compensation, esteem valence and job performance: an empirical

24
Int. J. of Human Resource Management 13:4 June 2002 697-719 i j Routledge ^___^ __^_ S % Taylor & Francis GK Compensation, esteem valence and job performance: an empirical assessment of Alderfer's ERG theory CA. Arnolds and Christo Boshoff Abstract Improving the job perfomiance of employees has been the focus of many motivation theories, especially the need theories. These theories have however been questioned because of a lack of research on the causal relationship between need satisfaction and job perfonnance. Research on the link between the need satisfaction-job performance relationship and individual personality differences among people has also been neglected. This study addresses these research gaps as far as the intervening influence of personality variables on the need satisfaction-job performance relationship is concemed. The present study investigates the influence of need satisfaction (as suggested by the Alderfer theory) on self-esteem (the personality trait) and the influence of self-esteem on perfonnance intention (the surrogate measure for job performance) of top managers and frontline employees. The empirical results show that esteem as a personality variable exerts a significant influence on the job performance of both top managers and frontline employees. These and other findings provide important guidelines for managers on how to address the motivational needs of top managers and frontline employees in order to improve their job performance. Keywords Job performance; esteem valence; existence needs; relatedness needs; growth needs. Introduction One of the most important organizational goals of any business firm is the maximization of its return on investment by reducing production or service delivery costs. This goal, together with the competitive nature of global business today, has increased the importance of cost reduction and the delivery of quality goods and services. As a result, many firms have had to re-evaluate the effectiveness of their operations. Organizational effectiveness, defined as the extent to which an organization achieves its goals (Mullins, 1999: 861), is largely dependent on the productive utilization of the materials, machines, money and people power in the production and delivery processes of firms. It is particularly important that the people power input of the production process of firms is effectively and efficiently utilized, as people to a large degree control the other input elements in the production process. In other words, firms need effective and productive employees whose job performance is at an optimal level to be able to C.A. Arnolds, Department of Business Management, University of Port Elizabeth, PO Box 1600, Port Elizabeth, South Africa (tel: -F041 5042692; fax: +041 5832644; e-mail: [email protected]). Christo Boshoff, Department of Business Management, University of Port Elizabeth, PO Box 1600, Port Elizabeth, South Africa (tel: +041 5042577; fax: +041 5832644; e-mail: [email protected]). The International Joumal of Human Resource Management ISSN 0958-5192 print/ISSN 1466-4399 online © 2002 Taylor & Francis Ltd http://www.tandf.co.uk/journals DOI: 10.1080/09585190210125868

Upload: hoangnhan

Post on 14-Feb-2017

220 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Compensation, esteem valence and job performance: an empirical

Int. J. of Human Resource Management 13:4 June 2002 697-719 i j Routledge^___^ _ _ ^ _ S % Taylor & Francis G K

Compensation, esteem valence and jobperformance: an empirical assessment ofAlderfer's ERG theory

CA. Arnolds and Christo Boshoff

Abstract Improving the job perfomiance of employees has been the focus of manymotivation theories, especially the need theories. These theories have however beenquestioned because of a lack of research on the causal relationship between needsatisfaction and job perfonnance. Research on the link between the need satisfaction-jobperformance relationship and individual personality differences among people has alsobeen neglected. This study addresses these research gaps as far as the interveninginfluence of personality variables on the need satisfaction-job performance relationship isconcemed.

The present study investigates the influence of need satisfaction (as suggested by theAlderfer theory) on self-esteem (the personality trait) and the influence of self-esteem onperfonnance intention (the surrogate measure for job performance) of top managers andfrontline employees. The empirical results show that esteem as a personality variableexerts a significant influence on the job performance of both top managers and frontlineemployees. These and other findings provide important guidelines for managers on howto address the motivational needs of top managers and frontline employees in order toimprove their job performance.

Keywords Job performance; esteem valence; existence needs; relatedness needs;growth needs.

Introduction

One of the most important organizational goals of any business firm is the maximizationof its return on investment by reducing production or service delivery costs. This goal,together with the competitive nature of global business today, has increased theimportance of cost reduction and the delivery of quality goods and services. As a result,many firms have had to re-evaluate the effectiveness of their operations. Organizationaleffectiveness, defined as the extent to which an organization achieves its goals (Mullins,1999: 861), is largely dependent on the productive utilization of the materials,machines, money and people power in the production and delivery processes of firms.It is particularly important that the people power input of the production process offirms is effectively and efficiently utilized, as people to a large degree control the otherinput elements in the production process. In other words, firms need effective andproductive employees whose job performance is at an optimal level to be able to

C.A. Arnolds, Department of Business Management, University of Port Elizabeth, PO Box1600, Port Elizabeth, South Africa (tel: -F041 5042692; fax: +041 5832644; e-mail:[email protected]). Christo Boshoff, Department of Business Management, University ofPort Elizabeth, PO Box 1600, Port Elizabeth, South Africa (tel: +041 5042577; fax: +0415832644; e-mail: [email protected]).

The International Joumal of Human Resource ManagementISSN 0958-5192 print/ISSN 1466-4399 online © 2002 Taylor & Francis Ltd

http://www.tandf.co.uk/journalsDOI: 10.1080/09585190210125868

Page 2: Compensation, esteem valence and job performance: an empirical

698 The International Journal of Human Resource Management

achieve their organizational goals (Kopelman, 1986; Spence, 1983). The link betweenemployee motivation and job performance is the focus of this study.

Statement of the problem

The need theories have been the focus of much of the research on motivation (Stahl,1986: 39), because they have been seen as among 'the most enduring ways tounderstand motivation' (Aram and Piraino, 1978: 79). Need theory suggests thatemployees are motivated to increase their job performance by their individual strivingto satisfy certain needs. Understanding what the needs are and how they are satisfiedwill enhance insight into work-related behaviours that increase job performance (Steinand Hollowitz, 1992: 20). To this end, the need theory of Alderfer (1967, 1969) hasbeen selected to assess the infiuence of need satisfaction on job performance of topmanagers and frontline employees.

The Alderfer need theory

One of the most widely used of the need theories, is Maslow's (1943) needs hierarchy(Bryan, 1983: 81; Hoffman, 1988: 79; Stahl, 1986: 39). Maslow contends that man hasfive basic categories of needs, which are ranked and satisfied in order of importance.The physiological needs, which are regarded as the most basic in all human beings,must be satisfied first. They are then followed by safety and security needs, affiliationor love needs, esteem needs and self-actualization needs respectively. AlthoughMaslow's theory is intuitively appealing, various criticisms have been levelled at it (DeCenzo and Robbins, 1988; Steers and Porter, 1991). One of the most important of theseshortcomings is that it is a broad theory of human development rather than a descriptionof work motivation (Landy, 1985).

Alderfer (1967, 1969) attempted to address the shortcomings in Maslow's theory byaligning the needs hierarchy with empirical research (Robbins, 1998). According toAlderfer, man is motivated by three groups of core needs, namely Existence,/?elatedness and Growth needs, hence the name ERG theory. The existence needsinclude the human basic needs necessary for existence, which are the physiological andsafety needs. The relatedness needs refer to man's desire to maintain importantinterpersonal relationships. These are man's social, acceptance, belongingness andstatus desires. The last group of needs is the growth needs, which represent man'sdesire for personal development, self-fulfilment and self-actualization.

Alderfer's ERG theory has not stimulated a great deal of research according toIvancevich and Matteson (1999). The ERG theory is however regarded as a more validversion of the need hierarchy (Robbins, 1998) and has elicited more support fromcontemporary researchers as far as motivation in the work situation is concerned(Luthans, 1998). One of the main strengths of the Alderfer theory is the job-specificnature of its focus. In the theory specific reference is made to pay fringe benefits,relatedness needs from co-workers and superiors and growth need satisfaction at work.Much confusion and lack of conclusive evidence are still evident in the motivationalimpact of these variables on the job performance of employees and particularly theinfluence of pay.

As recently as 1998, Pfeffer (1998) still refers to the six dangerous myths about pay.Pfeffer (1998) reflects on whether or not individual incentive pay improves employeejob performance and whether people generally work for money. Igalens and Roussel(1999) are of the opinion that 'postulates which underlie compensation policies inFrance and which appear in the discourse of chief executive officers, human resource

Page 3: Compensation, esteem valence and job performance: an empirical

Arnolds and Boshoff: Compensation, esteem valence and performance 699

managers, consultants and politicians have never been based on hypotheses tested byfield research'. More importantly, research has emphasized what the content of thecompensation package should be in order for it to motivate employees, but no researchhas been done on the social psychology of the package (Heath, 1999: 26).

