Central Bankers Inflate the Numbers on QE

(Bloomberg Opinion) -- Quantitative easing has been the great monetary policy experiment of the last decade.

Most central bankers believe that the large-scale asset purchases have boosted growth and lifted inflation during repeated crises. Critics respond that the impact has been trivial, and they point to potential side effects, such as promoting excessive risk-taking and fostering inequality. The available evidence appears to support the optimistic take. But to what extent can we trust these results?

A new academic paper casts fresh doubt over the case for QE. Lubos Pastor, an economist at the University of Chicago, and three colleagues have parsed about 50 studies looking at the effect of asset purchases on growth and inflation. They found that researchers working in central banks tend to find a larger impact than academics and conclude that central bankers have incentives (such as possible promotions) to be too kind toward QE.

The first thing to know about such studies on asset purchases is that measuring their economic aftermath is problematic. It is impossible to tell how the economy would have behaved in their absence and to disentangle the effects of other events that occurred at the same time. So researchers have to make different assumptions and employ a variety of models. These choices have an enormous impact on the results they produce.

The economists find that, on average, the studies they reviewed show that QE has a positive effect on inflation and growth. This finding vindicates the central banks that have embarked on the monetary policy — a list that includes the U.S. Federal Reserve, the European Central Bank and the Bank of England.

However, they also find that, on average, papers written entirely by central bankers found an impact on growth at the peak of QE that was more than 0.7 percentage points higher than the effect estimated in papers written entirely by academics. (This is a sizable difference considering the effect found on average across all studies was 1.57% at the peak.) In the case of inflation, the difference in the effect of QE at its peak between the two sets of papers was more than 1.2 percentage points. Central bankers also tended to use more positive language in summarizing their results in abstracts.

The authors then investigate the reason for these discrepancies. They suggest that career concerns may have played a role and provide some evidence that central bank researchers who found the largest impact of QE had a better chance of receiving a promotion. They also show that not all central bankers acted in the same way: Those working at Germany's Bundesbank, which has been a stern opponent of asset purchases in the euro zone, found smaller effects on output than academics.

The paper does have some important limitations. For starters, the number of studies observed is tiny, which makes it hard to reach strong conclusions. It is also possible that academics are incentivized in the opposite direction: A paper saying that QE had little to no impact on growth and inflation goes against the dominant wisdom and therefore could be easier to publish in an academic journal. Academics may also be looking at a narrower impact of monetary stimulus — for example, one that discounts the role of people's expectations. This may explain the smaller set of effects they found. 

Still, the paper highlights a fundamental problem at the heart of central bank research. From QE to the introduction of negative rates to the impact of bank regulation on lending, central banks often look at the impact of their own policy choices. They provide an exceptional environment for researchers to study these questions, but they must also ensure that their reputation is preserved and their message stays consistent. This problem is even more relevant within the euro zone, where national central banks jostle against one another as they seek to shift consensus at the ECB.

Perhaps the way to truly assess the impact of monetary policy is to read all working papers from a central bank with an open mind and a touch of skepticism. The research methods and data are there for anyone to see and double-check. If a result is not convincing, it will not pass the test of time.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Ferdinando Giugliano writes columns on European economics for Bloomberg Opinion. He is also an economics columnist for La Repubblica and was a member of the editorial board of the Financial Times.

©2020 Bloomberg L.P.