Issue Position: Early Childhood Education

Issue Position

Early Childhood Education

Step 1 -- State the Evidence
(Sources for most of the data in this article are:
J.Mervis. Science, Vol. 333 pp. 952-956 and J.Mervis. Science, Vol. 333 pp. 956-957.)

The most frequently cited evidence for the effectiveness of early childhood education comes from three rather old studies. Many other projects have come and gone, but few of them have been well-studied with adequate control groups.

1. From 1962 to 1967, the HighScope/Perry Preschool Program enrolled impoverished children with low IQ scores, aged three to four. A random selection process placed 58 children in the program and 65 children in a stay-at-home control group. The half-day program hired certified teachers with advanced degrees at a cost of $17,600 per child. Children in the program entered kindergarten with better scores on school readiness tests than control children. However, by grade four the difference in test scores had almost disappeared. What surprised researchers were the results of long-term follow-up. As teenagers and young adults, students in the program had higher test scores, required less special education, did more homework, placed a higher value on learning, spent less time unemployed, and had fewer arrests. The resulting saving to taxpayers is estimated at $284,100 per child (16 times the cost). The delay in the appearance of benefits from the program is commonly attributed to the child having acquired better motivation and persistence which become more important in middle school.

2. The Abecedarian (pronounced a-b-c-darian) Study also involved a small group of students - just 57 in the treatment group. Beginning in 1972, it was more ambitious than the HighScope/Perry Program. Starting with children as young as six months, it was a full day program that lasted year-around through kindergarten. It showed somewhat better and more lasting improvements in student test scores, as well as most of the other benefits of the HighScope/Perry Program. However, it did not produce evidence for decreased delinquency. The cost per student was $70,700, and the estimated saving to taxpayers was only $176,300 (2.5 times the cost). The difference in savings to taxpayers between this and the HighScope/Perry Program was chiefly due to the failure of the program to demonstrate fewer arrests.

3. The Chicago Longitudinal Study was a much larger program with 1539 at-risk students. It was begun in 1985, a time when the benefits of early childhood education were widely accepted, and it was unlikely that parents would accept having their children in a stay-at-home control group. Accordingly, the "control group" was a matched school in which some students were already participating in Head Start or similar programs. It was a half-day program that served students aged three or four, with some students being kept in the program until age six. The program employed fewer teachers, but the teachers were professionally certified. Improved test scores of the treatment group were sustained through grade school. The treatment group was half as likely to need special education, had one-third fewer arrests, and had half the rate of substance abuse. The cost of the program was $8,200 per student and the savings to taxpayers (by age 27) was $83,500 (10 times the cost).

These are generally described as "high quality" programs as contrasted with Head Start. Head Start is the relatively low-cost, federal government program started in 1965. Few Head Start students get more than a year of treatment, the program uses few professional teachers, and teachers' pay is much lower than that of public school teachers.

Head Start is impossible to evaluate for two reasons. First, it is now impossible to find a control group. There are simply too many alternative state programs that impact students in similar ways. Second, Head Start is, itself, a moving target. A congressionally mandated study of Head Start in 2010 found that Head Start students do perform better on school readiness tests. However, by the end of first grade, no benefit is observable. It is still possible that Head Start students will show the same delayed benefits as were seen in the HighScope/Perry Program. Moreover, as poor reports of Head Start emerge, Head Start, itself, undergoes changes. The most recent change is a requirement by the Obama Administration that half of Head Start teachers be professionally certified.

Step 2 - Evaluate the Evidence

Early childhood education programs are highly variable in their academic emphasis, their involvement of parents, their concern with nutrition, the kind of schools they feed students into, the ethnic backgrounds of their students, and so forth. Accordingly, it is difficult to determine what works in early childhood education. Still, there is evidence that something of value is achievable.

Step 3 -- List the Options

There are now a large number of state and federal programs that impact early childhood. In this section I will focus only on the largest, Head Start. The options would seem to be as follows.

1. We could regard the failure of Head Start to provide evidence of its effectiveness as evidence of failure to achieve its goals. We could simply abandon Head Start.

2. We could continue Head Start with little or no change in funding in the expectation that hiring fewer, more highly paid teachers will introduce techniques into Head Start needed to make it effective.

3. We could increase the budget of Head Start with the same hopes as in choice 2.

4. Had the United States begun a serious study of early childhood education programs at any time prior to 1990, we would now have evidence regarding what works. Given a budget for Head Start of $7.56 billion per year, it seems reasonable to start such a program even now. $1 billion would fund 50 projects the size of the Chicago Longitudinal Study (the most expensive of the three discussed above). That is more than enough to investigate the numerous program variables. In 20 years, we would know what to do. Maybe. There is one strong reservation. In the process of scaling a small program up, important details may be lost. It is important to demonstrate the scalability of a program before its general adoption.

Step 4 - Prioritize

If we do not know what the payback is, we might well be cautious with our investment. However, when we know that we can get a payback of at least ten to one just by following the Chicago Longitudinal Study model, it seems foolish in the extreme not to invest whatever it takes in early childhood education to get that kind of payback. (This, of course, is just the financial payback seen from the point of view of taxpayers.) Seen in this way, choice 4 seems to be a high enough priority so that we should begin paying for it immediately.

Source