PERSONNEL PSYCHOLOGY

m1,54

THE EFFECTS OF EMPLOYEE SATISFACTION,

ORGANIZATIONAL CITIZENSHIP BEHAVIOR, AND
TURNOVER ON ORGANIZATIONAL EFFECTIVENESS:
A UNIT-LEVEL, LONGITUDINAL STUDY

DANIEL J. KOYS
Department of Management
DePaul University

This study addresses the issue of whether positive employee attitudes

and behaviors influence business outcomes or whether positive busi-
ness outcomes influence positive employee attitudes and behaviors.
We hypothesize that employee satisfaction, organizational citizenship
behavior, and employee turnover influence profitability and customer
satisfaction. Data were gathered from the units of a regional restau-
rant chain via employee surveys, manager surveys, customer surveys,
and organizational records. Cross-lagged regression analyses show that
employee attitudes and behaviors at Time 1 are related to organiza-
tional effectiveness at Time 2. Additional cross-lagged regression anal-
yses show no significant relationship between organizational effective-
ness at Time 1 and the employee attitudes and behaviors at Time 2.
These results add to the evidence that HR outcomes influence busi-
ness outcomes, rather than the other way around.

The management literature has been accumulating evidence that hu-

man resource outcomes are related to business outcomes. However,
most of the empirical studies have been conducted in the manufactur-
ing sector and most of them have used cross-sectional designs. This ar-
ticle presents evidence from the service sector and uses a longitudinal
design. The link between Human Resource (HR) outcomes and busi-
ness outcomes discovered in previous cross-sectional studies may have
occurred either because (a) good firm performance allowed companies
to engage in progressive HR strategies or (b) progressive HR strategies
encouraged good firm performance. A longitudinal design can help us
better understand the relationship between human resource outcomes
and business outcomes.
Human resource management and organizational behavior theories
suggest that the proper use of people enhances organizational effec-
tiveness. Most of the empirical studies testing this proposition have as-

Correspondence and requests for reprints should be addressed to Daniel J. Koys,De-

partment of Management, DePauI University, 1 E. Jackson Blvd., Chicago, IL 60604;
dkoys@wppost.depaul.edu.

COPYRIGHT 0 2001 PERSONNEL PSYCHOLOGY, INC.

101
102 PERSONNELPSYCHOLOGY

sumed a causal chain that says management practices (e.g., effective se-
lection) influence HR outcomes (e.g., employee performance and em-
ployee retention), which in turn influence organizational outcomes (e.g.,
customer satisfaction). Some of the studies have related HR activities
to organizational outcomes, leaving the intermediate step as a “black
box.” For example, some studies have shown significantrelationships be-
tween HR strategies (e.g., high performance and high commitment work
systems) and organizational performance (Arthur, 1994; Huselid, 1995;
nui, Pearce, Porter, & Tripoli, 1997). Other studies have shown rela-
tionships between specific HR activities (e.g., training, competitive pay,
advancement opportunities) and organizational effectiveness (Delaney
& Huselid, 1996; Ulrich, Halbrook, Meder, Stuchlik, & Thorpe, 1991).
Most of these studies used cross-sectionaldesigns. Longitudinal designs
can provide better evidence of causality.
Ryan, Schmit, and Johnson (1996) performed a longitudinal study in
multiple branches of a financial service organization. In each of 2 suc-
cessive years, their regression analyses showed that unit-level employee
satisfaction was related to unit-level customer satisfaction, employee
turnover, and a 60-day delinquency rate. This was as they expected. Un-
expectedly, their structural equation model showed that customer satis-
faction at Time 1 was related to employee satisfaction at Time 2, but
employee satisfaction at Time 1 was not related to customer satisfaction
at Time 2. Thus, they suggested that customer satisfaction causes em-
ployee satisfaction.