The social psychological impact of the compensation package refers to the questionof how well the principal (the manager who motivates) understands the agent (theemployee being motivated). The social psychology of compensation is important,because if the principal infers wrong motivations about the agent, compensationpackages can be misdirected. Heath (1999) found that these lay theories of motivationoften result in an overemphasis of extrinsic incentives in motivation strategies. Socialpsychology also suggests that external rewards, such as pay, will affect behaviour onlyif people have an intemal desire for these rewards and that intrinsic rewards willinfluence behaviour only when something in the external environment makes thatbehaviour worthwhile (Heath, 1999: 27). This means that the influence of rewards onvariables within people, such as their personality, beliefs, values, etc., could play asignificant role in whether they will be motivated or not.

It appears that the last word on the motivation of employees has not yet been spoken.Ettorre (1999: 8), for instance, still asks the question: is salary a motivator? Manycompanies are still struggling with motivational issues and bemoan the job-hopping andconcomitant brain drain of skilled people from their companies (Bennett, 2000: 2). Toovercome this problem and following the notion of different motivational strategies fordifferent people, the idea of customized compensation packages for employees is oftensuggested (Scharge, 2000: 274). Igalens and Roussel (1999: 1016), however, could notfind empirical support for the proposal that employees on various organizational levelsdiffer in how they value flexible pay and fringe benefits.

Against this background, the present study attempts to make a more completeassessment of the influence of need satisfaction on work behaviour by revisiting theAlderfer ERG theory. To achieve this purpose the interrelationship between theAlderfer needs, personality and job performance is investigated.

Empirical assessment of the Alderfer theory

Previous empirical evaluations of Alderfer's theory have focused primarily on theinvestigation of the correlational relationship between elements of the theory, on theone hand, and work behaviours, on the other hand (Fox et al., 1993; Salancik andPfeffer, 1977; Wanous and Zwany, 1977). It has been suggested that a more appropriateway of testing motivation theories would be to investigate the causal linkages betweenthe content of these theories and work behaviours (Wahba and Bridwell, 1976). Forinstance, a better empirical test of the Alderfer needs theory would be to ascertain whatneeds caM.se certain behaviours, rather than just showing the correlation between needsand behaviour. Results of causal investigations can, for example, enable those in chargeof motivating employees to compile motivation packages that are tailor-made forspecific individuals or groups of employees. Investigating causal linkages should alsocontribute to addressing the concern of Luthans (1998: 170) that need theories do notnecessarily predict work behaviour. As needs are often rooted in the social selves,beliefs, attitudes and values of people (Schein, 1971), need satisfaction has the potentialof consistently predicting work behaviour. The need satisfactions that cause the desiredwork behaviour, however, need to be identified first. The present study addresses thisresearch gap by empirically assessing the causal relationship between the Alderfer

Page 4: Compensation, esteem valence and job performance: an empirical

700 The International Journal of Human Resource Management

motivation theory, on the one hand, and employee job performance (as work behaviour)of top managers and frontline employees, on the other hand.

One of the most recent investigations into the causal relationship between variousneed satisfactions, as proposed by the Alderfer motivation theory, and employee jobperformance, was conducted by Arnolds and Boshoff (2000). Job performance wasmeasured by eliciting the performance intentions of top managers and frontlineemployees. Frontline employees are employees who interact directly with clients orcustomers and who are not supervisors or managers. Examples of frontline employeesare bank tellers, secretaries, sales clerks, security personnel, office staff, libraryassistants, catering staff, etc. Several researchers have described performance intentionsas a potent measure or predictor of employee job performance (Carkhuff, 1986; Shoreet al., 1990; Sumerlin and Norman, 1992).

Arnolds and Boshoff (2000) found that the satisfaction with pay exerts a significantpositive influence on the job performance of both top managers and frontlineemployees. The satisfaction with fringe benefits and relatedness needs from superiorsdid not infiuence the performance of top managers and frontline employees sig-nificantly. The influence of relatedness need satisfaction from peers, however, had asignificant positive infiuence on the job performance of frontline employees. Thesatisfaction of growth needs impacted significantly on the job performance of topmanagers, but not significantly on the job performance of frontline employees.

The findings of the Arnolds and Boshoff (2000) study contradict certain generalbeliefs on motivation. For instance, the fact that lower-order need satisfaction (fringebenefits) does not influence the job performance of frontline employees is dissonantwith Maslow's theory, as lower-order needs are supposed to motivate lower-levelemployees. The empirical results also contradict the findings of Hong et al. (1995), thatlower level employees are more concemed about security (which includes fringebenefits) need satisfaction. The results are further in disagreement with Alfred's (1991)findings that blue-collar employees are motivated to increase their job performanceby the satisfaction of higher-order needs (respect, recognition and personaldevelopment).

The Arnolds and Boshoff (2000) study brought to the fore important questions, as faras the motivation of top managers and frontline (lower-level) employees is concerned.First, does the satisfaction of security (fringe benefits) needs not influence the jobperformance of frontline employees as suggested by Maslow's theory? Second, arefrontline employees not motivated by the satisfaction of higher-order needs (growth) asAlfred's (1991) findings suggest? Third, does the satisfaction of fringe benefit needsand the respect from peers and superiors play no important role in the motivation of topmanagers?

The answers to these questions seem to be present in the elements of valence andinstrumentality as proposed by the expectancy theory. According to Victor Vroom's(1964) expectancy theory an employee's job performance is determined by the extent towhich first-level outcomes (higher levels of performance) lead to second-leveloutcomes (need satisfaction, such as praise, friendship, wages etc.). The level ofperformance will, however, be determined by the extent to which these second-leveloutcomes (need satisfaction) are valued by the individual (Steers and Black, 1994). Thisstudy considers whether higher job performance will be attained if needs to which anemployee attaches high valence are satisfied.

Against the background of the above-mentioned questions, another importantvariable, namely individual differences among people, needs to be considered. Anexpressed shortcoming in the research on need theories is the fact that no provision is

Page 5: Compensation, esteem valence and job performance: an empirical

Amolds and Boshoff: Compensation, esteem valence and perfonnance 701

made for individual differences among people. Although a theory such as Maslow's, forinstance, recognizes individual differences in what motivates people (Mullins, 1999:417), research on his theory has often neglected this aspect. No comprehensive studyhas as yet been done on the intervening effect of individual characteristics on therelationship between need satisfaction and job outcomes. The relationship between needsatisfaction and personality variables and their effect on performance outcomes havenot been investigated. This is necessary because according to Sharma (1991) anindividual is said to be an individual because 'he has acquired a distinctive pattern ofcognitive style as a trait of his personality, and a specific level of motivational process'.It is believed in some quarters that knowledge on personality traits, such as locus ofcontrol, authoritarianism, self-monitoring (Smit and Cronje, 1992: 273) and self-esteem(Schein, 1971) enables management to predict the work behaviour of employees moreaccurately.

Personality, need satisfaction and job performance

Schein (1971) suggests that individual attributes, such as personality structure and moreespecially one's self-concept, are important determinants of career perfonnance inorganizations. In combination with social self, beliefs, attitudes and values, self-concepthas a significant impact on an individual's needs (Mullins, 1999: 430). For this reasonthe infiuence of self-esteem, as a personality variable, on people's behaviours is oftencited in the literature (Fox et al, 1993; Levine, 1994).

Self-esteem is:

the evaluation that the individual makes and customarily maintains with regard tohimself; it expresses an attitude of approval or disapproval and indicates the extent towhich the individual believes himself to be capable, significant, successful and worthy. Inshort, self-esteem is a personal judgement of worthiness that is expressed in the attitudesthe individual holds.

(Bums, 1979: 55)

Self-esteem has a significant impact on an individual's needs. An individual's positiveor negative self-esteem will affect needs such as affiliation, esteem and self-actualization. Vroom (1964), for instance, found that the self-esteem of employeescould be reinforced by letting them feel accepted in the firm. Alfred (1991) found thatemployees placed a high premium on dignity and respect as elements of their self-esteem. Enhancing an individual's self-esteem with a continuous outpouring ofappreciation for work well done and allowing an employee ample room for self-determination (autonomy) has been found to increase job performance (Levine, 1994;Shouksmith, 1989).

