Testing error serial correlation in fixed effects nonparametric panel data models

In this paper we consider the problem of testing serial correlation in fixed effects panel data model in a nonparametric framework. Using asymptotic results developed in Su and Lu (2013), we show that our test statistic has a standard normal distribution under the null hypothesis of zero serial correlation. The test statistic diverges to infinity at the rate of root N under the alternative hypothesis that error is serially correlated, where N is the cross sectional sample size. Simulations show that the proposed test works well in finite sample applications. (C) 2015 Elsevier B.V. All rights reserved.