Measuring Employment : Experimental Evidence from Urban Ghana

Using a randomized survey experiment in urban Ghana, this paper demonstrates that the length of the reference period and the interview modality (in person or over the phone) affect how people respond in labor surveys, with impacts varying markedly by job type. Survey participants report significantly more self-employment spells when the reference period is shorter than the traditional one week, with the impacts concentrated among those in home-based and mobile self-employment. In contrast, there is no impact of the reference period on the incidence of wage employment. The wage employed report working fewer days and hours when confronted with a shorter reference period. Finally, interviews conducted on the phone yield lower estimates of employment, hours worked, and days worked among the self-employed who are working from home or a mobile location as compared with in-person interviews.