No relief for ECB’s status quo headache from rotation

Mika Widgrén

25 February 2008



The most recent ECB Governing Council (GC) meeting on 7 February left the key ECB interest rates unchanged. The meeting took place in the aftermath of the Federal Open Market Committee’s (FOMC) hyperactive moves to reduce the target for the federal funds rate 75 basis points to 3.5% at an unscheduled FOMC meeting on 21 January and then another 50 basis points to 3.0% on 30 January.1

In the follow-up press conference to the Governing Council’s decision, ECB President Jean-Claude Trichet described the newly-made decision as a mixture resulting from careful consideration of existing upside risks to price stability confirmed by monetary analysis and downside risks surrounding the outlook for economic activity. Despite the dramatic decrease in US rates, the unanimously agreed ECB decision emphasised the GC’s commitment to preventing second round effects of upside risks to price stability and aim to anchor inflation expectations in line with price stability.

Is this stark contrast due to policy judgements or it is institutional difference in decision-making procedures?

How do European and American monetary policy differ?

The roots of the Fed’s monetary policy decisions in the latter half of January lie in the FOMC’s focus on high uncertainty and currently exceptional high probability of extreme outcomes. Consequently, monetary policy calls for policymakers to be anticipatory in responding to macroeconomic implications of incoming information from financial markets and, moreover, in reducing the likelihood of vicious circle stemming from financial markets’ reaction to negative macroeconomic information that further increase the downside risks to economic activity.2 The key elements of monetary policy are flexibility, decisiveness and low inertia. Fed’s policy actions during the latter half of January, indeed, demonstrate these, at least to one direction.

What would have been the ECB’s reaction under exactly the same circumstances as Fed is facing? The question is relevant since it is most likely that the economic effects of financial disruption in the US will spill over to Europe and the global economy. Without even attempting to assess the economic effects of the Fed’s decisions in January, a comparison of the ECB decision-making system to that of the Fed makes one conclusion easy to draw: due to its institutional design, the ECB is likely to suffer from inflexibility, inertia and indecisiveness.3

The Governing Council, the key decision-making body of the ECB, consists of the Executive Board (EB) of six members who do not represent interests or views of any particular EMU country or group of countries and the Governors of EMU countries’ national central banks (NCBs). In today’s EMU15 that means the Governing Council has 21 voting members. Formally, decisions require a simple majority, though in practice the GC arrives at consensus decisions and does not vote. Nonetheless, NCBs form a majority in the Governing Council, and their local information advantage, preferences, and views influence what policy finally wins the consensus approval. The NCBs’ equal representation and the simple majority requirement constrain the set of feasible decisions that can be “shadow-voted” and approved by common consent.

The US Federal Open Market Committee consists of twelve members - the seven members of the Board of Governors of the Federal Reserve System, the president of the Federal Reserve Bank of New York, and four of the remaining eleven Reserve Bank presidents, who serve as voting members in one-year terms on rotating basis.4 If united, the members of the Board of Governors always have a majority.5

In the Governing Council, the one-national-central-bank-one-vote principle was intended to ensure that governors of national central banks would participate as independent actors, not as national stake-holders. Nothing guarantees that, though. Moreover, expanding EMU membership increases the voting share of the Central Bank Governors and makes the consequent numbers or inefficiency problem more severe. Indeed, in earlier studies (see footnote 3), co-authors and I argued that in an expanding Euroland it would become highly unlikely that Governing Council could pass optimal policies correspond to Euroland’s aggregated preferences. Moreover, there would be substantial risk of sticking to status quo when facing asymmetric shocks. In sum, Governing Council decisions might be too conservative and biased towards the status quo.

The soon materialising Governing Council reform…

In 2002, the foreseeable EMU enlargement caused EU leaders to reform their voting rules to avoid ECB decision-making paralysis. In 2004, they adopted a rotating arrangement in which the number of voting Central Bank Governors is fixed at 15 while the membership of Euroland expands to 16 and beyond.

In the rotation model, EMU countries will be divided into three groups of 1) five biggest countries, 2) 14 medium sized countries and 3) eight small countries, with each group respectively having four, eight and three voting national Central Bank Governors (NCBers) at a time. The model fixes the distribution of representation between the Executive Board and NCBers after the entry of the 16th member, which, in terms of power, benefits the EB in the long-run as EMU membership expands,6 but the reform does not affect the primacy of NCBers in the system.

The desired division of labour of the EB members and NCBers would give the former the role of contributing their knowledge about the state of financial markets and the latter the role of contributing local knowledge about the real economy. The rotation model, which will maintain the NCBers’ majority and unequal representation and introduce inequality of member states’ representation at any given phase of the rotation,7 might then undermine the idea that the interests of the entire Euro-zone, not local concerns, should matter.

… does not offer a proper solution either

In sum, the rotation model does not solve ECB’s status quo bias.

