Knowledge that Transforms

This paper reports a Monte Carlo study of the Box-Cox model and a nonlinear least squares alternative. Key results include the following: the transformation parameter in the Box-Cox model appears to be inconsistently estimated in the presence of conditional heteroskedasticity; the constant term in both the Box-Cox and the nonlinear least squares models is poorly estimated in small samples; conditional mean forecasts tend to underestimate their true value in the Box-Cox model when the transformation parameter is not equal to one; and conditional heteroskedasticity tends to worsen the bias in the Box-Cox predicted values. Copyright 1994 by MIT Press.

To facilitate maximum likelihood estimation for Box-Cox models, several authors have suggested dividing the dependent variable by its sample geometric mean. This paper points out previously unmentioned drawbacks of this 'recalling.' First, the 'resealed' model is not actually equivalent to the untransformed one, so that the procedure involves more than a unit change. Second, there is no clear interpretation of the parameters after such resealing. The authors suggest an interpretation but find that the usual formulas for standard errors and confidence intervals are not asymptotically valid. Only tests for zero coefficients are valid. Thirdly, the authors discuss the appropriate way of measuring elasticities in such models. Copyright 1994 by MIT Press.

A specification test is proposed to test the adequacy of the implicit conditional mean specification in a transformation model. Regardless of the transformation under the null, the statistic is computable from ordinary least squares output via auxiliary ordinary least squares regressions. Further, the test presumes no particular form for the conditional variance of the transformed and untransformed regressands. A simple goodness-of-fit measure that can be used to compare across transformations is a natural byproduct of the analysis. An application to a well-known fisheries data set is included. Copyright 1994 by MIT Press.

This paper compares the geographic dispersion of employment in manufacturing industries across U.S. metropolitan areas with an imputed measure of local industry demand. The results indicate that local demand has significant long-term and short-term location effects in many industries, including some with negligible transportation costs. Variation in location patterns across industries indicates that demand-side agglomeration economies and technological intensity strengthen the pull of local demand, while supply-side economies of scale and agglomeration effects weaken it. Implications are discussed with regard to international trade, technological change, industrialization and development, urban and regional issues, and industrial organization. Copyright 1994 by MIT Press.

Teofilo Ozuna, Jr., Irma Adriana Gomez, Estimating a System of Recreation Demand Functions Using a Seemingly Unrelated Poisson Regression Approach, The Review of Economics and Statistics, Vol. 76, No. 2 (May, 1994), pp. 356-360

John A. Bishop, John P. Formby, Paul D. Thistle, Convergence and Divergence of Regional Income Distributions and Welfare, The Review of Economics and Statistics, Vol. 76, No. 2 (May, 1994), pp. 228-235

This paper analyzes the technical efficiency of group home care for the developmentally disabled in the Detroit metropolitan area. Linear programming methods are used to construct a production frontier that allows measurement of relative technical efficiency among homes in the sample. A sensitivity analysis is performed to identify influential observations in the data that might result from measurement error that could distort the efficiency measures. A regression analysis is undertaken to examine the effect of various factors on measured efficiency. Policy implications are discussed in the concluding section. Copyright 1994 by MIT Press.

Janez Prasnikar, Jan Svejnar, Dubravko Mihaljek, Vesna Prasnikar, Behavior of Participatory Firms in Yugoslavia: Lessons for Transforming Economies, The Review of Economics and Statistics, Vol. 76, No. 4 (Nov., 1994), pp. 728-741

A controversial aspect of the management buyouts that were popular throughout the 1980s is the potential for a conflict of interest to arise when a manager bids to acquire the firm he manages. This study examines 184 management buyouts and reports three findings. First, returns to prebuyout shareholders are greater when managers must bid against outside acquirers. Second, bid revisions in the face of competition exceed revisions due to shareholder litigation and negotiations with boards. Third, the incidence of competition is negatively related to the prebuyout share holdings of managers. Coauthors are Ronald F. Singer, Anju Seth, and Darla F. Lang. Copyright 1994 by MIT Press.

This paper describes a low variance simulator of the normal distribution function. The probability integral is evaluated exactly at an initial point specified with a factor analytic covariance structure, so that the integral can be derived using dimension reduction techniques. The line integral between the initial point and the desired point is evaluated using Plackett's identity. A Monte Carlo simulation of this line integral simulator (LIS) with the Geweke-Hajivassiliou-Keane (GHK) simulator demonstrates that for dimensions of ten or less, the LIS outperformed the GHK simulator, typically by an order of magnitude and in some cases by two orders of magnitude. Copyright 1994 by MIT Press.