I can't say I appreciate my position in this discussion being referred to as a "lie", "stupid", etcetera.
However, putting that stuff aside -- numbers don't lie, and the Bureau of Labor Statistics data provided in my earlier post speaks for itself (of course, Google can help you find plenty of data from both sides of the argument). Using simple arithmetic, any person can analyze this BLS data forwards and backwards and, in my opinion, arrive at only one "rational" conclusion: women are discriminated against when it comes to compensation. The disparities are more than merely "detectable differences". The disparites are significant and they occur between median pay for women working full-time vs. median pay for men working full-time, in occupation after occupation (these are as apples to apples as such comparisons can get). Such disparities, also consistently applied from industry to industry, age group to age group, racial group to racial group (e.g. white women make less than white men, black women make less than black men), marital status to marital status, education level to education level can't be explained away by time in the field, hours worked, non-salary benefits, attitude, or anything else. (The BLS document also indicated that hourly female workers are paid lower hourly wages than hourly male workers).
As I wrote in a previous post, I also have personal experience with this issue, and I can't know why my wife's employer paid the men more than all the full-time women with more education, more experience and more hours worked. I can only speculate.
With regard to the question of why any "rational" employer would ever hire a man if they knew that they could get a woman to do the job cheaper -- this question assumes that employers who choose to pay women less for the same position, same experience, same performance are rational. I'd argue -- they're not. However, since this question seems to be an obstacle for some, I will speculate:
1. Hiring on the basis of sex is illegal (and such illegal activiity is easier to spot than pay discrimination), and/or
2. generally, those who make the hiring decisions believe that men provide a better return on their payroll investment than women, even at higher salaries (i.e. they're sexists), and/or
3. there aren't enough women in the job market to fill all vacancies, and/or
Again -- just thinking out loud. I would just reiterate, that as the man who made millions off of the Pet Rock demonstrated, people don't make rational economic decisions in the way that Thomas Sowell would have us believe. Personal and subjective emotions, opinions, bias and judgements (often poor) play a significant role.
In his article, "Morality vs. Sanctimoniousness" Thomas Sowell makes some of the same straw man arguments that others have made here: comparing the compensation of part-time female workers to full-time male workers; comparing female workers of one occupation with male workers in a different occupation. In addition, he accused those who see the plain truth behind statistics such as those provided by the U.S. Department of Labor as being "intellectuals" (i.e. those who disagree with Thomas Sowell), "sanctimonious" (which, by Sowell's own definition, is sanctimonious on his part), being irrational when referring to what our "imagination reads into the statistics" and having a "greed for power".
Sowell's article did not take on the issue at hand: all else being equal, are men compensated more than women? Instead, among other things, he took a "kill the messenger" approach that, I believe, should not pursuade.
(One more thing -- On the "mutual agreement" argument, when negotiating with a job applicant, the employer has complete information regarding compensation and benefits provide within his department or company -- the job applicant does not. In my wife's case, it was a fluke that she learned the truth. Few individuals know how their compensation compares with that of their colleagues. Pay disparities among employees are the reason that employers have policies prohibiting employees from telling each other how much they make. If employers were the meritocracies they want us to believe, such confidentiality would be unnecessary.)