One University of Minnesota study that began in the 1970s followed 267 children of first-time low-income mothers for nearly four decades. It found that whether a child received supportive parenting in the first few years of life was at least as good a predictor as I.Q. of whether he or she would graduate from high school… Yet the cycle can be broken, and the implication is that the most cost-effective way to address poverty isn’t necessarily housing vouchers or welfare initiatives or prison-building. Rather, it may be early childhood education and parenting programs.
Kristof’s entire editorial in the NYT.