In 1991, the U.S. Council of Economic Advisers undertook an initiative to increase the quality of economic statistics. One specific objective was to reduce the size of revisions in GNP estimates. The authors present evidence that one straightforward and inexpensive way of forwarding this objective is for the Department of Commerce to utilize better publicly available information released by other governmental agencies. An important caveat, however, applies: the relationship appears to be nonlinear. Specifically, the inefficient use of information is concentrated in those quarters where the change in the preliminary GNP estimate is large in absolute value.

We extend our 1985 test of restrictions on Layson and Seaks' (1984) Box-Cox difference transformation to models with autoregressive terms and moving average errors. The tests are based on the Lagrange Multiplier principle. Several empirical examples are given.

The expectati on of the excess holding yield on a long bond is postulated to depend upon its conditional variance. Engle's ARCH model is extended to allow the conditional variance to be a determinant of the mean and is called ARCH-M. Estimation and infer ence procedures are proposed, and the model is applied to three interest rate data sets. In most cases the ARCH process and the time varying risk premium are highly significant. A collection of LM diagnostic tests reveals the robustness of the model to various specification changes such as alternative volatility or ARCH measures, regime changes, and interest rate formulations. The model explains and interprets the recent econometric failures of the expectations hypothesis of the term structure.