STRATEGIC HUMAN RESOURCE MANAGEMENT AND ORGANIZATIONAL PERFORMANCE: TESTING A TYPOLOGY OF FIT

To be urilizeddfecrively and to have a po~irive effect on the mukirude of complex challenges facing South Mrican organizations. human moure" (HR) have to be managed in a srntegically CO( \'C'Ct manner. This study investigated the performance implicariom of tv .. o types of fit in stntegi< human resource management (SHRM). The results indicated that an organiurions HR strategy had a modeming effect on the relationship berv .. ecn business unreg)' and performant:e. Furthermore, the survey indicated that organizations could be classified according to clusters ofim~mal ly consistent HR pnctiCC'S. with some perfo~mancc: differen«s across th~ dust~l'S. Finally. C'\;denc~ for the validir),ofthe propoJed typology offit were found, thus maldng a contribution to theory building in the field of SHRM.

In order to make stnregic changes effective and conducive to organizational performance.they have to be approached.developed and implemented in a str.l.tegically correct manner (Grobler. 1993).HR managers are increasingly being called upon to develop HR strategies and prognmmes targeted at improving the eff'ccti\'enes! of organizations.Top management is also increasingly turning to HR practitioners to act as agents of change in facilitating the com plex tnnsformation process in organizations (Swanepoel.Erasmus, Van Wyk & Schenk.1998).South African org:rn izations are in a tr.l.nsition phase char.l.cterized by inslability and the human resource specialists hence lire drawn tight into the broader dynamics of a society in transformation.
Employers in South Africa seem to lack the necessary str.l.tegk orientation to ensure the integntion of HR issues with the business aspects of orgtnizatio ns (Swanepoel ct ai., 1998).Research done by Grobler (1993) led him to conclude that th~ time has come for South African organizations to face the challenge of the futu re and adopt the recommended str.l.tegic approach of H RM . .The approach to enabling H R managers to contribute signifi-Rt'lU(Srsp 1rJIri.1IshoIJd buddrtSStd r~: .-tsE.rdbrttlu [)(pl .'" loid.lSfrialPsl"iwJ"l}: us.Pri .... &g Xl MlJtitla"J, ~. '0 candy to the success of the organization is known as stnl~ic human resource management (SHRM).SHRM.then. is ~ process which is used to anticipate future HR dem~nds b~std on business strategy and to plan for the provision.moti\'ation and development of human resources in such a way mat th~ mission and the business stntegy of the organization o.n best be realized (Smit. 1993).As such.SHRM represents a mort holistic focus on organizational performa nce, as well .s a m2jor shift from the more traditional view of hum.n resource management primarily as a staff function.
SH RM includes considering the degree of 'fi t' between an organizations human resou rce and business srntegies (Dyer &: Holder. 1988).The notion of 'fit' is essential to explain the rtb• tionship bttween SHRM and org:rnizational perform~nce.This could be done by examining the match between HR strategies and business stntegics in organizations.or by exami• ning the internal coherence among a set of HR dimensions (i.e.selection.tnining and compensation~ Sevenl moods have been developed to indicate how the company St tlt~y :rnd the HRM function can be linked (Arthur. 1992: Dyer &: Holder, 1988;Miles & Snow, 1984;Sivasubramaniam I>: Kroeck. 1995).Sivasubramaniam and Kroeck (1995) proposed a new f)'pology of fit in SH RM :rnd determined the effect it ultimately has o n organizatio nal performance.They tCSted the differCill fo rms of fit (see Figure 1) using data collected from i broad nnge of organizations in the USA.There exists a need 10 replicate this study in different industry contexts and count• ry/cultural settings, like in South Africa.to examine the genenlizability of the fr.l.me\',•ork.
A potential limitation of the study of Sivasubramaniam and Kroeck (1995) aro~ from the fact that information on HR stntegy and bwiness strategy was sought from a single rt• spondent, the HR executive.raising the possibility of (om• mon method bias.Sivasubnm:rniam and Kroeck (1995) th~refor~ ass~n that future r~arch should b~ design~d as multiple-informant studies 50 that respons~s from chief executive ~fficers (CEOs) could be related to responses from HR execuuves.
Research in SHRM is at a crossroads today, and fo r the field to progress beyond descriptions of HR practices in different settings to theory testing, explicit consideration of fit among HR practice1, organizational contingenci~s and organizational p~r formance.is essential (Sivasubramaniam & Kroeck. 1995).

