Job Satisfaction and the Good Soldier: The Relationship between Affect and Employee

"Citizenship"
Author(s): Thomas S. Bateman and Dennis W. Organ
Source: The Academy of Management Journal, Vol. 26, No. 4 (Dec., 1983), pp. 587-595
Published by: Academy of Management
Stable URL: http://www.jstor.org/stable/255908 .
Accessed: 02/03/2011 06:04

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=aom. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Academy of Management is collaborating with JSTOR to digitize, preserve and extend access to The Academy
of Management Journal.

http://www.jstor.org
cAcademy of Management Journal
1983, Vol. 26, No. 4, 587-595.

Job Satisfaction and

Good the Soldier:
The Relationship
Between Affect and
Employee "Citizenship i 1
THOMAS S. BATEMAN
Texas A&M University
DENNIS W. ORGAN
Indiana University

A measureof a wide array of employee activitieson

thejob was completedby employees'supervisorsat two
points in time; employeesreportedtheir own job satis-
faction via theJob DescriptiveIndex.Implicationsof rela-
tionshipsmuchhigherthantypicallyfound in thejob sat-
isfaction-performanceliteratureare discussed.
Apparentlythe dusthas settledoverwhatonce was a controversialissue:
the satisfaction-performancelinkage.Organizationalpsychologists(Lawler
& Porter, 1967) generallyendorsethe view that any covariancebetween
job satisfactionandjob performanceemergesonly whensatisfactionresults
from performance-contingentrewards. Any notion that satisfaction
"causes")performanceis regardedas naive folk wisdom, not supportable
by the empiricalrecord.
Organ(1977) has cautionedthat such a position might prematurelyre-
ject somethingof value in lay psychologythat endorsedthe satisfaction-
causes-performanceproposition.He suggestedthat a clue to the possible
reconcilabilitybetweenthe phenomenologyof countlesspractitionersand
the noncorroboratingempiricalrecordmight lie in the meaningof "per-
formance.'"Defined narrowlyas quantityof output or quality of crafts-
manship-as perhapsoperationalizedin most of the formal researchad-
dressedto this issue-performance does not consistentlyor appreciably
follow from satisfactionin a direct functionalrelationship.But there are

1An earlier version of this paper was presented at the 42nd National Academy of Management meet-
ings, New York, 1982.

587
588 Academy of Management Journal December

other conceptionsof performancethat often may be more salient to the

practitioner.
Katzand Kahn(1966)havenoted the manyoccasionsin whichorganiza-
tional functioningdependson supra-rolebehavior-behavior that cannot
be prescribedor requiredin advancefor a given job. These behaviorsin-
cludeany of thosegestures(oftentakenfor granted)thatlubricatethe social
machineryof the organizationbut that do not directlyinherein the usual
notion of task performance.Examplesthat come to mindinclude:helping
co-workerswith a job relatedproblem;acceptingorders without a fuss;
toleratingtemporaryimpositionswithout complaint;helping to keep the
workareacleananduncluttered;makingtimelyand constructivestatements
about the work unit or its head to outsiders;promotinga work climate
thatis tolerableand minimizesthe distractionscreatedby interpersonal con-
flict; and protectingand conservingorganizationalresources.For lack of
a betterterm, the presentauthorsshall referto these acts as "citizenship"
behaviors.
Supervisorspresumablyvaluesuchbehaviors,in partbecausethey make
theirown jobs easierand free theirown time and energyfor more substan-
tive tasks. One suspectsthat they value it all the morebecausethey cannot
"require"such supra-role,citizenshipbehaviors,exceptperhapsto some
minimallyacceptableor enforceablestandards.
Thereare two distinctconceptualbases for thinkingthat such behavior
wouldbe influencedby job satisfaction(or, at least, the affective state un-
derlyingjob attitudes).Firstof all, social exchangetheory (Adams, 1965;
Blau, 1964)predictsthat, givencertainconditions,peopleseekto reciprocate
those who benefit them. To the extent that a person's satisfactionresults
fromthe efforts of organizationalofficials and such efforts areinterpreted
as volitionaland nonmanipulativein intent, the personwill seek to recip-
rocate those efforts. The personmay not have the ability or opportunity
to reciprocatewith greaterworkoutputor creativesolutionsto workprob-
lems. Citizenshipbehaviorsof the sort describedabove are more likely to
be under the person's control and thus more likely to be a salient mode
of reciprocation.
A second basis for predictingthis relationshipderivesfrom a series of
social psychologicalexperiments(Rosenhan,Underwood,& Moore, 1974)
whichstronglysupportthe contentionthatprosocialgesturesaremost likely
to occurwhena personexperiencesa generalizedmood state characterized
by positiveaffect (Clark& Isen, 1982).To the extentthatjob satisfaction,
as conventionallymeasured,reflectsthis positiveaffectivestate, it is likely
that more satisfiedpersonsdisplaymore of the prosocial, citizenshipbe-
haviors. As Rosenhanet al. (1974)phrasedit, positive affect tends to de-
creasethe psychologicaldistancebetweenself and others, and positive af-
fect tends to generalizefrom whatevercaused it to other stimuli(notably
persons)in the temporaland social context.
In sum, it is predictedthat there is a causal connectionbetweenprior
overallsatisfactionand subsequentdisplayof a host of citizenshipbehaviors
(as specifiedbelow). Moreover,a strong connectionfor satisfactionwith
1983 Bateman and Organ 589

