Abstract

This paper analyzes the measurement error arising from limited sample sizes in the estimation of gross and net mobility rates using simulated data and a two-sector stochastic stock-and-flow model. It further investigates how the magnitude of these errors changes with increasing sample size and in response to reductions in the volatility of underlying stochastic shocks. The results demonstrate that sampling variability in mobility statistics declines substantially with increasing sample size, but the rate of improvement diminishes beyond a certain point. I identify a sample size of 10,000 observations as a critical point beyond which the marginal reduction in standard error becomes negligible.