Using the outcome as a conditioning variable

This is Joseph.

Felix Salmon has a great comment on Facebook and Female Coders:

To spell it out: The more your code is rejected, at Facebook, the less likely you are to rise up the ranks. So the fact that women suffer from significantly higher levels of code rejection is a big problem. The evidence for this being a problem is precisely the fact that Facebook’s female engineers are disproportionately found at lower levels rather than higher levels. 

And yet somehow, Facebook has contrived to use that fact in support of its claim that there isn’t a difference in how male and female engineers at the company are treated. Instead of treating the prevalence of men in the upper engineering ranks as prima facie evidence that there’s something amiss, they use it to exonerate themselves of sexism.

As Thomas Lumley notes, conditioning on the outcome rather ensures that the model is unhelpful (I found this article reading his most excellent site).

What would be interesting to see would be what would happen if you randomized some gender swapping on the code authorship.  Or were able to blind reviewers to the gender of the code author; that would likely have some pretty positive consequences.  Because another consequence of this finding is that either some sub-par code gets accepted or some good code is being rejected.  You would assume that a private company would want to correct this issue, either way.    

Josh Marshall has my favorite ready-to-mix metaphor of the whole sordid affair

As a fan of the extravagantly mixed metaphor, I wish Marshall had jammed these two together, but they are both good enough to stand alone, funny and remarkably apt.

Last night came word that Deputy Attorney General Rod Rosenstein reportedly threatened to resign Tuesday night when he realized that the White House was making him the key decision-maker in the firing of James Comey. This tells us a lot, but not necessarily what it might appear on first notice. The main significance is that no more than 30 hours in, the White House’s absurd cover story about firing Comey over his misdeeds toward Hillary Clinton – a lie that virtually everyone at the White House has now publicly repeated and vouched for – is coming apart at the seams. We can also see the staggering fact that after no more than two weeks on the job, Rosenstein’s public reputation, which was formidable, has been destroyed. He now joins a legion of Trump Dignity Wraiths, men and women (though mainly men) of once vaunted reputations or at least public prestige who have been reduced to mere husks of their former selves after crossing the Trump Dignity Loss Event Horizon.

…

What Rosenstein seems not to have realized was that Trump would blame him firing on him. To put it in mafia terms, ‘I said I’d help you whack Carlo. But you didn’t say you’d tell everyone it was my hit!’ This is why Rosenstein’s threat to resign rings hollow and indeed why I suspect he hasn’t resigned. What’s his argument? That he knowingly participated in the bad act and put his legal knowledge to work justifying it but is outraged that he’s being asked to take the blame?

And now our video accompaniment. First on the subject of event horizons. Before buying the Star Wars franchise, Disney first tried to make their own knockoffs. It did not go well.




Going from the ridiculous to the sublime... When the subject is mafia hits, particularly hits with nasty unexpected consequences, there really is no other choice.

Charles Pierce continues to hammer away at the differences between 1974 and 2017

We've been making the point for ages now that, while the legacies of Goldwater and Watergate are important, the analogies to today are deeply problematic. The differences are, if anything, more important than the similarities.

As noted previously, veteran political observer Charles Pierce was pointing out this contrast even as the Trump White House was preparing to announce the firing of Comey. The next day, Pierce made the same point even more forcefully. 

On July 25,1974, the House Judiciary Committee opened debate on the articles of impeachment. A congressman named M. Caldwell Butler said:

    "For years we… have campaigned against corruption and misconduct…But Watergate is our shame."

And a congressman named Lawrence Hogan said:

    "After reading the transcripts, it was sobering: the number of untruths, the deception and the immoral attitudes. By any standard of proof demanded, we had to bind him over for trial and removal by the Senate."

And a congressman named William Cohen said:

    "I have been faced with the terrible responsibility of assessing the conduct of a President that I voted for, believed to be the best man to lead this country, who has made significant and lasting contributions toward securing peace in this country, throughout the world, but a President who in the process by actor acquiescence allowed the rule of law and the Constitution to slip under the boots of indifference and arrogance and abuse."

And a congressman named Thomas Railsback said:

    "I wish the President could do something to absolve himself."

