Thoreau complains about a grad student in his upper division biophysics class:

One of the problems involves the entropy change from evaporating a cubic centimeter of water. She asked me how she’s supposed to know the number of atoms in a cubic centimeter of water. Um, this is basic freshman chem stuff.

No, that's not right. It's junior chem stuff. High school junior, that is. At least, that's where I learned it.

The same student apparently also had a problem converting joules to electron volts. Well, I don't know the conversion myself. But I typed "joules electron volts" into Google and got the answer in .33 seconds. So I guess at least some grad students don't know how to use Google either. This does not bode well for our coming economic war with China, does it?

Charles Duhigg has a fascinating story in the New York Times Magazine this week that's all about the way retailers use data mining and microtargeting to sell you more stuff. Among other things, he tells the story of how Target exploited a pile of clever statistical relations to predict when women were pregnant so that they could send out coupon books full of items that pregnant women might want to buy. As it turns out, Target was unamused by Duhigg's curiosity about how this all worked. When he asked Target to comment, they refused. When he offered to fly out to company headquarters, they told him not to come. When he did anyway, a security guard escorted him off the premises. Quite plainly, Target was concerned that their customers would freak out if they discovered just how much Target knows about them and how accurately Target can aim its marketing bazookas in their direction.

And it turns out Target was right: pregnant women did freak out. So they fine-tuned their coupon books to contain a bunch of random stuff (lawnmowers, videogames) among all the pregnancy-related items. Women who got those coupon books just figured this was the stuff on sale at Target this week and had no idea that it was more than a coincidence that half the offers were for diapers and onesies.

Longtime readers will be unsurprised to learn that I'm not thrilled by this kind of thing. But Felix Salmon challenges people like me to explain why:

Nowadays, computers have made it increasingly possible to fine-tune personalization down to the individual level, where it can sometimes get “spooky”....If sophisticated corporations manage to make their marketing materials less spooky, I don’t think there’s going to be much popular opposition to continued targeting — at least not in this country. Germany is different: Germans care a lot about their privacy, and fight hard for it.

Here, however, I’ve never received a good answer to the “why should I care?” question.

I'll admit up front that my reaction to this is mostly emotional: I just don't like the idea of Target or anyone else knowing so much about me. But the truth is that Felix is right. I don't really fear Target. The worst thing they can do with all their data mining wizardry is send me personalized coupon books. If I don't like it, I can toss them in the trash.

(As it happens, I actually think the increasing power of modern marketing in the hands of giant corporations is something we should all take a little more seriously than we do. But that's a topic for another day. For now, let's assume that Target's use of my personal data is pretty benign.)

So then: what's a good reason to be uneasy about this? Well, if Target can use statistical relations to predict pregnancy, I'll bet they can predict other things too. For example, the early stages of Alzheimer's. And much of the data that Target and other retailers collect is also available to other marketers at a price.

Sadly, not all of them have the scruples that Target does. If you were, say, a semi-shady company hawking dubious life insurance schemes, who would be your best prospect? Answer: someone still in enough control of their faculties that they live on their own without lots of supervision from their children or other caregivers, but just infirm enough that their judgment isn't so good anymore. They'd be perfectly suited to be scared/bullied into buying a bunch of crap they don't need. Hooray for data mining!

Alternately, keep in mind that it's not just corporations that can get their hands on this stuff. The federal government can buy it too for whatever profiling schemes the whiz kids at NSA come up with. I'm not super excited by this prospect either.

Now, it's true that scammers have preyed on the elderly for a long time, and governments have collected information about their citizenry for a long time too. But a problem that's small and controllable when the targeting data is both diffuse and arduous to collect can become cancerous when the data is cheaply and broadly available. Roughly speaking, then, that's my answer to the question of why I care. It's not Target per se. It's not my local supermarket. It's the fact that everyone is doing this, the data is all getting collected in centralized databases, and this stuff is then accessible to just about anyone, not just folks who want to send me coupons.

I suspect there's not much we can do about this. And maybe the good outweighs the bad in any case. After all, the internet is still a good thing even if pedophiles occasionally use chatrooms to stalk their prey. Still, I think there are legitimate reasons for concern. Life is more than just coupons.

Have you been following the great North Carolina school lunch totalitarianism story? Me neither. But as near as I can tell, a little girl who was enrolled in Raleigh's "More at Four" Pre-K program came to school a few days ago without any milk, so someone — teacher's aide? cafeteria worker? who knows? — told her to go through the lunch line and get some. (It was free.) But she got confused and thought she had to get a whole new lunch. When she told her mother about this, the mother decided that school bureaucrats were forcing her child to eat their lunch instead of the one she packed for her daughter. A local TV station picked up on this, and before long the conservo-sphere was reporting that jack-booted thugs in North Carolina were forcing every child in their schools to eat only their Obama-approved chicken nuggets instead of nutritious family-packed meals.

Seriously. That seems to be approximately what happened — though I'll grant that the school system's communication apparatus seems to bear a wee bit of fault here too. Mark Thompson has the entire bloody blow-by-blow here.