supervisionin particularis predicted.The rationalefor this predictionis

that the immediatesupervisorrepresentsthe most directsourceof variance
in events that arouse a felt need to reciprocateor that influence positive
affect. Also, citizenshipbehaviors,it is suspected,are more often seen as
"for" (i.e., benefiting)the supervisorthan for any othersinglepersonrep-
resentingthe organization.

Method
Subjectsand Procedure

Data were obtainedat two separatetimes from a sample of employees

in a major midwesternstate university.A wide variety of jobs in non-
academic,administrativedepartmentswere surveyed,includingdata pro-
grammers,loan collectors,studentcounselors,fund-raisers,accountants,
and other professional,supervisory,and technical(nonclerical)positions.
The initialdata collectionwas performedvia questionnaireadministra-
tion at the host departments.One of the researchersmet with all members
of each department(groupsrangedin size from 6 to 16) who completed
the measuresof job satisfaction.Concurrently,departmentalsuperiorswho
had frequentcontactwiththe subjectswerecompletinga scale, ratingeach
subject'srecentbehaviorat work. Subjectswere not awarethat they were
beingevaluated;theyalso wereassuredof the anonymityof theirresponses.
Questionnaireswere identifiedwith a code numberknown only to the re-
searcherand the subject. The second administrationemployedthe same
procedureand took placefrom five to sevenweeksafterthe firstcollection.
Completedata (includingsubjects' and supervisors'ratings)were col-
lected for 82 employeesat time 1. For variousreasons,severalindividuals
were unavailableat time 2. Data were collected at both times for a total
sampleof 77, whichincludedsubjectsand their superiors.Of the sample,
29 (38 percent)were female and 48 (62 percent)were male. Their average
departmentaltenure was 6.8 years.
Measures

Citizenshipbehaviorwas measuredvia the responsesof each subject's

immediatesuperioron 30 7-pointitems. The items tappeda varietyof be-
haviorssuch as compliance,altruism,dependability,housecleaning,com-
plaints,waste,cooperation,criticismof and arguingwith others,and punc-
tuality.
A preliminarymeetingwith a small group of managersprovidedinfor-
mationabout any ambiguitiesor irrelevancieson the originalitems. Inter-
nal reliabilitycoefficientsfor the finalscalewerec = .92 at time 1 anda = .94
at time 2, and the test-retestreliabilitywas .80. Althoughindividualitems
on the scale appearto be behaviorallydistinct, the psychometricproper-
ties of the scale indicate that it provided a composite criterionthat is
590 Academy of Management Journal December

