Horace Mann hoped schools would be the great equalizer. Nearly two centuries of evidence suggest a more complicated picture. Schools can open doors for children from poor families, and many adults credit a teacher with changing their lives. Yet in nearly every country, children from richer and more educated families do better in school on average, and the gaps appear before the first day of kindergarten. A society that takes the equal worth of every child seriously has to ask why, and what can be done.

The Coleman Report

In 1966 the United States government published Equality of Educational Opportunity, known after its lead author, the sociologist James Coleman, as the Coleman Report. It surveyed hundreds of thousands of students and was expected to show that schools serving Black children had far fewer resources. Differences in resources turned out to be smaller than expected, and the report's headline finding surprised many: family background and the backgrounds of classmates predicted achievement more strongly than school spending or facilities. Some read this as proof that schools make little difference. Later research shows that some things schools control matter a great deal, especially teachers.

What schools can change

Studies that follow teachers over many years find large differences between them. The economists Raj Chetty, John Friedman and Jonah Rockoff, linking school records for millions of students to later tax records, found that students assigned to more effective teachers were more likely to attend college and earned more as adults. Class size also matters in some settings. In Tennessee's Project STAR experiment, begun in 1985, children in the early grades were randomly assigned to small or regular classes. Those in small classes did better, with the largest gains for Black and low-income students. Random assignment matters because it rules out the chance that richer or more motivated parents simply chose the small classes.

Money and law

In the United States a large share of school funding comes from local property taxes, so districts with valuable homes can spend more while taxing less. Courts have been asked to fix this. In San Antonio v. Rodriguez (1973), the Supreme Court ruled 5 to 4 that education is not a fundamental right under the federal Constitution, which left the fight to state courts, many of which have since ordered fairer funding. Earlier, Brown v. Board of Education (1954) had ended legal school segregation, yet many schools remain divided by race and income because neighborhoods are.

Does money matter? For years some researchers argued that spending had little link to outcomes. Newer studies of court-ordered funding increases find that sustained extra spending improved outcomes, especially for children from low-income families, though how the money is spent clearly matters.

The early years

Many gaps open before school starts. A 1995 study by Betty Hart and Todd Risley famously estimated that by age four, children in professional families had heard about 30 million more words than children in families on welfare. The study was small, and later research disputes the size of the gap, but the finding that early conversation matters has held up.

Preschool is the usual remedy, and the evidence is mixed in an instructive way. The Perry Preschool Project in Ypsilanti, Michigan, enrolled a small group of poor Black children in the 1960s and followed them for decades. Participants were more likely to finish school and hold jobs, and less likely to be arrested. The economist James Heckman has argued that such programs repay their cost many times over. But a large randomized study of Tennessee's state pre-kindergarten program found that by sixth grade, children who had attended were doing slightly worse than those who had not. Quality, not just enrollment, seems to decide the outcome.

Tracking

Schools often group students by measured ability, a practice called tracking. Supporters say it lets teachers pitch lessons at the right level and lets advanced students move faster. Critics note that lower tracks often get less experienced teachers and less demanding work, which can lock in early disadvantage. Much depends on whether students can move between groups as they improve.

Worked example: two schools, one test score

School A has an average score of 80 on the state reading test; School B has 65. A newspaper ranks A far above B. But A serves mostly children of professionals, while B serves many children learning English and many from poor families. When researchers measure growth, meaning how much each student improves over the year, B's students gained more than A's. Which school is better? The raw average mostly measures who walks in the door. Growth measures come closer to what the school adds, though they are noisy for small schools and can still be skewed. A fair judgment uses both: the average shows where children stand, which matters for their futures, and growth shows which school is doing more with what it receives. The lesson of the Coleman Report applies: before praising or blaming a school, ask whom it serves.