To Answer or Not to Answer: What Determines Telephone Survey Response Rates in Times of Lockdown?

Abstract

Surveys in low- and middle-income countries evoke much higher response than those in richer countries. This has led to a situation wherein there is much less interest by researchers on various factors that might be impacting survey response in the former. However, many factors, including increasing urbanisation and industrialisation and accompanying demographic and socio-economic changes in countries of the global South, may lead to a situation where the survey response is no longer high warranting researchers to pay greater attention to it. In this article, the authors undertake this exercise by drawing on the panel the Delhi Metropolitan Area Telephonic Survey (DMAS-T) that was conducted in the National Capital Region during the pandemic from April 2019 to June 2021 (this includes periods of the COVID-induced lockdown). The results indicate that underneath the initial high response to the survey, there is much heterogeneity by the lockdown phase, survey month, and other individual and socio-economic background characteristics of the respondents. The results reinforce the argument for paying greater attention to survey response as a means of understanding survey data as also of designing better surveys with the intention to reduce non-response and/or bias. JEL Code: C830