Leapfrog estimation of a fixed-effects model with unknown transformation of the dependent variable

摘要: Abstract This paper considers a fixed-effects panel version of the linear transformation model, in which dependent variable is h(yt) for an unspecified, strictly monotonic h. Examples model include multiple-spell proportional hazards and dependent-variable models (e.g., Box–Cox model) with fixed effects. A semiparametric estimator, called leapfrog introduced shown to be n -consistent asymptotically normal. The estimator allows h vary over time heteroskedasticity across observational units. Related estimators are considered, general covariance result based on second-order U-processes presented.