題目: Robust Inference for Dyadic Data
Professor, Department of Economics
University of California, Davis
時間😨：2017年5月18日（星期四） 16🚵：00-17：30
Abstract: In this paper we consider inference with paired or dyadic data, such as cross-section and panel data on trade between two countries. Regression models with such data have a complicated pattern of error correlation. For example, errors for US-UK trade may be correlated with those for any other country pair that includes either the US or UK. We consider models with regressors treated as predetermined and stationary. The standard cluster-robust variance estimator or sandwich estimator for one-way clustering is inadequate. The two-way cluster robust estimator is a substantial improvement, but still understates standard errors. Some studies in social network data analysis have addressed this issue. The network in international trade studies is much denser than in typical network studies, so it becomes especially important to control for dyadic error correlation. In applications with the gravity model of trade we find that even after inclusion of country fixed effects, standard errors that properly control for dyadic error correlation can be several times those being reported using current methods.
返回頂部
