Terry G. Seaks

Functional Form in Regression Models of Tobin's q

The Box-Cox transformation is used to compare alternative functional forms of market value equations. Based on evidence from a panel of 480 publicly-traded U.S. manufacturing companies and two additional data sets used previously in the literature, the semilog form of a Tobins q equation is found to be strongly preferred to the commonly estimated linear form. The authors provide illustrations in which inferences can be affected by the choice of functional form. Copyright 1993 by MIT Press.

A Comparison of Two New Measures of Tax Progressivity [Measurement of Tax Progressivity: An International Comparison]. [Measurement of Tax Progressivity]

Recently N. C. Kakwani (I977) and Daniel Suits (I977) independently and simultaneously developed new measures of the degree of tax progression. Both measures are related to the Gini ratio of income inequality and are distinct from conventional elasticity measures of progressivity in that each yields a summary statistic, while standard elasticity measures of progressivity vary as the tax base changes. In this paper we briefly describe the measures, identify their important characteristics, and establish the analytical relation between them. Empirical calculations of the progressivity of the U.S. income tax system for the period I 962-76 reveal that the two measures display different magnitudes of change in progressivity. More importantly, in three out of fourteen years the measures moved in opposite directions.

A Hausman test for a dummy variable in probit

A new Hausman test is presented for the exogeneity of a dummy variable in a probit model. It is very easy to implement because of the equivalence of the log likelihood functions for bivariate probit and recursive probit. The procedure is applied to a model of student loan default due to Knapp and Seaks (1992).

Estimation and Testing for Functional Form in First Difference Models

A maximum likelihood method for estimating and testing for the proper functional form in first difference regression models is developed. The parametric transformation of the regression variables we propose includes simple first differences and percentage changes as special cases. The method has a simple relationship to the familiar Box-Cox test, and the coefficient estimation and LR testing are easily implemented with standard regression packages. We apply the new method to three published studies: the St. Louis equation, a money demand model, and a model relating poverty to economic growth.

A multinominal logit analysis of the influence of policy variables and board experience on FOMC voting behavior

Previous studies have used probit or logit models to analyze two states of monetary policy (tighter or looser). In this paper we employ multinominal logit to permit Federal Reserve monetary policy to assume one of three alternative states (tighter, looser, or no change) as a function of three independent economic variables (unemployment, real growth, and inflation) and the amount of experience of the Board of Governors. The results indicate that the Federal Reserve reacted differently under Burns, Miller and Volcker and between Volckers two operating procedures in the formulation of monetary policy.

Estimating the Functional Form of the Independent Variables in Probit Models

Researchers estimating probit models often face the choice of expressing the right-hand-side variable(s) in either natural units or in logarithms. The Box–Cox transformation familiar in classical regression analysis is simple to apply to the regressors in a probit model to distinguish whether xor log(x) is sufficiently close to the optimal, as judged by a likelihood ratio test. A numerical example is provided based on a model of student loan default.

SIMULATIONS WITH ECONOMETRIC MODELS AND ALTERNATIVE METHODS OF ESTIMATION

Despite the increasing importance and increasing use of macroeconometric models, there is an econometric problem of fairly substantial proportions that remains relatively unexplored and frequently unacknowledged in macroeconometric model building. The problem concerns the simulation properties of estimators for simultaneous equations, and it arises from the asymmetry of treatment given to lagged endogenous variables during the estimation process and during the simulation process. For the former process, the lagged endogenous variables are treated as predetermined and fixed at the time of model estimation. However, once the model is complete and is simulated over several time periods, lagged values of the endogenous variables are usually derived by using previous period solutions of the endogenous variables. The lagged endogenous variables that were fixed at the time of estimation have become

A Monte Carlo Evaluation of the Box-Cox Difference Transformation

The Box-Cox difference transformation permits the selection of either the first difference or percentage change form of a time series regression model. Monte Carlo evidence on the small sample properties of the transformation parameter A indicates that the difference transformation works quite well even in samples of size 30. Likelihood ratio testing is compared to an asymptotically equivalent alternative Lagrange Multiplier test. It is shown that values of R2 can often be higher for the incorrect transformation.

The computation of test statistics for multivariate regression models in event studies

Abstract Event studies are often estimated with Zellners seemingly unrelated regressions (SUR) estimator and hypothesis tests are applied to cross equation combinations of the coefficients corresponding to event dummy variables. A number of recent event studies fail to point out a useful computational simplification for obtaining the necessary covariance matrix and test statistics.

Nixon and Ford as Ex-Post Keynesians

Reestimation of a growth rate version of the St. Louis equation suggests that the equation does believe in fiscal policy, but only for the Nixon-Ford administrations.