Estimation of Poverty Transition Matrices with Noisy Data

This paper investigates potential measurement error biases in estimated poverty transition matrices. We compare transition matrices based on survey expenditure data to transition matrices based on measurement-error-free simulated expenditure. The simulation model uses estimates that correct for measurement error in expenditure. This dynamic model needs error-free initial conditions that can not be derived from these estimates. We provide bounds on the initial-conditions parameters, when these initial conditions are obtained by projection, and we also obtain initial conditions on the assumption that there is no time-constant measurement error. We ?nd that for both estimates of the initial conditions measurement error in expenditure data magni?es economic mobility in and out of poverty. Roughly 44% of households initially in poverty at time t??1 are found to be out of poverty at time t using expenditure data from the Korean Labor and Income Panel Study (KLIPS). However, when we remove measurement error through a model-based simulation, only between 32 and 40% of households initially in poverty are found to be out of poverty.