Here we go again

By Kathy G.

Last week, Megan McArdle briefly interrupted her union-bashing fiesta to serve up the following confused and confusing hot steaming pile o' truthiness:

I'm familiar with the research on parental skills and early childhood intervention. I just don't know what to do with it. So far I have not seen a single successful early childhood intervention that is even arguably scaleable: you're talking about intensively monitored programs using top-notch personnel, all of whom are deeply committed to the project's goals and procedures. The longest data set we have is, AFAIK, the Perry Pre-School Project, which produced exceedingly modest gains for a pricetag of more than $20,000 per child per year in today's dollars. Again, that's with a small program of highly committed staff.

What we got instead was Head Start, which produces small gains that most evidence suggests disappear a few years after the kids exit the program. Even if we wanted to do Perry Pre-School, or something even better, nationwide, where would we find the staff? Pre-K sounds great, but it's very likely to be slightly glorified baby sitting outside of affluent school districts that don't need it in the first place.

Since we're not (I hope) going to take kids out of the disadvantaged homes they are born to, the schools are what we're stuck with. And there are programs that work--at least, better than what we have now. They just bore the hell out of the teachers.

She concludes, "I'm not against early childhood intervention, if it works."

It's an odd post. First she says that the only (marginally, according to her) effective programs (like Perry) are too expensive and not scaleable. Next she says the large-scale programs like Head Start don't work. Then she says well, if they did work, such programs would bore the hell out of the teachers. Even if that's true, does she seriously believe that's a reason why we as a society should not do it? Memo to Megan: in the real world, most people have jobs that are frequently -- sometimes completely -- tedious. That is why they are paid to do them.

And then there's that sentence about "I'm not against early childhood intervention, if it works." This after a long list reasons of why such programs don't work, and if even they do, why we shouldn't enact them anyway. Wtf?

Okay, let's take on those factoids one by one.

But first, some background: as some of you may  know, the leading researcher in this country on the impact of early childhood intervention is the Nobel prize-winning University of Chicago economist, James J. Heckman. Heckman is a University of Chicago economist in every sense of the word -- a very conservative dude. He's also a quant god -- econometric nerds know him for his famous "Heckman two-step" technique, which is a method of controlling for selection bias.

Anyway, Heckman has done groundbreaking work on the impact of early childhood intervention. And even he -- a hardcore conservative who's none too fond of government spending -- thinks it is something very much worth doing. Unlike we lefty types, he doesn't support it for equity and social justice reasons. He supports it because it's economically efficient. As he has shown, investing in early childhood education produces a substantial rate of return. In cost/benefit terms, it is very much worth our while to spend money on early childhood programs. Heckman, writing for that well-known organ of leftist command-and-control propaganda, the Wall Street Journal opinion page, sums up the case for early childhood intervention:

It is a rare public policy initiative that promotes fairness and social justice and, at the same time, promotes productivity in the economy and in society at large. Investing in disadvantaged young children is such a policy. The traditional argument for providing enriched environments for disadvantaged young children is based on considerations of fairness and social justice. But another argument can be made that complements and strengthens the first one. It is based on economic efficiency, and it is more compelling than the equity argument, in part because the gains from such investment can be quantified -- and they are large.

There are many reasons why investing in disadvantaged young children has a high economic return. Early interventions for disadvantaged children promote schooling, raise the quality of the work force, enhance the productivity of schools, and reduce crime, teenage pregnancy and welfare dependency. They raise earnings and promote social attachment. Focusing solely on earnings gains, returns to dollars invested are as high as 15% to 17%.

Traditionally, when social scientists analyzed the effectiveness of early childhood education programs, they'd look at the impact on IQs and test scores. And admittedly, on that score, the results of these programs tended to be extremely disappointing, especially in the long-term.

But then researchers began to look at other outcomes, such as the impact on school completion, crime, welfare dependency, teen pregnancy, employment, and earnings. And once those outcomes were factored in, and quantified, the early childhood education programs began to look a whole lot more successful. Certainly, they've been proven to have a much higher rate of return than later interventions such as "reduced pupil-teacher ratios, public job training, convict rehabilitation programs, tuition subsidies or expenditure on police."

