The International Journal of Human Resource Management 7:3 September 1996 Catching up on competitors: how organizations can motivate employees to work harder Birgit Benkhoff Abstract The paper is an attempt to explain why some organizations are more competitive than others in terms of their human resources. It establishes under what circumstances some employees exert extra effort beyond the level required to keep their job and to get promotion by exploring the role of motivational forces that may complement the calculative considerations as suggested by expectancy theory. In a multivariate analysis proxies for five motivational mechanisms were tested, based on need theory, positive work disposition, intrinsic motivation, behavioural commitment as implied by organizational roles and social exchange theory. It emerged that all of them make their own contribution to the explana- tion of extra work effort. Keywords Competitiveness, extra effort, test of motivation theories, management techniques Many firms stress the importance of human resources for their success, but very few seem to be successful at developing the potential of their work-force and using human assets as part of their strategy. This is surprising. Considering that human resource practices such as selection methods, performance evaluation sys- tems or rewards systems can be purchased on the market in a similar way to technology, one would expect that firms would catch up with their more success- ful competitors by imitating their methods. However, we find that certain per- sonnel policies are effective in some firms, yet not in others. According to Wright, McMahan and McWilliams (1994) two alternative issues may be responsible for that, and both are challenged in this paper. The first one has to do with evaluation problems. Because of the difficulties associated with measuring performance and because of the complex relationships between the factors contributing to performance, rival firms cannot identify the way in which human or organizational assets provide a competitive advantage. For example, it may not be the team work they tried to imitate that contributes to the success of the superior company, but the communication system. Another reason could be that the way the human resource pool operates is inimitable. Every firm has its unique history which unites its members in terms of shared values and meanings. Its policies work on the basis of this context and through a net of personal relationships between employees and with personal contacts outside the organization. Since the history of the firm with the competi- tive advantage and the exact attributes of its sub-groups can hardly be copied, human resources, unlike technology, may be highly specific to every firm. Rival 0985-5192 © Chapman & Hall 1996 Motivating employees to work harder Til firms cannot benefit from imitating personnel practices and therefore face a sus- tained competive disadvantage. This argument is intriguing, yet so far untested. To enhance competitiveness, we need to identify general rules for management that apply across individuals and organizations. The purpose of this paper is an attempt to do this. It introduces a third approach that avoids both the huge problems associated with evaluating the effectiveness of individual policies, on the one hand, and the multitude of organizational and interpersonal relation- ships specific to each firm, on the other. In order to escape the evaluation problem associated with employee perfor- mance, this study focuses on effort as the intermediate outcome. Whether effort leads to increased organizational performance is not normally under the control of individual employees, but is determined by the quality of managerial strategy and organizational processes. Measuring effort rather than performance avoids contamination by those factors. Rather than taking a contingency view, refusing to draw general conclusions, it seems worthwhile to try and identify the small building blocks of the assumptions and motivational mechanisms on which the policies rely. What seem to matter among the confusing variety of policies which work in some cases, but not in oth- ers, as illustrated by literature reviews, for instance on job redesign (Kelly, 1992) or participation (Levine and Tyson, 1990), are the perceptions of the individual. However, findings on employee perceptions and attitudes have been largely ignored in the evaluation of management issues. Many researchers believe in 'facts', even though we know from bargaining and industrial disputes that individ- uals often find it difficult to agree on what they are. A pay system which in the eyes of some workers is inequitable or an attempt at manipulation or exploitation may be regarded as fair treatment by others and have a motivating effect. Consequently, not only is it very difficult to evaluate particular management policies or techniques, it may not even make sense. What one can do instead is to trace the contributors to performance on a 'micro'-basis. If certain positive perceptions and attitudes are related to extra effort, we can be confident that policies or management techniques which create those perceptions and attitudes are likely to have a performance-enhancing effect. Even if the human resource pool as a whole is unique, there may be general principles associated with employee effort that are common to all individuals. On this basis it will be suggested that human resource practices, provided they sat- isfy certain motivational mechanisms, will induce employees to exert extra effort. The result may be a recipe that allows firms to catch up with the advantage cre- ated by the human assets of their competitors. Hard-working employees may not be a sufficient condition for organizational success. Expectancy theory suggests that at individual level high performance requires the necessary skills and a clear understanding of the role, as well as motivation. However, these two moderators do not seem to provide problems for managers. The main difficulties are stimulating motivation and knowing whether the costs are justified by the benefits. Limitations of expectancy theory Under the influence of economic research, work effort has traditionally been seen as the result of the incentive structure facing employees (Brockner, et at., 1992; 738 Birgit Benkhoff Mitchell and Nebeker, 1973). The framework most frequently applied is expectancy theory which assumes that individuals are rational economic agents who conserve energy and regard effort as a form of cost. Expectancy theory