• Welcome to the new Internet Infidels Discussion Board, formerly Talk Freethought.

The gender pay gap in Biden's White House

https://www.aei.org/carpe-diem/base...nder-pay-gap-at-the-biden-harris-white-house/


What are the cultural problems in Biden's White House that has caused women staffers to dominate? Or, is Biden's White House discriminating against men in hiring?



Why does Biden's White House pay women less than men? I think it's because society hates women but I was led to understand that Democrats hate women slightly less than wider society.
And the report from that rightwing thinktank confirms your observation since the gap is only 1% based on average salaries.

BTW, median salaries are better measures when there are lots of observations.
The median is a dumb value, in the sense that there is no processing to obtain it. It is merely the middle value.

[TABLE="width: 128"]
[TR]
[TD="align: right"]140[/TD]
[TD][/TD]
[TD="align: right"]140[/TD]
[/TR]
[TR]
[TD="align: right"]120[/TD]
[TD][/TD]
[TD="align: right"]120[/TD]
[/TR]
[TR]
[TD="align: right"]100[/TD]
[TD][/TD]
[TD="align: right"]100[/TD]
[/TR]
[TR]
[TD="align: right"]100[/TD]
[TD] Median[/TD]
[TD="align: right"]80[/TD]
[/TR]
[TR]
[TD="align: right"]80[/TD]
[TD][/TD]
[TD="align: right"]80[/TD]
[/TR]
[TR]
[TD="align: right"]60[/TD]
[TD][/TD]
[TD="align: right"]60[/TD]
[/TR]
[TR]
[TD="align: right"]40[/TD]
[TD][/TD]
[TD="align: right"]40[/TD]
[/TR]
[TR]
[TD="align: right"]91.4[/TD]
[TD] Average[/TD]
[TD="align: right"]88.6[/TD]
[/TR]
[/TABLE]


Using the median here is just brain fucking the numbers to get an apparent disparity that does not exist.

ETA: The numbers above are an example, not the real numbers, which should be obvious.
 
https://www.aei.org/carpe-diem/base...nder-pay-gap-at-the-biden-harris-white-house/


What are the cultural problems in Biden's White House that has caused women staffers to dominate? Or, is Biden's White House discriminating against men in hiring?



Why does Biden's White House pay women less than men? I think it's because society hates women but I was led to understand that Democrats hate women slightly less than wider society.
And the report from that rightwing thinktank confirms your observation since the gap is only 1% based on average salaries.

BTW, median salaries are better measures when there are lots of observations.
The median is a dumb value, in the sense that there is no processing to obtain it. It is merely the middle value.

[TABLE="width: 128"]
[TR]
[TD="align: right"]140[/TD]
[TD][/TD]
[TD="align: right"]140[/TD]
[/TR]
[TR]
[TD="align: right"]120[/TD]
[TD][/TD]
[TD="align: right"]120[/TD]
[/TR]
[TR]
[TD="align: right"]100[/TD]
[TD][/TD]
[TD="align: right"]100[/TD]
[/TR]
[TR]
[TD="align: right"]100[/TD]
[TD] Median[/TD]
[TD="align: right"]80[/TD]
[/TR]
[TR]
[TD="align: right"]80[/TD]
[TD][/TD]
[TD="align: right"]80[/TD]
[/TR]
[TR]
[TD="align: right"]60[/TD]
[TD][/TD]
[TD="align: right"]60[/TD]
[/TR]
[TR]
[TD="align: right"]40[/TD]
[TD][/TD]
[TD="align: right"]40[/TD]
[/TR]
[TR]
[TD="align: right"]91.4[/TD]
[TD] Average[/TD]
[TD="align: right"]88.6[/TD]
[/TR]
[/TABLE]


Using the median here is just brain fucking the numbers to get an apparent disparity that does not exist.

Unless there is a disparity. Not one caused by Biden Administration hypocrisy, or anti-female bigotry, or anything like that.

Rather a pay disparity resulting from different demographic groups(in this case men and women) choosing freely but differently. Both getting what they want, but wanting different things.

The OP made a dreadful hash of explaining that, but still.
Tom
 
The median is a dumb value, in the sense that there is no processing to obtain it. It is merely the middle value.

