Interview with James Heckman

by on February 14, 2013 at 9:47 am in Economics, Education | Permalink

More than just the usual, this is a real interview, recommended.  Excerpt:

James Heckman:  Well, the reason why I’m skeptical is that the most salient work on Head Start is this new evaluation which came out last October. It actually came out later than I responded to Deming. I am skeptical for the following reason. It’s really heterogeneous, and I’m sure there are some very high quality programs and some very weak ones. The latest study showed very weak effects. That was a short-term followup. Head Start has never had a long-term followup.

I was surprised by the extent to which he defends Head Start, and to the extent he sees part of that program as Perry follow-ups.

Gary Saturday February 14, 2013 at 9:58 am

Heckman is great, but I didn’t see where the link went before I clicked, and suddenly realized I have a sense of compunction giving Wonkblog page views.

Jan February 14, 2013 at 10:27 am

Well that’s a foolish reason not to read an interview.

Andrew' February 14, 2013 at 12:11 pm

Certainly not after you’ve clicked…but then again maybe that is sunk-cost fallacy.

Rob42 February 14, 2013 at 1:11 pm

It’s not sunk cost fallacy to read the article after you clicked even though you would not have beforehand, because the price of reading has been reduced once you clicked. The price of reading the article is the sum of (a) your time and (b) giving a click to Wonkblog. Since you were unaware of (b) when you clicked, you thought the article would be worth the cost of (a), but presumably before you clicked, you would not have paid the price of (a) + (b) had you known it. After you’ve clicked, you’re still faced with the choice of reading the article or not, but now, the price of reading the article has dropped to just (a), which you are willing to pay (you’ve already inadvertantly paid (b), and there’s nothing you can do about this). So, you should read the article.

This is different than rain at a football game where the cost to attend is sum of (a) ticket price and (b) sitting in the rain. Here, you’ve already paid (a), but the (a) amount isn’t what would keep you from watching the gam, its the (b) of sitting in the rain, which you would still need to pay. That is why you don’t go to a rainy game even though you’ve paid the ticket price.

prior_approval February 14, 2013 at 10:03 am

I think the U.S. to return to the glorious days of Dickensian child development policies – after all, that was what fueled an empire to its global spanning power.

And look what happened when the Victorians abandoned such policies – a long term decline of the empire, not to mention declining rates of child employment and childhood diseases and mortality.

Marian Kechlibar February 14, 2013 at 10:11 am

I am not really sure what do you aim your sarcasm at.

Is every governmental program of education spending automatically efficient, because the original intent was to Help The Children? Are Programs For The Children exempted from the usual bureaucratic inefficiencies and Murphy’s law?

Surely not.

Yet, this kind of sarcasm that you employ worsens the policy debate, as it attaches stigma of a Heartless Child Eater to the opposition.

prior_approval February 14, 2013 at 10:42 am

‘Is every governmental program of education spending automatically efficient, because the original intent was to Help The Children? Are Programs For The Children exempted from the usual bureaucratic inefficiencies and Murphy’s law?’

Anyone who honestly thinks that what the government in the U.S. (federal/state/local) currently does for children even approaches the level of discussing its efficacy needs to confront such things as the fact that American society seems incapable of providing a level of basic health care to children, universally. This contrasts to Germany, as a very concrete example, where children are automatically insured – and where ensuring children’s health is considered a fundamental aspect of keeping a society functioning over the long term.

And for a bit of anecdotal data – I know a librarian who worked in a public school in Norfolk, home of the world’s largest naval base. She had many children in her school who had literally never touched a book in their lives – and yet, on this blog, we are discussing whether a program like Head Start has any role to play in a society unable to allow a preschooler to even learn what a book is.

America needs to confront its failings as a society (and trust me, a certain sort of commenter is just waiting with bated breath to explain what the ‘true’ failure is – and yes, there is a German word for it), and part of that failure is how poorly many American children are served by a system they are born into it.

Foreigner February 14, 2013 at 10:50 am

“America needs to confront its failings as a society”

You are such an American. The neverending apocalypse complex runs strong in you.

prior_approval February 14, 2013 at 11:02 am

I am American – but where does ‘confront its failings as a society’ become a ‘neverending apocalypse complex’?

