Key message 2 – Well-designed financial rewards linked to job performance can help improve public sector outcomes

 
It is well known that performance-based incentives can help to motivate better behaviours from workers in the public sector just as in the private
sector. In order to limit political influence over salaries, however, civil service salaries are typically  based on rigid, formulaic pay scales with limited room for using financial incentives to reward  performance. Nonetheless, governments are increasingly developing creative schemes to attract more qualified staff and motivate good behaviour based on outcomes achieved. In places where a large scale restructuring of government salary structures may be difficult, leveraging even just a small portion of the wage bill toward merit-based incentives may yield surprising results, particularly when performance can be accurately measured.

A number of studies have documented the impact of policy experiments that reward performance based on outcomes in areas including health, education, and public finance. Policy findings emerging from this research show that it is important to ensure that financial incentives are simple, well understood, and linked to clear, measurable targets.

Research on health centres in Rwanda suggests that offering performance-based pay to health facility staff might lead to improved provision of pre- and post-natal care (Basinga et al., 2011; Gertler and Vermeersch, 2012). A study in Indonesia found that offering incentive payments to villages based on their performance on a set of health indicators improved outcomes such as prenatal visits and malnutrition rates, although these effects became less significant over time (Olken, Onishi and Wong, 2014).

In the education sector, Muralidharan and Sundararaman (2011) examined a large-scale programme in India, where public school teachers’ pay was partly determined by student test scores. After two years, they found that the relatively small incentives – equal to 3% of teachers’ annual salaries on average – substantially improved student learning in math and language.

In the area of public finance, an IGC study by Khan, Khwaja and Olken (2014) found that offering  incentive payments to tax collectors in Pakistan helped increase tax revenues – a major challenge for the government (See Figure 2). In areas where tax collectors were offered large incentives, revenue growth was 13 percentage points greater than in areas without incentives. Importantly, the intervention was cost-effective, with revenue gains significantly exceeding the cost of the incentives.

While a common fear is that incentives might ‘crowd out’ intrinsic motivation to get a job done well, there is in fact only limited empirical evidence to support this in practice. Instead, the latest evidence tends to suggest that incentives can in  fact leverage intrinsic motivation (Ashraf et al., 2014; Rasul and Rogger, 2016).

Another key lesson underlying all of this work is the importance of putting into place strategies  for measuring the results and cost-effectiveness of performance-based incentive schemes over time –including any unintended consequences – in order to continually inform policy decisions on whether these schemes should be dropped, adjusted, or scaled up.

Figure 2

Khan, Khwaja and Olken, 2014