Tyler Cowen points me to this from the Economist:

Most trading in bitcoin takes place in China: Huobi and OKCoin, two Chinese exchanges, are thought to account for more than 90% of transactions. The currency seems to have become an outlet for Chinese savers frustrated with their limited investment options and searching for high-yielding assets. The Chinese authorities are worried enough to have banned banks from dealing in bitcoin, but individuals are still free to speculate and have been doing so with gusto.

....China has also become the global hub for bitcoin mining, the process by which heavy-duty computing power is used to process transactions involving bitcoin, earning those doing the processing some new bitcoin as compensation. Over 80% of new bitcoin are now minted in data centres in places like Sichuan and Inner Mongolia.

One of the selling points of e-currencies like Bitcoin is that their decentralized nature makes them inherently free of government meddling. But is that really true? I've long thought that techno-evangelists show far less respect than they should toward meatspace assets like nuclear bombs, gun-wielding police forces, ownership of fiber optic networks, vast fortunes in physical goods, and so forth. This is, for example, why so many of them were naive enough back in the 90s to believe that the internet would spell doom for traditional marketing—only to wake up a few years later and discover that traditional marketers had adapted remarkably quickly to their supposed revolution. It turned out that high IQs aren't limited to Silicon Valley, and that websites and Google searches and Facebook advertising posed no more of a challenge to the existing order than television did in the 50s.

So is Bitcoin really safe from government meddling? It has been so far, but only in the same sense that an ant is safe from my boot as long as it doesn't annoy me. China, however, has already proved that a meatspace government can, in fact, crush the digital world if it's sufficiently motivated to do so. It's not even all that hard. So if e-currencies are now mostly a ploy for evading Chinese capital controls, I'd say we're about to learn pretty quickly whether (a) e-currencies can grow big enough to matter, and (b) national governments are truly helpless to do anything about them. I'll put my money on the meatspace men in Beijing if push ever comes to shove on this.

UBI Continues To Be Wildly Unpopular

The concept of an Unconditional Basic Income has become a hot topic on the interwebs. Conservative Charles Murray started things up in 2006 with the publication of In Our Hands, which created a brief stir and then sank into oblivion because (surprise!) conservatives were distinctly uninterested in cutting unconditional checks to lazy welfare bums.

Then it came out of hibernation a few months ago for reasons that escape me. At the time, I vaguely figured that much of the credit belonged to Vox's Dylan Matthews for his tireless advocacy of a UBI. It also, of course, had something to do with the explosion of Bernie mania. Bernie doesn't actually support a UBI, but he's said that it's "something that must be explored" and it pretty plainly fits into his general worldview. Nonetheless, after another 15 minutes of fame, it went into hibernation again. But it refused to die, emerging from its lair yet again a few weeks ago. Considering the fact that a UBI has less chance of being enacted than a law guaranteeing everyone a pet unicorn, this is a little odd. What's going on?

I'm still not sure, but much of it was probably due to an upcoming UBI referendum in Switzerland, engineered by the lefty Swiss community a couple of years ago. Today, after months of anticipation, they finally voted on it—and the results weren't pretty. Even in the heart of social democratic Europe, the mere concept of a UBI1 ended up with only 23 percent approval. It's still in pet unicorn territory.

But eventually it will become reality. We just have to wait for the robot revolution to evolve to the point where lots of middle-class white people are permanently put out of work. Then it swiftly will go from pet unicorn to "duh." I imagine this transformation will take a surprisingly short time and will happen sometime around 2030 or so.

1Despite endless headlines suggesting the Swiss were voting on payouts of $2,500 per month, the actual text of the initiative directs the Swiss parliament only to enact a UBI that "shall enable the whole population to live in human dignity and participate in public life." The actual level and financing of the UBI are not specified.

