Dear Stata Expert,

I have a question related to interpretation of interaction terms in a Probit model.

In my analysis, I have typed the following command:

probit crime foreign econ_stat foreign#c.econ_stat i.country, cl(country)

My intent is to estimate the combined impact of being a foreign-born with a lowering of the economic status (for instance being materially deprived) on committing a crime.
The dependent variable crime is binary variable taking value 1 if the individual has committed a crime and 0 otherwise, foreign is a dummy variable taking value 1 if the individual is foreign-born, econ_stat is a level variable, the higher this variable, the worse the socio-economic status of the individual.

Say that the coefficient is -0.02, can I interpret it in the following way: "for a given socio-economic status, being foreign born reduces the possibility of committing a crime by 0.02 times or by 2 percentage points" (as if the probability is defined as a percentage from 0 to 100)? Alternatively, can I also conclude that "worsening the socio-economic status is associated with lower than average probability of committing a crime when the individual is foreign-born"?

I hope that my question is enough clear.

Many thanks.

Gianluca