No, the law does not say otherwise. Here's the reason why...
You are giving your opinion on what the law should mean. But the courts have interpreted the law, case precedent, and regulations differently.
Even if you treat all ethnicities fairly, there are far fewer black people in the IT industry
Evidence for discrimination is not based on the general population. It is based on the qualified candidate pool.
... if you had to give people useless jobs just for a census count...
That means they were not qualified. Google is not arguing that women are less capable than men. If they made that argument, and had evidence to back it up, then they could justify the wage gap. But basing the wage gap on "ability to negotiate aggressively" is not legal because that is not an important qualification for an engineer. Or at least that is the opinion of the DOJ.
that a fair process exists is far, far more important than whatever the statistics afterwards say.
That is your opinion. The law says otherwise. Discrimination in hiring on the basis of gender is illegal, and it doesn't matter what process you use to justify that discrimination, or even if the discrimination is unintentional. If the END RESULT is systematic discrimination, then you are going to lose in court.
You want similar pay, you have to actually ask for it.
No you don't. If Google's hiring process results in equally qualified men and women being paid significantly and systematically differently, then it is illegal. They can't use the lame excuse that the qualified women "deserve" to be paid less because they are bad negotiators. Being a "good negotiator" is not relevant to being a qualified engineer. For large companies, the DOJ does not need to prove the hiring process is "unfair", they only need to show that the results of the process are unequal.
Nonsense. One programmer can be ten times as productive as the average programmer. Plenty of programmer are less than one-tenth as productive as the average. When I run an ad, I just don't know the range of candidates I am going to get. If I realize I am interviewing a superstar, then I will make a much higher offer, and I don't want to dissuade the best applicants from applying by (falsely) indicating a fixed salary. I am also often willing to take a chance on a bright self-taught kid, but I don't want to set unrealistic expectations that he (or she) is going to be pulling in the same starting salary as a CS grad.
Posting a fixed wage may work when you are hiring someone to work the french fry machine at McDonalds, but it makes no sense for hiring technical and creative people.
Testing only finds the problems you've already thought of. Peer review can (but is not assured to) find the problems and situations you haven't.
Which would be more reliable: 1. Code that has been tested, but not reviewed. 2. Code that has been reviewed, but not tested. #1 would be a thousand times more likely to work. In most cases, #2 would not even compile.
My experience is that code review is important for code that can be read, understood, and maintained by people other than the author. But it is not effective at ensuring proper functioning.
Scientific peer review is similar. It helps ensure that a paper is readable and understandable. It does NOT mean that the data is valid or that the conclusions are correct. That requires replication, which is far less common.
A job pays what the job is worth and the skill set the candidate brings to the table.
The problem is that a job interview only gives a very tiny peek into that "skill set". At my company, a candidate programmer will spend several hours writing code, and that is a good measure of their ability to write a 50 line function. But it is NOT a good measure of their ability to troubleshoot and patch a system with 200,000 lines of code written by semi-competent people that left five years ago.
Disclaimer: I never ask "What was your previous salary?" because that just encourages lying. Instead I ask "What are your salary requirements?" and if it is reasonable, I try to make an offer a little higher than their expectation. If it is unreasonable, I might try to lowball them and see if they accept, but usually they just won't be getting an offer.
Female candidates are about to become a protected class...
They already are. You can't ask about gender during the interview process. It is usually obvious when a candidate shows up for the interview, but it isn't always obvious during the screening process, especially with a name like Pat or Chris. You also cannot ask about pregnancy, marital status, number of children, or future childbearing plans, which are all issues that disproportionately affect women.
Several women have told me that they removed their wedding rings before job interviews.
Just mandate that the pay be stated upfront in the original job ad.
When I advertise for a "programmer", I usually don't have a specific salary in mind. If one applicant is more capable than another, then I will offer more. If I put a low salary range in the ad, the better candidates will not apply. If I put a high range, then I will be flooded with responses from lousy candidates that are not even remotely qualified for that salary.
If you pay less, you get less. I worked for a company that had about 20 programmers making $30k in the heart of Silicon Valley. How did we do it? We hired kids straight out of high school and trained them as code monkeys, to whip up Javascript or throw-away Perl scripts. Most of our projects were quick one-off stuff, and when we did need to maintain something for the long term we had one of our "real" programmers clean it up. This actually worked amazingly well, and the company was profitable for years. I kept in touch with many of those kids, and most of them went on to successful tech careers, and one of them even got a PhD from Stanford.
