I am struggling with putting a constraint in my biprobit model. I am not even sure if the constraint I am thinking about is even possible. It is a panel data setup. Here is the situation:

y1 = technology adoption (coded 0 for all months when adoption did not happen, coded 1 when adoption takes place)
y2 = bankruptcy (oded 0 for all months when bankruptcy did not happen, coded 1 when bankruptcy takes place)

Two biprobit equations: y1 = x1 x2, y2 = y1 x1 x2

all four combinations (0,0) (0,1) (1,0) (1,1) are possible.

I want constraint for first equation (adoption equation), where I want data only still y1 becomes 1 (like a hazard model situation). All the firms for which adoption happens, I do not want data after first '1'. However, for the second model (bankruptcy) I want all data. To describe it better, I want yellow highlighted data for the first equation, but I need full for the second one. Can someone please help?

Array