True state dependence and heterogeneity in off-farm labour participation
利用宾夕法尼亚农场两期面板数据，估计了考虑真实状态依赖或异质性的Probit模型，检验非农劳动参与持续性的来源，发现先前非农劳动状态对当前参与决策有显著影响。
This paper discusses models of off-farm labour participation estimated with panel data. With this kind of model, unexplained persistence in a labour state may be due to true state dependence (when a previous state modifies attitudes, constraints or parameters) and/or to heterogeneity (unobservable permanent individual characteristics). Probit models assuming either heterogeneity or true state dependence are estimated from a two-period panel survey of Pennsylvania farms, and compared with cross-section or pooled sample models. Statistical tests to infer the source of persistence are proposed. The results suggest that for Pennsylvania farmers and spouses, in addition to other personal, household and farm characteristics found to be important in previous studies, previous off-farm labour state is relevant in off-farm labour participation decisions.