satisfaction or job involvement). In turn, the affective
responses initiate withdrawal cognitions and decision processes (e.g., in-
tent to quit) that are then directly related to an individual’s likelihood
of voluntary turnover (e.g., Bannister & Griffith, 1986; Dalessio e t al.,
1986; Gerhart, 1990; Mobley et al., 1979; Steers & Mowday, 1981).
The model proposed in this study includes individual difference vari-
ables (e.g., tenure, gender, GMA, conscientiousness) and an affective
variable (job involvement) which have also been included in previous
studies of involuntary turnover. In developing the model, considera-
tion also was given to those variables the organization’s decision mak-
ers attended to while managing the workforce reduction. Although this
organization had no formally stated workforce reduction “policy,” deci-
sion makers nevertheless reported that they were significantlyinfluenced
MURRAY R.BARRICK ET AL. 519

by performance-related attitudes and behaviors. Therefore, job perfor-


mance was expected to play a substantial role in involuntary turnover.
Thus, we included two indices of performance (sales volume and super-
visor ratings) to predict involuntary turnover behavior.
The relationship between employee performance and turnover be-
havior is one area that has frequently been investigated in turnover re-
search. In their 1987 meta-analysis, McEvoy and Cascio concluded that
although three forms of turnover (voluntary,involuntary, and total) were
negatively related to performance, involuntary turnover had the highest
(negative) average true score correlation with performance ( p = -.51).
Based on these findings, it was hypothesized that sales volume and su-
pervisory ratings of performance would be negatively correlated with in-
voluntary turnover.
Another variable examined in this study, job involvement, assesses
the degree of psychological identification with one’s job. While organi-
zational commitment and job involvement have been used as measures
of affective responses in voluntary turnover research, job involvement
was believed to be most relevant to this reduction in force. Organiza-
tional decision makers were particularly concerned with retaining in-
dividuals who were actively involved in their jobs. Therefore, we be-
lieved job involvement would be an important antecedent of involun-
tary turnover because highly involved employees would be perceived as
more valuable, a critical factor given the limited promotional opportu-
nities in the foreseeable future. Support for this was found by La Rocco
et al. (1977) who reported a significant difference in job involvement ( d
= -.51) between those who were discharged or not recommended for
reenlistment (involuntary turnover) and those who reenlisted, with stay-
ers being more involved in their jobs. Job involvement was therefore
hypothesized to be negatively related to involuntary turnover.
Job involvement should also be related to employee performance.
Research on the relationship between job involvement and employee
performance generally indicates modest correlations (e.g., Rabinowitz
& Hall, 1977; Saal, 1978), although the results are not conclusive. More
recent evidence from a meta-analysis of organizational commitment
(Mathieu & Zajac, 1990) indicates that supervisor ratings are modestly
correlated with affective measures like job involvement. Therefore, job
involvement was posited to be positively related to both performance
measures (sales volume and supervisory ratings of performance) in this
study.
Two individual difference variables were included in the model pri-
marily because of their relationship with job performance. Over the past
15 years, research evidence of the predictive validity of general men-
tal ability (GMA) has accumulated which indicates that the true score
520 PERSONNEL PSYCHOLOGY

correlation with performance is around .56 (Hunter, 1986; Schmidt &


Hunter, 1981). Therefore, GMA was hypothesized to be positively cor-
related with sales volume and supervisory ratings of performance in
this study and, thereby, indirectly associated with involuntary turnover
through its effects on periormance.
Evidence for the predictive validity of the second individual differ-
ence variable, conscientiousness, has emerged recently (e.g., Barrick
& Mount, 1991, 1993; Barrick, Mount, & Strauss, 1993; Hough, 1992;
Mount, Barrick, & Strauss, 1991). In the last decade a five-factor model
of personality (the Big Five) has received wide attention as a compre-
hensive and parsimonious taxonomy of personality traits (see Digman,
1990, for an in-depth discussion). The five factors include: (a) Extraver-
sion (e.g., sociable, talkative, energetic, ambitious), (b) Agreeableness
(e.g., good-natured, cooperative, trusting), (c) Conscientiousness (e.g.,
responsible, dependable, planful, achievement oriented), (d) Emotional
Stability (e.g., calm, secure, not nervous), and (e) Openness to Experi-
ence (e.g., imaginative, artistically sensitive, intellectual).
Meta-analytic results suggest that there is a positive relationship be-
tween these personality constructs, particularly conscientiousness, and
various criteria (e.g., Barrick & Mount, 1991; Hough, Eaton, Dunnette,
a m p , & McCloy, 1990). Barrick and Mount (1991) meta-analyzed
available test validities and found true score correlations in the range
of .20-.23 for conscientiousnessmeasures for all job families and all cri-
terion types investigated. In contrast, validities for the other Big Five
factors generally were smaller or only predictive for a subset of occupa-
tional types or criterion categories. In another construct-oriented meta-
analysis, Hough (1992) reported observed validities of approximatelythe
same magnitude for two predictor constructs, achievement and depend-
ability (both aspects of conscientiousness),for job proficiency measures.
There are also theoretical reasons for including conscientiousness in
an involuntary turnover model in lieu of other personality factors. Re-
cently, theoretical models of job performance (Borman, White, Pulakos,
& Oppler, 1991; Hunter, 1986; Schmidt, Hunter, & Outerbridge, 1986)
have conceived of performance as comprising “can do” cognitive factors
(e.g., job knowledge) and “will do” motivational factors (e.g., awards
and commendations). Furthermore, results from a study by Barrick et
al. (1993) demonstrate that the “can do” performance components are
primarily predicted by GMA, while the “will do” factors are predicted by
conscientiousness. These results suggest there is a link between consci-
entiousness and motivational components of work performance. Thus,
conscientiousnessis viewed as a “trait motivation” variable (Schmidt &
Hunter, 1992) with important effects in the work domain. As with GMA,
MURRAY R. BARRICK ET AL. 521