indicativeof an employee'soverallevaluationby the supervisorin terms

of behavioraleffectiveness.
Job satisfactionwas measuredwith the Job DescriptiveIndex(JDI) de-
velopedby Smith, Kendall,and Hulin (1969). The JDI containsfive scales
pertainingto employees' satisfaction with work, pay, promotions, co-
workers,and supervision.The instrumentconsistsof 72 items-1 8 in each
of the work, supervision,and co-workerssubscales,and 9 in the pay and
promotionssubscales.All of the scalesarereportedto havecorrectedsplit-
half internalconsistencycoefficientsexceeding.80, andtest-retestreliabilities
averaging.57 (Schneider& Dachler, 1978).

DataAnalysis

Subsequentto descriptiveand correlationalstatisticalanalyses,the issue

of causalprioritieswas testedvia cross-laggedregressionanalysis(Rogosa,
1980).Basiccontextualinformationis affordedby six correlationsbetween
the two variablesx andy-static correlationsfrom time 1 and time 2 (rX1y,
and rX2Y2), two autocorrelations(rX112 and ryly2)that indicatethe test-retest
reliabilitiesor stabilitiesof the two variables,and two cross-laggedcorre-
lations (rx1Y2and ry,,2)between the time 1 value of one variableand the
subsequenttime 2 value of the other variable.
Causalanalysisand tests of spuriousnesstypicallyare conductedby sta-
tisticalcorrectionsand comparisonsfor the magnitudesof the cross-lagged
correlations(Kenny,1975).However,recentcriticismof cross-laggedcor-
relation(Rogosa, 1980)alternativelysuggeststhe use of structuralregres-
sion modelsfor the analysisof longitudinalpaneldata. This approachwas
appliedin the presentstudy. For a pair of variables,x and y, the causal
influencefromx to y is representedby the regressionparameterof the path
from x at time 1 to y at time 2. In like manner,the causalinfluencefrom
y to x is representedby the regressionparameterof the path from a prior
y to a subsequentx. Thus, where:

X2= 30 + 3IxI + -y2Y1+u, and

Y2=-yo+ 2XI+ yiYI + V

the parameterst3 and 'y representthe time-laggedinfluenceof a variable

on itself and I2 and 'y2representthe time-laggedcausal effects betweenx
andy. Underthe usual assumptionsgoverningregressionanalysis(linear-
ity, additivity,etc.), a nonzerovalue of a relevantparameteris indicative
of a significantcausal effect.
Staticand cross-laggedanalyseswereperformedto test the relationships
betweencitizenshipbehaviorand each of the separatefacets of satisfac-
tion, as well as with a summatedmeasureof overall satisfaction.
1983 Batemanand Organ 591

Table 1
Means and StandardDeviations at Time 1 and Time 2
ti t2
Variable X SD X SD
Citizenshipbehavior 142.1 27.3 141.1 27.0
Overallsatisfaction 143.5 28.3 138.0 29.4
Work 36.3 8.7 34.1 8.4
Co-workers 42.4 9.9 41.6 8.4
Supervision 42.0 10.6 41.2 10.1
Promotions 10.4 7.6 8.9 7.4
Pay 13.0 5.8 12.3 6.1

Table 2
Static Correlations(t1 and t2)
Between Facets of Job Satisfaction and CitizenshipBehaviors
Job Satisfaction
Work Co-worker Supervision Promotions Pay Overall
ti t2 ti t2 ti t2 ti t2 ti t2 ti t2

Citizenship
behaviors .09 .19* .24* .18 .46** .36* .37** .40** .16 .25* .41** .41**
*p<.05
**p <.01

