Orwell Award Announcement SusanOhanian.Org Home

Preschool study finds bright side, dark side

Susan Notes:

Wouldn't you think a news article could provide information on how to get the study?


How Much is Too Much?

The Influence of Preschool Centers on Children's Development Nationwide

Susanna Loeb
Margaret Bridges
Daphna Bassok
Bruce Fuller
Russ Rumberger

Stanford University
University of California

Presentation at the Association for Policy
Analysis and Management
Washington, D.C. -- November 4, 2005

This paper has undergone peer review and will appear in Economics of Education Review.

How Much Is Too Much?
The Influence of Preschool Centers on Children's Development Nationwide

Stanford University and the University of California

Children from poor families have displayed solid gains in their cognitive skills when attending preschools. Evidence for these effects stems from small, controlled experiments and from non-experimental data on children enrolled in a variety of local programs. Less evidence is available on whether youngsters from middle-class homes benefit from attending early childhood programs. In addition, we know little about possible impacts from the duration (the age at which children enter preschool) or the intensity (hours enrolled each week) of attendance. These empirical questions have become more pressing as several governors and state legislatures press forward to offer publicly funded preschool to all children.

Our new analysis utilizes rich data gathered from 14,162 kindergartners and their parents and teachers by the National Center for Educational Statistics. Survey administrators assessed children to determine their early language and pre-reading skills and their understanding of numbers and mathematical concepts. Teachers gauged various forms of children's social and emotional development; and parents answered an extensive set of questions regarding the type of preschool or child care they used, the age their child entered care, and the intensity of attendance.

The scope of this information is unprecedented in terms of the number of participating children and the nationally representative character of the family sample. Never before have researchers been able to examine the effects of preschool and other forms of child care across differing socioeconomic and ethnic groups and to generalize the findings to the nation's child population.

We found that almost two-thirds (64%) of all children nationwide attended a preschool center in the year prior to kindergarten, typically at age four. When all children under five years of age are included, we estimate that about 5.2 million youngsters currently attend a center program. Disparities persist in terms of which children gain access. Hispanic children remain less likely to enroll in a center than White and African American children (Figure A). Clear progress has occurred since the 1960s to raise the enrollment rates of Black children.

The effect of center attendance on children's cognitive skills. Accurately estimating the effect of children's exposure to preschool centers requires taking into account home factors, such as income and parenting practices, which may predict both whether a child enters a center and the child's pace of early learning. Without controlling for these selection factors, we would risk inaccurately specifying how center attendance contributes to child development.

We statistically adjusted for 32 different features of the child's home and family. We also statistically controlled for 13 characteristics of each family's surrounding community. We employed statistical techniques, including propensity score matching, to verify the robust character of our estimates.

We find that attending a preschool center prior to kindergarten raises early language and pre-reading skills and math skills by about 10 percent of a standard deviation (SD) on average (Figure B2). The magnitude of this benefit is more than double for English-proficient Hispanic children (0.23 SD), compared with White children. To set the magnitude of these estimates in context, we can compare them to the magnitude of other interventions. For example, the benefit accruing to children randomly assigned to Tennessee's smaller kindergarten classes of less than 17 students equaled about 0.21 SD, compared to the control group which remained in larger classes.

Comparing the effects of preschools for children from families with differing incomes, we find that children from extremely poor families display the strongest gains in pre-reading and math skills after attending a center program (Figure C2). Specifically, those children from the nation's poorest families show a 0.20 SD gain in pre-reading and a 0.22 SD advantage in math concepts, compared with their counterparts who remain at home with a parent. This translates to an 8 and 9 percentile point gain on a standardized test, respectively, equaling the magnitude of effects observed in sizeable early childhood programs (such as, Chicago's Child-Parent Centers).

We did not observe statistically significant gains in language and pre-reading skills for other children from lower-income families (the complete bottom quartile of family income), compared with their counterparts who remain at home with a parent, although we did see a significant gain in math proficiency.