After all, Germany was forced to confront its failings as a society, and since that nadir, has shown it is possible to have a strong democratic system exist, as the best prevention against such failings. Of course, it takes a certain devotion to historical honesty to face what one’s society did, and instead of denying it, to work against it in the present.

But you are right in one sense – reading places like here, you would think that America is doomed, as Americans are incapable of either learning or implementing successful ideas or policies from other societies.

Marian Kechlibar February 14, 2013 at 11:20 am

“… we are discussing whether a program like Head Start has any role to play in a society … ”

Every program has a role to play, but this still does not mean that every program is worth keeping. Pros and cons must be evaluated, otherwise the debate is reduced to magical thinking.

It may well be that Head Start is using money which could be spent on something else, more efficient.

I am also sympathetic to the federalism argument. The USA is too big and diverse for one-size-fits-all programs. Even Bundesländer in federal Germany have many schooling pecularities; and compared to the USA, Germany is compact and socially more homogeneous.

prior_approval February 14, 2013 at 12:09 pm

‘It may well be that Head Start is using money which could be spent on something else, more efficient.’

Or it could be that the very idea of having such a program is so unattractive to a certain segment of American society, that regardless of the costs to that society over the long term, it is better to have no program than to explore alternatives.

And that some people, such as Prof. Cowen in his previous post, even dispute that daycare/pre-school programs play a role in successful educational programs in other countries, even when the people responsible for those programs insist they are critical to successful educational outcomes over the long term.

Cliff February 14, 2013 at 12:27 pm

Or you could be completely wrong and your political opponents could be honorable and have the same good intentions as you.

And if your test of whether something is critical is whether the people responsible for it say it is critical, then I guess there is little in the world that is not critical.

JWatts February 14, 2013 at 12:28 pm

“Or you could be completely wrong and your political opponents could be honorable and have the same good intentions as you.”

Ideologue’s don’t tend to think that way.

Andrew' February 14, 2013 at 2:05 pm

People running programs insist they are important. Whuppedy doo.

prior_approval February 14, 2013 at 3:18 pm

‘People running programs insist they are important. Whuppedy doo.’

The people running what is considered to be the best educational system on the planet just, possibly, might, actually know what they are talking about.

But who cares what they think, right? – America is too exceptional to actually care about such things as actual results.

derek February 15, 2013 at 1:47 am

The Department of education was created in the Carter era. Is there any empirical evidence that it has had any positive effect on education in the US? Increased test scores?

The only rationale for a Federal bureaucracy like this is to make sure that areas that are less prosperous have enough resources to educate their children. Otherwise it is far away experts meddling where local initiatives would be far better.

bjssp February 14, 2013 at 10:10 am

Why are you surprised he defends Head Start so much?

scrooge February 14, 2013 at 10:10 am

“not to mention declining rates of child employment and childhood diseases and mortality”

But don’t forget, child obesity has gone through the roof

Marian Kechlibar February 14, 2013 at 10:14 am

Thanks, scrooge, for drawing attention to this strangely neglected problem.

How can the politicians seriously talk about “bending the curve cost” of healthcare, when the obesity trend predicts the very opposite? When the whole nation, on average, fattens 2% every year, the resulting epidemics of obesity-related illnesses, from high blood pressure through diabetes to arthritis, will in itself cause significant healthcare cost growth.

prior_approval February 14, 2013 at 10:27 am

It sure as hell has – the Victorians never needed to worry about that problem. But then, back in the glory days of a young queen’s reign, the poor houses and orphages were truly lean institutions, without any fat at all.

Cliff February 14, 2013 at 12:28 pm

Ah, so is it better-funded poor houses and orphanages that have led to the obesity? I think not.

prior_approval February 14, 2013 at 3:23 pm

No, I think it was mainly corn subsidies that have led in large measure to American obesity, with a sedentary life style playing a large role – and the pre-Victorians weren’t fans of cheap food either. Though the Victorians did get around to repealing this –

‘Overview. The Corn Laws were a series of statutes enacted between 1815 and 1846 which kept corn prices at a high level. This measure was intended to protect English farmers from cheap foreign imports of grain following the end of the Napoleonic Wars.