Results
Table 1 presentsdescriptivestatisticsfor the studyvariablesat both times
surveyed.Table 2 then shows the static correlationsbetween citizenship
behaviorsand the specificfacets of satisfaction.Thereare indicationsthat
each dimensionof job satisfactionmay be positivelyrelatedto citizenship
behavior, with two facets-supervision and promotional opportunity-
reliablymore importantthan pay, co-workers,and the work itself.
Subsequentcross-laggedanalysiswas conductedbetweencitizenshipbe-
haviorand eachmeasureof satisfaction.The patternsof relationshipswere
virtuallythe same in all instances.As a summaryexample, Figure 1 dis-
plays the cross-laggedanalysissurroundingthe relationshipbetweenjob
relatedcitizenshipbehaviorsand overalljob satisfaction. The test-retest
reliabilitiesare fairly high for both variables.Both static correlationsare
positiveand stronglysignificantand are particularlysubstantialwhencom-
pared to most previousstudies of the satisfaction-performancerelation-
ship. Inspectionof the cross-laggedstatistics,however,failedto discriminate
a single causal direction.Both raw correlationsare highly significant,the
relativemagnitudesarein the predicteddirection,and the predictedcausal
correlationis slightly greaterthan the two static correlations.However,
the two cross-laggedcorrelationsare fairlycomparableto one another.Fur-
ther, both path coefficients(shown in parentheses)are positive, yet much
smallerthan the correlations;they also are comparablein magnitudeto
one another and are statisticallyinsignificant.
592 Academyof ManagementJournal December

Figure 1
Cross-Lagged Relationships Between Overall Satisfaction
and Citizenship Behaviorsa
.71*
SATISFACTION, SATISFACTION2
43* (.12)b
.41* .41*
.39* (.11)
BEHAVIOR, .80* BEHAVIOR2

aCross-laggedpatternsof relationshipsbetweencitizenshipbehaviorand specificfacetsof job sat-

isfactionrevealessentiallythe sameresultsas overallsatisfaction.Thesedata are availablefrom the
first authoron request.
bPathcoefficientsare in parentheses.
*p<.001

Thus, evidencefor the predicteddirectionof causalitywas not obtained.

However,the resultsdo reliablysuggestthatjob satisfactionis indeedstrong-
ly and positivelyrelatedto a "citizenship"dimensionof role performance.
Discussion
The statisticalrelationshipsobtainedhere betweengeneraljob satisfac-
tion and the aggregatemeasureof citizenshipbehaviorsare considerably
strongerthan those typicallyreportedbetween satisfactionand "perfor-
mance."Of course,the samplesize (77)limitsthe confidencethatone could
attach to comparisonsbetweencorrelations.However, when the 95 per-
cent confidenceintervalsare computedfor the correlationsinvolvingover-
all satisfactionand satisfactionwith supervision,the lower limits of these
intervals(.15-.26) still exceedthe r of .14 from Vrooni's(1964)review(the
upper limits of the confidence intervalsrange from .54 to .62).
The strongerrelationshipfound heremay be becausethe citizenshipbe-
haviors of interesthere generallyrepresentactions more under the voli-
tional control of workersthan conventionalproductivitymeasures.Pro-
social gesturesare less likely to be constrainedby other situationalforces,
and they pose very little in the way of ability requirements.
Consider,for example, Smith's (1977) study, which found that job at-
tendanceon a given day was predictedby satisfactionmuchmore strongly
in a location hit by a severewinterstormthan in a differentlocation expe-
riencingclementweather.In extremelybad weather,absenceis somewhat
moredefensiblethan usual;to attemptto show up for workbecomesmore
a matterof intent. This attenuationof the situationforce "requiring"at-
tendanceallows more varianceof the behaviorin question and increases
the likelihoodthat suchvariancecan be attributedto "internal"(i.e., atti-
tudinal or dispositional)forces. It might be added that the act of strug-
glingthroughbad weatherto reportto workrepresentsmoreof a prosocial,
citizenshipgesturethan does attendanceon other days, and it probablyis
more likely to be valued and appreciatedby responsibleofficials.
1983 Bateman and Organ 593

