Economic and Political Weekly November 15, 2003 4859 T he non-performance of performance agreements in the US government remains a mystery in the ever-expanding universe of ‘reinventing government’ initiatives world- wide [Osborne and Plastrik 1997, Osborne and Gaebler 1992, Kettl 2000, OECD 1994 and Scott 1996]. This paper goes beyond the explanations based on institutional analysis to identify con- ceptual flaws in the design of performance agreements and derive key lessons for developing nations attempting similar reforms. The article is written from the perspective of belief in the value of performance evaluation in the public sector in general and the concept of performance agreements within government in particular. Thus its objective is not to question the rationale for the US government’s performance agreements policy. Rather, it is intended to improve the quality of implementation efforts in the future and make a good policy even better. I Background The concept of a performance agreement is straightforward. It is intended to be a written agreement between the president of the US and a cabinet department secretary that describes the mutual responsibilities of the two parties to the agreement. The following paragraph from the introduction to various perfor- mance agreements 1 signed by the president with a select group of departmental secretaries succinctly describes the concept. The American people deserve a government that works better and costs less. The departments and agencies of the federal government hold vital keys to improving performance and to restoring the faith of the American people in their government. Many changes will need to take place for this broad goal to be realised. The purpose of performance agreements with senior officials is to establish clarity and consensus about the priorities for departmental management. They are intended to improve the management of the Executive Branch and are not intended to create any legally enforceable rights. From these agreements should flow the programme management priorities of the departments. These agreements represent a beginning, a basis of continuous improve- ment as we reinvent our government to meet the needs and expectations of the American people. The origins of this policy lie in the pioneering Government Performance and Results Act of 1993 (GPRA). This act requires that at least 10 federal agencies launch three-year pilot projects, beginning in fiscal 1994, to develop measures of progress. Each pilot project was to develop annual performance plans that specified measurable goals. The agencies were to produce annual reports showing how they were doing on those measures. At least five pilots would also test so-called managerial flexibility waivers – which exempt them from some administrative regulations – to help them perform even better. In exchange for greater flex- ibility, these agencies were to set higher performance targets. This is exactly the process of measured deregulation (we agree to deregulate you if you agree to be held accountable) that must be the basis of an empowered and accountable government. GPRA requires that, after learning from the pilot programmes, all federal agencies must develop five-year strategic plans linked to measurable outcomes. By 1999, every agency was expected to craft detailed annual performance plans – that is, plans that describe what they intend to achieve, not plans that detail how many pencils they will buy or how many people they will hire. The agencies will have to report their successes and failures in meeting those goals; however, GPRA allows the Office of Management and Budget (OMB) to exempt very small agencies from these requirements. II Importance of Performance Agreement Approach Performance agreements represent the culmination of a long tradition of the central role played by evaluation in public policy [Carter 1983, Wholey 1983, Picciotto and Weisner 1998, Popovich 1998, Sarji 1996 and Gore 1993]. Whether from theory or from practice, the message is the same: what gets measured gets done. In large institutions, public and private, things are counted, and whatever is counted, counts. Typically, public agencies are either not clear about their goals or are aiming at the wrong goals. The lack of clarity of goals can be attributed to the fact that most public agencies have to deal with multiple principals who have multiple (and often conflicting) objectives. This leads to fuzziness in the agencies’ perception of what is expected from them. The simple act of Performance Agreements in US Government Lessons for Developing Countries A performance agreement is a document summarising the understanding between the US president and his cabinet secretary regarding the expected results and performance indicators for the concerned department. This article argues that performance agreements are an effective instrument for promoting a culture of accountability in the government and lauds the attempt by the US administration to introduce them systematically in the highest echelons of the government. However, due to technical flaws in their design, these instruments have proved to be ineffective. This experience provides valuable lessons for developing countries in search of effective instruments of accountability. A well-designed system of performance agreements can be a great instrument for transforming development goals into reality. PRAJAPATI TRIVEDI