implies that cognitive processes are major determinants of behaviour and that individuals are capable of calculating costs and benefits of potential courses of action. The results of those calculations are supposedly used to choose among alternatives. Expectancy has its merits but it is problematic both on theoretical and practical grounds. Research suggests that it applies only to certain individu- als (Landy and Becker, 1987). Its strength lies in predicting discrete choices, and it works best where individuals have a period for reflection on the opti- mal outcome (Wanous, Keon and Latack, 1983). In the work situation employees rarely have sufficient time to do that and are normally faced with too many outcomes to be able to conduct comparative evaluations. Contrary to the assumptions of expectancy theory, it seems that individuals are not always conscious of their values and expectancies and of their often complex and shifting motives. On a practical level, the theory does not often lend itself to application at the individual workplace. Managers find it difficult to stipulate in detail, for most types of employees, the kind and level of performance that is to be rewarded. It is not easy either to judge the size and combination of rewards that makes the extra effort worthwhile in the eyes of employees. A further problem with the application of incentive systems as suggested by expectancy theory is how to monitor and control performance in order to discourage workers from choosing the apparently favourable option of shirking. If one wants to explain everyday work behaviour which, to a large degree, consists of impulsive and habitual action, one needs to explore other sources of motivation. Those mechanisms can account for effort which is not based on cal- culation. They may also explain patterns of behaviour which are difficult to jus- tify in terms of expectancy theory, such as sustained effort over time when circumstances and incentives change, e.g. while the supervisor is absent. Alternative motivation theories There is no theoretical approach that excludes cognitive mechanisms of one kind or other, but none of the motivational mechanisms examined here is the result of detached calculation. The analysis considers need theory, positive work disposi- tion, intrinsic motivation, behavioural commitment as implied by organizational roles and social exchange theory. The five theories are important and regularly receive attention in literature reviews on motivation theories (for example, Kanfer, 1992; Landy and Becker, 1987). They also have, each individually, previously been applied in statistically based research. Unlike goal-setting theory, which can be tested only where the management technique is applied, and reinforcement theory, which is difficult to measure with questionnaire items, they lend themselves to cross-sectional research for all types of employees. Behavioural commitment is less popular as a motivational mechanism, but its operationalization in terms of hierarchical level and supervisory status is often used as an independent variable in the organiza- tional behaviour literature. Motivating employees to work harder 739 Need theory and positive work disposition stress non-cognitive individual dif- ferences while intrinsic motivation, behavioural commitment and social exchange theory focus on motives supposedly common to all individuals such as need for control, competence, social approval and fairness. None of the theories implies that working hard is necessarily a cost. Individuals may feel the desire or inclina- tion to exert effort or, as in the case of intrinsic motivation according to Deci (1980), may interpret it as a pleasant experience. The purpose of this article is both practical and theoretical. It is intended to advance management policies as well as research on motivation. To convince managers of the importance of employee perceptions, the existing qualitative studies need to be supplemented with quantitative data and statistical analysis. So far, organizational decision makers receive little guidance from man- agement research. Ideology, rather than evidence, determines whether people believe that it is better to have a low-paid work-force or a high-cost productive work-force. Many managers keep an eye on labour costs rather than the ratio of value added to labour costs. Pfeffer (1994) in his most recent book argues, for instance, that what distin- guishes the five top performing US companies is that they pay close attention to the needs of their employees and promote single status for avoiding 'them and us attitudes' and distrust of management. However, there is not enough convincing evidence on the benefits of this approach to discourage other managers from going too far in their delayering, imposing change rather than negotiating it, making employees feel they are dispensable and motivating them by fear rather than through the task. Apart from giving guidlines for decision making, this study should also help to fill gaps in the research agenda on motivational mechanisms. Motivation theories (for a review see Kanfer, 1992) are typically researched and applied separately with little consideration for the relationships between them (Landy and Becker, 1987: 6-7). Similarly, they have been related to effort independently of each other. French (1957), for instance, tested a connection between achievement motivation and problem-solving success, and Brockner et al. (1992) linked effort and intrinsic motivation. One has to remain sceptical about these results. Studies largely use bivariate statistical analysis, not allowing for the possibility of several different sources of motivation and not controlling for various influences on behaviour. Most motivational mechanisms overlap, and many aspects are related to each other. One cannot rule out, for instance, that intrinsic motivation is a form of satisfaction, that needs and their satisfaction are part of the effort-reward relationship, and therefore part of social exchange, or that the perceived importance of work is an expression of need for achievement. Therefore each measure should be expected to act as a proxy for others when analysed separately. Not surprisingly sources of motivation under scrutiny do not have an impres- sive track record as proven antecedents of effort or performance. Landy and Becker (1987) come to the conclusion that there is little support for the need sat- isfaction-behaviour relationship and no full evidence for Deci's model of intrin- sic motivation (Deci and Ryan, 1980). Organization