Monday assorted links

Comments

"5. Post-Damore, 46% of self-identified moderates feel less comfortable expressing their views in Silicon Valley."

I think the whole paragraph is informative:

"The impact on different ideological groups was starker: 70 percent of self-identified “very conservative” respondents, 64 percent of “conservatives,” and 66 percent of “libertarians” were “less” comfortable. Even 46 percent of moderates were less comfortable. By contrast, 13 percent of liberals and 26 percent of “very liberal” respondents felt more comfortable sharing their views."

The last line is odd. When has a liberal not been overly comfortable in sharing his/her views?

That tendency had now been ramped up to 11.

TMC is correct. You would have to be in the industry in the valley to understand.

I mean JWatts.

"I haven't seen black panther, but I heard it's reaaaally good"

- All white people right now

I heart it was just another "superhero movie" and as such not worth watching unless you have a friend in the business

It did $192 million in 5 days in the US alone breaking tons of records. Plenty of white people saw it.

Someone I know was called in by execs and told that their career advancement was over at their company because of their reaction to the announcement that a certain well known conservative figure was speaking on premises.This person came to tell me right after, so I am pretty darned sure of their authenticity. They would have probably gotten away with what they said had they been an individual contributor, but not as a manager. Said conservative speaker is quite high profile, but I don't think anyone would consider said speaker to be in the business of trolling.

Now, this is a place where, if I was a Trump supporter, I'd be uncomfortable of sharing my views at all, so it's not as if it's a conservative SV company. Very little tolerance of views on one side doesn't lead to infinite tolerance on the other side either. Regardless of where the center point is (and yes, in SV, a large majority of companies lean heavily left), there's still a variety of levels of tolerance to dissent past this center point, and some might be far narrower than you'd think.

I find this story implausible. If it happened then large gaps are being left out.

At the large, leading, SV companies (Google, FB, linkedin, Netflix, salesforce, etc) the left is generally free to run amok on the various groups/boards/memegens. They are free to shun any that disagree with their ideology, and are encouraged to report any slight to HR.

So, either the person in the story was not at a leading firm or they went full liberal and did something astonishing (publicly humiliating the company) or it didn’t happen.

I doubt even close to 46% of self-described moderates actually hold moderate viewpoints. Everyone wants to think they're middle class and have moderate political opinions, but most often they're wrong.

Moderate relative to which group is the question. Certain commenters here like to fancy themselves as being politically "moderate" (not me of course as a card-carrying member of the vast right-wing conspiracy).

Indeed. I'd venture to guess, based on being friends with a number of Silicon Valley types, that that group in particular would be unusually bad at locating itself on the mainstream political spectrum. Since the general public is pretty crappy at it, I'd wager tech people are absolutely awful.

I'd also guess that the error goes both ways, and on both sides.

I think the Damore episode should make people think twice about claiming diverse groups are more effective than homogenous groups.

I want to suck some Bill Krystal cock.

Because Damore pointed out that Google is not very diverse, and then Google quickly fired him (hence drawing most attention away from the fact that Google isn’t very diverse)?

Only 70% Asian!

1: The headline made me chuckle, but then I read the article and it covers very real issues: international students who graduate with a STEM degree in the US have an advantage in the work permits that they can get.

And although most people would agree that economics is not a STEM field broadly speaking, to the extent that the student studies econometrics (or some of the mathematical economics subjects) they are studying a subject that most people would agree is STEM. As some of the commenters to the article say, compared to say field biology or working in a chemistry lab, some parts of econ are a lot more STEM-y.

But is that enough for us to say that econ is a STEM field? Most people including I think most economists would be queasy at the idea, but the article makes it clear that there are strong incentives for both the departments and the international students to have a program categorized as STEM. That doesn't mean it's right, but it is happening.

I had a friend who was in a top biology grad program and when I showed her the students' homework that I was grading (intermediate micro) she said "wow that's a lot of math". I responded, jokingly, "yes this is Real Science". People have and will continue to argue about the "science" part, but the math part is there and no one questions whether math is part of STEM.

