THE BOOK--Playing The Percentages In Baseball

Very good stuff.

I find it interesting that people think this matters, especially with such a small sample size. 

And, of course, it’s possible that there just happen to be more qualified men than women in the pool.  This is especially true if the highest-placed staff need to be the most experienced, since we can probably assume that there was a lot more gender bias than now 25 years ago when the experience was being accumulated.

Can you link to the press stories on the subject?  I’m curious to see what spin is being put on these numbers.

If there are more highly qualified men then is McCain intentionally hiring subpar staff members? That seems unlikely.

This was the article that started it all:
http://www.humanevents.com/article.php?id=28510

And Palin is quoting that article.

Right, but what I’m saying that if the point is to show that either candidate is discriminating, the raw numbers aren’t enough.  You have to show that the hiring ratio doesn’t match the candidate pool.

It is possible that McCain and Obama are not dipping in the same pool.  Obama may be focusing on a pool that, by coincidence, has a larger share of males (say Fortune 500 execs as an example, which may or may not apply here).  And McCain may be focusing on a pool that, by coincidence, has more women.

It’s also possible that even if they are dipping in the same pool, a disproportionate number of men want to be part of the Obama camp, and vice versa for McCain.

The point is is that it’s irresponsible to make any claim anything without a proper analysis.  The uncertainty level around the conclusion is simply too great to dismiss any reasonable retort.  What I’ve done here is just skimmed the surface.

Phil, I searched for “obama staffers pay equal” on google news and blog search and found a ton of stories.

This actually makes me feel sick.  The population is numerically illiterate enough to believe the claims at face value.

From a conservative blog:

“It seems that Obama has been shortchanging his female Senate staffers :

Obama’s 28 male staffers divided among themselves total payroll expenditures of $1,523,120. Thus, Obama’s average male employee earned $54,397.

Obama’s 30 female employees split $1,354,580 among themselves, or $45,152, on average.”

I honestly feel sick to my stomach because this could be the type of thing that could cause a massive swing in the female voters.

I guess I have to give credit to McCain for coming up with something that can be rebuked, for all intents and purposes.  What’s Obama going to do?  Stand up and teach the country about statistical significance?  Quite brilliant, actually.

Sigh…

1.  If Obama is arguing “because women are making only 77% of what men do,” something has to be done, then it’s perfectly reasonable to ask him why, if there is a similar gap on *his own staff*, why something doesn’t need to be done there too.

2.  Obama can rebut the argument by arguing that because his staff is so small, compared to the US as a whole, it’s not statistically significant, and it’s just coincidence.

3.  In that case, republicans can argue that the “something that must be done” might catch a lot of Obamas in its net if the solution is quotas.

Overall, I think pointing out the discrepancy on Obama’s staff is fair argument, but only as a rebuttal to Obama, who brought up pay equity in the first place.  If the goal is to make Obama look like a sexist, it fails.  But, of course, the accusation might stick anyway, as the previous poster (Sorry, I can’t see his/her name while I’m posting) just pointed out.

(Disclosure: I am strongly opposed to almost all proposals for Pay Equity laws, but nonpartisan on the US election.)

I agree strongly with #5 and #1. There is absolutely no way we can conclude with any certainty that either candidate is discriminating or practicing gender bias, or whatever you want to call it, based on this sketchy data.

I’ll also add that if one or the other candidate told me to my face, “Yes, I have this unconscious (or even conscious) bias in that I pay my men (or women) more than I pay my women (or men),” that would probably be at or near the bottom of the list of my concerns or my criteria for supporting him or her.

I also don’t understand the point in identifying possible gender biases in pay for one individual or company.  It is a free market.  You pay someone exactly the same as everyone else would pay them.  If women get paid less for equal pay by everyone, then unless you are trying to make a statement (which isn’t a bad idea), then you will end up paying women less too.

If a woman walks into your office to apply for a job and you don’t already have the pay established and you know that everyone else will pay her $50,000, then you pay her $50,000.  If a man walks into your office for the same job, while you might also like to pay him $50,000, if he can get $55,000 from everyone else, then you pay him the 55.  You don’t offer the woman 5 more thousand because you think it is unfair that everyone else will only pay her 50.  Business is business.  Sure, you can make the argument that you should, on principal, pay her the 55, especially if you are a politician trying to take a stand on gender and racial equality.

