CPARorca.cf.ac.uk/107557/1/cpar.isopraxis.2018.pdfCPAR Review Attention has been given to the notion...

Isopraxis Leadership: Leader Confidence, Managerial Strategy, and Organizational Performance

Kenneth J. Meier *, Laurence J. O’Toole, Jr. †

* Texas A&M University, USA; Cardiff University, UK† The University of Georgia

Volume 8 Issue 1, June 2017

Chinese Public Administration ReviewCPAR

Attentionhasbeengiventothenotionthatorganizational leaders’expressedconfidenceandoptimismregardingtheirorganizations’performancecanaffectthatperformancebyincreasingthemotivationand/orself-efficacyofsubordinates.Thisidea,apartofvariousleadershiptheories,wecall“isopraxisleadership.”Thispaperexaminesthelogicoftheclaim,reviewsandcritiquesextantevidence,developsameasureofleaderconfidence(thestartingpointforisopraxisleadership),undertakesinitialvalidationofthemeasure,andthentestsforthelinkbetweenleaderconfidenceandperformanceamongseveralhundredpublicorganizations.Leaderconfidenceisfoundtobelargelyunrelatedtoperformance;someevidenceindicatesthatitcanhelponlyforthoseorganizationsthatarealreadydoingwellorhavemoreresourcesthanaverage–thatis,whereitisleastneeded.

Leadership is a central concern of publicmanagement(Morse,Buss,andKinghorn2007;Trottier, Van Wort, and Wang 2008; Rainey

2009;Fernández,Cho, andPerry2010;HansenandVilladsen 2010; Getha-Taylor et al. 2011) and maybe all the more important under today’s conditionsof constrained resources and heightened concernsaboutgovernmentalperformance.How leaders lead,however,isperceivedtobeahighlycomplexprocessthat contains numerous factors, contingencies, andconsiderations. The contributions of scholarshipto clarifying the role and impact of leadership onorganizationalperformanceatthesametimemustbecharacterizedasdisappointing.Inanextensivereviewof the literature some time agoYukl (1989, p. 253)concluded that “several thousand empirical studieshavebeenconductedonleadershipeffectiveness,butmostoftheresultsarecontradictoryandinconclusive.”Inthethreedecadessincethatassessment,studiesofleadershiphavecontinuedtoproliferate,butconsensuson key questions remains elusive (Fernández 2005;VanSlykeandAlexander2006;Yukl2010).

Still,progressonthisissuewouldseemtobeacentralconcern for the field. As Van Wart has observed,

“Although many types of leadership in the publicsector have been discussed extensively, such asleadershipbythoseinpolicypositionsandworkingincommunitysettings,administrativeleadershipwithinorganizationshasreceivedscantattentionandwouldbenefit from a research agenda linking explicit andwell-articulatedmodelswithconcretedatainpublic-sectorsettings”(2003,p.214).

Working toward this objective is the aim of thispaper. The present analysis steps back from thecomplexity of existing work on leadership to focusononeaspectgenerallycommon tomost leadershiptheories,particularlytransformationalleadershipandcharismaticleadership:theroleofleaderoptimismandconfidence regarding the leader’s own organization,or a leader’s efficacy regarding that organization.Leaders’ sense of organizational efficacy is part ofnumerousapproachesto leadership;while it isnotastand-alone,self-contained,andcomprehensivetheoryofleadership,itispartofsomanyleadershiptheoriesthat it merits study in its own right. This abridgedleadership “theory” that is focused on confidentleaders and high standards we term “isopraxisleadership.”The basic idea is that the leader serves

Address correspondence to the author Dr. Meier at [email protected]. View this article at cpar.net

Meier & O’Toole • 47

as a role model and motivator for employees; thatis,subordinatesadopttheconfidence,standards,andbehaviors(mimeticisomorphism)oftheleader.Asaconsequence, so the theoretical arguments indicate,others in the organization improve their individualand collective performance. As with other notionsofhowandwhyleadershipmightmakeadifference,of course, the key is whether isopraxis leadershipactuallyworksor,moredirectly,doesleaderoptimismresultinbetterorganizationalperformance?Cansuchahypothesisbevalidated?

Theanalysisproceedsinfourparts.First,theexistingleadership literature is briefly reviewed to outlinethe key aspects of isopraxis leadership and the roleof leader optimism. Second, a measure of leaderconfidence/optimism is developed, one based on aleader’s perceptions of organizational performancerelative to actual performance. Third, this measureis validated by comparing it to key measures ofmanagement,particularlymanagerialstrategy.Fourth,theheartofthepaperexamineswhetherorganizationsledbyconfidentleadershavehigherperformancethanthoseledbyothers.Finally,thearticleconcludeswithadiscussionoftheimplicationsofthesefindingsfortheoriesofpublicmanagementandforthetrainingofpublicmanagers.

TRANSFORMATIONAL AND CHARISMATIC LEADERSHIP

Theideaofisopraxisleadershipandleaderoptimismis not a stand-alone theory, but rather a conceptthat has its roots in the extensive literatures onleadership, particularly transformational leadershipandcharismaticleadership(Burns1978;Yukl2010).Ratherthanattemptingthedauntingtaskofafulltestofallaspectsofleadershipwiththerequisitemyriadvariables and multiple interactions, this analysisseekstoisolateasinglecomponent,commontomanytheories,andprovideasystematicassessmentofthatcomponent.Inlinkingtransformationalleadershiptocharismatic leadership, Rainey (2009, p. 332) notessome commonalities, particularly that both notionshold that leaders “express confidence in followers,set high expectations for them, andempower them”

(seealsoYukl1989,p.260;2010;Shamir,HouseandArthur1993,p.578).OnecanfindsuchexpressionsforleaderconfidenceinCongerandKanugno’s(1998)theoryofcharismaticleadership,Bass’(1998)theoryof transformational leadership, and Warren Bennis’general theories of leadership (Bennis and Nanus1985,p.59).

This setof leadership ideas is related tobutdistinctfrom the work of psychologist Albert Bandura(1977)andhisconceptofself-efficacyaspartofhisdevelopment of social learning theory (see Rainey2009,p.323).Wefirst characterize the literatureonself-efficacyandthendistinguishisopraxisleadershipfromthisgenerallineofwork.

Self-efficacy is essentially a “person’s estimate ofhis or her capacity to orchestrate performance on aspecifictask”(GistandMitchell1992,p.183).Self-efficacy is deemed important not only because itrelates to the leader’sconfidencebutalsobecause itmeansthatsuchleaders tendtosethighergoalsandcommunicate high expectations to their followers(see also the notion of “self-leadership” offered bySimsandLorenzi(1992,p.301-4).BennisandNanus(1985,p.59),whousetheterm“positiveself-regard”ratherthanself-efficacy,stateabouttheirsampleof90leaders,“Likeathletes,theyregularlysethighergoalsandobjectivesforthemselves.”

The link to athletics is important for two reasons.First,itgeneratesalargenumberofmetaphorsthatareoften used as symbols in leadership andmotivationseminars. Successful coaches and former athletesareoftenusedaspractical illustrations in leadershiptraining.Second,itlinkstheconcepttoanextensivebody of research on elite athletes and performancethat is presumed to be translatable to organizationalperformance.Thebasicliteratureonathletes,however,isfundamentallyflawed.Whilenumerousstudieshavedemonstrated a positive relationship between self-efficacy(orconfidence)andperformanceforathletesparticipating in team(FeltzandLirgg1998;Moritz,Feltz, Farhbach and Mack 2000; Taylor 2006) andindividualsports (Gould,WeissandWeinberg1981;HighlenandBennett1983;Taylor2006),noneofthe


48 • Chinese Public Administration Review

studiescontrolforthepastperformanceoftheathletes.So while there are positive correlations betweenself-efficacy and performance, this literature cannotdetermine if self-efficacy causes better performanceorbetterperformancecausesself-efficacy.