Employee Attitudes

Why should there be a relationship between unit-level employee job

satisfaction and organizational effectiveness? After all, the empirical ev-
idence has shown only a weak relationship (T = .17) between individual-
level employee satisfaction and individual-level performance (Iaffal-
dano & Muchinsky, 1985). Ryan et al. (1996) argued that organizational
performance is not simply a sum of individuals’performances; therefore
it may be influenced by factors other than those affecting individual-level
performance. One of these factors may be “shared values.” If a unit’s
employees share positive attitudes, they should have norms of coopera-
tion and collaboration, which in turn enhance unit productivity (Ryan et
al., 1996).
Schneider’s work has supported this proposition. His attraction-
selection-attrition model (Schneider, Goldstein, & Smith, 1995) said
that over time, employees in an organization become more homoge-
neous in disposition and develop shared attitudes. He argued that aggre-
gated employee satisfaction should be correlated with aggregated cus-
 DANIEL J. KOYS 103

tomer satisfaction because employees in boundary-spanning roles are

in positions to be aware of and responsive to organizational and cus-
tomer goals (Schneider & Bowen, 1985). More recently, he conducted
a longitudinal study about the relationship of employee attitudes and
customer attitudes (Schneider, White, & Paul, 1998). Using structural
equation modeling, he found that employee perceptions of a “climate
for service” lead to subsequent customer perceptions of service quality.
However, his cross-lagged analysis indicated that there was a recipro-
cal effect between employee perceptions of a “climate for service” and
customer quality perceptions.
A more practitioner-oriented rationale for the influence of aggre-
gated employee attitudes on organizational effectiveness was presented
in the “service profit chain” (Heskett, Sasser, & Schlesinger, 1997).
The model proposed that work force capability, satisfaction, and loy-
alty would lead to customers’ perceptions of value. That value percep-
tion would lead to customer satisfaction and loyalty, which would lead to
profits and growth. Heskett et al. (1997) presented cross-sectional evi-
dence that employees’ perceptions of their capabilities,satisfaction, and
length-of-service were correlated with customer satisfaction. Using an
“employee-customer-profit” chain, researchers at Sears analyzed ag-
gregated data from 800 stores. They found that employees’ attitudes
towards their company and their jobs lead to positive employee behav-
iors towards customers (Rucci, Kirn, & Quinn, 1998). Such a chain was
also apparent in an industry-wide study of restaurants that concluded
that restaurant managers need to do a better job of being an “employer
of choice” if they are to satisfy customers and produce financial results
(Foodservice Research Forum, 1997).

Employee Behaviors

Employee attitudes cannot influence organizationaleffectiveness on

their own; employees must also behave appropriately. Two employee
behaviors that are important to many managers are job performance
and retention. This study addresses that aspect of performance behavior
known as organizational citizenship and it addresses employee turnover.
We use Organ’s (1988) five categories of organizational citizenship
behavior (OCB). Conscientiousness means that employees carry out in-
role behaviors (i.e., individual task performance) well beyond the mini-
mum required levels. Altruism implies that they give help to others. Civic
virtue suggests that employees responsibly participate in the political life
of the organization. Sportsmanship indicates that people do not com-
plain, but have positive attitudes. Courtesy means that they treat others
with respect.
104 PERSONNELPSYCHOLOGY