conscientiousness was hypothesized to be positively related to sales vol-


ume and supervisory ratings of performance in this study and indirectly
associated with involuntary turnover. Given no research on the relations
between these two predictors and involuntary turnover, a direct relation-
ship was not hypothesized.
Two demographic variables are included in the present model, gen-
der and job tenure. Most voluntary turnover models include demo-
graphic variables such as age, gender, race, tenure, marital status, num-
ber of dependents, and educational experience (e.g., Cotton & Tuttle,
1986; Lee & Mowday, 1987). In the current study, data on gender,
age, and job tenure were available. However, the latter two variables
were highly correlated, therefore only one variable was included in the
analyses. Since job experience has been found to play a significant role
in models of job performance (Schmidt et al., 1986), job tenure was
included in the model. It should be noted that Stumpf and Dawley
(1981) found that tenure contributes over 7% explained variance be-
yond that explained by gender and age for both voluntary and involun-
tary turnover. Furthermore, the firm’s decision makers suggested that
when all other factors were equal, they tended to follow the “last hired,
first fired” rule of thumb. Turning to gender, two of the studies reviewed
previously (Stumpf & Dawley, 1981; Wells & Muchinsky, 1985) found
that women were less likely to be discharged. Thus, based on these re-
sults, we hypothesized that both job tenure and gender (i.e., women will
be less likely to turnover than men) will be negatively related to turnover.
Tenure, gender, and conscientiousnesswere expected to be indirectly
associated with turnover through job involvement. This is based on a re-
cent meta-analysis (Mathieu & Zajac, 1990)which shows that these vari-
ables (conscientiousness was assessed by one of its components, need
for achievement) are positively related to organizational commitment,
an affective variable highly correlated with job involvement (Brooke,
Russell, & Price, 1988; Mathieu & Zajac, 1990). In contrast, GMA
was not significantly related to organizational commitment. Therefore,
GMAwas expected to be unrelated to job involvement. Furthermore, no
strong theoretical rationale existed for positing a relationship between
GMA and job involvement. Based on these results, we hypothesized
that tenure, gender (women will be more involved than men), and con-
scientiousnesswould be positively related to job involvement.
Finally, sales volume was hypothesized to directly relate to perfor-
mance ratings. In support of this hypothesis, Borman et al. (1991), in
their model of supervisor performance ratings, found that task profi-
ciency measures (as measured by work sample tests) were positively re-
lated to performance ratings.
522 PERSONNELPSYCHOLOGY

The hypothesized relations are reported in parentheses in Figure 1.


To summarize, GMA and conscientiousnesswere expected to indirectly
relate to involuntary turnover through the two performance measures.
Furthermore, conscientiousness was also expected to indirectly relate
to turnover through job involvement. The two demographic variables,
tenure and gender, were expected to be directly linked to involuntary
turnover, and indirectly linked through job involvement. In turn, job in-
volvement was expected to be both directly associated with turnover and
indirectly associated with turnover through the mediating mechanisms
of sales volume and supervisory ratings of performance. Finally, both
performance measures were expected to directly relate to involuntary
turnover, while sales volume was also expected to be indirectly linked to
turnover through supervisory ratings of job performance.

Method

Sample

This study is an extension of a previous study (Barrick et al., 1993)


investigating the relationship between conscientiousness and job perfor-
mance. Subsequent to collecting the data for the Barrick et al. study,
the researchers learned of the reduction in force taking place within the
organization. Thus, many of the predictor variables used in this model
were actually collected as part of the Barrick et al. study. Our sample
consisted of 227 sales representatives in a large appliance manufactur-
ing organization; 86% of these were men (women were coded “ 1”; men
were coded “O”), the median age was 38 years, and the average tenure
in the organization was 8.7 years. The nature of the sales is best classi-
fied as wholesale because the customers were appliance dealers rather
than end users. It should be noted that over the course of the study, 33
subjects left the firm for reasons other than involuntary turnover (un-
avoidable leavers, N = 8; voluntary leavers, N = 16; and early retirees,
N = 9). These subjects were excluded from all analyses because they
were not relevant to the study.
Analyses were conducted on the data for the remaining 194 of the
227 original subjects. In mid-1991 subjects completed a demographic
data form, a measure of GMA, the Personal Characteristics Inventory
(PCI, a personality inventory which measures the Big Five personality
dimensions), and a job involvement questionnaire. Turnover data were
collected 18 months later. Performance data were available for a subset
of the subjects (N = 109) and consisted of supervisory ratings of overall
performance in the second half of 1991 and the total volume of sales over
MURRAY R. BARRICK ET AL. 523