Mike Konczal says the illustration on the right is his favorite graph of the week. It's based on data from the Welfare Transition program in Florida and it shows two things. The line shows tourism revenue: high in winter, lower in spring, and lower still in summer. Obviously the tourism industry in Florida needs more workers in winter than in summer.

The bars show the sanction rate in the WT program: higher in winter and lower in summer. Apparently, when the tourism industry needs more workers, welfare case workers are more likely to force their clients out of the system and into the job market (i.e., "sanction" them).

The authors say that the correlation between these two things is .95, which is, frankly, too good. If Florida's welfare bureaucracy was trying to match up workers with the needs of the tourism industry, they couldn't produce a correlation that perfect. So there has to be something more going on here.

Still, this is nonetheless pretty persuasive evidence that case workers do, in fact, calibrate sanction levels to the needs of the job market. So my next question is this: is this a bad thing? Mike doesn't really take a position, though he seems vaguely disapproving. And it's possible that the details of the sanctioning regime are objectionable. But just in general, is there anything wrong with welfare case workers trying to push clients into the job market when jobs are available, but being more lenient when jobs just aren't there? Offhand, I'm not sure I see a problem with this.

Today I defend Marco Rubio and Politifact. But Rubio more than Politifact.

Last weekend Rubio gave a speech at CPAC and said, "The majority of Americans are conservatives." Politifact checked up on this:

The Gallup Poll has been regularly asking Americans about their political ideology since 1992, and they compile the results of many polls each year and release an annual report. For 2011, Gallup found that the largest group of Americans identify as conservative, at 40 percent. Another 35 percent identify as moderate, while 21 percent identify as liberal.

....Rubio said that the majority of Americans are conservative. A respected ongoing poll from Gallup shows that conservatives are the largest ideological group, but they don’t cross the 50 percent threshold. So we rate his statement Mostly True.

Liberals pounced, and with good reason. This is mind-bogglingly dumb. 40% of Americans identify as conservative, so it's "mostly true" that a majority of Americans are conservative? Seriously?

But guess what? Gallup is not the only pollster in the world! Here's a Politico Battleground Poll from a few months ago that forced people to choose whether they leaned left or right:

In this poll, 61% of the country identified as conservative. That's a majority!

Now, this is a poll of "likely" voters. It forces a choice between liberal and conservative. And even though we all know that "independent" voters mostly lean left or right pretty reliably, I imagine this poll still understates the number of true centrists.

Nonetheless, it's a poll. And Rubio could quite reasonably point to it as evidence that a majority of the country is conservative. Liberals, just as reasonably, could point to other polls suggesting that a majority of Americans support liberal goals regardless of what they call themselves. It's kind of complicated!

Which is why Politifact probably shouldn't have pretended to fact check this in the first place. And if they insisted, they should have pointed to multiple polls instead of pretending that a single Gallup poll was gospel. It's a dog's breakfast.

But Rubio is off the hook. All he needs is one good poll to justify his statement, and he's got one.

Here is today's installment of "Government Bureaucrats Working To Make Your Life Better." Seriously. Behold the latest rule from the FCC:

After receiving thousands of complaints from consumers, the Federal Communications Commission clamped down Wednesday on unwanted robo-calling by approving sweeping changes to its telemarketing rules for wireline and mobile phones.

....Under the new FCC rules, telemarketers are required to obtain written consent, which can be in the form of an online approval, before placing autodialed or prerecorded calls to a consumer. Telemarketers also must provide an automated opt-out mechanism during each robo-call so that consumers can immediately tell the telemarketer to stop calling.

The FCC also eliminated the "established business relationship" exception, which had allowed robo-calls to be placed to the land-line home phones of consumers with "prior or existing" associations with companies represented by telemarketers.

I'm especially thrilled about the last bit. If my dentist (or whoever) wants to call to tell me something genuine (don't forget your checkup tomorrow!), that's fine. But if they just want to harass me endlessly with robocalls to sell me more stuff (we're having a sale on tooth whitening!) then forget it. That's an ordinary commercial telemarketing call and I want it stopped regardless of whether I happen to have set foot in their establishment before.

I have my doubts about how many robocallers actually care about the FCC's rules, but I'm sure at least some of them do, and thus my life will be made ever so slightly better. So hooray for the FCC. These are your bureaucrats watching out for you.

I finally got around to reading last week's New York Times story about the widening education gap between high and low-income students. The chart on the right tells the story: according to research by Sean Reardon, the gap in reading performance between black and white children has declined by about one grade level since 19701 (one "unit" on the chart is three grade levels). At the same time, the gap between rich and poor has gone up by nearly two grade levels:

“With income declines more severe in the lower brackets, there’s a good chance the recession may have widened the gap,” Professor Reardon said....One reason for the growing gap in achievement, researchers say, could be that wealthy parents invest more time and money than ever before in their children (in weekend sports, ballet, music lessons, math tutors, and in overall involvement in their children’s schools), while lower-income families, which are now more likely than ever to be headed by a single parent, are increasingly stretched for time and resources. This has been particularly true as more parents try to position their children for college, which has become ever more essential for success in today’s economy.