Bandura’s work and the empirical work on self-efficacyinathleticsandotherfieldswereintroducedspecifically to management byMarilyn Gist (1987;GistandMitchell1992).Thesubsequentapplicationsthen distinguished between individual self-efficacyandgroupself-efficacybasedonwhethertheefficacymeasure is for individuals or for the entire workgroup (that is, whether the question is asked aboutconfidence in thegrouporself-confidence,seeJungandSosik2003).Theempiricalliteratureispromisingin that self-efficacy has shown a relationship toperformance in a variety of areas ranging from lifeinsurancesales(BarlingandBeattie1983)tofacultyresearch productivity (Taylor, Locke, Lee, and Gist1984). An extensive meta-analysis of self-efficacy(covering nearly 200 empirical studies) found astrong relationshipbetweenself-efficacyanda largenumber of dependent variables that were eithermeasuresofemployeeperformanceorcouldbelinkedto performance, such as absenteeism (see Judge,Jackson, Shaw, Scott and Rich 2007). Judge et al.,however,criticizethisliteratureforfailingtocontrolforavarietyofotherindividualcharacteristicssuchasgeneralmentalability,experience,conscientiousness,etc. that are correlated with both self-efficacy andorganizational outcomes. When this is done in themeta-analysis, the impact of self-efficacy dropssignificantlyandisrarelyamongthemoreimportantdeterminantsoforganizationalperformance.Thesamequestionmightbe raisedabout leaderoptimismandconfidenceanditspossiblerelationtoorganizationalperformance.

Furthermore,whatwearetermingisopraxisleadershipisnotthesameasindividualself-efficacyonthepartofleaders,becauseitdoesnotrefertotheindividualleader’s“self”oreventoaworkgroup’ssenseofitsown efficacy, but to a leader’s optimism about andconfidence in the leader’sorganization.Most earlierattemptstodevelopaself-efficacyscalearebasedon

aconceptof the individual self --eithergeneralizedself-efficacy (e.g., Sherer et al. 1982; Chen, Gully,and Eden, 2001), or “particularized” judgmentsof individual capability for certain sorts of tasks(Pajares2006).Someoftheextantscalesseektotapmanagers’ self-efficacy regarding the leadership ofchange(PaglisandGreen,2002)--thisisalsoclearlyinapplicable.Ournotionofisopraxisleadershiprefersto how leader optimism and confidence about howthewhole organization contributes to organizationalperformance.

Whatisthemicrotheorybehindthenotionthatisopraxisleadershipanditscoreconcept,leaderoptimism,willlead tohigher levelsoforganizationalperformance?Thatis,howmighttheconfidenceandoptimismofaleaderaffecttheperformanceoftheentireorganizationwhichreliesontheactionsofmanyindividualsotherthantheleader?Twotheoreticalpathwaysarefoundintheliterature:(1)thatperceptionsofsuchleadershipleadtogreateridentificationwiththeorganizationandhigherlevelsofmotivationonthepartoforganizationmembers, and (2) that subordinates adopt similarattitudes(isomorphismor therole-modeleffect)andthattheheightenedsubordinatesenseofoptimismandconfidenceleadsdirectlytogreaterperformance.

In terms of motivation Bennis and Nanus (1985,p. 32) state, “What we observed was that our 90leaders induced (stemming from their own self-regard) positive other-regard in their employees.”Bass (1998)specificallycontends that theendresultof transformational leadership is that followers trustand respect the leader and are thenmotivated to domorethantheyoriginallyanticipateddoing(seealsoPillaiandWilliams2004,p.164).Somegosofarastocontrastthisleadershipstylewiththatassumedbyeconomicmodels of organization. Such leaders “gobeyond a simple performance-reward transactionby elevating their subordinates’ self-image and self-confidence and by arousing subordinates’ emotionalattachment to the leader’s espoused values andto the collective” (Javidan and Waldman 2003, p.229). Phrased differently, Boal and Bryson (1988,p.19)arguethat“theprimaryimpactofcharismaticleadershipis throughfacilitationof thecreationofa



new or different world that is phenomenologicallyvalidtothefollower.”

Shamir,House,andArthur(1993,p.578)raiseakeypointwhen theystress that“newleadership theoriesemphasize symbolic leader behavior, visionary andinspirationalmessages . . . display of confidence inselfand followers . . .”Shamiretal. (1993,p.580)specificallynote that thisstyleof leadershipstressesmotivation“byfaith,ratherthanrewards.”Thepointis worth emphasizing in thatmuch of the literaturehasamessianictoneandthatleadershiptrainingoftentakesonanexplicitly religiousovertone, sometimeswithbiblicalquotations.(AnunsystematicindicatorofthispointisthatfourofthetenpaidlinksforaGooglesearch of “leadership training” were also amongthe tenpaid links foundunder “Christian leadershiptraining.”)

Thisrole-modeleffectmightwellbemoreprevalentin some types of organizations than others. Yukl(1989, p. 277) argues that when organizations arefacedwithvaguegoalsandsignificantconstraints,andwhenmanagementdoesnothavemanydirectwaystoexert influenceonorganizationalperformance,“it isallthemoreimportanttomaintaintheimpressionthatorganizationalleadersknowwhattheyaredoingandaremakinggoodprogresstowardattainingorganizationalobjectives.”Suchasituationislikelyverycommoninpublicorganizations; thus, it isnosurprisethatParkandRainey(2008)findthatwhenfederalemployeesperceivetransformationalleadership,theyalsoexpresshigherlevelsofcommitmenttotheorganizationandgreaterjobsatisfaction.

In terms of impacts on performance, Bandura’swork plays a key role. Bandura’s social learningtheory holds that people “learn bywatching others,throughmodelingandvicariouslearning”(seeRainey2009, p. 323). When leaders express confidencein the organization’s performance, they are alsoexpressingconfidenceintheorganization’smembers.Subordinateswhoobservethisconfidenceincorporateit into theirownviewsof theorganizationand theirrole in that organization.An extensive literature inpsychologydealswithwhatiscalledthe“Pygmalion

effect,” the idea that if role models establish highexpectations for individuals, those individuals willperform better. The Pygmalion effect was firstdemonstratedwith school children inRosenthal andJacobson (1968) and was applied to organizationalsettingsbyEden(1984).EdenlinkedthePygmalioneffect to expectancy theory whereby the higher anindividual’s level of expectancy, the greater themotivationtoperform.“Conveyinghighexpectationsbyacredible,authoritativesourceevidentlymotivatessubordinatestomobilizemoreoftheirownresourcestoperformwell”(Eden1984,p.66).

Recent research shows that the Pygmalion effectworks not just for individuals but also for groupsof individuals. Since Eden (1990) first proposedgroup-level efficacy as a concept and linked it tothe performance of Israeli army platoons, severalother studies have also been conductedwith groupsor teamsas theunitsofanalysis.Ameta-analysisof53 group perceived-efficacy studies found a strongrelationship between self-perceived group efficacyand group performance (Gully, Incalcaterra, Joshi,Beaubien2002).Aswiththeindividuallevelstudies,however, theseinvestigationsdonotcontrolforpastperformanceorsimilarmeasuresofability.

The literature suggests, in sum, that what we callisopraxis leadership startswith leaderoptimismandconfidence in the organization’s performance, andthat leader confidence might affect organizationalperformancethroughacoupleofpathways–byeithera)establishingarolemodel,b)motivatingothers,orboth.Whethersuchatendencyisafairlystabletraitor a consciously adopted style is an open questionat this point, and one not addressed or answered inthe present study. Although the above discussionseparates motivational and role model effects, inpractice the twoare likely tobeconflated. Isopraxisleaders are unlikely to just become role modelsand not also seek tomanage the organization. Thiscombination is best summed up by Shamir, House,and Arthur (1993, p. 582): “Charismatic leadersincrease effort-accomplishment expectancies byenhancing the followers’ self-esteem by expressinghighexpectationsofthefollowersandconfidencein



thefollowers’ability tomeetsuchexpectations. . . .Bydoingso,theyenhancefollowers’perceivedself-efficacy.”Additionalresearchshowsthatcharismaticleaders then use different management practicesbecause “high expectations among supervisorscauses them to be both more supportive and moretaskorientedtowardstheirsubordinates,resultinginhighersubordinateperformance”(Eden1984,p.66).Managementpractices,thus,reinforcetherolemodeleffect.

