Pith. sign in

REVIEW 1 major objections 1 minor 8 references

Central Bank Communication with Public: Bank of England and Twitter (X)

T0 review · 1 major / 1 minor · reviewed 2026-08-07 · deepseek-v4-flash

Pith's one-line read A decade of Bank of England Twitter data shows content and timing beat posting volume for public engagement.

desk verdict Solid descriptive contribution on Bank of England Twitter engagement, but the headline determinants result needs exposure controls before it can carry the policy weight. read the letter →

arxiv 2506.02559 v1 pith:RSM2BUI7 submitted 2025-06-03 econ.GN q-fin.EC

classification econ.GNq-fin.EC
keywords centralbankcommunicationTwitterengagementofEnglandmonetarypolicysocialmediapublicreadabilitytweetelasticity
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks what makes the public engage with the Bank of England on Twitter, using 3.13 million tweets that mention the Bank from 2007 to 2022, including 9,810 official posts. It claims that content quality, timing, and relevance to current economic conditions matter substantially more than posting volume for public engagement. The evidence: evening hours draw peak engagement even though the Bank posts mainly at 9 a.m.; a Turing £50-note post drew 1,300 times the average engagement; videos raise likes by 1,700 percent, photos by 126 percent; and more readable language lifts all engagement metrics. The paper concludes that central banks should post accessible, media-rich content during high-attention moments rather than simply tweet more often.

What carries the argument

The paper's central machinery is a two-model empirical setup applied to the official BoE tweets. Model 1 regresses the logarithm of weekly engagement (likes, replies, retweets, quote tweets) on the logarithm of weekly tweet count, estimated month by month to produce a time-varying elasticity series; this is what yields the average total elasticity of 1.095 and the volatility contrast with the Federal Reserve. Model 2 is a Poisson count regression of each tweet's engagement on tweet characteristics: an MPC-announcement dummy, reply status, link and hashtag indicators, mutually exclusive media-type dummies (GIF, photo, video), and a Flesch Reading Ease readability score. The Poisson coefficients are exponentiated to give percentage effects, and the readability and media coefficients are the load-bearing numbers behind the 'content over volume' conclusion.

What would settle it

Re-estimate Table 4 with an exposure offset equal to each tweet's age (or with year fixed effects and time-since-posting controls) and check whether the video, photo, and readability coefficients survive; if they shrink toward zero, the paper's central claim that content characteristics drive engagement fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that the Bank of England's audience on Twitter responds far more to what the Bank posts and when it posts than to how many times it posts. Across the 2011-2022 official account, total engagement elasticity with respect to tweet volume averages 1.095, but with wide swings, including deep negatives around Brexit; unlike the relatively stable Federal Reserve pattern, the BoE's engagement is volatile and content-dependent. In the tweet-level Poisson regressions, posts on Monetary Policy Committee announcement days receive 122 percent more likes, 270 percent more retweets, 123 percent more replies, and 376 percent more quote tweets. Videos multiply likes by roughly 1,700 percent, photos by 126 percent, and a one-point increase in Flesch Reading Ease readability adds 0.6-1.8 percent across metrics. The paper therefore concludes that effective central bank communication on social media requires strategic content optimization, not higher posting frequency.

Load-bearing premise

The Poisson model treats the 9,810 official tweets as directly comparable in engagement even though they were collected in July 2022, so a 2011 tweet had up to eleven years to accumulate likes while a 2022 tweet had only weeks; if older tweets also differ in format, the media and readability coefficients are biased.

Editorial extensions

If this is right

  • Shifting a share of posts from 9 a.m. to evening hours when engagement peaks would raise total interaction without any increase in posting volume.
  • Reserving media-rich, plain-language formats for Monetary Policy Committee announcement days would concentrate engagement where the public is already paying attention.
  • Because reply tweets receive 44-85 percent less engagement, the Bank's move toward two-way conversation needs formats that keep replies visible, such as threads or scheduled Q&A.
  • Because engagement elasticity is volatile and often negative, simply tweeting more cannot be relied on to scale public engagement.
  • Cultural and banknote content produces exceptional spikes, so these posts are valuable but not a scalable template for routine policy communication.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the Bank shifted routine policy tweets to video format, the 1,700-percent video effect would likely shrink, since videos today coincide with the most newsworthy announcements; the paper itself flags this endogeneity.
  • The absence of tweet-age controls means the media and readability coefficients may partly measure how long a tweet has been live; a re-analysis with exposure offsets is the natural next test.
  • The elasticity spikes during Brexit suggest that public attention to central banks is state-dependent; communication strategies could be dynamically timed to economic uncertainty rather than fixed schedules.
  • If the same readability and media effects appear on other platforms or in other central banks' data, the 'content over volume' result becomes a general law of central bank social media; if not, it is a Bank-of-England-specific phenomenon.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