the same period for the sales representatives. It should be noted that
performance data for later periods were unavailable to the researchers.
GMA. GMA was assessed by the Wonderlic Personnel Test, Form 5.
Across forms, test-retest reliabilities reported in the test manual range
from .82 to .94. Alternate form reliabilities range from .73 to .95 while
other measures of internal consistency (e.g., a,KR-20) reported range
from .88 to .94 (see WonderlicPersonnel Test Manual, 1983).
Conrcientiousness. Conscientiousness was measured with the PCI
(for a more thorough description of item content and construct validity
evidence for the PCI, see Barrick & Mount, 1993, and Barrick et al.,
1993). Coefficient alpha was .85 for this scale, and in a previous study a
test-retest reliability of .84 was obtained.
Job involvement. Job involvement was measured with the 6-item
version of the job involvement scale developed by Lodahl and Kejner
(1965). Examples of items are: “The major satisfaction in my life comes
from my job” and “The most important things that happen to me involve
my work.” Corrected odd-even reliability coefficients for this scale have
ranged from .72 to .89 (Goodman, Furcon, & Rose, 1969; Lodahl &
Kejner, 1965), and in this study the alpha coefficient was .78. A 5-point
Likert scale anchored by “strongly agree” and “strongly disagree” was
used, with negative items recorded so that a high score on the scale is
indicative of high job involvement.
Job performance ratings. An 11-dimension measure of job perfor-
mance was developed by the researchers based on an analysis of the
sales jobs. The dimensions were job knowledge, quality of work, quan-
tity of work, initiative, customer communications, organizational com-
mitment, job commitment, planning and allocation, interpersonal ori-
entation, self-development7and account management. Each dimension
was defined by a one-sentence description, followed by three interpreta-
tive examples illustrating important facets of that dimension. The sub-
j e c t ~supervisors
~ rated their performance on a 5-point Likert scale from
“Consistently exceeds job requirements” to “Somewhat below job re-
quirements.” Raters were informed that ratings were being collected
for research purposes. Overall performance was the sum of the rankings
across all dimensions. The alpha coefficient was .86. Interrater agree-
ment could not be assessed but is assumed to be comparable to the mean
value (SO) reported in Rothstein (1990).
Sales data. The measure of sales performance was a z-score com-
puted in the following way. The mean sales of the geographic region
over the second half of 1991 was subtracted from the salesperson’s av-
erage monthly sales, divided by the standard deviation of that region. A
2-scorewas used in order to control for the wide variance in sales volume
in the 14 different geographical regions.
524 PERSONNEL PSYCHOLOGY

Turnover: Turnover data were collected 18 months after the other


data were collected, although the actual reduction in force was initiated
9 to 12 months after the other data were collected. Although no formal
“layoff’ policy was stated, the separation decisions may be best under-
stood by focusing on who is retained rather than who is dismissed. In
making these decisions, decision makers were guided by the policy of re-
taining those employees who were highly motivated and possessed the
necessary skills and abilities to do the work. In addition, the decision
makers were influenced by a preference for more experienced employ-
ees, all else being equal.
In this study, there were two turnover categories of interest: stayers
( N = 136) and involuntary leavers ( N = 58). To identify which employ-
ees were no longer with the firm, a list of employees in 1993 was com-
pared with a list from mid-1991. Reasons for leaving were obtained from
the director of training and not from employee records. The director
was selected because he had frequent access to all of the sales represen-
tatives and supervisors and had previously been a sales representative.
Due to this, the director was familiar with the sales representatives and
was expected to have more accurate knowledge of why sales represen-
tatives left than was available from company records. In addition, the
director called all of the relevant supervisors to verify the reasons for
leaving. If the supervisor was not sure of the employees’ reasons for
leaving, the director checked with the sales representatives. Although
the training director provided the turnover data (the researchers were
not able to interact directly with supervisors or sales representatives), it
should be noted that extensive information was elicited from the super-
visors and sales representatives about the reason for the turnover deci-
sion. Therefore, while it was not possible to directly assess the reliability
of the various sources of information, the turnover measure itself was
based on information from different perspectives, as suggested by Cam-
pion (1991).

