VoxEU Column Monetary Policy

The preferences of FOMC members

Classifying the preferences of members of policy committees has been a topic of intense debate and research. This column presents spatial analysis of the preferences of the Federal Open Market Committee (FOMC) members using transcripts from meetings. The results indicate that a political appointment channel was not active or effective, and there is little effect of career experiences. The overall lack of systemic preference among FOMC members is a reassuring with regard to the institutional design of the FOMC.

Monetary policy committees as doves and hawks

Central bank watching is commonly done in the financial press. News outlets such as Bloomberg, the Financial Times, the Guardian, Business Insider or The Economist regularly engage in reporting about the preferences represented in various monetary policy committees. Such news articles often classify committee members on a dove-hawk scale. These terms are rarely explicitly defined. Loosely speaking, a dove is a central banker who given the macroeconomic conditions is more inclined to ease compared to his peers, whereas a hawk is less inclined to do so.

The classification in news outlets is often based on official voting records – if available.  In case of the Federal Open Market Committee (FOMC) – the monetary policy committee governing the Federal Reserve – this is problematic. It has been well documented in the academic literature that central bankers at the FOMC show some restraint in dissenting (see Blinder et al. 2001).

While this restraint may have lessened in recent years, this still is an issue of concern. One way to circumvent this problem is to analyse the transcripts of FOMC meetings (Meade 2005). The transcripts are released with a lag of five years and provide interesting insight in the thought process of the committee members. Importantly, during a meeting there typically is a round where each committee member states his or her preferred policy rate. After this round there is further discussion before turning to an official vote on a (consensus) proposal formulated by the Chairman. The policy stated by each committee member provides an ‘unfiltered’, thus better measurement of the preferred policy of each individual member.

While the preferences provide better data, another concern is how one should use these data to construct a score on a dove-hawk scale. Counting the number of times a committee member favoured lowering the interest rate is unsatisfactory since that score would be influenced by the macroeconomic circumstances during the tenure of that committee member. And how can we compare the scores of committee members who were appointed in different periods? Furthermore, there might be interest in some statistical analysis on the preferences. Are there patterns in the preferences of central bankers?

The latter question is particularly relevant since a lot has been written on systematic differences between specific groups of central bankers. The Federal Open Market Committee has twelve members with voting rights. Seven are members of the Federal Reserve Board, whereas five of them are Federal Reserve Bank Presidents. One question is whether there is a difference between the preferences of board members and bank presidents. Earlier research suggested that there is in fact a difference and, moreover, that the preferences of FOMC members might be tied to the economic circumstances in their home region (see Jung and Latsos 2014 and Chappell et al. 2008). Other research considered patterns due to presidential appointment (see Chappell et al. 1993) or career concerns (see Adolph 2013).

The focus of the academic literature on systematic differences between groups of committee members is motivated by the concern that these might introduce harmful biases in the decision making process. For example, if the incumbent president would be able to influence the decisions by the FOMC by appointing board members, then this could give rise to political business cycles. Another example is related to career experiences and career prospects. If it would be the case that FOMC members are more concerned with career prospects, then this could have a negative effect on the workings of the FOMC.

Retrieving unobserved preferences among committee members

One way of retrieving the unobserved preferences is to use so-called spatial voting models. These models are nowadays commonly used to scale politicians on an ideological scale or the retrieve the ideologies of justices at the Supreme Court.

There have been a few efforts to apply such spatial voting models to monetary policy committees. The first attempt we are aware of is documented in a book by Kelly Chang (2003). However, observers found the results of that analysis, focused on the FOMC, not entirely convincing (Morris 2004). Moreover, the statistical analysis might be considered a bit crude given current standards in the analysis of spatial voting models. Recently there have been renewed attempts with more sophisticated statistical methods. Hix et al. (2010)  and Eijffinger et al. (2013a) analysed the Bank of England and Eijffinger et al. (2013b)  is work in progress on various continental monetary policy committees.

