Call Me Maybe: Experimental Evidence on Using Mobile Phones to Survey Microenterprises

Garlick, Orkin and Quinn (2017) run the first randomised controlled trial to compare microenterprise data from surveys of different frequency – weekly or monthly – and medium – phone or in-person. They show that neither the quality of survey data nor the probability of respondents remaining in the panel differ substantially by frequency or medium, though respondents are more likely to miss higher-frequency interviews.