Thomas Piketty Has Eric Hanushek's Number

I'm just starting Thomas Piketty's Capital in the 21st Century, which puts me months behind the mass of American public intellectuals who were obsessing about the tome this spring, but will give me the advantage of being able to say I actually read the thing.

Last night, I came across a quote that was so perfectly relevant to the current state of education policy research that I was prepared to type it in right here. But it turns out I'm not the only one who likes this particular passage; various forms of this Piketty quote are all over the web.

Here's one:

To put it bluntly, the discipline of economics has yet to get over its childish passion for mathematics and for purely theoretical and often highly ideological speculation, at the expense of historical research and collaboration with the other social sciences. Economists are all too often preoccupied with petty mathematical problems of interest only to themselves. This obsession with mathematics is an easy way of acquiring the appearance of scientificity without having to answer the far more complex questions posed by the world we live in. There is one great advantage to being an academic economist in France: here, economists are not highly respected in the academic and intellectual world or by political and financial elites. Hence they must set aside their contempt for other disciplines and their absurd claim to greater scientific legitimacy, despite the fact that they know almost nothing about anything. This, in any case, is the charm of the discipline and of the social sciences in general: one starts from square one, so that there is some hope of making major progress. In France, I believe, economists are slightly more interested in persuading historians and sociologists, as well as people outside the academic world, that what they are doing is interesting (although they are not always successful). My dream when I was teaching in Boston was to teach at the École des Hautes Études en Sciences Sociales, whose faculty has included such leading lights as Lucien Febvre, Fernand Braudel, Claude Lévi-Strauss, Pierre Bourdieu, Françoise Héritier, and Maurice Godelier, to name a few. Dare I admit this, at the risk of seeming chauvinistic in my view of the social sciences? I probably admire these scholars more than Robert Solow or even Simon Kuznets, even though I regret the fact that the social sciences have largely lost interest in the distribution of wealth and questions of social class since the 1970s. Before that, statistics about income, wages, prices, and wealth played an important part in historical and sociological research. In any case, I hope that both professional social scientists and amateurs of all fields will find something of interest in this book, starting with those who claim to “know nothing about economics” but who nevertheless have very strong opinions about inequality of income and wealth, as is only natural. [emphasis mine]

Guilty as charged. I took a couple of undergrad courses and read Paul Krugman -- that's the extent of my familiarity with economics. I suspect, furthermore, that I am a fairly typical student of education policy if you count me among those coming from the "education" side of things. Many of us have a background in teaching and came over to policy as a result of being interested in studying the dynamics of the systems around us.

But the folks who have been driving the direction of policy increasingly appear to have little practical understanding of how schools actually work; instead, they have training in economics, and are particularly attracted to a "quanty" view of the world.

Now, don't get me wrong: I love a good linear regression as much as the next guy. The mathematical modeling of phenomena in education can be very powerful and useful for informing policy. One of the reasons I have been reluctant to call for a ban on standardized testing is that the tests yield data, and that data can help us to develop theories as to why education is the way it is. We wouldn't, for example, be able to make the case that a student's socio-economic background is the most powerful predictor of his academic outcomes if we didn't measure those outcomes to begin with (although we could have all the data we need to develop these theories without being nearly as intrusive as we are now).

But it's clear that the education field has moved into a mode of thinking described quite accurately by Piketty: we are letting the theoretical mathematical models of economists dictate policy even though we have no idea how those models will function in the real world.

Take the Vergara case, which overturned teacher tenure laws in California. Eric Hanushek, the intellectual godfather of reforminess, testified for the plaintiffs that his research shows that removing the bottom 5 to 8 percent of "ineffective" teachers would generate a bazillion dollars and wipe away the national debt and leave everyone's breath minty fresh. It's a nice idea...

But there are so many practical considerations missing from Hanushek's framework that it's astonishing to think we'd even consider large scale policy changes based on his musings. For example: how do you identify the lowest 8 percent? Where's the cutoff between 8 percent and 9? Do you have any idea whether the teachers replacing these slackers are any better? What happens to the pool of teacher candidates when you change their compensation (tenure is part of a teacher's compensation)? How will this affect the distribution of teacher quality across and within districts? What's the effect of firing misidentified good teachers on the willingness of workers to enter the teaching corps? Will teacher shuffling affect the quality of community-building that goes on in schools? Are teachers equally "effective" in all environments?

