Type to search

Narrative monetary policy surprises and the media

Breaking News

Narrative monetary policy surprises and the media


Narrative monetary policy surprises and the media: How central banks reach the general public

Saskia ter Ellen, Vegard H. Larsen, Leif Anders Thorsrud 08 December 2020

In 2008, Alan Blinder et al. (2008) wrote that “nearly all the research to date has focused on central bank communication with financial markets. It is time to pay more attention to communication with the general public”. We know a lot about the transmission mechanism between central bank communication and financial professionals – as illustrated by the number of Vox columns on this topic (e.g. Hansen and McMahon 2016, Cieslak and Schrimpf 2018, Leombroni et al. 2018, Lustenberger and Rossi 2018, Ehrman et al. 2019, Schmeling and Wagner 2019, Altavilla et al. 2019, 2020, Baranowski et al. 2020). However, less is known about whether and how central bank communication reaches a wider audience. What we do know from recent evidence is that households are not well informed about monetary policy per se (e.g. Coibion et al. 2019), but that central bank communication makes it more likely for people to indirectly receive monetary policy related news (Lamla and Vinogradov 2019).

The news media transmission channel of monetary policy

In a recent paper (Ter Ellen et al. 2020), we develop a novel method to identify narrative monetary policy surprises using textual data on central bank communication and news media, and use it to study the news media transmission channel of monetary policy. In contrast to related research, our method does not require the public to read news with the purpose of acquiring information about the central bank or its policies explicitly. Rather, it captures transmission through news articles on topics that are important for monetary policy – inflation, labour market developments, the exchange rate, the oil market, financial stability, and uncertainty. Since such narratives are key ingredients in any interest rate decision and important for households (e.g. Larsen et al. 2020), central bank communication has the potential to affect media coverage, and thus expectations, even if households know little about monetary policy per se.

Constructing narrative monetary policy surprises with textual data

Figure 1 Word clouds for the six identified topics visualise the context of the keywords they are based on 

Narrative monetary policy surprises and the media 2

Combining methods from the natural language processing (NLP) literature with the factor identification scheme proposed by Bai and Ng (2010), we obtain six ‘topic factors’ from central bank communication, as well as from the news media prior to and after central bank announcements. The methodology we propose is general, while the economic application uses data specific to Norway. The time series of topic factors that emerge from media and central bank communication can be seen in Figure 2, alongside key macroeconomic aggregates for the Norwegian economy. Overall, the factors reflect the conventional narrative held about economic developments over the last two decades. For example, the estimates suggest that the media focused more on unemployment related issues around 2003, 2009, and 2015, which are periods associated with downturns, or recessions, in the Norwegian economy. Likewise, the enhanced focus on exchange rates and inflation during the earlier parts of the sample, relative to the latter part, is natural given that Norges Bank formally moved from a fixed exchange rate regime to inflation targeting in 2001.

Figure 2 Factors in news media alongside key macroeconomic aggregates

Narrative monetary policy surprises and the media 3  

Note: Figures a-f report, in grey and dotted black, the factors estimated from the news media dataset as well as key macroeconomic aggregates for the Norwegian economy. The narrative central bank factors are reported in black. The frequency on the x-axis is event dates and the quarterly observable variables are mapped to this frequency using the first event date in the quarter. 

Conceptually, our identification of narrative monetary policy surprises informally builds on the mechanism proposed in Nimark and Pitschner (2019) where no agent (i.e. household) has the resources to monitor all events that are potentially relevant for their decision, and thereby delegate their information choice to specialised news providers. We assume households have certain beliefs about macroeconomic conditions, and that such beliefs might be affected by central bank communication that reaches the public through the media if it is newsworthy. Then, narrative monetary policy surprises are identified as the differences in narrative focus between the media and the central bank’s communication along the dimensions of the aforementioned topics.

Using interest rate market data and the high-frequency identification framework pioneered by Gurkaynak et al. (2005) to identify what we label as conventional monetary policy surprises, we find that there is a weak and insignificant relationship between these surprise measures and the proposed narrative surprises. Hence, the narrative surprises identified here capture a different part of the central bank’s communication than conventional monetary policy surprises do – and suggest that the media channel might be complementary to the transmission mechanism traditionally studied.

Narrative monetary policy surprises capture central bank information

Our findings further suggest that central bank communication indeed finds its way to the general public by affecting news media coverage on these topics. First of all, our results show that the narrative monetary policy surprises lead to a significant change in media coverage after the interest rate meeting relative to before. In contrast, conventional monetary policy surprises do not significantly affect changes in media coverage. Subsequently, following narrative monetary policy surprises the interest rate, the stock market, consumer confidence, house prices, and industrial production all increase. These response patterns are not in line with conventional monetary policy shock interpretations, but rather in accordance with what the newer monetary policy literature identifies as the information component of monetary policy surprises. In fact, a positive co-movement between the interest rate and the stock market following monetary policy surprises has been the defining identifying feature of “central bank information shocks” in, for example, Jarocinski and Karadi (2018) and Cieslak and Schrimpf (2018). The common interpretation for this information component is that the central bank, through its communication, reveals private information about its views on current and future economic conditions. The narrative monetary policy surprise is a natural candidate for an information component, both in terms of its estimated impulse responses, and especially in terms of its construction.

