The literature on the inflation bias that can arise under discretionary monetary policy has emphasized the value of establishing a reputation for following credible policies designed to sustain low inflation. The ability to achieve such a reputation can depend on the institutional structure that governs the conduct of monetary policy. Because of this, a great deal of interest has been generated by the Reserve Bank of New Zealand Act of 1989. This Act was designed to give New Zealand's Reserve Bank greater independence from political influence, establish a single objective for monetary policy (price stability), and ensure accountability on the part of the Reserve Bank for the achievement of its policy objective. The basic structure of the Reserve Bank Act has been cited as a useful model for other countries (Willett 1994), but little formal analysis of the new law has yet appeared in the literature. Dawe (1993) provides a discussion of the Reserve Bank Act, while Fischer (1993) interprets New Zealand's monetary policy under the Act as an inflation-targeting policy and provides an excellent discussion of such policies. Buckle and Stemp (1991) formally analyze policy under the Act as equivalent to a price-level-targeting policy. These approaches capture some aspects of New Zealand's new policy process, but they neglect other aspects and thus may not provide a complete evaluation of the unique institutional changes contained in the Reserve Bank Act.

The Act specifies a complex set of relationships between the Reserve Bank and the government. These cover such issues as the appointment and reappointment procedures for the Governor of the Reserve Bank, the establishment of short-run inflation targets, and the conditions under which these targets may be modified. The Act essentially establishes a contract between the central bank and the government; the nature of this contract determines the incentives facing the Reserve Bank and, as a result, affects the conduct of monetary policy. Viewing central bank institutions as contracts leads one naturally to ask what sorts of incentives such contracts should establish for the central bank, and to ask whether the contract implicit in New Zealand's new Act is in some sense optimal. The purpose of this paper is to evaluate the Act using the principal-agent approach to central bank design developed in Walsh (1993, 1995) and Persson and Tabellini (1993).

An overview of the Reserve Bank Act is provided in section t. The key aspects of the legislation likely to affect the conduct of monetary policy are discussed, and it is argued that the Act establishes a dismissal rule for the Reserve Bank Governor. Section 2 then derives an optimal dismissal rule which is used in section 3 to evaluate the Reserve Bank Act. Conclusions are summarized in section 4.

1. THE RESERVE BANK ACT

The Reserve Bank Act took effect on February 1, 1990, and was part of a broader governmental reform program designed to increase accountability in the public sector. The Act's purpose was to establish a clear and achievable policy objective for which the Reserve Bank could be held accountable. New Zealand's central banking reforms were motivated by its poor inflation performance (see Fischer 1993). As Table 1 shows, relative to the rest of the OECD, inflation was higher in New Zealand prior to the first oil price shock in 1973, New Zealand's inflation rate rose more during the 1970s, and it fell less after the general shift by OECD countries to disinflationary policies in the first half of the 1980s. Except for the temporary impact of wage and price controls in 1982-1984, inflation fluctuated around the 12-15 percent level from the mid-1970s until 1986.(1) While piecemeal reform in the mids 1980 gave the Reserve Bank the tools to achieve price stability as early as 1985 (Walsh 1988), previous unsuccessful attempts to reduce inflation meant the Reserve Bank lacked the credibility necessary to minimize the output costs of disinflation and to maintain low inflation. …