ŠT—v

Much past panel data research has focused on estimating effects from nonstationary panels with a large time series T, and large individuals N, even though the practical sample size in time dimension is moderate. This paper introduces a regression limit theory for nonstationary panel data with large numbers of cross section but moderate time series observations. For such cases a nonstationary panel limit theory which allows for large N and fixed T is useful. We, therefore, introduce fixed-$T$ panel model with homoscadasiticity disturbance terms both in suprious and cointegrated cases.