The (home) bias of European central bankers: new evidence based on speeches

Abstract

Speeches are an important vehicle for central bankers to convey individual views on the preferred policy stance. In this paper, we employ an automated text linguistic approach to create an indicator that measures the tone of the 1,618 speeches delivered by members of the Governing Council (GC) during the period 1999M1−2014M4. We then relate this variable to euro-area and national macroeconomic forecasts. Our key findings are as follows. First, inflation and growth expectations have a positive and significant impact on the hawkishness of a speech. Second, the voiced preferences of national central bankers largely coincide with the level of independence their banks had at the time of the Maastricht Treaty. Third, country-specific macroeconomic conditions matter for speeches delivered inside the central banker's home country but not for those made abroad. Fourth, differences in central banker preferences are the key source of variation in their speeches before the financial crisis, whereas divergent national economic conditions are the main factor in the second part of the sample.

Adding rooms onto a house we love: Central banking after the global financial crisis

“…However, political psychologists with access to political leaders’ speeches and private notes have found surprising congruence between public and private speech (Renshon ; Dyson and Raleigh ). Researchers have relied upon public speeches to make ‘cognitive maps’ of individual central bankers’ beliefs and their evolution over time (van Esch and de Jong ; Bennani ), and central bankers themselves claim such congruence (Bernanke ; Bennani and Neuenkirch ; Rajan ). Moreover, as our primary interest is in broad ideational trends in the central banking community, it seems reasonable to turn to public speech acts to identify these trends.…”

“…That is, we draw on the Bank for International Settlements (BIS) database of 13,586 speeches given by central bankers worldwide from 1997 to 2017 (Figure ) to conduct both quantitative and qualitative text analysis . It is a rich trove of data, one that has increasingly been mined by scholars using automated text analytic techniques (Fontan et al ; Bennani and Neuenkirch ; van Esch and de Jong ). This line of work dovetails with a broader interest by academics and central bankers in using quantitative content analysis to study central bank‐related themes (Schonhardt‐Bailey ; Bholat et al ; Golub et al ; Moschella and Pinto ).…”

Adding rooms onto a house we love: Central banking after the Global Financial Crisis

“…However, political psychologists with access to political leaders' speeches and private notes have found surprising congruence between public and private speech (Dyson & Raleigh, 2014;Renshon, 2009). Researchers have relied upon public speeches to make "cognitive maps" of individual central bankers' beliefs and their evolution over time (Bennani, 2015;van Esch & de Jong, 2013), and central bankers themselves claim such congruence (Bennani & Neuenkirch, 2017;Bernanke, 2015;Rajan, 2017). Moreover, as our primary interest is in broad ideational trends in the central banking community, it seems reasonable to turn to public speech acts to identify these trends.…”

“…That is, we draw on the Bank for International Settlements (BIS) database of 13,586 speeches given by central bankers worldwide from 1997-2017 ( Figure 1) to conduct both quantitative and qualitative text analysis. 2 It is a rich trove of data, one that has increasingly been mined by scholars using automated text analytic techniques (Bennani & Neuenkirch, 2017;Fontan, Claveau, & Dietsch, 2016;van Esch & de Jong, 2017). This line of work dovetails with a broader interest by academics and central bankers in using quantitative content analysis to study central bank-related themes (Bholat, Hansen, Santos, & Schonhardt-Bailey, 2015;Golub, Kaya, & Reay, 2015;Moschella & Pinto, 2019;Schonhardt-Bailey, 2013).…”

Researching Technocracy: Actor-Centred Methodologies and Empirical Strategies for Studying EU Economic Governance in Hard Times

“…Geuijen, Hart, Princen, and Yesilkagit (2008); Suvarierol (2011) Experiments high yes Doberstein (2017) Recent advances in automated content analysis have made the quantitative analysis of speech data another viable option. For example, Bennani and Neuenkirch (2017) provide an indicator for inflation preferences of ECB policymakers based on the tone they reveal in their public speeches. In another attempt of 'dissecting the brains' of policymakers, Bennani (2015) constructs cognitive maps for European central bankers based on the economic reforms they advocate publicly.…”