Purpose of study
This study is the fint to explicate different types of fit in SHRM and their impacts on organizational performance in the South African context.The general research aim of this Study was to ~nhanc~ the limited body of knowledg~ on how stnt~g ic human resourc~ manag~m~nt can have a positive influence on organizational performance.The specific purpose of this study was to test the val idity of Sivasubramaniam and Kroeck's proposed typology of fit in SH RM in the South Afrian context.

Dtsc:: ription of6t typologies
The typology of fit proposed by Sivasubramaniam and Kroeck (1995) is adapted from the work of O razin and Van de Yen (1985) and Venkatnman (1989) and is based on nvo dimensions.namely the context of fit (internal or external) and the criterion-specificity of fit (criterion-frtt or criterion-specific).Si\'a5ubramaniam and Kroeck (1995) propose fit in sm.tegic human resource management to be in five fotms.namely: matching.inte-nction.g~stalt.s•ystem.and covariation (see Figure 1).This study conantrated only on the fit as interaction and fit as gestalt perspectives. . .
Cnteno n-sp eclfic C rnenon-free Criterio n-specificity of fit Figure 1: Typology of 6t in SHRM (Sivasubramaniam & Kroeck, 1995. p. 74) Fit as interaction: The 'fit as interaction' p~rspective explores how the relationship between strat~gic human resource manag~ment pnctices and external context impact on specific performance criteria (Sivasubramaniam & Kro~ck. 1995).Several studies have tested the contingency relationships between business strategy and specific human resource practices such as compensation and selection.but only Martell (1989) and Huselid (1995) have directly examined the interaction perspective.Manell (1989) found only some support fo r the modenting effect when whether a company's strategy has a significant moderating effect on the relationship berween human resource management practices was tested.Huselid (1995) found no support for the interaction between human resource practices and business strategy and only partial support for the interaction between human resource practic~s and the company's score on a stntegic human resou[c~ index.
Likewise.Sivasubramaniam and Kroeck (1995) obtained no support for the performance impact of an interaction perspec-ti\'~ of fit in SHRM.According to Si"•asubramaniam and Kroeck, the lack of support for the interaction perspective may be due to performance criteria choice.Th~y am:rt that interaction effects may be found when other criter:ia such as turnover rate.employee satisfaction and innovativeness are used.
Here it could be argued that the interaetion between Dyer and Holder's (1988) human resource strategy and the company's business strategy.operationalized in terms of Miles and Snows (1984) typology.could impact on the company's performance.Thus. it can be postulated that: Hypo thesis 1: HR strategy will moderate thl! relationship between a company's business strategy and perfor-manc~.
Fit as gestalt: Gestalts.or archetypes, represent clusters of attributes which ~xhibit a high degree of internal coherence.'Fit as gestalt' in strategic human resource managemem can be defined as the reference to the interrelationships among diff~rem dimensions of human resource management such as staffing.training.appraisal.compensation.job design and corporate involvement.A gestalt perspective could be suPPOrted to the degree that companies could be clustered.based on the similarities of their human resource practices (Sivasubramaniam & Kroeck. 1995).
Fit as gestalt implies that the key to effective human resource management Ii~s in finding an appropriate combination of practices.whether it be the synergistic combination. of all the practices o r the specific 'architecture' or culture which binds them together (Guest. 1997).This is often sec:n as the human resource-ba5C'd source of competitive ad".ntagewithin organizations.
Even though the fit perspective of gestalt is not explidtly identified.it is referred to as useful to identify types of HR systems (Arthur. 1992)."bundles-of HR practices (MacOuffie.1995).or taxonomiesofHR strategies (lchniowski in Sivasubramaniam & Kroeck.1995).Sivasubramaniam & Kroeck (1995) found support for the gestalt perspective by identifying three internally consistent H R strategic types.However.their hypothesis.that all three strategies are equally effective.was suPPOrted only when the criteria ,vere objective measures of pe:rformance and not , .... hen perceptual measures were used.
If the Dyer and Holder (1988) typology is used to derive measures for assessing human resource practices.clusters of companies whose profiles of human resource practices ar~ internally consistent could be identified.It could further be expected that these clusters resemble the three human resource strat~gies proposed by Dyer and Holder (1988).It can th~refore be postulated that: Hypothesis 2: a: Companies can be classifi ed acco rding ro clusters of internally consistent HR practices.b: There are no significant performance differences across the clusters, as all imernally consistent fo rms are ~ually effective.The sample is representative of a wide Sp&trum of sectors in the South African industry, with 72 organisations cbssified as manufacturing companies and 34 as services companies.Sales reven ue (mean of R3 706 million per yen ) indicated that the responding companies have relativdy large turnover.and "vas relatively large (mean of3 659 employees per company), with 38 employees on avenge operative in the HR depanment.