[TABLE="width: 128"]
[TR]
[TD="align: right"]140[/TD]
[TD][/TD]
[TD="align: right"]140[/TD]
[/TR]
[TR]
[TD="align: right"]120[/TD]
[TD][/TD]
[TD="align: right"]120[/TD]
[/TR]
[TR]
[TD="align: right"]100[/TD]
[TD][/TD]
[TD="align: right"]100[/TD]
[/TR]
[TR]
[TD="align: right"]100[/TD]
[TD] Median[/TD]
[TD="align: right"]80[/TD]
[/TR]
[TR]
[TD="align: right"]80[/TD]
[TD][/TD]
[TD="align: right"]80[/TD]
[/TR]
[TR]
[TD="align: right"]60[/TD]
[TD][/TD]
[TD="align: right"]60[/TD]
[/TR]
[TR]
[TD="align: right"]40[/TD]
[TD][/TD]
[TD="align: right"]40[/TD]
[/TR]
[TR]
[TD="align: right"]91.4[/TD]
[TD] Average[/TD]
[TD="align: right"]88.6[/TD]
[/TR]
[/TABLE]


Using the median here is just brain fucking the numbers to get an apparent disparity that does not exist.

Unless there is a disparity. Not one caused by Biden Administration hypocrisy, or anti-female bigotry, or anything like that.

Rather a pay disparity...
No pay disparity has been demonstrated to exist in the OP or the AEI piece.
 
The median is a dumb value, in the sense that there is no processing to obtain it. It is merely the middle value.

[TABLE="width: 128"]
[TR]
[TD="align: right"]140[/TD]
[TD][/TD]
[TD="align: right"]140[/TD]
[/TR]
[TR]
[TD="align: right"]120[/TD]
[TD][/TD]
[TD="align: right"]120[/TD]
[/TR]
[TR]
[TD="align: right"]100[/TD]
[TD][/TD]
[TD="align: right"]100[/TD]
[/TR]
[TR]
[TD="align: right"]100[/TD]
[TD] Median[/TD]
[TD="align: right"]80[/TD]
[/TR]
[TR]
[TD="align: right"]80[/TD]
[TD][/TD]
[TD="align: right"]80[/TD]
[/TR]
[TR]
[TD="align: right"]60[/TD]
[TD][/TD]
[TD="align: right"]60[/TD]
[/TR]
[TR]
[TD="align: right"]40[/TD]
[TD][/TD]
[TD="align: right"]40[/TD]
[/TR]
[TR]
[TD="align: right"]91.4[/TD]
[TD] Average[/TD]
[TD="align: right"]88.6[/TD]
[/TR]
[/TABLE]


Using the median here is just brain fucking the numbers to get an apparent disparity that does not exist.

Unless there is a disparity. Not one caused by Biden Administration hypocrisy, or anti-female bigotry, or anything like that.

Rather a pay disparity...
No pay disparity has been demonstrated to exist in the OP or the AEI piece.

I can understand why you edited my post as you did.
Tom
 
Swammerdami’s illuminating analysis of the data illustrates the problem of drawing a conclusion about the distribution if a population from a single measurement. It makes the accusation of hypocrisy on the part of the Biden administration on this issue hard to maintain by anyone with intellectual integrity. It makes it painfully clear that Mr. Perry is an ideological hack who gives support to the old adage “Figures don’t lie, but liars will figure.”

I find it astonishing (except that I don't) that you can make this accusation. The calculation of median differences is the established method of talking about the gender pay gap--there is nothing dishonest about it and using it does not make Perry an 'ideological hack'. Indeed, if Perry had done something different and bizarre--like talked about the 55th percentile, I'd have good reason to think the numbers suspect.
 
You are simply wrong about what the point of the thread was. I'm in a good position to know because I wrote the OP. The point was to highlight the idiotic wrongness of the gender pay gap narrative.

Ya know @Metaphor; , here's a problem.
I agree with you on this part. I agree that to an enormous extent the "pay gap" is now mostly about women choosing their own priorities like men do. But they tend to prioritize differently from men. They tend to prioritize flexibility, secure working environment, transportable skills and other things more than men. The tradeoff is lower income. The wage gap is primarily the collective choices women make compared to the ones men make.