Note: in this context “corn” means grain of all kinds, not simply the vegetable corn.’

http://www.britainexpress.com/History/victorian/corn-laws.htm

Andrew' February 14, 2013 at 10:16 am

Quality really matters: No kidding. The proposal is essentially the parents are so bad that putting them in a 7-kid ‘family’ is superior.

JL February 14, 2013 at 10:20 am

As I said in the other thread, I don’t understand why Heckman puts so much stock in these few small experiments. Aside from lack of replication, the generalizability of e.g. the Perry program is dubious. Even if it worked in the 1960s, poverty in the 1960s was qualitatively different from today’s poverty, and these days even the poorest parents tend to have a lot more formal education than they had half a century ago. There’s no guarantee that the effects would the same today. And his Head Start comments seem like wishful thinking.

Rich Berger February 14, 2013 at 11:24 am

I agree with the small size experiments, but I believe poverty and parents in general are worse than they were in the 1960s. Out of wedlock births have skyrocketed since the 60s, especially among non-whites.

JWatts February 14, 2013 at 11:39 am

” Out of wedlock births have skyrocketed since the 60s”

While this is true, it is mitigated to a great degree by social acceptance of out of wedlock birth. We have two nephews/cousins that are living with their girlfriends and kids. The situation barely raises an eyebrow, let alone ostracization.

So I think an apple to apple comparison is the number of single parents now vs then, not the out of wedlock birth which may not mean anything at all if it still involves a nuclear family.

Rich Berger February 14, 2013 at 12:00 pm

Married couple does not equal boyfriend living with mother of children. I do not believe the current casual attitude toward creating children is beneficial for the children. I think was a major observation in Murray’s book “Coming Apart”.

JWatts February 14, 2013 at 12:33 pm

Agreed.

However, I think children living with cohabitating parents are better off than children living with a single parent. I think you are better off considering cohabitating parents as a subset of nuclear families and single parents as another set. Vs creating a set of married nuclear families vs unmarried families.

Rich Berger February 14, 2013 at 12:29 pm

Here is a piece from the Heritage Foundation with statistics on births legitimate/illegitimate over time, by group – http://www.heritage.org/research/reports/2012/09/marriage-americas-greatest-weapon-against-child-poverty

John B. February 14, 2013 at 12:08 pm

It may be true in US that wedlock births are skyrocketing among non-whites, but I consider it a racist comment anyway. If you compare it to 60s in any ethnic or racial group, it skyrockets everywhere. So there is definitely no causality among color and wedlock birth. And from my experience when I lived in Canada, I can tell you they handle their immigration policies well and have pre school service available for everyone without restricting it to any social group. I think it is a better approach since it always helps children to hang out together without being divided among poor/rich/black/white etc.

Jan February 14, 2013 at 11:36 am

You realize that the onus is on those who distrust the findings of the Perry experiment to replicate it and disprove the results, right? It was a study, there were findings, it has been the basis for even quite politically conservative states and locales to adopt publicly supported pre-school. This seems so obvious to me that the only real explanation I can think of is that the doubters are actually driven more by ideology than science and are afraid to actually do the study.

JWatts February 14, 2013 at 11:42 am

“You realize that the onus is on those who distrust the findings of the Perry experiment to replicate it and disprove the results, right?”

Are you ignoring the studies that indicate the effects of Head Start don’t last? Or are you arguing that we should replace Head Start with more intensive programs?

Andrew' February 14, 2013 at 11:48 am

“You realize that the onus is on those who distrust the findings of the Perry experiment to replicate it and disprove the results, right?”

No. It doesn’t.

Jan February 14, 2013 at 12:18 pm

Sure it is, the Perryists are winning. Armchair researchers aren’t going to end support for pre-school by just saying there isn’t good evidence whey they’ve nothing of their own.

Cliff February 14, 2013 at 12:31 pm

The Perry system is impossible to implement, so that particular argument is already “won”.

Andrew' February 14, 2013 at 2:07 pm

No it isn’t. We can keep going. For starters, maybe it would be better to test another idea and see what its order of magnitude of benefit would be with a small study. That’s kind of how science works.

Andrew' February 14, 2013 at 2:34 pm

Or, the ‘onus’ could be on the proponents to replicate the study to determine exactly how what they think worked actually worked. The “well, we spent $11k/kid/year and 40 years later we get some results” is not very convincing without mechanisms and intermediate proxy variables. Are kids ‘blind’ to this intervention? If not, does being special cause a placebo effect to motivation? What about once everyone is special? Even more concerning, are the control kids blind to the intervention?

