Organisation de Coopération et de Développement Économiques
Organisation for Economic Co-operation and Development
English – Or. French
PUBLIC GOVERNANCE AND TERRITORIAL DEVELOPMENT DIRECTORATE
PUBLIC MANAGEMENT COMMITTEE
Working Party on Human Resources Management PERFORMANCE-RELATED PAY OF GOVERNMENT EMPLOYEES: ASSESSING REFORMS ACROSS OECD MEMBER COUNTRIES
Paris, 7 October 2003
For further Information, please contact Dorothée Landel, OECD Secretariat Tel: + 33 1 45 24 82 43, E-mail: firstname.lastname@example.org
EXPERTS MEETING – 7 OCTOBER, 2003
PERFORMANCE-RELATED PAY OF GOVERNMENT EMPLOYEES:
ASSESSING REFORMS ACROSS OECD COUNTRIES
Understanding the scope of the meeting
. The meeting aims to review the different types of performance-related pay policies (PRP) that exist, 1
and to assess their strengths and weaknesses. Presentations and discussions at the meeting will focus on implementation and outcomes issues linked to PRP.
2. The terms “performance-related pay” (PRP) and “performance pay” are used to refer to a variety of systems linking pay to performance. PRP systems refer to the variable part of pay which is attributed each year (or on any other periodic basis) depending on performance. PRP systems are generally applied at the individual employee level, but are also sometimes applied at the team/unit level, or at the organisational level. PRP excludes: (i) automatic pay increase by grade promotion and in step (in grade) pay increases (not linked to performance), (ii) various types of allowances which are attached to certain posts or certain working conditions (i.e. overtime allowances).
3. Presentations and discussions at the meeting will rely on country reports. Each session of the meeting will consist of two country presentations, followed by a lead discussant, and a general discussion. 4. The meeting is a starting point for the project on pay for performance. Conclusions of the meeting will fit into a report on “Performance-related pay of government employees: assessing reforms across
OECD countries”, prepared by the Secretariat in collaboration with Professor David Marsden, Director of the Industrial Relations Department at the London School of Economics and Political Science. Delegates are invited to express ideas and recommendations on the scope and focus of the final report on PRP. 5. A preliminary version of the report will be sent to countries for review in the second quarter of 2004. The draft version of the report will be presented at the HRM Working Party meeting in October 2004.
Main points to be discussed at the meeting:
6. The experts’ meeting has two main objectives:
- To present an overview of performance-related pay policies of government employees in OECD
- To discuss the following points in detail and to draw conclusions regarding them: ; What is the relative importance of the following reasons for introducing performance related
pay systems? Improving performance is usually one of the main reasons, what are the other reasons
for introducing PRP?
- Introducing organisational or cultural change
- Improving recruitment and retention of staff
- Improving accountability of high civil servants
- Containing salary costs and reducing the pay bill of the organisation, through better control over
- Other types of managerial or political reasons
; What implementation problems were met? (difficulty in the measurement of performance,
homogeneity and inflation in the ratings and rewards, resistance of staff, lack of transparency,
incentives too low to really be motivating, insufficient budget dedicated to PRP), and has it been
possible to overcome these problems? If yes, how has it been possible? If no, has the solution
sometimes been the withdrawal of the PRP scheme?
; What has been the impact of PRP policies? Have evaluations of PRP policies been undertaken and
what were their conclusions? How have public organisations been able to measure change linked to
PRP, in a context where individual and collective performance is difficult to assess? ; What conditions are most favorable for the successful implementation and impact of
performance related pay policy?