1 major / 1 minor

Summary. This paper studies the Bank of England's use of Twitter/X by assembling a large dataset of 3.13 million tweets mentioning the BoE and all 9,810 official BoE tweets from July 2011 to July 2022. It documents descriptive patterns in posting volume, timing, and engagement, estimates a monthly elasticity of engagement with respect to tweet volume (Model 1, Eq. 1), and estimates a tweet-level Poisson model of engagement determinants (Model 2, Eq. 2, Table 4). The headline conclusions are that content quality, timing, and relevance matter more than posting volume; that MPC announcement days, video/photo content, and more readable language are associated with substantially higher engagement; and that the BoE should therefore prioritize accessible, media-rich content at high-attention moments rather than increasing tweet frequency.

Significance. If the causal interpretation of Table 4 were warranted, the paper would offer concrete, actionable guidance for central bank digital communication, extending a literature that has mostly relied on surveys and experiments. The descriptive contributions are valuable: the 11-year official-account panel, the documentation of the posting-time/engagement-time mismatch, and the quantification of engagement differentials for cultural content such as the Alan Turing banknote are all useful and credible. The paper also makes appropriate comparisons to the Federal Reserve evidence of Gorodnichenko, Pham, and Talavera (2024). However, the policy implications rest on the Poisson coefficients in Table 4, and the manuscript does not control for tweet age or calendar-year effects in an 11-year sample with cumulative engagement counts. This is a load-bearing identification problem that affects the central claim, so the paper needs substantial revision before the headline results can be accepted.

major comments (1)
  1. [Section 4.2.3 and Section 5 (Implications)] The manuscript acknowledges in Section 4.2.3 that video coefficients 'should not necessarily be interpreted as causal effects, as videos are typically reserved for the most important announcements,' but the policy implications in Section 5 and the abstract present media-rich content as a strategy that 'would likely yield greater public engagement.' This asymmetry is problematic: once the confound between media type and announcement importance is admitted, the same concern applies to photos, GIFs, and readability, since these features also vary systematically over time and across announcement types. The authors should either estimate specifications that address selection on announcement importance (e.g., controlling for tweet topic, MPC day, and year) or substantially soften the causal language in the abstract and conclusions.
minor comments (1)
  1. [References] Several references appear inconsistently: Ehrmann and Wabitsch (2022a) and (2022b) have identical titles in the reference list, and Haldane (2017) and Haldane (2018) refer to essentially the same speech. Please correct these entries.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: all headline results are regression estimates from scraped tweet data; the borrowed econometric methodology is external and not self-referential.

full rationale

This paper makes no derivational claim from first principles, and no central result is defined in terms of another result it is supposed to support. The elasticity estimates in Model 1 and the Poisson coefficients in Model 2 are all estimated directly from the 9,810 official BoE tweets and 3.13 million mention-level tweets; they are not fitted parameters relabeled as predictions. The econometric specification is explicitly borrowed from Gorodnichenko, Pham, and Talavera (2024), which is an independent published source, and the paper applies it to a new dataset with stated modifications; that is methodological borrowing, not a load-bearing self-citation. The authors do cite their own institutional affiliation, but they do not rely on any prior work by themselves as evidence. The paper also candidly acknowledges in Section 4.2.3 that video effects may not be causal because videos are reserved for important announcements; that is a limitation, not a circular reduction. Concerns about missing exposure offsets, tweet-age controls, or year fixed effects are threats to causal identification and robustness, not circularity under the stated criteria. No equation reduces to another by construction, no uniqueness theorem is imported from the authors' own work, and no empirical pattern is renamed as a new organizing principle. Accordingly, the appropriate finding is no significant circularity, with score 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper's conclusions rest on several unstated statistical assumptions. The most important is comparability of tweets across time without an age control; the others concern statistical power, readability measurement, and the interpretation of engagement as effectiveness.

assumptions (4)
  • domain assumption Tweet engagement counts are comparable across the 2011-2022 sample without controlling for tweet age or calendar time.
    Entering at Section 3.1.2 and Table 4; older tweets have more time to accumulate likes, retweets, replies, and quotes, so this assumption is required for the reported coefficients to be unbiased.
  • domain assumption Monthly elasticities from Model 1 are estimated from enough weekly observations within each month to produce meaningful estimates.
    Entering at Section 3.1.1 and Figures 8-9; with roughly four weeks per month, extreme elasticity values may reflect estimation noise rather than true volatility.
  • domain assumption Flesch Reading Ease calculated on tweet text is a valid readability measure for short messages containing links, hashtags, and emoji.
    Entering at Section 3.2.2; tweets are much shorter than typical documents for which Flesch scores were designed.
  • domain assumption Engagement metrics (likes, retweets, replies, quotes) are a meaningful proxy for public engagement or communication effectiveness.
    Entering at Section 2.2; the authors acknowledge Twitter users are not representative, but the analysis still interprets engagement as a success metric.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Central Bank Communication with Public: Bank of England and Twitter (X)." pith.science (2026). https://pith.science/paper/RSM2BUI7

