Identification and Estimation of Nonseparable Single-Index Models in Panel Data with Correlated Random Effects

The identification of parameters in a nonseparable single-index models with correlated random effects is considered in the context of panel data with a fixed number of time periods. The identification assumption is based on the correlated random-effect structure: the distribution of individual effects depends on the explanatory variables only by means of their time-averages. Under this assumption, the parameters of interest are identified up to scale and could be estimated by an average derivative estimator based on the local polynomial smoothing. The rate of convergence and asymptotic distribution of the proposed estimator are derived along with a test whether pooled estimation using all available time periods is possible. Finally, a Monte Carlo study indicates that our estimator performs quite well in finite samples