Evaluation of Education Projects: the Case of the
Perry Preschool Program
EDWARD M. GRAMLICH
“How all these good things could happen as a result
of a brief preschool program which only lifted IQ
scores for a short time is something of a mystery.
High Scope’s own rationale focuses on the interaction
between a person’s capabilities and the
environment. In this view, it is critical that students
be well-prepared for the highly demanding school
experience when it starts.”

Maybe black kids with low IQs just aren’t ready for our arbitrary everyone-starts-kindergarten at the same age system. Maybe red-shirting would achieve the same results at no cost.

Jan February 14, 2013 at 12:26 pm

I’m not endorsing any particular changes to Head Start, nor am I saying that those studies are wrong. But if people want to talk about the weaknesses of the Perry study as a surrogate for pre-schools generally, then they need to replicate the study and disprove the findings. If you want to rail against Head Start and there is good evidence that hasn’t been challenged by contradictory findings, that is fair game.

JWatts February 14, 2013 at 12:41 pm

Then you are making a strawman argument. Who exactly is talking about the weaknesses of the Perry study as a surrogate for pre-school programs generally?

It’s widely acknowledged that no one is talking about replicating the type of pre-school that the Perry study covered. So the comments are along the line that the Perry study is irrelevant to the conversation. Which it obviously is.

The topic is primarily about ‘Head Start’. There have been studies of ‘Head Start’. The general conclusion is that the program has no long term benefits.

Jan February 14, 2013 at 4:11 pm

Criticism of the Perry study comes up. every. single. time. You know this, man.

JL February 14, 2013 at 4:28 pm

You realize that the onus is on those who distrust the findings of the Perry experiment to replicate it and disprove the results, right?

Heckman wants tax-payers to fork out billions of dollars every year from now till eternity (cf. Head Start which goes on despite lack of positive results) based on a small study from the 1960s. It’s ridiculous to suggest that the onus is on his opponents to prove that this would not be a reasonable investment.

Recently, there’s been a crisis in psychology because it turns out that the results of many famous experimental studies cannot be replicated. Why should we believe that the small, never-replicated Perry experiment, which was conducted in a very different society and time, would, if replicated today on a mass scale, produce the same results? What we need is a replication study in a large modern sample.

Steve Sailer February 14, 2013 at 6:42 pm

“You realize that the onus is on those who distrust the findings of the Perry experiment to replicate it and disprove the results, right?”

Uh, no, when there’s a politically popular finding 40 years ago, it’s only natural to expect some replication.

Jan February 14, 2013 at 9:21 pm

Which is why righter-wingers and pre-school skeptics are welcome to fund the experiment. They don’t.

john_d February 14, 2013 at 11:05 am

Are we missing the issue here? Isn’t the true value of public preschool that it frees up low-income and single mothers so that they can return to the work force and develop vocational skills, and our politicians couch the program in terms of education so that it seems like a handout to people who deserve it (kids who had no say in their circumstances of birth) and not to those who don’t (low income mothers who are having kids they can’t afford)?

Nylund February 14, 2013 at 12:22 pm

There’s definitely truth to the argument that early childhood education programs are, in part, subsidized baby-sitting. For low wage single parents, the cost of childcare can be prohibitively expensive. Practically all the money you earn at your low-wage job goes straight to paying for the childcare (if you don’t have a free family baby sitter, like a retired grandparent). The incentive to work is low.

Two of the proposals in the SOTU address were related to that. Up the wage rate and lower the childcare costs. The idea is to increase the opportunity cost of not-working and lower the cost of working, hopefully tilting the scales in a way where the incentive to work is larger than the incentive not to.

Or, to put it in plain English, “Go out and work! We’ll pay you more and we’ll watch your kid for free!”

Of courses, there are costs and someone will bear them, namely, the non-poor. It’s a roundabout way of redistributing resources from the non-poor to the poor in order to incentivize low-skilled single-mothers to work more. But yes, politically there are people who hate the idea that their money is being used to give free stuff to poor people, so you couch it in terms of educating our little angels. In reality though, whether or not it helps their educational attainment is secondary. Policy makers hope there’s a positive effect there, but it may not be the main justification for the program.

