The lean startup crowd exalts the power of data to overturn shadow beliefs about what is good for the business that more reflect the personal opinions of the people in charge rather than the reality on the ground.
I say this with the utmost respect: "diversity leads to innovation/meritocracy" is a shadow belief (just like "all good programmers are geeky white males from MIT or Stanford" is a shadow belief). It may be true. It may also be catastrophically untrue. I'm agnostic. I can afford to be agnostic about this since I'm one guy in a rice field and nobody's opinions on diversity will change the race/gender/etc composition of my business (kinda hard to do with a one-man team), but if you're making hiring decisions based on this, you might want to start thinking of process design decisions which place more emphasis on the data and less on shadow beliefs.
(Possible solution not accounted for above: I suppose if you value diversity as an end goal to itself, then you could run your business in such a way as to maximize it even if it was not optimal among other axes.)
I don't have data on "diversity" as such, but I tend to give extra points to candidates who speak multiple languages.
It's not just an indicator of intelligence. A team with broad "cultural literacy", in my experience, tends to create software with broader appeal. At Terespondo we kicked Google, Yahoo and Microsoft's ass for years in search advertising in South America because we understood and catered to our customers.
This probably does not hold for all kinds of software but for consumer stuff it's been a real advantage.
Do you sell/support BCC in Japanese? Not many people here could do that.
As a "man on the ground" with a team to manage/hire/fire I have found what Eric says in his post to be quite true. It has been very helpful for my team have a wide range of races, disciplines and of course genders.
I don't see anything revolutionary in what Eric espouses and in practice I've found them to be quite successful in helping my team keep moving in the right direction.
ps. I come from a technical not business background but have taken on more of that role as the years have progressed.
He addresses the point head-on: "Diverse teams make better decisions than homogenous ones. I won’t recap the academic research that underlies this assertion; for that, you should read James Surowecki’s excellent Wisdom of Crowds.
I am not unfamiliar with that research. I'm just not very impressed by it, particularly with the relevance of it to the issue at hand.
An awful lot of academic writing about diversity starts with the conclusion and then finds results to justify it. I mean, hypothetically supposing one were to do a double-blind study of pick-your-favorite-improvement and find, to your surprise, that it resulted in more skew, not less. ("We scrubbed all the resumes of any indication of race/gender and ended up inviting more white males to interviews?! Dude, WTF.")
Would you want that study to be in your CV when tenure decisions were being made, knowing what you do about the shadow beliefs of the people on your tenure committee? Bury that data and bury it deep if you value your career.
Incidentally, some sources of repeatable bias in academic studies are actually documented in the literature (whoa, meta!) For example, people have a tendency to not publish null results and just file them away in a desk drawer, which means the published literature in e.g. marketing systematically exaggerates the magnitude of effects. If your team is doing A/B testing and every test moves the needle something is almost certainly wrong, but in the literature something almost invariably moves the needle.
Larry Summers would publish that paper. What about all those researchers who already have tenure? There are plenty of academics who are willing to challenge the status quo on thinking if they believe they found evidence to the contrary. It would then be their duty to explain why something like that might happen. In your example perhaps it could be associated with a white people generally still having more opportunities throughout their lives in education (better schools on average), job experience (better jobs on average) and perhaps other factors. They could even spin that into people aren't being given an equal chance based on race and there is lasting effects throughout people's lifetime. The RESULT is the same: people hire more white people in blind tests, but the call for action is different.
Larry Summers as in "forced out of Harvard with the precipitating event being that he suggested there might be an outside possibility that differences in female participation in science are caused by differences in female's motivation to pursue science" Larry Summers? He would not be the first example I'd bring up in defense of the academy's willingness and ability to tolerate heretics. He had tenure, he has status, he was the flipping President of Harvard, he had just about everything going for him an academic possibly could. And that still didn't save his job when he said something the academy didn't like.
I thought about not writing that but decided it would probably be good bait to see if you were going to actually look at the rest or not and actually comment on substance. You make it sound like he lost his status and his job. He simply got a different position at Harvard and works with Obama, real step down.
You really think that unbiased data for an issue as complex and multi-layered as this is possible? There is never going to be a paper that proves once and for all that the best people to hire are white american males.
So don't talk about "we need the data" for situations where it is clear that there will never be any data. In such cases, just do what you feel is the right thing: if you feel your business will only do well with a particular type of person, stick with that type of person.
Personally, I found the argument as constructed convincing - intellectual diversity is neccessary. And a homogenous crowd is not likely to be intellectually diverse.
> In such cases, just do what you feel is the right thing: if you feel your business will only do well with a particular type of person, stick with that type of person.
I don't think that's legal in the United States for a sufficiently large company.
There are dueling definitions of "diversity" going on. That book seems (judging from the amazon description) to be talking about diversity of opinion. Cultural/religious/racial diversity is a different thing.
“Our diversity, not only in our Army, but in our country, is a strength. And as horrific as this tragedy was, if our diversity becomes a casualty, I think that’s worse,” Casey said.
Read: "In our particular organization, diversity is more important than not being killed." Also read: "I'm a 4-star general, and anyone who disagrees with this particular insanity is going to be in a shit load of trouble."
"To be clear, though, this diversity refers only to diversity of opinion, not necessarily to demographic diversity. So why is demographic diversity important?
"Demographic diversity is an indicator. It’s a reasonable inference that a group that is homogeneous in appearance was probably chosen by a biased selector."
"Demographic diversity is an indicator. It’s a reasonable inference that a group that is homogeneous in appearance was probably chosen by a biased selector."
How about some supporting evidence?
Better yet, how about some evidence that demographic diversity (ethnicity, age, gender, income/wealth) is a useful indicator of useful diversity. After all, that's your claim.
Of course, if you do show that such diversity is an indicator of useful diversity, you've just shown that one should select against certain such characteristics in (albeit not necessarily the same characteristics) in almost every situation. Why? Because in each situation, some characteristics are better than others.
That doesn't imply that a team won't be diverse, just that, if you're correct, successful teams will be more likely to have certain "looks" in each role.
I think that that's bunk, but I can't explain why there hasn't been a white DB in the NFL for several years.
I say this with the utmost respect: "diversity leads to innovation/meritocracy" is a shadow belief (just like "all good programmers are geeky white males from MIT or Stanford" is a shadow belief). It may be true. It may also be catastrophically untrue. I'm agnostic. I can afford to be agnostic about this since I'm one guy in a rice field and nobody's opinions on diversity will change the race/gender/etc composition of my business (kinda hard to do with a one-man team), but if you're making hiring decisions based on this, you might want to start thinking of process design decisions which place more emphasis on the data and less on shadow beliefs.
(Possible solution not accounted for above: I suppose if you value diversity as an end goal to itself, then you could run your business in such a way as to maximize it even if it was not optimal among other axes.)