HomeEconomyCentral Bank on Social Media

Central Bank on Social Media

Central Bank on social media: Experts and non-experts accept ECB exchanges on Twitter

Central banks are increasingly using social media to communicate with target audiences—experts have long been the focus of central bank communication, but they are also increasingly communicating with non-experts. The latter is a special challenge until some commentators predict that this effort by the central bank is doomed to fail (Blinder 2018). Non-experts generally have little knowledge of central banks and monetary policy, and they have little knowledge of central bank issues (van der Cruijsen et al. 2015). Consistent with this, there is ample evidence that the central bank has limited communication with the general public. This is not only reflected in the limited understanding of more complex issues such as monetary policy strategies (Coibion ​​et al., 2020), but also in the general lack of understanding of central bank inflation targets (Christelis et al., 2020).

Non-experts rarely obtain information about the central bank directly from sources. Instead, they learn about central banks and monetary policy through media reports (ter Ellen et al. 2020). This kind of intermediary not only means that the media can play an important role in shaping economic results, but it may also weaken the central bank’s ability to communicate. Corbion et al. (2019) shows that consumers have a much greater reaction to inflation expectations when reading the central bank statement than reading news articles about the same statement.

Therefore, communicating via social media may be a promising way to communicate directly with experts and non-experts. In a recent paper (Ehrmann and Wabitsch 2021), we analyzed the ECB-related traffic on Twitter to understand how monetary policy issues are discussed on social media, and how Twitter traffic affects ECB communications Responsiveness and method.

Tweets about the European Central Bank

To perform this analysis, we collected a data set of tweets about the European Central Bank or its president in English (the main language of financial markets and economics) and German (the largest language used in the Eurozone). These data are from 2012 (when the use of Twitter became more stable) to 2018 (in order not to confuse the analysis with the replacement of the European Central Bank President from Mario Draghi to Christina Lagarde, the latter has its own, Very active, Twitter account). These tweets must contain one of the keywords “ECB”, “European Central Bank” or “Draghi” (or their German translation), and are subsequently extensively cleaned up to ensure that we only keep the tweets that are actually related to ECB. For example, we filter out tweets related to “England and Wales Cricket Board” (also abbreviated as ECB). For all the remaining tweets, we then calculate the favorability (i.e., the degree of positive or negative expressed opinion), the absolute value of favorability (indicating the strength of the language used) and subjectivity (a measure of the degree of a tweet) The measurement is mainly factual or quite subjective) using the dictionary method.

In addition, we distinguish between expert and non-expert Twitter accounts based on simple classification. First, let’s assume that Twitter users who tweet on the day of at least every two ECB press conferences are experts on monetary policy issues. Second, we impose that non-experts not only tweeted less frequently at ECB press conferences, but their tweets related to the ECB accounted for only a small portion of their Twitter activity.1 Although the expert group is much smaller than the non-expert group, it generates a disproportionately large share of tweets related to the European Central Bank.

  • Tweets related to the European Central Bank by non-experts are more subjective, use stronger words, and more diffuse emotions

Figure 1 compares the tweets posted by non-experts (blue) and experts (red) by plotting the average subjectivity and average language strength of each Twitter account. The figure clearly shows that non-expert tweets are more diverse than expert tweets, the average subjectivity of non-experts is stronger, the language used is also stronger, and there is a positive correlation between the two concepts, indicating more Subjective opinions also tend to be expressed in stronger language.

figure 1 Emotions expressed in tweets by experts and non-experts

notes: This graph shows the average opinion strength and average subjectivity (both measured on a scale of 0 to 1) of ECB-related tweets published in English by expert accounts (red) and non-expert accounts (blue).

  • More subjective tweets and stronger language tweets are retweeted more frequently

For every tweet in our dataset, we know how often they are reposted. Researching the determinants of retweets helps to understand which views influence the debate on Twitter. Although only about one-seventh of the tweets are reposted, some of them may have a high impact. On average, on the condition of being reposted, the number of reposts for a tweet is about 4.5 times, but the 99th percentile is 43, and the maximum is 4,868. Tweets that are more subjective and those that use stronger languages ​​are much more likely to be retweeted. This shows that the discussion on the European Central Bank on Twitter has been disproportionately affected by strong language and relatively subjective views.