A secondreasonfor the unaccustomedstrengthof the attitude-behavior

relationshipfound here probablyderivesfrom the broad arrayof citizen-
ship gesturessampledby the behavioralmeasure.As Fisher(1980)has ap-
propriatelypointed out: attitude-behaviorlinkages are attenuatedif the
specificitiesof the constructsare not matched;narrowlydeflned attitude
measuresaremorepredictiveof theircorrespondingspecificbehaviorsthan
of generalmeasuresof behavior;and global measuresof affect or mood
predictaggregativemeasuresof broadlysampledbehaviorsbetterthandoes
any single behavior.Too often the interpretationof resultsfrom studies
of satisfactionand performancehas not takeninto accountthe imbalance
betweengeneralizedattitudemeasuresand a measureof a very narrowly
conceivedfacet of job behavior(e.g., quantityand/or qualityof task per-
formance).The resultsof this study demonstratethat generalpatterns of
employeebehavior,ratherthan a single-actproductivitycriterion,arepre-
dictable from a generalmeasureof satisfaction.
However,the strengthof the staticrelationshipsfound heredid not pass
the test of causalinference;the cross-laggeddifferentialswerenot reliable.
Whywas a causalconnectionbetweensatisfactionand citizenshipbehaviors
not supported?One possible explanationconcernsthe time interval(six
weeks)chosen. Conceivably,this was too shortan intervalfor a functional
relationshipto manifestitself. The currentstateof conceptualand empirical
work in this area offered no clearcues as to what the appropriateinterval
mightbe. Researchin anotherarea(leadership) indicatesthe appropriateness
of relativelyshort time lags (Sims Szilagyi, 1979), suggestingthat the
&
six-weekintervalis not inherentlyweak methodologically.However, the
test-retestcorrelationof the time 1 and time 2 behavioralmeasureswas
.80, approachingthe limitset by the internalreliabilityof the measureitself.
Withsuchstability,then, behaviorat time 2 was best predictedby behavior
at time 1, with initial job satisfaction unable to add significantly to
prediction.
Kenny(1975)statesthat failureto find a significantcross-laggeddiffer-
ential means that one cannot reject the null hypothesisof spuriousness,
that is, cannotrule out the possibilitythat the variablesare correlatedbe-
cause of theirdependenceon a common antecedentvariable.In this case,
any such"spuriousness" does not originatefromcommon-method variance.
The citizenshipmeasurecamefrom a respondent(the supervisor)indepen-
dentof thosewho reportedtheirsatisfaction;thusthe "same-source"prob-
lem does not apply.
Supervisorybehavior-especially that generallydescribedas "suppor-
tive" or "considerate"- could representa commoncause. It has beenwell
establishedthat consideratesupervisionaffects job attitudes, even when
attitudesare indexedby nonself-reportmeasures-for example,turnover
(Fleishman,1973). Perhapssupportivesupervisionalso elicits citizenship
behaviors,independentlyof its effects on job satisfaction.
Personality,or somedimensionof stableindividualdifferences,also could
account for the covariancebetweenmood and citizenship.The literature
594 Academyof ManagementJournal December

does not offer eithera consistentor a preciseclue as to what dimension(s)