This all seems pretty believable, but it's a little odd nonetheless. Here's another chart, this time from the long-term NAEP test, showing reading performance among 9-year-olds at the top and bottom of the performance spectrum:

These two charts aren't necessarily contradictory. One shows income levels and the other shows performance levels. Still, common sense suggests that the top percentiles of performance are mostly made up of high-income kids while the lowest percentiles are mostly made up of low-income kids. And on the performance chart, the gap hasn't been growing at all. In 1971, the gap between the top 10% and the bottom 10% was 108 points (on a scale where ten points is roughly one grade level). In 2008 the gap was down to 94 points. The gap shrunk by about one and a half grade levels during the exact same period that Reardon says the income gap increased by about one and a half grade levels. If you compare the top 25% to the bottom 25% you see about the same thing.

So....I dunno. As I said, these are two different measurements, and the NAEP test doesn't break down scores by income. But it still seems surprising that there's such a difference. At the very least, I wish the Times had explored this and tried to explain it.

1Note that the Reardon chart is by birth year. So look at the numbers in 1960 to see the gap between 9-year-olds in 1970.

I really and truly didn't mean to make this into copyright week here on the blog, and I promise to shut up about this soon. But there's a piece of this conversation that seems to be perpetually ignored, and Jerry Brito gives me an excuse to mention it today. He's responding to the back-and-forth between Tim Lee and me on recent legislative attempts to enforce copyright in the digital era:

Tim points out that copyright owners have, as a matter of fact, received greater and greater enforcement powers—almost on an annual basis. As a result, Tim says, “most of us are not anti-copyright; we just think enough is enough, and that the menu of enforcement tools Congress has already given to copyright holders is more than sufficient.”

Sufficient for what, though? Sufficient to significantly reduce piracy online? That’s certainly not the case. Piracy is rampant on the net. Some would say, though, that the only meaningful ways left to enforce copyright would (dare I say it?) break the Internet as we know it.

So I think that when Tim says that the powers copyright holders now have are “more than sufficient,” I think he means sufficient to provide an incentive to create. After all, the purpose of copyright is to “promote the progress of science,” not to protect some Lockean notion of property. It may be the case that while owners’ rights are no doubt being violated, a further reduction in piracy won’t affect the incentive to create.

I am, as always, speaking only for myself, but I think this is too cramped. The Constitution says that the purpose of patents and copyright is to "promote the Progress of Science and useful Arts," but the fact that the Constitution says this doesn't mean it's the only reason to grant patents and copyrights. There's another reason too: because creators have a moral right to profit from their works. In real life, pretty much everyone acts as if they believe this, and I suspect that for most of us it's the real underpinning of our support for IP law.

In any case, it's certainly part of the reason I support reasonable enforcement of IP law. The Sonny Bono Act was a travesty, software patents deserve to die a quick death, and a lot of recent attempts to police the internet have been cretinously wrongheaded, but the reason this hasn't soured me on continuing to look for solutions to widespread piracy is because I'm not solely concerned with promoting innovation and creation. I'm also concerned with protecting authorial rights to dispose of (and profit from) their creations as they see fit, and this is as true for digital creations as it is for physical creations. I'd hate to live in a world in which authors found it nearly impossible to make money from their works.

Now, I might eventually find myself in such a world anyway. We'll have to wait and see. But I'm certainly not willing to shrug my shoulders and give up on it quite yet.

I don't really have anything insightful to say about this, just that it's an astonishing chart. Mitt Romney's favorability rating has been in complete free fall since early January. If the Republican primary lasts another few months, he's going to be about as popular as your average dinnertime telemarketer.

Here's an interesting tidbit of research. Many states place substantial restrictions on the work that nurse practitioners are allowed to perform. Allowing them to do more might help relieve the shortage of family doctors and primary care pediatricians, but most doctor groups oppose this. Part of the reason is the fear that it would reduce physician pay, so Patricia Pittman and Benjamin Williams set out to see if this was true. Their answer: not really. In states that ditched their restrictive SOP (scope of practice) laws — the blue dots in the charts on the right — pediatricians actually made more money, while GPs made slightly less. In both cases, the difference was small and not statistically significant. Aaron Carroll comments:

Bottom line — there was no difference. Allowing more mid-level practitioners to practice freely and independently was not associated with physicians earning less.

In the interest of full disclosure, I’m married to a nurse practitioner. So I may be biased in my assessment that she’s amazingly talented. But for those physicians who are worried that increasing the ability of mid-level practitioners to work independently might negatively impact their income, that doesn’t necessarily seem to be the case.

Incumbent professionals often promote restrictive occupational licensing schemes because they want to limit competition. In many cases, this is probably rational in a self-serving way. But in others, it probably doesn't even make sense from a purely selfish perspective. This seems to be one of them.