Tests for the presence of cross-sectional or time random components in a panel equation. estimated using pooled least squares.
Computes the conventional LM (Breusch-Pagan, 1980, uniformly most powerful LM (Honda, 1985), standardized Honda (Moulton and Randolph, 1989; Baltagi, Chang, and Li, 1998), locally mean most powerful (LMMP) (King and Wu, 1997), Standardized King-Wu, and Gourieroux, Holly, and Monfort (1982) test statistics.
Note that the equation must be estimated with pooled least squares for this test to be applied.
Syntax
equation_name.rcomptest
Options
G
Examples
equation eq1.ls @log(gsp) c @log(p_cap) @log(pc) @log(emp) unemp
eq1.rcomptest
will estimate a panel model using pooled least squares and will compute and display the panel random effects test results.
Cross-references