From a Chronicle of Higher Education review of Richard E. Nisbett's Intelligence and How to Get It: Why Schools and Culture Count:
Opinions have changed over the last few years, and many scientists would now agree, "If you were to average the contribution of genetics to IQ over different social classes, you would probably find 50 percent to be the maximum contribution of genetics," says Nisbett, a professor of psychology at the University of Michigan at Ann Arbor. Class is a crucial determinant of intelligence; adoption studies, for example, have indicated that "raising someone in an upper-middle-class environment versus a lower-class environment is worth 12 to 18 points of IQ—a truly massive effect," he says. Children of middle-class parents are read to, spoken to, and encouraged more than children of working-class parents, all experiences that influence intellectual development.
Intelligence and How to Get It also examines how better schooling boosts IQ scores and how school systems can improve. Nisbett cautions that more money does not always equate to higher-quality education, and that parents who take advantage of vouchers to move their children to better schools are a self-selecting group of people who are motivated to help their children excel academically, which leads some researchers to overestimate the vouchers' effectiveness. On the other hand, he finds that class size and teachers' experience and skills can make a big difference, especially for poor and minority children. He notes, too, that children who are exposed to "instructional technologies" in the classroom benefit intellectually; working with word-processing programs, for example, can help students learn to read faster, which leads to further advantages.
A couple of notes regarding education effects. First, a number of studies that control for parental behavior in voucher settings (usually by tracking students who got into voucher programs and those who applied but didn't receive them) consistently show better results for the kids in the voucher schools. Second, the benefits of smaller class sizes are not really clear unless you're talking about specific situations. Overall class size in K-12 public schools has never been lower in the past century or more; yet outcomes are basically flat at least over the past 30 years.
That said, I think Nisbett gets it right that environment matters when it comes to fulfilling potential (however defined).
Read Reason's review of Richard Herrnstein and Charles Murray's The Bell Curve, which covered similar ground. Written by future Nobel Prize winner James J. Heckman, here's a snippet:
Accounting for ability weakens but hardly eliminates the role of education in raising earnings. On average, an extra year of schooling still increases earnings by at least a substantial 6 percent to 8 percent. So there is room for social policy to eliminate earnings differentials between persons of the same ability level. Neither The Bell Curve nor the literature on schooling provides much evidence on the all-important question of the efficacy of education as a tool for equalizing the earnings of persons of different ability levels.
What little is known indicates that ability—or IQ—is not a fixed trait for the young (persons up to age 8 or so). Herrnstein noted this in IQ and the Meritocracy. Sustained high-intensity investments in the education of young children, including such parental activities as reading and responding to children, stimulate learning and further education. Good environments promote learning for young children at all levels of ability. In this sense, there is fragmentary evidence that enriched education can be a good investment even for children of low initial ability, because it stimulates cumulative learning processes and may raise ability. There is much more negative evidence for adults, where ability is a more stable trait. For low-ability adults, there is little evidence that educational investments are socially profitable.
That education matters so much is, I think, a really strong case for taking it out of the essentially exclusive hands of government and experimenting with many different models geared to many different students. You can get prepared food a million different ways in contemporary America. There's something really screwy when virtually all education gets delivered in the same manner.