share interactive transcript request transcript/captions live captions
download
|
MyPlaylist
SPEAKER 1: This is a presentation by Human Development Outreach and Extension at Cornell University.
SPEAKER 2: It's one of these paradoxes that the people who least need introductions are the ones who get the longest introductions. And today's speaker certainly falls in that category. But she warned me beforehand that I had to keep my introduction down to one minute. So that means I'm throwing away 90% of it. Rich mentioned that the previous speakers in the [INAUDIBLE] symposium series have been luminaries, and today's speaker falls squarely in that category.
And like the last two, Eleanor Maccoby and Al Bandura, she's won the James McKeen Cattell award from APS for a lifetime contribution to the scientific application of psychology. She's also won a major lifetime contribution for policy application from APA. She's won a number of major divisional awards, including the Urie Bronfenbrenner Award from division 7 for developmental psychology, the Nicholas Hobbs Award for lifetime contribution for division 37, which is children, youth, and families. She's won others. I can't even remember.
She's also been recognized, not only in our field of psychology, but she's won major awards from the Association of Political Science, the Association of Policy Analysis, and probably some others. Just the quantitative stuff is daunting. She's published over 400 articles and 17 books. She's very highly cited. Her 97th book with Greg Duncan I think is fast becoming a citation classic, Growing Up In Poverty. She, if you Google her, comes up with I think 29,000 entries. I learned this the other day in preparing these remarks. I didn't read all of them. But the first few looked very positive.
[LAUGHTER]
But anyway, enough of these quantitative decoration things. Let me just cut to substance right away and say that there are few problems in all of the social sciences more important than this, more important than achievement gaps between majority and minority students, between rich and poor students. If you look at surveys of school district superintendents, of members of Congressional staffs, and so on, they all point to this and they all say, what do we know about it. More importantly, what can we do to narrow the gap. Because that gap is so predictive of a whole basket of important developmental outcomes. It predicts lifetime earnings. It predicts involvement with juvenile criminal justice systems, welfare dependency, longevity, mental health, and a whole bunch of other things that I've forgotten. It's exceedingly important.
And today's speaker was one of the first, I think, to recognize that the real action is preschool. It's there because the achievement gap that is so often lamented when people look at national assessment of educational progress scores for eighth graders, or 10th graders, or 12th graders, it's already existing before kids start school. There's some debate about whether it grows at all from kindergarten through 12th grade. But if it does grow, it doesn't grow much. Because you're already saying 2/3 to a full standard deviation difference before kids start formal schooling.
So people like Jim Heckman, who won the Nobel Prize in economics five years ago, and [INAUDIBLE] and these folks, Grismer, they've done this cost benefit analysis saying, you really need to invest in preschool readiness to narrow the gaps. And today's speaker was there a long time before, I think, this became a generally condensed view. So without further ado-- was that fast enough, by the way?
JEANNE BROOKS-GUNN: Thank you.
SPEAKER 2: Brooke, it's really a pleasure and an honor to introduce you.
JEANNE BROOKS-GUNN: In any case, again, I'm thrilled to be here, and I'm thrilled to be lecturing at this particular named lectureship. As Steve said, I'm interested in looking at equity issues. I've been interested in equity issues since I was in high school, but never dreamed I'd be doing work on equity and gaps in school readiness. And I'm really pleased that I had the opportunities to do this kind of work.
So what I want to do today, if this makes sense to everybody, is talk about this particular topic. And what I'm going to do, rather than going through a lot of my research on this topic, what I want to do is outline a premise which undergirded a special issue of Future of Children that I co-edited with CC Rouse, who's an educational economist at Princeton. Future of Children is a journal. Have any of you read that journal? Yeah. It's a journal that Packard Foundation ran for years, and then they did a pass off to Princeton and Brookings.
The first issue that Princeton and Brookings took over happened to be on this topic. Why? Because I live part of my life at Princeton, which is a long story. And when these guys got the grant, we were all talking about what we ought to do. And I said, well this is what I'd like to do. And if Columbia had gotten this grant, that's what we would have done our first issue on. So my pals at Princeton said, OK, that's a really good topic. We'll take that as our first issue. And the first issue that they did was on this particular topic. And again, I've been doing work in this area for probably 20 years at this point.