But that wasn't at all clear in the OP. You knew what the point was, but even I(who agrees with you) didn't see anything like that. Given your OP history, like Cornell and Vermont and such, you kinda need to be very clear.

Tom

I thought I was notorious enough on the board that it would be clear, but I can see it was not.
 
Swammerdami’s illuminating analysis of the data illustrates the problem of drawing a conclusion about the distribution if a population from a single measurement. It makes the accusation of hypocrisy on the part of the Biden administration on this issue hard to maintain by anyone with intellectual integrity. It makes it painfully clear that Mr. Perry is an ideological hack who gives support to the old adage “Figures don’t lie, but liars will figure.”

I find it astonishing (except that I don't) that you can make this accusation. The calculation of median differences is the established method of talking about the gender pay gap--there is nothing dishonest about it and using it does not make Perry an 'ideological hack'. Indeed, if Perry had done something different and bizarre--like talked about the 55th percentile, I'd have good reason to think the numbers suspect.

Did you not understand posts #2 and #9?
 
Ok, so now you're saying there is no meaningful gender pay gap, and if there is, it's Biden's fault.

No, I'm not saying that. I'm saying Biden perpetuates the false and harmful narrative around the gender pay gap.

While the other side of your mouth laments the horrible inequity wrought upon Whitehouse employees...
Oooookay, Meta.

The "other side" of my mouth was uttering what I thought to be screamingly obvious sarcasm. I do not for a second believe that, for example, men stayed away from Biden's White House because Biden had created a toxic atmosphere for them.


Swammerdami's bizarre analysis succeeds in illustrating that the distributions of salaries in the White House are not normal or approximately normal, and therefore a comparison of means rather than medians is even less appropriate.
 
The median is a measure of central tendency. There is no one measure of central tendency. Reliable and fair analyses of distributions do not rely on one measure of central tendency or variation for that matter. When the data is clustered around a few values, looking at a measure of central tendency has little useful meaning. The notion that the any particular measure is "established" and is therefore automatically the best or most reliable measure is an appeal to authority because it ignores that possibility of special circumstances.

Mr. Perry's analysis is deeply flawed and therefore misleading based on its methodology. Not only does it ignore the clustering problem, but it does not even bother to compare similar jobs with each other because the equal pay argument is equal pay for equal work not equal pay for work.

But, I am no longer surprised by the sloppiness, ignorance and possible intellectual dishonesty of analysts such as Mr. Perry or how tightly their dupes hang on to the dishonest conclusions of such "analysts".
 
I've never seen someone use the 45%ile for anything in any context. But in terms of the gender pay gap, the median is used.
...
The median is a measure of centre, and it is special. It is the best representation of the typical earnings. I mean, why do you think the gender pay gap uses measures of centre and not the 90th%ile?



I think you need to make a case that these other percentiles make any sense at all to use.

Good! We have reduced the question to a straightforward one involving only intuition and grasp of statistics and mathematics. (Unless your claim is that gender pay gap is a special case where one must first discard all one's knowledge of relevant math and statistics.)

Swammi said:
The Board has several posters with excellent grasp and intuition of mathematics. Bomb#20 comes to mind (not to diminish others among the board's mathematicians).

Will you nominate Bomb#20 or some other mathematician to arbitrate our dispute (which now seems to be about statistics, not politics)?
This was off-topic. Still I'm a relative newcomer and am curious who the Board's best systems analysts are. Messrs Ipetrich and Bomb are my nominees. Who else? I nominate Mr. Bomb because I've seen a number of his correct analyses and don't recall an incorrect one. (I think he failed to admit that his claim about pi assumed what it was trying to prove, to wit that pi is a Normal Number, but that error, if any, is irrelevant to any practical engineering.)

Swammi said:
I want to ask the nominee whether, in the presented context, the 50th percentile value has some intrinsic mathematical importance that would not be possessed by, say, the average of the 45th and 55th percentile values.

OK?

Is gambling allowed on this Board?
RUDE! Knock it off.

