Measuring Farm Labor : Survey Experimental Evidence from Ghana

Gaddis, Isis ; Oseni, Gbemisola ; Palacios-Lopez, Amparo ; Pieters, Janneke

World Bank, Washington, DC

关键词: RECALL BIAS;    MEASUREMENT ERROR;    FARM LABOR;    AGRICULTURAL PRODUCTIVITY;    GENDER;   

DOI  :  10.1596/1813-9450-8717
RP-ID  :  WPS8717

学科分类：社会科学、人文和艺术（综合）

来源: World Bank Open Knowledge Repository

【 摘 要 】

This study examines recall bias in farmlabor by conducting a randomized survey experiment in Ghana.Hours of farm labor obtained from a recall survey conductedat the end of the season are compared with data collectedweekly throughout the season. The study finds that therecall method overestimates farm labor per person per plotby about 10 percent, controlling for observable differencesat baseline. Recall bias in farm labor per person per plotis accounted for by the fact that households in the recallgroup report fewer marginal plots and farm workers, denotedhere as listing bias. This listing bias also creates acountervailing effect on hours of farm labor at higherlevels of aggregation, so that the recall methodunderestimates farm labor per plot and per household andoverestimates the labor productivity of household-operatedfarms. Consistent with the notion that recall bias is linkedto the cognitive burden of reporting on past events, thestudy finds that recall bias in farm labor has a strongeducational gradient.