The literature touching in one way or another onisopraxis leadership, efficacy, and organizationalperformanceisimpressive.Thereareseveralhundredstudies in a wide range of organizations that showthat the style of leadership we term isopraxis leadsto increases in subordinate efficacy. Increased self-efficacy, then, leads directly to higher levels ofperformance. The two issues are rarely discussedbut are important in terms of public management–causalityandpossiblenegativeeffects.

Thecausalityquestionhasbeenalludedtointheabovediscussion.Self-efficacyisrelatedtoperformance,butquiteclearlyhighlevelsofperformancealsoleadtohigher self-efficacy (Jung and Sosik 2003; Hannah,Avolio, Luthans, and Harms 2008) and vice versa(Campbell and Hacket 1986). Without recognizingthis reciprocal relationship, one might be lulledinto thinking that all it takes to be amore effectiveorganizationis tohaveamoreconfidentcheerleaderin charge. This is not as absurd as it sounds.Former Secretary of Education Margaret Spellingsspecificallydescribedhercontributionintheseterms.Indeed,muchofthetrainingliteratureseemstoimplythat self-efficacy is exogenous to performance andthat manipulating self-efficacy will generate betterperformance.Although a few studies recognize thereciprocal relationship between self-efficacy andorganizational performance, there are no systematicinvestigations that incorporate this relationship in a rigorousstatisticaltest.

The management literature that recognizes thereciprocal causation relies on techniques that havelong been abandoned as inadequate in other parts

of the social science literature, suchas cross-laggedcorrelations. Bandura has influenced the formerliterature to accept the notion of controlling forresidualizedpastperformancetodeterminecausality(seeHeggestadandKanfer2005,p.90).Thistechniqueinvolvesregressingpastperformanceattime1onself-efficacyattime1andusingtheseresidualsasacontrolwhen linking self-efficacy at time 2 to performanceat time 2. This technique essentially assigns all thecommonvarianceofthetwoconceptstoself-efficacyandthusbiasestheresultsinfavoroffindingthatself-efficacymatters.

Thenegativeeffectsarealsorarelydiscussedorstudied.Inastudyofstudents,VancouverandKendall(2006)actuallyfoundthatgreaterself-efficacyresultedinlesseffort,whichinturnledtolowerperformance.Intheirtheoreticalworkonmanagementandself-efficacy,GistandMitchell(1992)notedthattheemulationofleadersbyfollowerscouldleadtogroupthinkandaresultinginabilitytoreacttochange.Perhapsmostdirectlyonpoint, Romzek and Ingraham’s (2000) examinationoftheaircrashthatkilledCongressmanRonBrownfoundthattheAirForce’s“cando”orientation(verysimilartoself-efficacy)ledtotakinggreaterrisksandwasafactorcontributingtothecrash.Aninflatedviewoforganizationalperformancecanalsomeanthattheorganization is not aware of performance gaps anddoesnotaddressexistingproblems(seeDowns1967).

MEASURING LEADER CONFIDENCE

Measuring isopraxis leadershipwouldmeancomingup with ameasure of leader confidence, a measureof the role model effect, and a measure of groupefficacy,aswellasgoodindicatorsoforganizationalperformanceandtheappropriatecontrols.Thatwouldbe ahighly ambitiousproject that is notpossible inanyexistingpublicorganizationdatabase.Oureffortis more limited and focuses on measuring leaders’positiveorientationandtheirconfidentassessmentoftheir organization’s performance. Established scalesmeasuring such a concept are not available. Indeed,evenfor thedifferentnotionofself-efficacy there isnoagreementonastandardmeasure. Inconsideringthispoint, it isuseful tostartwithBandurahimself,



who has indicated that “there is no all-purposemeasureofperceivedself-efficacy.The‘onemeasurefitsall’approachusuallyhaslimitedexplanatoryandpredictivevalue”(2005,p.307). Itwouldseemthatwhat is needed here is something like self-efficacyrelatedtothetasksathand,but(1)focusednotontheindividualmanagerbuton theoverallorganization’sfunctioning, and (2) with past performance purged.This latter point is elaborated upon shortly, andwe need to include such an adjustment in ourdevelopmentofameasureofleaderconfidence,eventhough thathasnotbeendone in the large literatureonself-efficacyandrelatedconcepts.Fortheformerpoint,weneed items that taphow themanagerseesthe organization’s functioning, not the manager’sownfunctioning.Soleaderconfidencehastodowithmanagers’ performance-related confidence aboutor optimism in the organization. One scale offeredto tap public managers’ organizational confidencehas been offered by Feeney and Boardman (2011).Unfortunately,thethreeitemscomprisingtheirscaleareunrelatedor,atbest,distantlyrelatedtomanagers’assessmentoftheorganization’sperformanceintermsofactuallydeliveringoutcomes.Instead,theirmeasureincludesmanagers’assessmentoftheirorganization’sethicalstandards,the“overallqualityofwork”intheirorganization,andtheirsenseofprideinworkingfortheirorganization.Wedevelopanalternativemeasure-- one tapping managers’ assessment of multiplemeasuresoforganizationalperformanceinthepublicsectorfieldofinterest,andonewhichfiltersoutpastperformancefromthemeasure.

Suchameasureneedsconvergentvalidity,butthatisSuchameasureneedsconvergentvalidity,butthatisnotall.Giventhedifficultiesevidentinearlierresearchthatdidnottakecaretoseparateoutpastperformancefromasenseofefficacyandpositiveassessment,anappropriate measure must also have discriminantvalidity. It is just as important to establish whatisopraxisleadershipisnotandhowitdiffersfromothervariablesasitistodeterminepreciselywhatitis.Inparticular,withoutseparatingconfidenceandoptimismfromability,ameasurecanofferlittletothestudyofthis aspect of leadership in public management. Totheextent thatconfidenceandoptimismarenothing

more than a reflection of the talent and skills ofthe individual or the organization, the prescriptionto enhance confidence reduces to “improve theorganization by improving the organization.” DizzyDeanoncestated,“Itain’tbraggingifyoucanbackit up”; similarly, leader confidence needs to meansomething more than the prior performance of theorganization(BaseballAlmanac,n.d.)(ThequotewasfirstusedbybaseballpitcherDeanin1934concerninghowmanywins his brother Paul and hewouldwininthe1934season–hetoldthereporter they’dwin45.Thequote is also attributed later toMuhammadAliinregardtohisfightwithSonnyListon).Withoutthis distinction, isopraxis management suffers fromthesameflawasself-efficacyresearch insportsandprivatesectororganizations.Inshort,wemightsay“Itain’tisopraxisleadershipifyou’realreadydoingverywell.”

Ameasure of leader confidence, therefore, needs toAmeasure of leader confidence, therefore, needs topurge prior performance out of any assessments oforganizational efficacy. In the present paper, this isdonebyusing auniquedata set that combinesbothsurveys of top-level managers and an elaborateperformanceappraisalsystem.

Data for this analysis come from two sources, theAcademicExcellence IndicatorSystemof theTexasEducation Agency (TEA) and an original surveyofTexas school superintendents.The state ofTexasoperates an elaborate accountability system forTexas schools that collects information on a varietyofperformanceindicatorsaswellasdataonstudentsand finances. All data other than administrators’perceptions of performance and management styleare taken from this source for the academic years2008-2009 and 2009-2010, the year immediatelypreceding and the year immediately following thegathering of the perceptual data. For the perceptualdata, school system superintendents were surveyedviaa four-wavemail surveybetweenJuly2009andNovember2009.Theresponserateforpublicschoolsuperintendents was 58%; the survey also includedsomecharterschoolsuperintendentsforatotalof642respondents. (The survey contains responses from



595 public superintendents and 47 charter schoolsuperintendents.Theresponserateforcharterschoolsisdifficult todetermine since eachcharter school istreatedbythestateasaseparatedistrict;butiftwoormoreschoolsareoperatedbythesameorganization,theywouldhave the same superintendent. Inclusionof the charter school superintendents hadno impacton any of the results presented in this paper.) Thesample is representative of the entire populationwith no apparent biases.Therewere no statisticallysignificant differences between the respondents andthenon-respondentsintermsoftheTexasAssessmentof Knowledge and Skills (TAKS, the highly salientstatewidestandardizedexamination,explainedfurtherbelow) scores, college-bound scores, the racial andincome distribution of students, and instructionalexpenditures. Respondents did receive $480 less inperpupilrevenuethannon-respondentseventhougheducationalexpendituresweresimilar.

