"Performance-based Evaluation" - A Magic Wand?

TSURU Kotaro
Senior Fellow, RIETI

One point of view toward structural adjustment is in respect to how economic entities become efficient. Organizations that were previously breeding grounds for inefficiency have been restructured, and "performance-based evaluation" (the approach whereby people are evaluated and rewarded based on their performance) has become recognized as a means to provide incentives to the people within these organizations. There are moves to apply this system to public institutions and public servants, as well as organizations and employers in the private sector, where performance is relatively easy to measure (e.g., see Economics Review No. 7 for the relationship between independent administrative institutions and performance-based evaluation). And yet, is "performance-based evaluation" really such a decisive factor in improving efficiency? In this article, I will examine this issue principally from the perspective of "personnel economics."

- Problems associated with "performance-based evaluation"

Let us consider a scenario in which "performance-based evaluation" is applied to the workers of a company. The simplest form of "performance-based evaluation" is piece rate pay, whereby people are rewarded in line with their output. However, as will be described, there are a number of problems with piece rate pay. First of all, performance often depends on the results of teamwork, and it is difficult to clearly delineate and verify the contribution made by each worker. Where performance is difficult to verify, the incentive arises for the employer to hold back on rewards by undervaluing performance.

Even in cases where performance is verifiable, the kind of piece rate pay in which workers entirely bear the risk of fluctuation in performance is, at the same time, not necessarily the best choice for them, given that risk cannot be diversified in terms of human capital (the trade-off relationship between risk-sharing and incentive). If "performance-based evaluation" is applied to a particular job among the several a worker performs, the increased incentive for the worker in doing the evaluated job will mean the other jobs end up being neglected. In other words, a bias is produced in the sense that effort goes disproportionately into jobs and aspects of work (quantity over quality) where performance is more easily visible.

One approach to the difficulty of verifying performance is to adjust the method of evaluation. For example, a relative evaluation could be used when it is difficult to make an absolute evaluation due to the substantial margin of error in producing the evaluation. A typical example of this is the tournament-type approach in which workers compete with one another and are rewarded according to the ranking of their results. However, although the tournament-type approach avoids the problems of undervaluing production inherent in piece rate pay, it does have a drawback in that effort is not always rewarded, since winners and losers are determined by tiny differences in performance while the rewards vary considerably. Where it is difficult to achieve objective evaluations that are verifiable (i.e., quantifiable), an alternative approach is to use subjective evaluations. Subjective evaluations mean the recipient of the evaluation is closely monitored by his or her superior. Yet, the scope for this evaluation to become distorted due to "favoritism" on the part of one's superior grows as the evaluation becomes more subjective and this can easily generate inefficiency as staff take pains to "ingratiate" themselves with their superior in the attempt to win "favor" (a kind of rent-seeking activity). If aggravated, circumstances such as these will substantially inhibit the level of morale and efforts of workers due to a sense of unfairness with the evaluations. Viewed in this light, we can see that "performance-based evaluation," however simple it seems to be, brings with it a number of different problems. "Personnel economics" instead emphasizes the idea that the "bureaucratic approach," including evaluations based on seniority and job rank-based promotion, has advantages for curbing opportunistic behavior and "favoritism" on the part of the evaluator.

- The need for organizational restructuring before the introduction of "performance-based evaluation"

What I describe above looks mainly at the scenario where "performance-based evaluation" is applied to the workers of a private sector company. What then are implications for employers in the public sector (i.e., public servants)? (For information on personnel and incentive systems in the bureaucracy, see RIETI Discussion Paper 01-E-05 and "Why Bureaucrats Choose Forbearance Policy.") The central theme of the Outline Reform of the Public Servant System published at the end of last year is the application of a system based on the ability and performance of public servants. Nevertheless, it is difficult to imagine such a system working given the limitations of "performance-based evaluation" described above. This is because performance is even more difficult to measure and verify in the case of public servants than in the private sector, and the wide-ranging objectives targeted by organizations mean that the detrimental effects of bias resulting from the application of the "performance-based evaluation" to specific objectives are even greater than in the private sector. Still, more serious is the potential of a disproportionate emphasis on subjective evaluations that would be expected as a matter of course to allow "favoritism" and "ingratiation," from which these organizations have up to now been relatively immune thanks to the seniority system and a centralized, long-term system of evaluation (rewarding people with positions), to thrive. Were this to happen, people's sense of the unfairness of evaluations would increase and this would clearly be detrimental to operations that require the coordination and cooperation particularly requested of the public sector.

Public sector employers considering the introduction of "performance-based evaluation" to even a small degree must, therefore, first reform their organizations themselves. In other words, if we accept that the diverse and vague nature of the objectives to be pursued by public institutions contributes to the difficulty of evaluating those who work in them, the first priority is to divide these organizations into modules in order to clearly identify their missions and make it easier to evaluate people justly. As such, the perspective that workers' incentives vary according to the organizational form is an important one.

June 18, 2002

June 18, 2002