The use of independent committees for the setting of interest rates, such as the Monetary Policy Committee (MPC) at the Bank of England, is quickly becoming the norm in developed economies. In this paper we examine the issue of appointing external members (members who are outside the staff of the central bank) to these committees. We construct a model of MPC voting behaviour, and show that members who begin voting for similar interest rates should not systematically diverge from each other at any future point. However, econometric results in fact show that external members initially vote in line with internal members, but after a year, begin voting for substantially lower interest rates. The robustness of this effect to including member fixed effects provides strong evidence that externals behave differently from internals because of institutional differences between the groups, and not some unobserved heterogeneity. We then examine whether career concerns can explain these findings, and conclude that they cannot.