I love the irony of this. Accused of underpaying women - their reaction is "let us get some data". Nope. underpaying men. This is the way you do things. You collect the data, analyze and react. Wash rinse repeat.
Given that Google's workforce is 69% men, it's quite possible that they found 3 men underpaid for every 2 women who were underpaid (60%/40%), which would mean they gave more men than women salary equity adjustments while still having a bias in the direction of the original complaint.
It's not a contest of who is more underpaid. This would indicate that Google systematically underpays, regardless of gender. THAT is the real issue here.
Lots of people want to work in game development. Lots of people do. Some people say game devs are underpaid. I tend to think that they're paid what they agreed to accept and that is based in large part on market forces (supply and demand, value over replacement, etc.). I also left game development because it wasn't for me long-term. (I was underpaid relative to what I could make in another industry, but I wasn't underpaid by my game-dev company. They quickly hired a replacement for me, I'm sure.)
A company like Google, paying substantially over minimum wage, is probably not underpaying most employees systematically, almost by definition IMO.
> I was underpaid relative to what I could make in another industry, but I wasn't underpaid by my game-dev company.
If a company underpays all its employees equally, it's still underpaying. Companies that do this are taking advantage of how people are not rational actors.
In the context of video game companies, it's the comparison to other software companies and positions requiring about the same level of experience and prestige. Game producers are well-known at this point for recruiting naive grads and paying them peanuts compared to what they could get even just making CRUD web apps.
It's likely that many charities underpay software developers, because the developers are working there voluntarily in part for the mission.
Political campaigns have the same property in a lot of cases.
So long as people want to express their preferences in where they work based on the content of the work, the mission of the organization, the laptops they issue, or any other set of criteria, I think it's fair game that those choices tip the supply/demand relationship in or away from their favor.
I think a key difference here is that a greater awareness of market-rate salaries preemptively affect those "want to work on video games" decisions much more than working for a volunteer organization. Corporate organizations in general put a lot of effort and culture into suppressing open knowledge and comparisons of pay rates.
Not everybody is underpaid, but among the underpaid, they found men as well as women.
As random illustration, they might pay 200k+-5% to 80% of all level X developers, and the other 20% might only earn 150k+-5%.
I still agree that 'underpaid' is a difficult term here. Maybe the 80% are overpaid? The only hard fact in this example would be that the salary is more skewed than it should be for a 'fair' compensation.
What does the ideal salary distribution curve look like, anyway? Should it be a bell curve? A long-tail distribution? Should it have distinct peaks, i.e. everyone with the same YOE is paid exactly the same?
Should the salary distribution mirror the talent distribution, or no?
Good questions. I think that compensation should mirror the value an employee provides to the company. However, in many cases, that is very hard or even impossible to measure.
In large companies with many career levels, the role level indicates the perceived value for the company. Then, the distribution within the level doesn't matter that much, as long as the difference between minimum and maximum pay is relatively small, and there is not much overlap to the neighborhood levels.
Of course, there is no objective criterion what "relatively small" means in this context.
And, as the article already points out, this only covers fairness within the same level. Maybe even more important is if everybody is evaluated in the same way when it comes to career progress.
If you believe that developers are within a factor of 33% of productivity (defined notionally as “value created per week”), then there is excessive skew in the numbers you propose.
I believe the productivity figures are at least a factor of 2, and probably a tail out to a factor of 10 or more versus the median developer (and then a left-going distribution under the median where a substaintial number of developers do not create any value at all). Under that set of beliefs, 20% making ~$150K and 80% making ~$200K is actually insufficient skew vs a “fair” system.
I wrote this in the context of the article, where a Google developer with 33% more productivity would already have progressed to the next role level. At least, that is what I would assume for large companies.
Promotion is not centered primarily around productivity, at least not beyond SWE2 or whatever the second engineering level is in a given company. Ability to lead/influence at a scope larger than self is largely disjoint from individual productivity in coding.
Not literally everyone can be underpaid, but most people can be underpaid, with a very small percentage paid fairly (and a smaller percentage overpaid)
If you're underpaid, that means there is an opportunity for you that will pay more. Aside from H1B, I struggle to see how anyone at google is underpaid.
I'm not sure, but you seem to be asking many questions in a pointed manner on this topic, while offering no answers or even interpretations of your own.
I don't claim to have any of the answers, and it is a really hard problem. I'm just pointing out that it doesn't mathematically make sense to claim that everyone can be underpaid relative to each other^, and that the claim that software engineers are underpaid in general relative to the value they create is a separate issue.
^ That's like saying that two people can each be shorter than each other. It's not possible according to the way the "less than" function is defined.
I'm not implying that you are purposely doing anything. You like most people on HN are not sexist.
What is happening though is a flaw within our logical frameworks. Our cultural biases are making something evident.
If more women are underpaid than men, then that it is obviously and overtly a form of sexism... logically this implicates that If more men are underpaid than women then it is also sexism.
Instead of calling out the logical implication why do people jump to an illogical conclusion?
The question is now, do women experience sexism in terms of compensation at google or do men? Or do neither?
> Men account for about 69 percent of the company’s work force, but they received a disproportionately higher percentage of the money.
So they were compensated out of proportion with their demographic makeup of the company--they made up a larger share but needed more money to "catch up."
I think there's not enough information in the article to determine conclusively. Suppose men represent 75% of the compensation (due to senior ranks being skewed male more than lower ranks), 69% of the workforce, and received 72% of the salary equity increases. Were men or women "more" underpaid previously (assuming the equity increases themselves were "fair")?
A similar thing has happened with companies that go to great lengths to make sure they don't have any gender bias in their hiring process (no face-to-face during interviews, voice altering during phone interviews, etc.) only to find that they still selected more men than women. The conclusion, of course, is that they must be doing something wrong because it's simply impossible that they wouldn't be selecting a perfect 50/50 mix of men and women.
Similarly, I think imagine Google will find similar flaws with how these numbers were generated.
It's extremely difficult to actually hide all signs of gender, even with the measures you mentioned. Even simple things like the way a job advertisement or a resume is worded can suggest the genders of the people involved in a statistically predictable way. For example: http://gender-decoder.katmatfield.com/static/documents/Gauch...