Themeasureofleaderconfidenceinpubliceducationthatisadoptedherebeginsbyaskingmanagerstoevaluatethe quality of their own organization’s performancecompared to similar districts. Superintendents wereasked“compared tosimilardistricts,myassessmentofour____performance is . . . .”Theywereaskedto rate their district on a five-point scale using thecategories “excellent,” “above average,” “average,”“below average,” and “inadequate.” (The phrase“similardistricts”wasused toallow themanager toadjust for context – that is, to give the organizationmore credit formore difficult tasks or less credit ifresourcesareample.Analysisnotshownindicatesthattherespondentsdidnotmakeanyadjustmentsbasedontherace,ethnicity,orpovertylevelofthestudentsorfortheresourcesofthedistrict,teacherexperience,andawidevarietyofother factors.)Threedifferentstimulus items were used: “TAKS performance,”“college-boundperformance,”and“overallqualityofeducationinthedistrict.”Suchperceptualmeasuresasthesearecommonlyusedinavarietyofmanagementsurveys,includingtheFederalHumanCapitalSurveyandtheMeritPrinciplesSurvey,amongothers.Thesemeasures are frequently used as actual performancemeasures without any attempt to determine if themeasuresareobjectivelyrelatedtoperformance.Such

an approach fails to distinguish between how wellthe organization does and themanager’s confidencein the organization’s performance (see Andrews,Boyne, andWalker 2006). In the case of these topmanagers,thereisnoquestionthattheyareinterestedinperformancemeasures(thegreatmajorityofthemindicatebysurveyresponsesthatscoringwellontheTAKS is their organization’s most important goal;college-boundperformanceisalsohighlyrankedasanorganizationalgoal)andthat theyhavereadyaccesstodetailedperformanceinformationontheirownandallotherdistrictsandareupdatedonanannualbasisby theTEA.To convert these responses into a puremeasureofconfidenceintheorganizationandthustapaportionofisopraxisleadership,actualperformanceispurgedfromthemeasures.Thatis,fortheperceivedTAKS performance measure, one can predict thatmeasurewiththedistrict’sactualTAKSperformancefor thepreviousyear (2009)via regressionand taketheresidualofthisregressionastheindicatorofself-efficacy.Positiveresidualsreflectahigherperceptionofperformancethanwouldbeexpectedfromobjectivemeasures. For the college-bound performanceindicator, a similar regression is performed byusing as the independent variable the percentage ofstudents who score above 1110 on the SAT or itsACT equivalent (equal to the top 20%nationwide),astandardthatisdefinedbythestateofTexasasanindicator of college readiness. (Regression residualsare used in a wide range of policy research fromeconomiccrises(Rattsø1999)topublicattitudes(deBoefandKellstedt2004).Theseregressionresidualsarenormallydistributed(Martinez-Iglewicztest)andmeet the assumptions required for factor analysis(Thompson2004;Yates1987).)Fortheoverallqualityof educationmeasure, bothTAKS performance andthe1110indicatorareusedasindependentvariables.Theseresidualmeasuresareuncorrelatedwithstudentcharacteristics (race, income), teacher experienceand turnover, and revenues per student (see below).They are positively correlated with superintendentexperience (+.21),weakly correlatedwith class sizeand, by definition, uncorrelated with the previousyear’s performance score. The residuals from eachof these three equations are then factor analyzed toget the common variance. The results of the factor



analysis are presented in table 1, which shows aconceptwithahighdegreeofreliability;theloadingsrangefrom.78to.89,accountingfor70%ofthetotalindividualitemvariance(theCronbach’salphais.78).The factor scores are used here as the measure ofleaderconfidence.

Table 1. Measuring Leader Confidence: The Factor Analysis

Measure LoadingResidual TAKS Performance .83Residual College-bound Performance .78Residual Overall Quality of Education .89

Eigenvalue 2.10 Percentage of the variance 70.0

Although the factor measure has face validity as ameasureofleaderconfidence,additionalinformationcan be provided concerning the concept’s convergent anddiscriminantvalidity(ZellerandCarmines1980)bycorrelatingitwithothermeasuresofmanagement,particularlymanagement strategy.This is especiallyimportant for a new measure of organizationalbehaviortoensurethatanewanddistinctelementofmanagementisbeingtapped.Inparticular,onewouldexpecttheconfidentleaderstoengageinwhatMilesandSnow(1978)term“prospecting”–thesearchfornewideasandnewstrategiestoimplementinandthroughtheorganization.Table2 shows that themeasureofisopraxisleadershipispositivelycorrelatedwithtwocommonmeasures of prospecting – agreementwithstatementsaboutbeingamongthefirstorganizationsto adopt new ideas and about continually searchingfornewopportunitiestoprovideservicestoclientele.

Thisprospectingstrategyofleadership,however,Thisprospecting strategy of leadership, however, shouldnot be taken to indicate that the confident leader isexcessively focused on the external environment ofthe organization. Further analyses reported in table2 determine that there is no relationship betweentheconfident leadermeasureandsomeother surveyitems, including a factor score of networkingwith environmental actors or with initiating those



Table 2. How Do Confident Leaders Manage?

Correlations with Other Management MeasuresMeasure Correlation ProbabilityManagerial Prospecting First to adopt new ideas .21 <.0001Search for new opportunities to provide .18 <.0001

Measure Correlation ProbabilityBufferingControl outside factors that affect district .10 .01

Try to limit external events impact .02 ns

Managerial Networking -.01 nsNetwork Initiating .01 nsManagerial Stability .15 .001Internal Management/Human Resources (Quality of teachers, principals, etc.)

.41 <.0001

Employee Stability .05 nsTeacher experience (years) -.11 .01

Delegate Authority (principal’s discretion) .11 .01

Organizational Correlates of Leader ConfidenceMeasure Correlation ProbabilitySize (enrollment) .01 nsCentralization (percent central administration) .16 <.0001

Revenue per student .04 nsInstructional expenditures per student

.05 ns

Low income students (percent) -.03 ns

African American students .03 ns

Latino students .00 nsCharter school .19 <.0001

interactions. (The networking measure is a factorscoreofthefrequencyofcontactfromdailytoneverwith eleven actors in the environment. The singlegeneral factor generates an eigenvalue of 3.48.Theinitiationmeasureissimplyacountofwhichoftheseinteractions were initiated by the superintendent.)Thisconclusiongenerallyextendstoeffortstobuffertheorganizationfromtheenvironment.Althoughtheresponsetotryingto“controloutsidefactorsthataffectthe district” has a significant positive relationship,thecorrelationisweaker;andthereisnorelationshipwith a desire to limit the impact of external events.Confident leaders aremore likely to have served intheorganizationalongertime(seeGistandMitchell1992,p.191),albeitmodestly,andareslightlymorelikely to endorse delegating authority to mid-levelmanagers.Thelatterisconsistentwiththerolemodel/isomorphicaspectsofcharismatic leadership.Totheextent that subordinates adopt the isopraxis leaderasa rolemodel, the leadercanbemoregenerous indelegatingauthority.

Thestrongestrelationshipintable2isapositiveonewith an internal management or human resourcesfactor at .45, thereby suggesting that the impact ofconfidentleadershipcouldwellbethroughtheprocessofeitherdevelopingqualityemployeesormotivatingothers in the organization. (This factor (eigenvalue2.18) combines an assessment of the quality ofteachers,thequalityofprincipal’smanagementskills,the quality of professional development programs,and agreementwith the statement that “with peopleinthisorganization,wecanmakeanyprogramwork,”andthewillingnesstorecommendasubordinateforasuperintendentpositioninanotherorganization.)Thismeasure has a strong evaluative component zeroinginon theperceivedqualityof subordinatemanagersand line personnel. This pattern clearly reflects aleader’sconfidenceinorganizationalpersonnel.Ifthisrelationshipreflectsmotivation,however,onewouldexpect a strong relationshipwith employee stability(100minustheturnoverpercent);butthatcorrelationisstatisticallyinsignificant.Morestrikingly,confidentleadershipisassociatedwithlessteacherexperience,arelationshipthatisinconsistentwiththeideathatthismanagementapproachworksviamotivation.