Still I am proud that my brief effort with the data found a fatal (or at least fatalish) flaw in Professor Perry's argument.[
Start another thread for details, e.g. raw data./I] Unless of course, Metaphor is right, and we should ignore the equalities across a very broad range except for a very thin slab near 50% ! (Like the Butterfly ballots almost 21 years ago, Serendipity worked against the truth-tellers!)


And this is sincere. You will earn much respect, Mr. Metaphor, if you can study the posted table and admit that,
Yes, the insistence solely on 50% — single parameter no less — gives an exaggerated look at the more general statistic — perhaps a smoothed median! BTW I myself have witnessed successful solutions and for a strong non-trivial objective like this at least two statistical parameters would be used.

. I wonder if some staffer charged with assigning ranks and salaries, tried to hit a jackpot! Perry's Bulge at 50% is rather weird. Why not the bulge at 56% instead? Or whatever? Is it an anomaly of the algorithm used in hiring while meeting a pay equality objective. OTOH, that the data has this specific peculiar shape MIGHT, just might be evidence of sabotage!

Or, sincerely, — maybe you, Metaphor, have good math intuition — forget our dispute and study the table I showed without regard to its application. Can you understand the point I'm making about the "bulge" near 50% being like happenstance? (I am rooting for you!)


I don't have time to fully respond to this now, but I can say the following

* the idea that measuring a median gap and not reporting other percentiles makes Perry's analysis a hack job is false and bizarre. Every analysis of the gender pay gap I have encountered talks about either means or medians (medians are a better comparison for a variable like income, but it is less common to get a test of statistical significance from). Indeed, Perry made similar analyses for the Trump and Obama White House (using the median) so it would make his analysis less honest had he deviated from his method.

* Differences (or lack of them) at other percentiles cannot and does not change the fact that pay gaps are reported as a function of the median or mean (again, noting that the mean is a more problematic measure of centre than the median for data like income)

* Although your analysis shows the two distributions are not normal (that was known anyway, since they had different medians but nearly the same mean), the same is probably true of the gender gap data in wider contexts

* Even if it would be fair to conclude that there is "no difference" in pay by gender in Biden's White House (and I don't think it is fair to conclude that) - I am not criticising Biden's White House for a gender pay gap, but the promulgating of a false narrative around it. If it was fair to conclude that there was no evidence of a pay gap based on your analysis, it would lessen the charge of hypocrisy--but perhaps raise the charge of conscious wage manipulation.
 
Mr. Perry's analysis is deeply flawed and therefore misleading based on its methodology. Not only does it ignore the clustering problem, but it does not even bother to compare similar jobs with each other because the equal pay argument is equal pay for equal work not equal pay for work.

That is not the case. If it were, the WGEA and other institutions would not report a headline figure that did not control for equal pay for equal work--just a raw difference in overall (full time) work. Yet they do report unadjusted figures, year after year, and they measure progress by reporting shrinking of the unadjusted figures.
 
Swammerdami's bizarre analysis succeeds in illustrating that the distributions of salaries in the White House are not normal or approximately normal, and therefore a comparison of means rather than medians is even less appropriate.

Actually it succeeds in DEMONSTRATING that the median is not representative in this case. And you know it. There is a GPG right at the median, and virtually nowhere else.
 
The median is a dumb value, in the sense that there is no processing to obtain it. It is merely the middle value.

Oy gevalt. Are you suggesting that ordering the values is not a process?

Using the median here is just brain fucking the numbers to get an apparent disparity that does not exist.

No numbers have been 'brain fucked'.
 
Swammerdami’s illuminating analysis of the data illustrates the problem of drawing a conclusion about the distribution if a population from a single measurement. It makes the accusation of hypocrisy on the part of the Biden administration on this issue hard to maintain by anyone with intellectual integrity. It makes it painfully clear that Mr. Perry is an ideological hack who gives support to the old adage “Figures don’t lie, but liars will figure.”

I find it astonishing (except that I don't) that you can make this accusation. The calculation of median differences is the established method of talking about the gender pay gap--there is nothing dishonest about it and using it does not make Perry an 'ideological hack'. Indeed, if Perry had done something different and bizarre--like talked about the 55th percentile, I'd have good reason to think the numbers suspect.