And a congressman named Walter Flowers said:

    "This is something we just cannot walk away from. It happened, and now we've got to deal with it."

All of these congressmen voted to send the articles of impeachment to the full House.

On August 7, 1974, two senators named Hugh Scott and Barry Goldwater, along with a congressman named John Rhodes, went to the White House and told Richard Nixon that his removal from office was inevitable. Nixon resigned the next day. Now, looking back from the swamp in which we currently find ourselves, there is one remarkable thing about all the people whose actions in that perilous time showed what stuff of which they and the country were made.

They were all Republicans.

Every damn one of them, from Sirica to Goldwater and back again. They all did their duty, as best they saw that duty and, as a result, a Republican president was forced to give up an office he'd won in a landslide only one year earlier.

"Why doesn't anybody have enough anymore?"

One of the questions we've been discussing on and off for a while now is how the Republican Party has changed since Watergate, and what the implications of those changes might be as the Trump administration continues a seemingly inevitable spiral of scandal. With recent events, the question is even more relevant.

A few hours before the Comey firing, Charles Pierce (who knows a thing or two on the subject) provided us with and illustrative clip.




And a characteristically sharp commentary.

That's Senator Lowell Weicker, Republican of Connecticut, telling John Ehrlichman, a thug in the employ of the Nixon White House, that, whatever alibis Ehrlichman had concocted to excuse himself and his boss for their blatant criminality, Weicker wasn't having it any more. This was in July of 1973, a full year-plus before Nixon finally was run out of power. At this point, there was no telling whether the White House stonewall was going to hold or not. (The number who knew was pretty much limited to the guilty, and the shell-mouthed operatives of the Watergate Special Prosecutor's office.) Being an independent-minded Republican still carried a considerable political risk. Weicker did not care. He'd had enough. Why doesn't anybody have enough anymore?

Information from hacking scandals

This is Joseph.

I assume everyone has read Nate Silver's take on the Comey letter by now.

In that context, Josh Marshall has a very good set of points as to how information leaks hurt the recent US presidential election:

Second: The FBI Director broke all precedent and DOJ guidelines to announce a criminal investigation into what proved to be the losing candidate just over a week before the election. There was little reason to believe the purported new evidence would lead to any criminal charges or indeed even any substantial new evidence. And it turned out that the ‘investigation’ was based on nothing. The entire blow up turned out to be based on nothing and knowing what we know now about what investigators and Comey knew at the time suggest he had little reason to think there was anything there.

Third: A rival foreign power ransacked the computer files and email logs of the losing candidate and strategically leaked them out over the final months of the campaign with the intention and the effect of distracting and damaging what proved to be the losing candidate.

Now look at France and their episode of hacking.

This suggests to me that we need to move past the nice but naive idea that leaks can ever be non-political.  The idea of more information always being good isn't false, but we need to acknowledge that a selective information leak affecting one party is a very partisan act.  It isn't like we saw the emails and logs of both sides being released so that people could use "behind the scenes" information to make a more informed electoral choice.

Instead, the decision to hack or leak data should always be filtered via a question of what is the agenda of the person(s) who are releasing this data.

Trying to imagine the pitches that get turned down...

One of these days, we need to have a serious conversation about all of the Soylents and the Juiceros. We need to think about the opportunity costs of diverting hundreds of millions (billions?) into obviously laughable products and services. We need to consider the income inequality implications of a business culture (particularly in Silicon Valley) where having the right connections counts for more than viable business plans. We need to ask credulous journalists how they justified the shameless puff-pieces that invariably preceded the collapses.

For now though, I'm just here for the snark and schadenfreude.

Silicon Valley Can’t Stop Puking Money All Over Soylent by Adam Clark Estes

Soylent, a substance, is about to be everywhere. A team of queasy venture capitalists just invested $50 million into the company. This, despite the fact that Soylent is perhaps best known for lying about its ingredients and giving people fits of relentless vomiting and uncontrollable diarrhea.
...

The new round of funding is led by GV (formerly Google Ventures) and includes big shot funds like Andreessen Horowitz and Lerer Hippeau Ventures. That extra $50 million brings Soylent’s total pile of cash from investors up to $74.5 million. Rob Rhinehart, the original startup bro who hates food and invented Soylent, told Bloomberg that he soon hopes to sell Soylent “pretty much anywhere you can get a coffee.” That means you might win more opportunities to spend money in order to celebrate the dismal, dangerous future we’re all living in, whether we like it or not. Let’s hope those coffee shops have bathrooms, though. We’re gonna need it.

Preexisting conditions

This is Joseph

It looks like preexisting conditions may be back as a way to increase the costs of health insurance, and this is a bad thing.  Why?

In traditional insurance, if your house burns down then the insurer makes whole the loss.  In medical insurance, a major medical event may have sequels for many years.  If you change insurers (or your insurer goes out of business) then the original insurer is off the hook for continuing costs.  The new insurer looks at the ill participant as a "house already on fire" and would like to not cover these ongoing costs.

There is also an information issue.  If a house burns down twice then that might just be bad luck.  But if a person has two heart attacks, is the second one a random event or a sequel of the first one.  It is pretty clear that we don't want courts adjudicating this question on a case by case basis (or people being denied care because nobody can figure this out).

The most logical solutions are regulatory (like the Affordable Care Act implemented) or structural (like implementing single payer health care).  A portable system of health insurance might also work, although do you want to bet on your insurer being solvent for forty-five years?:

But there is a reason that preexisting condition clauses are unpopular.  Not because people want a free lunch (although some of that may always be present) but because it opens people up to random events (employer bankruptcy, anyone) leaving them unable to afford health care.

And before anyone talks about market solutions, please try and actually use "self-pay" as an option in a modern medical setting.  I actually tried this and it was very difficult to be seen by an MD (and only MDs can do things like prescribe antibiotics -- which can be life saving).

I have a small confession...

 For an embarrassingly long time, I thought that "epitome" and ep-i-tome were two different words.

The Bret Stephens story is about more than climate change denial

The conversation has gotten muddled and off-topic on both sides (which is not to say that both sides are equally wrong, just that most of the critics of Bret Stephens are making a valid case badly). A great deal of the discussion has come to center around a not particularly relevant debate about whether or not certain positions are acceptable in the pages of a publication like the New York Times.

It is true that, though the exact boundaries are inevitably hazy, there are certain positions that are so innately offensive that they should have to clear and extremely high bar before making it into the public discourse. Two obvious examples are defenses of the Holocaust and slavery. There is, of course, and inherent conflict between condemning offensive ideas and defending freedom of expression, but that's not really what we should be focusing on here.

The issue here is not that Stephens took a position (or even a string of positions) that you or I strongly disagree with; the issue is that he took this string of indefensible positions (effectively taking the pro side on things like racism, torture, religious bigotry, income inequality, sexual harassment, and the war on data) and defended them by recycling tired and hackneyed arguments that are logically flawed, dishonest and/or incoherent , not to mention almost always badly written.

We are not talking about lapses in an otherwise outstanding career or bad traits that are counterbalanced by notable strengths. Bret Stephens did not succeed despite these things; his entire career was built on being an apologist hack.

Here is an illustrative example from an excellent summary compiled by Hamilton Nolan;

The campus-rape epidemic—in which one in five female college students is said to be the victim of sexual assault—is an imaginary enemy. Never mind the debunked rape scandals at Duke and the University of Virginia, or the soon-to-be-debunked case at the heart of “The Hunting Ground,” a documentary about an alleged sexual assault at Harvard Law School. The real question is: If modern campuses were really zones of mass predation—Congo on the quad—why would intelligent young women even think of attending a coeducational school? They do because there is no epidemic. But the campus-rape narrative sustains liberal fictions of a never-ending war on women.

When you strip away all of the nonessentials, the underlying claim is that an institution or group cannot be accused of prejudice or harassment if the victims willingly choose to join. This is very closely related to the classic argument: "you don't have to be here."  For a long time, this was one of the default responses to charges of discrimination. It was applied to African-Americans trying to break the color barrier, women entering traditionally male professions, gay athletes trying to be open about their sexuality, you name it.

Stephens is careful to couch his arguments in terms of some shadowy liberal conspiracy rather than an attack on victims of sexual assault, but when you claim that a woman's decision to attend an institution or pursue a profession precludes the possibility of a culture of harassment and assault, you have made this an argument about women, one that is neither sound nor original.

