Nuisance vs. Substance: Specifying and Estimating Time-Series-Cross-Section Models

In a previous article we showed that ordinary least squares with panel corrected standard errors is superior to the Parks generalized least squares approach to the estimation of time-series-cross-section models. In this article we compare our proposed method with another leading technique, Kmenta's "cross-sectionally heteroskedastic and timewise autocorrelated" model. This estimator uses generalized least squares to correct for both panel heteroskedasticity and temporally correlated errors. We argue that it is best to model dynamics via a lagged dependent variable rather than via serially correlated errors. The lagged dependent variable approach makes it easier for researchers to examine dynamics and allows for natural generalizations in a manner that the serially correlated errors approach does not. We also show that the generalized least squares correction for panel heteroskedasticity is, in general, no improvement over ordinary least squares and is, in the presence of parameter heterogeneity, inferior to it. In the conclusion we present a unified method for analyzing time-series-cross-section data.

[1]  D. Freedman,et al.  Bootstrapping a Regression Equation: Some Empirical Results , 1984 .

[2]  James H. Lebovic.  Riding Waves or Making Waves? The Services and the U.S. Defense Budget, 1981–1993 , 1994, American Political Science Review.

[3]  J. Stock,et al.  Variable Trends in Economic Time Series , 1988 .

[4]  James C. Clingermayer,et al.  Disentangling Patterns of State Debt Financing , 1995, American Political Science Review.

[5]  H. White.  Asymptotic theory for econometricians , 1985 .

[6]  Micheal W. Giles,et al.  Racial Threat and Partisan Identification , 1994, American Political Science Review.

[7]  R. Parks,et al.  Efficient Estimation of a System of Regression Equations when Disturbances are Both Serially and Contemporaneously Correlated , 1967 .

[8]  A. Blais,et al.  Do Parties Make a Difference? Parties and the Size of Government in Liberal Democracies* , 1993 .

[9]  Brian M. Pollins.  Does Trade Still Follow the Flag? , 1989, American Political Science Review.

[10]  Stephen Nickell,et al.  Biases in Dynamic Models with Fixed Effects , 1981 .

[11]  H. White.  A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity , 1980 .

[12]  James A. Stimson.  Regression in Space and Time: A Statistical Essay , 1985 .

[13]  H. White,et al.  Some heteroskedasticity-consistent covariance matrix estimators with improved finite sample properties☆ , 1985 .

[14]  Michael S. Lewis-Beck,et al.  Comparative Democracy: The Economic Development Thesis , 1994, American Political Science Review.

[15]  David F. Hendry,et al.  Serial Correlation as a Convenient Simplification, not a Nuisance: A Comment on a Study of the Demand for Money by the Bank of England. , 1978 .

[16]  Nathaniel Beck,et al.  Comparing Dynamic Specifications: The Case of Presidential Approval , 1991, Political Analysis.

[17]  R. Engle.  Wald, likelihood ratio, and Lagrange multiplier tests in econometrics , 1984 .

[18]  A. Buse,et al.  Elements of econometrics , 1972 .

[19]  G. King,et al.  On Political Methodology , 1990, Political Analysis.

[20]  Cheng Hsiao,et al.  Analysis of Panel Data , 1987 .

[21]  Andrew Harvey,et al.  The econometric analysis of time series , 1991 .

[22]  Jonathan N. Katz,et al.  What To Do (and Not to Do) with Time-Series Cross-Section Data , 1995, American Political Science Review.