The Italian Job – How Passports Influence Monetary Policy Decisions

German Logo of the ECB.

Logo of the ECB (Image via Wikipedia)

The nationality of the ECB’s managerial staff matters for the monetary policy of the central bank, two German and Austrian economists show in a new paper. The findings are so politically charged that the detailed results of the study are kept secret.

Today is the first working day of Mario Draghi as the new president of the European Central Bank. Apart from Draghi, two more Italians have a seat in the ECB council: Lorenzo Bini Smaghi and Ignazio Visco, Draghis successor at the helm of the Bank of Italy.

Currently, no other Euro zone country has as much influence in the ECB council as Italy has. Ironically, the country is also at the centre of the Euro crisis.

Does the nationality of the ECB’s policy maker matter?

„Not at all“, would be the official answer of the central bank to this question. Formally, members of the governing council do not represent their home country. According to the rules of the ECB, they have to base their decisions on the needs of the whole currency area. In a metaphorical sence, they hand in their passports when they join the central bank.

Is this a realistic description of what’s going on in the central bank?

Harald Badinger (Vienna University of Economics and Business) and Volker Nitsch (University of Darmstadt) addressed this question a remarkable paper. According to their results, the official view is a delusion. In reality, the nationality of the policy makers does influence the monetary policy of the ECB.

The results of the paper entitled “National Representation in Multinational Institutions: The Case of the European Central Bank” are so explosive that Badinger and Nitsch refrain from publishing the most juicy parts of their work.

Badinger and Nitsch meticulously – and apparently to the consternation of the ECB – collected data about the nationality of all top-level employees of the central bank. They were able to get this information for 190 of the 210 senior managers who have been hired by the central bank since 1999.

The economists reveal that high-ranking ECB officials have a tendency to hire their fellow countrymen. Additionally, the nationalities of the policy makers do influence the monetary policy decisions in a meaningful way.

As Badinger and Nitsch assert:

“Looking at the European Central Bank, we show that nationality is indeed relevant for both hiring and decision-making.  (…) There is evidence for the existence of national networks between adjacent management layers.

Finally, monetary policy decisions seem to be linked to national representation in the core business areas of the ECB.”

The analysis is based on a common rule of thumb for monetary policy – the so-called Taylor rule. As the economists explain in their paper, “Taylor rules specify the nominal interest rate as function of inflation and the output gap”.

Badinger and Nitsch compare two different specifications of the Taylor rule. The first one is based on the actual economic weights of the different Euro zone countries. The second version uses the national representation weights in the management of the ECB instead. According to the paper, the latter specification explains the actual interest rate decisions of the central bank better than the first one.

As the economists conclude:

“there is convincing evidence that the empirical fit of the Taylor rule based on national representation weights in the management of the ECB clearly exceeds the fit of the rule that is observed for euro area aggregates based on economic weights.”

This is where the paper really gets interesting. Which member states are too heavily represented in the ECB and which are marginalised? Which countries benefited from the biased monetary decisions and which were?

Unfortunately, the answers to those questions are not given in the paper. The economists do not publish any country specific results at all. In a lengthy telephone call with Volker Nitsch, I tried to convince him to reveal country specific details. However, I wasn’t able to make him change his mind.

Both researcher are afraid that the media could take the results the wrong way and might initiate a nationalist campaign against the ECB.

Between the lines I also gleaned that the authors do not want to offend the ECB. Apparently, from the central bank’s perspective, even the current version of the paper is very hard to digest. Allegedly, a senior ECB representative was completely startled when he listened to a presentation of the paper at a conference.

However, shooting the messenger seldom turns out to be a successful strategy.

Update: Bernd Hayo (University of Marburg) and Pierre-Guillaume Méon have recently published a  related paper entitled “Behind closed doors: Revealing the ECB’s Decision Rule”. Here’s the abstract:

“This paper aims at discovering the decision rule the Governing Council of the ECB uses to set interest rates. We construct a Taylor rule for each member of the council and for the euro area as a whole, and aggregate the interest rates they produce using several classes of decision-making mechanisms: chairman dominance, bargaining, consensus, voting, and voting with a chairman. We test alternative scenarios in which individual members of the council pursue either a national or a federal objective. We then compare the interest-rate path predicted by each scenario with the observed euro area’s interest rate. We find that scenarios in which all members of the Governing Council are assumed to pursue Euro-area-wide objectives are dominated by scenarios in which decisions are made collectively by a council consisting of members pursuing national objectives. The best-performing scenario is the one in which individual members of the Governing Council follow national objectives, bargain over the interest rate, and their weights are based on their country’s share of the zone’s GDP.”