What I'm going to do is go through kind of the arguments and how we structured this particular issue. And I want to do it this way because it shows how you can take our very basic research that many of us have done in this field and put what you might say an economic policy twist on it to come up with implications about what you ought to do to reduce this gap. And that's a slightly different way than I would have thought before I collaborated with demographers and economists. So I hope you guys find this useful. If you don't, of course you can tell me. I love good arguments.
As Steve said, we're concerned about racial gaps. Here is an example of what he was talking about. I'm using NAEP data here from fourth grade and 12th grade. So theoretically, these would have been the same kids given the years that I took the data for. And I looked at the number of kids, percentage of kids on NAEP, who were above what they call the proficiency level by race and ethnicity. And what I want you to look at is the fact that at grade 4 and grade 12, the percentages of proficient are about the same, a, and b, the gaps, the racial ethnic gaps, are about the same at both grade levels.
So the question from these kinds of data and from the book that was done by Jenks and Philips in-- now I can't remember-- thanks, '98-- which we did one chapter in Young Children-- what this is really saying, as Steve was saying, is that these gaps exist early. I will show you some data from children at age 5, and age 4, and age 3 where the same gaps exist.
In general, there is a lot of debate about whether these gaps get bigger over time, or whether they reduce. I think the best evidence to date is that you do not see a reduction in these ethnic test score gaps over the years of school. And in a way, I've always found this very surprising. Because you would think possibly that the schools might be reducing these inequities. The case is that our schools aren't. That's another reason to go back and start looking at young children.
So we take this as our given. And I should say I'm going to present all data on race and ethnicity. I could tell essentially the exact same story for social class. But we decided, when CC and I decided to do this special issue, we wanted to focus on race and ethnicity. Because we think that not enough is being said about these gaps. And that if you talk about social class, sometimes you can kind of just sweep under the rug some of the ethnic disparities that exist in this country that are very alarming. But we could have. If you want to think, if your frame is more a class frame, these data will pretty much, what we're showing, support the same things that you see in class. Why are they so, so similar? Because black and Hispanic kids are so much more likely to be poor. So that's why the two stories overlap so much.
What are we looking at for school readiness? We're looking, in this particular volume, at academic skills and social and emotional skills. I really think a lot more effort has to be put into looking at disparities in social and emotional skills. I have less data on that. I'm not going to present a lot of that today. We have some. The reason we don't have such great data is one of the things we wanted to do was use national data sets or very large data sets for these analyses. Most of our large data sets in the United States on children and adolescents have not done a terrific job at measuring emotional skills. We measure behavior problems, aggression, and anxiety or depression.
But we don't look at impulse control particularly well. We don't look at things like following directions. We don't look at regulation across a variety of settings, and we don't look at attentional processes. I'm actually starting some work now on attention-- not talking about that today, but if there are any questions, I'd love to talk about it. Because I think that attention is a fascinating process because it really is-- in a way, it has elements of regulation as well as learning and cognition. And so it kind of bridges the cognition and emotional very nicely. And we have some kind of cool data coming out. But clearly, we can identify these different ways that this is where we have the data. So that's what I'm using here.
Why are we looking at the preschool? Besides the fact that we see these ethnic gaps early, the fact is we know that there are correlations or associations in longitudinal studies between having low school readiness scores and the four different outcomes that I've listed here. Clearly, the effect sizes are somewhat different. But you could do a whole lecture just on this. That's not where I'm going.
Basically, the kind of argument that people have put forward, especially with the early childhood education programs, is that if you get a child's school readiness score higher than it would have been through intervention, what you've done is alter the trajectory a child is on. You've kind of just bumped the trajectory up, which should have, one would hope, long-term impacts over time. Interestingly, there is less data on that from people actually doing longitudinal studies and looking at trajectory changes, and then if the kid stays on it.
Frank Herzenberg and I did some analyses in the early '80s using a longitudinal study of poor black teenage mothers and their offspring in Baltimore. No, I'm in New York. That was in Baltimore. And there are some really good data now coming out of some of the English longitudinal studies. You know, they have four birth cohort studies where you have thousands and thousands and thousands of kids. So you can kind of look at what happens to these trajectories in a large sample.