One of Heckman's singular contributions to this research is his focus on non-cognitive skills. As he has shown, success in life depends on a lot more than IQ. He writes that "noncognitive traits like motivation, persistence, time preference, and self control" are, like cognitive traits, important in "explaining schooling and socioeconomic success." There is evidence that, even if early intervention programs do little over the long term to increase cognitive skills, they have lasting and significant effects on the noncognitive front. And it's the improvement in noncognitive skills that seem to be driving the positive outcomes of the early childhood interventions.

Okay, back to the McArdle post. Her first argument is that even the best of the early childhood intervention programs, the Perry Pre-School Project, produced "exceedingly modest effects."

Well, in short -- no. Here's a table from one of Heckman's papers that looks at Perry's costs and benefits:

As you can see right there in black and white, when you factor in the costs that are saved in terms of crime, child care, welfare, K through 12 education (i.e., reduced special ed costs), plus the extra tuition costs of participants who went to college, plus the beneficial effect on participant earnings and future generation earnings (the earnings of the participants' descendants), the benefits-to-costs ratio of Perry was approximately 9 to 1, with rates of return around 15% to 17%. Program participants were significantly more likely than a control group to graduate from high school, to be employed, and to earn more, and teen pregnancy and crime rates were substantially lower. How this program could be considered anything less than an outstanding success is beyond me.

And btw, Heckman says that if the program were administered today, it would cost about $9,785 per participant per year in 2004 dollars, as compared to around $7,500 per pupil per year for "ordinary public education." I don't know where McArdle got her $20,000 figure, but it's misleading, to say the least.

Okay, I think we've established that the gains from Perry were far from "exceedingly modest." McArdle's next tack is to imply that large-scale programs like Head Start have been failures. This, again, is a crock. Go back to the table above. The Chicago CPC program is the kind of large-scale program that McArdle implies could not be successful, but as you can see, it had a benefits-to-costs ratio of well over 7 to 1. Program participants had significantly higher rates of high school graduation, and significantly lower rates of special ed, grade repetition, and juvenile arrests than did a control group.

And speaking of Head Start, McArdle says that it "produces small gains that most evidence suggests disappear a few years after the kids exit the program." More truthiness! The gains in IQ and test scores that that program produces are indeed short-lived, but studies suggest that there are other positive effects. For example, this paper, which was published in 2002 in the premier economics journal, the American Economic Review, produced the following results:

Among whites, participation in Head Start is associated with a significantly increased probability of completing high school and attending college, and we find some evidence of elevated earnings in one’s early twenties. African Americans who participated in Head Start are significantly less likely to have been charged or convicted of a crime. The evidence also suggests that there are positive spillovers from older children who attended Head Start to their younger siblings.

This paper by Janet Currie published in the 2001 Journal of Economic Perspectives found that Head Start has a favorable benefits-to-costs ratio at least over the short term (via improved health and nutrition, and prevention of abuse and neglect) and medium term (via preventing special ed and grade repetition). Long-term effects, however, are hard to assess, says Currie, mainly because the "practical problems" of tracking large numbers of individuals over many years have caused there to be "very few well-designed studies of longer-term effects."

I dunno about you, but the extent to which Megan is, as she claims,"familiar with the research on parental skills and early childhood intervention," seems highly doubtful. Once again she appears to have ignored many important facts, and then gone ahead to twist and retrofit others so as to make them ideologically convenient. And if she responds to this post, I expect that, as per usual, she'll claim that she's been misunderstood, and that the words I quoted in her post connote something other than their plain meaning. And no doubt, she'll dig up some brand new reasons not to support early childhood intervention.

But it's worth pointing out that some of the more good faith, intellectually serious comrades of hers on the right do support early childhood intervention. There's Heckman, of course. And there's also Brink Lindsey, libertarian and vice president for research at the Cato Institute fer chrissakes, who had this to say in a recent article in The New Republic:

The idea that class-based cultural differences contribute to academic underachievement is cause for consternation across the ideological spectrum. Let's start with me and my fellow libertarians. We insist on the central importance of individual responsibility for the healthy functioning of a free society. Yet, by the time people become legally responsible adults, circumstances not of their own choosing--namely, how they were raised and whom they grew up with--may have prevented them from ever developing the capacities they need to thrive and flourish. Which raises the possibility that government intervention to improve those circumstances could actually expand the scope of individual autonomy.