11 Comments

Filed under Financial Crisis, Monetary Policy

11 Responses to The Italian Job – How Passports Influence Monetary Policy Decisions

  1. J

    I am not sure the paper tells us anything with regard to the conduct of monetary policy. Since when are we interested in R^2 which tell us nothing about the conduct of monetary policy (except for potentially the size of accidental monetary policy shocks). We should be interested in the reaction coefficients of Taylor rule. And while the fit of the regression with representation weighting may be better, are the reaction coefficients between the different specifications significantly different from each other? From the wide standard errors I would guess they are not!
    Second, there is no indication of the difference in R^2 being statistically significant between the specificiations. Add to that the inflated type 1 error due to the multiple comparisons which requires an alpha- correction. If I run several different regression with different regressors, chances are that one will provide a better fit. Second, the authors study one particular structural form of the Taylor rule and find differences in fit with differently weighted regressors. But we don’t know if the underlying structural rule, i.e. the assumed forward-looking Taylor rule really describes ECB behavior optimally. What if the ECB actually used a Taylor rule with output growth instead of output gap? For this rule, probably the unweighted scheme works better. But does this tell us anything? Finally, the paper does not discuss the new coefficient estimates which imply a stronger reaction to both output gaps and inflation. Unless you specify your underlying model of the economy and particularly the distortions monetary policy tries to counteract, you cannot say who benefits from this stronger reaction function.

    • Thanks for you’re comment. I’ll ask the authors of the paper to comment on your points.

    • Jasper

      In addition to the comments of J:
      The authors find that the ECB capital share and the staff number at the national central bank are a major determinant of the nationality of ECB top level officials. So if the authors compare Taylor rule where the individual country contributions to the output gap and the deviation from the inflation target are GDP weighted, with a Taylor rule where the individual country contributions are weighted according to the nationality of ECB top level officials, you are maybe just testing if the difference between GDP share and capital ratio share or the staff number at the national central bank improves the explanatory power of the model.

      The capital share is an equally weighted average of the GDP share and the country’s population share**. My guess would be that the staff size of the national central bank is a function of bank sector size and the country’s population size. Which means you could be testing whether the country’s population share in deviation of its GDP share is significant. If this is the case, it benefits low GDP per capita countries. From an aggregate welfare perspective, maybe that is not so bad. In my view, some additional robustness checks might be appropriate.

      ** http://www.ecb.int/ecb/orga/capital/html/index.en.html

  2. David

    Would be an interesting thing to know why these results should be secret? There is many research on the question if ECB decision makers nationality influences monetary policy. This if often discussed in the context of national inflation differentials in the euro zone. As defined in the goals of monetary policy, the decision maker should focus on the average inflation rate of the whole euro zone. Contrary to this there are some papers saying that the focus is possibly closer to the inflation rate of the decision makers home country. So, what’s the new thing?

  3. Thanks for bringing this interesting and revealing research to our attention. As you say, with the number of Italians presently on the ECB council and the market focus on Italy’s debt, this issue is especially relevant now.

    I was puzzled by the logic of Axel Weber and Jurgen Stark resigning from the ECB council (reportedly) over bond buying – to me it seemed to make more sense to stay and fight your corner, and this research suggests that that is what most countries’ ECB staff do.

  4. Pingback: The Eurozone Crisis’ End Date Reconsidered « Publius Abroad

  5. What if someone has two passports?

  6. As the saying goes, “posession is nine-tenths of the law”. Draghi got the job and Draghi eased at his first opportunity. I imagine that Weber would not have approved, but if you leave the job to someone else, you have to accept what they do.

  7. Guillem Aranda

    I bet the Taylor fit between the German figures and ECB rates beats both from the paper.

  8. icantbelieveitsnotbutter

    “The results of the paper … are so explosive that Badinger and Nitsch refrain from publishing the most juicy parts of their work ”

    They also refrain from writing down a model.

    “Both researcher are afraid that the media could take the results the wrong way and might initiate a nationalist campaign against the ECB.”

    So what’s the point of the paper? And how should we take the results? The problem is that, absent a model, there is no clear economic interpretation of the estimated coefficients.

    To me this is “hit-and-run” economics. Badinger and Nitsch present some “explosive” results without making explicit their assumptions. And then they are surprised if the readers (or the media) try to make their own assumptions in order to interpret the results.

  9. After his comments in the FT over the weekend, one hopes that Jens Weidmann is not the resigning type.