We already know that Donald Trump thinks Judge Gonzalo Curiel is biased against him because he's "Mexican." But what about other judges with non-white backgrounds? John Dickerson asked him about this on Face the Nation today:

Later, when asked if he believed a Muslim judge would treat him unfairly because of another controversial proposal to temporarily bar Muslims from entering the U.S., Trump replied: "It's possible, yes. Yeah. That would be possible, absolutely."

"Isn't there sort of a tradition though in America that we don't judge people by who their parents were and where they came from?" Dickerson asked. "I'm not talking about tradition," Trump replied. "I'm talking about common sense, okay? He's somebody, he's proud of his heritage."

OK then. No Hispanics and no Muslims. I wonder which non-white ethnicities are allowed to pass judgment on Trump? He's been pretty rough on China and Japan, after all. And Trump was king of the birthers a few years ago, so blacks probably don't think much of him. This brings up an obvious question:

When questioned on whether he would instruct his lawyers to ask that Judge Curiel get thrown out of the Trump University case, Trump said: "Well, I may do that now—We're finding things out now that we didn't know before."

"Because of his Mexican heritage though?" Dickerson pressed. "No, but because of other things," Trump responded. "I mean because of other things."

Hmmm. He wants Curiel tossed off the case, but not because of his Mexican heritage. Why is that? A reader points me toward a piece by Garrett Epps in the Atlantic this morning about a 1998 case overseen by federal judge Denny Chin:

Eventually, Chin dismissed Klayman’s client’s case....Not long after, the judge got a letter from Klayman and his co-counsel, Paul Orfanedes, asking a few “questions” about the judge’s Asian American background....In a written response, Chin...lowered the boom. Klayman and Orfanedes were required to withdraw as counsel from the case and would not be permitted to appear in Chin’s court on any matter ever again. They would be required to show his opinion to any other judge in the district in any future case. The court clerk would also report the sanctions to every court where they held bar membership.

....The Second Circuit briskly affirmed Chin’s order. “Courts have repeatedly held that matters such as race or ethnicity are improper bases for challenging a judge's impartiality,” wrote the chief judge, Ralph Winter, a Reagan appointee.

In public, Trump can rant about anything he wants. But in court, if his lawyers so much as mention Curiel's Mexican heritage in a recusal motion they risk nuclear sanctions. Even for Trump, they aren't willing to do that.

Still, there are always those "other things." My own guess is that this is a blustery Trumpian fiction, just like all the evidence of Barack Obama's Kenyan birth that Trump insisted his private investigators had been digging up back in 2011. We'll see.

Bottom line: Donald Trump apparently believes that the only judge qualified to try his case is a white Christian. I guess this is the new, more presidential Trump that his backers keep insisting will show up any day now for the general election.

I recommend that the following two words be officially removed from the English language:

  • Neoliberal
  • Fascist

The purpose of words is to facilitate the communication of thoughts between human beings. I have conducted exhaustive research1 on Twitter and other popular social networks which conclusively proves that these two collections of phonemes now do the opposite of that. They are therefore not words. Let's stop using them.

1Forthcoming in the Journal of Demotic Memes and Dialectics.

The Guardian claims that "at least 33" large American cities use testing methods deliberately designed to undercount the presence of lead in tap water:

Of these cities, 21 used the same water testing methods that prompted criminal charges against three government employees in Flint over their role in one of the worst public health disasters in US history.

....Testing methods that can avoid detecting lead include asking testers to run faucets before the test period, known as “pre-flushing”; to remove faucet filters called “aerators”; and to slowly fill sample bottles. The EPA reiterated in February that these lead-reducing methods go against its guidelines, and the Flint charges show they may now be criminal acts.

....The EPA has warned since 2008 that pre-flushing is problematic and goes against the “intent” of regulations designed to detect lead....Further distortion is achieved through the removal of “aerators” — the small metal filters at the tip of faucets. These filters can collect lead particles and add to lead detected in tests.

I don't know how serious this is. I suppose no one will know until these cities collect data properly and compare it to their old results. The EPA says it plans to release new testing rules in 2017.

Muhammad Ali and the Abuse of Ellipses

In February 1966, Muhammad Ali said:

I ain't got no quarrel with them Viet Cong.

In March 1967 he said:

My conscience won't let me go shoot my brother, or some darker people, or some poor hungry people in the mud for big powerful America. And shoot them for what? They never called me nigger, they never lynched me, they didn't put no dogs on me, they didn't rob me of my nationality, rape and kill my mother and father.

In popular culture, this has become:

I ain't got no quarrel with them Viet Cong...They never called me nigger.

I have to say that this is a pretty breezy employment of ellipses. Using them to indicate the passage of a few sentences? Fine. Using them to indicate the passage of 13 months? I have to cry foul on that, no matter how good it makes the quote.

Quoctrung Bui of the New York Times writes today about perceptions of massive unemployment among young college grads:

We asked: “What would you guess is the current unemployment rate for four-year college graduates between the ages of 25 and 34?”...The most common answers for college graduates were between 20 and 30 percent. Perhaps an understandable mistake....But what surprised us was that the majority of people thought that unemployment rates for those with college degrees were higher than for those without.

....We posed the same question to our friends and parents. Many have college degrees themselves; some are educators. They, too, mostly guessed that college graduates would be more likely to be unemployed than nongraduates....We ran the quiz one last time with the same question and anchor, structured as a multiple-choice quiz. This time, nearly half of the people in the survey guessed that college graduates had higher unemployment rates. We had to concede that we weren’t witnessing a mirage.

Are we — the news media — to blame?

Yes! Yes you are!

But I'll cut you some slack. The range of 20-30 percent seems to be the American public's go-to guess for just about everything in the news. What's the percentage of gay people in the US? 20-30 percent. The inflation rate? 20-30 percent. Illegal immigrant population? 20-30 percent. Amount of the federal budget dedicated to foreign aid? 20-30 percent. Bird deaths from wind turbines? 20-30 percent.

As near as I can tell, anytime something becomes familiar enough to intrude on the public consciousness, it falls into the 20-30 percent trap. That seems to be the all-around perception of "a smallish but still newsworthy amount."

That said, the news media still shares a lot of the blame for this, because they're the ones who collectively decide how much to cover stuff. By over-covering the alleged employment woes of college-educated millennials, they encourage people to think the problem is worse than it is—and they distract attention from where the problem really is. The truth is quite different: even at the height of the Great Recession, the unemployment rate of college-educated millennials never cracked 5 percent other than momentarily. It was young high school grads who suffered from astronomical joblessness:

But wait! Maybe college grads got jobs, but they were all crappy jobs that paid peanuts. Not really. College-educated millennials took a beating during the Great Recession, just like everyone, but rebounded to their 2003-05 level after three years and have rebounded even further since. Young high school grads, by contrast, are still making about 10 percent less than they did in 2003-05:

(This is from Census table P-28 here if you feel like checking it out yourself.)

College-educated millennials get all the attention, but that's not because they have it so bad. It's largely because they loom large in the minds of the press corps—who are all college educated themselves—and because they're verbal enough that they write a lot about themselves. High school grads, not so much. But they're the ones who were really hit hard by the Great Recession.

Muhammad Ali's Fist — Life Size?

This is from the great Pictures on a Page, by Harold Evans. Originally published in Esquire in 1974.

But is it really life size? On the printed page, yes it is. On the web, who knows? It all depends on how your device scales it. That's something that goes missing in the digital world. For the record, "life size" in this case is 173 mm wide from the left margin of the picture to the right. If you want to compare your fist to Ali's—and yes, it really is kind of irresistible—zoom in and out until that's how big the image is.

Weekly Flint Water Report: May 21-27

Here is this week's Flint water report—late but not forgotten, due to some kind of weird screw-up at the Michigan DEQ site. First they didn't have the data. Then eventually they did, but the spreadsheet was unreadable. Then they took everything down entirely. Then they finally reposted last week's data. As usual, I've eliminated outlier readings above 2,000 parts per billion, since there are very few of them and they can affect the averages in misleading ways. During the week, DEQ took 392 samples. The average for the past week was 11.10.

For some reason I got curious about the whole Trump University thing this afternoon, so I started googling sort of randomly to learn more about it. I came across several interesting items that seemed worth sharing, but I don't really have any special narrative to put together about them. Instead, here's a semi-random three-part Trump U dump. Be sure not to miss footnote 2!

1. The Evolution of Trump's Racist BS Against Judge Curiel

A few days ago the whole world suddenly went ballistic because Donald Trump went after the judge in the Trump U case. But this is nothing new. Three months ago he railed against judge Gonzalo Curiel at a rally in Bentonville, Arkansas: "We have a very hostile judge....tremendous hostility, beyond belief. I believe he happens to be Spanish, which is fine, he's Hispanic, which is fine....but we have a judge who's very hostile." The next day he told Fox News Sunday: "I think it has to do perhaps with the fact that I'm very, very strong on the border, very, very strong at the border, and he has been extremely hostile to me."

Curiel has since gone from Spanish to Mexican to "of Mexican heritage," but Trump has basically been peddling this racist BS for months. So why is it that no one seemed to care much about it back in February, when it might have mattered?

2. The 98 Percent Satisfaction Rate

Back in November Steven Brill wrote a really good piece about Trump U for Time. If you have any interest in this case, you should put aside a few minutes to read it.

One of my favorite bits is about Trump's most treasured defense, which is that all the suits against him are the work of a tiny number of malcontents. "Listen to me," he said in February, "98 percent of the people that took the courses, 98 percent approved the courses, they thought they were terrific." Trump has a whole website devoted to this, 98percentapproval.com. Here is Brill:

Trump’s director of operations Mark Covais...declared that the satisfaction percentages were taken from “about 10,000” surveys of Trump University customers. Yet in the same affidavit Covais said that there were 7,611 tickets sold to Trump University programs....How could Trump have 10,000 “rave” surveys from paying customers if there were only 7,611 paying customers?

....The more apparent inconsistency is that Covais...declared that the company had issued 2,144 refunds to 6,698 attendees of the $1,495 three-day program, or 32%. That a third of the customers demanded refunds is hard to reconcile with a claimed 98% satisfaction rate....Similarly, the refund rate for the $34,995 program, which according to the lawsuits was tougher on giving money back, was 16%. If at least 31% of one group and 16% of the other were so instantly dissatisfied that they immediately demanded refunds, how could 98% have been satisfied?

How indeed? Perhaps it's just the usual Trump "puffery," his second-favorite defense for the lies he told about Trump U?

3. Is Judge Curiel Biased Against Trump?

I Am Not A Lawyer™—about which more later—but I was still curious about whether Judge Curiel has, in fact, ruled against Trump overwhelmingly. So I googled a bit until I got bored. Here are the rulings I could find:

March 1, 2014: After years of legal maneuvering, Curiel allowed two cases against Trump to go forward. "Curiel certified one case, Tarla Makaeff vs. Trump University, as a class-action in California, Florida, and New York. The plaintiffs had hoped it would get certified in all 50 states. Curiel also narrowed down the pursuable actions from 14 to 5. In a second related case, Art Cohen vs. Donald J. Trump, Curiel denied Trump's motion to dismiss the case." I'd score that as two anti-Trump rulings (both cases were allowed to move forward) and one pro-Trump ruling (the class-action was limited to three states).

October 29, 2014: Curiel certified Cohen's suit as a class action. That's one anti-Trump ruling.

