When all speak, should we listen? A cross-country analysis of disagreement in policymaking and its implications

This study proposes an ‘implicit measure’ of disagreement to investigate presence of any latent discord in the policy briefings of central banks across nations. We analyse Monetary policy statements of the Bank of Japan (Japan), Banco Central do Brasil (Brazil) and Central Bank of the Republic of Turkey (Türkiye) on both specific as well as overall economic outlook by using novel text mining tools. Subsequently, we explore the relationship of the proposed measure of implicit disagreement with the accuracy of growth and inflation forecasts. Our findings confirm that disagreement/discord among policymakers over varied economic aspects provides signals to forecasters and helps in aligning their forecasts better. The results of this study indicate that disagreement too can be considered as a plausible criterion that can explain the forecaster's belief thereby playing a positive role in ‘forecaster learning’.