Abstract: This paper develops a complementary estimation approach for functional coefficient panel data models with sample selection and fixed effects. We propose a pairwise differencing strategy that avoids imposing restrictions on the fixed effects at the cost of not directly estimating level coefficients on time-invariant regressors. We derive the estimator's asymptotic properties and show via simulations that it performs competitively relative to existing methods. Eliminating the need to estimate individual fixed effects yields considerable computational savings, making the approach suitable for large datasets.

Autoría

Descargar referencia