Over the past decades, communication has become an important instrument for monetary policymakers (Woodford 2013). By communicating clearly and frequently, central banks provide accountability to the public. Also, using various communication channels, central banks aim to steer private-sector expectations of macroeconomic developments and potential policy responses (Blinder, Ehrmann, Fratzscher, De Haan and Jansen 2008). It has also been argued that communication is useful in the context of macroprudential supervision (Born, Ehrmann and Fratzscher 2010).
From the start of its operations, the European Central Bank has extensively communicated about its monetary policy. Most prominently, since January 1999, the ECB discusses its interest rate decisions in monthly press conferences, which are closely followed by markets participants (Ehrmann and Fratzscher 2009). At the same time, many observers were critical of ECB communications, especially during the early years of the ECB’s operations. Cechetti and Schoenholz (2008) note that "the ... record is filled with outside complaints about ECB communication with financial markets", while Gros and others (2000) discuss how “such comments have often expressed conflicting views of the policy outlook and of policy priorities, making it harder to understand collective decisions inside the ECB”. ECB officials themselves have also acknowledged the challenges of communicating their policies (Hämäläinen 2001).
Yet, the apparent inconsistencies seem at odds with evidence that ECB communications influence market expectations (Ehrmann and Fratzscher 2007) and help in predicting future policy moves (Sturm and De Haan 2011). We resolve this paradox by showing that ECB introductory statements were, in fact, quite consistent over the first decade of its operations (Jansen and De Haan 2013).
To investigate the paradox between inconsistency and effectiveness, we assess the similarities in the ECB’s vocabulary over time. How alike was the way in which the ECB communicated about its monetary-policy stance in 1999 and 2009? The thought experiment is as follows. Suppose someone had closely followed and analysed ECB communications during the early years of the Economic and Monetary Union. How well would this person have been able to understand monetary policy during subsequent years, without heavily investing in any further kind of analysis?
We conduct this thought experiment using a methodology called Wordscores. Laver, Benoît and Garry (2003) developed this method to study manifestos of British political parties. We apply Wordscores to the introductory statements by the ECB president at the monthly press conferences, as that is the prime communication device for the ECB. The idea underlying Wordscores is the following: from texts of which the policy position is known (the so-called reference texts), Wordscores extracts information using word frequencies. Then, this information is employed to estimate the policy positions of subsequent texts about which nothing is known (so-called ‘virgin texts’).
Our basic analysis uses ECB introductory statements from the years 1999-2001 as reference texts, and statements from subsequent years as virgin texts. In all cases, the policy position is the ECB’s monetary policy stance. To use Wordscores properly, the reference texts should be used in the same context as the virgin texts. This implies that the reference texts relate to a similar context as the virgin texts. We use this requirement as the basis for studying consistency. If the introductory statements from the first three years of the Economic and Monetary Union are useful to understand statements from later years, one can conclude that the two sets of communications are internally consistent.
Figure 1 illustrates the gist of our findings. As noted, we use introductory statements between 1999 and 2001 as reference texts, and statements from 2002 to 2009 as virgin texts. The green line, obtained by applying the Wordscores approach, measures the ECB’s monetary-policy stance as explained in the press conferences. Positive (negative) values denote a hawkish (dovish) policy stance. The red bars denote the actual changes in the ECB’s main refinancing rate, where positive (negative) values denote tightening (easing) of monetary policy.
Figure 1. The ECB’s policy stance (2002-2009) measured by applying Wordscores
Note: The black dots connected by the green line denote the Wordscores for each introductory statements by the ECB presidents between January 2002 and July 2009. Positive (negative) values denote a hawkish (dovish) policy stance. The red bars denote the change in the ECB policy rate, where positive (negative) values indicate monetary policy tightening (easing). The reference texts are taken for the period 1999–2001 and coded using four ECB communication indicators from the literature.
It is striking how the estimated policy stance obtained through Wordscores broadly follows the actual ECB interest rate policy choices. Solely on the basis of the introductory statements during the first three years of the EMU, the Wordscores method is able to present a fairly accurate picture of the ECB’s policy stance in subsequent years up to mid-2009. On the basis of these and other results, we conclude that ECB communication, as laid down in its introductory statements of the last decade, may be considered as internally consistent.
Disclaimer: The views expressed here are those of the authors and do not necessarily represent those of the institutions with which they are affiliated.
Blinder, Alan S, Michael Ehrmann, Marcel Fratzscher, Jakob de Haan and David-Jan Jansen (2008), “What we know and what we would like to know about central bank communication”, VoxEU.org, 15 May.
Born, Benjamin, Michael Ehrmann, and Marcel Fratzscher (2010), “Macroprudential supervision – can central bank communication be an effective policy tool?”, VoxEU.org, 29 November.
Cecchetti, SC and KL Schoenholtz (2008), “How Central Bankers See It: The First Decade of ECB Policy and Beyond”, NBER Working Paper 14489.
Ehrmann, M, and M Fratzscher (2007), “Communication by Central Bank Committee Members: Different Strategies, Same Effectiveness?”, Journal of Money, Credit and Banking 39(2–3), 509–541.
Ehrmann, M and M Fratzscher (2009), “Explaining Monetary Policy in Press Conferences”, International Journal of Central Banking 5(2), 41-84.
Hämäläinen S (2001), “The ECB's monetary policy: accountability, transparency and communication”, speech at the conference ‘Old Age, New Economy and Central Banking’, organised by CEPR/ESI and Suomen Pankki.
Gros, D, O Davanne, M Emerson, T Mayer, G Tabellini, and N Thygesen (2000), Quo Vadis Euro? The Cost of Muddling Through, Second Report of the CEPS Macroeconomic Policy Group, Centre for European Policy Studies, Brussels.
Jansen, David-Jan and Jakob de Haan (2013), “An assessment of the consistency of ECB Communication”, in Pierre S Siklos and Jan-Egbert Sturm (eds.) Central bank communication, decision-making and governance, MIT Press, 183-201.
Laver, M, K Benoît, and J Garry (2003), “Extracting policy positions from political texts using words as data”, American Political Science Review 97(2), 311-31.
Sturm, J-E, and J De Haan (2011), “Does Central Bank Communication Really Lead to Better Forecasts of Policy Decisions? New Evidence Based on a Taylor Rule Model for the ECB”, Weltwirtschaftliches Archiv/Review of World Economy 147(1), 41–58.
Woodford, Michael (2013), “Inflation targeting: Fix it, don’t scrap it” in Lucrezia Reichlin and Richard Baldwin (eds.) Is inflation targeting dead? Central Banking After the Crisis, VoxEU.org ebook, 14 April.