Just to show you some of the data from 3 and 4-year-olds, this is using the vocabulary test score, the PPVT, which is a receptive vocabulary test, as most of you know. And all I want to do here is to show you the distributions from a national sample for PPVT scores for black and white kids. Essentially, we're talking about 3/4 to 1 standard deviation of a difference between blacks and whites on this particular test. This test and full-scale IQ tests usually show this 3/4 to 1 standard deviation difference.
I'm going to talk about standard deviation effects throughout this talk. You guys all know what this is, but some people don't, so I usually put this up. Most of our tests are normed to have a standard deviation of 15. If whites are scoring 12 points higher, you would have a gap of 12. And that would be 80 standard deviation units, or 80% would be the effect size.
So if people talk about effect sizes or standard deviation differences, if you are more comfortable thinking about points, just do the math in your head. Remember that the whole story has to do, when you talk about this with the standard deviation, not the mean level. And we can talk about that too. Actually there is some very interesting data about whether you look at standard deviations, or you look at mean levels. But for this talk, I want to talk about standard deviations.
Now what does a 1 standard deviation difference mean? Can you guys all read that so I don't have to read it? Good. At our conference, we have this really neat guy come who has developed a program where you can take things like there's a standard deviation difference between this group and that group, or half a standard deviation difference. And you can convert it into the kinds of statistics that you might want to use when you're talking to people at your state or federal government level, or if you're talking to the New York Times. New York Times is not fainting when you come in and start talking about standard deviations, as you can imagine, nor is Congress.
So if anyone wants, I can get you this program. Because it's a really great way, when you're talking to people outside of academia, and they say what do you mean by a standard deviation. Or in our early intervention program, the infant health and development program, at age 18, 15 years after the program ended, there was a four point difference favoring the intervention kids compared to the control kids, which is about 30% of a standard deviation.
So people go, what does four points mean. Why should we care about four points or five points if our intervention has changed something? This is a way to translate it into terms that almost anybody could understand. So I really like this, and I think some of these are really surprising to people, that a standard deviation-- the one that gets to me the most is that white students would be 13 times more likely than minority students to score in the top 5% or enroll in a gifted class. OK, that's a huge difference, if you're talking standard deviation. And if you're talking a half a standard deviation, you're still talking five or six times. Whoops. Oh. OK.
Some of the early work we did is summarized here. And we did slightly different analyses for our Future of Children journal than I've done in my research articles, but the same arguments work. And what you want to do here is look at the standard deviation difference between blacks and Hispanics compared to whites on a math test given in kindergarten, and a reading test. This was with 20,000 kids, so this is a national sample.
On this particular test, what you're seeing is test score gaps ranging from about 40% to 70%-- not as big as that standard deviation that you get on the vocabulary tests, but getting close. I might say, for those of you who are really into data, the ECLS Early Childhood Longitudinal Study kindergarten data-- I've talked to the people who developed the test because I kept saying, this seems low. Excuse me, this actually makes things seem better than I actually think the state of the condition of black and Hispanic kids is in our country. And indeed, the guy who developed the tests actually thinks that part of this here, the 39 and 45 is due to the way the test was constructed.
When you look at these same kids in first and third grade, there's a bigger gap between whites, blacks, and Hispanics that fits more true to what I would have expected, and they're more in this range. So sometimes people say, what is that? It's so small. Is that really true? Well, according to [INAUDIBLE], it really probably has to do with the kinds of things they measured and the way ETS did the IRT. Well, and then people are now taking the data and saying that the gap increases, when I actually think what we've got is a test construction problem. You always need to talk to those test constructors. They know all these things that many of us don't know.
Now what we have over there is what the differences are between black and white, and Hispanic and white kids in the United States in terms of SES. And there are two points I want to make here. First point is what do we have in that index. We have parental education. We have income poverty. We have age of parents. And we have family structure. I had to think what we have in this particular one. OK, so we have about four things. That's a. So we don't have a lot of things in that index.