I disagree, Math, when used to produce some useful relationship, is indeed "STEM". however Math, when used to simply churn numbers, is not really STEM. <> /s

Nonetheless, an employer who needs a mathsy mind for some reason can get an Econ student who has done a lot of maths, whether or not it was useless. And since mathsy tasks in the real world often involve vaguely defined operations on noisy, difficult to interpret datasets -- an economist might be better suited than most "real" STEM graduates.

So if you were a professor of Mathematics who studief pure mathematics, that wouldn't count as STEM? I think STEM is an arbitrary, overly broad, useless categorization, nonetheless if defining STEM matters for legal reasons a course of study depending on advanced knowledge of math has to count.

Also, if you're talking about how people thin l about STEM instead of a technical definition, it seems like everyone pretty much means engineering.

Why do we care about STEM? Presumably, it's because people with math abilities are in shorter supply. In that sense, quantitative economics should be part of STEM. Those people may have quant skills that are in shorter supply than non-STEM.

Some may argue that quantitative economics majors are not in short supply. That highlights the shortcomings of allowing politicians and bureaucrats to design some sort of central planning "point system" for immigration. We already have a point system --- it's called the labor (and housing) market.

"1. Is econ STEM?" It's a sort of trans-STEM. Intersectional STEM.

Economics isn’t science, tech, or engineering. Arguing that it fits into math because it makes use of math is to misunderstand what it means to work in the field of mathematics. Physics isn’t computer science even when we write long codes and use up a lot of cpu time. Political science, sociology, etc... aren’t STEM either - even when they use sophisticated statistical models to interpret their result.

Biology and wetbench chemistry are science- even when the math is pretty basic.

Econometrics is statistics with greater emphasis on structural modeling and theory. You might dismiss econometrics when say it's used to estimate somebody's macroeconomic model, but the skills, techniques, estimators, theorems, etc. of econometrics are unquestionably STEM: it's a form of statistics.

The issue is not whether econometrics is STEM; the issue is that there are very few, perhaps no, undergraduate programs where the students can major in econometrics. Instead they major in economics, where the STEM label is indeed questionable.

1. Is econ STEM?
Yes, except labor markets.

How many Econ grads are STEM in the sense that can/do go on to do non-Econ STEM?

Economists are allowed to publish in Stats journals from what I hear.

It is remarkable that in 2018 anyone is surprised by #4, that "indirect" genetic effects outnumber direct effects, in any multicellular creature. We've got 20,000 genes to do millions of different tasks at different times, and so it's a given that the combinatorics of many genes account for any specific activity. There are a handful of one gene / one action correspondences that were figured out long ago and became the canonical textbook examples of genetics, but this has sadly influenced lots of people, including lots of biologists, to think that this is how life works.

Routinely I attend seminars that focus on how perturbing one gene has an effect on something, avoiding the fact that a hundred other genes would have done so also. When pressed, most would acknowledge that pleiotropy (one gene -> multiple effects) is a thing, but they'd prefer to ignore it. The field of Systems Biology has aimed to counter this, to treat genes as part of networks whose overall stability and dynamics are important. Though this has been influential, it's less influential than it should be; in brief, it's hard.

Re: Damore... it's important here that Damore wasn't fired for holding a "conservative" viewpoint, he was fired for being wrong, empirically wrong, in a way that revealed his inability to objectively evaluate his coworkers. Both sides seemed to miss this. The left preferred to be outraged rather than rebut his points, but on many points he was in fact incorrect. The Economist, imo, did the best job on that news cycle:

https://www.economist.com/news/international/21726276-last-week-newspaper-said-alphabets-boss-should-write-detailed-ringing-rebuttal

In any case, it's problematic that conservatives feel less likely to express their beliefs... but it's also problematic that 60%+ of self-described conservatives find their views close enough to Damore's to consider his firing significant.

"The Economist, imo, did the best job on that news cycle:"

TLDR. What was he actually wrong about?

Heh. Talk about your motivated reasoning.

There is a reasonable defense to Google's actions, though. I reckon they get sued all the time by protected groups (anyone who isn't a white guy under 40), and firing Damore would be very helpful in defending against such suits, while the risk of suits from white guys under 40 is comparatively minor. That's a hard-headed business decision, and it might be the real calculation, but you can't just come out and say it.

