The Gates Foundation has spent about $200 million since 2010—in addition to other sources who kicked in about $400 million—on an education initiative designed to increase student performance:

The school sites agreed to design new teacher-evaluation systems that incorporated classroom-observation rubrics and a measure of growth in student achievement. They also agreed to offer individualized professional development based on teachers’ evaluation results, and to revamp recruitment, hiring, and placement. Schools also implemented new career pathways for effective teachers and awarded teachers with bonuses for good performance.

They helped out all teachers; fired bad teachers; promoted good teachers; and paid bonuses to effective teachers. So how did it work out? They contracted with RAND to perform an evaluation, and here are the results for the three public school systems:

The difference between the blue and green bars is basically zero or negative everywhere. The one exception is a sudden and massive improvement in reading in the Shelby County School District during the last few years of the program. This is sort of inexplicable, though it might have something to do with its merger with the Memphis School District during this period.

In any case, let’s take a closer look at student performance in all three public school districts plus the CMO (charter) districts:

Long story short, there was no improvement at all in student achievement, despite the fact that funding was far greater than it would be in any real-life reform of this nature. There may have been some other successes in this program, but if the ultimate goal is better students, it was a complete failure. Whatever the answer is, rewarding good teachers and firing bad ones sure doesn’t seem to be it.

POSTSCRIPT: The program may have been a failure at improving student outcomes, but it’s worth pointing out that this doesn’t make it a failure as an experiment. The whole point of an experiment is that you don’t know the answer beforehand, and this one helps us understand what doesn’t work. That’s valuable, since it lets us move on to other things.

Also, kudos to the Gates Foundation for running an honest test. They designed it well; they funded it properly; and when it was over they hired a third party to provide an honest assessment of what happened. That’s the way it should be done.

Last year Donald Trump cut taxes on corporations and the rich by about $2 trillion over the next ten years. But what about us ordinary schmoes? For us he’s levied the following taxes:

• 25 percent tariff on steel imports
• 10 percent tariff on aluminum imports
• 25 percent tariff on $50 billion in imports from China

By my calculation, this comes to about $250 billion over ten years, something that Republicans have grumbled about slightly but done nothing to stop. And make no mistake: this is basically a sales tax that will ultimately be paid by consumers. So there you have it: a $2 trillion cut cut for the rich and a $250 billion tax hike for the middle class—so far. Welcome to the Republican Party’s vision for America.

This is the highly regarded VW Amarok pickup truck. Would you like to buy one? Too bad! You can't.Volkswagen

The German auto industry has suggested that the US and Europe become a free-trade zone in light vehicles:

That would mean scrapping the EU’s 10% tax on auto imports from the U.S. and other countries and the 2.5% duty on auto imports in the U.S….“Germany has the right approach to resolving this trade disagreement among friends,” said U.S. Commerce Secretary Wilbur Ross. “If the EU were to reduce its 10% tariff on U.S. cars and trucks, that would be a positive first step toward trade that was more fair and reciprocal.”

One catch is that the Europeans also want a 25% U.S. tax on imports of light trucks—pickup trucks, sport-utility vehicles, and big vans—scrapped. Abolishing this relic of the Johnson administration could alienate U.S. auto workers, a core constituency for Mr. Trump in the midterms this fall. Mr. Ross didn’t comment on whether the U.S. would be willing to cut tariffs on trucks.

Well, this would sure call Trump’s bluff. As near as I can tell, scrapping the EU’s 10 percent tariff on American cars would have zero effect on European carmakers. Why? Because no one in Europe wants American-made cars in the first place. On the other hand, scrapping the 25 percent US tariff on light trucks would be a boon for European carmakers. I’m sure Trump has no idea that the US has a 25 percent tariff on European pickup trucks, but hey, that’s the problem when you mouth off on stuff you’re ignorant about, isn’t it?