B, when Herrnstein's book came out with Murray, The Bell Curve, they had some sentences in there saying, the differences between blacks and whites in terms of social class could not possibly be large enough to account for the differences in IQ test scores. When Greg Duncan and I read that, we went, that's not true because we've done analyses and published on this. So we went back and actually took what Herrnstein and Murray said and we calculated if indeed the gap in living conditions was large enough to account for the test score gap.
And this is an example of what we found. We have data sets with better SES data that Greg and I have published on, but we like this because it's 20,000 children. And what you see is, in fact, the difference between the living conditions of black and white children, and Hispanic and white children are large enough to account for the entire gap. Herrnstein and Murray basically were saying things couldn't be that bad for black and Hispanic children in America. They're wrong. They are that bad. The disparities are huge. This is what this is telling you. Things are very similar to when Harrington wrote his book-- 60s, 70s, whenever that book came out-- in terms of the lives of black and white children.
So the SES index, basically from this analysis we did in Future of Children and the analyses that Greg and I have done over the years show at this point, I think very, very starkly-- just as a parentheses, we, Greg and I, when the bell curve came out were so outraged we wrote our stuff up supposedly in regular English and sent an op-ed to The New York Times because we thought they should publish this.
And we got a letter back, which you never get a letter when you send an op-ed to The New York Times-- they don't respond-- saying, loved it, but we've already had Myron Hofer write something on brain development, so sorry guys. But it was always so nice they at least wrote us back and thought it was important. And it was nice they did Hofer's piece. But it was a very different point than ours. Ours was much more on lives of black kids in America today. Lives aren't great, in terms of socioeconomic status.
So since we're psychologists and we're interested in poverty, but we're also interested in how poverty might translate into lower test scores, I've spent a lot of my time looking at why social class might matter. So what I do as a psychologist, as many of the people do here, is look at the ways social class may translate into something that's occurring, that could be related to lower test score gaps. And I've just listed three up here. Again, people have devoted their entire careers to looking at some of these different pathways.
The first really has to do with learning materials in the home, better childcare. We could also put there living in a neighborhood that's safe. Takes money to purchase housing in neighborhoods that are safe. This all could work also through what's called the family stress model that [INAUDIBLE] wrote about so eloquently, and this is really if you are not poor, you experience less stress, less depression and may lead to less harsh parenting. We've got three or four articles on these pathways, straight research articles, if anyone is interested in them.
Then our health folks say, gee, well we may be able to purchase better health care as well. And you guys could put up other pathways if you want to. So what we did from there in our book is we said, OK, let's look at health, parenting, and child care as they relate to racial gaps. And we asked three questions. And then we asked our authors to answer all three of those questions. And we asked authors to do this in four different chapters.
The first question is, where are their race gaps in these conditions? How much of the racial gap in test scores would be explained by each circumstance? And then what policies might reduce the racial gaps in these circumstances? And we actually asked everybody to try to estimate how much of the racial test score gap is explained by their particular circumstance in standard deviation unit. So we said to people, OK, do a back of the envelope estimate. You don't have to be fancy. You don't have to be an economist to do this. But let's translate this into something so we can then see what percentage, how much we'd reduce that gap if we put into place particular policies or changed some of these conditions in which black and Hispanic kids are more likely to live in than white kids.
So this is a summary of two chapters. I decided not to put up everything. I'm doing this fairly simply. We had one chapter on low birth weight. And here I've just put a summary for low birth weight. Are there gaps in prevalence by ethnicity? The answer is yes. Are there effects on school readiness of low birth weight? The answer is yes.
If you reduce the difference in birth weight between black and white babies-- in other words, if we were able to get the low birth weight rate for blacks down to the same rate that we have for white babies born today, how much of that black-white gap would be reduced? And what you can see is one could say gee, at least it's reducing it some. But some would say it's not reducing it a lot. It's 2% to 4%. And again, I can send you all the stuff to show you how we got to these. To go through every one of these, we'd be here all day, and we'd be unconscious, or I'd be unconscious.
Asthma-- yes, yes. Asthma looks like if we reduce the rate of asthma for black children, we wouldn't affect the readiness test scores, math and reading, but we would reduce behavior problems. ADHD-- yes, yes. This would only reduce the gap by 1% to 2%. Iron deficiency-- there are gaps. It's not clear whether iron deficiency affects school readiness. Therefore, we have to say there'd be no effect on the gap.
