cnBlog – The CNB’s experience with research

Research in central banks has value added along multiple dimensions. It can aid policymakers in dealing with current pressing issues that the central bank has to tackle, but it is also important for spreading the latest economic findings both inside and outside the institution. Hence, the question is not whether to conduct research in central banks, but how to organise it, how much to invest in it and how to ensure effective transmission of research outcomes into the policy decision-making process. For such considerations, one needs to have a good idea about how other central banks conduct research. For this reason, we decided to prepare an overview of various aspects of in-house research in 50 central banks in Europe and the US and to construct a novel data set containing information about more than 20,000 research papers published by those central banks over the last two decades. It will probably come as no surprise that we identified significant differences between regions, not only across the Atlantic, but also in Europe. Even so, we managed to find some important features and trends common to all central banks.

Published in the CNB Research and Policy Notes No. 3/2020

Economic and financial research plays a pivotal role in central banks around the world. It is now the norm for central banks have their own research units, and some of them even have affiliated research institutes. The role of research is to expand the knowledge base that central banks can draw on when taking various measures. This system reflects the general belief that sound monetary, macroprudential and overall economic policy must build on cutting-edge economic thinking. Not only can high-quality research outputs foster better decision-making, they can also enhance a central bank's reputation and increase its credibility and make it easier to communicate policy actions. Moreover, research in central banks has established itself as a prominent contributor to the academic literature over time. Given the fact that central banking tasks continue to grow in number and scope, the importance of research and the demands placed on are growing too. In our paper, which we summarise in this blog article, we aim to document various aspects of the research conducted in central banks and identify significant differences and similarities. We add to the existing but still quite modest literature.

We compiled a unique data set of central bank research publications

We constructed this large dataset on the basis of publicly available information stored in the IDEAS/RePEc website, using the web-scraping technique. We obtained the paper's title, the authors' names and affiliations, the abstract, keywords, JEL codes, an indication of whether the paper has been published in a scientific journal, and the impact factor (a measure of quality). Based on the authors' first names, we determined whether they were male or female with some degree of probability. This allowed us to analyse the gender structure of research teams over time and across central banks. The main summary statistics for our resulting dataset in Table 1 show that there is significant regional heterogeneity, largely reflecting central banks research traditions, but also their size and importance. The US Federal Reserve Banks, the Eurozone central banks, the ECB and the BIS publish more articles on average and also achieve a high proportion of publications in academic journals. By contrast, smaller central banks, especially those from Central and Eastern Europe, employ more women in their research teams.

Table 1 Summary statistics for 2000-2019

	11 original euro area members	Western and Northwestern Europe	Eastern and Southeastern Europe and Baltic Countries	USA	ECB & BIS	Total
No. of papers	6,692	1,880	1,892	8,161	3,193	21,908
Papers per central bank	608	313	105	628	1,597	438
Papers per author	1.6	1.4	1.3	2.2	1.9	1.6
(0.5)	(0.3)	(0.3)	(0.4)	(0.6)	(0.5)
Share of papers published in journals	32.8	27.4	15	40.4	30.8	27.7
	(6.5)	(14.3)	(9.2)	(7.5)	(7.1)	(13.4)
Male-to-female authors ratio	2.5	3.3	2.1	3.4	3.5	2.7
	(0.8)	(1.0)	(2.3)	(0.6)	(1.0)	(1.6)

Note: Standard deviations in parentheses. Higher standard deviation means greater heterogeneity within region.

Which research topics prevail?

In terms of JEL codes (a method of classifying scholarly literature in the field of economics), the broad research areas change minimally over time. More than half of the research papers fall into two categories - E: Macroeconomics and Monetary Economics and G: Financial Economics. This is not surprising, as it reflects the two main objectives usually pursued by central banks - price stability and financial stability. However, when we look at the evolution of topics in terms of keywords and words in abstracts, we see in Figure 1 quite a sharp change following the Global Financial Crisis (GFC) of 2008-2009. Before 2008, the research focused mostly on monetary policy issues and macroeconomic topics (the blue part of the figure). After 2008, it related mainly to financial stability issues - the behaviour of the financial sector, systemic risk, crises, credit growth and prudential regulation (the red part of the figure).

Figure 1 Word cloud before and after the GFC of 2008-2009

Is there a gender gap when it comes to research in central banks?

In absolute terms, men make up about two thirds of all the authors of the central bank research papers published since 2000 (Figure 2A). All-male authorship teams continue to dominate, with a more than 50% share of all teams in 2019. However, the share of female authors has been rising continuously over time (Figure 2B). Turning to the heterogeneity between central banks, we find that smaller central banks in terms of the total number of authors engage more women. On the other hand, larger central banks in the US and the Eurozone have a substantially lower share of female authors. This gap may have a significant effect on research outcomes, because women may have a tendency to favour different research topics than men. Systematic under-representation of women in research teams might thus lead to under-investment in certain important areas and issues.

Figure 2 Distribution of authors by gender

Differences in quality as measured by impact factor

The quality of research publications, as measured by the impact factor (derived from the number of citations of the publication series in question), varies substantially in regional terms (Figure 3). Not surprisingly, the US Federal Reserve banks, the ECB and the BIS receive the highest impact factors for their in-house research publication series and for their papers published in scientific journals. This is significant, because it may indirectly indicate how many end-readers the research reaches - it is more likely that someone will read your article if you publish it in a prestigious journal or in a publication series cited heavily by other economists. Other factors contributing to a higher impact factor for a central bank include a higher share of researchers from academia involved in the research and the fact that research papers on major world economies generally have a higher probability of being published in a prestigious journal and of being cited. In this respect, though, we shouldn't lose sight of the real purpose of the research conducted in central banks: to expand the knowledge base that central banks can draw on when making policy decisions. Consequently, even a research paper published by a small central bank with relatively few citations can be highly relevant to the central bank's decision-making.

Figure 3 Regional distribution of the impact factor (a measure of a paper's quality)

Public link

Please use the above public link if you want to share this noodl on another website.