The bottom portion of Table 2 examines a set oforganizational-level correlates with the measureof leader confidence. Gist and Mitchell (1992, p.194) hypothesize, for example, that self-efficacy isassociatedwithtaskdifficulty.Theshortstoryisthatwith the exception of administrative centralizationandcharterschools,therearenoorganizational-levelcorrelates of the leadership confidence, includingsize (enrollment), resources (revenues per studentand instructional expenditures per student), or taskdifficulty (the percentage of black, Latino and low-incomestudents).Confidentleadersarefoundinmorecentralizedorganizations,inthiscasemeasuredasthepercentageofthetotalnumberofemployeeslocatedincentraladministration.Thisbehavioralcentralizationshouldbecontrastedwiththeabovecorrelationwithdelegation.

Thecharterschoolscorrelationisquiteconsistentwithwhatisknownaboutleadershipandcharterschools.Charter schools are started by entrepreneurs whoperceive that they can educate children better thanexistingpublic schools.Charter school leaders needtorecruitparentsandstudents,findadditionalsourcesof funds, and build a coherent educational system.Generally, this process relies on identification witha specific philosophy of education (e.g.,Montessorior “back to basics”). Because charters often havelower financial resources per student than publicschools,theyneedtorelyonmorenormativeformsofmotivation,somethingthatconfidentisopraxisleadersaremorelikelytoprovide. Thissetofvalidationefforts, in short, indicates thatthe measure of leader confidence is a reasonableone. The article next turns to the performance-related hypothesis: does leader confidence boostorganizationalperformance?

THE IMPACT OF LEADER CONFIDENCE ON PERFORMANCE

Todetermine if leader confidence has an impact onthe performance of public organizations, analysisfocusesonthreekeyoutputindicatorsforTexaspubliceducation–performanceonthestatewideexam,daily



student attendance, and ameasure of college-boundperformance.

The Texas Assessment of Knowledge and Skills(TAKS)isastatewide,high-stakestestthatallstudentsin grades3 through11must take.The exams coverestablishedlearningcriteriaforindividualgradesandinsomecasesforspecificclasses.Themeasureisthepercentageofstudentswhopassedallexamsthattheytookforthe2009-2010academicyear(theyearafterthe survey).This score is a central part of the stateaccountabilitysystem–schoolsreceiveperformanceratingsbasedonthesetestscores.Theresultsarehighlysalientbothtothepublic(thereleaseoftheresultsisfrontpagenews)andthesuperintendents,whooftenhaveperformanceclausesintheircontracts.ThemeanTAKSpassratewas75withastandarddeviationof12andanormaldistribution.

Student attendance is a basic-level performanceindicator, but one crucial for theorganization, sincestate funding is based on the number of studentsattendingclasses.Themeanis95.8%withastandarddeviationof1.6,withamodestnegativeskew.

Thefinalmeasure is thepercentageofstudentswhoscoreabove1110ontheSAToritsACTequivalent,anofficialstatedefinitionofa“collegeready”student.ForTexasdistricts themean is21.3witha standarddeviation of 13.7; the distribution is truncated atthe low end and has a relatively long positive tailrepresentingthestate’swealthydistricts.

Leaderconfidencewillbeusedtopredictthesethreeperformance indicators for the year following thesurvey that tapped managers’ perceptual data. Inaddition, standard practice in education productionfunctionsistocontrolforresourcesandtaskdifficulty(FinnandAchilles1999;Hanushek1996;HedgesandGreenwald1996;Nye,Hedges,andKonstanopoulus1999;Wenglinsky1997).Fourmeasuresofresourceswillbe included–averageteachersalary,classsize,teacherstability(100minustheturnoverrate),andperpupil instructionalspending.Threemeasuresof taskdifficultyfocusongroupsofstudentswithadditionaleducational needs that are likely to be reflected in

overalltestscores–thepercentagesofblack,Latino,andlow-incomestudents.

Beforeproceedingtotheanalysis,onemightquestionifthetopmanagerofaschooldistrictwouldbeableto influence these performance indicators given thatthere are a large number of factors that influencestudentperformance.Theeducationliteraturestronglyendorses the belief that superintendent leadershipmatters both from the perspective of policymakers(Hess 1999) and on the basis of managerial theory(Ouchi and Segal 2003). A systematic quantitativestudyof themanagementofschooldistricts, in fact,attributed20percentofthecross-districtvariationtotop-levelmanagement(seeO’TooleandMeier2011).

Table 3 presents two regressions involving TAKSscores. The first replicates the type of analysisgenerallydoneintheself-efficacyliterature–thatis,itdoesnotcontrolforthepreviouslevelofperformance(thelattervariablewouldbethefunctionalequivalentof talent levels in regard to athletes).This equationshows a positive relationship for leader confidence(significant at the .05 level with a one-tailed test).Because theconfidencemeasure is a factor score, ithas an effective range of -3 to +3.Thismeans thatthemaximum effect size of confident leadership onTAKS performance in this under-specified equationcouldbeashighas3.7percentagepoints.Whilethatmightnotatfirstglanceappeartobealargeimpact,suchanincreasewouldbehighlyvaluedintheTexashigh-stakes performance system. Unfortunately, asthe second regression demonstrates, this impact isspurious.When one controls for prior performance,the impact of leader confidence is effectively zero.Confident leadersappear tobeechoingperformanceratherthaninfluencingit.

Wedonotdiscussinanydetailheretherelationshipsforthecontrolvariables.Allaresignificantand,exceptforinstructionalspending,areinthecorrectdirection.Care should be taken in any implication that morefundsspenton instructionareassociatedwith lowerperformance.Thisapparentrelationshipistheimpactof instructional funds after controlling for teachersalaryandclasssize,twofactorsthatmakeupthebulk





Table 3. Leader Confidence and TAKS Performance

Dependent Variable = Pass rate on state standardized test (TAKS) in 2010Without Past Performance With Past Performance

Variable Slope t-score Slope t-scoreLeader Confidence .616 1.84 -.023 0.122009 TAKS Pass Rate — — .797 37.24% Black Students -.069 2.15 -.003 0.12% Latino Students -.049 2.39 .009 0.66% Low Income Students -.239 8.32 -.048 2.86Teachers Salary (000s) .522 4.61 .125 1.97Class Size -.501 2.69 -.107 1.03Teacher Stability .239 6.23 .017 0.77Instruction Spending (000) -1.204 3.17 -.208 0.98

Standard error 8.21 4.55R-square .49 8.44F 72.81 364.94N 622 622

Table 4. Leader Confidence and Student Attendance

Dependent Variable = Average daily attendance in 2010Without Past Performance With Past Performance

Variable Slope t-score Slope t-scoreLeader Confidence .007 0.11 -.031 1.512009 Attendance Rate — — .758 66.23% Black Students -.001 0.10 -.001 0.46% Latino Students -.006 1.68 .000 0.12% Low Income Students -.002 0.49 -.005 2.64Teachers Salary (000s) .069 3.52 .009 1.24Class Size -.244 7.55 -.023 1.91Teacher Stability .031 4.83 .003 1.14Instruction Spending (000) -1.46 2.20 -.001 0.04

Standard error 1.43 0.50R-square .24 .91F 24.43 663.36N 625 625

ofinstructionalfunding.