And then the last one we had in this particular group was lead. There are gaps still. There are huge effects on school readiness. And you would reduce the black-white test score gap by about 2% to 3%. Interestingly, this would have been much larger 25 years ago because we had many more babies, especially black babies living in inner cities, who had high rates of lead. We've actually done a pretty good job of reducing lead rates in the United States, in general. Although of course, as that reduction has happened with some of the clean air acts, we've realized that lower and lower concentrations of lead are still related to cognitive functioning. So it's kind of a-- but we're basing this not on where we would have been 20 years ago, but where we are now.
OK so this is basically a summary in about two minutes of two of the chapters in our Future of Children. I might add parenthetically, that the person that wrote on health care conditions also looked at maternal depression. I personally disagreed with her in terms of, I don't believe maternal depression is a health condition of a child, so while she has it in the chapter, I couldn't convince her to take it out. She's a colleague. She's the chair of economics at Columbia. So I can argue with her all the time, but-- we're good friends-- it does absolutely no good. She left depression in. And maternal depression differences would reduce the test score gap by quite a bit, more like 6%. But I still don't think it's a health condition.
So this could be a half empty-- depends on how you want to look at this. If you took all of these factors, you know, you'd be dealing with 10% to 15% of the racial gap. Any one in and of itself is going to account for, if we're right here and these things are causal, a small bit. But if you're focusing on health, we could do 10% to 15% of the gap.
Next chapter was on parenting behavior. I wrote this chapter. Our rule in Future of Children is the editors don't write a chapter. But the person who was going to write it thinked out. So I wrote it. But luckily, I have done something on parenting behavior so it wasn't impossible for me to do this. The first thing I did is went to the national data sets and tried to figure out what dimensions of parenting have actually been measured. And a lot of the people here who look at parent-child interactions in much more detail could think of much more nuanced ways to cut the data on parenting. These work for a variety of reasons.
We could quibble about how you want to divide up parenting. But you can see the categories that I came up with-- nurturance, discipline, teaching, language. I separated out language and teaching because there's so much literature on language, and I wanted to keep it separate. Some people might put those together.
Monitoring, keeping track of kids. We do a terrible job of measuring that, even though all parents know we spend an awful lot of our time doing that with kids. Management of the household or routines. We'll be talking about some of these issues in the chaos conference in the next three days. And materials-- what kind of stimulating materials do you have in the home. So these were the categories that we were looking at. Eleanor Maccoby may have talked some about nurturance and discipline last year when she was here. And we went again to the data that exists on large enough samples that you feel comfy that whatever you're getting is reasonable.
So this is all from pretty much large-scale studies. You find, in general, across these studies white mothers are more likely to exhibit higher rates of language teaching and the provision of stimulating materials. The effect sizes are large, 20% to 40% of a standard deviation for each of these three parenting dimensions.
Now the first thing people will say is, oh, what about social class. Well, we look at the effect sizes with and without social class. One I control for income, education, family structure, maternal age. But we reduce the sizes of the effects, but they don't disappear. They reduce by about a half. So if you control for social class, you're still getting 10% to 20% of a standard deviation difference between black and white mothers on these characteristics. It is important when you go and look at the big data sets, that you do not get differences in terms of nurturance or discipline, even though people talk a lot in our literature. I've done some of the discipline studies. Those studies are typically not nationally representative. And they're not able to control for SES.
When you put all those controls in, you don't see the big discipline differences that some people talk about between black and white families. We don't have good enough data from our national samples to look at management, unfortunately. So I can't tell you if people are managing their households differently, even though it seems in terms of black-white differences, it does look like poor families in general have much less predictability in household management and routines.
So when you look at racial and ethnic differences in parenting, if, basically, you could put in intervention programs that reduce the differences that we've found, you could reduce the racial and ethnic differences by 20% to 40%. Again, we're not controlling for SES.
