The Bureau of Labor Statistics looked at weekly wages instead of annual salaries, and determined that women earn 81 percent what men earn. That gap narrows to 86 cents on the dollar when hourly wages are examined.
And hourly wages among those with college degrees are at parity: The BLS figures show there is no gap between men and women.
So I looked up what the BLS had to say:
Highlights of Women’s Earnings in 2010 - U.S. Department of Labor, U.S. Bureau of Labor Statistics - July 2011
Women earned 81 percent of the median weekly earnings of their male counterparts [...] Sixty-two percent of women and 56 percent of men employed in wage and salary jobs were paid by the hour in 2010. Women who were paid hourly rates had median hourly earnings of $11.83, 86 percent of the median for men paid by the hour ($13.76). (See tables 9, 10, and 18–21.)
That is where the first 2 are from, even though I am not sure if "people who were paid hourly rates" = "when hourly wages are examined". The last fact, I could not find in the statistics. Table 9 tells us that the gap by people who were paid by the hour for "Some college or associate's degree" is 83.4% while for "Bachelor's degree and higher" it is 96.9% which also isn't exactly parity but close.
These other tidbits were also interesting:
Earnings differences between women and men were widest for Whites and for Asians. White women earned 81 percent as much as their male counterparts in 2010, while Asian women earned 83 percent as much as their male counterparts. By comparison, Hispanic women had earnings that were 91 percent of those of their male counterparts, while Black women earned 94 percent as much as Black men. (See tables 1 and 14.)
Among younger workers, the earnings differences between women and men were not as great. Women earned 91 percent of what men earned among workers 25 to 34 years old and 95 percent as much as men among 16- to 24-year-olds. (See table 1.)
Among full-time workers (that is, those working 35 hours or more per week in their sole or principal job), men were more likely than women to have a longer workweek. In 2010, 25 percent of men in full-time jobs worked 41 or more hours per week, while 14 percent of females in full-time jobs worked the same number of hours. Women were more likely than men to work 35 to 39 hours per week: 13 percent as opposed to 5 percent. A large majority of both male and female full-time workers had a 40-hour workweek; among these workers, women earned 87 percent as much as men earned. (See table 5.)
The solution to figure out the wage gap usually is to keep in mind that the main reason is often women, going out of the work force due to childbirth / taking care of family obligation. Which of course correlates with younger workers who have a lower gap and men that work more hours to support their family. I looked into the data and tried to figure out if marriage rate and race could explain the racial wage gap (the idea is that if a fewer percentage of black women are married than white women this might explain the gap) but this didn't really work out.
And finally a summary of two tables in the BLS report of items where the gap is lower than 10%
Table 1. Median usual weekly earnings of full-time wage and salary workers, by selected characteristics, 2010 annual averages - Women's earnings as percent of men's
16 to 19 years 94.6 20 to 24 years 93.8 25 to 34 years 90.8 Black or African American 92.1 Hispanic or Latino ethnicity 90.7 Never married 97.2
Table 9. Median hourly earnings of wage and salary workers paid hourly rates, by selected characteristics, 2010 annual averages - Women's earnings as percent of men's
16 to 19 years 97.8 20 to 24 years 91.7 25 to 34 years 91.9 65 years and older 93.8 Black or African American 92.1 Never married 95.1 Bachelor's degree and higher 96.9