Did you not understand posts #2 and #9?

What is there to understand?

If the median is better on 'larger' samples and therefore not suitable for this analysis (a claim laughing dog makes but does not evidence), then so much the more is examining the data at every possible percentile inappropriate.

Indeed, since most workplaces have fewer than 560 employees, any examination of percentiles might show a gender pay gap (in either direction) at some percentiles and no gap at other percentiles. Are you suggesting that there is no way to demonstrate a gender pay gap in these other workplaces if it is not apparent at every percentile along the distribution?
 
Women DO tend to value flexibilty, etc. At least women with children do, or those who are planning to have children. Why is this less a concern for men?

Why shouldn't it be? Is there a correct way to be a man?

Traditionally, women were compensated at lower rates than men were because they were women. Likewise, professions that are heavily female staffed tend to be compensated at lower rates, despite requiring as much or more education, skill, etc. WHY?

Because there is a labour market, and employers will pay the lowest price a labour market will bear for their employees. If men do not want to go into those lower-paid professions because they value money more than women, so what? It means women will stay in lower-paying professions and men will pursue alternatives.

There is a current nursing shortage--has been for years. WHY? This is a case where the obvious market solutions (higher compensation, more flexibility, etc). has not been instituted to encourage more nurses although the nursing shortage has been ongoing for years. WHY?.

Oh, I got this one. It's because society hates women. It's a shame only that male nurses have to suffer alongside female ones, but the patriarchy has to be upheld.

EDIT: I have some counter questions

Women earn less money when they become mothers, and men earn more money when they become fathers. Does this seem like a problem to be rectified to you? IF so, why?
 
Code:
    Percentile    M       F      F+U
        90%    $155k   $155k   $155k
        80%    $130k   $130k   $130k
        70%    $110k   $110k   $110k
        60%    $104k   $100k   $100k
        55%    $100k   $100k   $100k
        50%    $100k    $80k    $80k
        45%     $80k    $80k    $80k
        40%     $80k    $80k    $80k
        30%     $77k    $62k    $62k
        20%     $62k    $62k    $62k
        10%     $58k    $48k    $48k
As you see, my "research" agrees with Perry's as to the 50%. BUT if we had chosen any other reasonable percentile — 40%, 45%, 55%, 60% — the peculiar bulge that Perry and Metaphor are so fond of would disappear!

In the interest of intellectual honesty, I show the 10th and 30th percentiles also, even though they may support the Perry-Metaphor thesis. Raise your hand if you think a "scholar" like Professor Perry would have been so bold, were the effect reversed.

Smells like cherry picking to me.
 
You are simply wrong about what the point of the thread was. I'm in a good position to know because I wrote the OP. The point was to highlight the idiotic wrongness of the gender pay gap narrative.

Ya know [MENTION=103]Metaphor[/MENTION]; , here's a problem.
I agree with you on this part. I agree that to an enormous extent the "pay gap" is now mostly about women choosing their own priorities like men do. But they tend to prioritize differently from men. They tend to prioritize flexibility, secure working environment, transportable skills and other things more than men. The tradeoff is lower income. The wage gap is primarily the collective choices women make compared to the ones men make.

But that wasn't at all clear in the OP. You knew what the point was, but even I(who agrees with you) didn't see anything like that. Given your OP history, like Cornell and Vermont and such, you kinda need to be very clear.

Tom

That's only part of the story.

Women DO tend to value flexibilty, etc. At least women with children do, or those who are planning to have children. Why is this less a concern for men?

Because a lot of men leave that burden to the women. The best economic answer would probably be for one parent to maximize income and one to give up some income for flexibility.

Traditionally, women were compensated at lower rates than men were because they were women. Likewise, professions that are heavily female staffed tend to be compensated at lower rates, despite requiring as much or more education, skill, etc. WHY?

There is a current nursing shortage--has been for years. WHY? This is a case where the obvious market solutions (higher compensation, more flexibility, etc). has not been instituted to encourage more nurses although the nursing shortage has been ongoing for years. WHY?.