We then want to preschool attendance, enrollment in preschool at ages 3 and 4, enrollment in Head Start, and the quality of programs. First, Hispanic kids are the kids who are not in preschool. I just gave you one example. And you can just look at from 2000 on. The percentages of four-year-olds in preschool for blacks and whites are pretty similar. We did some of our estimates, the back of the envelope estimates. And we tried different scenarios here. So the first is if you increase Hispanic and black enrollment to 80%, we would reduce the test score gaps by how much. For blacks, 4% to 20%, and I can go into why we have such a big range here. For Hispanics, 12% to 52%. It's larger for Hispanics because many fewer Hispanics are in preschool. We have a problem here.
Another way to think about it from a policy standpoint is what happened if we have preschool for every single kid that was below the poverty threshold. And you can see what the reductions would be there. Then we had our authors estimate, let's say that we could increase preschool so all kids at twice the poverty level or less were in preschool. And you can see fairly substantial reductions, especially for Hispanics. The reason we had our authors do number two and three is because if you think of policy, you're much more likely to be trying to target by income than by race for preschool.
We also look at Head Start in a little bit different way than most people have, and I really love how this turned out. So think, about 10% of our kids are in Head Start. Of black children, about 20% of the black children are in Head Start, 15% of Hispanic, and 4% of white. You get these differences because there are fewer poor white kids. And Head Start's for poor kids.
So we had our authors go kind of backward and say, if Head Start didn't exist, and you pulled all those kids out of preschool, out of our preschool numbers, the gaps in preschool enrollment would increase a lot. 9 percentage points for black children, and 31 percentage points for Hispanic children. We then had the back of the envelope estimates done to show that the gaps in school readiness would increase if we didn't have Head Start. It's kind of a very different way to think about Head Start than most of the analyses that people like Janet Currie and others have done.
We then turned to quality. That was just enrolling in preschool and not changing quality in America at all. And that is enrolling three and four-year-olds. Remember that we have higher quality care for three and four-year-olds than we do for one and two-year-olds. And I'm not even dealing with one and two-year-olds today. So we looked at the quality of preschool and school readiness literature, and there are several different literatures, the randomized trials Steve talked about, the cost benefit analyses that have been done by the economists, and the major studies that have been done following kids who were in high quality early childhood education through adolescence.
There are also studies that have been done. Henry's been involved in the NICHD study, which is a natural history study, where people have tried to estimate the effect of being in a better quality preschool than not which, of course, is not the same as an experiment but does give you some estimates that you can use. When we looked at this literature, clearly there is a relationship between high quality and outcomes. From my perspective, it's not clear how large that is because the natural history studies. you have terrible selection bias problems. And in a weird way, it's hard to estimate the effect sizes from the experiments for a variety of reasons.
But in any case, there is some association-- how big it is-- we can quibble. Actually, we quibble a lot. But also, when we go back to the literature on the experimental literature, it is looking from the few trials that we have where we can look at this, that the effects are larger for kids whose parents are poor or whose parents have less education. These findings are mostly from a study that I've done of 1,000 children in eight sites. Children in the treatment group got early childhood education in the second and third years of life. Those in the control group did not. And we had a range of social class.
So unlike almost all of the early childhood education literature, which focuses just on poor kids, we were actually able to show that you got no effects of our wonderful program, and it was a wonderful program, if mothers had a BA. Now this may not surprise any of us theoretically, but it is kind of nice there's one study out there that shows what we all, probably most of us, believe theoretically.
We also got much bigger effects for poorer moms than non-poor moms for the kids. And we also found in terms of cumulative risk that if you're poor, if you have lots of risk, seven or more, there was no treatment effect. But there was for poor families who had from one additional risk up to about five or six, suggesting that some of our early intervention programs may not work for kids who are particularly in disorganized households. The effects from the most recent studies seem to be larger for black and Hispanic than white children. We're finding that in Early Head Start. And I do have some thoughts about that, but this isn't the time to talk about them.
What are indicators of high quality preschool? Everyone here knows what these are. I'm just listing these here. But what I want to do is just show if you improve the quality of Head Start, the one thing I want you guys to notice is that the effect sizes are smaller than if you increase access to preschool for three and four-year-olds. And I think that's very important. And that's because the effect sizes of quality aren't as big as people often think they are. In some of our national studies just looking at preschool versus not, controlling for 200 covariance, you get an effect of going to a preschool at age 4 and age 3. And that's why you're seeing smaller effect sizes here, which doesn't fit with what a lot of people think.
