E. Principle 5: The macro-supervisor should be less independent than central banks are now in their monetary policy responsibilities

Over the past two decades, the establishment around the world of independent regulatory agencies has ignited a debate on how to reconcile technocratic independent institutions with the demands of democratic legitimacy. How can giving freedom (i.e. independence) to unelected officials be reconciled with a society remaining democratic? The answer is: through accountability.

Performance accountability requires that there are objectives or standards (criteria ofassessment) according to which an action or decision might be assessed. The extent to which a particular type of accountability may be preferred can be a function of the type of supervision that is at stake.[1] In the case of supervision, input or process monitoring should be preferred, because as discussed above, performance or outputs on the supervisory activity are hard to measure. The fact that inputs, rather than output monitoring should be chosen also suggests that providing a monetary authority (with a clear performance objective) with independence is not the same as providing independence to a supervisor: if delegation and output measurement cannot be used, then independence must be more restricted with regard to financial supervision than with regard to monetary policy.

Transparency is a complement of accountability—information needs to be observed for the agent to be made accountable.[2] However, the provision of information is hardly ever a neutral account of what happened or of what is happening, as the agent is likely to provide it in a self-serving way. Essentially, as

Prat[3] has argued, if the action is transparently observable, the risk is that agents will behave in a conformist way by doing what is expected of them.[4] A related theoretical argument has been made by Amato, Morris and Shin[5] who argue that too much transparency can actually reduce policy effectiveness. If the central bank signal is noisy relative to private signals, they show that attaching too much weight to the noisier signal may distort the quality of the market’s treatment of information. While this argument is of unclear validity for monetary policy where most relevant information is public,[6] it may be very relevant to supervision, where private information is important.

A second downside of transparency concerns panics. While the macroeconomics literature[7] argues that transparency in the decision making of central banks is useful, the need for covert assistance in the case of lender-of-last-resort operations (which is recognized in the new Banking Act of2009 in the UK) is crucial to a crisis, since the belief in a panic is self-fulfilling. These considerations put transparency for supervisory decisions in a different category from transparency for monetary policy decisions, where the arguments are overwhelmingly in favour of disclosure.

Thus our review of the organizational economics and macro-literature on independence and accountability leads us to three conclusions. First, that the difficulty in making supervisory performance measurable means independence of supervisors should be limited with regard to certain supervisory decisions. Second, that input (or process) monitoring rather than output monitoring should be preferred. In other words, accountability cannot simply rely on whether or not crises are taking place; instead, mechanisms must be put in place that ensure that supervisors have to explain the actual decisions and the process leading to them. Third, transparency, itself a complement of accountability, must be minimized with regard to certain crisis-sensitive decisions in a supervisory agency to avoid career-based decisions by experts, informational distortions by the market and bank panics.

  • [1] See Canice Prendergast, ‘The Tenuous Trade-Off between Risk and Incentives’, 110(5) JournalofPolitical Economy 1071 (2002).
  • [2] See also Kaufmann and Weber in this volume at Chapter 13.
  • [3] Andrea Prat, ‘The Wrong Kind of Transparency’, 95(3) The American Economic Review 862(2005).
  • [4] This is arguably the case also for monetary policy tasks: if minutes of the meetings werepublished, then board members would be more likely to take the actions that are expected of them,such as acting in their national interest rather than the common interest.
  • [5] Jeffery D. Amato, Stephen Morris and Hyun Song Shin, ‘Communication and MonetaryPolicy’, 18(4) Oxford Review of Economic Policy 495 (2002).
  • [6] Alan Blinder and Charles Wyplosz, ‘Central Bank Talk: Committee Structure and Communication Policy’, Mimeo, Prepared for the session ‘Central Bank Communication’ at the ASSAMeeting in Philadelphia, 9 January 2005, (visited on 14 February).
  • [7] Alan S. Blinder, The Quiet Revolution: Central Banking Goes Modern (New Haven: Yale University Press, 2004).
< Prev   CONTENTS   Source   Next >