I am conducting a difference-in-difference regression where the probability of being unemployed is my dependent variable. I think I have interpreted the difference-in-difference parameter correctly in terms of percentage points (based on what existing papers I have studied have said), but I'm not sure how to interpret my control variables- percent or percentage points?

For example, if the education (years of schooling) variable results in a coefficient of -0.004, does this mean an additional year of education reduces the likelihood of being unemployed by 0.4% or 0.4% points? The same for race- if the coefficient for black people is 0.03 (relative to a base group of white people), does this mean that being black increases your likelihood of being unemployed by 3% or 3% points relative to white people?