We have enough evidence to expand quality pre-K

Professor Dale Farran of Vanderbilt University has a new policy brief at the Brookings Institution website, entitled “We need more evidence in order to create effective pre-K programs”. This policy brief makes the skeptical researchers’ case for collecting more research evidence on how to create effective pre-K programs. As a social science researcher, I hardly can quarrel with the need for improved research knowledge. However, I believe the policy conclusion that many will draw from this policy brief – that we should wait for more research before undertaking significant expansion of pre-K programs – is mistaken.

Professor Farran’s case is that we do not know as much as we would like about what really constitutes quality in pre-K programs, how to measure that quality, and how to improve quality. For example, we are not sure which skills increases in the Perry Preschool program were the most responsible for its large long-term benefits. We think that soft skills are in part responsible for the long-term benefits of Perry Preschool, and the long-term benefits found in other pre-K programs. But we’re not sure how best to measure soft skills. And the existing measures of classroom quality, including classroom observation measures, do not always seem consistently related to the skill gains in pre-K we can measure. Professor Farran then adds that the strong benefits for the Perry Preschool Program do not appear to be matched in some Head Start research studies, or in the Tennessee Pre-K study that she has helped direct.

From this, many readers might reasonably infer that we should wait until we have more research findings before undertaking any large-scale pre-K expansion. I don’t know if that is actually Professor Farran’s policy stance. She does say the following:

“The approach the field should take… is to begin a rigorous research effort to determine which malleable competencies in early childhood are most related to the developmental trajectories of poor children, which experiences within pre-K settings actually facilitate the development of those skills, and how the success of pre-K programs in transmitting those skills can be validly measured for purposes of accountability and improvement.”

This statement will not be interpreted by most readers as an endorsement of pre-K expansion.

Despite Professor Farran’s arguments, why do I think the case is still strong for pre-K expansion? First, long-term benefits of high-quality pre-K are attested to by a number of good studies. These good studies include not only the Perry Preschool and Abecedarian studies mentioned by Professor Farran, but also good studies of the Chicago Child-Parent Center program and the Head Start program. For example, the numerous studies by Arthur Reynolds and his colleagues of the Chicago Child-Parent Center program suggest that its pre-K program increases long-run earnings by 8%. Deming’s study of Head Start, which compares siblings who differed in whether they participated in Head Start, suggests long-term earnings effects of 11%.

Second, we have good evidence that high-quality pre-K can increase short-term test scores, and that early educational experiences that increase short-run test scores will predict better adult outcomes.  These studies of pre-K’s test scores effects include studies that show benefits of large-scale pre-K programs run by urban public school districts, for example there are good studies of both Tulsa and Boston.  Furthermore, there are good studies that suggest that early educational experiences that increase test scores in the short-term would be expected to increase adult earnings. (In contrast to what Professor Farran says, these studies have provided evidence that these effects of test score gains are due to causal effects of the early education interventions, and are not due to unobserved child characteristics.) In fact, these studies suggest that early test score gains of pre-K are a better predictor of long-term adult benefits than the medium-run test score effects in elementary and middle school.  The fading of pre-K test score effects by third grade found in Tennessee and in the Head Start experiment occurs almost universally, and in particular occurs in programs such as Perry Preschool that show very large adult benefits.  (See my review of this evidence in pp. 31-34 of my 2014 book, From Preschool to Prosperity.)

Why this short-term fading followed by long-term recovery of pre-K’s effects? The most plausible hypothesis has been most prominently identified by Nobel-prize-winning economist James Heckman, who argues that many of pre-K’s long-term benefits are mediated by improvements in so-called “soft skills”, which include social skills, planning skills, and self-confidence skills.

Now, as Professor Farran argues, we don’t have great measures of soft skills, or rigorous evidence on how best to improve soft skills. But this brings me to my third point: we don’t need to have rigorous evidence on how soft skills affect adult outcomes, or on how pre-K affects soft skills, to justify expanding pre-K.  We know that programs like Perry, the Chicago Child Parent Center program, Tulsa pre-K, and Boston pre-K can make a difference. We know that these programs employ certified teachers with a background in early childhood education, and that these teachers are paid sufficiently well to attract and retain teachers. We may not know all the intervening causal mechanisms that go between these program characteristics and strong outcomes.  But we don’t need to know all these intervening mechanisms for the intervention to make sense.

As I argued in a previous blog post, the best way to accommodate Professor Farran’s concerns that we don’t know enough about what quality pre-K is and how to create it is to do the following:

  1. Expand pre-K;
  2.  Fund it sufficiently well to over-invest in quality, for example by making sure the teacher salaries, qualifications and training are more than adequate;
  3.  As part of this expansion, experiment with different approaches to implementing quality pre-K, to allow for us to learn more.

The problem with delaying pre-K expansion until social science research has reached some stage of perfection is that this has a potential big cost: the loss of human potential by NOT providing quality pre-K to a cohort of young children. Policymakers should side with the preponderance of evidence that supports the high benefit/cost ratio for high-quality pre-K. (For example, see the research review by a group of early childhood researchers, and the research review by the Washington State Institute for Public Policy.)  In my view, this evidence makes the risks of not expanding quality pre-K greater than the possible risks of expanding pre-K that turns out not to be high-quality. And the latter risk can be minimized by over-investing in quality when we expand pre-K, and by testing out the relative effectiveness of different pre-K models.

The truth is that we will never get every social scientist to be fully satisfied with the research evidence on any public policy item. I sometimes wonder how Horace Mann in the 19th century would have fared if there had been social science researchers demanding  many rigorous studies showing that free “Common Schools” through 8th grade would benefit the public. Maybe we wouldn’t have free public schools today.

Having said that, in the case of pre-K, we probably have as rigorous research evidence for effectiveness as we are likely to have for most realistic public policy interventions to influence someone’s life course. We know that pre-K can work. While I agree with Professor Farran that we need to know more about the best ways to improve pre-K quality, we can gain such knowledge while still expanding pre-K programs so that we can benefit’s today’s children.


About timbartik

Tim Bartik is a senior economist at the Upjohn Institute for Employment Research, a non-profit and non-partisan research organization in Kalamazoo, Michigan. His research specializes in state and local economic development policies and local labor markets.
This entry was posted in Early childhood program design issues, Early childhood programs, Uncategorized. Bookmark the permalink.