Friday, October 9, 2015

The literature advances. Political debates don’t.

Do teachers earn less than the minimum wage?
When “teaching hours” are quantified, one may argue that teachers’ salaries are adequate for the hours they work. But I can attest that if anyone ever quantified the hours that I, and many of my colleagues, submit to create and maintain cutting-edge instruction, we would be making less than minimum wage and we’d be better off flipping burgers.

Read more here:
That quote comes from a teacher making the federal minimum wage of $7.25 an hour would need to work about 21 hours per day, 365 days a year, to reach the average annual teacher salary of $56,000.

Road map for navigating political debates.

But what's bothersome isn't the exaggeration. It's this part: "...if anyone ever quantified the hours..." If? People have been quantifying teacher working hours for years. (See our working paper on the topic, and note the literature review in particular.) Instead of repeating the talking point that teachers work some extremely large but unknowable amount of time outside of the classroom, why not look into the research? We actually have data that measure teacher work time as definitively as anyone could hope for, and the paper linked above has all of the details. (TL;DR: Teachers work about 41 hours per week during the school year.)

I've noticed a depressing pattern in Washington. Even as the literature advances, politicians and reporters still have the same old debates featuring the same old talking points. Sometime soon, there will be another person who says that teachers have it easy because the school day is less than eight hours. Then someone like Ms. Futterman will respond that teachers work a lot at home. And then the debate will stagnate, without anyone bothering to investigate whether those work hours at home could be quantified. Lather, rinse, repeat.

It's true of so many issues. (Don't get me started about a certain topic in psychology.) Each side has their talking points, and the reporters seem uninterested in moving beyond them. Another example is the Cato Institute's new report on federal pay, which mentions that Feds make 78 percent more than private-sector workers before any adjustment for skill differences. Predictably, critics responded that the greater average education and experience of federal workers could explain the difference. And then... stalemate.

I've seen this movie before. Five years ago, the same events transpired, and no one in the media seemed interested in testing whether education and experience actually do account for the difference. Andrew Biggs and I decided to do that analysis, and it inspired several other reports that compare similarly-skilled workers in each sector, including one from the CBO and a summary of each report from the GAO. The new literature established that federal workers do receive, on average, greater compensation than their private-sector counterparts.

So the debate in Washington moved on to how to reform pay, right? No. The new Cato point about a 78 percent premium without skill controls, and the mechanical reaction to it about the need for such controls, indicate that we've come full circle. Federal pay is back to being he-said-she-said. I noticed a few years ago how the careful studies summarized by the GAO were dismissed by the media:
The takeaway from the [GAO] report has often been depicted by news sources as something like: “All these studies give different answers, so we just don’t know how federal pay stacks up.” Or: “All of the studies have limitations of some kind, so we shouldn’t trust any of them.”

Is that really the best interpretation? Let’s compare the results of the five studies that attempted apples-to-apples comparisons by matching workers or jobs between sectors. These are the aforementioned AEI, Heritage, and CBO studies, a report by the Project on Government Oversight (POGO), and the federal government’s official annual wage comparison conducted by the President’s “Pay Agent.”

In summary, four of the five studies find that total federal compensation is higher than private-sector compensation. The fifth study (the Pay Agent report) cannot come to any conclusion because it does not consider benefits. So all of the studies reviewed by GAO that actually reached a conclusion found that federal workers are overcompensated...
When not a single study has yet even claimed to overturn the basic conclusion drawn by Heritage, AEI, POGO, and the CBO, one would think the media would sound less discouraged about ever getting a definitive answer to the federal pay question. One would think.
So why are so many issues stuck in this perpetual cycle of talking point, counterpoint, rest, repeat, even as the literature advances beyond those disputes? I can think of a few reasons, but this post has already used up my quota of cynicism for the day.