Table4examinestheimpactofleaderconfidenceonstudent attendance, a rather different performancemetric. If the motivational aspects of isopraxisleadership are effective, then this positive schoolatmospherecouldwellmakeschoolmoreinterestingforstudentsand,atthemargins,playaroleinenticingmore students to attend classes.While there is notmuch variation in student attendance, even modestchanges in the rate affect the level of state fundingand overall performance and, thus, are well worthpursuing.Leaderconfidence,however,isunrelatedtostudentattendanceinthefirstregressionandactuallynegativelylinked(albeitatthe.10levelwithonetail)in the autoregressive equation.The clear conclusionis that leader confidence has no impact on studentattendance. Theimpactofleaders’confidenceonhigh-endcollegeTheimpactofleaders’confidenceonhigh-endcollegeprep performance is examined in table 5. For thefirsttime,itcanbeseenthatconfidencemattersafter

one controls for past performance. A one standarddeviation increase in confidence is associated witha .93 percentage point increase in students scoringabove the 1110 mark on the SAT. Across the fullrangeofthevariable,isopraxisleadershipcouldhaveanimpactaslargeas5.6percentagepoints.Thisisasubstantialimpact,a26%increaseovertheaverageforalldistricts(21.3).Regressiondiagnosticsconfirmtherobustnessofthefinding;itisnotaffectedbyextremevalues,thesizeofthedistrict,ortheexclusionofanyofthecontrolvariables.Whatispuzzlingaboutthisrelationship,however,isthatitisnotreflectedinthecomponentparts.Thatis,whenoneexaminesaverageSATscoresandaverageACTscores(detailedresultsnotshown),neitherisaffectedbyleaderconfidence.The impact of leadership appears only to affect thehighestsetofscoresbutdoesnotchangetheaverageatall.

Giventhesizeoftheimpact,itisimportanttoprobewhetherornotleaderconfidencecouldbeexpectedtobenefitalltypesoforganizationsonsuchaperformancecriterion or, alternatively, whether it might be



Table 5. Leader Confidence and College-bound Performance

Dependent Variable = Students Above 1110 on SAT in 2010Without Past Performance With Past Performance

Variable Slope t-score Slope t-scoreLeader Confidence .656 1.43 .930 2.202009 College Boards — — .450 10.07% Black Students -.060 1.24 -.055 1.27% Latino Students -.016 0.54 -.023 0.85% Low Income Students -.401 9.49 -.208 4.87Teachers Salary (000s) .407 2.40 .209 1.31Class Size -.190 0.66 -.083 0.29Teacher Stability .069 1.13 -.053 0.92Instruction Spending (000) -.663 1.00 -.591 0.09

Standard error 10.63 9.47R-square .41 .52F 47.42 63.04N 557 541

applicable only to certain types of organizations.OnepossibilitymightbetheDizzyDeanhypothesis,thatonlyorganizationsthatarealreadytalentedhavethe ability to use isopraxis leadership to improvefurther.Thisnotionessentiallysuggestsaninteractionbetweenpastperformanceandleaderconfidence.Theclearestwaytoshowsucharelationshipistosimplydividethesampleintotwogroups–thoseabovethemean in terms of 2009 college board performanceand those below the mean. If the Dean hypothesisis correct, leader confidence should be stronglycorrelatedwithperformanceforthosewithhighlevelsof past performance and unrelated for those withlowperformance.Table6reportsthecoefficientsfortheseregressionsandshowsexactlythatpattern.Fordistricts above themean in2009, thecoefficient forleader confidence jumps to1.665and is statisticallysignificant; for thedistrictsbelowthemean, there isnorelationshipbetweenconfidenceandperformance.

Table 6. Leader Confidence and College Performance: The Need for Positive Past Performance and Resources

Dependent Variable = Students above 1110 on SATVariable Slope t-scorePoor Past PerformanceImpact of Leader Confidence -.021 0.04Good Past Performance*Impact of Leader Confidence 1.665 2.68Below Average ResourcesImpact of Leader Confidence .171 0.34Above Average Resources**Impact of Leader Confidence 1.705 2.55

Note: All equations control for 2009 college boards, percent black students, percent Latino students, percent low income students, teachers’ salary, class size, teacher stability, and instructional spending.*Good past performance = greater than 19.5% above 1110 **Above average resources = more than $10,000 in revenue per pupil

The issue can be pursued still further. If sports area source of inspiration for management research,thenperhapsitmightbeworthtestingtheNewYorkYankees’hypothesis–thatis,thatsomedistrictssimplyhavealotmoreresourcesandthuscanthrowmoneyat anyproblem that arises.Again, this suggests thatconfidenceismorelikelytopayoffwhenaccompaniedbyampleresources–onceagain,ahypothesisaboutaninteraction.Toinvestigatethispossibility,ananalysisis performed by splitting the sample at themean interms of revenues per student. The lower portionof Table 6 shows the confirming results. For thosedistricts with more plentiful resources, confidencehasastrongpositivebenefit(asignificantcoefficientof1.70),butforthosewithbelowaverageresources,such a leadership style has no impact on studentperformanceonthecollegeboardexams.

Table 6 raises the question as towhether talent andresourcesareessentiallythesamethingorsubstitutesforeachotherintermsofallowingconfidencetoworkitsmagic. If one examines the 112districts that areabove the mean in terms of past performance andabovethemeanintermsofresources,onefindsthattheeffectisadditiveratherthansubstitutive.Theslopewithin this subsample (regressionnot shown) jumpsdramatically to 3.41 and is strongly significant. Fortheother428districts,thereisnorelationshipbetweenleader confidence and performance on the collegeboards. This pattern of relationships holds only forthecollegeboard indicator.Similar regressionswithTAKSratesandattendancedidnotfindanysubgroupswhere isopraxis leadershipwaspositivelyassociatedwithgreaterperformance.

IMPLICATIONS

Thestrategyofanalysisherewas to stepback fromthe complexity of current leadership theories andfocus on one element common to many theories,the idea that managers need to be self-confident,optimistic individuals who serve as role modelsto subordinates.While this valued trait is found incharismatic leadership, transformational leadershipandotherleadershiptheories,itisonlyoneaspectoftheories that specifynumerousvariousandcomplex



interrelationships. The present analysis starts withthe idea of isopraxis leadership – the process bywhich leader confidencemight affect organizationalperformancethroughrolemodelingandbuildingtheconfidenceof subordinates – and then examines thekeyconceptofleaderconfidenceanditsrelationshiptothreemeasuresoforganizationalperformance.

The overall results, as sketched in the precedingsection, present a rather negative assessment ofleader confidence. For some important performancemeasures, this leadership style has no influence onresults.Thisiseventhecaseforstudentattendanceasameasure,althoughthisoneinparticularwouldseemtobeespeciallyamenabletoinfluencebyaconfidentand possibly motivating top manager. Further,althoughthisleadershipapproachdoesseemtomatterfor college-bound performance, it does so only fororganizationsthatarealreadyperformingwelland/orluckyenough tohavebetter thanaverage resources.Therich(infundsorachievement)getricher.Forthisperformance criterion, additionally, the subsequentperformance effects are additive. Confidence byorganizational leadersseemstodonothingfor thoseschool districts that most need improvement. Thisfindingisaparticularlyperverseone.TheDizzyDeanhypothesisandtheNewYorkYankeeshypothesisfindsomesupport,itistrue;butevenifthereisnocryinginbaseball,onecanbe saddened todiscover suchapattern.

These results pertain directly only to the severalhundredpublicorganizationsthatareincludedinthissample and for the time period under investigationhere.Thefindings, furthermore,derivefromusinganewmeasureoftheconceptofleaderconfidence.Themeasure has some obvious strengths, especially thepurgingofpriorperformance,butitisstillrelativelyuntested.Furthervalidationiswarranted.Theresultsherearemostlikelytobevalidforthoseotherpublicorganizationsthatsharekeycharacteristics–thosethatarehighlyprofessionalizedwithsubstantialdiscretionlodged at the street/classroom level. It could be thecasethatperformanceinothertypesoforganizationsrespondstoleaderconfidencemore(orperhapsless)favorably that in the sample analyzed here. It will

thereforebe important to replicate this investigationinotherempiricalsettings.

Ofcourse,doingsoinanappropriatewaywillrequiredata in timeseries that includesensiblemeasuresofarchivalperformanceandalsoperceptualassessmentsof performance by leaders to tap the degree ofconfidence and optimism that is present. Purgingalready-established levels of performance must beanessentialpartof suchstudies, even though–andespeciallybecause–takingthisstephasnotbeenpartof any earlier studies of self-efficacy in leadershipand individual (e.g., athletic) or organizationalperformance.Thepresent study shows thatomittingthecontrolforestablishedperformancebiasesresultstowardpositivefindingsregardingconfidence.

Theimplicationsofthestudyreportedherecertainlyhave the potential reach to additional publicmanagementsettingsandalso–becauseofthefailureuntilnowofresearchersinmanyotherfieldstopurgepastperformance–intoanumberofotherrealms,fromthedeterminantsofathleticperformancetothepracticeof leadership in large organizations to the content of motivational/leadership training seminars forpublic,private,andnonprofitmanagers.Thispointdoesnotentail any sort of overall indictment of training anddevelopment initiatives.Butwhile it is undoubtedlytrue that investments in leadershipdevelopmentandtraining can bear substantial dividends, it seemsequallylikelythatsomemuch-toutedversionsofsuchtrainingproducelittleintermsofoverallperformance.The trick will be in separating the wheat from thechaff.

Muchhasbeensaidaboutthevalueofconnectingtheresearchagendasandfindingsofpublicmanagementscholarstothepressingneedsofpublicmanagersandpublicorganizations.Thisstudy,whilerepresentingapreliminaryinvestigation,underscoresthatpoint.Largeamountsoftimeandfinancialresourcesarecurrentlybeing expended on behalf of training programs andefficacyinitiativesthataresoldtoclientsonthebasisof virtually no systematic empirical evidence asidefrom anecdotes and testimonials. One should beappropriately suspicious of such claims when they



appearinotherfieldsofknowledge,suchasmedicalpractice,wheretheevidence-basedmovementhasmadeimportantstridestowardperformanceimprovements.Oneshouldalsobecautiousinacceptingsuchclaimson behalf of isopraxis leadership, an approach tomanagement with many adherents, an industry oftrainers and consultants, but essentially no properlyvalidatedresults.

This is one of those topics onwhichone should beskeptical of the findings of interpretive research.Most people would prefer that isopraxis leadershipworks. Many people believe that it should. Thosetrained in thevalueof confidence andoptimism formeetingpublicobjectivesmaycarryunderstandable,implicit biases about its influence in terms of theirownexperience.Theymayevennoteanassociationbetween efficacy and performance on the basis oftheirandothers’experiences.Theroleofsystematicresearch in such a circumstance can be to examinethedata,conductresearchthatcangetpastthedangerofspuriousfindings–thusdistinguishingassociationfromcausality,andtherebyassisttheworldofpractice.

Thecredibilityofacademicsandtheirresearch,nottomentionthesocialvalueoftheirfindings,islikelytobeconsiderablyenhancedbyconductingresearchthatbears direct relevance to, and actually assists, someofthekeydecisionsthatpublicmanagersmustmake.How to prepare suchmanagers to lead, and how toexpendpublic fundsonbehalf of this objective, arecertainlyamongthesekeydecisions.

REFERENCES

Andrews, Rhys, GeorgeA. Boyne, and RichardM.Walker.2006.“SubjectiveandObjectiveMeasuresof Organizational Performance. An EmpiricalExploration.” In George A. Boyne, Kenneth J.Meier, Laurence J. O’Toole, Jr., and RichardM. Walker, eds. Public Service Performance: Perspectives on Measurement and Management.Cambridge:CambridgeUniversityPress:14-34.

Bandura, Albert. 1977. “Self-Efficacy: Towarda Unifying Theory of Behavioral Change.”Psychological Review84:191-215.

Bandura,Albert.2005.“GuidesforConstructingSelf-EfficacyScales.”InFrankPajaresandTimUrdan,eds.,Self-Efficacy Beliefs of Adolescents.Charlotte,NC:InformationAgePublishing:307-335.

Barling, Julian and R. Beattie. 1983. “Self-EfficacyBeliefs and Sales Performance.” Journal of Organizational Behavior Management5:41-51.

BaseballAlmanac,n.d.DizzyDeanQuotes,retrievedfrom http://www.baseball-almanac.com/quotes/quodean.shtml

Bass,BernardM.1998.Transformational Leadership.Mahwah,NJ:Erlbaum.

Bennis, Warren and Burt Nanus. 1985. Leaders: Strategies for Taking Charge. New York:HarperCollins.

Boal, Kimberly B. and John M. Bryson. 1988.“CharismaticLeadership:APhenomenologicalandStructuralApproach.” pp. 11-28 in JamesGeraldHunt, B. Rajaram Baliga, H. Peter Dachler, andChester A. Schriesheim, Emerging Leadership Vistas.NewYork:LexingtonBooks.

Burns, JamesMacGregor Burns. 1978. Leadership.NewYork:HarperCollins.

Campbell, Nancy K. and Gail Hackett. 1986. “TheEffects of Mathematics Task Performance onMathSelf-EfficacyandTaskInterest.”Journal of Vocational Behavior28(Number2),149-162.

Chen,Gilad,StanleyM.Gully,andDovEden.2001.“ValidationofaNewGeneralSelf-EfficacyScale.”Organizational Research Methods 4,1 (January):62-83.

Conger, Jay A. and Rabindra N. Kanungo. 1998.Charismatic Leadership in Organizations.ThousandOaks,CA:SagePublications.

deBoef,SuzannaandPaulM.Kellstedt.2004.“ThePolitical (and Economic) Origins of ConsumerConfidence.”American Journal of Political Science 48(Number4),633-649.

Downs,Anthony.1967.Inside Bureaucracy.Boston:Little-Brown.

Eden, Dov. 1984. “Self-Fulfilling Prophecy as aManagement Tool: Harnessing Pygmalion.”Academy of Management Review 9 (Number 1),64-73.

Eden,Dov.1990.“PygmalionWithout InterpersonalContrastEffects:WholeGroupsGainfromRaising



Expectations.”Journal of Applied Psychology 75(Number4),394-398.

Feeney, Mary K., and Craig Boardman. 2011.“Organizational Confidence: An EmpiricalAssessmentofHighlyPositivePublicManagers.”Journal of Public Administration Research and Theory21,4(October):673-97.

Feltz,DeborahandCathyD.Lirgg.1998.“PerceivedTeamandPlayerEfficacyinHockey.”Journal of Applied Psychology83(Number4),557-564.

Fernández, Sergio. 2005. “Developing and Testingan Integrative Framework of Public SectorLeadership: Evidence from the Public EducationArena.”Journal of Public Administration Research and Theory15,2(April),197-218.

Fernández,Sergio,YoonJikCho,andJamesL.Perry.2010. “Exploring the Link between IntegratedLeadership and Public Sector Performance.”The Leadership Quarterly21:308-323.

Finn, Jeremy P. and Charles M. Achilles. 1999.“Tennessee’s Class Size Study: Findings,Implications and Misconceptions.” Education Evaluation and Policy Analysis21,97-110.

Getha-Taylor,Heather,MajaHusarHolmes,WillowSA. Jacobson, Ricardo S. Morse, and Jessica E.Sowa. 2011. “Focusing the Public LeadershipLens: Research Propositions and Questionsin the Minnowbrook Tradition.” Journal of Public Administration Research and Theory 21,Supplement1(January):i83-i97.

Gist, Marilyn E. 1987. “Self-Efficacy: ImplicationsforOrganizationalBehaviorandHumanResourceManagement.” Academy of Management Review 12(Number3),472-485.

Gist,MarilynE.andTerenceR.Mitchell.1992.“Self-Efficacy:ATheoreticalAnalysisofItsDeterminantsandMalleability.”Academy of Management Review 17(Number2),183-211.

Gould, Daniel, Maureen Weiss, and R. Weinberg.1981.“PsychologicalCharacteristicsofSuccessfulandNonsuccessfulBigTenWrestlers.”Journal of Sport Psychology3,69-81.

Gully,StanleyM.,KaraA.Incalcaterra,AparnaJoshi,andJ.MathewBeaubien.2002.“AMeta-Analysisof Team Efficacy, Potency, and Performance.”Journal of Applied Psychology 87 (Number 5),


819-832.Hannah,SeanT.,BruceJ.Avolio,FredLuthans,and

P.D.Harms. 2008. “LeadershipEfficacy:ReviewandFutureDirections.”Leadership Quarterly19:669-692.

Hansen, JesperRosenberg andAndersR.Villadsen.2010. “Comparing Public and PrivateManagers’Leadership Styles.” International Public Management Journal13(3):247-274.

Hanushek,Erik.1996.“SchoolResourcesandStudentPerformance.”InGaryBurtless,ed.,Does Money Matter?Washington,DC:Brookings.

Hedges,LarryV.,andRobGreenwald.1996.“HaveTimes Changed? The Relation between SchoolResources and Student Performance.” In GaryBurtless, ed., Does Money Matter? Washington,DC:Brookings.

Heggestad, Eric D., and Ruth Kanfer. 2005. “ThePredictive Validity of Self Efficacy in TrainingPerformance:LittleMoreThanPastPerformance.”Journal of Experimental Psychology11,2(June):84-97.

Hess, Frederick M. 1999. Spinning Wheels: The Politics of Urban School Reform.WashingtonDC:TheBrookingsInstitution.

Highlen, Pamela S. and Bonnie B. Bennett. 1983.“EliteDiversandWrestlers:AComparisonBetweenOpen-andClosed-skillAthletes.”Journal of Sport Psychology5:390-409.

Javidan, Mansour and David A. Waldman. 2003.“Charismatic Leadership in the Public Sector.”Public Administration Review63(Number2),229-242.

Judge, Timothy A., Christine L. Jackson, John C.Shaw, BrentA. Scott, and Bruce L. Rich. 2007.“Self-EfficacyandWork-RelatedPerformance:TheIntegralRole of IndividualDifferences.” Journal of Applied Psychology92(Number1),107-127.

Jung,DongI.andJohnJ.Sosik.2003.“GroupPotencyandCollectiveEfficacy.”Group and Organization Management28(Number3),366-391.

Miles, Raymond E., and Charles C. Snow. 1978.Organizational Strategy, Structure, and Process.NewYork:McGraw-Hill.

Moritz,SandraE.,DeborahL.Feltz,KyleR.Fahrbach,andDianeE.Mack.2000.“TheRelationofSelf-


EfficacyMeasurestoSportPerformance:AMeta-AnalyticReview.”Research Quarterly for Exercise and Sport71(Number3),280-294.

Morse, Ricardo S., Terry F. Buss, and C. MorganKinghorn, eds. 2007. Transforming Public Leadership for the 21st Century. Armonk, NewYork:M.E.Sharpe.

Nye, Barbara, Larry Hedges, and SpyrosKonstantopoulos.1999.“TheLong-termEffectsofSmallClasses.”Education Evaluation and Policy Analysis21:127-142.

O’Toole, Laurence J. and Kenneth J. Meier. 2011.Public Management: Organizations, Governance and Performance. New York: CambridgeUniversityPress.

Ouchi,WilliamG.andLydiaG.Segal.2003.Making Schools Work.NewYork:SimonandSchuster.

Paglis, Laura L., and Stephen G. Green. 2002.“Leadership Self-Efficacy and Managers’Motivation for Leading Change.” Journal of Organizational Behavior23:215-35.

Pajares, Frank. 2006. “Current Directions in Self-efficacyResearch.”InM.MaehrandP.R.Pintrich,eds., Advances in Motivation and Achievement.Volume10.Greenwich,CT:JAIPress:1-49.

Park, S. M. and Hal G. Rainey. 2008. “Leadershipand Public Service Motivation in U.S. FederalAgencies.” International Public Management Journal11(Number1),109-142.

Pillai, Rajnandini and Ethlyn A. Williams. 2004.“TransformationalLeadership,Self-Efficacy,GroupCohesiveness, Commitment and Performance.”Journal of Organizational Change Management17(Number2),144-159.

Rainey,HalG.2009.Understanding and Managing Public Organizations. 4th ed. San Francisco:Jossey-Bass.

Rattsø, Jørn. 1999. “Aggregate Local Public SectorInvestment and Shocks.” Applied Economics 31(Number4),577-584.

Romzek,BarbaraS. andPatriciaWallace Ingraham.2000. “Cross Pressures of Accountability:Initiative,CommandandFailureintheRonBrownPlane Crash.” Public Administration Review 60(Number3),240-53.

Rosenthal, Robert and Lenore Jacobson. 1968.

Pygmalion in the Classroom: Teacher Expectation and Pupils’ Intellectual Development.NewYork:Holt,ReinhartandWinston.

Shamir,Boas,RobertJ.House,andMichaelB.Arthur.1993. “The Motivational Effects of CharismaticLeadership.” Organizational Science 4 (Number4),577-594.

Sherer,Mark,JamesE.Maddux,BlaiseMercandante,StevenPrentice-Dunn,BethJacobs,andRonaldW.Rogers.“TheSelf-efficacyScale:ConstructionandValidation.”1982.Psychological Reports51:663-71.

Sims, Henry P., Jr., and Peter Lorenzi. 1992. The New Leadership Paradigm: Social Learning and Cognition in Organizations. Newbury Park, CA:Sage.

Taylor,M.Susan,EdwinA.Locke,CynthiaLee,andMarilyn E. Gist. 1984. Organizational Behavior and Human Performance34(Number3),402-418.

Thompson, Bruce. 2004. Exploratory and Confirmatory Factor Analysis. Washington, DC:AmericanPsychologicalAssociation.

Trottier, Tracey, Mongtomery Van Wart, andXiaoHu Wang. 2008. “Examining the Nautreand Significance of Leadership in GovernmentOrganizations.”Public Administration Review68,2(March-April):319-33.

Van Slyke, David M. and Robert W. Alexander.2006. “Public Service Leadership: Opportunitiesfor Clarity andCoherence.”American Review of Public Administration36:362-374.

Vancouver, Jeffrey B. and Laura N. Kendall.2006. “When Self-efficacy Negatively Relatesto Motivation and Performance in a LearningContext.” Journal of Applied Psychology 91(Number5),1146-1153.

Wenglinsky,Harold.1997.“HowMoneyMatters:TheEffect of School District Spending onAcademicAchievement.”Sociology of Education 70 (July),221-237.

Yates, Allen. 1987. Multivariate Exploratory Data Analysis.Albany,NY:SUNYPress.

Yukl,Gary.1989.“ManagerialLeadership:AReviewofTheoryandResearch.”Journal of Management 15:251-289.

Yukl, Gary. 2010.Leadership in Organizations. 7th



Edition.UpperSaddleRiver,NJ:PrenticeHall.Zeller, RichardA. and Edward G. Carmines. 1980.

Measurement in the Social Sciences. NewYork:CambridgeUniversityPress.

ABOUT THE AUTHORS

KennethJ.MeieristheCharlesH.GregoryChairinLiberalArtsandDistinguishedProfessorofPoliticalScience at Texas A&M University. He is also aProfessorofPublicManagementintheCardiffSchoolofBusiness,CardiffUniversity(Wales).Hisresearchinterests includepublicmanagement, representation,race and politics, and public policy.He is currentlyworking on building and testing theories of publicmanagement and representative bureaucracy acrossdifferentnationalcontexts.HeisthefoundingeditorofPerspectivesonPublicManagementandGovernance.

Laurence J. O’Toole, Jr. is Margaret Hughes andRobert T. Golembiewski Professor of PublicAdministrationandDistinguishedResearchProfessorintheDepartmentofPublicAdministrationandPolicy,SchoolofPublicandInternationalAffairs,UniversityofGeorgia,USA.Heisalsoprofessorofcomparativesustainability policy studies in the Department ofGovernanceandTechnologyforSustainabilityattheUniversityofTwenteintheNetherlands.Hisresearchinterestsincludequestionsofpublicmanagementandperformance, as well as policy implementation innetworks.



CPARorca.cf.ac.uk/107557/1/cpar.isopraxis.2018.pdfCPAR Review Attention has been given to the notion...

Documents

Transcript of CPARorca.cf.ac.uk/107557/1/cpar.isopraxis.2018.pdfCPAR Review Attention has been given to the notion...