justice theories such as equity theory and social exchange theory have lost their popularity mainly because there is still no support for the hypothesis that overpayment leads to an increase in effort. In order to find out whether the theories themselves are at 740 Birgit Benkhoff fault or only their operationalization and the research methods, further investiga- tions are required, such as provided by this study. Since effort and motivational mechanisms are difficult to measure, many authors have avoided dealing with motivation theories directly and have treated extra effort as an expression of attitudes such as job involvement and commit- ment. They typically use some attitudinal measure referring to extra effort and test the statistical relationships with a diverse selection of variables such as bio- graphical data, particular attributes of the work situation and characteristics of the organization in the hope of finding a connection (see Mathieu and Zajac, 1990; Rabinowitz and Hall, 1977, for reviews). The results tend not to be very robust, except for showing a stable link with work satisfaction. There is obviously no short cut that allows us to leave motivation theories aside. Merely identifying single positive work attitudes and hoping to find an association with easily measurable selection criteria that help pick the respective employees or beneficial working conditions does not seem to lead very far. What we want to know is what makes people work hard. If it is a matter of attitudes, we need to know the causes of these attitudes. It may turn out that positive atti- tudes implying discretionary effort originate from the same motivational sources that cause extra effort. In that case, Scholl (1981) argues, attitudes as intermedi- ate variables would lose their analytical utility. It may hence be advisable to con- centrate on the explanation of effort until we have a clearer idea of what could be gained by including attitudes in a model. This analysis is an attempt to shed more light on motivational mechanisms, both in their relationship to one another and their relationship to extra effort. Proxies of five motivational forces are used to answer two questions. First, do the motivation theories which differ in focus but do not necessarily contradict each other, represent genuinely distinct sources of motivation? Or is there too much overlap to justify treating them as separate mechanisms of equal impor- tance? The second and key question to be answered is: does each, or do only some, of the motivation theories make their own contribution to the explanation of discretionary effort. Data and methods The variables used here are taken from an employee survey held by a large chemical company in Germany (from now on referred to as ChemAG). It was conducted in the form of questionnaire-based interviews in the homes of 1,931 randomly selected employees. The purpose of the survey which consists of more than eighty questions, including twenty open ones, plus demographic items, was to establish what employees thought about their company and how they felt about their work situation. A variety of measures which approximate the rele- vant motivational mechanisms were among the questionnaire items. Before ran- dom selection of the sample, which was derived from the personnel records of the company (with the exception of apprentices), the population was stratified to make sure that all business areas and departments, and all levels of the hierarchy were represented. The use of statistical analysis for testing motivation theories creates consider- able problems. Whether a measure matches the concept has to be decided at face value since there is no generally accepted benchmark for judging convergent or Motivating employees to work harder 741 discriminant validity. Researchers in the past have been more concerned about reliability than about validity. This has led to the widespread use of multi-item questionnaires without sufficient care being taken that all items measure the same concept (see, for example, Rabinowitz and Hall (1977) on the multi-dimen- sional nature of Lodahl and Kejner's (1965) Job Involvement Scale and Benkhoff (1994) on organizational commitment instruments). But, unless the homogeneity of these scales is established through factor analysis, we cannot be sure that they are measuring only one concept. This analysis is partly based on single-item measures, partly on short instru- ments of two to five items. As argued by Kiesler (1971), among others, short scales are not necessarily less valid than longer tests. Short scales are common among instruments measuring job motivation, such as Vroom's (1960) three-item scale tapping 'Attitude Towards the Job', Seashore et al. (1982) Michigan Organizational Assessment Questionnaire measuring 'Intrinsic and Extrinsic Rewards Satisfaction' and 'Social Rewards Satisfaction' with three items each, and Quinn and Shepard's (1974) four-item scale on 'Self-esteem at Work'. Single items may be not as reliable as multi-item scales, but they are less likely than longer scales to be contaminated by different attitude dimensions and hence are probably not less valid. Many studies (for example, Allen and Meyer, 1990; Buchanan, 1974; Dubin, Champoux and Porter, 1975; Thompson, Kopelman and Schriesheim, 1992) have at least partly used the single-item approach. The measures Extra effort over and above perceived job requirements is captured in two differ- ent ways. The first measure is based on the responses of employees who were asked with which of four statements they agree most: (a) 'I put myself out in my work and I often do more than is demanded of me. My job is so important to me that I sacrifice much for it.' (b) 'All in all I enjoy my work and every now and then I do more than is required. But this should not be a permanant situation.' (c) 'In my job I do what is demanded of me. Nobody can criticize me there. But I cannot see why I should exert extra effort beyond that.' (d) 'I often have to force myself to go to work. I therefore only do what is absolutely necessary.' Agreement with statement A (22.4 per cent = 1) versus the rest (= 0) forms the variable Extra work. Its flipside. Lack of motivation, is captured by the small groups of respondents who agreed with statements C (4.3 per cent) or D (1.9 per cent). The categories were combined to leave a sufficient number of cases for detailed statistical analysis. The second effort measure. Voluntary task, represents employees who have voluntarily taken part in at least one company work group (problem-solving