Does performance-related pay work in the public sector?
Authors: Beth Foley
, Kathryn Ray and Tiffany Tsang
04 December 2014
George Osborne’s announcement in last year’s spending review that reforms to progression pay would be taken forward to 'ensure that public sector workers do not receive pay increases purely as a result of time in post' represented a further step in an ongoing trend towards performance pay in the public sector. Performance-Related Pay (PRP) schemes — pay systems where an element of remuneration is based on performance — have long been of interest to economists and managers alike. Under the simplest logic, PRP aims to strengthen the link between rewards and productivity, motivating individuals to work harder to ultimately improve outcomes.
But can PRP schemes improve performance in the public sector? Forms of PRP began to be introduced into the UK public sector in the 1990s and, as this process accelerates today, our recent evidence review for the Office of Manpower Economics examines the evidence.
In the public sector, as in the private sector, evidence suggests PRP can lead to improvements in those outcomes which are directly incentivised. In education, this can mean increases in student test scores; in health, improvements in care quality. Yet, despite some positive impacts, our findings suggest a need for caution when applying PRP to the public sector. Not only is the evidence inconsistent – with much dependent upon organisational context and scheme design – positive effects are often small, sometimes short-lived and cost effectiveness data is rare. Moreover, a number of challenges and unintended consequences have been found, highlighting the complexity of public sector PRP.
1) Misallocation of effort occurs when PRP encourages employees to focus on incentivised tasks at the expense of others. The latter may be equally desirable but not incentivised, usually because they are not easy to measure. This is a significant problem in public services due to the complexity of desired outcomes, some of which may only be visible in the long-term. This raises questions about the feasibility of meaningful performance measures within public sector PRP. For example, in education, PRP schemes have raised student test scores, but have had less positive effects on co-operative behaviour among teachers, suggesting that PRP encourages ‘teaching to the test’ and a focus on ‘borderline’ pupils. Misallocation of effort may be minimised through introducing a broader range of targets, but there are trade offs here as schemes become complex and costly to administer. An example is the QOF in UK primary care, which has a range of targets across different domains of care, but is costly to monitor with modest positive impacts.
2) Collaborative activity: The delivery of public services tends to be a collaborative activity, which makes attributing individual responsibility for performance difficult. There is a risk that PRP schemes can act as a disincentive to teamwork and, if distribution of rewards is viewed as unfair, demotivation can occur. Responses to PRP may also relate to employee characteristics; for example, in education, male teachers are more likely to support PRP and respond to it positively than their female counterparts. This has wider implications for the acceptance and success of PRP in female-dominated public services.
On the other hand, some PRP schemes can result in improvements in team relations and work organisation. Trials of team performance targets in the UK civil service have seen productivity improvements as a result of managers distributing tasks among team members more efficiently. Indeed it has been suggested that successful PRP schemes may operate via better management practice. Designing PRP systems in consultation with staff and allowing input on the targets they consider most important can help overcome perceived unfairness in PRP systems, and there is some evidence that collaborative PRP schemes are more likely to be effective.
3) Questions of motivation: PRP’s focus on financial reward alone risks neglecting the wider range of factors that motivate employees, including peer effects, management practices, perceptions of fairness and intrinsic motivation. The latter is particularly significant in the public sector, where many staff derive motivation from belief in the intrinsic value of their services. Evidence on whether PRP ‘crowds out’ intrinsic motivation, ultimately resulting in poorer performance, is inconclusive and partly depends on the extent to which performance targets align with professionals’ own goals. Nonetheless, other ways of motivating performance may utilise intrinsic motivations more effectively than financial rewards; thus more research is needed on the relative cost-effectiveness of different performance improvement strategies.
4) Scheme design: Finally, those designing and implementing PRP should consider that different ways of measuring and rewarding performance have different effects on behaviour. For example, relative measures are more likely to stimulate improvement among poorer performers, while threshold measures stimulate improvement among the middle range. Measures reliant on a manager’s subjective assessment of performance can be perceived as unfair, leading to diminished motivation. Ensuring targets align with overall improvement goals, employing a balance between objective and subjective measures, and involving employees in discussions about the most appropriate metrics help to address these issues.
So, what does the future hold for PRP in the public sector? Our review suggests that such policies are far more complex than they first appear, that evidence is frequently inconsistent and that much depends on the design and context of the scheme. Full consideration of alternative means of performance improvement, such as developing good management practice, is also vital when weighing up the costs and benefits of public sector PRP.
All blog posts for this author