Research on self-esteem, especially as far as job performance is concerned, isunderpinned by the proposition that individuals with positive self-esteem have higherself-perceived competence, self-image and success expectancy (Miner, 1992). Peoplewith low self-esteem, on the other hand, adhere to norms of low job performance whichinhibit creativity, performance and effective interpersonal relations and conflictresolution in the organizations where they are employed (Korman, 1971; Tharenou,1979).

The extent to which self-esteem can be enhanced through the satisfaction of certainneeds can improve job performance significantly. Levine (1994: 77), for example,found that by allowing ample room for self-determination and in so doing enhancing theemployee's self-esteem, the job perfonnance of such an employee could be increased.

Page 6: Compensation, esteem valence and job performance: an empirical

702 The International Journal of Human Resource Management

NEED SATISFACTION PERSONALITY JOB PERFORMANCE

ALDERFER NEEDS

• Growth• Relatedness• Existence

Self-esteemPerformance

intentions

Figure 1 The hypothesized model

It was also reported by Fox et al. (1993: 690), that pay, when perceived as instrumentalto improving self-esteem, was significantly related to increased job performance.

The preceding review suggests that a positive self-esteem is a personality trait thatpeople attach much value to, because it is associated with valued outcomes such ascompetence, self-image, success, creativity and effective interpersonal relations, amongothers. It is therefore hypothesized that:

HI: Need satisfaction (as measured by the Alderfer needs) exerts a positiveinfluence on self-esteem, which in turn, exerts a positive influence on employeejob perfonnance (as measured by performance intentions).

The hypothesized relationships are graphically depicted in Figure 1.

Objectives

The main objective of this study is to investigate whether the job performance of topmanagers and frontline employees can be improved by satisfying the human needs theyvalue in the work situation. More specifically, the study investigates to what extent apersonality variable (self-esteem) acts as intervening variable moderating the relation-ship between need satisfaction (as modelled by Alderfer) and the performanceintentions (as surrogate measure of employee job performance) of top managers andfrontline employees.

Methodology

The sample

The sampling procedure used in this study was a combination of random sampling, onthe one hand, and convenience and judgemental sampling, on the one hand. Addresslists of a sample of business organizations from the manufacturing, trade and servicesectors in South Africa, which employ fifty and more employees (managementincluded) were randomly selected from the database at the Bureau of MarketingResearch at the University of South Africa (UNISA). One thousand five hundred(1,500) questionnaires were mailed to the chief executive officers of these firms.

Organizations with sufficient numbers of lower level/frontline employees wereidentified on a convenience basis. Employees who fit the definition of a frontlineemployee were selected on a judgemental basis. One thousand (1,000) questionnaireswere distributed in this manner to frontline employees in the banking, retail, securityand legal industries in various South African metropolitan areas. Frontline employees

Page 7: Compensation, esteem valence and job performance: an empirical

Amolds and Boshoff: Compensation, esteem valence and performance 703

were regarded, for the purpose of this study as lower level employees and include thosenon-management groups of employees who are at the frontline of service provision,such as bank tellers, secretaries, sales clerks and security personnel.

Five hundred and seventeen (517) usable questionnaires (an overall response rate of20.7 per cent) were retumed. This was made up of 304 top managers (response rate =20.2 per cent) and 213 frontline employees (response rate = 21.3 per cent). The overalland individual response rates exceeded the absolute minimum response rate of 8 per centneeded in both finite and infinite populations (Martins ef a/., 1996: 271). For the purposesof applying structural equation modelling, the sample size also comfortably exceeds theminimum of 100 respondents per model suggested by Hair et al. (1995: 637).

Out of 304 top management respondents only thirty-five (11.5 per cent) werefemales, while the frontline sample consisted of 114 females and 99 males. This is a fairreflection of the composition of the workforce at these occupational levels in SouthAfrica. The gender sub-samples were, however, too small (Hair et al., 1995) to conducta multivariate analysis based on gender. Such an analysis was therefore not done in thisstudy. For the same small sub-sample reason and due to the highly sensitive nature ofracial classification in South Africa an analysis based on race was not conducted in thisstudy. Respondents often refuse to complete questionnaires that refer to race.

The measuring instruments

The instrument used to measure the extent to which the respondent's job permits thesatisfaction of needs according to the ERG theory was developed by Alderfer (1967),who reported favourable results on the convergent and discriminant validation of theinstrument. Spearman-Brown reliability coefficients ranging from 0.80 to 0.88 werereported for the five ERG sub-scales.

Performance intentions of respondents were assessed using the instrument developedby Shore et al. (1990). One item, namely 'I could do a lot more work if I tried a littleharder', from the Cranny et al. (1992) scale was added to make the performanceintention scale a 4-item one. This was done for reasons of consistency as all other scalescomprised four or more items. Cronbach (1951) reliability coefficients ranging from0.60 to 0.84 were reported for the performance intention scale (Shore et al., 1990).

Rosenberg's (1965) scale was used to measure the self-esteem variable. This measuretaps a uni-dimensional index of global self-esteem and has been reported to have areproducibility index of 0.93, an item scalability of 0.73 and test-retest reliability of0.85 (Burns, 1979). Lopez and Greenhaus (1978) reported adequate convergent anddiscriminant validity levels, and a reliability coefficient of 0.78 for this instrument.

It is evident from the preceding review that instruments with acceptable validity andreliability were used to investigate the variables under scrutiny. Respondents wererequested to respond to all questions in the above-mentioned instruments on a seven-point Likert scale.

Data analysis

Internal reliability and discriminant validity

The internal reliability of the measuring instruments was assessed by calculating theirCronbach alpha coefficients. To assess the discriminant validity of the measuringinstruments exploratory factor analyses were conducted, using the computer programmeBMDP4M (Frane et al., 1990). Maximum likelihood was specified as the method of

Page 8: Compensation, esteem valence and job performance: an empirical

704 The International Journal of Human Resource Management

Table 1 Rotated factor loadings: Alderfer's model: top management sample'

ALFBlALFB2ALFB3ALFB4ALRSlALRS2ALRS3ALRS4ALPYlALPY2ALPY3ALPY4ALRPlALRP2ALRP4ALGRlALGR2ALGR3ALGR4

Eigen values

Factor 1Existencebenefits

0,7440.7650.8260.940

-0.0130.0740.095

-0.0670.0700.1120.0140.0570.044

-0.0510.0700.0900.040

-0.013-0.057

2.770

Factor 2Growth

-0.007-0.0140.060

-0.0320.1000.018

-0.1220.0160.0830.277

-0.028-0.054-0.136

0.1610.1090.5840.7940.6860.648

2.035

Factor 3Relatednesssuperiors

-0.005-0.0600.114

-0.0260.8270.6930.6910.5270.065

-0.0540.0510.109

-0.0420.0520.0770.156

-0.025-0.0610.157

2.023

Factor 4Existencepay

0.0070.124

-0.0520.012

-0.0200.1150.0050.0250.6440.3950.8620.7150.189

-0.025-0.119-0.0700.041

-0.0370.214

1.963

Factor 5Relatednesspeers

0.0810.073

-0.064-0.043-0.054-0.0350.0530.0580.092

-0.0350.047

-0.0180.8230.6240.4100.137

-0.0370.0610.039

1.302

Note' Loadings greater than .03 were considered significant

factor extraction and a Direct Quartimin oblique rotation of the original factor matrixwas used (Jennrich and Sampson, 1966) in all instances.

The extraction of five factors (existence needs-pay, existence needs-fringe benefits,relatedness needs-respect from superiors, relatedness needs-respect from peers andgrowth needs) was specified in the factor analyses of the Alderfer needs for both the topmanagement and frontline employee samples. The empirical evidence supported thesurmized five separate and distinct variables as suggested in the literature.

Based on the factor analyses (and the resultant Tables 1 and 2), Table 3 identifies theitems which were regarded as measures of the individual latent variables retained in thetheoretical models. Only these items (see Table 4 for description) were used in allsubsequent statistical procedures.

Table 3 also indicates the Cronbach alphas of the latent variables that were includedin the final theoretical models. All the Cronbach reliability coefficients were above0.550 and thus above the cut-off point needed for basic research (Pierce and Dunham,1987; Smith et al., 1983; Tharenou, 1993).