Monday, September 21, 2015

The costs of immigration are real

A recurring theme of my writing is that, first, immigration has both benefits and costs, and, second, that the costs are systematically downplayed by immigration boosters. For example, back in the spring I wrote about the fiscal cost of immigration in "The Amnesty Numbers Game":
Consider a congressional hearing held March 17 by the House Oversight and Government Reform Committee. The purpose was to determine the fiscal impact of President Obama’s executive actions on immigration, specifically the granting of work authorization via executive order to millions of illegal immigrants. Witnesses opposed to amnesty presented serious reasons why the president’s actions could impose a cost on taxpayers. At every stage of the hearing, however, Democrats responded with dismissive grandstanding and appeals to authority.
Earlier this summer I discussed the cultural cost of immigration in a piece for Real Clear Policy called "Are Low-Skill Immigrants Upwardly Mobile?":
The desire to increase social mobility has taken center stage in recent years, as lower-skill workers and their families struggle to join the middle class. A recent New York Times poll found that only 35 percent of Americans agree that "everyone has a fair chance to get ahead in the long run." The explanations for insufficient mobility are many and varied.... But whatever the root causes of class stratification, the political class tends to ignore a major policy that worsens the problem — namely, the mass immigration of low-skill workers.
Two of my most recent pieces focus on the distributional costs of immigration. Increasing the supply of labor lowers production costs, but the savings come in the form of lower wages for the workers competing with immigrants. The direct connection between the economic benefits (lower consumer prices) and distributional effects (lower wages) is sometimes denied by immigration boosters. But, as I wrote in a National Review piece, other times they are honest -- perhaps inadvertently honest:
The [farm lobby's] report is clear about the ... desire to keep wages low by increasing the supply of labor. It describes wage increases as “a strain on many U.S. farms” that other industries have managed to avoid. It shows that real wages for food preparers, housekeepers, cashiers, and other low-skill workers outside farming have decreased since 2002. The reason, according to the report, is that “employers in non-agricultural industries have been able to find enough workers to fill job vacancies without upward pressure on wages.” Farm owners wish they had the same privilege.
Finally, I wrote about an important new paper from George Borjas (my advisor in graduate school), who has re-examined the wage impact of the Mariel boatlift:
There is perhaps no economic study more often cited by immigration advocates than economist David Card’s 1990 analysis of the “Mariel boatlift.” After Fidel Castro announced in 1980 that anyone wishing to leave Cuba could do so via the port of Mariel, 125,000 Cuban immigrants came to Miami in a matter of months that summer. The sudden influx of young, able-bodied workers generated an unusually good test of how immigration affects wages. Economic theory predicts wages should have declined in Miami after the boatlift, but Card was surprisingly unable to detect any wage impact at all.
Borjas found that there was a decline in wages, concentrated among the least skilled natives.

Wednesday, June 10, 2015

A Closer Look at science-gilding

Earlier this week I did a radio interview to promote my Public Discourse article, "Science-Gilding the Marriage Debate." The interview was with Sheila Liaugminas on her program, A Closer Look. Listen below:

Friday, May 29, 2015

The "science-gilding" of public policy

A prominent study of public opinion on same-sex marriage was recently retracted after revelations that the data were fabricated. Academic fraud of this sort is contemptible, but at least it is relatively rare and publicly condemned once discovered.

Much more common – but less often recognized – is the misuse of legitimate research for political purposes. That's the subject of my new essay for the journal Public Discourse:
As someone who deals with numbers for a living, I want to see public policy informed by data to the greatest extent possible. When empirical evidence helps policymakers better understand an issue, that’s terrific.

But misusing science for political ends is just as bad as ignoring it altogether. Sometimes the misuse comes from portraying weak evidence as conclusive. (“A tiny demonstration project from the 1960s proves that universal preschool would be a great investment.”) Other times, it involves lumping objective findings with a subjective policy prescription. (“To fight global warming, Science demands that we pass the president’s cap-and-trade plan.”)
I call such misuse the "science-gilding" of public policy, since advocates are trying to gild their political positions with the veneer of science. My essay uses the alleged "public health case" for same-sex marriage as an example, but I encounter science-gilding in many different areas of public policy, from preschool to climate change to immigration. Why isn't it exposed?
Science-gilding has occurred on both sides of the same-sex marriage debate, but it is generally the side favoring redefinition of marriage that gets away with it. One reason is that most of the mainstream media are sympathetic to same-sex marriage. The more important reason, though, is that reporters are easily swayed by statements from respected organizations like the [American Public Health Association]. They usually have neither the time nor the expertise to dig into the research and discover that the emperor has no clothes.
Hence my essay. If it helps the media develop some much-needed skepticism about "authoritative" scientific pronouncements, perhaps science-gilding will become less prevalent.

Tuesday, March 31, 2015

"The amnesty numbers game" published in National Review

In the spring of 2013, the Social Security Administration declared that the Schumer-Rubio amnesty bill would have a positive effect on its budget. Supporters of amnesty immediately seized on the report as further evidence that immigration is a win-win-win proposition.

But it was deeply frustrating for immigration skeptics. SSA would not say how it did its calculations. We were simply given a table of projections and asked to accept them. What assumptions did SSA make about the lifetime earnings of immigrants? We don't know. How about immigrant skill levels? We don't know. What would happen beyond the 10-year budget window? We don't know.

The amnesty debate is filled with authoritative-sounding numbers, but just a bit of unpacking often reveals that the emperor has no clothes. That's the point of my new essay for National Review that came out this week. A sample:
[T]rade-offs rarely make for good political messaging. So when the Obama administration began its push for amnesty for illegal immigrants — first through the legislative process and then more recently through “executive action” — it needed a slew of ready-made talking points to counter every objection. The result was a politicization of statistics on a grand scale and, inevitably, a degradation in the quality of debate. Numbers from seemingly authoritative sources have been stripped of context and presented as argument-stoppers, with little attention to the actual data or methodology involved. 
Consider a congressional hearing held March 17 by the House Oversight and Government Reform Committee...
Read the whole thing here.