Figure 3 Key indicator performance following monetary policy surprise 

a) Three-month interest rate

Narrative monetary policy surprises and the media 4

b) Stock market index

Narrative monetary policy surprises and the media 5

(c) Industrial production        

Narrative monetary policy surprises and the media 6 

d) Consumer confidence index

Narrative monetary policy surprises and the media 7

Note: Figures a-d show how the three-month interest rate, stock market, industrial production, and consumer confidence move in the 24 months after a narrative monetary policy surprise (black lines) as compared to a conventional monetary policy surprise (grey lines). 

Can central banks communicate with the general public?

Blinder (2018) is not very optimistic about central banks’ ability to reach the general public, writing that “[c]entral banks will keep trying to communicate with the general public, as they should. But for the most part, they will fail”. However, there seems to be light at the end of the tunnel. Bholat et al. (2019) show that making communication relatable to people’s life enhances the public’s understanding of central bank statements. This emphasises the importance for central bank communication to focus on the inputs to monetary policy decisions, such as employment conditions and prices, rather than the monetary policy tools itself. After all, households have very little affiliation with the key policy rate or its projections, but face consumption decisions on a daily basis.  As Haldane and McMahon (2018) put it, “the main policy implication is that, while more still needs to be done, there are compelling reasons for central banks to continue trying [to communicate with the general public]”. Our study suggests that central banks indeed should not give up, as their statements, be it indirectly, reach their audience. 

Authors’ note: The views expressed here are those of the authors and do not necessarily represent those of the institutions with which they are affiliated.


Altavilla, C, L Brugnolini, R S Gürkaynak, R Motto, and G Ragusa (2019), “Monetary policy in action: Multiple dimensions of ECB policy communication and their financial market effects”, VoxEU.org, 04 October.

Altavilla, C, R S Gürkaynak, R Motto, and G Ragusa (2020), “Financial market reactions to monetary policy signals”, VoxEU.org, 03 August.

Bai, J, and S Ng (2020), “Principal components estimation and identification of static factors”, Journal of Econometrics, 176, 18-29. 

Baranowski, P, W Doryń, T Łyziak, and E Stanisławska (2020), “Words versus deeds: How central banks manage expectations”, VoxEU.org, 22 October.

Bholat, D, J Broughton, N Ter Meer, and E Walczak (2019) “Enhancing central bank communications using simple and relatable information”, Journal of Monetary Economics, 108, 1-15.

Blinder, A, M Ehrmann, M Fratzscher, J de Haan, and D J Jansen (2008), “What we know and what we would like to know about central bank communication”, VoxEU.org, 15 May.

Blinder, A (2018), “Central bank communications. Through a crystal ball darkly: the future of monetary policy communication”, AEA Papers and Proceedings, 108, 567-571.

Brubakk, L, S ter Ellen, and H Xu (2017), “Forward guidance through interest rate projections: does it work?”, Norges Bank Working Paper.

Cieslak, A, and A Schrimpf (2018), “Monetary and non-monetary news in central bank communication”, VoxEU column, 22 October.

Coibion, O, Y Gorodnichenko, and M Weber (2019), “Monetary policy communications and their effects on household inflation expectations”, VoxEU.org, 22 February.

Ehrman, M, G Gaballo, P, and G Strasser (2019), “Signalling a future path of interest rates: The international evidence on forward guidance”, VoxEU.org, 01 August.

Ellen, ter S, V H Larsen, and L A Thorsrud (2020), “Narrative monetary policy surprises and the media”, Journal of Money, Credit and Banking, forthcoming.

Haldane, A, and M McMahon (2018), “Central Bank Communications and the General Public”, AEA Papers and Proceedings, 108, 578-583.

Hansen, S, and M McMahon (2016), “The nature and effectiveness of central-bank communication”, VoxEU.org, 03 February.

Jarocinski, M, and P Karadi (2018), “The transmission of policy and economic news in the announcements of the US Federal Reserve”, VoxEU.org, 03 October.

Lamla, M, and D Vinogradov (2019), “Central bank announcements: Big news for little people?”, Journal of Monetary Economics, 108, 21-38.

Larsen, V H, L A Thorsrud, and J Zhulanova (2020), “News-driven inflation expectations and information rigidities”, Journal of Monetary Economics, forthcoming.

Leombroni, M, A Vedolin, G Venter, and P Whelan (2018), “Central bank communication and the yield curve”, VoxEU.org, 18 October. 

Lustenberger, T, and E Rossi (2018), “More frequent central bank communication worsens financial and macroeconomic forecasts”, VoxEU.org, 04 March.

Nimark, K P, and S Pitschner (2019), “News media and delegated information choice”, Journal of Economic Theory, 181, 160-196.

Schmeling, M, and C Wagner (2019), “Central bank tone moves asset prices”, VoxEU.org, 22 February 2019.

Leave a Comment

Your email address will not be published. Required fields are marked *