- An already well-functioning performance management system
- Strong collaboration with unions
- Involvement of staff in the reform
- Clear communication of the objectives and functioning of the PRP scheme to staff
- Linkage between individual objectives and organisational ones
- Measurement of performance against specific goals or with quantitative measures of
- Clear feedback to staff on the allocation of rating and rewards
- Specific budget dedicated to PRP
- High level and well differentiated reward
- Possibility of internal and external review
- Combination of pay with other incentives such as promotion
7. Public sectors in OECD countries have come under pressure from four main sources since the late 1970s - early 1980s:
i) labour market pressures have resulted in problems for recruiting and retaining staff in parts of the
public sector in many countries;
ii) macro-economic considerations have created a general pressure to improve wage flexibility in
both the public and private sectors and a specific pressure to hold down public sector pay costs as
an element in reducing public expenditure and government deficits. The public sector pay bill is
indeed the biggest budgetary item in almost all countries.
iii) there are pressures to improve managerial efficiency in the public sector so that better quality
services are delivered to users in a cost-effective manner.
iv) there appears to have been some lowering of public confidence in public institutions, and a
popular perception that public servants are poorly motivated compared with their private sector
8. An important feature of most public sector modernisation plans is a move towards management control by emphasising performance rather than rules and procedures. Most OECD countries have introduced a variety of new policies to enhance public sector performance which have taken different shapes and paths. They have been implemented at the governmental, organisational, team and individual levels. Performance management strategies include corporate and unit planning, the introduction or re-design of employee performance appraisal systems, the linkage between performance and promotion and/or between pay and performance.
9. The linkage of pay to performance targets is not a new idea. Incentive pay approaches can be traced
to Taylor's scientific management at the beginning of the century. Performance pay approaches have been developed more extensively in the private sector since the economic crisis of the mid 1970s, in a context often referred to as the “new managerialism”. In the 1980s and 1990s, these approaches have been
progressively extended to public sector organisations.
10. Reasons for introducing performance-related pay policies in the public sector are diverse, some being more official than others. The main rationale for establishing performance pay policies is to improve performance and motivation of managers or employees. This objective is coupled with the need to provide ministries/departments and individual managers with greater flexibility to recognise and reward individual or team performance contributions. The introduction of PRP at the employee level is also a way to forge a closer link between individual job goals and organisational goals.
11. Another reason for introducing performance-related pay policies, especially at the managerial level, is the need for the public sector to be able to attract and retain talent currently being rewarded for performance in the private sector. This is especially important in the context of an increasing gap in pay differentials between high-skill workers in the private and public sector, which constitutes a serious threat for the recruitment of high quality managers in the public sector.
12. PRP can be used to assist or stimulate organisational change. It might help to achieve cultural change by letting employees know the values and expectations of the organisation. PRP create alternative career paths, where performance is more important than grade and seniority. This has an important impact on strategic human resources management, and constitutes strong incentive for improved staff development and training.
13. In some cases, improving performance can be a secondary objective for introducing PRP. The real motivation can be to use PRP as an instrument to obtain political support for more funding or more flexibility over pay. PRP may also be seen as a way to have better control of the pay bill. The introduction of PRP strategies can indeed be a way to contain salary costs by reducing the incidence of automatic progression through salary levels. The individualisation of pay may sometimes also be a way to circumvent the collective bargaining process, thus reducing the influence of trade unions and re-establishing managerial control.
History of PRP policies in governments of OECD countries
114. Results of the recently completed OECD survey on strategic human resources management show
that more than four-fifths of OECD countries have developed a performance appraisal system for employees. Two thirds of countries have a system that, at least in formal terms, links individual performance with organisational goals. In 80 % of cases, performance management systems are “somewhat linked” or “strongly linked” with pay. In countries where no reforms of the pay system have taken place,
priority is often given to promotion policies, position re-classification and other non-pay instruments. 15. The first experience with pay for performance precedes the 1980s-1990s period. In Canada, for example, a formal system of merit increments for civil servants was first introduced in 1964. In France, the earliest version of the General Regulations for civil servants, enacted in 1946, specified that individual or collective bonuses could be granted periodically to civil servants in the recognition of exceptional performance. In Japan, there has been a provision for a “diligence allowance” for national public employees since the early 1950s. However, most of these plans (except in Canada) were ad hoc granting of rewards and did not allow financial rewards to be included in salaries – or in the calculation of pensions -
on a longer term basis.
