Brookings: Social Mobility


Early Childhood Achievement Gaps and Social Mobility

By Richard V. Reeves, September 24 2013.

Part 1

Policy-makers have developed an obsession: early childhood. Obama is struggling to make pre-K education accessible to all. Governors around the country—Democrat and Republican alike—are investing in the 3- and 4-year-olds in their state. Despite some disappointing evaluations of the Head Start program, significant political and financial capital is being invested in preschool.

Why? Because of the mountain of evidence that gaps in ability open up long before kindergarten. That’s true for cognitive skills, like math and reading, and for non-cognitive skills, like application and self-control.

The link between income and the acquisition of cognitive skills—for which the best data exists—is strong, consistent, and clear. The graph below shows what percentage of children by family income are top performers on cognitive ability tests before starting school, when starting high school, and during adolescence. The three tests are the Peabody Picture Vocabulary Test (PPVT), the Peabody Individual Achievement Test (PIAT), and the Armed Forces Qualification Test (AFQT).

As you might expect, children from affluent homes—measured by income quintile (equal fifths of the distribution)—are much more likely to be scoring in the top third of the ability range.

graph of top third of test scores, by family income

Two points:

  1. The gaps by income are big. Half of the children from the most affluent homes are in the top third of test scores, compared to just one in seven of those from the poorest families.
  2. More surprisingly, perhaps, there is essentially no change in the relationship between income and test scores at the different ages. The shape of the bar chart is in fact spookily similar for the three tests. (The adolescent scores are from a different survey to the early two scores: but if anything, this makes the similarity of the distributions even more startling.)

Children from more affluent backgrounds are the A-grade students before school even starts, and then simply maintain their position through the school years. Addressing the divide as early as possible looks necessary, and urgent.


Part 2

Deep divisions between rich and poor emerge in early childhood, with lifetime implications for social mobility. Yesterday, I showed the way affluent children dominate the top of the test score distribution, from pre-school onwards. The story for those scoring poorly on tests is, as you’d expect, a mirror image:

Top Third of Test Scores, by Family Income

There is a particularly sharp break towards the bottom of the income distribution: six out of ten of those from the bottom quintile are in the bottom third of test scores, compared to around four in ten of those from the next quintile up.

So what? Well, for one thing it looks as though the administration’s focus on the early years is fully justified. The gaps in cognitive ability between rich and poor that have opened up by the age of 3 or 4 are apparently replicated, almost automatically, up through adolescence. Closing the gap early is the right priority, especially between the poorest and the rest.

Four caveats:

  1. The administration’s pre-k proposals are aimed at 4-year-olds, but by this time the gaps have already emerged. Tamara Halle and her colleagues at Child Trends find that ability gaps by income actually widen quite dramatically between the ages of 9 months and 2 years, when pre-k is still a distant prospect.
  2. The push on pre-k education should be heavily supplemented by programs to support parents. Parents are the ultimate pre-k educators, and many need help to do a better job, not just in the early years, but in the early months. Parenting should rank alongside pre-k education as a priority for policy-makers, as Kimberly Howard and I argue in our recent paper The Parenting Gap.
  3. The effectiveness of pre-k programs cannot simply be assumed. Some work much better than others. Ongoing, rigorous evaluation is required. The administration is right to be focused on quality assessment.
  4. The biggest beneficiaries of early childhood interventions—whether in the form of pre-k education or parenting support, or both—are from low-income families. Given our fiscal constraints, the targeting of programs on those most likely to benefit is vital. That’s the subject of tomorrow’s post.


Part 3

Achievement gaps open up in early childhood, damaging chances of upward mobility—especially for those from poor backgrounds. The question is: what can we do about it?

The leading policy to address these early childhood gaps is pre-k, for which support is high but evidence is mixed. We know that a couple of very intensive early years programs, namely the Perry Preschool Project and the Abecedarian Project, showed big payoffs but also that it is not often appropriate to generalize from these programs. We also know that Head Start, our current federal early childhood program for low-income children, has no apparent lasting impacts, at least in terms of academics. In short, policy focused on early childhood education has promise but federal investment so far has had disappointing results.

President Obama has now proposed universal pre-k for 4-year-olds. Is this the solution for closing early achievement gaps and improving social mobility?

In a paper presented at BPEA last week, Cascio and Schanzenbach provide new evidence on the potential effects of Obama’s proposal. They found universal pre-k programs in Georgia and Oklahoma made it much more likely that disadvantaged children were enrolled in pre-k, but there was also a big shift of more advantaged children from private pre-k to public, meaning higher costs for taxpayers. There is modest evidence that these programs improved test scores in later years, but only for the disadvantaged children. The implication is that Obama’s proposed program could create some cognitive advantages for low-income children—the principal beneficiaries—but may not be the most cost-effective approach.

Pre-K enrollment by Maternal Education

Intriguingly, Cascio and Schanzenbach also found that with universal pre-k in place, less-educated mothers spent less time with their 4-year-olds overall but spent better quality time with their children when they were together: specifically, 25 more minutes per weekday doing activities with them such as reading, playing, and talking. In other words, pre-k may not simply substitute for parental investment in their children but may also improve it.

This is good news. We recently argued in “The Parenting Gap” that more investment should be focused on improving the home environment of disadvantaged children. We concentrated on direct parenting interventions like home visiting programs, which are a small part of Obama’s proposal. Cascio and Schanzenbach’s work suggests that in fact pre-k might improve parenting, too, albeit indirectly.

High-quality pre-k and high-quality home visiting programs can both help to close early childhood gaps in parenting and in the development of cognitive skills. But quality is king.

See Richard V. Reeves, Early Childhood Achievement Gaps and Social Mobility, Brookings, September 24 2013.

(Emphasis added)