بهره وری کار وقتی که دانش سیستم های پاداش مختلف متغیر است: گزارش از آزمایش های اقتصادی

Recent evidence suggests that both buyers and sellers of labor services prefer uniform wage payments. This study further examines this topic, by offering different monetary rewards for identical activities, and by varying knowledge of the alternative reward systems. The mode and proffered pay scale(s) had little or no impact on work productivity when subjects were ignorant of the alternative reward systems. However, work productivity varied significantly by mode and pay scale when subjects were aware of the alternatives.

It has always been difficult to develop a wage payment system that is tailored to each individual worker's productivity. Sir John Hicks recognized this problem long ago when he suggested that most collective work processes require incentive systems that have collective motivational elements. Current evidence from the fields of economics and psychology suggests that Hicks' insights – and those of “old-fashioned” labor economists as well – remain salient. For example, economic studies examining ultimatum and trust games suggest that subjects do not always act to maximize individual economic returns (Thaler, 1998; Camerer and Thaler, 1995; Fehr et al., 1993). Psychological studies also find factors such as perceived `fairness', or `equity' interact with personal motivators to determine workplace productivity (Ambrose and Kulik, 1999; Adams, 1963; Pritchard et al., 1972). Fehr and Falk's (1999) experimental research built on these findings by examining employer wage-setting practices. They found that subjects who hired labor services in the experiment clearly understood the value of positive reciprocity, and often refused to accept low wage bids of workers fearing unemployment. The emerging story is that buyers and sellers alike prefer relatively uniform wage payments because the incentives implicit in specific labor contracts depend on employer and employee perceptions of their impact on collective work effort. Contrary to simplistic neoclassical labor economics, it would be unreasonable to expect that this type of wage setting policy would clear the labor market. This paper contributes to the literature by examining in a controlled environment the effect of knowledge of alternative reward systems (i.e., uniform vs. non-uniform reward systems) upon work productivity. For identical work activities (collating and clipping stacks of 15 sheets of paper numbered from 1 to 15), subjects were compensated according to one of three different reward systems ($.27 or $.81 per unit individually produced, or an $8.10 flat rate, regardless of individual work productivity). Half of the subjects were informed of the alternative reward systems, and half of the subjects were not informed of these alternatives. It was found that the mode and amount of reward had little impact on work productivity, when subjects were uninformed of the alternative reward systems. However, when subjects were informed of the alternatives, those paid $.27 per produced unit or the $8.10 flat rate decreased their work productivity, while those paid $.81 per produced unit increased their work productivity.

When Sir John Hicks wrote the statement that opened this article, the future Nobel Laureate economist was sharply criticizing his own 1930's work on wages (Hicks, 1932). He took issue with his previously narrow focus on the characteristics of individual workers. As an alternative, his revised analysis embraced some of the findings of the dominant labor economics paradigm of the 1950's and 1960's (Kaufman, 1988). This school of thought focused on efforts to devise predictable reward systems that induced effective work, and were perceived as fair by workers and employers alike. By the end of the 1960's, this focus on group incentives gave way to human capital theory and, later, efficiency wage theory. Even economists influenced by Marxist economic reasoning began to concentrate on individual workers' response to the “cost of job loss” (Bowles, 1985). Our work suggests that it would be useful to return to a consideration of variables that might influence both individual and collective work performance. Such a return to the concerns of earlier institutionalist economists need not contradict human capital or even efficiency wage theory, but such contemporary approaches concentrate too intensively on the individual to provide a complete description of the personal and situational factors that determine workplace productivity. Our results indicate that work productivity (at least within the context of this experiment) is, to an appreciable degree, independent of the piece or flat rate reward system by which subjects are paid, as long as there exist no referents that permit comparisons of alternate reward systems. Indeed, the experimental task – collating and clipping randomly numbered sheets of paper – was specifically selected because of its low intrinsic value, rendering it difficult for subjects to spontaneously decide what an appropriate wage rate might be in such a situation. Experimenter observations confirm that subjects who were uninformed of the alternate reward systems performed the experimental task with humor and enthusiasm, resulting in statistically identical levels of work productivity. However, when subjects were informed of the alternate reward systems, work productivity levels changed dramatically. Those paid the 30 Dirhams flat rate showed a significant decrease in their work productivity, while those paid 3 Dirhams per produced unit showed a significant increase in work productivity, and those paid 1 Dirham per produced unit showed no significant change at all.3 It is tempting to conclude that the absence of alternative reward systems enabled subjects to more easily develop `team spirit' (regardless of their respective reward systems), thereby enabling them to maintain relatively constant levels of work productivity in all three uninformed conditions. That is, when subjects are unaware of alternative reward systems, work productivity is primarily determined by the degree of `team spirit' among subjects (which can be assumed to be equal in all three uninformed groups), and not by the reward system itself. However, knowledge of the alternative reward systems allowed informed subjects to perceive more clearly the link (or lack thereof) between individual work effort and pay. Informed subjects paid 1 Dirham per produced unit realized the inequity of the available reward systems, and this served to inhibit their motivation to perform the experimental task. In a similar manner, informed subjects who were paid 3 Dirhams per produced unit, well aware of their more advantageous reward system, were highly motivated to perform the experimental task. Those subjects paid the 30 Dirhams flat rate, aware (from their perspective) of the complete lack of relationship between individual effort and reward, were also inhibited in their performance of the experimental task. The results of our experiment imply that uniform reward systems of whatever kind will yield approximately similar work productivity results. However, if workers are aware of alternative compensation possibilities, then their productivity will vary across reward systems. In particular, high-end piece rates will promote higher productivity levels. This is consistent with the empirical findings of Lazear (1996). The implementation of a relatively generous piece rate system may do more than raise work productivity, however. More poorly paid piece rate workers and flat rate workers might respond by reducing their work performance. Perhaps the piece rate system has the effect of inducing all workers to think of themselves as individualistic, self-seeking actors, rather than as team players. Given the positive and negative effects that stem from piece rate systems, it is not obvious that such compensation incentives will actually increase total labor productivity. This may especially be true if some tasks produce measurable productivity that permit performance-based pay, while others do not. It is important to note that the subjects' personal characteristics, the unusual locale of this study, and the brief, non-repetitive nature of the experiment itself limit the external validity of the obtained results. Further, the trivial nature of the experimental task is clearly not applicable to many work processes in the “real world”. The results, however, are suggestive. At the very least, further exploration of the complex links between knowledge of alternative compensation systems, pay uniformity, notions of collective fairness, and workplace productivity would seem to be a potentially fruitful area of investigation for both economists and psychologists.