Promotion as Motivation Presented by Pei-wen Huang Tournament theory suggests that the larger the spread between the winner's and loser's prizes, the higher the effort exerted by both players, and the larger the raise associated with a given promotion, the greater the incentive to be promoted. Salary Structure and EffortA salary structure has two features--level and spread. The spread refers to the difference between the salaries that high and low level employees receive. The higher the spread, the more anxious is any given employee to obtain a promotion to the higher-level job. Then each worker puts forth more effort in an attempt to secure the higher-level job. These results in a higher overall level of effort as workers are forced into a rat race to attempt to win the higher position. The benefit of the rat race is that it produces higher levels of effort and therefore higher levels of output for the firms as a whole. The cost is that since workers are putting forth more additional effort, they must be compensated accordingly. Salary of Total Total Difference between Research Salary of earnings of earning of Winner and Loser Structure Engineer Project Head Loser Winner Earnings Structure A $50,000 $200,000 $500,000 $1,250,000 $750,000 Structure B $50,000 $100,000 $500,000 $750,000 $250,000 Structure C $100,000 $250,000 $1,000,000 $1,750,000 $750,000 Under structure A, each worker puts forth more effort in an attempt to secure the higher-level job. This results in a higher overall level of effort as workers are forced into a rate race to attempt to win the higher position. If workers find effort especially painful, they would prefer the lower paying job that induces less effort than the higher paying, higher effort job. Luck Noise, or luck, has adverse effects on effort. When the outcome of a promotion race is affected by extraneous factors, workers reduce their effort because effort is less likely to affect the outcome. To offset the effects of luck, it is necessary to increase the spread. Thus, salary spreads are greater in work environments where measurement errors or production uncertainties are large. New firms and new industries are most likely to have large luck components and should therefore use 1 larger spreads. The amount of luck of noise inherent in the promotion process affects the appropriate salary structure. Firms setting up salary structure should take into account that the wage spread affects worker incentives differently, depending on the amount of luck involved. If the environment is one where luck is extremely important, then a firm should select a higher wage spread. Tournaments and the Organizational Chart Structure 2 provides more incentive than structure 1... In the corporation, being promoted from file clerk to clerical supervisor not only provides the workers with a higher salary, but it also allows him to compete for subsequent jobs. Thus, the promotion has value even if file clerks and clerical supervisors earn exactly the same amount. As the individual moves toward the top of the hierarchy, the option value diminishes. In the very last round, the winner gets the CEO’s salary and only the CEO’s salary. Individuals will not work hard to win the CEO’s prize on the basis of the additional opportunities it affords. Firms that have symmetric structures, like those shown in structure 1, are likely to be using some other form of worker motivation such as piece-rate pay or compensation on the basis of effort. Firms that reward workers on relative performance rather than absolute performance are more likely to have structures like the one shown in structure 2. 1 2 When too many workers compete for a promotion that carries with it a small raise, effort will suffer. Incentives can be produced either by relative comparisons or by absolute measurements. Piecework provides absolute incentives while promotions, based on relative performance, provide relative incentives. The advantages of relative incentives are twofold. First, relative comparisons are often easier and cheaper to make than absolute judgments. Second, relative comparisons reduce the effects of common noise, like business conditions or 2 supervisor attitude, that would affect pay based on absolute standards. The disadvantage of relative incentives is that they may foster either collusion, or induce negative competition between workers which could be harmful to the firm. Internal promotion produces favorable incentive effects, but lends itself to worker collusion. One way to reduce the likelihood of collusion is to hire from the outside. Therefore, outside hiring should only be used when the outside candidate is significantly better than all internal candidates, or when insiders have colluded to put forth too little efforts in the past. The economic analysis of internal labour markets The ILM raises a number of issues for economic analysis. First, to what extent is the pricing and allocation of labor within firms controlled by non-economic variables? Secondly, to the extent that non-economic variables are important, what are the implications for economic theory? Thirdly, are the conventional methods of economics appropriate for empirical research on the ILM or is it necessary to develop alternative approaches? A major theme concerns the efficiency aspects of the ILM. Much of the early work suggested that ILMs, by introducing rigidities which constrain labor market adjustments, prevent allocative efficiency being achieved and deliberately trade efficiency for equity. However, later work has viewed the ILM as an efficient response to uncertainty in a market containing an idiosyncratic factor of production; the firm's internal market is seen to operate more efficiently than thee external market. Other models introduce competitive elements and examine the implications of alternative labor contracts for Pareto efficiency. Nevertheless, some inefficiency is found to exist due, for example, to over-investment by individuals in signals' to employers concerning their quality. A closely related aspect concerns the motivation of workers, although different authors use their own special vocabulary, making precise comparisons difficult. Some stress the `negative` aspects of moral hazard and the enforcement of compliance with the firm's rules, while others stress more `positive` aspects whereby individuals identify their own interests in terms of the performance of the enterprise. Despite these differences of emphasis, most authors agree that questions of work effort and motivation are central to the analysis of the ILM and its rationale. A related issue concerns the inappropriateness of aggregate data for ILM analysis; this imposes the need for researchers on the ILM to collect their own data. The ILM concept was developed by economists in the institutionalist tradition. It represents the integration of a wide range of ideas on labour market structure, labour 3 mobility, wage determination. Such observations could be explained in a number of different