Money_Does Matter - school cost's in the United States

School Administrator, May, 1996 by Gerald W. Bracey

Even given the restrictions of range, some empirical investigations have found that money improves achievement. Howard Wainer of the Educational Testing Service found a significant correlation between money and state-level scores on the National Assessment of Educational Progress. Robert Lockwood of the Alabama Department of Education and James McLean of the University of Alabama found that districts that spent more on instructional materials had higher test scores. And Ronald Ferguson of Harvard found that districts that used additional money to reduce class size and to hire more experienced teachers also showed gains in test scores.

Some studies have deliberately used misleading statistics to purportedly show the absence of any impact of money. William Bennett, former U.S. education secretary, correlated the per-pupil expenditures of each state with that state's SAT scores and found essentially no correlation. Pundit George Will leapt at Bennett's conclusion by noting that the top five SAT states--Iowa, North Dakota, South Dakota, Utah, and Minnesota--were all low spenders. New Jersey, on the other hand, spent more money per student than any state and yet ranked only 39th in SAT scores.

What Will may not have known--but surely Bennett did--was that in the five high-scoring states, virtually no one takes the SAT. In Minnesota, only 10 percent of the college-bound seniors took the SAT. In the other four states, the percentages were half that. In New Jersey, on the other hand, 76 percent of the senior class huddled in angst on Saturday mornings to bubble in answer sheets for ETS. Earlier studies had shown that most of the differences among states came from the different proportions of seniors who sat for the test. Surely Bennett knew this. The "wall charts" he produced as education secretary divided states into two categories before ranking them: SAT states and ACT states, depending on which was the more frequently used admissions test. Neither test can represent all states.

Logical Flaws

Certainly, the best-known study failing to find a relationship between money and achievement is that by Eric Hanushek, author of Making Schools Work. It is likely the best known because Hanushek and those who politically approve of the findings keep citing them even though other, more sophisticated analyses of the same data have shown that Hanushek's data do not support Hanushek's conclusions.

Hanushek's famous dictum is that "there appears to be no strong or systematic relationship between school expenditures and student performance." But as consultant Keith Baker later observed, Hanushek never tells you how he formed this conclusion. When Baker reanalyzed Hanushek's data using a formal set of "decision rules," he found, indeed, a positive relationship between money and achievement. More recently, Larry Hedges and colleagues at the University of Chicago reanalyzed Hanushek's collection of studies with a much more sophisticated methodology and found, as Baker did, that money makes a big difference.