One characteristic feature of central banks today is that policy decisions are almost exclusively made by a committee rather than by a single policy maker. Another is that central banks are considerably more transparent than they used to be. Together, this has brought to the fore an important but so far unresolved issue: to what extent should a central bank's communication reflect the full spectrum of opinions among its committee members? Does information on all members views make monetary policy easier to understand and predict, or does it make it harder? We address this issue by employing a novel method. We measure the sentiment and tone of the minutes of the Swedish central bank using an automated content analysis that converts the qualitative information in the minutes to a quantitative measure. We find that this measure is useful in predicting future policy rate decisions.
Central Bank Communication Minutes Content Analysis