To make the argument more properly, consider a simple example. Suppose that the Governing Council has three alternatives: (1) to increase or (2) decrease interest rates or (3) keep them as they are at status quo. Call these ‘+’, ‘-‘ and ‘0’ respectively. Suppose that the Executive Board acts in unison representing Euroland’s general interest (having 6 votes), while NCBers have preferences based on their economic situation at home. And suppose that there is an even mix of higher than average inflation countries and lower than average growth countries.8 In the illustrative example, suppose that the Executive Board and NCBers rank the alternatives according to the table 1.

Table 1 An example of the EB’s and NCBs’ preference orderings in EMU15 or under agreed rotation system
The number of NCBs having the preference ordering
+ - + - 0 0
- + 0 0 + -
0 0 - + - +
0    0   4+6 3   4  

Now, there is a majority of eight NCBers that prefer a decrease to an increase. It is thus natural to assume that the Euroland average would be ‘decrease’, which is also the first preference of the Executive Board in this example, In pair-wise comparison between decrease and increase, the former would get a clear majority. However, in a pair-wise comparison of ‘decrease’ and the status quo, ‘decrease’ is defeated by the status quo, which would get 11 NCB votes and defeat the 4 NCB votes plus 6 EB votes in favour of ‘decrease’. In this case, the status quo also beats ‘increase’ in a pair-wise comparison but only seven of 15 NCBs support that as their first preference.

The key point of the example is that in this hypothetical case (and there are many more), the Governing Council is not able to agree on socially optimal policy.

As a follow-up, consider the EMU of 27 countries and suppose that a majority of 12 non-voting NCBs prefer ‘decrease’ over ‘increase’. That would mean that in pair-wise comparison ‘decrease’ would defeat ‘increase’ among the EMU countries by an even clearer margin but it would be defeated by the status quo in the GC. This would make the social losses even more severe. More examples of such preference orderings can be easily constructed. In sum, the Governing Council suffers from a potentially great status quo bias and is unable to guarantee socially optimal policies based on Eurozone-wide interests.

A solution?

A solution for all this is, following the logic of 1972 Nobel Laureate Kenneth Arrow’s impossibility theorem, to delegate monetary policy decisions to a benevolent and accountable dictator. That makes me wonder if the Executive Board could take the job.

1 Frederic S. Mishkin was absent from the 21 January meeting. William Poole voted against the action of 21 January, and Richard W. Fisher voted against the action of 30 January.
2 See Governor Fredric S. Mishkin’s speech Monetary policy flexibility, risk management and financial disruptions at Federal Reserve Bank of New York January 11, or The Federal Reserve's Tools for Responding to Financial Disruptions at the Tuck Global Capital Markets Conference, Tuck School of Business, Dartmouth College, Hanover, New Hampshire February 15, 2008
3 See e.g. Baldwin, Berglöf, Giavazzi and Widgrén (2001): Nice Try – Should the Treaty of Nice be Ratified, Monitoring European Integration 11, CEPR, London, Baldwin, Berglöf, Giavazzi and Widgrén (2001): Preparing the ECB for Enlargement, CEPR Policy Paper No. 6 or Baldwin, R., Berglöf, E., Giavazzi, F. & Widgrén, M. (2001): Eastern enlargement and ECB reform, Swedish Economic Policy Review Vol. 8, 17-49.
4 The rotating seats are filled from the following groups of Banks, one Bank president from each group: Boston, Philadelphia, and Richmond; Cleveland and Chicago; Atlanta, St. Louis, and Dallas; and Minneapolis, Kansas City, and San Francisco. Non-voting Reserve Bank presidents attend the meetings of the Committee and participate in the discussions.
5 In Canada, Sweden and New Zealand, monetary policy committees consist of executive members only; in the UK and Australia committee includes independent experts who form a minority in the former and majority in the latter.
6 For the evaluation of the distribution of power among EB and NCB governors see e.g. Anskar and Belke (2006): The Allocation of Power in the Enlarged ECB Governiong Counciul: An Assessment of the ECB Rotation Model, Journal of Common Market Studies 44, 865-897.
7 See e.g. Gros (2003): Reforming the Composition of the ECB Council in View of Enlargement: an Opportunity Missed! Intereconomics: Review of European Economic Policy, 38, 124-129 or Baldwin, Berglöf, Giavazzi and Widgrén (2001): Preparing the ECB for Enlargement, CEPR Policy Paper No. 6 for general critique against the agreed rotation model.
8 This division corresponds roughly with the current situation. Note that higher than average growth and inflation countries are small with an exception of Spain. Note, however, that the main point of the example is to demonstrate the vulnerability of the rotation system in terms of GC’s capacity to act.





Topics:  Monetary policy

Tags:  ECB, Fed, Governing Council

Professor of Economics at Turku School of Economics, Finland and CEPR Research Fellow