DO NOT CITE OR QUOTE WITHOUT AUTHOR’S WRITTEN PERMISSION Missing(ness) in Action: Selectivity Bias in GPS-Based Land Area Measurements 1

Land area is a fundamental component of agricultural statistics, and of analyses undertaken by agricultural economists. While household surveys in developing countries have traditionally relied on farmers ’ own, potentially error-prone, assessments to collect land area information, the availability of affordable and increasingly reliable Global Positioning System (GPS) units has recently made GPS-based area measurement a practical alternative. Nonetheless, survey implementing agencies typically require interviewers to record GPS-based area measurements only for plots within a given radius of dwelling locations, in order to reduce survey costs, keep interview durations within reasonable limits, and avoid the difficulty of asking respondents to accompany interviewers to distant plots. It is, therefore, common for as much as a third of the sample plots not to be measured, and available research has not shed light on the possible selection bias in analyses that rely on partial data due to gaps in GPS-based area measures. We explore the systematic patterns of missingness in GPS-based plot areas, and investigate their empirical implications in the context of the inverse scale-land productivity relationship. Using Multiple Imputation (MI) to predict missing GPS-base