Structural equation analysis

It has been suggested that an empirical investigation of the causal relationships betweenneed satisfaction and certain behavioural outcomes could provide a more appropriatetest of motivation theory (Wahba and Bridwell, 1976: 231). As the present study is an

Page 9: Compensation, esteem valence and job performance: an empirical

Arnolds and Boshoff: Compensation, esteem valence and performance 705

Table 2 Rotated factor loadings: Alderfer's model, frontline employee sample'

ALFBlALFB2ALFB3ALFB4ALRSlALRS2ALRS3ALRS4ALPYlALPY2ALPY3ALPY4ALRPlALRP2ALRP3ALRP4ALGRlALGR2ALGR3ALGR4

Eigen values

Factor 1Existencebenefits

0.7400.7650.8160.8650.0450.0380.136

-0.1140.055

-0.0340.0660.059

-0.1460.107

-0.0780.1260.019

-0.0390.1260.069

2.672

Factor 2Relatednesssuperiors

Q.Qll-0.0010.029

-0.0630.7730.8300.6200.4610.0120.032

-0.005-0.016

0.2660.022

-0.1760.170

-0.003-0.064

0.0370.182

2.064

Factor 3Growth

-0.0330.0060.0950.076

-0.032-0.061

0.0590.183

-0.0480.191

-0.136-0.026

0.0780.0180.093

-0.0500.5430.8560.6160.662

1.979

Factor 4Existencepay

0.1090.022

-0.0150.0540.044

-0.024-0.020

0.0970.6990.7150.6060.6740.1190.0600.0010.024

-0.0840.0780.0380.040

1.883

Factor 5Relatednesspeers

-0.0580.0370.037

-0.033-0.067-0.028

0.184-0.025-0.068-0.075

0.1390.1010.4440.5270.7380.3550.226

-0.077-0.040

0.131

1.307

Note' Loadings greater than 0.3 were considered significant

Table 3 The empirical factor structure

Latent variableMeasuringvariables

FinalCronbachalpha

Alderfer existence needs (pay)Alderfer existence needs (fringe benefits)Alderfer relatedness needs (superiors)Alderfer relatedness needs (peers)Alderfer growth needsPerformance intentionsSelf-esteem

ALPYALFBALRSALRPALGRPERFESTE

1,2,3,41,2,3,41,2,3,41,2,3,41,2,3,4

1,2,3,41,2,3,4

0.7880.8920.7880.6530.7890.7270.556

attempt to assess the causality of need satisfaction according to the Alderfer theory,structural equation modelling was considered as the appropriate technique for thispurpose.

Structural equation analysis, or modelling (SEM), is a multivariate techniquecombining aspects of multiple regression and factor analysis to estimate a series of

Page 10: Compensation, esteem valence and job performance: an empirical

706 The International Journal of Human Resource Management

Table 4 Description of scale items

Alderfer existence needs (pay)ALPY I I get enough money from my job to live comfortablyALPY 2 My pay is adequate to provide for the basic things in lifeALPY 3 Considering the work required the pay is what it should beAPLY 4 Compared to the rates for similar work here my pay is goodAlderfer existence needs (fringe benefits)ALFB 1 Our fringe benefits cover many of the areas they shouldALFB 2 The fringe benefit programme here gives nearly all the security I wantALFB 3 Compared to other places, our fringe benefits are excellentALFB 4 The fringe benefit programme here is adequateAlderfer relatedness needs (superiors)ALRS 1 My boss encourages people to make suggestionsALRS 2 My boss takes account of my wishes and desiresALRS 3 My boss keeps me informed about what is happening in the companyALRS 4 My boss lets me know when I could improve my performanceAlderfer relatedness needs (peers)ALRP I 1 can count on my co-workers to give me a hand when I need itALRP 2 My co-workers will speak out in my favour if justifiedALRP 3 1 can tell my co-workers honestly how I feelALRP 4 My co-workers welcome opinions different from their ownAlderfer growth needsALGR 1 I always get the feeling of leaming new things from my workALGR 2 My job requires that a person use a wide range of abilitiesALGR 3 My job requires making one or more important decision(s) every dayALGR 4 1 have the opportunity to do challenging things at workPerformance intentionsPERF 1 I often think of improving my job performancePERF 2 I will actively try to improve my job performance in the futurePERF 3 I intend to do a lot more at work in the futurePERF 4 I will probably do my best to perform well on the job in the futureSelf-esteemESTE 1 I wish I could have more respect for myselfESTE 2 I feel that I am a person of worth, at least on an equal plane with othersESTE 3 1 take a positive attitude toward myselfESTE 4 I certainly feel useless at times

interrelated dependence relationships simultaneously (Hair et al., 1995). The benefit ofSEM is that it is a technique that allows for 'stating theory more precisely, testingtheory more precisely' (Hughes et al., 1995: 129). The technique has three particularlysalient characteristics, namely the ability to estimate multiple and interrelateddependence relationships; the ability to represent unobserved concepts or variables inthese relationships; and the ability to account for measurement error (Hair et al., 1995:622; Hoyle, 1995: 10). SEM is therefore regarded as a more comprehensive, advancedand rigorous statistical technique to analyse attitudinal data than other techniques, suchas analysis of variance (ANOVA) and multiple regression (Hoyle, 1995).

It is true that causation is less well defined in the behavioural sciences than inphysical processes (Hair et al., 1995: 626-7). Thus, although some of the strict classicalprovisions for causation may not be met in the behavioural sciences, 'strong causalassertions can possibly be made if the relationships are based on a theoretical rationale'(Hair et al., 1995: 627). It is for this reason that SEM is recommended for testing and

Page 11: Compensation, esteem valence and job performance: an empirical

Arnolds and Boshoff: Compensation, esteem valence and performance 707

confirming theoretical relationships rather than for exploratory purposes or thegeneration of new theories. In other words, SEM can confirm that a theoretical modeldoes (or does not) approximate the data. It does not imply that no other plausiblemodels exist.

Despite possible limitations, SEM is therefore generally accepted as measuringcausal relationships (Hair et al., 1995: 626-7; Hoyle, 1995: 10), because of the above-mentioned unique improvements, for instance, on multiple regression and, moreimportantly, because SEM models are based (or should be) on sound theory. It is alsobelieved that a claim of causality can be made in the present study, because the criteriafor causation (Bollen, 1989: 40) have been met. In other words, there is an associationbetween the hypothesized variables; there is evidence of at least what Bollen (1989)calls 'pseudo-isolation'; and there is strong theoretical support for the directionality ofthe need satisfaction-performance relationship.

Tbe empirical results

The influence of the satisfaction of the Alderfer needs on self-esteem and self-esteemon job performance of top managers

Figure 2 depicts the causal model constructed to investigate the influence of thesatisfaction of the Alderfer needs on employee job performance via the effect of theformer on self-esteem as an intervening variable. The computer programme RAMONA(Browne and Mels, 1996) was used to analyse the causal model and the results thereofare reported in Figure 2.

The empirical results confirm the importance of self-esteem as a significantdeterminant of job performance. Figure 2 shows that self-esteem significantly (0.583,p<0.0\) influences the performance intentions (the surrogate measure of jobperformance) of top managers.

The results indicate that growth need satisfaction was significantly (0.760, p < 0.01)related to self-esteem. The satisfaction of existence needs (pay and fringe benefits) andrelatedness needs (to peers and superiors) has no significant influence on self-esteem.The results reveal that the job performance of top managers can be increased byenhancing their self-esteem via the satisfaction of their growth needs. Providing foropportunities for creativity, self-fulfilment, advancement and autonomy for topmanagers in the firm could thus significantly improve their job performance.

The results also suggest that the positive self-esteem of a top manager is notreinforced by the manager's perception of the extent to which the firm satisfies his/herneed for pay and fringe benefits. The satisfaction of relatedness needs from peers andsuperiors also does not improve a top manager's self-esteem. It appears that the resultssupport the theory of Herzberg et al. (1959) that lower-order needs, such as pay, fringebenefits and relatedness (affiliation) needs, are all hygiene factors, which eliminatedissatisfaction, but do not motivate top-level employees. According to Herzberg et al.(1959), top-level managers are motivated by higher-order needs, such as growth,advancement and achievement.

The empirical results (path coefficients) indicate that the hypothesis (HI), that needsatisfaction (as measured by the Alderfer needs) exerts a positive infiuence on self-esteem, which, in tum, exerts a positive influence on employee job performance (asmeasured by performance intentions), is supported as far as growth need satisfaction is

Page 12: Compensation, esteem valence and job performance: an empirical

708 The International Journal of Human Resource Management

1 Unless indicated otherwise, all the relationshipsare significant at the/;<0.01 level of significance.NS denotes non-significant relationships.