PS. I think the same argument somewhat also applies to the ideas of year-round schooling. That too is probably more about providing year-round day care than anything to do with educational attainment.

JWatts February 14, 2013 at 12:51 pm

“Two of the proposals in the SOTU address were related to that. Up the wage rate….Of courses, there are costs and someone will bear them, namely, the non-poor.”

In an environment of high unemployment with background of automation slowly eroding the viability of low knowledge workers, I would expect increasing the minimum wage to have the opposite effect.

You won’t end up with less poor people, you’ll end up with more non-working poor and less minimum wage employees.

Ed February 14, 2013 at 3:07 pm

This is a decent point. The thinking here is a step or two behind, and we should be considering whether it is a good idea to give low income, low education people more incentives to take jobs that simply won’t exist.

And while I favor a citizens’ dividend/ guaranteed income or some equivalent, if automation really does advance as much as predicted, subsiding child raising by people who will be on welfare most of their lives is probably not the best idea either.

And actually I think this will apply to lots of (ex-) workers in other categories. But the conventional wisdom that automation will be removing low skill jobs permanently doesn’t match up well with the conventional wisdom that we really have to more low skill workers participating in the workforce.

maguro February 14, 2013 at 4:48 pm

If what want is fewer poor people, maybe we should stop importing them by the million.

Jan February 14, 2013 at 9:23 pm

I’m as opposed to foreigners as the next white guy, but we haven’t been importing them and recently they’ve done a good bit of the old self deportation. I think we are ready to rock.

Cliff February 14, 2013 at 12:33 pm

Even for “wealthy” parents, the cost of child care is a huge disincentive to working. I’m not opposed to the idea of free daycare (probably a voucher system makes the most sense), though I would want to look at what you really gain in terms of additional working parents and additional children.

John B. Chilton February 14, 2013 at 12:19 pm
ziel February 14, 2013 at 1:56 pm

I’d like to know the source of Heckman’s contention that the black/white IQ gap had dropped to 2/3rds of a standard deviation.

Jason Malloy February 14, 2013 at 4:59 pm

Heckman is clumsily alluding to this research by Flynn and Dickens, estimating that young black children (not the black population!) have an IQ of about 90 since the 2000s.

Here’s an analysis which suggests much more minimal gains since the 1960s.

Jason Malloy February 14, 2013 at 5:26 pm

Heckman: “The black population was one standard deviation below and now it’s about 2/3. ”

It’s disappointing to see Heckman be this sloppy with important facts (and he is one of the few sociologists who emphasizes these gaps as very important).

From James Flynn’s own ‘What is Intelligence?’ (2007): “By 2002, the mean IQ of black American children aged 4 had risen to 95.4. This puts them less than 5 points below white 4-year-olds at 100. However, by the age of 24, blacks lose fully 12 points and sink to 83.4, almost 17 points below whites. In other words, they lose 0.60 points per year as they age.” (p. 123).

ziel February 14, 2013 at 6:00 pm

Thanks Jason.

Scoop February 14, 2013 at 2:04 pm

Anyone aware of any research on what effect, if any, the creation of these programs has on how many kids people have?

Stuart Buck February 14, 2013 at 3:37 pm

Heckman says, “A small sample would actually work toward not finding anything. You have a limited number of observations. You would argue that the statistical observations would not be very great, and there would not be much of them.”

This is a bit tricky. Given a small sample size, you’d have a harder time finding statistical significance unless there were (or seemed to be) a large effect size. But Heckman seems to be implying that given the large effect size here, the fact that it came from a small sample makes it more reliable — which wouldn’t be true, right? Small samples have more sampling variability and less precision than large samples, and thus present a higher chance that a large effect size might be spurious.

Michael Bishop February 17, 2013 at 4:38 pm

Yes.

Edward Burke February 14, 2013 at 6:19 pm

Not unduly opposed to WaPo but I have no immediate plan to read the article, as I just heard Heckman interviewed by NPR’s Robert Siegel.

The Nobel laureate argued that while Head Start poses a modest institutional challenge to the institution known as “the family”, the service it renders is what (I paraphrase) “most parents want for their children”.

Yet many in the “most parents” population seem ready and perfectly willing not to want for their children what they want Head Start to give their children.