In any case, this is precisely the kind of “level playing field” that Trump keeps insisting he wants, so he could hardly object. Europe ought to push hard on this, just as a way of embarrassing Trump. He’d have a hard time saying no, but the US auto industry would go nuts. The last thing they need is more competition in the lucrative light truck market from the likes of Fiat and Mercedes and Volkswagen. This could be fun.

The internet’s tax-free era has come to an end:

Internet retailers can be required to collect sales taxes in states where they have no physical presence, the Supreme Court ruled on Thursday….The vote was 5 to 4. Justice Anthony M. Kennedy wrote the majority opinion and was joined by Justices Clarence Thomas, Ruth Bader Ginsburg, Samuel A. Alito Jr. and Neil M. Gorsuch.

That’s not a Supreme Court lineup you see very often. But this is what happens when the court decides a case that doesn’t really split along partisan lines. They actually look at the law dispassionately.

Anyway, that’s that. Within a year or less, every state will collect sales tax on internet and mail-order sales. And even though this affects everyone, I can hardly wait for Donald Trump to find out. I expect an FU tweet aimed at Amazon within a few hours.

The CDC’s latest Youth Risk Behavior Surveillance survey is out, and as you already know, the kids are alright these days. But there’s always something, isn’t there? The Washington Post uses the CDC report as a hook to tell us that risky teen behavior is down across the board except for one thing: they aren’t eating their veggies:

Do kids ever get a break from our nation’s news media? Apparently not. The Post kinda sorta plays this for laughs, but they also take it seriously enough to lie with statistics, informing us that since 1999 there’s been a 71 percent increase in the share of kids who don’t eat vegetables.

This is technically true, but check out the scales on the Post’s charts. Most of them go from zero to about 40 percent—except for the veggie chart, which goes from zero to 5 percent. I have helpfully redrawn their graphic to put everything on approximately the same scale:

The veggie line has barely moved.  But I guess that doesn’t make for a very good story. The kids may be alright, but I sure have my doubts about the adults these days.

Anyway, while we’re on the subject I might as well point out that in addition to all the stuff the Post mentions, teen arrest rates are down, pregnancy rates are down, drunk driving is down, math and reading scores are up, graduation rates are up, and college attendance is up. Given all this, I recommend that we switch to a Logan’s Run sort of society except that instead of killing people at age 30 we just take away their right to vote.¹ Let’s give the vote to everyone aged 21-30 and see how they do. It can’t be an awful lot worse than what we have now.

¹But not to hold office. Only 20-somethings can vote, but they can vote for any old fogey they want to.

Here are some findings from a couple of recent Pew surveys. First, a pair of questions related to voter enthusiasm in midterm elections:

The left-hand chart shows that, starting in 2006, midterm votes became much more a vote against the sitting president than anything else. I wonder if that’s entirely a good thing? In any case, anti-Trump sentiment is high this year, but it’s worth noting that it’s still not quite as intense as anti-Bush sentiment was in 2006. Of course, by then Bush had had six years for us to get sick of him.

The right-hand chart shows voter enthusiasm. Democrats usually have trouble turning out the vote for midterm elections, but this year enthusiasm is sky-high—far higher than it was even in the wave election of 2006. It’s even as high as Republican enthusiasm in the tea-party wave of 2010. If this keeps up, it’s bad news for Republicans.

So who’s responsible for this enthusiasm? Young women:

Every other demographic group has changed hardly at all, but millennial women have increased their identification with Democrats by 15 points in just a few years. What’s interesting is that this started in 2015, so it’s probably not especially related to either Donald Trump or Bernie Sanders. But it might be! I’d have to get hold of the individual surveys throughout 2015 to know for sure.

This is the picture they're talking about.John Moore/Getty Images

From an “outside White House adviser” on White House immigration zealot Stephen Miller:

Stephen actually enjoys seeing those pictures at the border. He’s a twisted guy, the way he was raised and picked on.

Meanwhile, Trump fixer Michael Cohen—Michael Cohen!—has resigned as deputy finance chair of the RNC Finance Committee in an act of conscience:

As the son of a Polish holocaust survivor, the images and sounds of this family separation policy is heart wrenching. While I strongly support measures that will secure our porous borders, children should never be used as bargaining chips.

