When the Labour Party came to power in the UK in 1997, one of their first decisions was to establish the Monetary Policy Committee (MPC) – an independent committee of experts whose remit is to vote monthly to determine interest rates. More recently in May 2010, the new coalition government established the Office for Budgetary Responsibility – an independent committee of experts charged with providing the economic forecast for the government’s budget, as well as an assessment of whether the government is on track to achieve its fiscal targets. In both cases, the motivation was to remove these important economic decisions from the control of politicians, who might distort them for political ends, and to take advantage of the expertise of those appointed to committee.
One design dimension along which many committees differ is the use of outside experts. The Bank of England MPC is composed of five internal members (the Governor, two Deputy Governors and two Executive Directors), and four external members who are appointed by the Chancellor and have no responsibility within the Bank beyond their role as committee members. At the same time, monetary policy committees such as the Federal Open Market Committee in the US and the Riksbank are made up of only internal members (although experts from outside the central bank are sometimes hired for the role of the internal members as they are at the Bank of England).
This leads to two questions. Should a government set up a committee for policy decisions? And, if so, should external experts be part of it? Both questions obviously depend on how appointees will behave, but while there is a significant literature concerning the behaviour of politicians taking policy decisions (see Drazen 2001 for a survey), the behaviour of experts, and particularly outside experts, is much less well understood.
An existing literature documents voting differences between internal and external members on MPC (for example, Gerlach-Kristen 2003, Bhattacharjee and Holly 2005, Spencer 2006, Besley et al. 2008). However, voting differences can arise for different reasons. For example, members may differ in terms of how much information they have about economic conditions and in terms of what we call their “economic philosophy” – their beliefs about how interest rates translate into inflation outcomes. The existing literature provides no way to disentangle these effects, but doing so is crucial. According to the Bank of England, the purpose of external appointments is to “ensure that the MPC benefits from thinking and expertise in addition to that gained inside the Bank of England.” If external members have more information, their appointment is consistent with this view and potentially raises welfare since the committee can make a more-informed decision. If instead external members have different philosophies, their appointment may or may not harm welfare depending on the government's belief about the correct philosophy.
New insight on decision-making in the MPC
In a new paper, we propose a way to separately identify a member's philosophy from his or her private information and apply it to examine differences between internal and external members. We first construct a standard voting model in which each member's vote is the result of combining information with the person’s economic philosophy. We assume that members vote over two rates since the vast majority of meetings feature at most two distinct votes. The information is actually composed of two aspects – a prior belief about the correct decision and a private signal that the member receives.
A key insight is that a member's private information (labelled σ) and philosophy (labelled θ) have different impacts on the probability she votes for the higher rate under consideration depending on her prior belief (labelled q) that the high interest rate is correct. These are shown in Figure 1 and 2. In all cases, the probability of voting for the higher rate increases with q. If θ falls, a member becomes more dovish and will vote for the higher rate less often for all prior beliefs (Figure 1). In contrast, a member with more expertise (lower σ) will vote for the higher rate less often when her prior belief favours the high rate, and more often when her prior belief favours the lower rate (Figure 2). This makes philosophy and expertise empirically distinguishable.
Figure 1. Theoretical model of voting – effect of greater expertise
Figure 2. Theoretical model of voting – effect of more dovish philosophy
Since a member's prior belief is unobservable, we construct a measure of each member's prior in each period by combining a measure of the information garnered in the meeting from other members' votes with survey data that polls leading financial institutions in the City of London in the days leading up to MPC meetings about how likely they consider different interest rate movements. We then structurally estimate our voting model for internal and external members.
We find that external members are, on average, more dovish than internal members. While there are many papers that have concluded this in the past, they were unable to isolate the effect of private information (in fact most did not even consider it), and therefore truly expose differences in economic philosophy. We also find that the external experts have an informational advantage relative to internal members in the form of more precise private information. Figure 3 shows how these differences translate into the probability of voting for the higher interest rate; external members are less likely to vote for the higher rate (when the prior is balanced (q = 0.5) the probability is 54% if internal and 30% if external). We also show that the internal members display activism in the sense of having distaste for changing interest rates and external members display activism in the sense of having distaste for leaving them the same.
Figure 3. Estimated voting behaviour
Tenure, we find, has an important effect on external experts’ behaviour; external members become less likely to vote for the high rate the more time they spend on the committee. The surprising finding is that this dynamic difference is driven by philosophical divergence rather than increasing expertise. Internal members display philosophical consistency over their tenure, while externals become more dovish with experience. The estimated expertise parameters remain constant. These results are illustrated in Figures 4 (for externals) and 5 (for internals). The solid line is the estimated probability of voting for the high rate for members serving in their first year, and the dashed line is the estimated probability for members who have served more than one year.
Figure 4. Estimated voting behaviour– effect of experience on externals
Figure 5. Estimated voting behaviour– effect of experience on internals
We find evidence that the estimated informational advantage of external members may arise, at least partly, from early information exaggeration resulting from career concerns. Exploring a quasi-natural experiment, we show that external members serving in an era in which reappointment was relatively difficult initially behave as if they have less expertise; the estimated precision of their private information during their first year is over three times larger than that of externals in their first year in other periods. We also highlight the fact that non-academic externals show more precise private information at the beginning of their tenure than at the end. Career concerns also appear to influence the degree of philosophical divergence. External members' philosophies shift less over time when reappointment is more salient and when they are non-academics.
There are a number of important consequences of these findings:
- Finding that external members are more dovish and more activist may or may not favour their appointment, depending on what the committee designer's belief is about the correct approach. For example, interest rate inertia may be viewed as beneficial to the stability of the money market (Goodfriend 1987).
- If external members indeed have more information than internal members, their appointment is a good idea. But, if our estimates of externals' private information reflects their contradicting the prior belief to gain reputation, they are not behaving efficiently and in fact harming welfare. In order to avoid this possibility, the committee designer may want to consider ruling out reappointment and appointing academic members.
- The effect of time spent on the MPC raises the issue of the correct length of tenure for external appointments. The optimal length depends on whether the committee designer wishes to incorporate different approaches to monetary policy as different philosophies only appear to arise among experiences members.
We believe that these results have implications beyond monetary policy since the use of external experts on committees is increasingly popular. One example relates to the use of external experts by the US Food and Drug Administration (FDA); the FDA has recently been criticised for bringing new drugs to market too quickly and it is argued that the use of external experts is the cause of this activism. Our results indicate that external members do indeed display behavioural differences that committee designers must consider when deciding on their appointment.
Besley, T, N Meads, and P Surico (2008), “Insiders versus Outsiders in Monetary Policymaking”, American Economic Review, 98(2):218-23.
Bhattacharjee, A, and S Holly (2005), “Inflation Targeting, Committee Decision Making and Uncertainty: The Case of the Bank of England's MPC", Discussion paper, Department of Applied Economics, University of Cambridge.
Drazen, A (2001), “The Political Business Cycle After 25 Years”, in NBER Macroeconomics Annual 2001, MIT Press.
Gerlach-Kristen, P (2003), “Insiders and Outsiders at the Bank of England”, Central Banking, XIV(1):96-102.
Goodfriend, M (1987), “Interest rate smoothing and price level trend-stationarity”, Journal of Monetary Economics, 19(3):335-348.
Hansen, S. and McMahon, M (forthcoming), “What Do Outside Experts Bring To A Committee?: Evidence From The Bank of England”.
Spencer, C (2006), “The Dissent Voting Behaviour of Bank of England MPC Members”, Department of Economics Discussion Papers 0306, Department of Economics, University of Surrey.