The New York Times knew what it was getting  with their latest hire and they have been full-throated in their defense of the choice. Stephens is a tired, derivative hack of no apparent insight or talent (at least, David Brooks is good at being David Brooks). There is nothing to be learned from reading his new column, but their is a great deal to be learned from the way that America's best known "liberal" paper instinctively genuflects to the worst of American conservatism.

[Note: some typos have been corrected.]

Just substitute "dropping out of college" for "buying lottery tickets" and Peter Thiel for the stick figure...

On a related note, I've been arguing for years that a winning lottery ticket is the world's best investment (though there is that one little catch).

PhD Vouchers

This is Joseph.

Frances Woolley asks about why we don;t have vouchers for PhD programs.  As a part of this discussion she points out the issues of misalignment of incentives for K-12 school vouchers:

It's odd: school vouchers are frequently advocated for primary and secondary school students. Yet there are good reasons why they would not be expected to work well, especially for primary school, as the ones exercising the vouchers (the parents) are not the ones experiencing the education (the children). Also there are very large costs associated with switching schools, and especially going to a school outside your immediate neighbourhood, if you are six years old and not able to drive, or 12 years old and in a tight friend network. Hence the possibility for effective competition between schools is limited at the K to 12 level. 

I think a lot of the same issues would apply to PhD vouchers as would apply to school vouchers, especially in terms of the issues of coordination. So I am not a fan of the PhD vouchers idea, doubly so when there are provincial funding differences for the schools themselves.  I can see too many ways that we could end up with all viable students being at the University of Toronto, for example, with a clever use of network effects.

But the real issues are the barriers to change.  These are high -- just differences in the material that is being taught can be brutal to overcome when changing mid-semester.  Market forces will always be inhibited by the challenges in overcoming these barriers, especially since the service is being marketed to a proxy for the consumer.

But my biggest question is why the economist can rattle off real issues that don't even appear in the current conversation.  Everything is about quality, but this all presumes that the new system (at scale) will be better than the old system.  Which might be true, but pure market forces will suffer an uphill battle given transaction costs.  After all, how do you prevent "we will just under-serve a little bit, but not enough to make the costs of changing schools worthwhile" becoming a "race to the bottom" for most schools (with a few high priced and elite exceptions).

ESPN and the content bubble

It is easy to forget that, though it does occupy a unique niche, sports is part of the entertainment economy.

For years now, we have been discussing the increasingly apparent content bubble. Even though people's ability to consume content is approaching saturation (even taking into account multiple screens) and the competing sources of content are exploding, prices and production slates continue to skyrocket. This is not to say that there isn't a lot of money to be made or that the amount will not continue to grow, just that the probable rate of growth is nowhere near what would be needed to justify the current level of investment and hype.

We have mainly focused the discussion on things like scripted television and streaming services, but the recent news from the cable sports industry might provide a better example:

ESPN's Diminished Future Has Become Its Present by Kevin Draper

The causes of the layoffs are clear. As ESPN’s subscriber base, and the rate those subscribers paid monthly, grew in the late aughts and early 2010s, Bristol spent flagrantly. They created the Longhorn and SEC Networks, built a massive new SportsCenter studio, hired hundreds of writers to cover specific teams, and, most importantly, spent billions of dollars on live sports rights. They made big bets. They made wrong bets.

Right around the time the ink dried on a $15.2 billion deal to broadcast the NFL, subscribers began fleeing cable television in droves—not because of anything the Worldwide Leader did wrong, but because of secular changes in the way broadcast and video works. Phones, Twitter, and YouTube began instantaneously delivering highlights and entire games to fans, obviating the need for anyone to watch SportsCenter, or any other news shows, to catch up on what happened in sports, or even, in some cases, to watch live games. Terrestrial ad revenue never migrated online, and the revenue to be found there was largely eaten up by Facebook and Google, leaving little to pay those new ESPN.com reporters.

ESPN is still wildly profitable—the operating income of Disney’s media networks (of which ESPN plays the largest role) was $1.36 billion in the 2016 fourth quarter—but it’s less profitable than it used to be, and projects to be far less so in the future. With its latest cuts, ESPN isn’t just trying to stanch the bleeding and/or to be seen by investors as attempting to do so: They’re also laying out what the network will look like over the next five years and beyond.