16. The first wave of introduction of PRP policies across OECD countries took place in the late 1970s and early 1980s. PRP schemes were introduced in the United States in 1979 for the Senior Executive Service and in 1981 for middle-level managers. Denmark, the Netherlands, New Zealand, Spain, Sweden and the United Kingdom, all adopted some form of performance-related pay policy in central government during the 1980s. Experimental policies were tried in the 1980s in selected agencies in Finland, Italy and the German postal service. A second wave of introduction of PRP schemes occurred in the late 1980s and early 1990s, with Australia, Finland, Ireland or Italy adopting PRP policies. A more recent wave took place in the late 1990s, with countries such as Germany, Korea, Switzerland and Hungary adopting PRP policies. Norway also adopted performance-related pay elements for top level managers at the end of the 1990s. 17. It is likely that the use of performance-related pay will continue to spread in the public sectors of OECD countries. Yet, there has been very little evaluation of whether improvements have actually occurred and whether benefits justify costs in public sector organisations. Studies reveal a similar lack of evaluation of many private sector schemes. Despite this lack of evaluation, there seems to be widespread support both among elected officials and public sector managers themselves for the principle of linking pay to performance.
PRP mechanisms in OECD countries
18. An important contrast in the performance pay systems in central/federal governments concerns the degree of centralisation or standardisation. Some longer running schemes, particularly in Canada, the United Kingdom, and the United States, have been centrally imposed on line departments and agencies, and have adopted fairly standardised approaches to performance appraisal and to the allocation of bonuses
1 OECD, Survey on Strategic Human Resources Management, PUMA, HRM, 2002 (3)
and merit increments. However, long-running standardised PRP schemes have in some cases tended to evolve towards more flexible approaches to PRP in the past 10 years. In Denmark, the Netherlands, New Zealand and Sweden, schemes are less standardised and allow individual departments and agencies greater latitude in the design and implementation of policies.
19. The extent of PRP application and its target group differs from country to country. Initially, PRP policies were essentially directed only at senior management and middle management levels. It is still the case in Canada, Ireland, and Italy. Progressively PRP policies have tended to include all categories of public employees, even if policies are often more effectively applied for managerial staff. In the Survey, the following countries stated that their performance-related pay policies include all categories of government employees: Australia, Denmark, Finland, Germany, Hungary, Korea, Spain, Switzerland, United Kingdom, United States. Countries have either developed a single PRP policy for both managers and non managers, or two separate schemes, such as Denmark and the United States. In some cases, such as in Finland, top civil servants are excluded from the scheme.
20. The form of payment also varies considerably across countries. Two main types of system can be identified (in many cases the two are combined in a single policy):
; Merit increments, which take the form of either a fixed or variable increase which is added to
and can become a permanent part of the basic pay. Pay progression is therefore linked, wholly or
in part, to individual performance rather than seniority.
; Bonuses, which are one-off payments which are not consolidated into basic pay. They may be
expressed in either cash terms or a percentage of the basic pay. They can be distributed
independently of the level of salary.
21. The OECD HRM survey shows that in many countries, i.e. Australia, Denmark, Hungary, and the
United States, there is no special fund for PRP. Rather, all pay increases must be funded from within agency budgets. Canada, Finland, Ireland, Korea, and Spain on the other hand have provided a special budget for PRP.
22. In the right circumstances, incentives such as pay lead to enhanced motivation, effort, and performance. In the real world, however, incentives often have dysfunctional effects, and do not necessarily lead to improved motivation and performance. If the incentive system is badly designed, or badly communicated to staff, it can have serious adverse effects that can lead to degradation in individual or collective performance.
23. Previous OECD reports on PRP show that there are recurring problems with the implementation of 2performance-related pay policies, notably a lack of discrimination and an inflation in performance ratings, and a failure to establish a clear relationship between pay and performance. In many cases, PRP policies have been found to fail to satisfy key motivational requirements for effective performance pay. Overall,
previous OECD reports revealed widespread dissatisfaction with PRP schemes and raised important questions as to the impact of these schemes on the motivation and performance of managers. 24. A number of academics have written on the weaknesses and shortcomings of performance-related 3pay systems. Other experts tend to assert that performance pay systems have a positive impact on
2 OECD, Private Pay for Public Work, 1993 and OECD, Performance Pay Schemes for Public Sector Managers, an
evaluation of the impacts, 1997.
