Hello.
I'm using Stata 15 and ran the following cmp (https://econpapers.repec.org/softwar...de/s456882.htm) command to account for a self-selection bias since only firms that engage in "AVESIAE" are in my sample. Please note that I have an unbalance panel and all variables are continuous.
In addition, AVESIAE as well as SME are choice variables. I do have industry-level observations for both of these variables and used a control function approach to correct for endogeneity (Wooldridge 2015, http://jhr.uwpress.org/content/50/2/420.refs); e.g. for AVESIAE:
I'd like to include this control function in my cmp code. Is this possible. I've not found any reference to this.
Thank you for any assistance.
I'm using Stata 15 and ran the following cmp (https://econpapers.repec.org/softwar...de/s456882.htm) command to account for a self-selection bias since only firms that engage in "AVESIAE" are in my sample. Please note that I have an unbalance panel and all variables are continuous.
Code:
gen byte ind2 = cond(Select, cond(l.AVESIAE, $cmp_cont, $cmp_out), $cmp_out) cmp (SALESGR = c.l.AVESIAE##c.l.SME CV i.gvkey) (l.AVESIAE = CV i.gvkey), ind($cmp_cont ind2) vce(cluster gvkey)
Code:
xtreg l.AVESIAE l.AVEIndSIAE CV , fe vce(robust) predict CFSIAE, res
Thank you for any assistance.
