The CPANEL Procedure

The CORRELATED statement selects the subset of effects from the MODEL statement that are correlated with the individual (or cross-sectional) error term. This statement is valid only when you specify the AMACURDY or HTAYLOR model estimation method in the MODEL statement.

If you specify an effect in the CORRELATED statement but not in the MODEL statement, then the effect is added to the list of model effects and treated as correlated with the cross-sectional error term. If you have multiple MODEL statements, then each CORRELATED statement is associated with the MODEL statement that immediately precedes it.

The following statements fit a Hausman and Taylor (1981) model with model effects X1, X2, Z1, and Z2. The effects X2 and Z2 are correlated with the cross-sectional (firm-level) error term.

proc cpanel data = mycas.a;
   id firm year;
   model Y = X1 X2 Z1 Z2 / htaylor;
   correlated X2 Z2;
run;

You can also specify classification variables, interactions, and nested effects. For more information about constructing model effects, see the section Specification and Parameterization of Model Effects in Chapter 4, Shared Concepts.