groups, project groups for special tasks, quality maintenance and other discussion groups) and intend to do so also in the future (16.5 per cent = 1, no participation and/ or no intention do so in the future = 0). The effort measures are not ideal, but they compare favourably with the effort measures used in the literature. The most frequently used approach goes back to 742 Birgit Benkhoff Vroom (1964) who suggested the use of hours spent to capture effort, and many authors have followed him (for example, Kopelman and Thompson, 1976; Mitchell and Nebeker, 1973). The problem with using hours is that their number says little about the amount and quality of the work done. Brockner et al. (1992: 417) used a three-item index of self-report statements from which they quote the two items 'I try to work as hard as possible' and 'I intentionally expend a great deal of effort in carrying out my job.' Due to the qualifiers used in statements, the responses are likely to be ambiguous in their meaning. An advantage of the variable extra work used here is that it is less likely to be contaminated by social desirability than many self-report items on effort, e.g., 'In my work I like to feel I am making some effort, not just for myself but for the organization as well' (Cook and Wall, 1980). The wording of statement A is fairly extreme and leaves the choice of the alternative positive response category B which would not lead to social disapproval either. Since the statements labelled as lack of motivation go against social desirabil- ity, the responses are probably honest. The variable matches one of the reverse items in Buchanan's (1974) organizational commitment scale 'I do what my job description requires; the organization does not have the right to expect more.' It is phrased somewhat more strongly than Cook and Wall's (1980) items relating to extra effort ('I am not willing to put myself out just to help the organization') and should therefore be more discriminating. Need theories predict high performance from employees with a high need for achievement. One would also expect greater effort from employees with a high need for esteem (McClelland, 1961), provided that approval and recognition at the workplace depend on high performance. Need for achievement is measured here with two combined variables. Respondents were confronted with a list of work characteristics and asked to indicate their importance by stating whether they 'must' rather than 'should' or 'can' be present in a company. Those who said that a) training opportunities or b) promotion opportunities were a 'must', and who also said in a subsequent question that those prospects were 'fully' or 'partly' present at their workplace are regarded as having a need for achievement. They are labelled as 1 (28.1 per cent for training, 23.5 per cent for promotion), while employees who either do not value advancement very much and/ or who say that these prospects are not at all realized form category 0. The measure implies that needs lead to higher performance only if individuals perceive that their needs can be satisfied. Need for esteem is measured with three statements. (a) 'Personal approval of my work is an incentive to me to do even better.' (56.6 per cent) (b) 'I find my work interesting because I have the feeling that I am needed.' (67.3 per cent) (c) 'I find self-confirmation and recognition at work.' (64.5 per cent) Agreement is coded as 1, disagreement as 0. In the literature needs tend to be measured in terms of specific satisfaction or the degree to which individual expectations are met, similarly to the need for achievement items used here. Cook and Wall (1980), for instance, in their ques- tionnaire refer to work experiences such as 'Recognition received for your achievements'. Respondents are asked: 'Do you have as much of this characteris- Motivating employees to work harder 743 tic in your job and work life as you would like, ideally?' Response categories on a five-point scale range from 'I have more than I really want' to 'I would very much like more'. Positive work disposition: past research on personal characteristics suggests that individuals may have a greater propensity to work hard on their jobs if work is their central life interest (Dubin, Champoux and Porter, 1975), or if they con- sider high work effort as the right behaviour or norm, often referred to as Protestant Work Ethic (Kidron, 1978). The importance or centrality attached to work is measured with the statement: 'I regard work as the main purpose of my life.' Employees who agreed (19.2 per cent of respondents) form category 1, the others category 0. The statement is very similar to an item in Dubin, Champoux and Porter's (1975) questionnaire measuring central life interest (CLI). Respondents are asked to select among three phrases: 'I believe that: a) helping my fellow men is more important than anything else, b) my job is more important than anything else, c) most things are about equally important.' Intrinsic motivation is based on the need for competence and freedom, accord- ing to Deci (1980). Employees are predicted to be motivated if they are success- ful and do not feel constrained by outside control. The sense of enjoyment and freedom is captured by the item 'My work is almost like a hobby to me' (22.9 per cent agreed = 1, disagreement = 0). An almost identical item, albeit with a five-point scale, can be found in the job satisfaction questionnaire by Brayfield and Rothe (1951): 'My job is like a hobby to me'. Behavioural commitment is used here in terms of Kiesler's (1971) approach. He proposes that people try to establish an image of themselves as being competent and in control and that they want to appear to be consistent in terms of their attitudes and behaviour and in terms of their behaviour over time. At work this implies that employees who assume responsibility and are openly involved in decision making are likely to 'deliver the goods'. They often exert maximum effort in order not to disappoint other people's expect- ations. Employees may work extremely hard not to reveal any deficiencies in competence to their superiors and colleagues so that they can maintain their positive self-image. Two proxies are used to capture behavioural commitment: (a) position in the hierarchy with five categories: blue-collar workers (1), white- collar workers, middle managers, senior management and top management (5) (proportions representative of company structure) and (b) being a supervisor (34.3 per cent = 1, others = 0). The assumption is that employees who have applied to become supervisors or to be promoted have made the implicit claim that they are able and prepared to cope with the demands of the job. Further, it is probably not unreasonable to assume that other people's expectations increase as employees rise up the hierar- chy. The impact that visibility and responsibility have on behavioural commit- ment has been established by Hollenbeck, William and Klein (1989) in connection with commitment to difficult goals. Those who have doubts about using hierarchical level and supervisory status as proxies for behavioural com- mitment may consider these variables at face value and regard behavioural com- mitment as an explanation for their impact. 