For example, preschool enrichment programs--along the lines of Head Start, but more intensive and beginning with even younger kids--offer some promise in counteracting the negative influences of a disadvantaged upbringing.

Of course, James Heckman and Brink Lindsey would disagree with me over how to implement pre-K programs. They would support vouchers, which I wouldn't, and they probably wouldn't want a universal program, whereas I (like Hillary Clinton and Barack Obama) would. A universal program, it is true, wouldn't have such an impressive benefits-to-cost ratio, because it would include a lot of kids who are doing pretty well and wouldn't benefit to the extent that poorer kids would. But as I've written before, I find means-testing problematic, and I think it tends to hurt the effectiveness and political viability of many programs, and also to create perverse economic incentives. So I think universality is the way to go.

At any rate, it's encouraging to see that some folks on the right are at least down with the general concept of investing in early childhood intervention programs. Personally, I believe that such programs are among the most important solutions to the problem of inequality that plagues this nation. By the time many disadvantaged children begin school, they are already so far behind that they can never catch up. I think many middle class people would find it difficult to imagine the profound depths of cultural and intellectual deprivation many poor children experience. Lindsey quotes the following results from a famous study:

Professional parents directed an average of 487 "utterances" per hour toward their children, as compared to 301 for working class parents and only 176 for welfare parents. The quality of those utterances was also very different: Among professional parents, the ratio of encouraging to discouraging utterances was six to one; for working-class parents, the ratio slipped to two to one; and welfare parents made two discouraging utterances for every encouraging one. The consequences were predictable: By the time the children in the study were around three years old, the ones from professional families had average vocabularies of 1,116 words; the working-class ones averaged 749; the welfare kids, 525.

I believe that in November, we'll elect a Democratic president and substantial Democratic majorities in Congress and the Senate. I'm hopeful that we'll end up with a fully funded Head Start at the very least, and I think we have a decent shot at universal pre-K as well. The evidence as to the positive effects of these programs continues to accumulate, the costs are modest relative to the benefits, and the political climate will be friendly. It's long past time for us to take action on this issue.

Comments

[Uncle Vinny]

Outstanding. Think the Atlantic might dump Megan for you? At the least, I hope Bloggingheads will throw you an invite soon...

[Daniel S. Goldberg]

Kathy,

This is a great post, and you might also be interested in the robust evidence suggesting that one of the best ways to improve population health is to invest substantial resources -- whether public or private -- in early childhood development.

I've posted on this here:

http://www.medhumanities.org/2008/03/on-disparitie-1.html

and here:

http://www.medhumanities.org/2008/02/on-brain-imagin.html

There is a wealth of resources on the subject, and the fact that McArdle does not even mention the convergence between social policy and health policy is troubling, indeed. That is, even if one rejects the notion that investment in early childhood development is good social policy, inasmuch as we are concerned with improving population health, the best evidence suggests that one of the most promising means for doing so is through such early childhood development.

[Michael Greinecker]

Great post Kathy, I´ve learned a lot.

[Barry]

Kathy, I posted a link to this post on Megan's blog.

Uncle Vinny, the Atlantic certainly didn't hire Megan for facts, analysis or anything else but blithe glibertarian hackjobs.

[Retief]

I believe the boring methods coment is refering to "Direct Instruction". It's a curiously authoritarian hobby horse for a libertarian to be riding, but hobgoblins and little minds and all that.

[Ed at AFT]

Isn't Perry a two year program? That might explain some of the cost difference.

(And this is a great blog).

[Suzanne]

I love Head Start and universal pre-k, but early learning starts at birth. If these programs are great, imagine how much more effective they'd be if not only were they expanded so more kids could access them, but aged down so that infants and toddlers could get affordable, quality care and education, too. I guess I can just dream on, given how much resistance there is to the programs we have today...

Thanks for the excellent post.