But, I don’t think it works that way.  You don’t make a conscious decision one way or another.  If women are getting paid less than men in the workplace in general, you will end up paying them less too, without even knowing it (you’ll look at their former salary, etc.).

Perhaps the only thing we should be asking or demanding of the candidates is to audit their own staff and see if in fact women are getting paid less than men for the same work, and to rectify the situation.  That would be reasonable.  I DON’T think it is reasonable to criticize them if it is true but they likely are not aware of it (which is likely the case).

Without reading the article, someone is getting desperate here.  What a thing to pick on, of all the things to look into…

#8 - Agree with everything you said except the last sentence.  I think it’s a brilliant thing for McCain to jump on, and it’s the type of thing that can spread like wildfire.  Can’t you just picture the mothers sitting around the playground in Ohio or Pennsylvania? 

“I can’t believe it - Obama seemed so fair, but did you hear that he pays women 20% less than men!”

Before we know it, there will be mass chain emails going around with these “facts.”

A good friend is in the internet marketing (affiliate links and such) business, and one of the wisest things he ever said was “never underestimate the power of a housewife with an email account.”

Phil: I think the 77% claim is based on “equal work”.  That is, men and women, given “similar” work has the women paid 23% less.

It’s the same kind of thing that I show in my article.  The researchers identify similar types of work, and then they see how much men and women get.  But, it took me an hour to do the analysis and I had to go through 60 people, with scant information.  Imagine going through tens of thousands of people over hundreds or thousands of industries.

I don’t know that they can necessarily capture “similar work”.

The 77% does NOT (unless I am mistaken) simply take total wages divided by total hours worked, and split by men and women.  That would be ridiculous to do, and no researcher would ever present anything to that effect.

I’m not nonpartisan and won’t claim to be (I’m going to hold my nose and vote for McCain, because he at least pretends to be a Republican from time to time). But here’s how I see it - if you are going to campaign upon the idea of pay/hiring equality for women, it probably makes sense to make sure that you’re doing that. If thee are legitimate reasons that Obama would pay his male staffers more on average, it should follow that people who aren’t Obama would also have legitimate reasons, and so maybe Obama should reconsider his position.

And if there are people out there who were going to vote for Obama but now aren’t, well, I seriously doubt that their reasons for voting for Obama were any more informed than their reasons for voting against him.

The “problem” with democracy is that you end up with people end up chosing their ballots based on all sorts of reasons, not all of which are good or even barely competant. This of course increases as voter turnout goes up. From what I’ve seen in casual news reports, we’re looking at massive voter turnout, so this election is likely going to be decided by people who know little that’s worth knowing. I’ve made my peace with this already, and it’s not like there’s strong consensus among the well-informed anyway.

Tango/10: Are you sure?  I’d bet you that the 77% is overall, without adjusting for what type of work.  It *might* adjust for education or age, *maybe*.

I’ve seen other studies that showed if you adjust, the number is in the 90s.

Can we find out?

I don’t know about the 77% claim, but from the article you linked to:

“The most recent statistics are for the half-year from October 1, 2007 to March 31, 2008. Excluding interns focuses attention on full-time personnel. For someone who worked only until, say, last February 29, extrapolating up to six months’ service simplifies this analysis. Doubling these half-year figures illustrates how a year’s worth of Senate employees’ paychecks should look.

Based on these calculations, Obama’s 28 male staffers divided among themselves total payroll expenditures of $1,523,120. Thus, Obama’s average male employee earned $54,397.

Obama’s 30 female employees split $1,354,580 among themselves, or $45,152, on average.”

So clearly here they are simply adding them, splitting them up, and dividing by the # of male/female employees.

And while he does qualify it later on with “the explanation is that men hold more high positions with Obama”, this quote is the one that’s making the rounds.

All the studies I’ve seen that make adjustments for age, education, hours of work, year of experience, etc. find that women earn about the same (or perhaps even slightly more) than men. Women simply quit careers more frequently and at a younger age (presumably to have children) thus missing out on peak earning years and having less experience. Women also tend to work fewer weekly hours (~10 hours less than men).

Tango/#10 - Actually, it seems that the 77% figure is totally ridiculous. I could be wrong, but I believe this oft-cited figure is based on the US Census figures for “Median Earnings of Full-Time, Year-Round Workers 15 Years and Older.”