Thursday, March 19, 2015

Men dominate the ranks of elite Jeopardy players

Ever heard a contestant give a ridiculously bad answer on Jeopardy? Here are three recent favorites:
Category: Of the Game
Yakutsk and Irkutsk are two territories to conquer and control in this game of attrition.
Contestant: "What is Clue?"
Category: "A"cademy Award Nominees (correct response must start with "A")
11 nominations, 1984; it won 8 Oscars, including Costume Design
Contestant: "What is Titanic?"
Category: It's at the Smithsonian
Logically, one of these weapons from Star Trek
Contestant 1: What is a stun gun?
Contestant 2: What is a lightsaber?
One reason these are humorous is that Jeopardy contestants aren't random people off the street. They are all trivia geniuses. I'm pretty good with trivia myself, but I've repeatedly failed the show's entrance exam. There's really no shame in that. Think about it: There are roughly 245 million American adults, but only 350 or so get to play on Jeopardy each year. They are the crème de la crème.

The three finalists from Jeopardy's Battle of the Decades tournament.

Which provides another opportunity to look into the gender imbalance in elite intellectual pursuits. Should we expect men to perform better than women on Jeopardy? On average, men have superior visual-spatial skill, but Jeopardy requires little of that. More relevant would be mean sex differences on general information tests (male advantage) and verbal fluency (often a female advantage).

But here I'm interested less in means than in variances. The ability distributions for men generally have "fatter tails," meaning men outnumber women among both the worst of the worst and the best of the best, even when the averages are the same. If this is the case with trivia, we should see men disproportionately represented among Jeopardy contestants, and we should see greater disproportions as we look at more exclusive groups of contestants. That's exactly what we do see.
Percentage of women among...

Jeopardy contestants:          40 %

Jeopardy winners:              30 %

tournament winners:            12 %

“best of the best" winners:     0 %
The first two percentages are courtesy of a Slate article from last year. I calculated the last two percentages based on the winners of the annual Tournament of Champions and the five "best of the best" tournaments, respectively. (I did not look at lesser tournaments for teens, teachers, college students, etc.) Only three women -- Rachael Schwartz, Robin Carroll, and Celeste DiNucci -- have won a Tournament of Champions. When it comes to the "best of the best" tournaments, which bring together the strongest players in the show's history, just one woman has even made the finals. That was Leslie Frates, who finished third in the Tenth Anniversary Tournament back in 1993.

Robin Carroll is an interesting story. After winning her Tournament of Champions in 2000, she was the American representative in a special international tournament (not included in my best-of-the-best calculations), which she won handily. However, she had a significant advantage as a native speaker in an English-language competition. She has since competed in three best-of-the-best tournaments and failed to advance beyond the quarterfinals in any of them.

In my estimation, Pam Mueller is the best female Jeopardy player of all time. In the "Battle of the Decades," Jeopardy's most recent best-of-the-best tournament, she was the only woman among the nine semi-finalists. She did not make the final round.

So what's with this right-tail domination by men? I'm speculating here, but obsessiveness seems to be more common in males. People who immerse themselves in a passion, who devote their lives to it, who become almost Rain Man-like in their knowledge of it, are more often men. And the odder the passion, the more male the obsessives seem to be. Even the world's leading experts on My Little Ponies are men! A weird obsession with trivia is probably what it takes to be successful in a game like Jeopardy.

Friday, March 13, 2015

Single payer, many costs

Reihan Salam wrote recently for NR about how containing costs through price controls -- a  supposed selling point of single-payer systems like Medicare -- can severely distort the way doctors and hospitals treat their patients. It seems obvious that below-market reimbursement rates will adversely affect treatment in one way or another, but the trade-off is not often acknowledged by Medicare’s boosters.

What’s always puzzled me is that the same people who use the cost-control argument in favor of single-payer healthcare would be horrified to see it applied to other types of government monopolies. Take public education. There is essentially a “single payer” for education within school districts. But public-school advocates would never argue that the system should be used to push teacher salaries down below market levels. In fact, raising teacher pay to be on par with the salaries of other college graduates is a perennial goal of progressives. What happened to monopoly cost control?

Imagine Republicans proposing to reduce school spending -- just as Obamacare reduced Medicare’s budget -- but insisting that it’s not a cut to services, it’s merely “savings” generated by squeezing education providers a little more. Democrats would cry foul, but that’s exactly how they sold cuts to Medicare.

The prevailing view on the Left seems to be that underpaying teachers is bad, but underpaying doctors is good. I understand that a wealthy doctor may be less sympathetic in the public mind than a school teacher, but they both respond to financial incentives. Underpaying either one is going to have consequences.

I first made this point in an op-ed for Forbes.