Appointment procedures and FOMC voting behavior.

• Author: Tootell, Geoffrey M.B.
• Position: Federal Open Market Committee

Introduction

An important issue for the conduct of monetary policy in the United States is whether the various appointment procedures of members of the Federal Open Market Committee help determine their voting behavior. It is usually asserted that the method by which district bank presidents are appointed leads to more independence and/or more conservative behavior. Increased independence is generally believed to reduce suboptimal short-run manipulation of monetary policy, but at the cost of reducing the central bank's accountability to the country's short-run and long-run objectives. The social desirability of any potential differences in monetary policy resulting from these appointment procedures, and the optimality of discretionary policy in general, are not addressed in this paper, however. Instead, the effect on monetary policy of the different appointment procedures currently in place within the Federal Reserve System is examined. Does the increased independence that the district bank presidents supposedly possess produce significantly different monetary policy from that of the politically appointed board governors?

Many studies have examined the effects various appointment procedures may have on voting at the FOMC, and the results have been somewhat mixed; Belden [2], Chappell, Havrilesky, and McGregor [3], Havrilesky and Gildea [6], and Puckett [8] find that bank presidents tend to vote for tighter policy than Board governors, while Tootell [10] finds no differences between the two groups. Most of the literature has concluded that in some unclearly defined way, and for no clear theoretical reason, the more independent district presidents are more "conservative" than the Board governors. In this paper, "conservative" is defined over several clear dimensions in order to test whether district presidents vote more conservatively than Board governors. Some slight variation in the voting behavior of bank presidents and board governors is found, but the causes of any dissimilarities are shown to be much different from those asserted in the earlier literature.

Alesina and Sachs [1] examine an alternative distinction among FOMC members, where partisan heritage may produce different policy. The importance of this distinction is tested in Chappell, Havrilesky, and McGregor [3], and Havrilesky [4; 5]. These previous examinations of partisan behavior at the Fed have analyzed only board governors' behavior, since their party affiliation can be easily deduced from the party of the White House that appointed them. Because only the affiliation of board governors has previously been examined, partisan association could not be used to explain any differences in voting behavior between governors and presidents. This paper, however, includes the partisan affiliations of district bank presidents; the inclusion of this variable sheds new light on the old debate over the differences between presidents and governors. The view that district presidents are independent of political considerations is shown to be naive, which helps explain why board governors and bank presidents do, in fact, act so much alike.

Unlike the previous work in this area, this paper analyzes the importance of the appointment procedure to FOMC voting when the partisan affiliation of every member of the FOMC is included in the model. In the first section, the different models of voting behavior and their implications for comparisons of bank presidents and board governors are discussed. The second section describes the data and the statistical tests used to analyze the potential difference in voting behavior between these two groups. The next section uncovers new evidence that helps explain the apparent differences between these types of FOMC members found in the previous literature. Then, the hypothesis that the selection of bank presidents is immune to national political influences is examined. It is shown that, as with most important government appointments, partisan considerations seem to play a role in these appointments, and this role explains the differences that appear to occur between the voting behavior of presidents and governors. A conclusion follows.

2. Models of FOMC Voting Behavior

   The hypothesis that bank presidents are more conservative than board governors is rooted in a belief that bank presidents are more independent. To assess the relative independence of these two groups, it is necessary to examine their exact appointment procedures. Since both bank presidents and board governors are relatively protected from political pressure once they are appointed to office, and their terms are long, the hypothesis that the two groups vote differently when serving on the FOMC rests on the assumption that they are selected from different distributions.

   Board governors are nominated by the president and these appointments are ratified by the Congress. Bank presidents are nominated by the district bank's board of directors, but the Federal Reserve Board, particularly the chairman, must approve each nomination. Thus, the chairman has significant influence over the selection of the district bank presidents. Since the Federal Reserve Board chairman is a political appointee, it can be argued that the chairman's appointees are equivalent to political appointments. In part, this paper examines whether the FOMC voting behavior of the district presidents tends to reflect the politics of the chairman who helped to select them. Upon close inspection of the appointment process, it is not clear that the selection of district bank presidents should be any more removed from political considerations than that of board governors.

   Furthermore, the effect of politics on the FOMC is more complicated than the issue of whether presidents are more conservative than governors, as the previous literature has concluded. Although the board governors, who are directly appointed by an elected official, may vote differently from the district presidents, it is not clear why greater independence of bank presidents, if it does exist, should produce more "conservative" policy. One common explanation for the increased conservativeness of bank presidents is that their more indirect political appointment procedure removes them from the political concerns faced by board governors; these political considerations supposedly compel board governors to care about the reelection prospects of the party that appointed them, while the more independent bank presidents have no such concerns. This hypothesized extra political sensitivity of board governors is assumed to manifest itself as a bias toward more expansionary policy since a stronger economy increases votes for the incumbent.

   However, studies of the influence of politics on the economy traditionally examine the timing of policy, not simply its expansionary bias. The political business cycle, as found in Nordhaus [7], predicts a cycle around elections, not a base shift in preferences. If the political business cycle were the correct behavioral model for board governors, and if the distaste for inflation and unemployment were assumed to be identical for bank presidents and board governors, then the politically appointed governors and the independent presidents would differ only over the timing of policy. On average, their votes would look alike; the governors would appear more conservative in the year after an election and more liberal in the year before the election. This hypothesis does not, however, explain why bank presidents would be more "conservative" on the whole.

   The partisan political model does offer an explanation for why some members of the FOMC would be, in the aggregate...