In terms of the current pay gap, right now I think women earn on average about 87¢ for every dollar earned by men, and that's an hourly wage rate estimate.
In the article that's cited, “Why has the Wage Gap Narrowed ”, we looked at differences in education, job tenure, shifts in industry, shifts to higher-paying occupations, and occupational shifts of different age groups of workers. Some of those things account for about 60% of the reduction in the gap over time.
In another particular study with Michael Baker we calculated the adjusted pay gap over time. We saw that in the late 1980s the adjusted gap was much higher than the unadjusted gap, and by 2008 there was no difference between the adjusted gap and the unadjusted gap. That has to do with the fact that the differences in explainable characteristics between men and women are no longer explaining the gender wage gap by 2008, and there are more differences in how the labour market compensates men and women.
There are several interpretations along with this finding. First of all, men and women could be paid more equally in 2008 compared to 1988. We've seen a widespread convergence in unobservable characteristics between men and women. This would definitely be the case in work experience, which was not included in that particular study. We see that the attitudes towards women at work have also changed in that time. Those are some leading factors explaining why the gender wage gap has changed between 1988 and 2008.
Another potential factor is related to what we call a selection effect or a selection bias. Here we want to compare wages of men and women at two very different points in time, and the characteristics of people have been changing. One case in point here is that the employment rates of women have increased substantially between 1988 and 2008, so you can imagine a situation in which women in the late 1980s have above-average skills, and then there's a massive influx of women with average skills. This tends to decrease the wages of all women. Then when you're looking at the wage gap over time, it looks as if the wage gap didn't reduce as much as if you compared the exact same characteristics of workers.
When you control for the selection effect, the wage gap would be narrowed further by about 1.6 percentage points over that same period. If you do the same analysis on unobserved components of the wage gap, it could be as high as an additional 5%, so the wage gaps cited here could be a very conservative estimate about the reduction in the wage gap over time.