OK, so what we did then is we moved to our summary. And we said, OK, what intervention strategies do we have for reducing these gaps. I'm just going to run through real quick the ones that we identified and then talk about where the evidence is strongest. And again, there's a lot more to be said in our journal. Clearly, you can try to change socioeconomic through income supplementation, parental education, and marriage promotion.
We've had some modest effects with income supplementation, which is great. Health, we could try to do something to prevent low birth weight. We can try to increase health care, both the number of kids getting it and the quality. Parenting, we have four different kinds of programs, four or five, that we've looked at-- we meaning the field of educational psychology and developmental psychology. Visiting programs, like David [INAUDIBLE] that was started here at Cornell at Elmira. Center-based programs with a parenting component. Parental language and literacy programs. And parent behavior training programs.
And I can talk about some of these later. But I'm worried about time. Preschool programs-- expand access, expand quality. Get the states to do more pre-kindergarten programs. We are serving more children in America today through the state pre-K funded programs than we are through Head Start. We could increase Head Start as well. And of course, we have child care block grants to the states for subsidies for low income parents, which allows them to purchase child care on the open market.
So from our review, what things are unlikely to reduce these gaps? The first is public health insurance coverage. I'm not saying this is not good. I'm not saying that I don't believe we ought to expand SCHIP and that it's too bad that our president vetoed it last week or the week before. But it looks like from our review of the literature, expanding public health care insurance is not going to reduce the gaps that I'm talking about. Programs for low birth weight have been a dismal failure. All of us who believed that we could change these through social support. There is no good data suggesting that we're going to reduce test score gaps that way.
And last, and this is maybe controversial to this group, is that home visiting parenting programs do not seem at this point to be likely to reduce the test score gap at ages four and five. Please see the two Cochrane collaboration studies that just came out on home visiting for parenting programs.
Slight reduction-- education programs for low education moms seem to be doing something from the experiments we have. Income supplementation, these seem to work sometimes. Income tax credits seem to work as well. As you know, that's our most expensive program for reducing poverty in America. It's what we all call the stealth program because most people didn't realize when we increased the income tax credits, how many families this would help. It's a great program.
Modest-- enrollment in health programs. This is different from the public coverage. What we found is that there are huge disparities in people who are covered, who are eligible to be covered, who are actually in programs. And so we try to make the distinction between these two. There's some really cool natural experiments done out in California that show how this works by Janet Currie. Quality of health care-- our group disagrees on this. I think the idea of quality of health care is great, but nobody knows how to measure it. WIC nutrition programs work. Data looked pretty good there.
These are three programs that show more of an effect. These would be the early childhood education programs and then the two very specific programs that are sometimes linked to center-based programs. One is reading and literacy for low literate mothers, and I'm thinking here of Grover Whitehurst's work, Russ Whitehurst's work. He has this dialogic reading curriculum that he developed. And he would use them with moms in Head Start centers, and then the teachers in the Head Start centers would use them too. And he got huge effects on kids' test scores using a very focused reading program, not these general Even Start programs, very focused.
Same is true for kids with moderate to severe behavior problems. I'm thinking of Carolyn Webster-Stratton's work here. And Cybele Raver-- was Cybele here? She was here. She was one of yours. Her work is showing the same thing using a slightly different intervention. These interventions are ones that again are linked to Head Start programs or other preschool programs, and work with teachers and the mothers, and basically are doing training of the parents, not the kids. They're showing amazing effects, and I really would like to see a lot more work on those as well.
And I'm actually going to stop there. And so I would like to end here with the list of programs that we suggested when we finished this Future of Children that we would suggest be paid attention to, if the goal is to focus on reducing this test score gap. And as you might imagine, some people were not happy with some of our findings, particularly the ones that we said were unlikely to reduce readiness gaps. And right there, I'd like to open it for questions. Thank you.
[APPLAUSE]
SPEAKER 1: This has been a presentation by Human Development Outreach and Extension at Cornell University.
Watch the full presentation with speaker footage and slides .
Jeanne Brooks-Gunn discusses the effects of income, health conditions, parenting and preschool on racial and ethnic gaps in school readiness, and summarizes evidence supporting intervention strategies most likely to reduce the gaps.