Of course, society at large wants proper educational, social, and economic outcomes; yet society at large has every reason to expect families to take the initiative FIRST to provide for themselves, instead of unburdening themselves of the responsibilities of rearing children, without a Federal or state or local government robbing them further of incentives to value education for all the things education confers, without governments foisting further dependency on government social guidance, without governments turning more and more families into government clients.

prior_approval February 15, 2013 at 12:13 am

‘yet society at large has every reason to expect families to take the initiative FIRST to provide for themselves’

And yet, the society with the world’s finest educational system explicitly disagree with you. In the Finnish view, members of society have every reason to think they are participants in society, entitled to share in both its benefits and its responsibilities.

Yeah, it would never work in the America of this comment section.

derek February 15, 2013 at 1:52 am

The US has a problem with it’s existing schools. The educational authorities have children for 12 years and can’t seem to do a decent job. What we really need it so start them at 3 years old and finish them at 26, then for sure they would be well educated.

Fix the damn primary school system first. That is hard, very hard, and it seems that those who pontificate the most about education are the least willing to do anything about it. And no, I don’t consider spending money doing anything about it.

The Anti-Gnostic February 15, 2013 at 3:51 pm

In the Finnish view, members of society have every reason to think they are participants in society, entitled to share in both its benefits and its responsibilities.

That’s because they’re Finns living in their Finnish homeland. Is there even anybody in the whole place beyond the sixth degree of consanguinity?

America by contrast is just a giant flea market. You come here, you hustle, you make money, but it’s not really “home,” the people hustling in the stall next door aren’t really “neighbors,” much less “family.” It’s more a giant commons, and not much incentive to keep it clean, efficient or sustainable.

Diversity and a functional welfare state: chose one.

Steve Sailer February 14, 2013 at 6:45 pm

Obama’s view is that welfare moms are so incompetent at raising their own children that the taxpayers should pay to take their toddlers off their hands.

I propose that Obama start a major campaign to encourage welfare mothers to “Stop at One.” Obama should praise welfare mothers who have devoted all their resources to raising one child better than they could have raised a whole passel and he should shame welfare mothers who have more than one child.

Jan February 14, 2013 at 9:49 pm

Maybe his view is that more moms could go to work if they wouldn’t have to immediately put most of their earnings toward childcare.

While we’re shaming poor black women for having kids, let’s also make sure Planned Parenthood is de-funded and that insurance doesn’t have to cover birth control.

The Anti-Gnostic February 15, 2013 at 3:16 pm

They can find the time and resources to afford a living human child who will be with them for two decades but a condom, no way! (NB: masturbation is even cheaper.)

But I agree with your basic point. If we’re going to subsidize anything, it should be birth control. In fact, if you have to rely on public assistance to feed your family, birth control should be mandatory. I can’t think of a single reason why this should not be the case.

The best thing that conservatives could ever do for their cause is to take the Left up on its “reproductive choice” shibboleth.

Steve Sailer February 15, 2013 at 12:04 am

President Obama is a big college basketball fan, so he’s heard the phrase “One and Done.” He should recommend that to all welfare mothers and high school dropouts: One and Done. If you can’t get through high school or off welfare, than you owe it to society to have no more than one child.

One and Done.

D February 15, 2013 at 3:17 pm

“It’s disappointing to see Heckman be this sloppy with important facts”

Flynn did the same thing in his book “What is Intelligence”.

People get sloppy when the facts seem to bend a certain way.

Gideon February 15, 2013 at 10:27 pm

“prior approval”
“And yet, the society with the world’s finest educational system explicitly disagree with you. In the Finnish view, members of society have every reason to think they are participants in society, entitled to share in both its benefits and its responsibilities.
Yeah, it would never work in the America of this comment section.”

Remember that we all now live in an America in which we confuse equality of opportunity with equality of outcome. We guarantee the former and we have liberty. We guarantee the latter and we have tyranny.

prior_approval February 16, 2013 at 10:24 am

‘Remember that we all now live in an America in which we confuse equality of opportunity with equality of outcome’

Except that the ‘outcome’ in America in terms of measurable things like the Gini co-efficient show that the above belief is false, at least when seen through a factual perspective.

Comments on this entry are closed.

Previous post:

Next post: