related to earnings (Card and Krueger, 1996), but their approach cannot deduce an adequate level of earnings from which correlated academic achievement can be observed. As noted, absolute earnings levels are related to a wide variety of macroeconomic and labor market policies, of which educational investment is only one. While we know that better-educated adults have higher earnings than less-educated adults, this may be partly because education is used as a sorting mechanism.

These difficulties in defining adequate outcomes are logically prior to the challenges of attaching input prices to these outcomes. Yet we know very little about how to address them. Meanwhile courts, legislators, and the public will continue to demand that we "put the cart before the horse" and estimate the price of adequacy before we truly know what it is.

None of the preceding is intended to choke debate or impede the efforts of policymakers and analysts to design adequacy-based systems of school finance. Rather, these caveats are extended to those who might otherwise believe that determining what is adequate is a simple task of finding an able cost accountant and thereafter placing the numbers in the correct columns.



As we note below, however, the second approach has been utilized without controls for student background characteristics. And one statistical analysis illustrative of the first approach (Duncombe and Yinger, Chapter 8 in this volume) utilizes a "voter preference/tax price" model to estimate adequate outcomes, without reference to tests or other measures of student performance.


How substantial are these differences? Let's make the simplistic assumption that the only important resources in education were teachers and other classroom inputs. Then, if a typical district in Wisconsin required resources for an adequate education sufficient to fund class sizes of 20, and if Reschovsky-Imazeki's analysis were correct, then Chamber's adjustment would give Milwaukee suburbs 28 percent more resources than needed, or resources sufficient for class sizes of about 14 students. If Chamber's analysis were correct, then Reschovsky-Imazeki's adjustment would give Milwaukee suburbs 28 percent fewer resources than needed, and would only give these suburbs resources sufficient for class sizes of about 26 students.


This methodology, based on an empirical search for districts with adequate outcomes, is the implicit theory of the New Jersey Supreme Court in Abbott v. Burke (1990), in many ways the most radical of state court adequacy decisions. The Court's reasoning suggested a requirement that (poor) districts with low outcomes (including test scores) must be able to achieve high outcomes by spending what (rich) districts with high outcomes spend.


Personal communication, R. Greenwald, March 19-20, 1998.


Although the Augenblick method in Ohio and the Cooper method in

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement