Do bodycams reduce the use of force by police officers? In the previous post, I wrote about a recent study in which the headline said bodycams had no effect but the study itself said they had a huge effect. What's going on?

Answer: there were two studies by the same group. The first one looked at the overall data from ten police departments and concluded that bodycams had little effect. The second study broke the police departments into those that followed the experimental protocol and those that didn't. This is from the second study:

We present results from preplanned subgroup analyses on the efficacy of the treatment for particular groups of interest, in a pre-specified manner. We did not perform these analyses when presenting the preliminary main effects in Ariel et al. (2016), as the data we were interested in—police officers’ discretion—were not available at the time.

Here are the basic results. The treatment groups are supposed to keep their bodycams on at all times when they're interacting with the public. The control groups are supposed to keep their bodycams off at all times:

So there you have it. The authors conclude:

Given our preliminary findings, we think that there is a clear route for...improving the implementation of BWCs around the world: cameras should remain on throughout the entire shift—that is, during each and every interaction with citizens—and should be prefaced by a verbal reminder that the camera is present. We argue that the verbal reminder delivered by the officer wearing the camera provides a mechanism to remind that ‘rules of conduct’ are in play—common courtesy from officer and citizen for one, and potentially a legal requirement given the weight of privacy sensitivities in the public domain.

I continue to have some doubts about the sheer size of the effect here, which is pretty unprecedented in real-world interventions like this. It's also worth noting that the big difference in use of force comes from only three police departments: those that followed the experimental protocol faithfully. This small sample size opens up the possibility that these police departments were different in some way other than the fact that they followed the experimental protocol. Perhaps they had better leadership to begin with, and that's why officers did what they were supposed to do?

Nonetheless, interesting stuff. It certainly provides an obvious avenue for follow-up studies, and it deserves more play than the first study. Perhaps someone should ask Donald Trump what he thinks of bodycams. That seems to be the only sure-fire way of getting the media's attention these days.

Well, this is a chin-scratching result from a recent study:

Wearing body cameras increases assaults against officers and does not reduce police use of force

....Averaged over 10 trials, [body cameras] had no effect on police use of force, but led to an increased rate of assaults against officers wearing cameras.

More accurately, this would be a chin-scratching result if it were true. But even though this is directly from the journal article, it's a wildly misleading headline. Here are the details:

The researchers set out a protocol for officers allocated cameras during the trials: record all stages of every police-public interaction, and issue a warning of filming at the outset....Researchers found that during shifts with cameras in which officers stuck closer to the protocol, police use-of-force fell by 37% over camera-free shifts. During shifts in which officers tended to use their discretion, police use-of-force actually rose 71% over camera-free shifts.

Yikes! For officers who kept the cameras rolling all the time, use-of-force fell 37 percent. That's not just huge, it's positively gargantuan. Use-of-force only increased among officers who turned their cameras on and off. In other words, the real headline result is: keep the cameras rolling all the time and use-of-force plummets. So why on earth does the headline suggest that body cameras have no effect? In fact, their effect ranges from -37 percent (rolling all the time) to zero (control groups) to +71 percent (cameras on and off). That's gigantic beyond belief.

As for the 15 percent increase in assault against officers, that's very weird. It's a big increase, and apparently no one has a good idea of what might have caused it. I wonder if they also got different results here depending on whether officers followed the protocol and kept their cameras on at all times?

Bottom line: these results are too big. I don't have access to the paper, but something has to be going on. It's hardly believable that body cameras truly have an effect this gigantic. Something went wrong with this study.

UPDATE: It turns out there were two separate papers involved. Details here.

As Britain Goes, So Goes the EU?

A week from today, Brits will flock to the polls to vote on whether they want to remain part of the EU. Justin Hughes, the chief US negotiator for two recent international treaties, writes in the LA Times that British voters should ignore what Americans are saying about all this. That's fair enough. We did just nominate Donald Trump to lead one of our two major parties, after all. We're hardly in any position to be giving anyone else political advice at the moment.

