How to pay public servants: Rewarding ability and performance
Why the public sector should end flat pay schemes. In today’s blog, Sheheryar Banuri and Philip Keefer use experimental evidence from Indonesia to illustrate that in cases when monitoring job-performance proves too costly, paying for ability can be as effective, but cheaper to implement, as pay for performance schemes.
This post forms part of a cross-blog series on the 2030 Agenda for Sustainable Development run by the IGC, Africa at LSE, and South Asia at LSE blogs. As part of the series, this blog takes a closer look at the mechanisms that support #Goal16 - the promotion of peaceful, just, and strong institutions. View more posts in this series.
How should we pay public servants? Pay schemes in all organisations, including the public sector, have two objectives. First, they need to attract high quality workers; second, they need to provide incentives for these workers to exert effort. In the public sector, this is easier said than done.
Consider common perceptions of the public sector and public sector jobs. Public sectors are inefficient, uncaring leviathans, where customer service is an alien concept and hardly any work gets done. Public service jobs are stable and risk-free career tracks, where effort is divorced from compensation and shirking has few repercussions. These perceptions persist in a variety of settings and compensation schemes. We therefore conducted an experiment with a novel subject pool in Indonesia to precisely address the question, what effect does changing compensation schemes have on the recruitment of high quality workers and on the effort of public sector workers?
We examined common compensation systems that incorporate the unique characteristics of most public sector pay systems. First, the public sector tends to privilege equity among public sector workers, with a secondary focus on efficiency. Second, unlike private sector firms, where raising revenues is a primary objective, the public sector usually places citizen welfare as its primary objective. This is as murky and ill-defined as it sounds, but no less important. Nevertheless, as a consequence, the measurable outcomes and clear targets that are taken for granted in the design of private sector compensation systems are not always present in the public sector. Third, public sector jobs tend to pay significantly less than similar jobs in the private sector, particularly in developing countries.
Compensation schemes
With these constraints in mind, we examined stylised versions of three major compensation systems commonly found in the public sector. The most basic system (and indeed, the most frequent in developing countries) exhibits flat pay, giving every public sector employee performing similar functions the same compensation, with no adjustments for effort. Employees then progress in their career through seniority and promotions, but new entrants performing similar functions are paid the same base wage.
The second compensation scheme contains two essential elements: differentiated rewards according to office, and assignment to office according to merit. This ability-based or “Weberian” system is associated with Max Weber. Common to some developing countries, this pay scheme provides differentiated pay based on some ability measure (usually entrance examinations), with higher ability candidates getting higher pay. For example, in Indonesia, candidates who demonstrate greater ability on exams can enter the higher-paid Ministry of Finance rather than the lower-paid Ministry of Agriculture. As in the flat system, however, once pay grades are assigned, they do not depend on effort.
The final public sector pay scheme is commonly known as pay-for-performance, where pay is explicitly linked to effort, much like in the private sector. This pay scheme is increasingly evident around the world. For instance (perhaps the most notorious example), up to half of civil servant pay in Singapore is based on performance bonuses.
At first blush, the superiority of pay-for-performance scheme seems obvious. Why would any organisation not choose a system that provides direct extrinsic incentives for effort? However, it has two potential drawbacks. First, it is expensive in terms of monitoring costs, particularly for a large range of jobs in the public sector that lack well-defined goals. Second, increasing extrinsic incentives may “crowd out” intrinsic motivation to exert effort (Frey and Oberholzer-Gee, 1997). In the public sector, public service motivation has been found to be an important source of motivation for workers (Perry and Wise, 1990; Ashraf, Bandiera, and Jack, 2012), and managers are (rightly) concerned that performance-pay may undercut public service motivation yielding sub-optimal outcomes.
In a recent paper, we measure the effort of college students who have committed to joining the Indonesian public sector under these three different pay schemes (flat, “Weberian”, and performance pay). These future public officials are asked to engage in a computerised effort task, popular in the economics literature (Gill and Prowse, 2014). All participants are randomly assigned to one of these three tasks, from which we measure their effort. We then ask them to choose between a “public sector” pay scheme, and a “private sector” piece rate. Effort exerted under the public sector pay scheme generates donations to a well-known Indonesian charity (to induce public service motivation).
Is it better to motivate or compensate?
As one would expect, performance pay generates the highest amount of effort, significantly higher than both flat and ability-based pay. This is not really a surprise, since individuals typically respond to incentives. However, when we unpack this result, we find that the effects of the performance incentives operate solely on under-motivated participants – those who attached little value to the social mission that we gave to subjects. Participants exhibiting higher than average prosocial motivation were completely unaffected by the compensation scheme. However, participants exhibiting lower than average prosocial motivation exerted greater effort in the performance pay scheme.
This result suggests that the effects of performance pay on public sector workers is likely to be concentrated among under-motivated employees, dampening the effect of performance pay overall (depending on prosocial motivation levels). Furthermore, this means that selection effects should be important: if public sectors attract motivated applicants, performance-based pay will matter less.
Conclusion
So which of the three pay schemes is effective in attracting high-ability, motivated applicants? The answer is that both the differentiated pay schemes (pay for ability and pay for performance) are equally effective: the ability and motivation of subjects who prefer the public sector are indistinguishable across the two pay schemes. In contrast, flat pay attracts more low ability workers. Moreover, once sorting takes place, pay-for-ability and pay-for-performance workers exert statistically identical levels of effort.
So, despite the Sturm und Drang surrounding pay-for-performance pay systems, there is little to choose between pay-for-ability and pay-for-performance. Both perform well once workers are allowed to switch between private and public sectors. However, pay-for-ability schemes are cheaper to implement: the salary costs per unit of effort were lower under the pay-for-ability schemes.
Thus, it would seem that the preferred pay scheme really depends on how difficult it is to measure performance. In particular, the evidence strongly supports pay-for-ability schemes when performance is difficult to measure. Even more clearly, however, is the importance of moving away from compensation schemes that reward neither ability nor performance.
Further Reading:
Ashraf, N., Bandiera, O., & Jack, B. K. (2014). No margin, no mission? A field experiment on incentives for public service delivery. Journal of Public Economics, 120, 1-17.
Banuri, S., & Keefer, P. (2015). Was Weber Right? The Effects of Pay for Ability and Pay for Performance on Pro-Social Motivation, Ability and Effort in the Public Sector. (May 11, 2015). World Bank Policy Research Working Paper, (7261).
Frey, B. S., & Oberholzer-Gee, F. (1997). The cost of price incentives: An empirical analysis of motivation crowding-out. The American economic review, 746-755.
Gill, D., & Prowse, V. (2012). A structural analysis of disappointment aversion in a real effort competition. The American economic review, 469-503.
Hasnain, Z., Manning, N & Pierskalla , J. (2012). Performance-related pay in the public sector: a review of theory and evidence. World Bank Policy Research Working Paper, (6043).
Perry, J. L., & Wise, L. R. (1990). The motivational bases of public service. Public administration review, 367-373.
This article gives the views of the author, and not the position of the IGC.