Whether central banks should publish individually attributed voting records from monetary policy committee meetings is among the most contested areas of debate in monetary policymaking today. Though more and more banks are making this shift in the name of individual accountability and transparent policymaking, many central banks continue to shield their voting policymakers from such public scrutiny. In this paper, I argue that while unattributed vote outcomes are potentially beneficial as a communication mechanism to coordinate market expectations, publishing complete voting records undermines both the accountability of central banks and their ability to manage market expectations. Empirically, I leverage the implementation of a freedom of information law in Brazil which required the release of individually attributed central bank voting records. The evidence shows that not only has this policy change worsened the accuracy of market expectations, but it also reduced the ability of the central bank to influence market expectations with their official communications. This result has important implications for the ongoing debates over the merits of central bank transparency and provides a cautionary tale for the application of broad political reform efforts to monetary policy institutions.