Indeed. This was almost certainly the reason Google’s management fired him so publicly. Google has a tremendous target on their back, since US labor laws are insane, and they know it. There was an insane amount of pile-on of virtue singaling within Google during the affair, and that will likely hurt Google during Damore’s lawsuit.

What exactly are you arguing that he was he wrong on?

I'm not agreeing with everything he said, but it's not a good argument to post a link to an article is a bunch of childish insults.

For example, their first claim (before the explicit list of six other claims), is that Damore leaped to the conclusion that women can't code. Where in the memo did Damore say that women can't code?

Damore was right that medians of populations differ, but he makes the bigot's mistake. "Since X tend to be a bit Y, I will assume the next X in the door is Y."

Maybe blame schools for not teaching statistics, and the limits thereof.

What are you talking about? Where did he say something like that?

He specifically said two populations can be different and have significant overlap and that "you can’t say anything about an individual given these population level distributions.". He drew a diagram to illustrate the point.

You should probably read the memo before you start name-calling.

You need to provide a direct quote from him where he does that.

You are dreaming awake Anon. X won't provide any source for its statement, any more that Hitler provided any proofs for his affirmation that Poland provoked Germany to war in 1939. And the method of fight against people like X ought to be the same as the ones that eventually worked against Hitler.

X,

Clearly you did not read it, unless you have a reading comprehension problem. He said no such thing.

You are incorrect, but even if it were true that he was fired for being wrong, it would still be terrible. It takes a lot of trial and error to build a model of the world, and for that to be possible you can't severely punish people for being wrong. Or do you think theories and models spawn fully formed and correct in the heads of their originators?

Joël: "the method of fight against people like X ought to be the same as the ones that eventually worked against Hitler." Exactly: send our F35's to Dresden and incinerate that city again.

"he was fired for being wrong, empirically wrong, in a way that revealed his inability to objectively evaluate his coworkers"

He wasn't wrong. At best you could say that some of his statements are currently under debate.

http://quillette.com/2017/08/07/google-memo-four-scientists-respond/

I think Damore's argument held up pretty well considering that he isn't a professional writer or intellectual and that we wasn't anticipating his argument to be subject to unprecedented public scrutiny.

It was certainly much stronger than the median rebuttal that came about in the aftermath and better than the silly economist piece that you linked.

It was a bit wrong and very confrontational. Fine for comments, less good at the day job. Any day job. Anywhere.

I think Damore’s argument held up pretty well considering that he isn’t a professional writer or intellectual and that we wasn’t anticipating his argument to be subject to unprecedented public scrutiny.

It was surprisingly well written and would have been reasonable for the National Review in which makes me think possibly:
1) The goal was to be fired and become a conservative tech hero.
2) Had assistance from somebody with experience in labor law. It was well written but simply crossed the line for acceptability with a corporate structure.

"It was surprisingly well written and would have been reasonable for the National Review in which makes me think possibly: 1) The goal was to be fired and become a conservative tech hero."

It's possible, but consider the wage differential in writing for NRO vs working at Google.

I meant for my preceding reply to be directed towards Milo, not you.

Good one soy boy.

Damore wrote a literature review. He was factually correct.

However, he was wrong in not noting that Google is in a very tight box. The US labor laws are insane. Google hires Asians far in exceed of their proportion in the public, and it hires slightly above proportion of males, but slight less proportion of males compared to the rest of Silicon Valley. Google has attempted to skirt their very large potential liability through brand, and through very extensive hiring policies, but they know that if certain political parties get their way that Google will be in great jeapordy. Due to this Google had no choice but to fire damore. This is a political fight and damore should have simply continued to draw his *substantial* salary and donated to the Republican Party.

Want to avoid diversity? Just make a really, really big show of how much you love drinking Diversity flavored Kool-Aid!

It's a great message for the rest of us who are socially clueless. When we hear "diversity" we incorrectly think it means something bad for whites and asians.

"bad for whites and asians"

WRONG.

