Working Paper 937
This paper combines high-frequency monetary policy surprises with weekly Google search data in South Africa (2010–2026) to evaluate how central bank communication drives credit-related information acquisition. Using variation in shock magnitudes and pre-existing attention states, the results show that transmission is highly non-linear. Small policy surprises generate little measurable behavioural response, whereas large, salient shocks trigger significant credit search activity. Moreover, this behavioural response is highly conditional, heavily amplifying when baseline public attention is already elevated. This behavioural channel is concentrated within forward-looking formal credit searches, while distress-related and informal credit searches exhibit no systematic response to policy innovations. Ultimately, findings support the view that rational inattention operates as a selective filter on monetary policy transmission, suggesting that strategic central bank communication must actively navigate the public’s shifting capacity to process policy surprises.
Keywords: monetary policy surprises, rational inattention, salience, High-Frequency Data, Google Trends
