The federal-private wage differential is reexamined with recent data. Using specifications similar to those used in three previous studies, I find that the federal-private wage differential has narrowed since the 1974-79 period analyzed in those studies. In addition, I discover that wage differentials based on specifications used in the literature are biased upward because the previous studies did not control for detailed occupational and locational characteristics of workers. Corrected point estimates of the wage gap are much smaller than previous estimates in the literature. Differences in the gap across demographic groups, occupations, and locations are also discussed.

Many economic researchers have attempted to measure the effect of aggregate market or public policy variables on micro units by merging aggregate data with micro observations by industry, occupation, or geographical location, then using multiple regression or similar statistical models to measure the effect of the aggregate variable on the micro units. The methods are usually based upon the assumption of independent disturbances, which is typically not appropriate for data from populations with grouped structure. Incorrectly using ordinary least squares can lead to standard errors that are seriously biased downward. This note illustrates the danger of spurious regression from this kind of misspecification, using as an example a wage regression estimated on data for individual workers that includes in the specification aggregate regressors for characteristics of geographical states.

The error Components regression model is now widely applied in econometics and statistics. Given the potentially high costs of incorrectly excluding the component from the model, an error components test should have high power. In addition, if a test is to gain acceptance from practitioners, the test should be computed easily. To obtain improved critical-value approximations, we introduce a standardized Lagrange multiplier (SLM) test statistic, which is centered and scaled to have a zero mean and unit variance under the null hypothesis. We also examine the F test, which is easily computed and has a well-known exact distribution under the null hypothesis if the regression errors are normally distributed

Integrated GDP-Productivity Accounts by Michael J. Harper, Brent R. Moulton, Steven Rosenthal and David B. Wasshausen. Published in volume 99, issue 2, pages 74-79 of American Economic Review, May 2009