Hanushek's models can't account for all of this ; how could they? And yet Judge Rolf Treu stepped in and overturned state law on tenure based on what is really nothing more than speculation. This strikes me as more than a little rash.

The same can be said for the Chetty, Rockoff, Friedman study on the effects of "good" teachers on students' future earnings. Set aside how these economists ran their numbers through the Mountain-Out-Of-A-Mollhill-Inator to pump up the size of their outcomes. Set aside the problems of using an error-ridden system like Value-Added Modeling to identify the "good" teachers. Set aside the problem of whether the gains in earnings are reflective of increased worker productivity, or whether they reflect a shuffling in a distribution of incomes that is actually pre-set (to my mind, a far more likely possibility).

As a practical matter, how are you going to assign students to "great" teachers as opposed to "good" ones? Because, if the effects are as great as you claim, isn't it wrong not to deliberately distribute teachers based on their effectiveness?

Are you going to force "great" teachers to teach the neediest students? How do you know they'll still be "great"within that different teaching context? Is it not possible a "great" teacher of affluent students would be a lousy teacher of students who are at-risk?

Are we not going to allow teachers any autonomy in choosing their own assignments? Because teacher qualification is unevenly distributed between districts, how are you going to get teachers in high-performing districts to teach in low-performing ones? And again, what does that do to the pool of workers willing to teach?

When the President himself cites Chetty, Rockoff, Friedman in his State of the Union, and his Secretary of Education promotes policies that demand the use of test scores in teacher evaluation, we've left the realm of theoretical speculation: actual policies are being made and sold to the public on the basis of economic modeling. But it's only modeling -- and these models are, to be generous, incomplete. They do not take into account the many practical considerations school leaders must deal with every day.

I've heard Piketty's critique of his field more than once: economists, desperate to be taken seriously as "real" scientists, gussy up their work in a quanty patina. Again, I certainly have no problem with mathematical modeling, but it has to be viewed in context. We shouldn't rush to policy conclusions on the basis of limited and theoretical frameworks.

I'd ask these very smart and very accomplished men to perhaps put down their slide rules for a while, leave their offices, and follow the sage advice of Jonathan Kozol (as quoted here by Bob Somerby):

KOZOL (page 163): You have to go back to the schools themselves to find an answer to these questions. You have to sit down in the little chairs in first and second grade, or on the reading rugs with kindergarten kids, and listen to the things they actually say to one another and the dialogue between them and their teachers. You have to go down to the basement with the children when it’s time for lunch and to the playground with them, if they have a playground, when it’s time for recess, if they still have recess...You have to do what children do and breathe the air the children breathe. I don’t think there’s any other way to find out what the lives that children lead in school are really like.

Or, for that matter, their teachers.

Coming to a peer-reviewed journal near you!

ADDING: OK, this is a little gossipy, but it is pretty funny:

As a final day of testimony, it was the plaintiffs’ chance to put on a rebuttal case, and it played out with an inordinate level of contention and almost as many objections to questions as answers. One witness under cross-examination had no questions to answer because Judge Treu sustained objections to every one posed.

The plaintiffs called just two witnesses — Eric Hanushek, a senior fellow at theHoover Institution at Stanford and an expert in using economic analysis in education, and Anthony Smith, a former superintendent of the Oakland Unified School District.

Both were used to amplify testimony of other plaintiffs’ witnesses and to refute testimony of defense witnesses. Hanushek spent most of his time on the stand defending his belief that using value-added measures of teachers is critical for evaluating their effectiveness and supporting a recent Stanford report that showed LA Unified charter schools did a better job educating students than traditional district schools.

While plaintiffs attorney Marcellus McRae steered clear of asking about an even more recent Stanford report, a survey of all California charters that showed they performed about the same as regular public schools, a lawyer for the unions, Peder Thoreen, went right at it, only to be swatted back by objections because Hanushek said he had been out of the country when the report was released, and he was not entirely familiar with it.

That struck some of the lawyers as a bit disengenuous in that Hanushek’s wife,Macke Raymond, wrote both reports.

“It is a pretty outrageous statement given that it was written by his wife,” Jonathan Weisglass, a lawyer for the defense, said during an afternoon recess. “And it, in fact, says that for the entire state of California, that the performance of charter schools is equal to or worse than traditional public schools. It’s far less favorable to charter schools than the LA study. So it’s really kind of odd and surprising that he was only familiar with that one.”

If I ever claimed in open court that I had never read Mrs. Jazzman's work, she'd make me sleep on the couch.