There are several reasons why these behaviors can influence orga-
nizational performance. Based on social exchange theory (Blau, 1964),
aggregate citizenship behaviors would improve group performance be-
cause they help people work together (Organ, 1988; Podsakoff,Ahearne,
& MacKensie, 1997). Employees who help each other would not have to
go to supervisors for help, leaving the supervisors free to do more impor-
tant things. Organizational citizenship behavior would also help coordi-
nate activities among team members and across groups (Podsakoff et al.,
1997). For example, courteous people would inform each other about
nonroutine demands, allowing them to take steps to mitigate problems.
The empirical evidence supports this theory. In one cross-sectional
study, civicvirtue and sportsmanshipwere positively correlated with unit
performance (Podsakoff & MacKenzie, 1994). A meta-analysis showed
that citizenshipbehaviors correlated with job satisfaction, perceived fair-
ness, organizational commitment, and leader supportiveness (Organ &
Ryan, 1995). Almost all of the studies in that meta-analysis used cross-
sectional designs.
Organizational citizenship behavior can also contribute to customer
satisfaction (Morrison, 1995). Conscientious employees would go be-
yond customer expectations. Altruistic workers would help internal and
external customers. Those exhibiting civic virtue would make sugges-
tions to improve quality and customer satisfaction. Sportsmanship and
courtesy would create a positive climate among employees that spills
over to customers. In another cross-sectional study, civic virtue, sports-
manship, and altruism were positively correlated with financial results
and customer satisfaction (Walz & Niehoff, 1996).
Employee retention can influence organizational effectiveness be-
cause more experienced employees would have greater knowledge of
organizational and customer goals (Schneider & Bowen, 1985). Costs
would be lower because a low turnover rate means less hiring and train-
ing activities. Empirical studies have shown that employee turnover does
have a negative correlation with organizational effectiveness. For exam-
ple, a study at Sears showed that as voluntary turnover decreased, finan-
cial performance (i.e., return on controllable assets) increased (Ulrich
et al., 1991). Ostroff (1992) reported negative relationships between
high school teacher turnover and student academic achievement, stu-
dents’ satisfaction, administrative performance, and the percent of stu-
dents who dropped out of high school.
Our study looks at three HR outcomes: employee satisfaction, orga-
nizational citizenship behavior, and employee turnover. We investigate
the relationship between unit-level measures of those HR outcomes in
one year and unit-level measures profitability and customer satisfaction
in the following year. Specifically,we test the following hypotheses:
 DANIEL J. KOYS 105

Hypothesis 1: There is a significant relationship between Year 1’s unit-

level employee satisfaction (positive), organizational citizenship behavior
(positive), and employee turnover (negative) and Year 2’s unit-level prof-
itability.
Hypothesis 2: There is a significant relationship between Year 1’s unit-
level employee satisfaction (positive), organizational citizenship behavior
(positive), and employee turnover (negative) and Year 2’s unit-level cus-
tomer satisfaction.

The second hypothesis goes against the results reported by Ryan et

al. (1996). We expect that our results will differ from Ryan et al.’s (1996)
results because our investigation differs from theirs in three ways. First,
their study focused on attitudes. Our study examines an attitude (em-
ployee satisfaction) and two behaviors (OCB and turnover). We sus-
pect that employee behaviors have a more direct impact on organiza-
tional effectivenessthan do employee attitudes, especially when the con-
cept of organizational effectiveness includes profitability as well as cus-
tomer attitudes. Second, the Ryan et al. (1996) study used employee
turnover as a measure of organizational effectiveness; we use turnover
as an employee behavior and thus a predictor of organizational effec-
tiveness. High turnover means that the workforce is inexperienced; that
inexperience should have a negative effect on business outcomes. Third,
Ryan et al. (1996) studied employees who handled all customer transac-
tions by phone or mail; we study employeeswho meet customers face-to-
face. This may enhance the effect of employee attitudes and behaviors
on business outcomes.
A longitudinal design can help us see whether HR outcomes influ-
ence business outcomes or if business outcomes influence HR outcomes.
Most of the literature on employee attitudes and behaviors implies the
former. Thus, we propose that the relationship between Year 1’s HR
outcomes and Year 2’s business outcomes will be stronger than the rela-
tionship between Year 1’s business outcomes and Year 2’s HR outcomes.