Nurses don't bill. Thus many of the people who employ nurses like the shortage--it lets them get away with providing less service.
 
You are simply wrong about what the point of the thread was. I'm in a good position to know because I wrote the OP. The point was to highlight the idiotic wrongness of the gender pay gap narrative.

Ya know [MENTION=103]Metaphor[/MENTION]; , here's a problem.
I agree with you on this part. I agree that to an enormous extent the "pay gap" is now mostly about women choosing their own priorities like men do. But they tend to prioritize differently from men. They tend to prioritize flexibility, secure working environment, transportable skills and other things more than men. The tradeoff is lower income. The wage gap is primarily the collective choices women make compared to the ones men make.

But that wasn't at all clear in the OP. You knew what the point was, but even I(who agrees with you) didn't see anything like that. Given your OP history, like Cornell and Vermont and such, you kinda need to be very clear.

Tom

That's only part of the story.

Women DO tend to value flexibilty, etc. At least women with children do, or those who are planning to have children. Why is this less a concern for men? ...

Call me an old fogey with lingering sexism, but I agree with a point TomC and Toni (and Metaphor!) are driving at. There are valid reasons for pay gaps; trying to impose pay equality can lead to silliness.

In fact, when I compare the WH pay by gender looking at several percentiles I see the OPPOSITE of what Perry and Metaphor see. The identity of mean pay, and pay at almost every percentile EXCEPT 50% makes me suspect that some staffing official was running special software or looking at spread-sheets to artificially impose pay equality on the WH roster! (I'm guessing, but I think Metaphor — and presumably Perry — would find this to be silly, at best. I tend to agree with them!)

If the pay grades and hiring were done to deliberately create pay equality, why the bulge at 50%? I don't know whether there's a special reason for that bulge; maybe it's just evidence that the Holy Grail which Metaphor envisions, in which statisticians look ONLY at the Blessed Median — praise be unto it — is not as universally worshiped as Mataphor thinks.


I find it amusing that some of us (TomC and I) might agree with Metaphor that working statistics to ensure that an organization like the WH has "pay equality" may be an example of excessive wokeness! I look at the data (same 40% salary, same 45%, same at 55%, same at 60%, etc.) and it leads me to guess that WH staff and salaries were carefully tuned for that equality.

Yet Perry and Metaphor, who presumably would despise such tuning far more than I do, are promoting a statistic that suggests those salaries were NOT tuned: the discrepancy near 50% specifically. A discrepancy they should applaud but, like an action-movie hero jumping from one train to a train moving in the opposite direction, they grab onto that statistical quirk and advance a charge of hypocrisy!

The whole thing really is amusing!

As I've said, I don't care about WH pay disparity or lack thereof. For me the thread is a case study in the Right-Wing Bullshit machine. And definitely a black-eye on AEI research ... and on those who use that site for "information."

Professor Perry EITHER was unaware of the remarkable equality of WH salary stats (shown in the table I posted) OR saw that equality and suppressed it in his write-up because it didn't fit his theme. In the first case, Perry is an incompetent statistician; in the latter case he's dishonest. My guess is he's both.

I don't know Professor Perry. For all I know if I e-mailed him that table he'd reply "Gosh, you're right! I was so busy that week, getting my family vaccinated etc. that I didn't run all the statistical tests I usually do. I'll do an edit on that webpage. Thanks for pointing this out!"

I really do NOT think I'd get that response from Perry, but I can't rule it out. However Metaphor HAS been presented with the table and has NOT retracted.

I am pleased and proud that I spent an hour or so on the WH salary data (the hardest part of the task was ungarbling the result of copy/paste from the pdf file) and refuted Perry's bulge!

Attention Good Googlers: Is Perry's Bulge making the rounds of political talkers? (Did it really get mentioned in the Fucking Guardian for God's Sake as Metaphor seemed to claim, or is AEI just a thing with [sarcasm on] Parler and QAnon?) I ask because I'd like to e-mail my result to an appropriate commentator. In fact I think I will: I've exchanged e-mail with Robert Reich in the past and some others. ... Off to Gmail.
 
Last edited:
Back
Top Bottom