Read the stats that Google published. It has nothing to do with whites. Whites are underrepresented. Yet our "media" doesn't pick this up one iota.

56% of Google is white, 53% in tech positions. The US population is 62% white.

Does the change your virtue signaling position, if so why? It shouldn't.

That’s interesting. I wonder how underrepresented white gentiles are.

That economist 'rebuttal' is total dross. It spends its time either A) asserting that Damore is wrong on his facts without providing any evidence whatsoever, or B) refuting what it *think* he inferred rather than what he actually said.

If thats the best job rebutting his points, then the left should hope no one with any reasoning power reads Damore or that economist article.

This is one reason why I asked yesterday if there were dead tree papers and magazines that are more balanced when it comes to “conservativism” than the Economist.

I’m not saying I’m conservative—whatever that is exactly—on all issues but I’d appreciate it if conservative views are given their best possible spin, and a fair shake. If only to improve the left’s arguments.

Has Google fired every single employee who made an objectively incorrect statement on its internal listserv? I somehow doubt it.

Companies screw this up all the time. Somebody makes a complaint, and management gets pissed off and decides to fire him. The next time he shows up 3 minutes late, they grin with glee and can his ass. "It can't be retaliatory - he showed up late, and that violates the rules! We had good cause to fire!" Well yeah, but unless you can show that you fired every single employee who ever showed up three minutes late, it sure as hell looks like you were singling him out.

As noted Damore was not wrong. Funny thing is how fast a HR manager, and later a labor board lawyer, feel free to correct a Biology Phd about biology.

https://www.wired.com/story/the-pernicious-science-of-james-damores-google-memo/

From the Wired article:

"That said, Damore’s assertion that men and women think different is actually pretty uncontroversial, and he cites a paper to back it up, from a team led by David Schmitt, a psychologist at Bradley University in Illinois and director of the International Sexuality Description Project."

Also from the article:

"The memo is a species of discourse peculiar to politically polarized times: cherry-picking scientific evidence to support a preexisting point of view."

Not really. Most of it is pretty well accepted. Among experts of course

That Wired story is laughable.

After calling his essay “dangerous,” the writer embarks on a long chin-stroking explanation that while Damore cited actual research, that research is not representative or applicable and therefore, yes, dangerous.

But wait, what does the writer present when calling upon their own favourite researcher?

“From infancy, boys get footballs and girls get dolls, so is it that surprising? We’ve been socializing them. It doesn’t mean there’s anything innate,” says Janet Hyde, director of the Center for Research on Gender and Women at the University of Wisconsin.”

You see, Damore is dangerous because he cites mainstream research, and he’s wrong because we found a complete utter who denies there are innate differences between boys and girls.

You need a heart of stone not to laugh at these people.

Hyde wants to deny innate differences between boys and girls, but Damore is a living refutation of her beliefs.

He’s autistic and autism affects boys between 2:1 and 16:1 more often than girls. And people on the austistic spectrum are over-repped in tech, ergo a partial explanation why more boys at Google.

You need to be overdosing on denial to argue otherwise.

If you see a journlolist dance around the subject using vague words like "pernicious", "problematic", "troublesome", "dangerous", "creepy", "ewww coooties", then you know they're playing games and don't dare to actually make a factual statement. This particular article is disqualified from its very link. If they have a case they should be bold in stating it and showing their work, not dancing around the subject while using association fallacies to instill enmity.

dance around a subject*, since I was speaking generally

#2: The podcast they reference in the article is Chetty's convo with Tyler, check it out here:
https://medium.com/conversations-with-tyler/raj-chetty-tyler-cowen-inequality-mobility-american-dream-d5ea7f4742b1

5. Cowen's friend Peter Thiel has parted ways with Silicon Valley. One will recall that Mr. Thiel made his fortune as a facilitator (he's a founder of both Pay-Pal and Palantir, the former in the business of facilitating internet sales transactions and the latter in the business of facilitating snooping by government intelligence agencies). I've been thinking about facilitators a lot this past week since Donald Trump's lawyer revealed that he "facilitated" a $130,000 payment to Stormy Daniels. It seems that facilitators are what drive today's economy, and politics. Mr. Thiel can be credited with having facilitated the election of Mr. Trump. Now that he's left Silicon Valley for Los Angeles, I'm curious what facilitating he will do down there. As for Trump's lawyer, his facilitating reminds me of the facilitating that is provided by guys named Vinnie.