Figure 2 The influence of need satisfaction on employee job performance with self-esteem asintervening variable: Alderfer's theory - top management

concemed. The hypothesis (HI) is, however, not supported as far as the satisfaction ofrelatedness needs (peers and superiors) and existence needs (pay and fringe benefits) isconcemed.

Page 13: Compensation, esteem valence and job performance: an empirical

Amolds and Boshoff: Compensation, esteem valence and performance 709

The influence of the satisfaction of the Alderfer needs on self-esteem and self-esteemon job performance of frontline employees

The causal model, constructed to investigate the influence of the satisfaction of theAlderfer needs on employee job performance of frontline employees via theenhancement of self-esteem as an intervening variable, is depicted in Figure 3. Thecomputer programme RAMONA (Browne and Mels, 1996) was again used to analysethe causal model and the results are reported in Figure 3 as well.

Similar to the results of the top management sample (Figure 2), self-esteem emergedas a significant determinant of employee job performance. The empirical results (0.362,p < 0.01) concur with previous research that the job performance of employees can besignificantly increased if a positive self-esteem is inculcated and nurtured.

The empirical results reveal that the satisfaction of growth needs had a significantinfiuence (0.243, p < 0.05) on self-esteem and thus an indirect impact on performanceintentions of frontline employees. It is important to note that Amolds and Boshoff(2000) reported that growth need satisfaction had no direct infiuence on theperformance intentions of frontline employees. The results therefore suggest thatteaching frontline employees new things on the job, using a wide range of their abilities,allowing them to make more decisions everyday and providing for opportunities to dochallenging things at work, will enhance their self-esteem and consequently theirperformance intentions.

Figure 3 shows that satisfying the need for respect from superiors is negatively(—0.314, p < 0.01) related to the self-esteem of frontline employees. It would seem thatthe frontline employees in this sample do not seek the acceptance of their superiors, butrather the respect and acceptance of their peers. It could also mean that employees areafraid of being ostracized by their co-workers for seeking the favour of management.The satisfaction of relatedness needs from peers had a significant positive effect (0.484,p<O.Ol) on self-esteem and therefore on performance intentions. The respect of co-workers seems more important to frontline employees than the acceptance ofsupervisors. Endeavours to improve commitment to superiors would not increaseperformance intentions among frontline employees. Strategies to improve work groupcohesion and the image of the employee among his/her co-workers could be the recipefor increased job performance. Self-esteem can be reinforced by effective recognitionsystems geared at highlighting employee achievements.

For frontline employees, the satisfaction of the need for pay does not affect self-esteem. Amolds and Boshoff (2000) found a direct relationship between the satisfactionwith pay and the job performance of frontline employees. The empirical results alsoshow that the satisfaction of fringe benefits does not have a significant influence onperfonnance intentions via self-esteem as an intervening variable. Amolds and Boshoff(2000) also reported that fringe benefit satisfaction does not influence performanceintentions significantly. The Alderfer model therefore appears to support Herzberg etal.'s (1959) theory that pay and fringe benefits are hygiene factors, which prevent jobdissatisfaction, but do not motivate employees to increase job performance.

The results show that the satisfaction of the related needs and growth needs assuggested by the Alderfer theory significantly influences the self-esteem of frontlineemployees. Self-esteem, in tum, influences the job performance of frontline employees.The hypothesis (HI) that need satisfaction positively influences employee jobperformance via its effect on personality variables (self-esteem in this case) is thussupported as far as the satisfaction of growth needs and relatedness needs from peers isconcemed. Hypothesis HI is, however, rejected as far as the satisfaction of relatedness

Page 14: Compensation, esteem valence and job performance: an empirical

710 The International Journal of Human Resource Management

1.0

I Unless indicated otherwise, all the relationshipsare significant at the/)<0.01 level of significance.NS denotes non-significant relationships.

Figure 3 77ie influence of need satisfaction on employee job performance with self-esteem asintervening variable: Alderfer's theory -frontline employees

needs from superiors is concerned, because the results show a negative relationshipbetween the satisfaction of relatedness needs from superiors and self-esteem as a

Page 15: Compensation, esteem valence and job performance: an empirical

Arnolds and Boshoff: Compensation, esteem valence and performance 111

personality variable. Hypothesis HI is also rejected with regard to the satisfaction ofexistence needs for pay and fringe benefits. There was no significant causal relationbetween these need satisfactions and the self-esteem of frontline employees.

The indices of fit

The fit indices of the causal models were examined to establish the extent to whichthese models represent acceptable approximations of the data. To this end various fitindices (Hair et al., 1995) were calculated.

Table 5 shows that the relationship between the chi-square, degrees of freedom andexceedance probability values of the models indicates that the Alderfer model does notrepresent an acceptable fit between the observed and predicted matrices. An exceedanceprobability (significance) level of above 0.1 shows a non-significant difference betweenthe observed and predicted matrices (Hair et al., 1995: 683) and therefore an acceptablemodel fit. Table 5 indicates that the Alderfer model shows no statistical difference(significance level = 0.001) between the observed and predicted input data. Against thisbackground the hypothesized fit of the Alderfer model must be rejected. Hair et al.(1995: 683-4), however, caution that the chi-square exhibits the shortcoming ofindicating significant differences between the actual and estimated matrices as samplesizes become larger than 200 respondents. The sample size in the present study is 213respondents. Due to this shortcoming more improved fit indices, namely the root meansquare error of approximation (RMSEA) and the goodness-of-fit (GFl) indices wereconsidered.

The RMSEA and GFl indices reported in Table 5 meet the minimum acceptablestandards proposed by Hair et al. (1995: 689). The RMSEAs of 0.068 and 0.065(MacCullam et al., 1996) of the two models, for example, indicate that both modelsrepresent a reasonable fit of the responses from the sample from which they weredrawn. The GFl, believed to be one of the best absolute indices of model fit (Hoyle,1995), indicates the overall degree of fit of the hypothesized model on the data. HighGFl values in the range from nil (0) to one (I), such as those reported in Table 5, areindicative of reasonable or acceptable model fit.

The normed chi-square (x^) index, which provides an indication of the extent towhich model fit has been achieved by 'over fitting' the data, shows acceptableparsimony. Although the normed x^ is subject to the same shortcoming of the chi-square as discussed above, the Alderfer (1.906) model produced a parsimony levelwhich falls within the acceptable limits of more than 1.0 and less than 2.0 as suggestedby Hair era/ . (1995: 687).

Table 5 The absolute fit indices

Index

RMSEALISREL GFlNormed x̂Chi-squareDegrees of freedomExceedance probability

Alderfertopmanagement

0.0680.8492.396

737.88308

0.000

Alderferfrontlineemployees

0.0650.8321.906

636.80334

0.001

Page 16: Compensation, esteem valence and job performance: an empirical

712 The International Journal of Human Resource Management

Summary of empirical findings

The empirical results revealed that top managers are primarily motivated by growthneeds, in other words, higher-order needs. This means that a challenging workingenvironment that provides opportunities for creativity, self-fulfilment, advancement andautonomy is a key motivator of the job performance of top managers. The need forautonomy, creativity and advancement can be satisfied by putting the manager in chargeof projects that have to be driven from the conceptual phase to the completion phase. Itis believed that a person often experiences the feeling of self-fulfilment and creativityenhancement when a person starts and finishes a task.

Creativity must be actively managed by engendering an organizational culture thatpromotes a fair and constructive judgement of ideas, rewards and recognizesprogrammes for creative work and provides for mechanisms for the development andactive flow of ideas (Amabile, 1997). Role models and mentors who set goals, supportcolleagues, value the contributions of top managers and show confidence in topmanagers need to be identified and inspired to play this role. Creativity teams, whichinclude top managers and diversely skilled employees from various organizationallevels, need to be established and supported with the necessary resources in terms offunds, materials, facilities and information. Impediments to creating a culture ofcreativity, such as unhealthy internal politics in the firm, destructive criticism of newideas, negative intemal competition, the tendency to avoid risk and the over-emphasison maintaining the status quo, must be avoided, reduced or eliminated.