Click on my name if you want to see the PDF (the relevant chart is on page 14 of the PDF)

http://www.womensmedia.com/new/Lips-Hilary-gender-wage-gap.shtml

Some people think that if women move into male-dominated occupations in larger numbers, the wage gap will close. However, there appears to be a gender-related wage gap in virtually every occupational category. In researching this issue at the Center for Gender Studies, we found only four occupational categories for which comparison data were available in which women earned even a little more than men: special education teachers, order clerks, electrical and electronic engineers, and miscellaneous food preparation occupations (Source: Bureau of Labor Statistics).

For example, among school principals, women have an average of 3 years longer as teachers than men do (Source: National Center for Education Statistics). So it is hard to argue that women’s lower earnings are simply a result of women putting in fewer hours per week, or even fewer years than men.

http://legalcareers.about.com/b/2008/09/27/legal-careers-are-lucrative-but-gender-pay-disparity-exists.htm

female paralegals and legal assistants earned a median of $42,600 which was 93.2 percent of the $45,700 median that men earned.

This group excludes lawyers and judges, where years of experience is obviously a factor.  Paralegals and legal assistants likely don’t have such external factors that would skew the results.  And in these cases, we still find a disparity, though not so large as to be too troubling (though a 3K difference is pretty big).

17/Tango: Maybe female paralegals go work for small firms that are more flexible in their hours, and those firms pay less.  Maybe male paralegals work more hours, because their sexist bosses think that men should be able to work harder without complaining.  Maybe ... well, you could probably think of lots more without trying too hard.

My point is not to say that bias in pay doesn’t exist, just to point out that it’s nearly impossible to know what’s causing it with the data given.

http://www.businessweek.com/careers/workingparents/blog/archives/2007/04/working_women_a.html

Particularly unexpected was the gap one year out of college, when men and women should be most likely to show pay equity. “If a woman and a man make the same choices, will they receive the same pay?” the study says. “The answer is no. These unexplained gaps are evidence of discrimination, which remains a serious problem for women in the work force.”

Don’t assume the gap exists because women choose lower-paying professions, either—the study found a pay gap in the same fields. In education, women earn 95% as much as their male colleagues. In math, women earn 76% as much as men. This despite the fact, as the study lays out, women outperformed men academically, and their grade point averages were higher in every college major.

However, the researchers do suggest that women themselves may be partly to blame for the gap: “Women expect less and negotiate less pay for themselves than do men.”

Maybe someone can go through the study:
http://www.aauw.org/research/behindPayGap.pdf

***

http://www.denverpost.com/news/ci_5735341

She cited a 2000 study by the University of Denver for the Colorado Women’s Bar Association that found female lawyers with 10 to 20 years of experience earned 74 percent of what men with comparable career paths and experience earned.

I think we can agree that there is gender-bias.  Now, whether it is an explicit gender-bias

“she’s a girl, so I’ll offer her 10% less”

or an implicit gender-bias

“I’ll offer everyone the same salary, and I’ll allow them to ask for 20% more; as it turns out, 95% of those who ask for more are men”

, I agree that we don’t know. 

We can factor out age, experience, and education, and what we are left with is women earning some 20% less than men. 

Gordie Howe, the best player in the NHL for over a decade, was given a blank check.  And every year, he would just put in a modest raise, with the owner assuring him that he was the highest paid player.  In the middle of his career, he was told by a teammate of his that Howe was earning less than he was.

Howe, basically, acted like “a girl” (using the term as it’s being used in all these articles I’m reading), in accepting that he was being treated fairly.  Most men would not act like Gordie Howe.  Most women probably would.

Mikal L. Sutton-Vereen is a woman.

19:

Different jobs pay different salaries, even if they are in “math” or “education.” Maybe women are choosing jobs with lower salaries for good reason.

I once took a job in “computers” with a lower salary because I liked the job better.  Didn’t mean they were underpaying me, it was just a different, lower-pressure job.

Even the same jobs pay differently in different companies ... Richard Thaler wrote one of his “anomalies” articles on how secretaries (and many other jobs) were paid more at big, profitable companies like Exxon.  He didn’t really know why.

So maybe men like to work for bigger companies, and women smaller ones?  Could be.  Could be almost anything.

I’m pulling for Obama, but I have little to no doubt that if the situation were reversed, his supporters would be making a huge deal out of it. I’ve seen similar evidence used in a similar way by people my much more “rigorous” contexts than Human Events.