....Trump’s lawyer revealed that he “facilitated” a.....

So much for attorney client privilege.

Back in undergrad I noticed there were two broad categories of economics students, comprised of a) eager students who were keen to learn the subject material ('strivers'), and the others ('slackers') who weren't necessarily lazier or less intelligent, but were only in the class because they just needed the hours.

Whereas math and physics departments have created two tracks with this in mind (think Calculus 2 "proper" vs. "Calculus 2 for business applications"), economics continues to try and cater to both groups in one track, to the detriment of both groups (although I sympathize more with the strivers).

It seems self evident that the quality of undergrad economics education would improve for the strivers under a two-track system (and likely the profession's reputation would follow), so I can't think of why econ departments haven't also pursued this route - a good theory of why that is would be interesting.

PS - my favorite STEM course ever was my Physics 2 lab - it started with making notes, recording evidence, and walking step by step through developing the theoretical model of electricity and magnetism - I think microeconomics 101 could be really improved by similarly exploring real-life (experimental?) empirical support as we progress through concepts, wouldn't it be great to discover how prices emerge from supply and demand through a real life experiment on the first day of class?

Actually, many schools have set up a separate track for the super ambitious econs -- like Northwestern's MMSS program which was originally to accomodate those shooting for grad school. But aside from places like MIT or Caltech where high math competence is assumed to be the norm, the majority of econ majors aren't interested in using econ as anything more than a steppingstone to business so it's hard to have two tracks. The future non-tech CEO's want the prestige of graduating in the same group with the quants but don't want to do the work. Econ's signalling value diminishes to the ambitious slackers (not an oxymoron) at most general schools by separating the tracks too strongly.

Is #4 just an item of general interest, or is there a Straussian message intended? Certainly one takeaway would be that genetic engineering for particular traits (like say height, sprinting ability, intelligence, conscientiousness) might be more difficult than some assume. What does the fact of pleiotropy in human genetics imply about heritability of desirable (and undesirable traits)? Anything?

It implies that genetic screening for most disorders won't work, at least not before the AI singularity.

The heritability is well-established by twin studies.

Biology is hard.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2782548/

1. This reminds me of Tabula Rasa, the young woman in my wife's law school tax class who, after the professor showed the class how to solve the computation of tax in the hypothetical presented, asked the professor if the numbers would always work out that way. Sure they will, one just has to find the answer one is looking for.

#7 excellent. What is it like to be so talented? The man is a treasure. Great shout out to Thomas Sowell. His work ethic is inspiring.

#5 no kidding. Conservatives/libertarians joke in dark conners of the valley, but at work they know to toe the line. It is frightening, but we stay since they pay is good.

4. Lol, twin studies.

2. Yawn. More signaling? Does he ever stop?

1. Clearly, it is applied math.

6. Caplin has been on a tear lately. His recent interview with Russ Roberts was also excellent.

Could you say a little more about why you wrote "4. Lol, twin studies." Is there a good piece you could link to explaining why twin studies aren't trustworthy?

7. While Taleb is a little bit off in his rhetoric, he explains again and again why Pinker does not understand the world. (When Taleb criticizes he is often near-autistic but both he and Pinker are at similar distances from the Scylla and Charybdis of enchantment by the cumulative effect of our own compiled memories of intellectual - intellectual, but incremental, which is what we forget - triumphs. Hence, being at a similar distance, he understands the criticisms Pinker needs to consider, as one mid-wit to another might understand, although of course they both are much more than mid-wits)

5. Sad! So many intelligent people and so quiet all the time! Even back in the day in Prague in the 60s you could tell who wanted to tell a joke at the expense of the zeitgeist.