April 20, 2015: Trump countersued the plaintiff in one of the cases, Tarla Makaeff, and she eventually won a motion to dismiss Trump's suit as a nuisance.1 She asked for $1.3 million in costs, and the case was sent to Curiel for resolution. "Curiel found the rate requests for associates and partners reasonable, but denied Makaeff's request for staff attorney and paralegal fees....Curiel scaled back the number of hours expended on most of the plaintiffs' 25 proceedings by 20 percent, others by 50 percent and some he declined entirely, reducing the fees award by $542,920.85 to $790,083.40." That sounds fairly pro-Trump.

July 3, 2015: Curiel ruled that Trump would have to testify about his net worth: "Publicly available figures of Trump's wealth have been the subject of wild speculation and range anywhere from $4 to $9 billion," Curiel ruled. "Simply stated, plaintiffs are entitled to answers made under penalty of perjury." I think we can safely call that one anti-Trump.2

November 18, 2015:  Trump lost a bid for summary judgment throwing out the suits. However, Curiel ruled in his favor against an injunction that would have prohibited further false advertising. That's basically one ruling against Trump and one for him.

March 22, 2016: Curiel allowed Makaeff to withdraw as a lead defendant in her case. He dismissed Makaeff's claims and denied Makaeff's request to bar Trump from any further litigation he may file against her. He allowed the defense to re-depose the other main plaintiff, Sonny Low, and ruled that Trump would "likely be entitled to some award of fees and costs." That's one ruling against Trump and four in favor.

April 26, 2016: Curiel issued an order setting a July 18 hearing on a motion by Trump's lawyers to toss out the Cohen v. Trump case. "The judge also plans to take up other motions, including one asking that the case be stripped of its class-action status." This seems neither pro nor anti-Trump.

May 6, 2016: Curiel set a November 28 date for one of the suits. "[Trump's lawyer] asked for a trial after Inauguration Day on Jan. 20, but the judge raised concerns about distractions if Trump wins the election."

May 27, 2016: Curiel ruled in favor of a Washington Post motion to unseal several exhibits in the Trump U lawsuits. This is the ruling that ignited Trump's most recent hailstorm of attacks on Curiel.

I dunno. Does this pattern show any kind of bias for or against Trump? Trump seems thoroughly convinced that the whole case against Trump U should have been tossed out on summary judgment instantly, and the fact that Curiel didn't do that means he's "tremendously hostile." Of course, that's Trump's view of anyone who does something Trump doesn't like. But summary judgment was never very likely in these cases, and aside from that Curiel seems to have ruled for and against Trump fairly evenly. Hell, even Trump's own lawyer blew off Trump's rantings: "He's got very strong views about everything and he expressed his own views," he said, before acknowledging that, no, he didn't plan to ask Curiel to recuse himself.

What's more, a local TV station interviewed a San Diego lawyer who pointed out the obvious: "Judge Curiel gave lots of his rulings before Trump made those comments about the border and illegal immigration." And: "NBC 7 checked federal court records and confirmed Von Helms' claim." Curiel has been on these Trump U cases off and on since 2010, and obviously Trump's views on a border wall couldn't have influenced him before June of last year.

Now, I realize this is all sort of pointless. I think we all know perfectly well that Curiel is just an ordinary judge, and Trump is ranting against him because that's what Trump does whenever something doesn't go his way. He whines. Endlessly. Still, I'm kind of curious. It would be interesting if some kind of qualified lawyer type went through the records of these trials and reported back on whether Curiel seems to be conducting things fairly. Maybe he's not! Maybe he really does hate Trump. Unfortunately, I suppose that would be a lot of work. Oh well.

1But only after an appeals court overruled District Judge Irma Gonzalez, who had ruled in Trump's favor. Another Hispano-Mexican judge!

2Wait. What? Trump was required to testify under penalty of perjury about his true net worth? Yes indeed. And I believe he's already given that deposition. Oddly, though, he's never mentioned that, and hasn't brought it up in any of his recent ravings either. I imagine there's a good reason for that: If this becomes common knowledge, someone will ask him what number he provided under oath. Does he really have any good excuse not to share that with us?