im estimating a simultaneous probit model. Is this a correct way to do since there is a bidirectional causality between poor and informal employment - in literature
* Step 1: Estimation of Equation 1 (Probit)
global y1 hhinformal
global y2 poor
global x1 hdmale i.hhage i.hheduc hhmarried hdsize tot_informal urban
global z1 child_under6 m_hseduc (instruments)
global z2 large_firm (instruments)
* Reduced Form
probit $y1 $x1 $z1 // Equation 1
predict phat_informal
probit $y2 $x1 $z2 // Equation 2
predict phat_poor
* Structural Form
eprobit $y1 phat_poor $x1, endogenous($y2 = phat_informal $x1, probit)
* Step 1: Estimation of Equation 1 (Probit)
global y1 hhinformal
global y2 poor
global x1 hdmale i.hhage i.hheduc hhmarried hdsize tot_informal urban
global z1 child_under6 m_hseduc (instruments)
global z2 large_firm (instruments)
* Reduced Form
probit $y1 $x1 $z1 // Equation 1
predict phat_informal
probit $y2 $x1 $z2 // Equation 2
predict phat_poor
* Structural Form
eprobit $y1 phat_poor $x1, endogenous($y2 = phat_informal $x1, probit)
