I am running a regression of a binary variable (proficient in math or not) on education expenditure ($) and child gender. After running a probit model, I am calculating marginal effects:

Code:
margins, dydx(*) post
I obtain the following estimates:

Code:
education expenditure 0.000109
Code:
 male 0.0198
Could someone help me with the interpretation? According to me, the interpretation is:

1. A one dollar increase in education expenditure increases the probability of being proficient in math by 0.01 percentage points. Hence, a 100 dollar increase in expenditure increases the probability of proficiency by 1% point.

2. Boys are 1.9% points more likely to be proficient in math than girls, other things equal.

Please indicate if this is correct, especially the interpretation for education expenditure.

Thanks a lot!