Toward the end, though, Hughes offers up this:

Obama’s activism may reflect a larger concern: that a British exit from the EU would commence a full unraveling of the European Union. If that is what he’s thinking, it is a bit insulting to Europeans because it fails to appreciate the full scope and depth of continental Europe’s post-1945 integration. The immediate refugee crisis and the long-term Euro debt crises are greater hazards to a united Europe than whether Britain leaves the club.

I'd put this differently: The refugee crisis and the Euro crisis are bad enough already for EU solidarity. Add to that the growth of right-wing nationalist parties throughout Europe, and are we really sure that Brexit might not be enough to spur the "full unraveling" of the EU that Hughes scoffs at?

I'd put the odds of that pretty low. Maybe around 5-10 percent. But not zero. Why am I so relatively pessimistic? Because:

  • The EU and its predecessors have been around for only about 60 years. That's a blink of the eye in historical terms. The EU is much newer and more fragile than people give it credit for.
  • Tight, EU-style integration of neighboring nation states is historically unprecedented.
  • World War II was the original catalyst for the EU, but it's a chapter in the history books for the current generation of voters.
  • The threat of the Soviet Union is gone.
  • The euro really does pose a problem that's going to have to be addressed sometime soon: either tighter or looser integration. Tighter integration would include a fiscal union and a much larger flow of money from rich to poor regions—just like the United States. If that's off the table, then eventually looser integration will be the only option. Greece hasn't forced this decision yet because Greece is small and everyone hates them. Someday, though, it won't be Greece that's in trouble.

Put all this together, and I'm not convinced that Brexit is as benign as Hughes suggests. But then again, a year ago I didn't think Donald Trump had a chance of winning the Republican primary, either. So what do I know?

Donald Trump wants to make TV great again:

The breakout media star of 2016 is, inarguably, Donald Trump, who has masterfully—and horrifyingly—demonstrated an aptitude for manipulating the news cycle, gaining billions of dollars worth of free airtime, and dominating coverage on every screen. Now, several people around him are looking for a way to leverage his supporters into a new media platform and cable channel.

....According to several people briefed on the discussions, the presumptive Republican nominee is examining the opportunity presented by the “audience” currently supporting him. He has also discussed the possibility of launching a “mini-media conglomerate” outside of his existing TV-production business, Trump Productions LLC....Trump, this person close to the matter suggests, has become irked by his ability to create revenue for other media organizations without being able to take a cut himself. Such a situation “brings him to the conclusion that he has the business acumen and the ratings for his own network.”

I don't know if this is true. But it should be true, don't you think? It deserves to be true. Other candidates have used a presidential run as a platform for selling more books (Herman Cain, Newt Gingrich), but that's small potatoes. Trump is bigger than those guys. He wants to use his presidential run as a platform for a new TV show.

No, wait. A new TV network. That's the ticket. That'll finally show everyone that he's a better businessman than Dad ever was. And that's what his entire career has always been about, hasn't it?

I remember we used to play this game with Sarah Palin, and now we can play it with Donald Trump. The game is "What the Hell Did That Mean?" Today's edition comes from a speech in Atlanta. Donald is in the middle of his usual whine about how other countries don't pay us enough tribute for protecting them, when he turns to Saudi Arabia:

We defend them. We defend them. Every time somebody maybe makes a threat, there go the ships, there go the planes, there goes everything. And every time you turn on one of those aircraft carriers it costs you probably a million bucks. I'd say, don't turn it on. The captain would say, we want to show you how great these engines are working. No, I don't want to hear it, just don't.

Wut? We'll never find out what he meant, though. In the same way that we're all still wondering where Ronald Reagan was headed on his trip down Highway 1 when time ran out in the 1984 debate, Donald got distracted by a protester he wanted to toss out. So that's all there is.