744 Birgit Benkhoff Social exchange theory proposes that employees will work hard only if they feel that the organization responds or will respond in the future with adequate financial benefits, e.g. pay and promotion, or non-financial rewards, such as respect and recognition, job security or interesting tasks. Unless one argues in terms of social exchange theory or other forms of balance theory, it is difficult to explain why satisfaction should lead to effort and performance rather than per- formance to satisfaction. According to social exchange theory a motivational impact should be expected from workers with a high degree of satisfaction who may feel an obligation to reciprocate and hence exert particular effort. Five survey items are suitable for capturing the balance between effort and rewards: (a) Satisfaction with the current position, categories: not satisfied (1), fairly satis- fied, satisfied, very satisfied (4) (b) Satisfaction with the task area, categories: not satisfied (1), fairly satisfied, satisfied, very satisfied (4) (c) General job satisfaction, categories: very dissatisfied (1) to very satisfied (7) (d) Expectations employee had when joining the company have been met, cate- gories: yes (60 per cent = 1) and no (= 0) (e) Satisfaction with pay, categories: not satisfied (1), less satisfied (2), satisfied (3) and very satisfied (4) Expectancy theory, or the calculative mechanisms it proposes to be the source of motivation, can only be approximated here. ChemAG not only pays its work- force according to the salary scales agreed with the union, but also adds a monthly performance bonus determined by the supervisors which can range between £35 and £175. The impact of the financial incentive cannot be held con- stant since data on the respondents' performance-related pay were not available. Additional tests (not reported here) suggest, however, that extra work and vol- untary task probably do not represent many employees with a money motive. The item 'Better pay would make me perform better' is not related to the extra effort variables, neither are responses concerning money where employees are asked in open questions about what they like or dislike about the company. Another potential calculative explanation for extra effort is the desire for promo- tion. It was possible to control for this incentive with an item asking respondents whether they wanted promotion (= 1) or not (= 0). The results Distinction between motivational mechanisms The first part of the analysis explored whether the measures from the employee survey actually represented conceptually different measures. Since it is unlikely that the motivation theories are completely unrelated, an exploratory factor analysis was conducted with oblique rotation. The statistical package used (SPSSX, version 4.0) by default extracts the factors through principal-compo- nents analysis and discards those with an eigenvalue smaller than one. It is striking that the pattern of factors identifies nearly all hypothesized theo- ries as distinct latent variables (Table 1). The factors that emerged seem to repre- sent social exchange theory, behavioural conmiitment (hierarchical position. Motivating employees to work harder 745 Table 1 Factor analysis of motivation measures (rotated factor loadings of each item, pattern matrix) Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 1 Social exchange Satisfaction with position .79145 .01458 .03395 -.00258 -.08372 Satisfaction with task area .71264 .11945 .07709 .01912 -.01574 Overall work satisfaction .65901 -.04595 .18564 -.02141 .03074 Expectations met .65136 -.12933 .06113 .07004 -.03242 Pay satisfaction .61745 .07686 -.24527 -.07183 .14640 2 Behavioural commitment Hierarchical level -.06954 .86898 -.00735 .00197 -.02791 Having subordinates .06047 .79277 .06842 .03713 .00951 3 Need for esteem Approval is incentive .00834 -.04467 .73387 .03939 -.02498 Feel needed .03040 .07142 .69259 -.13659 .06624 Find recognition .07434 .05701 .65959 .10289 .08837 4 Need for achievement Training opps a must -.00583 -.02667 -.04736 .80605 .06549 Promotion opps a must .01059 .06025 .02574 .77770 -.03592 Factor 5 Work main purpose -.04607 -.09828 .09128 -.02034 .79755 Work is like a hobby .02745 .07876 -.01136 .05360 .73448 supervisory role), need for esteem and need for achievement. The meaning of factor five is not clear. It combines the two items measuring positive work dispo- sition and intrinsic motivation. Leaving aside the fact that in factor analysis sin- gle-item measures rarely appear as separate factors, one reason for the two variables clustering together could be that intrinsic motivation partly represents a tendency in individuals to enjoy their work, an attitude stable over time, i.e. a disposition. 'Tournament' of the variables As a second step the analysis explored whether the theories, or the variables rep- resenting them, contribute to extra effort when other effects are held constant. The appropriate multivariate statistical technique that allows one to establish the contribution of several hypothesized independent variables for a dichotomous dependent variable is logistic regression analysis (SPSSX, version 4.0). For the purpose of the 'tournament' of the motivational approaches, summary scores were computed by multiplying the individual items with their factor load- ings, thus giving them different weights in terms of their contribution to the underlying factor, and by adding up the scores. To reflect the hypothesized motivational mechanisms, seven independent variables were entered into the logit analysis, four weighted summary scores based on the four factors and two single items measuring positive disposition towards work and intrinsic motivation. The 746 Birgit Benkhoff analysis was conducted three times, using each measure of extra effort. Extra work. Lack of motivation and Voluntary