Judith Scott-Clayton is an assistant professor at Teachers College, Columbia University.
A report released last week has drawn new attention to low degree-completion rates among college entrants, particularly among those who never attend full time. The organization that published the report, Complete College America, seeks to rally policy makers around the goal of substantially increasing completion rates by 2020.
Today’s Economist
Perspectives from expert contributors.
But here’s a question: if we want to increase college completions over the longer term, is it more cost-effective to direct resources to college students or to preschoolers and kindergarteners?
The influential economist and Nobel laureate James Heckman, among others, has asserted that early educational investments have the highest return, because of the cumulative nature of skill development (“skill begets skill”). By the time a high school student is on the verge of college (or an older worker is considering returning to school), this argument goes, it may be too expensive to try to fix skill deficiencies that trace back decades.
Yet the federal government continues to invest far more in higher education than in early childhood programs (see chart). For example, the Pell Grant program received more than $28 billion in 2009, compared with less than $10 billion for Head Start.
College Board, Department of Health and Human Services, Joint Committee on Taxation.
A new study presented last week at a Federal Reserve Bank conference on education and employment adds to the growing body of evidence that early educational experiences can affect outcomes in college and beyond. (Note: Susan Dynarski, one of the study’s co-authors, was my doctoral adviser.) The study found that kindergarteners randomly assigned to smaller classes through the Project Star experiment were 2.7 percentage points more likely to enroll in college and 1.6 percentage points more likely to complete a degree by age 30.
While the magnitude of these effects may sound modest, they are statistically and substantively significant compared with the control group’s attendance rate (38 percent) and degree completion rate (15 percent). And the effects were much larger among minorities and at-risk students.
Previous experimental and quasi-experimental studies have found large effects of both the Abecedarian and Perry Preschool programs and the Head Start program on college enrollments. A consistent and fascinating pattern in all of these studies is that early effects on test scores tend to fade over time, only to re-emerge as effects on college-going and other adult outcomes.
When it comes to cost-effectiveness, however, the new study challenges the theory that early investments are always better. The 33 percent reduction in class size evaluated by the Project Star experiment was not cheap, and the authors estimate that it cost more than $400,000 per additional student induced to attend college as a result of program participation (calculated by dividing the $12,000 per-student cost of the program by the estimated college impact of 0.027).
Using impact estimates from other studies, the researchers calculate that programs aimed at individuals on the brink of college have the biggest bang for the buck — if the goal is college attainment.
For example, a recent experiment that randomly selected people to receive assistance with completing and submitting a federal financial aid application found that this assistance, which cost only $88 per participant, increased college enrollments by seven percentage points.
An enormous caveat to this analysis is that good early childhood programs have been shown to improve a host of outcomes even for those who never attend college, including childhood health and mortality. Moreover, scholarship programs and financial aid application assistance may work only to the extent that there is a pool of people who can benefit from college but aren’t already enrolled — so their current cost-benefit ratios may not be infinitely scalable.
Finally, some interventions at the college level may increase attainment simply by handing out more degrees rather than by fundamentally increasing skills (an extreme example of this is a new effort to locate and award degrees to former students who earned enough credits but never asked for their diplomas).
Based on both theory and evidence, it is hard to argue that we’re not underinvesting in early childhood education. Even policy makers who care only about college completion rates should be looking for promising interventions in the earliest years of life.
But that’s no reason to write off later interventions. There appear to be opportunities for cost-effective investment all along the educational continuum.
Article source: http://feeds.nytimes.com/click.phdo?i=04722eda2f8195470050ea2eda520492