By the way, you should listen to it. I tried to figure out how to transcribe his bizarre tone and body language during this harange, but it was beyond me.

Over the weekend, Scott Winship commented on the latest CBO report about income inequality:

The best way to view the state of the middle class is to look at the median household incomes CBO reports. The pre-tax and -transfer median was only 12 to 20 percent higher in 2013 than in 1979....After accounting for [transfers and] taxes, which have declined significantly since 1979, median income rose 40 to 47 percent....This post-tax and -transfer measure is what really counts — it’s the closest measure we have to disposable income. From 1979 to 2013, the median grew by just shy of $20,000. Cite that number the next time the Economic Policy Institute tells you the middle class is doing terrible.

This kind of thing is really annoying. Post-tax-and-transfer is not what "really counts." It all depends on what you want to know. If you want to know how the market economy is doing, you have to look at market income before taxes and transfers. Conversely, if you want some insight into the lived experience of families, then you want to look at post-tax-and-transfer income. They're both useful for different things. Neither one is the "real" measure of income growth.

That said, many people, including me, are very interested in market income. That's because we're not big fans of letting the market go wild and then trying to make up for its worst excesses via social welfare programs. We'd rather try to regulate the markets to provide more equal outcomes in the first place, and you can only do that if you know what market incomes are to begin with. So here's a look at market income growth since 1979:

That's an increase of about half a percent per year for the non-rich. We ought to do better. In real life, however, there's a limit to how much balance you can squeeze out of the market, which means that social welfare programs are important too. So I'm also interested in post-tax-and-transfer income, which gives me an idea of how well we're doing to help the people who need it most. In other words, both measures of income are important, and neither one is "what really counts."

But there's more! Winship wants to focus on post-tax-and-transfer earnings because his life goal is to convince everyone that income growth for the middle class has actually been pretty good lately. Toward that end, he makes the point that baby boomers are retiring, and many of them have modest market incomes because they rely mostly on Social Security and Medicare. "As retirees become a bigger share of the middle class over time, their lack of pre-transfer income pulls down the median." So please ignore market income, which has increased a dismal 12-20 percent since 1979! It only looks stagnant because of all those retiring boomers. Look instead at average post-tax and transfer income, which has increased 40-47 percent.

Well, maybe seniors do pull down the average a bit. But we shouldn't stop there, should we? If seniors rely heavily on transfer income, it stands to reason that working-age families don't. In fact, it turns out that only about a quarter of all transfer income goes to non-seniors,1 and nearly all of that goes to low-income families. Middle-class families of working age get very little.2

This means that non-elderly middle-class families have had to rely mostly on market income since 1979—along with a few crumbs of tax cuts.3 Which in turn means the increase in middle class earnings for working-age families since 1979 is a whole lot closer to 12-20 percent than it is to 40-47 percent. Cite that the next time you decide to sneer at the Economic Policy Institute.

1Page 5 of the CBO report: "On average, households received $13,900 in government transfers in [2013]—$9,900 from Social Security and Medicare and $4,000 from other government transfers." In other words, of the $13,900 that households received on average, seniors got $9,900 plus, at a guess, an additional $1,000 in other benefits. That leaves $3,000 out of $13,900 for the average non-elderly household, or 22 percent of all transfers.

2Table 3 of the CBO report. Excluding benefits for the elderly, the middle quintile gets only 7 percent of its income from government transfers, mostly from Medicaid.

3Figure 2 of the CBO report.

David French doesn't want effete liberals telling him what kind of gun he can and can't use to defend himself:

Any person who breaks into my house or who threatens my family on my property will very soon find themselves staring at the business end of an AR-15 — specifically this one, made by the DoubleStar Corporation, a Winchester, Ky., arms manufacturer.