Children from middle- and upper-income families experience modest gains in pre-reading and math skills, stemming from preschool attendance, compared with counterparts who remain at home. For example, the estimates for children from middle-class homes equal 0.13 SD in pre-reading and 0.12 SD in math concepts. This is good news for middle-class parents, many of whom already enroll their three or four year-old in a preschool program of some kind. Yet since middle-class children experience a positive bump from preschool, early learning gaps between children from low income and middle income families are not likely to narrow with universal preschool. The exception to this is for children from the very poorest homes, those who appear to gain the most from preschool attendance.

The effect of center attendance on children's social development. We find that attendance in preschool centers, even for short periods of time each week, hinders the rate at which young children develop social skills and display the motivation to engage classroom tasks, as reported by their kindergarten teachers. We use a composite measure of social-behavioral growth which includes indicators rooted in three domains of development: children's externalizing behaviors (such as, aggression, bullying, acting up), interpersonal skills (such as, sharing and cooperation), and self control in engaging classroom tasks.

This slowing of typical rates of social-emotional growth is particularly strong for Black children and for children from the poorest families. For example, Black children who attend a center program scored 0.26 SD lower, compared with Black children who remain at home cared for by a parent (Figure B2). In contrast, Hispanic children show no signs of lagging social development when they attend center programs, compared with their counterparts who remain at home with a parent.

Our findings are consistent with the negative effect of non-parental care on the single dimension of social development first detected by the NICHD research team. That study compared attendance in any type of child care to remaining in parental care. Our study contrasts children's exposure to preschool centers, specifically, to remaining in parental care. And while the NICHD study looked at a single dimension of behavior, we find that the decrement to social development is not limited to one domain, externalizing or aggressive behavior. Our analysis reveals that this drag on the typical pace of social development is moderate in magnitude and extends across all three forms of social-emotional growth.

Does earlier enrollment in a preschool center advance development? Cognitive benefits in both pre-reading and math are generally stronger when children first enter a center program between two and three years of age. The overall benefit for this group is greater than the gains observed for children who enter before age 2 or for those who enter after 3 years of age.

The results for social development are different. On average, the earlier a child enters a preschool center, the slower their pace of social development. Children who enter a center before age 1, for example, display a marked lag in social development of 0.29 SD. This level of magnitude is somewhat smaller than the effect of growing up in a home with a depressed parent, a detrimental effect that ranges between 0.35 to 0.70 SD, according to a pair of recent national studies.

This pattern -- strongest gains in cognitive growth for children entering between 2 and 3 years of age, yet negative social-behavioral effects from such early entry -- are observed consistently across ethnic and family-income groups.

How much is too much? Among all children who attend a center program, just over one-quarter attend for more than 30 hours per week the year before kindergarten. This translates to more 1.3 million children nationwide. Our analysis examined whether this type of intense exposure to centers each week affects children's cognitive and social development. Researchers, until recently, have been unable to test for such effects with a nationally representative sample of children.

We find quite differing effects from center attendance depending on how many hours per week children are attending a center. In addition, these effects differ in important ways by the income level of the child's family.

On average, children who attend centers for 15 to 30 hours per week experience stronger cognitive gains but weaker social development, compared with those who attend for less than 15 hours per week (Figure D2). Attendance for more than 30 hours weekly does not appear, on average, to yield additional cognitive benefits for children but does further suppress typical rates of social development. For example, children attending 15 to 30 hours per week and those attending more than 30 hours score approximately 8 percent of a standard deviation (0.08 SD) higher on pre-reading skills than those who attend centers for fewer hours. On the other hand, while children attending 15 to 30 hours per week score 10 percent of a standard deviation (0.10 SD) lower on the behavioral index than those attending for fewer hours, children in centers for more than 30 hours weekly score, on average, fully 25 percent of a standard deviation (0.25 SD) lower.

These average differences mask important variation among children from different family-income groups. For children from lower-income homes, additional hours in center programs does not further slow social development while it does advance cognitive gains (Figure E2). But for children from higher-income families, additional hours in care do further slow behavioral growth, while at the same time failing to improve cognitive outcomes relative to moderate attendance of 15 to 30 hours per week. The benefits of longer hours of care are clearly greater for children from lower-income families.