task, in turn as dependent variable. The logit analysis (Table 2) revealed that the motivational measures, including the incentive of promotion, are all significantly related to at least two of the measures of extra effort, except for need for achievement. This Table 2 Motivation theories and commitment: logistic regression analysis (significant coefficients only) Extra Voluntary Lack of Work Task Motivation n=1701 n=1725 n=1701 Need for achievement Need for esteem .3190 -.5773 (.00) (.00) Disposition .9383 (.00) Intrinsic motivation .6994 (.00) Behavioural commitment .4980 .7904 -.7346 (.00) (.00) (.00) Social exchange .1726 -.2057 (.00) (.00) Wanting promotion .4162 .3459 -.7191 (.00) (.02) (.00) Significance of goodness of fit (comparing the present model to a 'perfect' model) .43 .77 .47 does not necessarily mean that need for achievement has no impact on extra effort. Its lack of significance may have been caused by its overlap with related variables in the model, such as wanting promotion, or by the way the variable has been measured here. More appropriate items might have pro- duced different results. Extra work emerged as being related to positive work disposition, explaining the biggest proportion of the variance, intrinsic motivation (of second largest importance), behavioural commitment and social exchange. If one uses the nega- tively phrased item Lack of motivation as the dependent variable, the pattern looks very similar. The absence of a positive work disposition, behavioural com- mitment, shortcomings in social exchange and lack of interest in promotion seem to contribute to employee behaviour. Intrinsic motivation no longer shows up as important. But need for esteem becomes significant. The desire for respect and approval may not motivate very much, but the absence of recognition seems to act as a powerful de-motivator. The results of the logit analysis with the second effort measure voluntary task as dependent variable look somewhat different. Only three motivational mecha- Motivating employees to work harder 747 nisms show up as significant: need for esteem, behavioural commitment and desire for promotion. The three factors combined provide a much better fit for explaining voluntary task (explaining 77 per cent of what would be a perfect fit) than the five independent variables do for extra effort and lack of motivation. Discussion The statistical analysis has revealed that employees have a variety of different responses for working harder than is strictly required. Their behaviour is partly determined by a disposition to regard work as a very important aspect of their lives, i.e. a characteristic employees bring into the organization, and partly by the particular working conditions created by management. An interesting job and satisfaction with various aspects of their work, especially their position, task area and pay, induce them to exert extra effort. Additionally, extra effort increases with each level in the hierarchy, with responsibility for subordinates and with the desire for promotion. The prospect of a better-paid position is obviously an incentive, as suggested by expectancy theory. The most striking result brought out by this study is that we now have statistical evidence that financial and other tangible incentives have only a limited role to play in motiva- tion. We cannot reject the hypothesis that particular non-calculative motivational mechanisms have an additional independent influence. The calculative approach alone is obviously too narrow to account for the variation in work effort. The results imply that the motivation theories tested here warrant further research with respect to employee behaviour in spite of their disappointing track record so far. The pessimistic categorization by Kanfer (1992: 4) of need theo- ries, dispositions and organization justice theories as 'distal constructs' in terms of proximity to action (as against 'proximal constructs') may be premature. She argues: 'To date, most distal theories of motivation have enjoyed their greatest success in predicting other distal constructs, such as predecision and decision processes and intentions, rather than behaviour or performance.' That certain theories appear as 'distal' may be the result of flawed methods rather than a matter of explanatory power. This study suggests that they are clearly related to behaviour. The findings have implications not only for the direction of future research, but also for methods. One of the issues that need to be addressed is that of causality. Since we are dealing with cross-sectional data it is, strictly speaking, not possible to say what causes workers to exert extra effort. Causation is suggested by the theories, but one cannot reasonably determine what is cause and effect. Instead of arguing that employees with a high need for esteem, a positive work disposition, an interesting or a high profile job or a high level of job satisfaction will therefore work particularly hard, one might assume the opposite direction of causality. Hard-working employees tend to be treated better, get more recognition, become supervisors and move up the hierarchy. Three of the theories support a two-way process anyway, social exchange theory by implication and intrinsic motivation and behavioural commitment because they assume that behaviour may lead to particular attributions and to further behaviour of a similar kind. The causal influences surrounding work effort may be almost impossible to disentangle, except by using laboratory studies whose results may not be valid in the actual work environment. An indirect method of gauging the causes of extra 748 Birgit Benkhoff effort could be derived from the stability of work behaviour itself since different causes suggest different degrees of stability. If dispositions and needs are of great importance, measures of extra effort should remain almost constant over time. The role of intrinsic motivation, behavioural commitment or social exchange should become obvious in greater fluctuation. It may be that the factors most critical for employee motivation vary across different career stages. This would show in systematic changes in the relevant coefficients across the whole popula- tion. If we want to be very confident about our results in effort research, we first need to have more accurate measures. The somewhat different pattern in the sta- tistical results for extra work and voluntary task suggests that one has to be very careful when operationaiizing the concept of extra effort. The general self-report variable extra work seems to perform better in terms of the number of significant motivation theories than the more objective measure voluntary task. The two effort measures also differ in their pattern of independent variables. While employees who take on additional tasks for the benefit of the organization are characterized by a need for esteem (not significant for employees who generally work hard on their jobs), they do not seem to be motivated by a positive work disposition, work satisfaction and intrinsic motivation. The reason for the dis- crepancies could be that voluntarily working in groups outside the actual job is one of many forms of exerting effort, and that the measure is therefore only par- tial. The results presented here are only a first step towards more refined research. Apart from better effort measures we also need carefully designed and tested instruments to capture the different motivational mechanisms. The items used here to measure the theories compare favourably with many instruments in the literature and cluster roughly as expected. This gives some support for their validity, but they could be improved. Further, we need to develop better controls for calculative motives. One could either try to measure, and hold constant, anticipated tangible rewards, or ask respondents only about behaviour that makes a calculative motive unlikely, e.g. exerting effort even in the absence of the supervisor or other observers. Implications for managers Even though the results of this analysis need further confirmation, it may not be premature to consider the practical implications. If a company wants to maxi- mize employee motivation, it should take a two-pronged approach, concentrating on selection as well as adjusting working conditions accordingly. Since personal characteristics seem to contribute to effort, it would make sense for organiza- tions to identify applicants with the relevant needs and dispositions. The com- pany would gain further if the job situation is right. Intrinsic motivation can develop where jobs are designed to provide variation and a challenge, which employees will be able to meet if they are also given the necessary amount of training. Regular feedback, whether as part of an appraisal system or on an informal basis, should not only provide employees with a sense of direction, but, if positive, should give individuals the acknowledgement and sense of esteem that has been shown to influence their level of effort. An internal labour market can be expected to contribute further to employee motivation Motivating employees to work harder 749 since it provides the necessary positive prospects for employees who want pro- motion. With every step they progress in the hierarchy they are likely to feel obliged to put in additional effort, over and above the effect that may come from the additional intrinsic motivation, esteem or satisfaction that the higher position might also provide. Where the organizational ladder does not have enough rungs to keep ambitious employees fully motivated, one can assume that having responsibilities and subordinates to look after provides at least a certain effort-enhancing effect. Too much work pressure and stress on the other hand would allow neither intrinsic motivation to develop nor the sense of job satisfaction that has a posi- tive effect on performance. For beneficial social exchange to develop, the theory demands trust and a long-term relationship as guaranteed by job security. When considering the human resource policies that may or may not satisfy motiva- tional requirements, one should keep in mind, however, that it is not enough to put policies in place. What matter for effort are employee perceptions, which may also depend on how policies are introduced and how they are run. Unless the messages received from management are consistent and stable over time, per- ceptions will not be sufficiently positive to result in discretionary effort. ChemAG, where the survey was conducted, offers several of the conditions that allow the motivational mechanisms to work. Job security has so far not been undermined, the company provides a considerable amount of training, making sure that employees are up to the challenges of their job. A fairly well- developed internal labour market and organizational hierarchy keep aspirations alive, and works councils give employees an opportunity for participation that is taken seriously by management and works well enough to communicate upwards when particular issues cause dissatisfaction. However, employees interviewed complained about increased work loads and stress which may signal danger for otherwise high levels of effort. Once the relevance of particular motivation theories for effort is confirmed through further studies and robust research results, the mechanisms can be used to improve human resource practices. The problem for managers who have heard success stories of certain management techniques is that labels such as 'job redesign', 'participation' and 'team work' each hide a wide range of practices with a variety of rules. Not all forms of job redesign or participation have moti- vational potential since workers in each case may be facing very different experi- ences. Motivation theories can provide the criteria that help to anticipate whether a new policy will have a performance enhancing effect, or not, as the case may be. The questions managers will need to ask before implementing the technique intended to increase competitiveness, for instance, are: Does the spe- cific form of job-redesign to be introduced increase intrinsic motivation? Does it promise to satisfy need of achievement or need for control? Does it trigger behavioural conmiitment because of increased responsibility? For behavioural commitment to work, individuals need some control over the way they do their job. The more choice they have the more they will feel respon- sible for their performance and the more their performance will affect their self- esteem. If there are no opportunities to transcend strict rules regulating job performance, commitment is stifled and cannot benefit the company. The message from this piece of research is that effort is a matter of general motivational mechanisms which are not inimitable. Managers can foster effort 750 Birgit Benkhoff O ( r' • * o"oo o « p — o O O ^ "^ "^ <N — p p p P P 2 4.' O 4.- O 4.- • . 4- 04.- O |- O'R I I 4_ V / 4_ * ' 4_ ^^.y 4 , ^^~y 4_ " ' I 4_ 3 0 0 o'oo o"^ f s o o O T r o T f o o o - - o M ^ O - * O O O O O O "^ CN P "^ p rn ^ 4.-O 4.-O 4: O 4.-O 4.-O 4.-£ |-O 4.