Results

Descriptive statistics and uncorrected correlations between variables


are presented in Table 1. As can be seen, the two performance indices,
sales volume and supervisory ratings of job performance, had the highest
correlations with involuntary turnover (T = -.21 and -.35, respectively).
The other variables were also significantly correlated with turnover, al-
though the magnitude of the relations were smaller (T ranged from .12 to
.16, all p < .05). Furthermore, GMA and conscientiousnesswere signif-
icant predictors of the two performance measures ( T ranged from .14 to
MURRAY R.BARRICK ET AL. 525

TABLE 1
IntercorrelationMatrix
Variables
Variable Mean (SD) 1 2 3 4 5 6 7
1.GMA 28.08 (5.90)
2. Conscientiousness 2.61 (0.23) -.08
3. Age 38.00 (9.58) -.07 .15*
4. Gender 0.14 (0.29) .06 -.07 -.19*
%Jobinvolvement 3.53 (0.63) -.19** .29** .16* -.19**
6. Salesvolume 0.00 (0.91) .14* .19**-.04 -.01 .08
7. Supervisor ratings 3.56 (0.69) .18** .25**-.07 .02 -.16* .20**
8. Involuntary turnover 0.299 (0.28) -.14* -.15* .12* -.13* -.16* -.21** -.35**
N=194
*p< .05, all tests one-tailed
**p<.O1, all tests one-tailed
.25, p < .05). These results for conscientiousnesscorrespond to those re-
ported for sales representatives in previous meta-analytic reviews (e.g.,
Barrick & Mount, 1991). In addition, job involvement was significantly
correlated with conscientiousness, GMA, gender, and supervisory rat-
ings of performance (T- ranged from .16 to 29, p < .05). Finally, as
expected, none of the other Big Five personality constructs were signifi-
cantly correlated with involuntary turnover or either of the performance
indices.
We assessed the proposed turnover model by using maximum likeli-
hood estimation as implemented in LISREL VII (Joreskog & Sorbom,
1989). All subsequent analyses were based on the covariance matrix. In
this model, all the variables were treated as single indicators of latent
variables. An alternative approach would have been to operationalize
all variables in the model as multiple indicator variables, although this
would have required a larger sample to derive stable estimates of pa-
rameters. Finally, to control for the effects of measurement errors, the
estimates were corrected for unreliability. The following reliability esti-
mates were used: GMA (.85), conscientiousness (.89), job tenure (.95),
gender (.95), job involvement (.78), sales volume (.67), supervisor rat-
ings ( S O ) , and turnover (.95).l
‘The three “objective” indices (tenure, gender, and turnover) were assumed to be very
reliable indicators, although the .95 estimate allowed for some measurement error. The
performance ratings reliability estimate is based on interrater reliability rather than coef-
ficient alpha because the most important source of error in performance ratings are those
attributable to differences between raters rather than to the content of the measurement
procedure. The average interrater reliability of a single supervisor’srating of .50 is based
on the average mean estimate reported by Rothstein (1990). The sales volume reliability
estimate is based on a sample-weighted reliability estimate for sales measures reported in
a recent meta-analysis for objective measures assessed over 26 weeks (Hunter, Schmidt,
& Judiesch, 1990).
526 PERSONNEL PSYCHOLOGY
TABLE 2
Fit Statistics ofHypothesized and Alternative Turnover Models
Models
Fit statistic Hypothesized Null A C
Xa 10.43 131.16* 8.15 45.41*
(df1 (7) (22) (5) (12)
Difference in x a over hypothesized model NA 120.73’ -2.280 34.98’
Goodness-of-fit index (GFI) ,987 .858 .990 .949
Adjusted GFI (AGFI) .933 .768 .926 .847
Root mean square residual (RMSR) ,079 ,079 .078 ,082
Tucker-Lewis index .91 NA .873 .439
Normed-fit index .92 NA ,938 .654
Note: Model A = the “next most likely” unconstrained model, and Model C = the
“next most likely” constrained model; NA = Not Applicable, the GFI, AGFI, and RSMR
are recommended by Joreskog and Sorbom (1989); the normed-fit index by Bentler and
Bonnett (1980); the ’hcker-Lewis index by Marsh, Balla, and McDonald (1988).
*p<.o1

The goodness-of-fit indices reported in Table 2 demonstrate that


the hypothesized model has an acceptable fit with the data; the chi-
square was nonsignificant (x2 = 10.43, # = 7, p =.17), the adjusted
goodness-of-fit index (AGFI) of .93, the Tucker-Lewis index of .91, and
the normed-fit index of .92 were all well above the .90 guideline, and
the root mean square residual (RMSR) was below .lo (RMSR = .OS).
These measures indicate that the theoretical model has a good fit to the
data. The second column in Table 2 shows the fit statistics for the null
model, and as shown, the hypothesized model fit the data significantly
better than the null model. To determine the relative adequacy of the
hypothesized model in comparison to two alternative models, the “next
most likely” constrained (Model C) and unconstrained (Model A) al-
ternatives from a theoretical perspective were compared to the hypoth-
esized model. The more unconstrained model (Model A) posited two
additional linkages that were not included in the hypothesized model.
That is, in Model A conscientiousness and GMA were also expected to
directly relate to involuntary turnover. In contrast, the more constrained
model (Model C) differs from the hypothesized model in that conscien-
tiousness, tenure, and gender were not expected to relate directly to job
involvement, and in turn, job involvement was not expected to be asso-
ciated with either performance measure. The “next most likely” con-
strained ,model,Model C, was developed by deleting relations from the
hypothesized model that were expected to be relatively small (i.e., cor-
relations were expected to be less than .20), based on the literature pre-
viously reviewed.
As shown in Table 2, there was a significant decrease in fit when using
the “next most likely” constrained model (Model C) in comparison to
MURRAY R.BARRICK ET AL. 527

I I

22’.