…

If ESPN is trying to significantly trim costs, things are going to get grim, because cutting the salaries of online writers isn’t going to cut it. And so the fundamental question is how long ESPN—or Disney, or Disney shareholders—can be content with diminishing profits, and at what point they admit that aggressively outbidding competitors for live rights at the peak of what was at the time clearly a bubble was a mistake. If they do so, the knock-on effect to the leagues that rely upon their money to pay salaries and fund operations will be immense.

No-excuse charters and collateral damage

From Valerie Strauss writing for the Washington Post:

“College or Die.”

That’s the motto of the Charles A. Tindley Accelerated School, a charter school in Indiana which, according to its website,  “expects 100% of its students to be accepted at a fully-accredited four-year college or university” and “to achieve exceptionally high levels of scholarship and citizenship.”

The words “College or Die” are posted in giant letters in a hallway of Tindley, an open-enrollment charter school for grades six through 12 that opened in 2004 in a former grocery store in a low-income area of Indianapolis. It became well known in school reform circles when it was visited in 2011 by then-Indiana Gov. Mitch Daniels (R) and then-U.S. Education Secretary Arne Duncan, who heaped lavish praise on the school for its success in getting students into college.
...

Here’s what it doesn’t say: A lot of students leave the school before they get to senior year. Here’s an enrollment chart from the Indiana Department of Education:

As noted by educator and blogger Gary Rubinstein, Tindley had 93 students in ninth grade in 2013-2014. By the time that cohort got to 12th grade, only 40 students were in the class. That’s a loss of 57 percent. Such a big rate of attrition is not exclusive to Tindley; a number of charter schools, especially of the “no excuses” variety, lose a lot of students and don’t replace them. Students who can’t cut it have to find another school to take them, sometimes in the middle of a school year.

One of the points some of us have been raising for years now is that movement reformers lack adequate concern about the collateral damage of their proposals. It isn't just that many of the schools that reformers hold up as models have disturbingly high attrition rates; it is that (as both a motivational and a PR tool) these programs build themselves up as students' best and even sometimes last hope for escaping poverty and having a good professional life.

To say that this gets families' hopes up is a grotesque understatement, but arguably even worse, the kids who are unable to make it into the programs or complete them are essentially told that they are doomed to failure. Between the emotional damage, the disruption and the danger of a costly self-fulfilling prophesy, that is a hell of a toll to inflict on already disadvantaged kids.

That's right, a Rube Goldberg machine as narrative medium

From Gizmodo's Andrew Liszewski:

Biisuke Ball’s Big Adventure is actually a sequel to an earlier Rube Goldberg machine featuring the adventures of Biisuke, Biita, and Biigoro, three colored balls who somehow have more personality than most action stars. This adventure, which involves sneaking through camps and foiling traps, only plays out for about three-and-a-half minutes, but if Hollywood is reading this, we’d gladly sit through two hours of this in a movie theater.

All of the great ones make sacrifices

I'm planning to come back and connect this to some larger points, but for now I decided to get a quick post in to beat the Gizmodo rush.

From CNBC [emphasis added]:

In fact, a crucial decision Elon Musk was forced to make in 2010 when, by his own account, the billionaire was broke, is one of the reasons Musk has been able to cash in on Tesla's rapid share rise this year: Musk held on to shares at the very moment when a sale to raise cash would have made financial sense.

Musk, who had $200 million in cash at one point, invested "his last cent in his businesses" and said in a 2010 divorce proceeding, "About four months ago, I ran out of cash." Musk told the New York Times' DealBook at that time, "I could have either done a rushed private stock sale or borrowed money from friends."

It's a dilemma that many entrepreneurs face, but there is a big difference between the options available to Musk and the options available to most business owners. Musk was able to live on $200,000 a month in loans from billionaire friends — while still flying in a private jet — rather than sell any of his Tesla stake. Though the root of the problem is the same: intangible assets or, in other words, a business owner who is "asset rich" and "cash poor." And it can lead business owners to the most difficult decision of all: having to sell a piece or even all of their company.