Methods

Our research was conducted in a regional restaurant chain. We col-

lected useable data from 28 stores in that chain over a 2-year period.
(This is a small sample size in comparison to most individual-level stud-
ies. However, we suggest that our use of longitudinal, unit-level data will
still contribute to the HR strategy literature.) To avoid common method
bias, all variables were measured via different methods. We used cross-
lagged correlation and regression analyses (using a l-year time lag, as
did Ryan et al., 1996) to test our hypotheses.
106 PERSONNEL PSYCHOLOGY

HR Outcomes

Employee satisfaction was measured via a survey of hourly employ-

ees. Organizational citizenship behavior was measured via a survey of
the employees’ managers. In Year 1, 774 hourly employees (average
of 28 per unit) and 64 managers (average of 2 per unit) responded to
the surveys. In Year 2, 693 hourly employees (average of 25) and 79
managers (average of 3) responded. For Year 1 and Year 2, respec-
tively, respondents were full-time (61% & 62%) and part-time (39% &
38%); females (62% & 60%) and males (38% & 40%); front-of-the-
house workers (76% & 78%) and back-of-the-house workers (24% &
22%). This is representative of the employee population, with a 50%
response rate.
The hourly employees’ satisfaction was measured via a 4-item scale
based on a restaurant industry study (Foodservice Research Forum,
1997). Example items are, “If I were to choose a restaurant career, it
would be with this company,” and “I would recommend this company to
my friends as a place to work” (1 = strongfy disagree, 7 = strongly agree).
The coefficient alphas are .86 for both Year 1and Year 2. The responses
were aggregated to the unit level. This aggregation was justified empir-
ically given that the within-group interrater reliability scores for each of
the unitswere acceptable (James, Demaree, & Wolf, 1994). The average
rWgfor the 28 units was .84 for Year 1 and .78 for Year 2.
The managers responded to five items measuring their employees’
organizational citizenship behavior on a 7-point scale (1 = strong& dis-
agree, 7 = strongly agree). The items were chosen based on the literature
(Organ, 1988; Podsakoff et al., 1997; Podsakoff & MacKenzie, 1994;
Walz & Niehoff, 1996). We included one item for each dimension of
organizational citizenship behavior: conscientiousness (“the employees
work to exceed each guest’s expectations”), altruism (“I can count on my
coworkers when I need help”), civicvirtue (“the employee team feels re-
sponsible for our success”), sportsmanship (“the people I work with have
a ‘can do’ attitude”), and courtesy (“the people here treat each other
with respect”). The items evoked aggregate judgments to keep with Or-
gan’s statement that organizational citizenship behavior is a concept that
“in the aggregate,makes for a more effective organization” (Organ, 1988,
p. 6; emphasis in the original). The mean score of the five items was
used as the measure of OCB. The internal consistency of this scale was
acceptable; the coefficient alphas were .86 for Year 1and .85 for Year 2.
Internal consistency may or may not be a virtue in this context. On
the one hand, we need a reliable measure. On the other hand, the five di-
mensions may be too diverse to justify a single concept of organizational
 DANIEL J. KOYS 107

citizenship behavior. Therefore, we looked at several articles on orga-

nizational citizenship behavior to determine the diversity of the dimen-
sions (MacKenzie, Podsakoff, & Fetter, 1991; Moorman, 1991; Moor-
man, Niehoff, & Organ, 1993; Niehoff & Moorman, 1993; Organ &
Konovsky, 1989; Podsakoff & MacKenzie, 1994; Podsakoff et al., 1997).
Although the intercorrelations of the five dimensions ranged from .15
to 36, the average correlation was .56. We decided that .56 was high
enough to justify a single measure of organizational citizenship behav-
ior.
Each business unit under study had a general manager and one or
more assistant managers, so we aggregated the managers’ OCB ratings
to the unit level. This aggregation was justified both theoretically and
empirically. Theoretically, the aggregate organizational citizenship be-
haviors should be apparent to all the managers in a unit (Organ, 1988).
Empirically, the within-group (i.e., within-restaurant) interrater relia-
bility scores were acceptable (James et al., 1984); the average rWgfor
the 28 restaurants was .89 for Year 1 and .95 for Year 2.
Turnover rates came from company records. Each unit’s rate was
a 12-month, rolling average (number of hourly employee separations
divided by number of hourly employees). The turnover rate used here
was the rate as of the time we collected the other data.

Organization Effectiveness

Customer satisfaction was measured by a survey conducted in 24

units. Surveys were distributed in the restaurants at predetermined
times by the restaurant host/hostess. Customers were asked to fill out
a survey and deposit it in a special box on their way out of the restau-
rant. In return for doing so, the customers were given a coupon for a
free appetizer on their next visit to one of the chain’s restaurants.
We collected 5,565 customer responses for Year 1 (an average of 232
per unit) and 4,338 responses for Year 2 (an average of 182 per unit).
The respondents’ demographics were the same for both years: male
(55%) and female (45%); White (93%), African American (4%), and
other races (3%); and primarily between 25 and 54 years old. The cus-
tomer satisfaction score was the mean of four items rated on a
7-point scale (1 = strongly disagree, 7 = strong& agree). Coefficient al-
phas for Year l and Year 2 were .81 and 36. Example items were, “This
is a great place to have a meal” and “This is one of the best casual dining
restaurants around.” Scale scores were aggregated to the unit level. The
mean of the units’ within-group interrater reliability coefficients ( r w g )
was .95 for Year 1 and .94 for Year 2.
108 PERSONNEL PSYCHOLOGY

Our two profitability measures came from company records. Top

management used “profits after controllable expenses” as a major in-
dicator of unit profitability. This was because many of the things that
influenced profits were outside the influence of the unit manager. The
measure helped the company control for such things as corporate ad-
vertising charges and taxes. The size of the unit (which is outside of the
manager’s control) also influenced profitability. To control for that, the
company used a second measure called “profits after controllable ex-
penses as a percent of sales.” This was simply “profits after controllable
expenses” divided by “total sales.” Using operating profit as a percent-
age of sales is often used in the marketing literature to take into account
differences in overhead costs (Burleson, 1997).

Results

Table 1 presents descriptive statistics and the correlation matrix.

Coefficients representing the correlation between an HR outcome in
Year 1 and a business outcome in Year 2 are shown in bold face. The
average of these correlation coefficients is .33 ( r 2 = .11). Coefficients
representing the correlation between a business outcome in Year 1 and
an HR outcome in Year 2 are underlined. The average of these correla-
tion coefficients is .18 ( r 2 = .03). These bivariate analyses indicate that
over three times more variance in organizational effectiveness in Year 2
is explained by HR outcomes in Year 1, relative to the other way around.
Table 2 shows the results of the regression analyses testing our hy-
potheses; that is, the three HR outcomes in Year 1 predict organizational
effectiveness in Year 2. After correction for shrinkage, the average R2
for the equations is .21 (the level of statistical significance ranges from
.01 to .08). Table 3 presents the results of regression analyses where
Year 1’s organizational effectiveness scores attempt to predict Year 2’s
HR outcomes. After correction for shrinkage, the average R2for these
equations is .02 (the level of statistical significance ranges from .19 to
.65).

Discussion

Our data provide support for the proposition that HR outcomes in-
fluence organizational effectiveness(rather than the other way around).
Our first set of regression analyses shows that Year 1’s HR outcomes ac-
count for 14% to 31% of the variance in Year 2’s organizational effec-
tiveness. A second set of regression analyses shows that Year 1’s orga-
nizational effectiveness accounts for 0 to 7% of the variance in Year 2’s
HR outcomes.
 TABLE 1
Descr@tive Statistics and Correlation Matrixn
N M S D 1 2 3 4 5 6 7 8 9 10 11
1. Employee satisfaction Year 1 28 5.51 .52
2. Employee satisfaction Year 2 28 5.41 .43 .30t
3. OCB Year 1 28 5.75 .62 .47** .32'
4. OCB Year 2 28 5.67 .60 .19 .61** .43"
5. Employeeturnover%Yearl 28 105 54.2 -.14 -.17 .05 -.04
6. Employeeturnover %Year2 28 86 38.3 -.18 .13 -.14 -.02 .27t F9
7. Profitas%ofsalesYear 1 28 26.8 4.97 .37* 2 .39* -22 .OO -3
8. Profitas%ofsalesYear2 28 26.7 4.22 35* .43** .46** . 5 S * -.20 -.28t .76**
9. Profit Year 1 28 19Ok 92k .10 -B .21 -0s .10 - a t .78'* .65** Q
10. Profit Year 2 28 259k 9k 37t .22 A** .39* -.22 -.24 .81** .94** .70** m
11. CustomersatisfactionYear1 24 6.30 .11 .49** &* .06 .29t -.lo -E .26 .44* .21 .33t
12. Customer satisfaction Year 2 24 6.28 .13 .61** .09 .ll -.21 -32t .08 .20 .12 .02 .lo .35t
"Bold coefficientsrepresent the correlation between an HR outcome in Year I and a business outcome in Year 2. Underlined coefficients represent
the correlation between a business outcome in Year 1and an HR outcome in Year 2.
1-tailedtests: tp < .10 'p < .05 *'p < .01
 110 PERSONNEL PSYCHOLOGY