Policy implications. Several states are now investing heavily in universal preschool efforts, including California, Florida, Georgia, New Jersey, New York, Oklahoma, and North Carolina. Our findings suggest that as policy makers move to offer all families access to free preschool, children from middle-class families will likely experience modest benefits in terms of cognitive growth, even at current levels of uneven program quality. Yet this also means that disparities in early learning and social development among social-class or ethnic groups will not necessarily narrow as a result of universal preschool, since benefits accrue to children across the range of family income. Children from the very poorest families may catch-up some if resources are focused on their communities.

These findings also inform the debate over half-day versus full-day preschool programs. Our results suggest that full-day programs may be a wise investment for children from poor families who gain cognitively from more intensive preschool but do not show strongly negative behavioral consequences associated with additional hours. Half-day programs may be sufficient for children from middle or higher-income families, given that for these children the cognitive benefits taper-off after 30 hours per week of exposure, and the negative social-developmental effects intensify .

How to remedy this detrimental effect on social development stemming from exposure to preschool centers is a question that researchers and policy makers might tackle more urgently. Whether this slowing of emotional and behavioral development is endemic to larger group settings, or can be remedied through quality improvements, remains a pivotal question. As the recent Yale University study of preschool expulsion practices showed, many programs are not well equipped to serve young children with significant behavioral problems.

by Carrie Sturrock, in San Francisco Chronicle

As taxpayers, parents and educators debate the value of public preschool for every child, a new study by UC Berkeley and Stanford researchers finds for the first time that middle-class children -- not just kids from the poorest families -- receive a boost in language and math skills from preschool.

But its darker findings bolster earlier, more controversial conclusions that preschool can hinder social development.

The study, 'How much is too much? The Influence of Preschool Centers on Children's Development Nationwide,' was released today and comes as Hollywood movie director Rob Reiner leads a group of universal preschool advocates pushing for a June 2006 ballot measure that would tax the wealthiest Californians to fund preschool for all who want it. The study, with its good and bad news, is likely to fuel arguments on both sides of the preschool debate. Universal preschool advocates can seize on the findings that preschool benefits most children in language and math. Those who think scarce preschool resources should continue to go to the poorest children can point to the negative effects on social development, especially for children from the wealthiest families. The study looked not only at aggressive behaviors but also at a child's ability to cooperate and negotiate tasks in a classroom.

'If preschool is expanded, more isn't necessarily better,' said UC Berkeley child development research director Margaret Bridges, an author of the study who expressed concern about the negative effects on social development. 'Cognitive benefits are great, but we have to pay heed to what's going on with kids emotionally and socially.'

Advocates of universal preschool argue that the study didn't examine the quality of individual programs and, therefore, can't make definitive conclusions about the effects of preschool.

'Absent any information about what kind of quality the kids in this study were getting, the findings are meaningless,' said Susanna Cooper, director of communications for Preschool California, a nonprofit advocacy group that supports the ballot initiative. 'There may very well be these social (problems), but I don't know what they're from. Maybe these kids were in mediocre low-quality childcare situations. Were they overcrowded? What was the ratio of kids to adults? What was the training of the adults they were with?'

The researchers respond to such criticism by saying that the study is based on an unprecedented set of data that gives a snapshot of preschool's impact on the average child across the socio-economic spectrum. While the study didn't examine the quality of individual programs, it gives a pretty clear picture of how children are affected by the programs available, said Bruce Fuller, a study co-author and associate professor of education at UC Berkeley.

Researchers looked at information on about 14,162 kindergartners, collected through face-to-face assessments with the children as well as interviews with their teachers and written surveys from parents. The data were gathered between 1999 and 2004 by the federal National Center for Education Statistics as the debate about universal preschool took off at the national level. The children had their math and language skills assessed; kindergarten teachers gauged the children's social and emotional development; parents answered questions about the type of preschool or childcare center used.

The study did not explore the quality of individual programs and didn't distinguish between structured preschool programs and more basic childcare centers.