3 See for instance Alfie Kohn, Punished by rewards : the trouble with gold stars, incentive plans, A's, praise and other bribes, 1993
4performance. What is more consensual is the fact that despite the growing use of performance-related 5. merit schemes, their true impact remains sadly under-researched
25. It seems from the available literature that both collective and individual PRP policies have risks:
; Under an individual performance pay policy, resources and information are more likely to be
hoarded than shared. Marsden and Richardson reported that many workers said that merit
produced a degree of staff jealousy and a decline in morale. Employees can see PRP as an
entitlement, not as a reward. Instead of trying harder, low performers may rationalise their poor
performance evaluations as merely a sign of incompetence or bias on the part of those
conducting the performance appraisals. On the other hand, high performers may sometimes
endure negative social sanctions. Other possible side effects are the neglect of job aspects not
covered in performance goals, or reporting of invalid data on performance.
; Group-based rewards can have dysfunctional effects as well. They can promote regression to the
mean rather than outstanding contribution. Low performers may have little incentive to obtain
training and thus raise their contribution. High performers may become vigilant police officers
who pressure the low performers to try harder. As a result, low performers may feel tremendous
pressure and scrutiny from other group members, which may further inhibit improvements in low
26. Risks associated with PRP are often exacerbated in public sector organisations. Greater individualisation of pay raises the issue of what performance or productivity criteria can be applied to public sector operations, especially service delivery and policy advice functions, as a basis for more differentiated pay. When performance is difficult to measure on the basis of quantitative criteria, there is a risk that its measurement may be based only on subjective assessment, which lowers the credibility and the impact of the system.
27. In the public sector, the problem of PRP funding can also be very critical to its success. Depending on the budget situation of countries, the amount of performance pay can vary a lot depending on the years. The insecurity of PRP funds can severely damage their credibility in the longer term. 28. The inflation of ratings and rewards often tend to produce increases in personnel costs, even if there are formal payment ceilings set. Performance-related pay schemes are thus often more costly than expected. 29. Also, different surveys show that pay ranks low in employees’ expressed preferences of work conditions in the public sector. Pay is not valued as much as many other work characteristics, such as
expectations of job stability, recognition for one’s achievements and respect and fair treatment from one’s colleagues. At the managerial level, pay seems to be less valued than challenging job opportunities and a sense of accomplishment.
30. However, despite all these potential risks and adverse effects, there is a strong body of opinion that performance pay policy remains a sound idea, under certain conditions. When it is well designed and applied, PRP can potentially be a tool for improving performance and strategic management, and a lever for organisational change. The question is really under what conditions a PRP policy can be successful in terms of improving staff performance, and whether a system of pay for performance can be implemented in all kinds of public sector organisations. Certainly, the political, institutional, and wider cultural
4 See for instance Locke, E. and Latham, G., A theory of goal setting and task performance, 1990 or Milkovitch, H.
and Wigdor, A., Pay for Performance: Evaluating Performance Appraisal and Merit Pay, 1991
5 Marsden, D. W. and Richardson, R., The motivational effects of Performance-Related Pay in the Public Sector; A
case study of the Inland Revenue, 1992
characteristics of each country and organisation have to be well taken into account before introducing any kind of performance pay strategy.
31. Available research, and previous OECD studies on the topic, show that several criteria are fundamental in applying PRP in an efficient way :
; PRP needs to be applied only where an effective performance management by objective and
feedback is well functioning, in order to reinforce a motivational system in which intrinsic (non
monetary) rewards exist, such as reorganization of work processes, training, employee
involvement/consultation in decision-making, two-way communication, opportunities to
contribute ideas, career development plans and goal setting.
; Organisations should provide clear guidance to participants in PRP schemes on the standards
which are to be used to discriminate between levels of award, and on the distribution of awards:
communication and involvement of staff in the reform are essential.
; The PRP policy needs to be seen as fair and equitable by all employees.
; Performance should be assessed against specific goals, or with quantitative measures of
; In some cases, consideration should be given to replacing individual PRP schemes with PRP
schemes that reward good performance at the team or organisational level.