When you have a 20% gap (say 40K and 50K), it would almost mean that almost 100% of the men are going toward one industry and 100% of the women are going toward another industry, under otherwise similar jobs.

The reality is that any industry-size choices would affect say 20% of the men one way and 20% of the women the other way.  So, these kinds of things would make a small percentage difference.

There are clearly many issues and potential issues going on here and one should not lump them all together.

One issue seems to be whether women themselves are more timid in asking for raises or negotiating their initial salaries or generally have a “lower expectation” of earnings.  If that is true, that is an issue for the women themselves to become more empowered. I don’t think you can or should “blame” the employers for that, although I suppose it wouldn’t hurt to educate them as to it.

Another issue may be that employers explicitly (not necessarily consciously though) discriminate against women when they hand out salaries.  If that is the case, it is not a good thing for women and probably not a good thing for society, but I am not sure what anyone is supposed to do about it.

Interestingly, in a free market, I am not sure that #2 can exist without number 1.  If women “demanded” equal pay for equal work, employers who refused to do so and continued to discriminate would simply end up with lesser quality of work from their men as opposed to from their women who make the same money.  As in “game theory” eventually they should realize that and end up paying everyone exactly the same amount of money.  If that does not happen, it almost has to be because women are not letting it happen, or perhaps that there is something intrinsic in men that allows them to be more aggressive in negotiating contracts and asking for raises.

In fact, that last point is probably the case. No matter what, because of inherent differences in the “aggressiveness” of the sexes, men will probably always make more money than women for the same quality and quantity of work.  Only if we become a truly androgenous society, which I don’t think will ever happen, will that change I would think.

This may explain Scott Boras.

Another factor I haven’t seen mentioned is the true talent level. We all know that given a large enough sample, or the entir population, this should even out, and I am very comfortable assuming that men have the same talent level as women in most jobs, especially ones not involving physical strength.

I work at a mapping company. We look at stereo photography, classify what we see, and digitize it into a graphics database. We have 4 women and 13 men. Maybe it’s not a job that interests women. I do not know how much anybody except myself makes. I am very familiar with their level of quality. Compared to their co-workers (because I do not know other companies very well), I would give one of the women A-/B+, while the other three are C-/D. I am cinfident that these three women don’t get paid as much as the rest of use, commensurate with the quality of their work.

I do not think this is anything more than sample size. I’m sure if you went to all companies that do what we do, the mean talent of the women will be the same as the mean talent of the men. Just not so with the four we have.

However, what happens when the government comes in and demands that we explain why the women aren’t getting paid as much as the men? You would end up taking the population means and then enforcing compiance to them on a small unrepresentative sample.

Jamia Jowers is a man.

Some of the lower paid employees do not appear to be salaried - e.g. Christopher Hickling: paid $376 over 52 days. Overtime pay or bonuses may obscure extrapolations of rates of pay.

I don’t know what the breakdown is in the relevant markets for Senatorial staff (Washington DC and Chicago or Phoenix), but women probably comprise somewhat less than 50% of the overall national labor force, so prima facie we shouldn’t expect a women to achieve a full 50/50 split in “positions of power”.

Thanks Joe.

Re #8

MGL, isn’t the proper hiring strategy in the situation you describe to simply not hire men? If you can get a man for $55K/year and a woman for $50k and they have the same “true talent” the choice is pretty easy if you are aware of it.

However I personally don’t think most potential employers would know. My guess is that much more employers resemble the Cubs than the A’s.

I do however agree that almost all of the “true” gender bias is unconcious rather than concious. (By “true” bias I mean once you factor in that women and men have different types of jobs, work different amount of hours and years etc etc.)

Hey, man, the Cubs just made the playoffs two years running - and it wasn’t entirely an accident! Can’t we use, I dunno, Orioles jokes in this context for at least a while?

I would also guess that men are more likely to leave a job, so paying them extra may be incentive to not leave their job.

MGL, isn’t the proper hiring strategy in the situation you describe to simply not hire men? If you can get a man for $55K/year and a woman for $50k and they have the same “true talent” the choice is pretty easy if you are aware of it.

Well, yes, of course, and then men would eventually lower their asking price, and women would demand more because they would “realize” that they were in heavy demand, until an equilibrium were reached.

In order for this pay dynamic to remain out of an equilibrium state, women must be complicit in the pay disparity (or some other force is acting on the market - for example, let’s say that there were a law that you had to pass women less for equal work).