The empirical results suggest that the performance intentions of top managers wouldbe affected by their perceptions of advancement opportunities and personal growth.This is an indication to the human resources directors in charge of identifying topmanagement potential in the firm, to pay particular attention to promotion policies andprogrammes of their firms. In this regard, personal and organizational plateauing mustbe managed effectively (McCampbell, 1996: 63).

The results show that frontline employees are primarily motivated by the satisfactionof relatedness needs from peers and existence needs and particularly monetarycompensation. Relatedness from peers need satisfaction is a direct motivator, as well asan indirect motivator via its effect on employee self-esteem. Remuneration and thesatisfaction with fringe benefits do not enhance the self-esteem of frontline employees.They seem to be hygiene factors that must be in place to prevent employeedissatisfaction.

According to Maslow's theory, higher-order needs, such as growth needs, do notusually motivate lower-level employees, including frontline employees. The empiricalresults reported in this study suggest, however, that higher-order needs such as growthneeds can motivate frontline employees via self-esteem enhancement of suchemployees.

In terms of the expectancy theory it appears as if the satisfaction with pay and fringebenefits does not influence the performance intentions of frontline employees, becausethese need satisfactions do not have any esteem valence for these employees. In otherwords, frontline employees do not have a higher regard of themselves if their firms ofemployment pay them enough to acquire basic necessities such as a house, furniture andclothes. Their self-esteem is also not enhanced by security associated with fringebenefits. These working conditions and fringe benefits that satisfy these needs must,however, be in place to prevent job dissatisfaction according to the empirical findingsand in concurrence with Herzberg's theory.

Page 17: Compensation, esteem valence and job performance: an empirical

Amolds and Boshoff: Compensation, esteem valence and performance 113

The second finding relates to the question of whether lower-level employees, such asfrontline employees, are never motivated by higher-order needs. Amolds and Boshoff(2000) found that the satisfaction of higher-order needs, such as self-actualization, doesnot influence the performance intentions of frontline employees significantly. Theresults of the Alderfer model, however, indicate that frontline etnployees associategrowth need satisfaction (learning new things on the job, optimal use of abilities,participation in decision making and opportunities for challenging work) with self-esteem enhancement. The results suggest that the satisfaction of these needs does notimpact directly on the job performance of frontline employees. Tbe satisfaction of thesegrowth needs will however improve job performance if they reinforce the self-esteem ofthese employees in a way that increases the acceptance, respect and status theseemployees receive from their peers.

The empirical results in the present study show that managers would err if theyunquestioningly accept that lower-level employees are not motivated by higher-orderneeds. The results suggest that higher-order need satisfaction does infiuence the jobperformance of lower-level employees, provided that the motivation strategies directedat these higher-order needs are correctly implemented. Motivational interventionsshould be clearly defined in terms of the goals and end results they are directed atachieving. In other words, any managerial intervention (training, job enrichment,recognition systems) aimed at higher-order need satisfaction will have to have 'esteemenhancing properties' to be successful in improving job performance. Efforts to usehigher-order need satisfaction as a motivational tool will simply fail if frontlineemployees do not believe that their participation will enhance their self-esteem.

Managerial implications

A holistic overview

The impact of individual traits, such as personality characteristics for example, hasoften been ignored in past studies of motivation. For this reason, managers' motivationinterventions based on these studies could therefore also have been ineffectual.

The empirical results of the present study show that an individual's personality (self-esteem in this study) can exert a significant intervening influence on the relationshipbetween need satisfaction and work behaviour. This indicates to tnanagers thatdifferentiation in rewards is necessary when motivation interventions are made. This isonly possible when managers know their employees. This finding therefore suggeststhat efforts must be made to understand employees at a more intimate level.Management by walk-around, open door communication policies, social events andpersonal interviews are activities management dare not neglect in order to rewardemployees effectively.

The results also show that the self-esteem of frontline employees is not enhanced bymonetary rewards. Frontline employees are therefore not motivated to increase their jobperformance because the pay they receive enhances their self-image. It appears thattheir job performance is directly linked to their remuneration package. These employeeswant to receive a fair and equitable pay for a fair day's work, in other words, share inthe gains of their productivity. This brings to the fore the new buzz word in jobperformance improvement, namely gainsharing. Gainsharing is an approach to enhanceorganizational effectiveness which embraces the whole firm through a formal system ofemployee involvement and financial bonuses based on perfonnance gains. Pay is butone way in which employees can perceive gainsharing. Hanlon and Taylor (1991:

Page 18: Compensation, esteem valence and job performance: an empirical

714 The International Journal of Human Resource Management

238-56) found that gainsharing improves employee productivity, organizationaleffectiveness, positive communication and problem-solving activities by individualsand groups. Managers therefore need to regularly evaluate employee remunerationpackages against the background of the gainsharing principle.

The empirical results also reveal that the satisfaction of relatedness needs from peersis a significant causal determinant of the employee job performance of frontlineemployees. This result indicates that frontline employees value the acceptance,belongingness and respect of their peers in their working groups. Downsizing has beenshown to be a destroyer of work groups and thus employee goodwill, loyalty andmorale. The effects of downsizing on survivors in enterprises include worries aboutlong-term job security, a deep sense of loss, grief and depression and a drop incredibility and trust in management on the part of survivors (Clark and Koonce, 1995).Negative perceptions about superiors were also evident in the Alderfer tnodel in thepresent study. The empirical results showed that the satisfaction with relatedness needsfrom superiors exerted a negative influence on the self-esteem of frontline employees.These findings possibly indicate the necessity of a rethink of the issue of job security onthe part of management in business enterprises. Slabbert (1997: 16) believes, ifunemployment goes unchecked its social, economic and psychological consequencescould spell the demise of mankind. Profit-maximization at the expense of job securityis a recipe for disaster. Slabbert (1996: 49) therefore suggests that 'reduced profit andredistribution of generated wealth over a wider spectrum of humanity are morally andeconomically sound' and should provide the philosophical foundation of managementthinking in the future.

The empirical results also show that the motivation of top managers must not beneglected. Top managers are responsible for determining the vision and future directionof an enterprise. They are the ones who must develop the long-term plans and strategicthrust of firms and are therefore immensely important to the growth and survival offirms. For this reason they are often headhunted by competitors. A survey by a humanresource consultancy firm, FSA-Contact, in South Africa, as reported in the BusinessTimes of 16 January 2000, showed that 47 per cent of engineering manufacturing top-level staff leave their jobs for improved career opportunities elsewhere. The empiricalresults of the present study suggest that the retention of top executives can besignificantly increased by the provision for growth need satisfaction in firms. Topmanagers need challenging assignments that provide for innovativeness, autonomy andpersonal growth. Matching these need satisfactions with the strategic goals of firms cansignificantly impact the growth and survival of firms.

Pay as a motivator

The empirical results show that the performance intentions of both chief executiveofficers and frontline employees are significantly influenced by their satisfaction withpay. It appears that this relationship suggests that employees want to see their monetarycompensation as fair and equitable reward for the effort they put in. Their satisfactionwith pay, however, has no direct causal influence on the respect the employee has forhimself, his perception of whether he is on an equal plane with others, whether he holdsa positive attitude towards himself or whether he feels useful or not. In other words, payhas no direct influence on an employee's self-esteem, which means that it does notmatter whether the employee is paid adequately or not, his self-esteem will not beaffected by pay.

Page 19: Compensation, esteem valence and job performance: an empirical

Amolds and Boshoff: Compensation, esteem valence and performance 115

An employee's compensation package is thus much more than a cash reward for apiece of work completed. A paradigm shift in management thinking may be required toimprove the effectiveness of motivational interventions. The empirical results suggestthat pay is only important to the extent that it is instrumental in satisfying higher-orderneeds, growth needs in the case of top managers and respect and acceptance by co-workers and growth needs in the case of frontline employees. To employees their payis a barometer which shows that the firm values their contributions and that their inputscontribute to the prosperity of the firm.

Pay is therefore a much stronger motivational force when linked to higher-orderneeds. The empirical results confirm the proposals by Glassock and Gram (1995: 97-9)that rewards, such as handwritten thank-you cards, tokens ('funny money', exchange-able in the firm's cafeteria, health centre, day care centre or store, but never for cash),team celebrations during working hours, nights on the town whereby achievers and theirfamily members are wined and dined by the firm, and I-made-a-difference-awardswhich allow team members or individuals to go on a shopping spree or an extravagantdining experience, could have a stronger motivational impact on the work behaviour ofemployees than cash rewards. The empirical findings also support the belief of AlfieKohn, as cited by Levine (1994: 78), that employees must be paid well and fairly andthat everything possible then be done to help them forget about money. Kohn suggeststhat this could be achieved by promoting teamwork, participation and genuine interestin the job.