ways, including measurement inaccuracies, imperfect information or worker inertia. However, those who believed that the explanation was more deepseated directed their attention towards the nature of the employing organization. Doeringer and Piore explained low mobility by the efforts of employers to reduce turnover because of the presence of skill-specificity. Internal mobility is seen to represent a compromise between promotion by seniority and by ability; it is suggested that the former is more acceptable to the worker and the latter to the employer. Doeringer and Piore suggested that there are three principal elements of the ILM wage structure;(i) the plant wage level; (ii) the vertical differentiation and (iii) the horizontal differentiation of the wage structure. Neoclassical Analyses In dealing with the ILM, neoclassical economists have concentrated on only a small number of the characteristics which distinguish it from wage competition, such as low turnover, internal promotion and seniority. While the bulk of the neo-classical literature has only indirectly examined the ILM, two related strands have examined it more directly. The most direct analysis is the work of Williamson (1975). This research has variously been called the transaction cost approach or the theory of idiosyncratic exchange. The second type of direct analysis is the more disparate literature on market signaling, selection models and principal agent analyses. Transaction Costs Two basic behavioural assumptions are made. First, individuals are assumed to have bounded rationality'. Secondly, workers are assumed to be ` opportunistic`; they would not honestly disclose all the information they have. Signalling and Incentives The fact that the seller of a good has more information about its quality than potential buyers involves the use of various signals. Such signaling involves self-selection processes -- whereby the course of action taken by the seller differs according to the quality of the good being sold. It must be unprofitable for low quality sellers to imitate the signals of the higher quality sellers. Despite the somewhat different focus of this literature many contributions have been explicitly concerned with labour markets and have described their results as producing features of ILMs. The disparate contributions to the signaling/incentive literature have been synthesized and termed efficiency wage models. Efficiency wage models explain why firms find it unprofitable to reduce wages when there is high unemployment. In 4 brief, wage cuts are said to harm productivity and therefore, Institutional and Radical Analyses The recognition of non-economic determinants of the ILM has evoked a mixed response from economists in terms of theory construction. Neoclassicists have suggested the ILMs slow down the basic processes which are outlined in the wage competition model but that the model when modified is still applicable. Institutionalists have searched for alternative models, such as job competition and dual labour markets, but have not been able to convince the bulk of economists of the utility of these constructs. Radicals view the ILM as confirming the superiority of an economic model grounded in Marxian rather than neoclassical concepts. The ILM has thus simply re-oriented the debate between the schools of thought. Empirical Studies The ILM poses special problems for empirical analysis. First, it is a complex phenomenon which cannot be reduced to a small number of measures. Secondly, its characteristics have been shown to vary considerably across space and time in a manner that is not readily understandable. Thirdly, many of its principal dimensions are longitudinal in nature and cannot be analysed using cross-sectional date. The major consequence of these problems is that the types of data normally used by economists are not useful for ILM research. In particular, ILM research requires that researchers collect their own data. Empirical research on the ILM has generally adopted three main types of approach. The first involves the analysis of highly aggregative data; the second is based on case studies of particular firms and industries, using data collected specially by researchers; the third type concentrates on the analysis of career histories, usually of particular occupational groups. Low turnover rates have been found in many sectors and occupations. The filling of upper level positions by internal promotion has been shown to be a very common practice. It was shown that the experience of internal mobility is a key determinant of earnings. Osterman found that movement between job ladders was relatively easy at the entry position but less so at higher levels. The empirical research has also shown that wages are tied to jobs within ILMs and that they are not very responsive to changing labour market condition. Crockett (1983) found that firms operating ILMs exhibited rigid salary structures but showed a great deal of flexibility in moving workers between jobs at differing points on the classification: that is, the wages for workers were more flexible than the wages for jobs. The key finding is that wages represent a highly constrained adjustment mechanism. 5 While they would reduce total labour costs, they may increase labour costs per efficiency unit. This simple hypothesis has been used to explain the rigidity of real wages, they existence of dual labour markets, the payment of differing wages to similar workers and discrimination between distinct groups. Institutional This hypothesis argues that the labour market can be divided into two sectors - a primary sector composed of 'good' jobs and a secondary sector composed of 'bad' jobs. Wage and employment competition which is postulated to promote labour market efficiency in the neoclassical model, is therefore seen by Tthurow to give workers an incentive to hoard knowledge. The job competition model therefore emphasizes quantity adjustment rather than wage adjustment. Edwards (1975), for example, has contended that firms established ILMs to prevent the development of class consciousness among workers and to allow employers to maintain control over the production process. The ILM is therefore said to maximize the amount of labour (effort expended in production) obtained from a given amount of labour power (the capacity to perform useful work). This requires bureaucratic control, involving the development of formal definitions for the direction of work tasks and the use of institutionalized power to enforce compliance. Thus work habits become consistent with the form of bureaucratic control rather than with the actual work tasks. Future studies may want to examine the general pattern of workers' career mobility across industries or organizations in Taiwan and the diversities of internal labor market formations. Althauser and Kalleberg suggested that there are three characteristics of the ILM: (i) job ladder (2) entry only at the bottem.(3)promotion represents on improvement of technologe. 6