- I . f^ ^5 fNj f*, (^ ^5 i/^ ^f 1/^ ^' ^^ c^ ^ ^ ^3 ^o ^5 f^ f<, ^5 ^5 ^5 .-* ^3 ^5 ^? ^5 ^5 ^? ^'^ f^ ^5 '"^ ^5 f ^ 4~-^4-~—' ** I — ' ' l~—' f.^ ^^ f.^ ^5 ^5 f^ ^5 ^^ t*^ ^5 ."^ ^5 Os C3 C^ ^3 r*^ ^^ ^" * 1"*^^-" ' + " ~ ^ l ^ ^ + i i o — p—"pfNp — p — p p p f N p p p — o 4- O 4- O 4.- O 4.- O 4.- O 4.- O 4.- w 4. w I w 4. • I ;5 8 § 5 g 8 2 8 2 8 i2 f :^ 8 S 8 ?3 8 S § S i 4: O I O 4- O 4.- w 4.- O 4.- O 4.- O 4.- O 4: O |- O 4.- • = 8:2 8 S § g 8 2 8 S f::: 8 2 8 p S S 8 2 8 g: 4- O 4- O 4- O 4- O 4.- O 4.- O 4.- O 4: w 4.- w 4. w I w 4. - 1 ;q 8 :s 8 2 8 S 2 2 8 S 8 S 8 22 f 1^ 8 o S::: 8 2 8 S 8 4- O 4- O 4- O 4- O 4- O 4.- O 4.- O 4.- O 4.- w 4.- w 4. —- 1 — 4. ?; 8;?; 8 2 S!:: 8 ^ 8 2 8 ^ 8 S 8 S 8 2 8 S f 2 8 S 8 2 8 4- O 4- O 4- O 4- O 4- w 4- O 4.- O 4.- w 4.- O 4.- w 4.- ^ 4.- w 1 w 4. • 0 4 - 0 |- o 4- o 4- o |- o 4- ^.> 4; ^-> 4; o^ 4: O' 4; ^.:- 4/ O' |- -o 4: ^.> i' 'o°d4-O4-O°O°O°O . . . . |- O 4- O 4- O 4- O I-O4.-O |- O 4" / ^ ^ ^N /* ^ ^ ^ ' ^ ^ ^"^^ / ^ ^ ^ ^ . ^1 1* » • •f • • ^i. ^ c ^ ^ ^ ^ ^ ^ ^ ^14 r o ^ ^ ^ ^ vO ^O ^D 8 :8S228SP;Sp§82:8p^pp8a;pppS3p8ppp -0 . . : w 4- o 4- o 4- o 4- o 4- o 4- o 4- w 4- w I w 4. w I w 4. • ^ I w 4. C 2 B l : ^I sill g •i ^ n S S '% a -^ ^ 2 O ^ _j ;S -g .•= <« .2 .S M O OS i'o- (S TT H " i Motivating employees to work harder 751 by selecting certain types of employees and providing the working conditions that trigger motivation. No doubt, organisations have to give incentives. But to a large degree their competitiveness comes from a resource that is given freely by their employees. Birgit Benkhoff London School of Economics and Political Science Acknowledgements I would like to thank John Kelly, Ray Richardson, Riccardo Peccei and an anonymous referee for their comments and advice on earlier versions of the arti- cle. References Allen, N. J. and Meyer J. P. (1990) 'The Measurement and Antecedents of Affective, Continuance and Normative Commitment to the Organization', Journal of Occupational Psychology, 63: 1-17. Benkhoff, B. R. D. (1994) 'Employee Commitment to Work: A Test of Motivation Theories and a Re-Conceptualization', unpublished PhD thesis, London School of Economics. Brayfield, A. H. and Rothe, H. F. (1951) 'An Index of Job Satisfaction', Journal of Applied Psychology, 35: 307-11. Brockner, J., Grover, S., Reed, T. F. and Dewitt, R. L. (1992) 'Layoffs, Job Insecurity, and Survivors' Work EfTort: Evidence of an Inverted-U Relationship', Academy of Management Journal, 35: 413-25. Buchanan, B. (1974) 'Building Organizational Commitment: The Socialization of Managers in Work Organizations', Administrative Science Quarterly, 19: 533-46. Cook, J. and Wall, T. (1980) 'New Work Attitude Measures of Trust, Organizational Commitment and Personal Need Non-Fulfilment', Journal of Occupational Psychology, 53: 39—52. Deci, E. L. (1980) The Psychology of Self-Determination. Lexington, MA: D.C. Heath. Deci, E. L. and Ryan, R. M. (1980) 'The Empirical Exploration of Intrinsic Motivational Processes', in Berkowitz, L. (ed.) Advances in Experimental Social Psychology. New York: Academic Press. Dubin, R., Champoux, J. and Porter, L. (1975) 'Central Life Interests and Organizational Commitment of Blue-Collar and Clerical Workers', Administrative Science Quarterly, 20: 411-21. French, E. G. (1957) 'Effects of Interaction of Achievement, Motivation, and Intelligence on Problem Solving Sucess', Journal of Applied Psychology, 12: 399-400. Hollenbeck, J. R., Williams, C. R. and Klein, H. J. (1989) 'An Empirical Examination of the Antecedents of Commitment to Difficult Goals', Journal of Applied Psychology, 74: lo—23. Kanfer, R. (1992) 'Work Motivation: New Directions in Theory and Research', International Review of Industrial and Organizational Psychology, 1: 1-53. Kidron, A. (1978) 'Work Values and Organizational Commitment', Academy of Management Journal, 21: 239-47. Kiesler, C. A. (1971) The Psychology of Commitment: Experiments Linking Belief to Behavior. New York: Academic Press. Kopelman, R. E. and Thompson, P. H. (1976) 'Boundary Conditions for Expectancy Theory Predictions of Work Motivation and Job Performance', Academy of Management Journal, 9: 1-38. 752 Birgit Benkhoff Landy, F. J. and Becker, W. S. (1987) 'Motivation Theory Reconsidered', Research in Organizational Behavior, 9: 1-38. Levine, D. I. and Tyson, L. D. (1990) 'Participation Productivity, and the Firm's Environment', in Bunder, A. (ed.) Paying for productivity, A look at the evidence. Washington D.C: Brookings Institute. Lodahl, T. and Kejner, M. (1965) 'The Definition and Measurement of Job Involvement', Journal of Applied Psychology, 49: 24-33. Mathieu, J. E. and Zajac, D. M. (1990) 'A Review and Meta-Analysis of the Antecedents, Correlates, and Consequences of Organizational Commitment', Psychological Bulletin, 108: 171-94.McClelland, D. C. (1961) The Achieving Society. New York: Van Nostrand. McClelland, D. C. (1961) The Achieving Society. New York: Van Nostrand. Mitchell, T. R. and Nebecker, D. M. (1973) 'Expectancy Theory Predictions of Academic Effort and Performance', Journal of Applied Psychology, 57: 61-7. Pfeffer, J. (1994) Competitive Advantage Through People. Boston Mass: Harvard Business School Press. Quinn, R. P. and Shepard, L. J. (1974) The 1972-73 Quality of Employment Survey Ann Arbor, MI: Institute for Social Research, University of Michigan. Rabinovitz, S. and Hall, D. T. (1977) 'Organizational Research on Job Involvement', Psychological Bulletin, 84: 265-88. Scholl, R. W. (1981) 'Differentiating Organizational Commitment from Expectancy as a Motivating Force', Academy of Management Review, 6: 589-99. Seashore, S. E., Lawler, E. E., Mirvis, P. and Cammann, C. (eds) (1982) Observing and Measuring Organizational Change: A Guide to Field Practice. New York: Wiley. Thompson, C. A., Kopelman, R. E. and Schriesheim, C. A. (1991) 'Putting All One's Eggs in the Same Basket: a Comparison of Commitment and Satisfaction Among Self- and Organizationally Employed Men', Journal of Applied Psychology, 11: 738-43. Vroom, V. H. (1960) Some Personality Determinants of the Effects of Participation. Englewood Cliffs, NJ: Prentice Hall. Vroom, V. H. (1964) Work and Motivation. New York: Wiley. Wanous, J. P., Keon, T. L. and Latack, J. C. (1983) 'Expectancy Theory and Occupational and Organizational Choices: A Review and Test', Organizational Behaviour and Human Performance. 32: 66-85. Wright, P.M., McMahan, G.C. and McWilliams, A. (1994) 'Human Resources and Sustained Competitive Advantage: A Resource Based Perspective', Internationai Journal of Human Resource Managemeni, 5: 301-26.