The self-esteem perspective

The present study has a significant contribution to the demystification of motivation.Managers are perplexed on a daily basis by why well-paid top and middle managersresign from firms or are just not performing as expected. It is also not alwaysunderstood why lower-level employees, who are not paid big salaries, are veryenthusiastic about their jobs and never leave them even when they receive better joboffers. The empirical results suggest that the answer could lie in the personalityvariable, self-esteem. In the present study the self-esteem of both top managers andfrontline employees has a significant positive infiuence on their performance intentions.Shifting compensation specialists' focus from what need satisfactions cause improvedjob performance to identifying what need satisfactions are important for the self-esteemenhancement of employees could be an important shift in motivational thinking.

For chief executive officers the satisfaction of growth needs influences their self-esteem considerably. These top managers need challenging assignments, continuoustraining and development and the freedom and space to plough back their skills andexpertise in the upliftment of their communities. The self-esteem of frontline employeesis strongly influenced by the respect and acceptance from their peers. Any reward thatwould enhance the image, respect and acceptance of the frontline employee by theirpeers would motivate them to higher job performance. This could be achieved by theimplementation of various recognition tools and job enrichment strategies.

Frontline employees also need to feel that there are always new things to learn atwork. They would also like to make one or more important decisions every day, use awide range of their abilities and have the opportunity to do challenging things at work.These are all growth needs which were shown to have a significant influence on theself-esteem of frontline employees in the study.

In summary, this study shows that perceptions of pay influence the performanceintent of both chief executive officers and frontline employees. Compensation via the

Page 20: Compensation, esteem valence and job performance: an empirical

716 The International Journal of Human Resource Management

satisfaction of higher-order needs has a stronger motivational impact on the jobperformance of these categories of employees. It is the higher-order needs that improvejob performance via self-esteem enhancement and not monetary compensation. Finally,self-esteem is a strong determinant of job performance. Identifying the need satisfactionthat enhances self-esteem will improve employee job performance and in so doingimprove the effectiveness and profitability of enterprises.

Limitations of tbe study

The present study has made significant contributions to the body of knowledge onmotivation and employee job performance. Certain areas, however, still need to beexplored or expanded. These areas include the application of the motivation models toblue-collar employees and the inclusion, respectively, of determinants or antecedents ofneeds (such as age, gender, marital status, education, personal values, culturalbackground, kinship responsibility, supervisory style, and promotion opportunities) andpersonality variables (such as locus of control, self-monitoring and risk proneness) inthe theoretical models to improve job performance. It could, for instance, behypothesized that older frontline employees would be more concerned about the needfor self-actualization (growth need) the satisfaction of which will affect their self-esteem and performance intent differently from younger frontline employees. Femalescan for example be more concemed about the satisfaction of working and materialconditions (the existence need for pay and fringe benefits) that would allow them tocarry out their maternal duties. Hong et al. (1995: 14), on the other hand, found malesto be more concemed with relatedness and growth needs, entertainment, education andtraining in particular. For males the satisfaction of these needs could therefore have astronger infiuence on their self-esteem and performance intentions. In a similar fashion,differences in cultural socialisation can cause different influences on self-esteem andjob performance of different cultural groups.

As far as the inclusion of personality variables in theoretical models to improve jobperformance is concemed, locus of control has been reported to be an importantdeterminant of needs. Cunningham et al. (1995: 41), for instance, reported that extemallocus of control is negatively related to the need for recognition (esteem need) andinnovativeness (growth need). Allen et al. (1997: 123), on the other hand, foundsignificant positive correlations between internal locus of control, on the one hand, andthe need for friendship of co-workers (affiliation need), self-respect, responsibility,achievement (esteem needs), growth and development of new skills (growth needs).Against this background, the impact of different need satisfactions on the locus ofcontrol of 'extemals' and 'intemals' could have a differential impact on their respectiveperfonnance intentions. The inclusion of other personality variables, such as locus ofcontrol, could therefore increase our understanding of the interactive nature of needsatisfaction, personality and job performance.

The measures of fit in Table 5 suggest that the data fit the theoretical modelreasonably well. In other words, the theoretical model is a very plausible explanation ofthe relationship between need satisfaction, self-esteem and performance intentions. Thisdoes not mean that it is the only model that explains these relationships. Other equallyplausible models can also be constructed. The present study cannot and does not claimthat self-esteem is necessarily an antecedent of job performance, or the only antecedentof job performance, or that job performance will not influence self-esteem. The study(based on the empirical findings) can however claim that, for the present sample, thesatisfaction of growth needs will enhance the self-esteem and the job performance

Page 21: Compensation, esteem valence and job performance: an empirical

Arnolds and Boshoff: Compensation, esteem valence and performance 111

intentions of top managers and frontline employees; that the satisfaction of the need forrelatedness to peers will enhance the self-esteem and job performance of frontlineemployees; and that the satisfaction of the need for relatedness to supervisors willnegatively affect the self-esteem and job performance of frontline employees.

References

Alderfer, C.P. (1967) 'Convergent and Discriminant Validation of Satisfaction and DesireMeasures by Interviews and Questionnaires', Joumal of Applied Psychology, 51(6): 509-20.

Alderfer, C.P. (1969) 'An Empirical Test of a New Theory of Human Needs', OrganizationalBehaviour and Human Performance, 4: 142—75.

Alfred, M. (1991) 'A View from the Other Side', Productivity SA, 17(3): 7-11.Allen, R.E., Lucero, M.A. and Van Norman, K.L. (1997) 'An Examination of the Individual's

Decision to Participate in an Employee Involvement Program', Group and OrganizationManagement, 22(1): 17-143.

Amabile, T.M. (1997) 'Motivating Creativity in Organizations: On Doing What You Love andLoving What You Do', California Management Review, 40(1): 39-58.

Aram, J.D. and Piraino, T.G. (1978) 'The Hierarchy of Needs Theory: An Evaluation in Chile',Interamerican Journal of Psychology, 12(2): 179-88.

Amolds, C.A. and Boshoff, C. (2000) 'Does Higher Remuneration Equal Higher Job Perform-ance? An Empirical Assessment of the Need-Progression Proposition in Selected NeedTheories', South African Journal of Business Management, 31(2): 53—64.

Bennett, J. (2000) 'Salaries Not the Only Drawcard', Sunday Times Business Times, 26 March:2.

Bollen, K.A. (1989) Structural Equations with Latent Variables. New York: Wiley.Browne, M.W. and Mels, G. (1996) Path Analysis: RAMONA, SYSTAT6.0for Windows: Statistics.

Chicago, IL: SPSS Incorporated.Bryan, L.A. (1983) 'From Whence the Theory?', Training, 20(June): 81-3.Burns, R.B. (1979) The Self-Concepts in Theory, Measurement, Development and Behaviour.

London: Longman.Carkhuff, R.R. (1986) Human Processing and Human Productivity. Amherst, MA: Human

Resource Development Press.Clark, J. and Koonce, R. (1995) 'Engaging Organizational Survivors', Training and Development,

49(8): 22-30.Cranny, C.J., Smith, P.C. and Stone, E.F. (1992) Job Satisfaction. How People Feel about their

Jobs and How It Affects their Performance. Lexington, MA: Lexington Books.Cronbach, L.J. (1951) 'Coefficient Alpha and the Intemal Structure of Tests', Psychometrika, 16:

297-334.Cunningham, B., Gerrard, P., Chiang, F.P., Yong, L.K. and Linn, C.S. (1995) 'Do Undergraduates

Have What It Takes to Be Entrepreneurs and Managers of Small Businesses in Singapore?',Journal of Asian Business, 11(4): 35—49.

De Cenzo, D.A. and Robbins, S.P. (1988) Personnel/Human Resource Management, 3rd edn.Englewood Cliffs, NJ: Prentice-Hall.

Ettorre, B. (1999) 'Is Salary a Motivator?', Management Review, 88(1): 8.Fox, J.B., Scott, K.D. and Donohue, J.M. (1993) 'An Investigation into Pay Valence and

Performance in a Pay-for-Performance Field Setting', Journal of Organizational Behaviour,14(7): 687-93.