  • Non-experts are responsive to the ECB’s communication, but not as good as experts

Figure 2 shows how the number of English tweets (logs) responds to ECB communications on the same day.2 The graph shows that the press conference and the “whatever it takes” statement made in 2012 by the then European Central Bank President Draghi constituted the most important event in our sample, and the reaction speed was much higher on the same day. More professional communication activities are less attractive on Twitter. It was found that the response of non-experts was not as strong as that of experts, with the exception of the obvious exception of “whatever it takes.”

figure 2 Response ability of English Twitter traffic to ECB communication events

notes:Estimation of the coefficient of the influence of ECB communication events on the number of English tweets logs on the same day. Response coefficients that are statistically significant at least at the 10% level are drawn in dark shades. “At all costs” refers to Mario Draghi’s corresponding remarks on July 26, 2012. “Others’ speeches” are speeches by members of the European Central Bank’s Executive Committee other than the President of the European Central Bank. “Tweets” refer to tweets originating from the Twitter account of the European Central Bank in the days when other ECB communications events were not captured in the analysis.

  • In response to most communication events, tweets have become more authentic and express a more moderate point of view in a more unified way

Figure 3 provides a typical example of how sentiments expressed on Twitter respond to ECB communications. Before and after the ECB press conference, non-experts’ tweets in English were less subjective, and on average expressed their views in more moderate language. In addition, the differences between different Twitter accounts are smaller, which indicates that the opinions expressed are less dispersed than other accounts. These models are consistent with Twitter as a medium for transmitting information from the European Central Bank to non-experts.

image 3 Changes in sentiment in English tweets before and after the ECB press conference

notes:Estimation of the coefficient of influence of ECB communication events on Twitter sentiment before and after the ECB press conference (including 4 days ago and 5 days later). Response coefficients that are statistically significant at least at the 10% level are drawn in dark shades.

  • In some cases, the ECB’s communication has triggered controversial discussions on Twitter

The most prominent example is “Whatever it take”, which not only leads to a strong and continuous increase in overall traffic, but also has a great impact on the sentiment expressed on Twitter. This is most prominent in German and non-expert tweets. It can be seen from Figure 4 that the subjectivity and language intensity of experts and non-experts have changed significantly, but the changes in non-expert groups have been even greater. In this case, the transmission of information is likely to be dominated by highly fragmented and controversial discussions on Twitter. Interestingly, Twitter users regard the President of the European Central Bank as a person on the one hand, and distinguish it as an institution or its policies on the other. In the discussion around “whatever it takes”, comments on Mario Draghi’s tweets have become more subjective and more diverse on average, while the subjectivity of the ECB’s tweets has not changed significantly.

Figure 4 Emotional changes in German tweets after “whatever it takes”

notes: After “whatever it takes” (the same day plus the following 15 days), an estimate of the coefficient of the influence of the ECB communication event on the sentiment of the German tweets. Response coefficients that are statistically significant at least at the 10% level are drawn in dark shades.

Impact on central bank communication

The results described in this article have a direct impact on central bank communication. Central banks can reach non-experts-as we have shown, they respond to the ECB’s communication. Although they tend to have more subjective opinions than experts and express them in stronger language, in some cases, the ECB’s communication will trigger a relaxation of opinions and more factual discussions. This is important because in general, more subjective tweets expressed in stronger language are more likely to be reposted and therefore tend to be more influential.

Author’s note: This column is also published as a SUERF policy brief. The views expressed in the column represent the personal views of the author and do not necessarily reflect the views of the European Central Bank or the euro system.

refer to

Blinder, A (2018), “Crystal Ball in the Dark: The Future of Monetary Policy Communication”, Papers and Proceedings of the American Economic Association 108, 567–571.

Christelis, D, D Georgarakos, T Jappelli, and M van Rooij (2020), “Trust the central bank and inflation expectations”, International Central Bank Journal 16(6): 1-37.

Coibion​​, O, Y Gorodnichenko, and M Weber (2019), “Monetary policy communication and its impact on household inflation expectations“, VoxEU.org, February 22.

Coibion​​, O, Y Gorodnichenko, ES Knotek II and R Schoenle (2020), “Average inflation target and household expectations“, VoxEU.org, September 30.

Ehrmann, M and A Wabitsch (2021), “The central bank communicates with non-experts: there is no way to go?“, CEPR Discussion Paper No. 16525.

ter Ellen, S, VH Larsen and LA Thorsrud (2020), “Narrative monetary policy surprises and the media: how central banks reach the public“, VoxEU.org, December 8.

van der Cruijsen, C, DJ Jansen and J de Haan (2015), “Public understanding of monetary policy“, VoxEU.org, August 23.


1 For each Twitter account, we calculated the share of the European Central Bank-related tweets in the total number of tweets, and required non-experts to be located at the bottom 25% of the indicator distribution.

2 Apart from the European Central Bank’s press conference and “whatever it takes”, the reaction of the day was sufficient to capture additional traffic. At the press conference, Twitter traffic increased 5 days ago and 4 days later. The impact of “whatever it takes” is estimated to have a significant impact in the following 15 days.

Source link


Most Popular

Recent Comments