Meanwhile, Donald Trump himself has finally reversed his policy of separating children from their families at the border. However, instead of simply ending the policy that started all this, he’s going to continue his “zero-tolerance” rule but detain parents and children together. I think Josh Marshall gets this right:

The President says he’s signing an executive order to end family separations. The actual aim seems to be to pick a fight with the courts and allow separations to continue while blaming judges. According to The New York Times, the President will sign an executive order allowing children to be detained indefinitely with their parents. The problem is that that violates a 1997 consent decree saying that you can’t detain/imprison children for more than 20 days (technically what’s currently happening isn’t detention). It straight up violates that order. So what will almost inevitably happen is that a court will step in, say you can’t do that and then Trump will announce that the judge is forcing him to keep separating families.

Trump has been trying to blame his policy on someone else ever since it started, but the non-Fox media has been pretty good about repeatedly calling this out as a lie. So now he’s going to try to make it into the truth—or something close to the truth, anyway. Just remember: if he really wants to end the separations, he can always go back to the policy he followed for the first 443 days of his administration. It’s entirely up to him.

Is this egret hunting for lunchtime morsels in a Lousiana bayou? Nope. This is Irvine’s very own swamp, located at the Audubon bird sanctuary on land owned by the water district. The swampy bits are toward the back, so be sure to stroll all the way out to the furthest reaches of the sanctuary if you come for a visit. Also: summer bat walks start in July, so sign up now if you prefer flying mammals to degenerate flying dinosaurs.

Bloomberg reports that in the first quarter of 2018 Deutsche Bank had a single day in which they suffered losses of nearly 12 times VaR:

VaR stands for Value at Risk, and it’s a measure of the maximum loss a portfolio is likely to sustain under normal circumstances. Exceeding VaR moderately once a quarter is not too unusual. Nearly everyone probably does it once a year. But:

The extreme day was one of four in the first quarter in which Deutsche Bank’s U.S. traders had a loss that surpassed the firm’s regulatory VaR estimate. No other bank required to file quarterly reports with the Fed detailing significant trading activities had more than two such days, and none had a daily loss that even doubled its estimate, according to a review of the 33 filings from U.S. lenders and the local units of foreign firms.

How unlikely is this? VaR is typically not a bell curve, but we can start there. As near as I can tell, VaR is normally two standard deviations, meaning you’d expect to exceed VaR—by a little bit—perhaps 2 percent of the time, or maybe a few days per year. Deutsche Bank’s bad day was 12x VaR, which is 24 standard deviations. That’s like having an IQ of 500. You’d expect to it to happen…

…well, if my calculation is correct, you’d expect it to happen once every 10¹²⁷ days. Just for reference, the universe is a piddling 10¹⁰ days old. But we’re not dealing with a bell curve, so let’s divide by a bazillion, which gets us back to, oh, let’s be generous and say 10⁵⁰ days. That’s still a billion trillion quadrillion gazillion times the age of the universe. So this leaves us with two possibilities:

• Holy cow, Deutsche Bank got really unlucky.
• Deutsche Bank’s VaR model is really, really bad.

I’m going with Door #2. That said, even a bad model shouldn’t be off by this much. It’s possible that Deutsche Bank’s VaR model might now hold the record for worst financial model in the history of the world. They should be proud.

I’ve had my fights with Ramesh Ponnuru, but he understands just how Orwellian the arguments in favor of Trump’s family separation policy are. Hell, this tweet could have been written by Atrios:¹

This column explains that calling family separation “cruel” hurts the feelings of people who support it. https://t.co/qKi56ViyQG

— Ramesh Ponnuru (@RameshPonnuru) June 19, 2018

In case you want to torture yourself by reading the column, this link should get you through the paywall.

¹Although the use of “feelings” instead of “fee fees” gives it away as counterfeit.