might be reflectedhere, but one could well imaginethat certaintempera-
ments(e.g., low vs. highneuroticism)determinecharacteristic affectivetone
as well as one's generalizedorientationto prosocial gestures.
The strengthof the relationshipsfound in the presentstudy-compared,
say, to the .14 reportedby Vroom (1964)to be the mean satisfaction/job
performancecorrelation-coupled with the problemof stabilitiesthat may
have precludedcausal demonstrations,suggestsfuture work with longer
timelags(andlargersamplesizes,whichwillenhancethe powerof the causal
tests and enable a more discriminatoryfactor analysison the citizenship
scale).Additionally,ratingsof behaviorcould be obtainedfrom observers
otherthanthe supervisor,includingco-workers.If the relationshipis more
convincinglyshownto be spurious,it wouldseemthat a morefully mapped
causal model is needed. Such a model should incorporatesome identifi-
able environmentaland personalityvariablesin orderto accountadequate-
ly for the connectionsbetweensatisfactionand citizenshipbehaviors.More-
over, one could betterascertainwhetherconventionalmeasuresof job at-
titudes, as used here, adequatelyoperationalizethe "felt need to recipro-
cate" invoked by exchangetheory as a predictorof behavior.
The resolutionof these issues would appearto be worthwhilein view
of the practicalimplicationsthat wouldensue.If both citizenshipbehaviors
and affect reflect enduringdispositions,then fairly drasticalterationsof
the workenvironmentwouldbe neededto effect muchchangein thesevari-
ables. If they resultfrom situation-specificcauses, then perhapsorganiza-
tional officials can addressthem in a proactiveposture.And if citizenship
behaviorsare shown to be causal effects of satisfaction,it would appear
that job satisfactionis, in fact, more importantto organizationsthanjust
in its often-mentionedrelationshipswith absenteeismand turnover.Citi-
zenshipbehaviorsrepresenta multiple-observation criterion,whichnot only
is of potentialvalueto managersand organizationsbut also can contribute
to less restrictiveorganizationalresearch(by expandingthe numberof out-
come variablesthat are studied)and therebyprovide for broaderunder-
standingof employee behavior.
References
Adams, J. A. Inequityin social exchange.In L. Berkowitz(Ed.), Advancesin experimentalsocial
psychology(Vol. 2). New York:AcademicPress, 1965, 267-299.
Blau, P. Exchangeand power in social life. New York: Wiley, Inc., 1964.
Clark,M. S., & Isen, A. M. Towardunderstanding the relationshipbetweenfeelingstatesand social
behavior.In A. H. Hastorf& A. M. Isen (Eds.), Cognitivesocialpsychology.New York:Elsevier
North Holland, 1982, 73-108.
Fisher,C. On the dubiouswisdomof expectingsatisfactionto correlatewith performance.Academy
of ManagementReview, 1980, 5, 607-612.
Fleishman,E. A. Twentyyearsof considerationandstructure.In E. A. Fleishman&J. G. Hunt(Eds.),
Currentdevelopmentsin thestudyof leadership.Carbondale,Ill.: SouthernIllinoisUniversityPress,
1973, 1-37.
1983 Batemanand Organ 595

Katz, D., & Kahn, R. L. Thesocial psychologyof organizations.New York: Wiley, 1966.
Kenny,D. A. Cross-laggedpanelcorrelation:A test for spuriousness.PsychologicalBulletin, 1975,
82, 887-903.
Lawler,E. E., III, & Porter, L. W. The effect of performanceon job satisfaction.IndustrialRela-
tions, 1967, 7, 20-28.
Organ,D. W. A reappraisaland reinterpretation hypothesis.
of the satisfaction-causes-performance
Academyof ManagementReview, 1977, 2, 46-53.
Rogosa, D. A critiqueof cross-laggedcorrelations.PsychologicalBulletin, 1980, 88, 245-258.
Rosenhan,D. L., Underwood,B., & Moore,B. Affectmoderatesself-gratification andaltruism.Journal
of Personalityand Social Psychology, 1974, 30, 546-552.
Schneider,B., & Dachler,H. P. A note on the stabilityof the Job DescriptiveIndex.Journalof Ap-
plied Psychology, 1978, 63, 650-653.
Sims, H. P., & Szilagyi,A. D. Time-lags;Inleaderrewardresearch.Journalof AppliedPsychology,
1979, 64, 66-71.
Smith,F. J. Workattitudesas predictorsof attendanceon a specificday. Journalof AppliedPsychol-
ogy, 1977, 62, 16-19.
Smith, P. C., Kendall,L. M., & Hulin, C. L. Themeasurementof satisfactionin workand retire-
ment. Chicago, Ill.: Rand McNally, 1969.
Vroom, V. R. Workand motivation.New York: Wiley, 1964.

ThomasS. Batemanis AssistantProfessorof Organizational

Be-
havior,Departmentof Management,Collegeof BusinessAd-
ministration,TexasA&M University.
DennisW.Organis Professorof OrganizationalBehavior,Grad-
uate School of BusinessAdministration,IndianaUniversity.