In the comment thread the other day on universal pre-K, InLightened (one of our conservatives in case you don't know) alluded to the report from last year about Head Start's failures. I'd missed that in real time I admit, so I went back and read it, or read through it. You can do the same; here it is.

It's pretty terrible, there's no denying it. Basically, the researchers found that Head Start worked fairly well, emphasis on fairly, for the time that little children were in it, and maybe a year afterward. But by the third grade and in many cases by the first grade there were no appreciable differences between the Head Start kids and the control group kids in terms of social and emotional development, in terms of health and well-being, in terms of learning development. No difference.

Here's a paper that's less pessimistic without being pollyannish and optimistic. But it's hard to deny it. Taxpayers spend $8 billion a year on this program, and we aren't getting our money's worth. You can read the links for the social-sciencey explanations. I would suspect that a good part of the explanation just lies in the simple fact that some people are really good at their jobs and other people are really lousy at their jobs.

The people who follow these things closely, for example, always talk about the problem of "scaling up." That is, there's a terrifically successful program in some county or city somewhere, and then people try to take it to the state or national level and results are mixed. And they ask: Why? Well, common sense suggests to me that at that local level, you probably had a great and energetic leader, and you just can't have one of those everywhere.

So far, we're in the observation stage, and our differences of opinion aren't that great. But now we start to discuss what should be done about these problems, and here, obviously, is where liberals and conservatives really diverge. Liberals will tend to say that these problems are no reason to end the program, and sure, we need to fix it. Conservatives don't trust that liberals can or on some level event want to fix it, as long as the money keeps coming.

I think this has been a big problem of liberalism, and I've written this in the past, especially back when I covering New York City and saw a lot of institutions, like the City University of New York for example, that weren't functioning well at all but that people in charge were loathe to reform. But I also think that the absence of certainty about whether a program will work is a pretty lousy reason to be against it.

Now I'm talking about risk. We accept the existence of risk in the private sector. Guy starts a restaurant, gal invents a device, company gambles its future on a certain innovation. Each of these may or may not work. We don't care. It's their money. But Head Start is our money. We're entitled to care. We're entitled to demand good results. However, we should also realize that risk of failure is real, as in any human endeavor. Not that that should ever be an excuse for any crappy administrator, but it should be recognized as part of the reality of trying to solve a problem as complicated as improving the cognitive abilities of poor children who grow up with parents who won't even read them a damn book.

David Brooks writes well about this today, and so does Jonathan Cohn, writing more specifically about the scaling-up problem and Obama's proposal. And speaking of Obama, in fairness to him, conservatives, this brief report from the New America Foundation suggests that with his universal pre-K proposal, Obama seems willing to acknowledge Head Start's problems and revamp it to target younger chldren and then sort of fold it into the new pre-K programs that would have stricter rules about the qualifications of teachers than Head Start has ever had. So give his Education Department credit. They're not pretending there are no problems here.