To conclude, the earnings of married men and married women are determined in distinctive ways, with married men obtaining a net advantage in terms of the coefficients on the independent variables, even ignoring the intercept term.
This means that not only is there a large, unexplained, discriminatory element in the wage differential for married men and women but that the relevant variables affect earnings in different ways for each group.
The difference in the intercept term could represent discrimination, an unmeasured link between marital status and productivity, or differences in preferences or opportunity costs between sexes.

(from a statistical study published by Oxford and graciously linked in by linca)

Free at last! Free at last! Thank God Almighty, we are free at last! (Martin Luther King)