The empirical analysis is based on a linear cross-sectional time-series regression model with panel-corrected standard errors
. Measures of entrepreneurship are taken from the Global Entrepreneur-ship Monitor and the World Bank Group's Entrepreneurship Survey, while the set of trade indicators are taken from the World Development Indicators, IMF World Economic Outlook, UN Conference on Trade and Development Statistics, and the Fraser Institute of Economic Freedom of the World.
Monte Carlo analysis shows that these "panel-corrected standard errors" perform well.
We therefore propose to retain OLS parameter estimates but re@ place the OLS standard errors with panel-corrected standard errors. Monte Carlo analysis shows that these new estimates of sampling variability are very accurate, even in the presence of complicated panel error structures.
We argue that this method, which combines ordinary least squares parameter estimates with panel-corrected standard errors, should perform well.
In addition, the Monte Carlo evidence shows that panel-corrected standard errors perform extremely well, even in the presence of complicated panel error structures.
Any serial correlation of the errors must be eliminated before the panel-corrected standard errors are calculated.
Expression 5, however, can still be used, in combination with that panel structure of the errors, to provide accurate, panel-corrected standard errors (PCSEs).
The Accuracy of Panel-corrected Standard Errors. OLS standard errors are accurate in the presence of either panel heteroscedasticity or contemporaneous correlation of the errors if the terms in the error covariance matrix, [omega], are not related to the squares and cross products of the independent variables.
Panel-corrected standard errors performed excellently in these experiments.
Panel-corrected standard errors are, as predicted, roughly three times the standard errors obtained by Hicks and Swank.(31) The Hicks and Swank data, when estimated with corrected standard errors, are not consistent with many of their conclusions.