Here's the product product pitch: "The Mil Spec Dragon is what black rifle purists dream about! The MSD is built as close to military specifications as possible while maintaining its semi-auto, civilian legal status. No frills, no flash, just a high quality, reliable, carbine that is ready for any mission." It comes standard with an A-2 flash suppressor and a forged “F” front sight tower with taper pins. Popular upgrades include a detachable carry handle, Hogue pistol grip, Ace SOCOM stock, “H” buffer, winter trigger guard, and DS 416 4-rail slimline handguards.

Sounds nice. Let's continue:

But now I’m told — largely by people who don’t know the first thing about firearms — that no American “needs” an AR-style rifle. But when your life is on the line, what do you want? More accuracy or less? More firepower or less? More recoil or less? More reliability or less?

Quite so. And French is right when he says later that banning assault rifles like this one wouldn't have much effect on international terrorists. I assume he'd agree with what I said on Sunday: "Banning Muslim visitors or ramping up the air war against ISIS would have done nothing to stop Omar Mateen. Banning guns would have done nothing to stop either the Paris attacks or any other act of international terrorism."

So here's my question. It's a real question. Like most AR-15 variants, the Mil Spec Dragon comes with a 30-round magazine. Is that really necessary for self defense? I can't pretend that there are literally no circumstances where it might come in handy, but it sure seems pretty unlikely in your basic home robbery scenario. Surely a 10-round magazine would be sufficient for virtually any real-world self-defense application? At a guess, three or four rounds would be more than enough. Or one.

So what's the argument against a ban on large-capacity magazines? Other than, "they're bitchin," that is. A ban on large magazines would surely slow down folks like Omar Mateen without infringing in any real way on anyone's Second Amendment right to defend themselves. Is there an answer that doesn't rely on absurdly unlikely scenarios?

It's sophisticated these days to remind everyone that national polls don't matter much in presidential races. You really need to dig down to the state level. That's true enough, and you can check out Sam Wang's detailed look at state polls here if you want to really dive into things. Spoiler alert: Hillary Clinton has a pretty clear lead right now, but control of the Senate is neck-and-neck.

All that said, national polls aren't totally useless. Voting patterns tend to be nationwide, and generally speaking, it's very unlikely for a candidate who's more than a couple of points behind in the national polls to win in November. So here's your good news for the day: Pollster says that Donald Trump is currently losing in the national polls by about 10 points. It's only June, and a lot of things can change between now and Election Day. Nonetheless, perhaps America is finally coming to its senses.

Weekly Flint Water Report: June 2-9

Here is this week's Flint water report. 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 103 samples. The average for the past week was 6.20.

One of Flint's peculiar problems has been the large number of residences with astronomical lead readings. This got me curious: the average lead levels in Flint's water seem to be under control, but how about the number of outliers? Is that going down? The sample size here is necessarily small, but the answer seems to be: apparently not. In fact, the number of very high readings has been going steadily up:

The Kaiser Family Foundation has a new report out suggesting that Obamacare premiums are going to skyrocket next year. Maybe so. But before everyone gets into another Trump-inspired tizzy about this, please note the highlighted section of the Kaiser table below:

Insurers always ask for more than they get. In the 13 states that Kaiser examines, insurers have asked for rate increases averaging 11 percent. And who knows? Maybe they'll get it. More than likely, though, they won't. This is an opening bid, and the final contract won't be set for several more months.

So, as usual, take this with a grain of salt. The truth is that Obamacare premiums started out lower than most analysts predicted, because (a) insurers turned out to be really interested in lowballing their prices in order to gain market share, and (b) they didn't have much data to base their rates on. As the market shakes out, real-world data will become more available and market shares will start to settle down. It's possible that this will drive a couple of years of semi-large-ish premium increases, but that's about it. And maybe not even that much.

So don't panic. We don't know yet if premiums are really going up 11 percent. But even if they are, it shouldn't be a huge surprise. The real test will be how the original Obamacare projections compare to real-life premiums in, say, 2019. My guess is that they'll be fairly close. More here from Charles Gaba, who basically says we just don't know yet what will happen in 2017.