Figure 1: Path Coefficients for Antecedents of Involuntary Turnover (Hypoth-


esized Relations Posited in Parentheses; * p < .05, one-tailed; **p <
.01, one-tailed).

the hypothesized model. Therefore, the hypothesized model explained


the data better than Model C. Conversely, there was not a significant dif-
ference in fit with respect to the “next most likely’’ unconstrained model
(Model A) in comparison to the hypothesized model. That is, there was
no difference in the explanation of the data between Model A and the
hypothesized model, and consequently, both models adequately account
for the observed data. However, because parsimony is preferred and the
two additional relations posited in Model A did not significantly add to
the explanation of the data, the hypothesized model was preferred.
The standardized solution for the hypothesized model is presented
in Figure 1. As expected, there were significant direct linkages between
involuntary turnover and performance ( p = -.15 and -.56, respectively
for sales volume and supervisory ratings of performance), job involve-
ment (p = -.37), and gender (0= -.24). Also as expected,,there was
a significant direct linkage between GMA and conscientiousness with
performance (sales volume, p = .22 and .24; supervisory ratings of per-
formance, p = .23 and .40,respectively) and between sales volume and
supervisory ratings of performance (p = .25). Conscientiousness was
also significantlylinked with job involvement (0 = .32), as hypothesized.
Contrary to expectations, three relationships were in the opposite di-
rection hypothesized: Tenure was positively associated with involuntary
turnover ( p = .16), gender was negatively linked with job involvement
( p = -.20), and job involvement was negatively related to supervisory
ratings of performance (p = -.36). Finally, job involvement was not
PERSONNELPSYCHOLOGY

significantly linked to sales volume ( p = .08), and tenure was not signif-
icantly associated with job involvement ( p = .Ol), as hypothesized.
For the entire model, the multiple correlation for involuntary turn-
over was .66 (R2= .440). As this suggests, 44% of the variability in
involuntary turnover was accounted for by the relationships posited in
the model. To enhance our understanding of the relations among vari-
ables, the sum of the direct and indirect relations of each antecedent
variable with involuntary turnover was calculated, based on the pattern
of correlations reported in Figure 1. Involuntary turnover was most
highly related to supervisory ratings of job performance (-.56), sales
volume (-.29), conscientiousness (-.33), and job involvement (-.18).
A major portion of the relationship with involuntary turnover for the
two performance variables and job involvement were direct. Two other
variables, tenure and gender, were also primarily related directly with
involuntary turnover (.16 and -.24, respectively). In contrast, GMA
and conscientiousness were associated only indirectly with involuntary
turnover through job involvement and the two performance measures,
and for both variables, the major portion of these relations was through
supervisory ratings (- .13 and - .22, respectively). In sum, the associa-
tion of GMA and conscientiousness with involuntary turnover was in-
direct, while the association of the other five predictor variables with
involuntary turnover primarily was direct. In fact, only two variables
other than GMA and conscientiousness had a substantial indirect rela-
tionship with turnover. They were job involvement and sales volume,
which indirectly affected turnover through supervisory ratings (.20 and
- .14, respectively).
A nonlinear maximum likelihood logistic regression of turnover was
also conducted in this study in order to ensure that the results were not
influenced by the use of a dichotomous dependent variable. Recent ev-
idence (Huselid & Day, 1991) demonstrates that turnover researchers
often adopt inappropriate estimation procedures, particularly ordinary
least squares, which provides biased parameters when the dependent
variable is binary and errors are not normally distributed. In contrast,
probit or logit analysis models have none of these limitations when pre-
dicting binary variables. Furthermore, these nonlinear procedures as-
sume the underlying distribution of the dichotomous variable is a con-
tinuous latent variable, which corresponds to the theoretical assumption
implicit in turnover research (Campion, 1991). Table 3 summarizes the
results of the logistic regression analysis for all the variables included
in the path model. The table contains the regression coefficients (and
t-values) reflecting the change in the logarithmic odds of turnover for
unit changes in the antecedent variables and the model chi-square, which
represents a comparison between the null model that all coefficients are
MURRAY R. BARRICK ET AL. 529

TABLE 3
Summaly of Results of Logistic Regression of Turnover

Variables Involuntary turnover t

Constant 2.91' 1.82


GMA -0.37* -1.68
Conscientiousness -1.55' -1.80
Gender -1.31' -1.30
Tenure 0.05' 1.67
Job involvement -0.42' -1.31
Sales volume -2.80'. -1.97
Supervisor ratings -1.20'* -2.86
Model xZ 14.29'
p-value -0.015
*p<.05, all tests one-tailed
**p<.Ol, all tests one tailed

zero and the estimated model. As shown in Table 2, the model chi-square
was significant (p < .05), indicating that the estimated model represents
a significant improvement over the null model. Furthermore, each of the
variables was significantly related to turnover, and similar to previous
results, the two best predictors of turnover were supervisory ratings of
job performance (p < ,005) and sales volume (p < .02).
In general, results from the two sets of analyses were very similar.
However, the relative magnitude of some of the regression coefficients
did differ. For example, the coefficient for supervisory ratings of perfor-
mance was larger than salesvolume in the path analyses while the reverse
was true for the logistic regression. There are two possible explanations
for this finding. First, these variables are scaled differently; that is, sales
volume is expressed in z-score form, whereas the other variables are not.
Second, the nonlinear nature of the logit analysis may account for these
differences. Nevertheless, taken together, these analyses demonstrate
that the pattern of relations depicted in Figure 1is supported irrespec-
tive of whether zero-order correlations, linear structural equations, or
a nonlinear logistic regression estimation technique was used to analyze
the data.

Discussion

Relatively little is known about involuntary turnover or its antece-


dents. Due to continued downsizings and layoffs, both organizational
decision makers and researchers should be interested in understanding
the antecedents of the involuntary turnover decision. Furthermore, be-
cause organizations have more control over turnover due to a reduction
in force than turnover due to employee choice (i.e., voluntary turnover),
530 PERSONNELPSYCHOLOGY

enhanced understanding should improve these decisions and thereby


increase organizational effectiveness. This exploratory study provides
initial evidence of the relations between relevant predictors, including
individual difference characteristics, affective responses, and employee
performance with involuntary turnover behavior due to a reduction in
force.
The strongest predictor of involuntary turnover was supervisory rat-
ings of job performance (a = -.56). The other performance measure,
sales volume, was also related to turnover (0 = -.15) and was indirectly
related to turnover through supervisory ratings ( p = .25). The large re-
lations between the two performance measures and involuntary turnover
were not surprising because anecdotal evidence indicated that organiza-
tional decision makers emphasized employee performance when making
decisions during the firm’s reduction in force. These results also corre-
spond to previous meta-analytic research (McEvoy & Cascio, 1987).
Because employee performance has such a large impact on involun-
tary turnover decisions, it would seem that any variables that are an-
tecedents to performance should be included in a model of involun-
tary turnover decisions because they could have indirect relationships to
turnover. The results of this study confirm this assumption. For exam-
ple, there is evidence that GMA is indirectly associated with involuntary
turnover through performance. As expected, high ability employees at-
tained higher performance, which influenced the organizational decision
makers when making dismissal decisions. Furthermore, conscientious-
ness was found to be associated with supervisory ratings of performance
and sales volume and, through these variables, was associated with in-
voluntary turnover. In addition, conscientiousnesswas indirectly related
to turnover through its relationship to job involvement. As this suggests,
the more organized, planful, responsible, dependable, hardworking, per-
sistent, and achievement-oriented people are, the higher their perfor-
mance and the more involved they tend to be in the job, which in turn
reduces their likelihood of being laid off during a reduction in force.
This finding may have relevance to the voluntary turnover literature
as well. Because conscientious employees are more responsible and de-
pendable, they are apt to be more involved in and committed to the or-
ganization, which in turn reduces the likelihood that they will voluntarily
leave, Thus, an important question for future research is whether con-
scientiousness is also significantlyrelated to voluntary turnover.
The findings pertaining to job involvement were surprising in some
respects. First, the effect size for the direct relationship on involuntary
turnover reported for job involvement was relatively large, even after
accounting for performance. That is, the more involved an employee
was with the job, the more likely it was that the employee would remain
MURRAY R. BARRICK ET AL. 531

with the organization, regardless of performance level. The magnitude


of this relationship (0 = -.37) suggests that affective responses (e.g.,
job involvement and organizational commitment) may be as important
in predicting involuntary turnover decisions as they are for voluntary
turnover decisions.
Even more surprising was the finding that, contrary to our expecta-
tions, job involvement was negatively related to supervisory ratings of
job performance. That is, employees who were more involved in their
jobs tended to get lower supervisory ratings ( p = -.36). These results
are counterintuitive. Given the negative relationship found between job
involvement and turnover in the current study, additional research is
needed to clarify the nature of the relationship with involuntary turnover
due to a reduction in force.
The findings for the two demographic variables provided mixed sup-
port for the hypotheses. As hypothesized, men were more likely to be
dismissed than women during a reduction in force. However, contrary
to our hypothesis, employees with greater tenure were more likely to
be dismissed than those with less tenure. The latter finding contradicts
those found in a previous study of involuntary turnover (Stumpf & Daw-
ley, 1981). This suggests that the relationships between antecedents and
involuntary turnover may vary depending on the reason the organization
dismisses the employee (e.g., for cause or a reduction in force). Thus,
employees with greater tenure, who have remained with the firm over a
period of years (not dismissed for cause), may be more likely to be ter-
minated when the organization is attempting to reorganize or downsize.
One explanation for this finding is that older employees are more highly
paid, therefore terminating them is more consistent with the cost-saving
purpose of a reduction in force. However, this may not necessarily be
the case when turnover is due to other reasons.
The relations between antecedents and involuntary turnover in this
study differ from those obtained in the voluntary turnover literature in
several ways. First, the largest correlation with involuntary turnover was
with supervisory ratings of job performance ( p = -.51; corrected for at-
tenuation), rather than the affective variable (job involvement) as sug-
gested in the voluntary turnover literature (e.g., Mathieu & Zajac, 1990;
Steers & Mowday, 1981). The finding for supervisoryratings, which cor-
responds to the correlation of -.51 for involuntary turnover reported
in the McEvoy and Cascio (1987) meta-analysis, illustrates the impor-
tance of performance on involuntary turnover decisions. Second, our re-
sults suggest that GMA and conscientiousnessare likely to have stronger
indirect relations with involuntary turnover through their associations
532 PERSONNEL PSYCHOLOGY

with performance than is the case with voluntary turnover. As por-


trayed in the model, this occurs because these two variables are impor-
tant antecedents of performance, and employee performance has a more
substantial influence on involuntary turnover than voluntary turnover.
Third, the results for tenure and gender were opposite those found in
the voluntary turnover literature, where departing employees are likely
to have less tenure and are female rather than male.
This research provides initial evidence of the nature and strength of
the antecedents and mediators of involuntary turnover. Nevertheless,
additional research that tests the relations posited in this study is nec-
essary for several reasons. First, the study focused on one occupational
group, sales representatives, and on the employees of a single organiza-
tion in a non-union environment. Although many firms have personnel
policies and practices similar to those of the present organization and
employees performing tasks similar to those of the sales representatives,
the results should be generalized with caution. Second, as with any ex-
ploratory model, more research is required to test the generalizability
and robustness of the relationships. Furthermore, all antecedent vari-
ables were collected at one point in time. A test of the model using lon-
gitudinal data will be necessary to substantiate the causal assumptions
suggested in the proposed model.
When interpreting the magnitude of the relationships reported in this
study, it should be noted that the independent variables have not been
corrected for restriction of range. For example, evidence for range re-
striction is found with GMA, where the standard deviation is 5.90 in this
study compared to the average of 7.80 reported for over 3,000 sales rep-
resentatives in the test manual (Wnderlic, 1983). As this suggests, the
correlations reported in this study may understate the actual relations
because our independent variables may vary less than in the reference
population.
This study examined an area of research not previously investigated.
Although it is an initial attempt to identi9 a model of the processes that
are associated with involuntary turnover due to a reduction in force, we
believe the results are important for several reasons. First, in support
of past research, performance was significantly related to involuntary
turnover decisions. Second, and more importantly, two individual dif-
ference variables, tenure and gender, and one affective measure, job in-
volvement, were also directly associated with involuntary turnover, even
after controlling for performance. This suggests that dismissal decisions
made during a reduction in force incorporate variables other than per-
formance. Finally, two variables known to relate to performance, GMA
and conscientiousness,were found to be indirectly correlated with invol-
untary turnover, through the performance measures.
MURRAY R. BARRICK ET AL. 533

In conclusion, researchers and practitioners have long been con-


cerned with whom to hire (i.e., selection). For many reasons including
those dealing with long-term organizational effectiveness, researchers
and practitioners should also be concerned with who to dismiss during a
reduction in force (i.e., involuntary turnover). We clearly have much to
learn about the factors that influence these decisions, as indicated by the
dearth of published research on workforce reductions. One approach
for enhancing our understanding of this process is to model the determi-
nants of these decision processes. The exploratory model examined in
this study provides a foundation for future research in this area.

REFERENCES

Abelson MA. (1987). Examination of avoidable and unavoidable turnover. Journal of


Applied Psychology, 72,382-386.
Bannister BD, Griffeth RW. (1986). Applying a causal analytic framework to the Mob-
ley, Horner, and Hollingsworth (1978) turnover model: A useful reexamination.
Journal of Management, 12,433-443.
Barrick MR, Mount MK. (1991). The Big Five personality dimensions and job perfor-
mance: A meta-analysis. PERSONNELPSYCHOLOGY,44,l-26.
Barrick MR, Mount MK. (1993). Autonomy as a moderator of the relationships between
the Big Five personality dimensions and job performance. Journal of Applied Psy-
chology, 78,111-118.
Barrick MR, Mount MK, Straws JF! (1993). Conscientiousness and performance of sales
representatives: A test of the mediating effects of goal setting. Journal of Applied
Psychology, 78,715-722.
Bentler PM, Bonnett DG. (1980). Significance tests and goodness of fit in the analysis of
covariance structures. Psychological Bulletin, 88,588-606.
Borman WC, White LA, Pulakos ED, Oppler SH. (1991). Models of supervisor job per-
formance ratings. Journal ofApplied Psychology, 76,863-872.
Brayfield AH, Crockett WH. (1955). Convergent anddiscriminant validity by the multitrait-
multimethod matrix. Psychological Bulletin, 52,396-424.
Brooke PP Jr, Russell DW, Price JL. (1988). Discriminant validation of measures of job
satisfaction, job involvement, and organizational commitment. Journal of Applied
Psychology, 73,139-145.
Campion MA. (1991). Meaning and measurement of turnover: Comparison of alternative
measures and recommendations for research. Journal of Applied Psychology, 76,
199-212.
Cotton JL, 'httle JM. (1986). Employee turnover: A meta-analysis and review with impli-
cations for research. Academy of Management Review, 11,55-70.
Dalessio A, Silverman WH, Schuck JR. (1986). Paths to turnover: A re-analysis and review
of existing data on the Mobley, Horner, and Hollingsworth turnover model. Human
Relations, 39,245-264.
Dalton DR, Krackhardt DM, Porter LW. (1981). Functional turnover: An empirical as-
sessment. Journal ofApplied Psychology, 66,716-721.
Digman JM. (1990). Personality structure: Emergence of the five-factor model. Annual
Review of Psychology, 41,417-440.
Gerhart B. (1990). Voluntary turnover and alternative job opportunities. Journal of Ap-
plied Psychology, 75,467-476.
534 PERSONNEL PSYCHOLOGY

