Nobel laureate James Heckman spoke to the St. Louis Federal Reserve on Monday about human capital investment. A theme of Heckman’s research is that it’s wisest to invest in children when they’re young. Programs that serve preschool-aged children have a high rate of return, in economists’ parlance, while programs that try to make up for deficiencies later in a child’s life are less successful.

I agree that starting early is the best strategy for improving people’s chances in life, with the caveat that not all policies targeted at this age group do equally well. Here are three things to steer clear of in the realm of child policy:

1. Preschools run by the government. Replicating the public school system for preschoolers will bring about the same inefficiencies it currently suffers from, only with younger students.
2. Costly interventions in the lives of middle-class kids who don’t need help. Programs that involve one-on-one interaction, like Parents as Teachers, should be means-tested. Sending professionals out into homes on the taxpayers’ dime may be worth it in a few extreme situations, if there’s no other way to provide services to a child (for example, if the parents have no way to transport their children to a central location). Intensive help for kids who are going to do fine anyway results in a low rate of return. It also skews the results of these initiatives, making mediocre programs appear successful because so many graduates, who were never at risk in the first place, go on to thrive in school.
3. Standardized tests for three-year-olds. A No-Child-Left-Behind-style assessment of alphabet mastery will yield meaningless data on preschool effectiveness. Heckman is right that early childhood education should stress emotional development rather than academic knowledge. It’s tempting for politicians to impose tests, but a better measure of preschool quality is whether parents choose to enroll their children. As a corollary, no one should be forced into preschool, and parents should have more than one option.