The various drug companies spend 3 1/2 x as much on advertising and marketing as research
The actual ratio is about 1.8 to one, but still an unreasonable amount considering all the taxpayer subsides that these companies receive. Much of that is spent marketing directly to doctors in ways that are nearly indistinguishable from just bribing them to write prescriptions.
rather than look for someone who will create cures at a more sustainable cost.
If you are offering BILLIONS of dollars, you don't need to "look for someone". They will find you. The fact that no one else is offering to undercut the pharma companies producing these drugs most likely means there is no excessive profit margin. The development costs are enormous, and since the diseases are rare the market is tiny, that means a high unit price.
All evidence shows that the wealthier we get, the fewer children we have.
Only some evidence shows that. Until about 50 years ago, wealth meant more children. So what changed? 1. Contraceptives 2. Rapidly falling infant and childhood mortality 3. Increasing urbanization
Journal publishers act as a gateway in most fields, but not in Physics, where most papers are freely available as preprints on Arxiv.org. The reason Physics is different is that Arxiv was established in 1991, long before journal publishers saw online preprints as a threat, and by the time they recognized the problem, online preprints were already established as the norm in the Physics community. But in other fields, they were able to stifle the move to open publishing by embargoing any papers that violated their no-preprint policies.
Bill Gates has recently said that his foundation will require open publishing of research that they fund. That is a big positive step, since his foundation is an important source of funding for biomedical research.
But I'm wondering about the authors of the article if they jump from "not due to calories" to a conclusion that it must have been ritualistic. Why not taste?
The most plausible explanation is that they were fighting wars over territory with other tribes, and if people are killed in battle, then hey, there is no good reason to let a perfectly good corpse go to waste.
Eating each other for dietary needs would seem self-limiting. Also any society that condones murder doesn't last long.
Cannibals don't eat "each other". They eat their enemies and outsiders. Few societies consider wartime killings to be "murder". Even the Bible (at least the Old Testament) condones killing outsiders.
No. Small animals taste like chicken. Humans taste similar to pork. My Appalachian in-laws call it "long pig". They say that once you get past the "yuck" factor, it isn't that bad.
Of course, collecting and selling the information gained from those conversations to advertisers, that I'm pretty sure they would (and do) do.
There is no evidence whatsoever that they have done this. They have stated unequivocally that they do not sell your data and never will.
Selling data to advertisers would be stupid, and expose them to consumer backlash and lawsuits. It would make no sense. They make a ton of money by using that data to target ads, so why would they sell it to someone else who could undercut their business? If you were in the milk business, would you sell your cows?
The question is, why are so many places "doing scrum wrong"?
The GPP is not "doing scrum wrong", he is not doing it at all. He is using the word "scrum" to describe the normal time-wasting "status" meetings that are common at dysfunctional companies. That is not what scrum is.
Is scrum just hard to understand and implement?
No. There are many parts of Agile that are hard, but scrum isn't one of them. Scrum is easy. Just give each person 30 seconds to talk, and they will be forced to focus on the most important issues. If managers interrupt, ignore the scrum master, and hijack the meeting, then YOU ARE NOT DOING SCRUM ANYMORE and you should call it something else.
No solution to this problem of automation breaking the cycle of work, wages and consumption...
Except that "this problem" doesn't actually exist. Automation is slowing down as service jobs are proving much harder to automate than the manufacturing jobs that are mostly already gone.
The real problem in our economy is the opposite: poor productivity growth because of low levels of automation adoption.
We hear that all the time, "you're not doing it right...".
I get annoyed when people say that about little details, but if your daily scrum meetings are lasting 75 minutes then you are doing it so wrong that it is inaccurate to even call it "scrum".
My scrum meetings last 5 minutes. The purpose is to raise issues, not to discuss them. The discussion happens elsewhere.
Slashdot Mirror
No, the law does not say otherwise. Here's the reason why ...
You are giving your opinion on what the law should mean. But the courts have interpreted the law, case precedent, and regulations differently.
Even if you treat all ethnicities fairly, there are far fewer black people in the IT industry
Evidence for discrimination is not based on the general population. It is based on the qualified candidate pool.
... if you had to give people useless jobs just for a census count ...