4. I agree. In order to do actual "twin studies" that tell us something we did not know before, you have to be near the top of the profession. Otherwise the "statistics" are too overwhelming and confusing. So most "twin studies" are at about the level with respect to this real world what "Reader's Digest Condensed Novels" (issued once per season and now widely available for a couple cents and mailing charges on Amazon used-book store) were to the art of words back in the day.

if you buy one of those old Condensed Reader's Digest Novels from Amazon make sure to use a little rubbing alcohol to wipe the mold off, and probably you ought to keep them away from your other books anyway (something in the way they were manufactured, they seem like Reunion Day for mold from every decade from the 60s on - trust me).

(by the way, with respect to 4 - "I agree with the poster at 4:29" - (who, by the way, should not be terrified, not today, not yesterday, not tomorrow - the Lord is our shepherd and we should have no fear even in the valley of death, as they used to say in Pittsburgh back in the 20s, and a few other places-); that is, I agree with the sentiment: LOL, twin studies. 95 are misleading for every one, two, three, four or five that advance our understanding.

sorry I misspelled your name I have never encountered autocorrect in the box where one puts one's name before

1. Questions like "Is econ STEM?" are not very meaningful unless you specify for what purpose. It's like asking if a motorized wheelchair is a vehicle. Without knowing the purpose of the various visa programs, there's no intelligent way to answer the question.

1: Very obviously not.

Also, Happy US President's Day, everyone!

I wonder how Hillary Clinton is celebrating it.

Trump is not my preident.

Trump is not my president.

President Temer is Brazil's president with a public approval rating of about 6%, which doubled from its recent low! Things are really looking up for him.

President Temer is actually one of the most respected snd admired world leaders.

You mis spell three letter conjunctions?

The keyboard is bad. It is better than misspelling "misspelling".

"I wonder how Hillary Clinton is celebrating it." Sobbing into her gin, I expect.

No, she's calling all her friends and reminding them she won the popular vote.

Are you repeating Trump's mispunctuation for comedic effect?

It is true that Presidents' Day is a contaction of the Washington and Lincoln birthdays, and it is stangely self-interested for Trump to punctuate it as if it were about him instead.

https://twitter.com/styleguide/status/965656022143520768

#2. “For a policy nerd like me, being able to see that quantifiable evidence about things lots of us have been debating in theory for a long time is absolutely huge,” said Cecilia Muñoz, who led the Domestic Policy Council in the Obama White House. “The notion that you can use data about people’s social interactions and begin to piece together, ‘OK, what is social isolation actually costing us?’ is a whole other ballgame.”

Color me skeptical that data in the hands of people whose rational process is based on assuming the conclusion is going to produce much of value.

#7: As Walter Scheidel argues in The Great Leveler (another superb book of 2017), the most effective ways of reducing inequality are epidemics, massive wars, violent revolutions and state collapse.

Well, to be fair, that's partly because at least one or two of those things are in fact a necessary condition for egalitarians to achieve their desired outcomes.

2. Reads like a press release.

I like Thomas Sowell As a young teen/preteen I was an avid reader of his columns and enjoyed some of his books But when a leading light intellectual like Pinker says it's important to grapple with Sowell's ideas it's the thirteenth stroke of the clock

I had Sowell for Econ 1 at UCLA back in '74. He was the second worst teacher I ever had (the first was an alcoholic German teacher I had in high school.) And he was a dick as well. Sowell that is, not my high school German teacher

A really bad combination.

Oh, and he made us buy his book, not Alchian and Allen's which everybody else used and commanded a much better price on resale.

Sowell argues that civil rights laws hurt black Americans in the sense that they allowed black Americans to shop in white-owned stores, which decimated black-owned stores, and, as a result, black neighborhoods. Sowell is half right: black Americans could shop in white-owned stores but couldn't work there. Blacks were offered the promise of integration, but only integration that mostly benefited white folks. Sowell's solution is to return America to segregation.

Where did Sowell make that argument?

5) Diversity is just liberal jargon for, you can say things I agree with and you cannot say things I disagree with. Ask students at Berkeley if it is ok to punch a "Nazi". Then ask them if it is ok to punch someone from "Antifa".

Comments for this post are closed