TABLE 2
HR Outcomes (Year 1) Predicting OrganizationalEffectiveness (Year 2)

Customer
Profit, Profit as % of satisfaction,
Year za sales, Year Za Year 2‘
Employee satisfaction,Year 1 (p) .06 .15 .62**
OCB, Year 1 (0) .41* .38t -.14
Employee turnover, Year 1 (p) -.19 -.16 -.13
RZ .24 .26 .40
R2corrected for shrinkage .14 .17 .31
F 2.501 2.79t 4.51..
~

O n = 28
b n = 24
tp < .10 * p < .05 *‘p < .01

TABLE 3
OrganizationalEffectiveness (Year 1) PredictingHR Outcomes (Year 2)

Employee Employee
satisfaction, OCB, turnover,
Year 2” Year Za Year 2”
Profit, Year 1 (0) -.13 .04 -.25
Customer satisfaction, Year 1 (8) .39 .29 .03
RZ .15 .09 .06
R~ Corrected for shrinkage .07 .oo -.03
F 1.80 1.01 .67
Profit as % of sales, Year 1 (0) .07 .14 -.21
Customer satisfaction,Year 1 (0) .34 .26 .03
RZ .14 .I1 .04
RZ Corrected for shrinkage .05 .02 .oo
F 1.64 1.24 .45

Our results show some support for the hypothesis that Year 1’s unit-
level employee satisfaction, organizational citizenship behavior, and
turnover predict Year 2’s unit-level profitability. However, only orga-
nizational citizenship behavior has a significant beta weight.
There is stronger support for the hypothesis that Year 1’s unit-level
employee satisfaction,organizational citizenship behavior, and turnover
predict Year 2’s unit-level customer satisfaction. However, only em-
ployee satisfaction has a significant beta weight. Although this implies
that employee satisfaction influences customer satisfaction, customer
satisfaction may still affect employee satisfaction (as Ryan et al., 1996,
reported). There may be a reciprocal relationship between employee
satisfaction and customer satisfaction similar to the reciprocal relation-
ship between employee perceptions of a “climate for service” and cus-
 DANIEL J. KOYS 111

tomer perceptions of service quality (Schneider et al., 1998). We can

only conclude that the direction of causality between employee satisfac-
tion and organizational effectiveness is still an open question needing
further research.
It may be that Year 2’s organizational results were affected by
Year 1’s organizational results. In post hoc analyses, we performed
three additional regression analyses to test this possibility. In the first,
we entered Year 1’s profitability into the equation predicting Year 2’s
profitability (R2= .49, p < .Ol), then we entered the three HR out-
comes (R2change = .17, p < .05; total R2adjusted for shrinkage = .60,
p < .01). In the second regression, we entered Year 1’s profits as a per-
centage of sales into the equation predicting Year 2’s profits as a per-
centage of sales (R2= .57, p < .Ol), then we entered the three HR
outcomes (R2change = .07, p = .25 [nonsignificant]; total R2adjusted
for shrinkage = .58, p < .01). In the third equation, we entered Year
1’s customer satisfaction into the equation predicting Year 2’s customer
satisfaction (R2= .12, p = .12 [nonsignificant]), then we entered the
three HR outcomes (R2change = .35, p < .05; total R2 adjusted for
shrinkage = .35, p < .05). Thus, even after controlling for Year 1’s or-
ganizational effectiveness,HR outcomes have an impact on profitability
and customer satisfaction.
An examination of Thble 1 reveals that there was a 105% employee
turnover rate in Year 1 of our study. It appears that no one who was
around in Year 1 was around to contribute to organizational effec-
tiveness in Year 2. However, demographic data showed that some of
Year 2’s hourly employees (38.4%) worked at the firm for more than
12 months; some hourly positions must have turned over two or more
times during the previous year. This is typical in the restaurant industry
where there are many temporary employees (e.g., students working as
servers). In addition, many of the unit managers and assistant managers
were with their company for some time (in Year 2,74.5% were with the
firm for more than 12 months). These core managers may have per-
formed certain HR and supervisory practices that produced somewhat
stable levels of organizational citizenship behavior and employee satis-
faction over the years. Thus, even with an unstable hourly workforce,
Year 1’s HR outcomes can influence Year 2’s organizational effective-
ness.
Our bivariate analysis shows a slight negative correlation between
Year 1’s hourly employee turnover rate and Year 2’s customer satisfac-
tion. However, this relationship is not strong enough to hold up in our
multivariate analyses. We think that the negative relationship would
have been stronger if the company had measured voluntary turnover.
112 PERSONNEL PSYCHOLOGY

Because it measured total hourly turnover, the negative effect of volun-

tary turnover is weakened by the positive effect of involuntary turnover
(e.g., termination for poor performance).
Our cross-lagged regression analyses indicate that profitability and
customer satisfaction are influenced by different HR outcomes. Orga-
nizational citizenship behavior had an impact on profitability but not on
customer satisfaction. Employee satisfaction had an impact on customer
satisfaction but not on profitability. Still, we do not think that profit-
oriented managers should ignore employee satisfaction. Recall that the
“service profit chain” (Heskett et al., 1997) says employee satisfaction
contributes to customer satisfaction, which contributes to profitability.
OCB theory and research also say that employee satisfaction is important
because it leads to organizational citizenship behavior (Organ & Ryan,
1995). Longitudinal research is needed on all the relationships among
employee satisfaction, OCB, customer satisfaction, and profitability.
Our results must be taken with some caution because the study has
limitations. First, we conducted our research in an industry where the
majority of the work force has direct contact with customers. This means
that our results may not generalize to low-customer-contact firms. Sec-
ond, we had a small sample size. We trust that the results of other unit-
level research soon will be combined with our results in a meta-analysis
to move the field of strategic HR management forward. Third, we were
not able to separate voluntary from involuntary turnover. The effects of
turnover may be more pronounced if future researchers are able to do
so. Fourth, although longitudinal, the study is still correlational. Better
evidence of causality requires a field experiment.
In summary, our results suggest that HR outcomes influence orga-
nizational effectiveness, rather than the other way around. Specifically,
they imply that organizational citizenship behavior influences profitabil-
ity and employee satisfaction influences customer satisfaction.

REFERENCES

Arthur JB. (1994). Effects of human resource systems on manufacturing performance and
turnover. Academy of Management Journal, 37,670687.
Blau I? (1964). Exchange andpower in social life. New York: Wiley.
BurlesonA. (1997). The serviceprofit chain: An investigationofthe linkages between employ-
ees, Customers, and organizational profitability. Unpublished doctoral dissertation,
DePaul University,Chicago, IL.
Delaney JT, Huselid MA. (1996). The impact of human resource management practices
on perceptions of organizational performance. Academy of Management Journal,
39,949-969.
Foodservice Research Forum. (1997). Industry of choice. Chicago: The Educational
Foundation, National Restaurant Association.
 DANIEL J. KOYS 113

Heskett JL, Sasser WE, Schlesinger LA. (1997). The service profit chain: How leading
companies link profit and growth to loyalty, satisfaction, and value. New York Free
Press.
Huselid MA. (1995). The impact of human resource management practices on turnover,
productivity, and corporate financial performance. Academy of Management Jour-
nal, 38,635-672.
Iaffaldano MT, Muchinsky PM. (1985). Job satisfaction and job performance: A meta-
analysis. Psychological Bulletin, 97,251-273.
James LR, Demaree RG, Wolf G. (1984). Estimating within-group interrater reliability
with and without response bias. Journal ofApplied Psychology, 69,85-98.
MacKenzie SB, Podsakoff PM, Fetter R. (1991).Organizational citizenship behavior and
objective productivity as determinants of managerial evaluations of salespersons’
performance. Organizational Behavior and Human Decision Processes, 50,123-150.
Moorman RH. (1991).Relationshipbetween organizationaljustice and organizational cit-
izenship behaviors: Do fairness perceptions influence employee citizenship? Jou-
nal of Applied Psychology, 76,845455.
Moorman RH, Niehoff BP, Organ DW (1993). Tteating employees fairly and organi-
zational citizenship behavior: Sorting the effects of job satisfaction, organization
commitment, and procedural justice. Employee Responsibilities and Rights Journal,
6,209-225.
Morrison EW. (1995). Organizationalcitizenship behavior as a critical link between HRM
practices and service quality. Human Resource Management, 35,493-512.
Neihoff BP, MoormanRH. (1993).Justice as a mediator of the relationship between meth-
ods of monitoring and organizational citizenship behavior. Academy of Manage-
ment Journal, 36,527-556.
Organ DW. (1988).Organizational citizenship behavior. Lexington, MA. Lexington Books.
Organ DW, Konovsky M. (1989).Cognitiveversus affective determinants of organizational
citizenship behavior. Journal of Applied Psychology, 74,157-164.
Organ DW, Ryan K. (1995). Ameta-analytic review of attitudinal and dispositional predic-
tors of organizationalcitizenship behavior. PERSONNEL PSYCHOLOGY,48,775402.
Ostroff C. (1992). The relationship between satisfaction,attitudes, and performance: An
organizational-level analysis. Journal of Applied Psychology, 77,963-974.
Podsakoff PM, Ahearne M, MacKenzie SB. (1997). Organizational citizenship behavior
and the quantity and quality of work group performance. Journal of Applied Ay-
chology, 82,262-270.
Podsakoff PM, MacKenzie SB. (1994).Organizationalcitizenship behaviors and sales unit
effectiveness. Journal of Marketing Research, 31,351-363.
Rucci AJ, Kirn SP,Quinn RT (1998,JanuaryFebruary). The employee-customer-profit
chain at Sears. Harvard Business Review, 76,82-97.
Ryan AM, Schmit MJ, Johnson R. (1996). Attitudes and effectiveness: Examining rela-
tions at an organizational level. PERSONNEL PSYCHOLOGY,49,853-882.
Schneider B, Bowen DE. (1985).Employee and customer perceptions of service in banks:
Replication and extension. Journal ofApplied Psychology, 70,423-433.
Schneider B, Goldstein HM, Smith DB. (1995). The ASA framework An update. PER-
SONNELPSYCHOLOGY, 48,747-774.
Schneider B, White SS,Paul MC. (1998). Linking service climate and customer perception
of service qualily: %st of a causal model. Journal of Applied Psycholow, 83, 150-
163.
l3ui AS, Pearce JL, Porter LW, Ttipoli AM. (1997).Alternative approaches to the employee-
organization relationship: Does investment in employees pay off? Academy of
Management Joumal, 40,1089-1121.
114 PERSONNEL PSYCHOLOGY

Ulrich D, Halbrook R, Meder D, Stuchlik M, T h o r p S. (1991). Employee and customer

attachment: Synergies for competitive advantage. Human Resource Planning 14,
89-103.
Walz SM, Niehoff BF! (1996). Organizational citizenship behaviors and their effect on
organizational effectiveness in limited-menu restaurants. Academy of Management
Best Paper Proceedings, 307-31 1 .