ECHOES OF POLICY: LEVERAGING AI/ML TO SUPPORT CENTRAL BANK COMMUNICATION STRATEGIES

Rudy Marhastari, Cicilia Anggadewi Harun, Retno Muhardini, Agatha Silalahi, Annes Nisrina Khoirunnisa, Rheznandya Arkaputra Azis, Sintia Aurida, Rahardian Luthfan Ihtifazhuddin, Citra Ayu Rossi Wulandari, Alvin Andhika Zulen and Amin Endah Sulistiawati
Additional contact information
Cicilia Anggadewi Harun: Bank Indonesia
Retno Muhardini: Bank Indonesia
Agatha Silalahi: Bank Indonesia
Annes Nisrina Khoirunnisa: Bank Indonesia
Rheznandya Arkaputra Azis: Bank Indonesia
Sintia Aurida: Bank Indonesia
Rahardian Luthfan Ihtifazhuddin: Bank Indonesia
Citra Ayu Rossi Wulandari: Bank Indonesia
Alvin Andhika Zulen: Bank Indonesia
Amin Endah Sulistiawati: Bank Indonesia

No WP/19/2025, Working Papers from Bank Indonesia

Abstract: This study evaluates the effectiveness of Bank Indonesia’s communication strategy by integrating computational linguistics, media sentiment analytics, and macroeconomic diagnostics within a unified empirical framework. Using advanced Natural Language Processing (NLP) techniques, BI’s press releases from 2019–2024 are transformed into quantitative indicators capturing clarity, sentiment, comprehensiveness, consistency, and economic appropriateness. In parallel, news articles on inflation and exchange rate developments are analyzed to assess how policy messages are transmitted or amplified through media channels. These linguistic features are further enriched using Named Entity Recognition to identify stakeholder-specific resonance and potential pathways of narrative distortion within the public communication ecosystem. To assess macroeconomic implications, a VARX model links communication characteristics to intermediary channels, market expectations, and macroeconomic outcomes under both normal and anomalous conditions. Complementing this analysis, an Early Warning System (EWS) employing a 12-month rolling window and IsolationForest anomaly detection identifies periods of inflation and exchange-rate stress, providing a diagnostic foundation for anticipating heightened communication demands. The findings show that central bank communication functions not only as an information conduit but also as an active policy instrument that shapes expectations and influences market behavior. Building on these insights, the study proposes a three-pillar framework: Features, Timing, and Channels; to strengthen clarity, responsiveness, and coherence in central bank communication. This research advances the literature by integrating AI/ML-based diagnostics with policy communication analysis, offering an empirically grounded approach to enhancing communication effectiveness, transparency, and expectation management.

Keywords: Central Bank Communication; Communication Feature; Sentiment Analysis; Communication Impact Analysis; Early Warning System (search for similar items in EconPapers)
Pages: 58 pages
Date: 2025
New Economics Papers: this item is included in nep-cba and nep-mon
References: Add references at CitEc
Citations:

Downloads: (external link)
http://publication-bi.org/repec/idn/wpaper/WP192025.pdf First version, 2025 (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:idn:wpaper:wp192025

Access Statistics for this paper

More papers in Working Papers from Bank Indonesia Contact information at EDIRC.
Bibliographic data for series maintained by Shinta Fitrianti () and Jimmy Kathon ().