That means they were not qualified. Google is not arguing that women are less capable than men. If they made that argument, and had evidence to back it up, then they could justify the wage gap. But basing the wage gap on "ability to negotiate aggressively" is not legal because that is not an important qualification for an engineer. Or at least that is the opinion of the DOJ.
that a fair process exists is far, far more important than whatever the statistics afterwards say.
That is your opinion. The law says otherwise. Discrimination in hiring on the basis of gender is illegal, and it doesn't matter what process you use to justify that discrimination, or even if the discrimination is unintentional. If the END RESULT is systematic discrimination, then you are going to lose in court.
Unequal results do not imply an unfair process exists.
Whether it is "fair" or not is irrelevant. All that matters is that it is illegal.
You want similar pay, you have to actually ask for it.
No you don't. If Google's hiring process results in equally qualified men and women being paid significantly and systematically differently, then it is illegal. They can't use the lame excuse that the qualified women "deserve" to be paid less because they are bad negotiators. Being a "good negotiator" is not relevant to being a qualified engineer. For large companies, the DOJ does not need to prove the hiring process is "unfair", they only need to show that the results of the process are unequal.
To argue otherwise is disingenuous.
Nonsense. One programmer can be ten times as productive as the average programmer. Plenty of programmer are less than one-tenth as productive as the average. When I run an ad, I just don't know the range of candidates I am going to get. If I realize I am interviewing a superstar, then I will make a much higher offer, and I don't want to dissuade the best applicants from applying by (falsely) indicating a fixed salary. I am also often willing to take a chance on a bright self-taught kid, but I don't want to set unrealistic expectations that he (or she) is going to be pulling in the same starting salary as a CS grad.
Posting a fixed wage may work when you are hiring someone to work the french fry machine at McDonalds, but it makes no sense for hiring technical and creative people.
Testing only finds the problems you've already thought of. Peer review can (but is not assured to) find the problems and situations you haven't.
Which would be more reliable:
1. Code that has been tested, but not reviewed.
2. Code that has been reviewed, but not tested.
#1 would be a thousand times more likely to work.
In most cases, #2 would not even compile.
My experience is that code review is important for code that can be read, understood, and maintained by people other than the author. But it is not effective at ensuring proper functioning.
Scientific peer review is similar. It helps ensure that a paper is readable and understandable. It does NOT mean that the data is valid or that the conclusions are correct. That requires replication, which is far less common.
A job pays what the job is worth and the skill set the candidate brings to the table.
The problem is that a job interview only gives a very tiny peek into that "skill set". At my company, a candidate programmer will spend several hours writing code, and that is a good measure of their ability to write a 50 line function. But it is NOT a good measure of their ability to troubleshoot and patch a system with 200,000 lines of code written by semi-competent people that left five years ago.
Disclaimer: I never ask "What was your previous salary?" because that just encourages lying. Instead I ask "What are your salary requirements?" and if it is reasonable, I try to make an offer a little higher than their expectation. If it is unreasonable, I might try to lowball them and see if they accept, but usually they just won't be getting an offer.
Female candidates are about to become a protected class...
They already are. You can't ask about gender during the interview process. It is usually obvious when a candidate shows up for the interview, but it isn't always obvious during the screening process, especially with a name like Pat or Chris. You also cannot ask about pregnancy, marital status, number of children, or future childbearing plans, which are all issues that disproportionately affect women.
Several women have told me that they removed their wedding rings before job interviews.
Just mandate that the pay be stated upfront in the original job ad.
When I advertise for a "programmer", I usually don't have a specific salary in mind. If one applicant is more capable than another, then I will offer more. If I put a low salary range in the ad, the better candidates will not apply. If I put a high range, then I will be flooded with responses from lousy candidates that are not even remotely qualified for that salary.
If you pay less, you get less. I worked for a company that had about 20 programmers making $30k in the heart of Silicon Valley. How did we do it? We hired kids straight out of high school and trained them as code monkeys, to whip up Javascript or throw-away Perl scripts. Most of our projects were quick one-off stuff, and when we did need to maintain something for the long term we had one of our "real" programmers clean it up. This actually worked amazingly well, and the company was profitable for years. I kept in touch with many of those kids, and most of them went on to successful tech careers, and one of them even got a PhD from Stanford.
The various drug companies spend 3 1/2 x as much on advertising and marketing as research
The actual ratio is about 1.8 to one, but still an unreasonable amount considering all the taxpayer subsides that these companies receive. Much of that is spent marketing directly to doctors in ways that are nearly indistinguishable from just bribing them to write prescriptions.
rather than look for someone who will create cures at a more sustainable cost.
