John Brandl: How to give kids a proven head start

Here's heartening news: Researchers looking at ways to address the achievement gap between black and white students have found programs that, dollar for dollar, yield far more benefit than they cost.

On Monday the local think tank Growth & Justice brought together some of the country's leading experts on the topic. These scholars have been measuring the dollar costs and dollar benefits of programs in the public schools.

University of Minnesota Prof. Arthur Reynolds reported his findings that early childhood programs for very needy children can have a beneficial influence that lasts into adulthood. That's not as obvious as it sounds. In the past 40 years, $100 billion has been spent on Head Start, America's largest early childhood program. A recent nationwide evaluation, though mostly upbeat about Head Start's contributions to the lives of many children, concluded "positive impacts were not found" among children of unemployed, high school dropout, single teen mothers on public assistance. That is, Head Start hasn't much helped those who need it most.

Reynolds and his colleagues studied a program, Child-Parent Centers, which was much more ambitious than Head Start, enrolling children from the poorest neighborhoods in Chicago for as long as six years. Services were provided to their families as well. The researchers located the enrollees 20 years later and noted the differences in their lives as against people from similar family backgrounds who had not been enrolled. They discovered that the people who had been enrolled had completed more schooling, had been in less trouble with the law, had attained higher levels of employment and rated better on a measure of mental health.

Those findings were bolstered by the Minneapolis Federal Reserve's Arthur Rolnick and Rob Grunewald. Their research found that some early childhood programs have had little effect. But they also found that some larger, more expensive programs, incorporating parental involvement and carrying the children for many years, are not only effective but yield dollar benefits that exceed costs.

Henry Levin of Columbia University and Clive Belfield of City University of New York also described programs whose dollar benefits they calculated as exceeding dollar costs.

The most-promising programs reported on by these and other scholars potentially yield spectacular returns, much greater than most private-sector investments. Early childhood programs extending into elementary school can return up to $10 in benefits per dollar of cost. Reducing class size for low-income children results: benefits four times the costs. Creating small schools with a rigorous curriculum and mentoring, with teachers who commit long-term: benefits seven times the costs.

So it's a no-brainer -- we should fund all those things tomorrow, right? Hold your horses.

Most of the scholars neglected to show why the success of their programs should be expected on a larger scale. Here's a cautionary tale.

A careful evaluation in Tennessee found that children in classes of 15 did better than those in classes of 22. Duh, you say. Well, California, influenced by the Tennessee experiment, set out to reduce class size. It hired 18,000 new teachers in the first year and spent billions. After a few years, evaluators found that reducing class size in California had little or no effect. Replication of successful government programs is not automatic.

Rolnick and Grunewald took up this challenge and concluded that designing markets for services that have been found to be successful is the answer. In the case of early childhood education, they proposed scholarships for children in low-income households. Parents could shop for the program that appealed most to them. The scholarships could be used only at programs that had the qualities shown to yield positive results: "well-trained teachers, relatively low ratios of children to teachers and research-backed curricula." The idea will be tried here in Minnesota starting next year, privately funded, with the clear hope that government money will follow.

We have long known that educational achievement is influenced mostly by characteristics of a student's family and only secondarily by school characteristics. This suggests that the hunt for ways of closing the gap can be understood as a hunt for supportive cultures in which to immerse students who do not receive adequate support at home. Except for Rolnick and Grunewald, the researchers at the conference seemed to be operating under the implicit assumption that those supportive cultures are to be found only in public schools.

This is puzzling, since the deepest source of inspiration for most people is not government but family or religion. We should consider the possibility that some such supportive cultures exist in institutions other than the public schools. For example, most (but not all) research finds that low-income blacks who attend Catholic schools graduate from high school in significantly larger numbers than do those from public schools. The Catholic schools accomplish their results at considerably lower cost. Perhaps for some children religion creates the supportive community they need.

For some children, overcoming the untoward consequences of street culture and unsupportive family life might very well be beyond the powers of government. Some people -- teachers, as well as students and parents -- are buoyed by government, inspired even, but very many are not.

There is an opportunity here for a grand accommodation of the political left and right. The accommodation would publicly fund early childhood education and other government programs whose benefits exceed their costs -- but in ways that would leave the poor free to use those funds at nongovernmental institutions, including religious schools.

John Brandl, a former legislator, teaches at the University of Minnesota's Humphrey Institute of Public Affairs.