Hi there,

I am using movestay to examine the impact of innovation (Innovation) on capital structure choice (CCC). The outcome for capital structure is measured by a binary variable (=1 for equity, =0 for debt), while the innovation outcome is a continuous variable. The code is as follows:

movestay (Innovation $controls), select (CCC=z1 z2) cluster (gvkey)

where $controls stand for control variables used in the innovation equation (including size, age, leverage, roa, and tobinq); z1 and z2 are instrumental variables used in the choice regression.

However, the regression results show that all the control variables used in the innovation equation have been automatically included in the choice equation. I wonder if it is possible to exclude some of them (e.g., age) from the choice model.

Thanks ahead for your attention and help.