Technology and the US labor market: Evidence from the sectoral and regional decomposition of the change in the US workforce skill mix

In time series regression models with "short-memory" residual processes, the Durbin-Watson statistic (DW) has been used for the problem of testing for independence of the residuals. In this paper we elucidate the asymptotics of DW for "long-memory" residual processes. A standardized Durbin-Watson stalistic (SDW) is proposed. Then we derive the asymptotic distributions of SDW under both the null and local alternative hypotheses. Based on this result we evaluate the local power of SDW. Numerical studies for DW and SDW are given.