Our recent research (Eijffinger et al. 2015) is, besides the analysis provided by Chang, the only spatial analysis of the FOMC we are aware off. In this column, we argue how one could estimate the preferences of FOMC members in the period 1989-2007 in a coherent framework from preference statements obtained from transcripts. We are able to provide measures for the preferences of FOMC members and rank them on a dove-hawk scale. Subsequently, we explore a few long standing topics in the literature:

We look at the influence of career experience and do not find strong evidence for a systematic influence. This stands in contrast to the results presented in a recent book by Adolph (2013), although the sample analysed by the latter differs substantially from the sample we use.
We find that the preferences of board members appointed by different presidents do not show any pronounced differences. This finding casts doubt on a political appointment channel.
We do find that, on average, board members exhibit more dovish preferences than Federal Reserve Bank presidents.

Furthermore, we calculate the evolution of the median preference as well as the evolution of the median preference among the group of Federal Reserve Bank presidents, and the evolution of the median preference of board members. In Figure 1, the black line indicates the median preference. The top line (pink) shows the most hawkish preference and the bottom line (blue) shows the most dovish preference. The shaded areas indicate the associated uncertainty. This figure tells us that the median preference has proved to be very stable, especially when compared with the most hawkish and most dovish preferences at the FOMC. This is interpreted as further evidence against a presidential appointment channel.

Figure 1. Overview of the evolution of the median, maximum, and minimum ideal point at the FOMC

Figure 2 presents the evolution of the median preference of Federal Reserve Bank presidents (on the top in pink) and board members (bottom in blue). This figure tells us that there is variation in the median preference of Federal Reserve Bank presidents and board members over time. While it is true that, on average, Board members are more dovish, this does not imply that this always has been the case. In the early 90s, we even briefly observed the opposite.

Figure 2. Overview of the evolution of the median ideal point among Board Governors and Bank Presidents


Our research (Eijffinger et al. 2015) shows how a spatial voting model can be used to analyse preferences of FOMC members – even if the official voting records are potentially unobtrusive.

The preliminary results of the paper show that there may be less evidence in favour of systematic patterns in the preferences of FOMC members than previously thought. The analysis suggests that a political appointment channel was not active or effective and also finds little scope for the effect of career experiences. The results are comforting for those concerned about the institutional design of the FOMC.


Adolph, C (2013), Bankers, Bureaucrats, and Central Bank Politics: The myth of neutrality, Cambridge University Press.

Blinder, A, C Goodhart, P Hildebrand, D Lipton, and C Wyplosz (2001), How do central banks talk? Geneva reports on the world economy 3, London: Centre for Economic Policy Research.

Chappell, W H Jr, T M Havrilesky, R R McGregor (1993), “Partisan Monetary Policies: presidential influence through the power of appointment”, The Quarterly Journal of Economics Vol. 108, No 1.

Chappell, W H Jr, R R McGregor, and T A Vermilyea (2008), “Regional Economic Conditions and Monetary Policy”, European Journal of Political Economy Vol 24.

Chang, K (2003), Appointing Central Bankers: the politics of monetary policy in the United States, Cambridge University Press.

Eijffinger, S, R Mahieu, and L Raes (2013a), “Inferring hawks and doves from voting records”, CEPR Discussion paper 9418.

Eijffinger, S, R Mahieu, and L Raes (2013b), “Estimating the preferences of central bankers: an analysis of four voting records”, CEPR Discussion paper 9602.

Eijffinger, S, R Mahieu, and L Raes (2015), “Hawks and Doves at the FOMC”, CEPR Discussion Paper 10442.

Hix, S, B Hoyland, and N Vivyan (2010), “From to Doves to Hawks: A spatial analysis of voting in the monetary policy committee of the Bank of England”, European Journal of Political Research.

Jung, A, and S Latsos (2014), “Do Federal Reserve Bank Presidents Have a Regional Bias?”  ECB working paper No 1731.

Meade, E (2005), “The FOMC: Preferences, voting, consensus”, Federal Reserve Bank of St. Louis Review.

Morris, I L(2004), “Review of Kelly Chang, Appointing Central Bankers”, Perspectives on Politics.

1,575 Reads