Frane, J., Jennrich, R.I. and Sampson, P.F. (1990) 'P4M-Factor Analysis'. In Dixon, W.J. andBrown, M.B. (eds) BMDP Statistical Software Manual, Vol. 1. Berkeley, CA: University ofCalifornia Press, pp. 311-37.

Glassock, S. and Gram, K. (1995) 'Winning Ways: Establishing an Effective WorkplaceRecognition System', National Productivity Review, 14(3): 91-102.

Hair, J.F., Anderson, R.E., Tatham, R.L. and Black, W.W. (1995) Multivariate Data Analysis:With Readings. Englewood Cliffs, NJ: Prentice Hall.

Page 22: Compensation, esteem valence and job performance: an empirical

718 The International Journal of Human Resource Management

Hanlon, S.C. and Taylor, R.R. (1991) 'An Examination of Changes in Work Group Communica-tion Behaviours Following Installation of a Gainsharing Plan', Group and OrganizationalStudies, 16(3): 238-67.

Heath, C. (1999) 'On the Social Psychology of Agency Relationships: Lay Theories of MotivationOveremphasize Extrinsic Incentives', Organizational Behavior and Human Decision Processes,78(1): 25-62.

Herzberg, F., Mausner, B. and Snyderman, B. (1959) The Motivation lo Work. New York:Wiley.

Hoffman, E. (1988) 'Abraham Maslow: Father of Enlightened Management', Training, 25(September): 79-82.

Hong, J.C., Yang, S.D., Wang, L.G., Chiou, E.F., Sun, F.Y. and Huang, T.L. (1995) 'Impact ofEmployee Benefits on Work Motivation and Productivity', International Journal of CareerManagement, 7(6): 10-14.

Hoyle, R.H. (ed.) (1995) Structural Equation Modelling: Concepts, Issues, and Applications.London: Sage.

Hughes, M.A., Price, R.L. and Marrs, D.W. (1986) 'Linking Theory Construction and TheoryTesting: Models with Multiple Indicators of Latent Variables'. In Hair, J.F., Anderson, R.E.,Tatham, R.L. and Black, W.W. (eds) Multivariate Data Analysis: With Readings. EnglewoodCliffs, NJ: Prentice Hall, pp. 694-707.

Igalens, J. and Roussel, P. (1999) 'A Study of the Relationship between Compensation Package,Work Motivation and Job Satisfaction', Journal of Organizational Behaviour, 20: 1003-25.

Ivancevich, J.M. and Matteson, M.T. (1999) Organizational Behavior and Management.Singapore: McGraw-Hill.

Jennrich, R.I. and Sampson, P.F. (1966) 'Rotation for Simple Loadings', Psychometrika, 31:313-23.

Kopelman, R.E. (1986) Managing Productivity in Organizations: A Practical, People-OrientedPerspective. New York: McGraw-Hill.

Korman, A.K. (1971) Industrial and Organizational Psychology. Englewood Cliffs, NJ: Prentice-Hall.

Landy, F.J. (1985) Psychology of Work Behaviour, 3rd edn. Homewood, IL: The DorseyPress.

Levine, H.Z. (1994) 'Why Incentive Plans Cannot Work', Compensation and Benefits Review,26(1): 77-8.

Lopez, E.M. and Greenhaus, J.H. (1978) 'Self-Esteem, Race, and Job Satisfaction', Journal ofVocational Behaviour, 13: 75-83.

Luthans, F. (1998) Organizational Behavior. Singapore: McGraw-Hill.McCampbell, A.S. (1996) 'Promoting productivity of plateaued managers by providing auton-

omy'. Production and Inventory Management Journal, 37(1): 61-4.MacCullum, R.C., Browne, M.W. and Sugawara, H.M. (1996) 'Power Analysis and Determina-

tion of Sample Size for Covariance Structure Modeling', Psychological Methods, 1(2):130-^9.

Martins, J.H, Loubser, M. and Van Wyk, H.D. (1996) Marketing Research: A South AfricanApproach. Pretoria: Unisa Press.

Maslow, A.H. (1943) 'A Theory of Human Motivation', Psychological Review, 50: 370-96.Miner, J.B. (1992) Industrial/ Organizational Psychology. New York: McGraw-Hill.Mullins, L.J. (1999) Management and Organizational Behaviour. Harlow, Essex: Prentice

Hall.Pfeffer, J. (1998) 'Six Dangerous Myths about Pay', Harvard Business Review, May-June:

109-19.Pierce, J.L. and Dunham, R.B. (1987) 'Organizational Commitment: Pre-Employment Propensity

and Initial Work Experiences', Journal of Management, 13(1): 163-78.Robbins, S.P. (1998) Organizational Behavior: Concepts, Controversies, Applications. Upper

Saddle River, NJ: Prentice-Hall.Rosenberg, M. (1965) Society and the Adolescent Self-Image. Princeton, NJ: Princeton University

Press.

Page 23: Compensation, esteem valence and job performance: an empirical

Arnolds and Boshoff: Compensation, esteem valence and performance 119

Salancik, G.R. and Pfeffer, J. (1977) 'An Examination of Need-Satisfaction Models of JobAttitudes', Administrative Science Quarterly, September: 427-56.

Scharge, M. (2000) 'Cafeteria benefits? Ha! You Deserve a Richer Banquet', Fortune, 3September: 274.

Schein, E.H. (1971) 'The Individual, the Organization, and the Career: A Conceptual Scheme',Journal of Applied Behavioural Science, 7(4): 401-26.

Sharma, P. (1991) 'Need Motivations as a Function of Age, Educational Status and CognitiveStyle of Managers and Workers', Indian Joumal of Psychometry and Education, 22(1):43-50.

Shore, L.M., Newton, L.A. and Thornton III, G.C. (1990) 'Job and Organizational Attitudes inRelation to Employee Behavioural Intentions', Journal of Organizational Behaviour, 2:57-67.

Shouksmith, G. (1989) 'A Construction of a Scale for Measuring Work Motivation', New ZealandJournal of Psychology, 18(2): 76-81.

Slabbert, A. (1996) 'Capitalism at the Crossroads', International Journal of Social Economics,23(9): 41-50.

Slabbert, A. (1997) 'An Exploratory Investigation of the Hypothetical Relationship betweenDownsizing and Employee Attitudes', unpublished research article. Cape Technikon, CapeTown.

Smith, A.S., Organ, D.W. and Near, J.P. (1983) 'Organizational Citizenship Behaviour: Its Natureand Antecedents', Joumal of Applied Psychology, 68(4): 653—63.

Smit, P.J. and Cronje, G.J. de J. (1992) Management Principles. Kewyn: Juta and Co.Spence, J.T. (1983) Achievement and Achievement Motives: Psychology and Sociological

Approaches. San Francisco: W.H. Freeman.Stahl, M.J. (1986) Managerial and Technical Motivation: Assessing Needs for Achievement.

Power and Affiliation. New York: Praeger.Steers, R.M. and Black, J.S. (1994) Organizational Behaviour, 5th edn. New York: HarperCollins

College.Steers, R.M. and Porter, L.W. (1991) Motivation and Work Behaviour, 5th edn. New York:

McGraw-Hill.Stein, M. and Hollowitz, J. (1992) Psyche at Work: Workplace Applications of the Jungian

Analytical Psychology. Wilmette, IL: Chiron.Sumerlin, J.R. and Norman, R.L. (1992) 'Self-Actualisation and Homeless Men: A Known-Group

Examination of Maslow's Hierarchy of Needs', Journal of Social Behaviour and Personality,7(3): 469-81.

Tharenou, P. (1979) 'Employee Self-Esteem: A Review of the Literature', Journal of VocationalBehaviour, 15: 316—46.

Tharenou, P. (1993) 'A Test of Reciprocal Causality for Absenteeism', Journal of OrganizationalBehaviour, 14: 269-90.

Vroom, V.H. (1964) Work and Motivation. New York: Wiley.Wahba, M.A. and Bridwell, L.G. (1976) 'Maslow Reconsidered: A Review of Research on the

Need Hierarchy Theory', Organizational Behaviour and Human Perfonnance, 15: 212^0 .Wanous, J.P. and Zwany, A. (1977) 'A Cross-Sectional Test of Need Hierarchy Theory',

Organizational Behavior and Human Perfonnance, February: 78-97.

Page 24: Compensation, esteem valence and job performance: an empirical