Goodman P, Furcon J, Rose J. (1969). Examination of some measures of creative ability


by the multitrait-multimethod matrix. Journal ofApplied Psychology, 53,240-243.
Herzberg F, Mausner B, Peterson RO, Capwell DF. (1957). Job attitudes: Review of research
and opinion. Pittsburgh, PA Psychological Services of Pittsburgh.
Hough LM. (1992). The “Big Five” personality variables-construct confusion: Descrip-
tion versus prediction. Human Performance, 5,139-155.
Hough LM, Eaton NK, Dunnette MD, Kamp JD, McCloy RA. (1990). Criterion-related
validities of personality constructs and the effect of response distortion on those
validities. Journal of Applied Psychology, 75,581-595.
Hunter JE. (1986). Cognitive ability, cognitive aptitudes, job knowledge, and job perfor-
mance. Journal of VocationalBehavior, 29,340-362.
Hunter JE, Schmidt FL, Judiesch MK. (1990). Individual differences in output variability
as a function of job complexity. Journal ofApplied Psychology, 75,28-42.
Huselid MA, Day NE. (1991). Organizational commitment, job involvement, and turnover:
A substantive and methodological analysis. Journal of Applied Psychology, 76,380-
391.
Joreskog KG, Sorbom D. (1989). LIShlEL VII user’s reference guide. Mooresville, I N
Scientific Software.
La Rocco JM, Pugh WM, Gunderson KE. (1977). Identifymg determinants of retention
decisions. PERSONNEL PSYCHOLOGY, 30,199-215.
Lee W,Mowday RT. (1987). Voluntarily leaving an organization: An empirical investiga-
tion of Steers and Mowday’s model of turnover. Academy of Management Journal,
30,721-743.
Lodahl M, Kejner M. (1965). The definition and measurement of job involvement. Journal
of Applied Psychology, 49,24-33.
March JG, Simon HA. (1958). Organizations. New York Wiley.
Marsh HW, Balla JR, McDonald RF! (1988). Goodness-of-fit indexes in confirmatory
factor analysis: The effect of sample size. Psychological Bulletin, 103,391- 410.
Mathieu JE, Zajac DM. (1990). A review and meta-analysis of the antecedents, correlates,
and consequences of organizational commitment. Psychological Bulletin, 108,171-
194.
McEvoy GM, Cascio WE (1987). Do good or poor performers leave? A meta-analysis
of the relationship between performance and turnover. Academy of Management
Journal, 30,744-762.
Mobley WH, Griffeth RW, Hand HH, Meglino BM. (1979). Review and conceptual anal-
ysis of the employee turnover process. Psychological Bulletin, 86,493-522.
Mount MK, Barrick MR, Strauss JP. (1994). Validity of observer ratings of the Big Five
personality factors. Journal of Applied Psychology, 79,272-280.
Rabinowitz S , Hall D T (1977). Organizational research on job involvement. Psychological
Bulletin, 84, 265-288.
Rothstein HR. (1990). Interrater reliability of job performance ratings: Growth to asymp-
tote level with increasing opportunity to observe. Journal ofApp1ied Psychology, 75,
322-327.
Saal FE. (1978). Job involvement: A multivariate approach. Journal ofApplied Psychology,
63,5361.
Schmidt FL, Hunter JE. (1981). Employment testing: Old theories and new research
findings. American Psychologist, 36,1128-1 137.
Schmidt FL, Hunter JE. (1992). Causal modeling of processes determining job perfor-
mance. Current Directions in Psychological Science, I , 89-92.
Schmidt FL, Hunter JE, Outerbridge AN. (1986). Impact of job experience and ability on
job knowledge, work sample performance, and supervisory ratings of job experi-
ence. Journal of Applied Psychology, 71,432-439.
MURRAY R. BARRICK ET AL. 535

Steers RM, Mowday RT (1981). Employee turnover and postdecision accommodation


processes. In Cummings LL, Staw BM (Eds.), Research in organizational behavio6
3, (pp. 235-281). Greenwich, CT: JAI Press.
Stumpf SA, Dawley PK. (1981). Predicting voluntary and involuntary turnover using ab-
senteeism and performance indices. Academy ofManagement Journal, 24,148-163.
Wanous JP, Stumpf SA, Bedrosian H. (1979). Job survival of new employees. PERSONNEL
PSYCHOLOGY,32,651662.
Wells DL, Muchinsb PM. (1985). Performance antecedents of voluntary and involuntary
managerial performance. Journal ofApplied PJychology, 70,329-336.
Wonderlic Personnel Test Manual. (1983). Northfield, I L E.F. Wonderlic & Associates.

You might also like