@misc{pith2026250602559,
  author       = {Pith},
  title        = {Pith review of: Central Bank Communication with Public: Bank of England and Twitter (X)},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RSM2BUI7}},
  note         = {Machine review of arXiv:2506.02559}
}
read the original abstract

Central banks increasingly use social media to communicate beyond financial markets, yet evidence on public engagement effectiveness remains limited. Despite 113 central banks joining Twitter between 2008 and 2018, we lack understanding of what drives audience interaction with their content. To examine engagement determinants, we analyzed 3.13 million tweets mentioning the Bank of England from 2007 to 2022, including 9,810 official posts. We investigate posting patterns, measure engagement elasticity, and identify content characteristics predicting higher interaction. The Bank's posting schedule misaligns with peak audience engagement times, with evening hours generating the highest interaction despite minimal posting. Cultural content, such as the Alan Turing 50 pound note, achieved 1,300 times higher engagement than routine policy communications. Engagement elasticity averaged 1.095 with substantial volatility during events like Brexit, contrasting with the Federal Reserve's stability. Media content dramatically increased engagement: videos by 1,700 percent, photos by 126 percent, while monetary policy announcements and readability significantly enhanced all metrics. Content quality and timing matter more than posting frequency for effective central bank communication. These findings suggest central banks should prioritize accessible, media-rich content during high-attention periods rather than increasing volume, with implications for digital communication strategies in fulfilling public transparency mandates.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

8 extracted references · 8 canonical work pages

  1. [7]

    Microblogging Money: Ex- ploring the World’s Central Banks on Twitter

    “Microblogging Money: Ex- ploring the World’s Central Banks on Twitter. ”BoF Economics Review 4/2024. Korhonen, Iikka, and Elisabeth Newby. 2019 . “Mastering central bank communication challenges via Twitter. ” Research Discussion Papers 7/2019 , Bank of Finland. Kryvtsov, Oleksiy, and Luba Petersen

  2. [8]

    Inflation targeting does not anchor inflation expectations: Evidence from firms in New Zealand

    “Inflation targeting does not anchor inflation expectations: Evidence from firms in New Zealand. ” Brookings Papers on Economic Activity Fall: 151–225. Kyriakopoulou, Danae, and Pierre Ortlieb. 2019 . “Central banks take on social media. ”Global Public Investor 2019 special report: Central Banks Communication: 57–62. Lamla, Michael J, and Dmitri V Vinogra...

  3. [143]

    The long journey of central bank communication

    Issing, Otmar. 2019 . “The long journey of central bank communication. ” , ECB Working Paper Series No. 2300, European Central Bank. Issing, Otmar

  4. [2004]

    Monetary policy alternatives at the zero bound: An empirical assessment

    “Monetary policy alternatives at the zero bound: An empirical assessment. ”Brookings papers on economic activity 2004 (2): 1–100. Bernanke, Ben S. 2007 . “Federal Reserve Communications. ” Speech given at the Cato Institute 25th Annual Monetary Conference, Washington, D.C. Bernanke, Ben S

  5. [2015]

    Inaugurating a new blog

    “Inaugurating a new blog. ”Blog Post: 1–50. Bholat, David, Nida Broughton, Janna Ter Meer, and Ed Walczak. 2019 . “Enhancing central bank communications using simple and relatable information. ”Journal of Monetary Economics 108: 1–15. Binder, Carola. 2017 . “Fed speak on main street: Central bank communication and household expectations. ”Journal of Macro...

  6. [2018]

    Central bank communications and the general public

    “Central bank communications and the general public. ”AEA Papers and Proceedings 108: 578–583. Hansen, Stephen, and Michael McMahon. 2017 . “How the Bank of England reconnected the public 34 with its purpose. ”Blog Post. Hansen, Stephen, and Michael McMahon

  7. [2021]

    Clear, consistent and engaging: ECB monetary policy communication in a changing world

    “Clear, consistent and engaging: ECB monetary policy communication in a changing world. ”ECB Occasional Paper 2021/2742. Bernanke, Ben, Vincent Reinhart, and Brian Sack

  8. [2024]

    Central bank communication on social media: What, to whom, and how?

    “Central bank communication on social media: What, to whom, and how?” Journal of Econometrics: 105869 . Haldane, Andrew. 2017 . “ A little more conversation, a little less action. ”Bank of England-Speech. Haldane, Andrew G

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.