RESEARCH MElHOD
Infonnation relating to HR strategy and HR outcomes was gathered from the HR executives, while information relating to business strategy and strategic performance indicators was obtained from the CEQ.Surveying the C EO and HR executives separately.and then matching their responses at the analysis stage only.reduced the roncept of common-method bi3.s.

Measuring instruments
This research utilized twO instruments of measurement developed by Sivasubramaniam and Kroeck (1995) to asse5S human resource strategy.business strategy and perceived organintional perform3.nce.The data pertaining to the human resource strategy part of the research was collected by means of the Human Resource PraCtices in Corporations Questionnaire.Content validity was secured and by using L1SREL.confirmarory factor analysis providtd support for rhe convergent validity of the six factor.31 -item instrument (Sivasubramaniam and Kroeck. 1995).In order 10 ensure that the South African pers pective in this field would emerge from the questionnaire.tWO items from the original questionnaire were dropped and five new items were added.The six strategic human resource m anagement dimensions represented by these thirty-four items were staffing.training.compensation.appraisal.job design and corporate involvement.D ata thus colleaed with this questionnaire provides information to classify an organiution's HR strategy as being inducement.investment o r involvement oriented (Dyer and Holder. 1988).
The data pertaining to business str:ategies was collected by me:lns of the Strategic Questionnaire.The questionnaire consisted of an 11 -item scale proposed by Conant. Mokwa and Varadarajan (1990) to oper:atlo nalize Miles and Snow's (1984) strategic [•fpology.For this study.the res ponding organizations were classified acco rdi ng to their business strategies as defenders.analyzers or prospectors.Conant et al. reported ~dequate internal consistency and inter-r:ater reliability of the Items.
Conceptualizing and measuring organizational performance reveals dill to be a r:ather complex matter.It is clear from the liter:ature that there are a number of W3.ys to measure company performance (Guest, 1997;Pu uw & Richardson. 1997: Van der Post. 1997).In this study, company performance was measured alo ng twO distinct dimensions -strategic outcomts and human /'t-Jouru OlltCOltlts.Ten items.which included sales growth nte.market share, return on sales, return on investments.return on asSets.new product development.product quality.bnnd image.customer services and the company's overall competitive posicion were used to measure straug;c outcamtJ.Eight item~ which included relative empl oy~ tumov~r, relativ~ ~m ploy~ absenteeism.employee turnover.employee absen-t~i sm, quality of work fo rce.number of employ~ grievances, labour productivity and occupational safety, were used to meas ure human trSOUf(t out(omes.

IUSULTS
After operationa.lizing business strategy and HR strategy from the data, the formulated hypotheses have been tested in both the external and intem3.lCOnt~xt of fit.The effects of HR str:ategy and business srntegy on organizational performance were investigated from two different perspea:ives of fit.