If you are offering BILLIONS of dollars, you don't need to "look for someone". They will find you. The fact that no one else is offering to undercut the pharma companies producing these drugs most likely means there is no excessive profit margin. The development costs are enormous, and since the diseases are rare the market is tiny, that means a high unit price.
If you can convince couple unaffiliated people in your field that this is decent-enough idea - then it is real science.
That is just normal peer review, which is the lowest bar. How do you decide which researchers receive funding, or who gets tenure?
How do you convince that your code works? You have someone else to review it.
Nonsense. You see if code works by testing it. You review code to ensure quality (readability, maintainability) not correctness.
All evidence shows that the wealthier we get, the fewer children we have.
Only some evidence shows that.
Until about 50 years ago, wealth meant more children.
So what changed?
1. Contraceptives
2. Rapidly falling infant and childhood mortality
3. Increasing urbanization
Journal publishers act as a gateway in most fields, but not in Physics, where most papers are freely available as preprints on Arxiv.org. The reason Physics is different is that Arxiv was established in 1991, long before journal publishers saw online preprints as a threat, and by the time they recognized the problem, online preprints were already established as the norm in the Physics community. But in other fields, they were able to stifle the move to open publishing by embargoing any papers that violated their no-preprint policies.
Bill Gates has recently said that his foundation will require open publishing of research that they fund. That is a big positive step, since his foundation is an important source of funding for biomedical research.
Instead of doing real science it is a popularity contest.
How do we determine what is "real science"?
But I'm wondering about the authors of the article if they jump from "not due to calories" to a conclusion that it must have been ritualistic. Why not taste?
The most plausible explanation is that they were fighting wars over territory with other tribes, and if people are killed in battle, then hey, there is no good reason to let a perfectly good corpse go to waste.
Eating each other for dietary needs would seem self-limiting. Also any society that condones murder doesn't last long.
Cannibals don't eat "each other". They eat their enemies and outsiders. Few societies consider wartime killings to be "murder". Even the Bible (at least the Old Testament) condones killing outsiders.
so, do we taste like gamy chicken?
No. Small animals taste like chicken. Humans taste similar to pork. My Appalachian in-laws call it "long pig". They say that once you get past the "yuck" factor, it isn't that bad.
Look, half the reason people come to NYC is the fact that you don't drive - you take cabs or the subways.
Then why are the streets full of cars?
Of course, collecting and selling the information gained from those conversations to advertisers, that I'm pretty sure they would (and do) do.
There is no evidence whatsoever that they have done this. They have stated unequivocally that they do not sell your data and never will.
Selling data to advertisers would be stupid, and expose them to consumer backlash and lawsuits. It would make no sense. They make a ton of money by using that data to target ads, so why would they sell it to someone else who could undercut their business? If you were in the milk business, would you sell your cows?
The question is, why are so many places "doing scrum wrong"?
The GPP is not "doing scrum wrong", he is not doing it at all. He is using the word "scrum" to describe the normal time-wasting "status" meetings that are common at dysfunctional companies. That is not what scrum is.
Is scrum just hard to understand and implement?
No. There are many parts of Agile that are hard, but scrum isn't one of them. Scrum is easy. Just give each person 30 seconds to talk, and they will be forced to focus on the most important issues. If managers interrupt, ignore the scrum master, and hijack the meeting, then YOU ARE NOT DOING SCRUM ANYMORE and you should call it something else.
No solution to this problem of automation breaking the cycle of work, wages and consumption ...
Except that "this problem" doesn't actually exist. Automation is slowing down as service jobs are proving much harder to automate than the manufacturing jobs that are mostly already gone.
The real problem in our economy is the opposite: poor productivity growth because of low levels of automation adoption.
We hear that all the time, "you're not doing it right...".
I get annoyed when people say that about little details, but if your daily scrum meetings are lasting 75 minutes then you are doing it so wrong that it is inaccurate to even call it "scrum".
My scrum meetings last 5 minutes. The purpose is to raise issues, not to discuss them. The discussion happens elsewhere.
Legally, at least in the US, if they deny you the opportunity to take your vacation they must pay it out in a use it or lose it system.
This is incorrect. There is no such requirement.
Some individual states require compensation for forfeited vacation, but there is no federal requirement.