We are pleased to announce the latest post on the SUARA SEACEN Blog by Meltem Chadwick, Senior Economist, MMPM, together with her co-authors Jaqueson Galimberti (Asian Development Bank) and Rennae Cherry (Reserve Bank of New Zealand).
Nonresponse Bias is Distorting Household Inflation Expectations: Evidence and Policy Fixes
Our new blog post shows that standard household inflation expectation series are distorted by item nonresponse. Using RBNZ microdata from 1998 to 2022, Chadwick, Cherry and Galimberti (2025) document that about 44 per cent of respondents leave the expectations question blank and that these missing voices are not random. Once corrected for selection with a Heckman selection model, the aggregate inflation expectation series shifts down by roughly 0.3 percentage points on average and most apparent differences by gender, income, and ethnicity fade. Age remains the outlier, with older respondents overpredicting inflation more strongly.
The blog also illustrates that design choices matter. The survey’s shift to online mode in 2018 cut nonresponse to about 24 per cent and narrowed gaps across groups. Response is also state-dependent, rising when inflation drifts far from target.
For Asia-Pacific central banks, the message is clear: Modernise survey modes with inclusion in mind. Mandate proper selection corrections and tailor communication to reach the citizens who are least likely to answer yet most in need of clarity.
Click here to read.
