Journal of Economic and Social Measurement

This editorial article describes the Journal of Economic and Social Measurement. It discusses the evolution of this journal during the past thirty plus years of its publication. In the process, it identifies the range of topics that have been historically addressed as a means of suggesting types of possible future articles. It includes both a statement of editorial policies and information about normal selection procedures. This article therefore serves to define the scope of this journal, but fundamentally should be understood as intended to be suggestive, rather than either prescriptive or restrictive. The journal's essential characteristic is both to act as a complement to the standard disciplinary journals and to be innovative in its acceptance of new ideas and methods.

This paper has two aims: first to describe methods, issues, and outcomes involved in matching data from the Insurance Component of the Medical Expenditure Panel Survey (MEPS-IC) to other business microdata collected by the US Census Bureau, and second to present some simple results that illustrate the usefulness of such combined data. We present the results of linking the MEPS-IC with data from the 1997 Economic Censuses (EC), but also discuss other possible sources of business data. An issue in any linkage is whether the linked sample remains representative and large enough to be useful. The EC data are attractive because, given the survey's broad coverage and large sample, most of the MEPS-IC sample can be matched to it. We use the combined EC/MEPS-IC data to construct productivity measures that are useful auxiliary data in examining employers' health insurance offering decisions.

This paper examines the problem of estimating real output and inflation using index numbers that are related to nominal output change through addition rather than multiplication. This approach to the development of index numbers is consistent with the manner in which nominal output actually evolves through time relative to some base year. Formulas for additive measures of real output and inflation that accurately capture the relative impacts of price and output change on nominal output are derived, and their use in separating real and nominal effects is illustrated and compared with standard indices. The additive approach is also used to understand the nature of formula bias in some traditional index numbers. Finally, through the test approach, the indices presented in this paper have a number of appealing properties.

An honest confidence interval for the error variance in a stepwise regression is a one-sided interval that adjusts for the effects of variable selection. The endpoint of this interval may be many times larger than the usual endpoint. Such adjustments are most important when selecting variables from a large number of available predictors, particularly in situations with more available predictors than observations. An illustration using a regression model of stock market returns illustrates the calculations.

Some well-known empirical work on the forgone earnings of mothers in Australia became a barrier to a new research project. Initially the existing results could not be replicated, but further investigations and replication attempts eventually revealed the serious errors in the earlier research. Once the results were re-estimated – in the spirit of the original research but with the errors removed – the new research project could proceed. Valid inferences could then be made about the changes in forgone earnings over time for different groups of Australian women.