It found that nearly two-thirds of all children nationwide attended preschool the year before kindergarten. The researchers found evidence supporting past studies that preschool has the greatest cognitive impact on the poorest children -- those whose families make $16,000 or less. Those children exhibited language and math skills that on average were 8 and 9 percentage points higher, respectively, than their stay-at-home peers. This sort of finding isn't new and has fueled support for the federal Head Start program for poor children.

Children from lower-income families -- an economic notch above poor families -- didn't see a statistically significant improvement in language but performed an average of 6 percentage points better in math than peers who didn't attend preschool.

What struck researchers was this: Middle-income children did 5 percentage points better in both language and math than those in that income bracket who stayed at home. And children from the highest income quartile -- those whose families made $66,000 or more -- also saw improvements, although small ones: They performed 3 percentage points above average in language. Their gains in math weren't statistically significant.

These findings -- while positive overall -- don't convince Fuller that universal preschool is the way to use scarce resources. If preschool gives everyone a leg up in either language or math, then a universal program wouldn't close the achievement gap between children from low-income and higher-income families, Fuller argues.

'Middle-class families are benefiting, but if we move toward universal preschool, it's not clear that universal preschool would close gaps in early learning because the gain experienced by low-income kids may not ever be enough to catch up with the gain by middle-class kids,' Fuller said.

Advocates of universal preschool say that's not the way to look at it.

'We want all children to be better prepared and be successful for life,' said Catherine Atkin, president of Preschool California. 'All children benefit from a quality preschool experience.'

Efforts to expand preschool are happening across the country, including in Georgia, Oklahoma, New Jersey and New York. California is one of three states -- including Massachusetts and Illinois -- where significant plans are being made to offer all 4-year-olds voluntary preschool.

In California, the proposed universal preschool ballot initiative would levy a 1.7 percent tax on couples who make at least $800,000 annually or single people who make $400,000.

The UC Berkeley-Stanford study found that all children who attended preschool at least 15 hours a week displayed more negative social behaviors such as trouble cooperating or acting up, when compared with their peers. The discrepancies were most pronounced among children from higher-income families.

Children from lower-income families lagged behind their peers who didn't attend preschool an average of 7 percentage points on the measure of social behavioral growth. But children from higher-income families lagged 9 percentage points behind their peers. These wealthier children did even worse when they attended preschool for 30 hours or more: They trailed their peers by 15 percentage points.

It's not clear why children from higher-income families exhibit more negative behaviors than their stay-at-home peers. Fuller speculated their peers might be in enriching home environments that include things like trips to the library as well as dance and music lessons. Other studies have found childcare centers negatively affect children's social development, said Jay Belsky, director of the Institute for the Study of Children, Families and Social Issues at Birkbeck University of London, in an e-mail interview.

'It is time to come to grips with what all too many have denied for all too long, namely, that all disconcerting news about adverse effects cannot be attributed to low-quality care, which has been more or less the mantra of the field of child development and the child-care advocacy community for decades,' Belsky said.

But advocates of universal preschool insist that quality is key. The Preschool for All initiative in California calls for offering every 4-year-old access to half-day preschool three hours a day, five days a week from teachers holding bachelor's degrees.

'You have to look at all the research together and draw from it the lessons that have been clear for quite a while -- that the quality of the program makes an enormous difference, and it makes a difference in terms of the social and emotional development of young children,' said Alan Simpson, spokesman for the National Association for the Education of Young Children in Washington, D.C., which accredits preschool programs nationwide. 'A quality program has teachers and staff that know how to nurture children's social and emotional development.'

E-mail Carrie Sturrock at csturrock@sfchronicle.com.

— Carrie Sturrock
San Francisco Chronicle


This site contains copyrighted material the use of which has not always been specifically authorized by the copyright owner. We are making such material available in our efforts to advance understanding of education issues vital to a democracy. We believe this constitutes a 'fair use' of any such copyrighted material as provided for in section 107 of the US Copyright Law. In accordance with Title 17 U.S.C. Section 107, the material on this site is distributed without profit to those who have expressed a prior interest in receiving the included information for research and educational purposes. For more information click here. If you wish to use copyrighted material from this site for purposes of your own that go beyond 'fair use', you must obtain permission from the copyright owner.