Fit as interaction
In this study, it ~ proposed that the inter:action effect ~~n a company's HR strategy and business str:ategy would impact positively on organizational performance.To test this hypothesis.hierarchical regression was employed.sin« this proved to be ~n appropriate statistical technique for determining w hether significant m odenting effects existed betWmI va riables (Tabachnick & Fidell. 1989).The aim was to deea-mine if addition of information regarding HR stralegy ."dbusiness str:ategy improved prediction of organizational per.
fo rmance beyond that affo rded by the interaction (modmting) effect betw~n HR and business strategies.HR strategy.
business Strategy and the inlenction effect between them wm measured 3.S independent va riables and the performanct measures as the dependent variables.In step one, the interaction effect w as added to the equation.In step two.business strouegy was added to the equation.l nd in step three.H R strategy was added.
After step I. w ith the inter:action between HR strategy and business stntegy in the equatio n, for HR performance.hiera rchical regression analysis is used.Hypothesis 1. namdy thlt the interaction betw~n HR st rat~gy and business mateg), will affect a company's performance, can be ~ccepted.To determine if there is a significant increase in R 2, business strategy :md HR str:ategy was added to the equation.
The results of an ANOVA sho\ved signifinnt differences {F (2, 148) = 300.65,p < 0,001) between the three cluster means (ctnters).indiC<llting cluster 1 to be on the opposite of a continuum to cluster 3, with cluster 2 in the middle.
In order to test for performance differences across the thrtt HR srntegies, :m SPSS'" ANOVA was used with fi"'e performance measures as dependent variables and the duster distribution as the independenr variable.The results in the ANOVA indicate that there could be a difference among the duster means (p < 0.05) of HR performance and perceived financial performance.Hypothesis lb.stating that all three HR stntcgies are equally effective, was thus only partially supported when the criteria were organ izational performance, strategic performance and employee productiviq: In order to identify which pairs of means differ significantly.the LSD (least significant difference) test of the multiple range testS (POSt Hoc) was conducted (SPSS Users Guide.1998), The differences among the above-mentioned performance measures entails a substantial possibility that a compan y thar follows the Inducement HR stntegy, ach i~ved lower (p< 0.05) HR p~rformance than companies follow ing Inv~stment or Involvement HR strategies.Companies can also have higher (p < 0.05) perceived fmancial perform a nc~ if they follow Involvement HR strat~g ies.as opposed to following Induce-m~nt or I nvestm~nt strategies.

DISCUSSION
The results indicate thar the interaction between a company's HR stntegy and business strat~g y does hav~ :t positive in-fluen~ on company performance.In ch~i r study, Sivasubnmaniam and Kroeck (1995) found no suppOrt for th~ performance impact of an interaction perspective of fie.Contrary to this study.they used an ANOVA to determine the interaction effect.Failure to support the hypothesis was ascribed to choice of criteria and operationalization of strategy variables.
The results of this stud\' indicate that the intemal consistent HR practices of the different companies do resem ble the ch anct~ristics of the three H R strategies proposed by Dyer and Hold~r (1988).Differences in aJlocation of training budgets betwe~n the three dusten provided some face validity to th~ classiflcation scheme that was used.The proposition that co mpani~s could be classified according to dusten ofintem:tl-Iy consistent HR practice was supported by th~ results of this study, which indicated a positive relationship between the identified dusters and the proposed HR scntegies.
The results of this study indicate that companies in different dusten differ from one another on HR performance and perceived fin anciaJ perfonnance.Thus.companies following Involvement H R stntegies may differ significantly from companies following Inducement H R stnt~gi es in tenns of HR performance and perceived fi nanci:tl performam:~.The three HR strategies were only equally effective in terms of the org~ nizational.stntcgic and ~mploy~~ productivity me:!.sures of performance.Sivasubramaniam and Kroeck (1995) found that al l thr~ strategies were o nly equally effective when th~ criteria.in their case.were objective measures of p~rformanc~.
R esults of this study indicated that both the internal and external context of flt were either related.or partially related.to organizational performance.It could thus be arg ued that fit among the HR dimensions and flt with business stntegy could be determinants of organizational perform anc~.

CONCLUSION
T his study addressed a relevant research field since SHRM is regarded as one of the current approaches to effectively tnnsform and develop organizations.This study tested the proposed typologyofflt in SHRM of Siva sub ram ani am and Kroeck (1995).and also address~d th~ question of performance implications of fit for C\\'O of their five different perspectiv~s of fit.To pro\'ide a more holistic view and understanding of this field, furur~ research could also incl ud~ the psychom~tric testing of the fit typologies of matching.systems (profile deviation) and conriation (Sivasubramaniam & Krocek, 1995) in the South African cont~x t.
The validation of the proposed typology of fit in SHRM can b~ taken even further if more dimensions were added to th~ actual database.More detailed information on labour union orientation, affirmative action and management of diversity within the company especially could fu rther the possible understanding regarding implementing specific HR ac tiviti ~s that could prove to be s ucc~ssful in dealing with these typical strategic HR issues in South Africa (Gerber et aL. 1998).
This study has provided ro me insights inco the SHRM-Strat~ gy-Performance relationships.Furthermore.the study made a contribution to theory building in the field ofSHRM since it substantiates the first ~v id~nces. in a broad spectrum of companies.for the generalization of Sivasu bramaniam's pOStulated new typology of fir to [he South African corporate world.The study.th~refore, furth~r elevated SHRM to a rheorerical levcl rather than merdy a descriptive model.