The Inmates Running the Asylum?
of their educational offerings. Programs that do not sufficiently increase their graduates’ performance on these exams could have their eligibility for federal funding terminated.
For simplicity’s sake in evaluating the impact these alternatives would likely have, we have grouped these proposals together because they tend to lead to improvement and deterioration in similar ways, though the magnitudes can vary.
What Would Improve? To begin with, we would see massive improvement in the definition of appropri- ate measures of quality. No longer would we need to rely on proxies of quality, like input usage. Rather, there would be a direct objective measure of a program’s quality, namely, the extent to which it increases students’ performance on the qualifying/licensing/certification exam. It should also be pointed out that adjusting for the quality of students admitted is fairly straightforward, and doing so would avoid biasing measures of success in favor of more selective institutions.
We would also see massive improvement in the certification of minimum standards, as quality would be defined in terms of meeting some minimum standard of educational value added. Revoking eligibil- ity of programs that did not improve their student’s performance (or do not improve it enough) would
provide clear certification that eligible programs meet some minimum standard. One big advantage of this approach is that accreditation decisions can be reached at a much lower level than the level of the entire institution. For example, an institution could be accredited to offer degrees in accounting and early American history, but not in chemical engineering, all based on their students’ performance on the rele- vant exams. Accreditation could even be offered at the course level, again, presuming a suitable exam could be devised.
This would drastically reduce the importance of any one accreditation decision, which would be ben- eficial. It is inconceivable that a large institution like the University of Texas would lose accreditation because of subpar performance in one field. This, of course, allows for subpar performance to persist. Making accreditation decisions at a lower level would allow for the termination of accreditation for only those areas with subpar performance without endangering the rest of the institution.
We would also see massive improvement in the information provided to the public. With an objective measure of student performance, it is fairly straightforward to determine the relative quality of education provided by each program. If program X and program Y start with similar students and teach in similar ways, but program Y’s graduates score higher on the certification exam, it is safe to conclude that they are providing a superior education. Since the performance of each program’s students on these exams would be publicly available, students, parents, and policy makers could make much more informed decisions. Magazines and newspapers would regularly publish pieces detailing not only what programs provide the best education, but also which ones provide the best value (cost adjusted quality).
Institutional autonomy would also increase. Because eligibility decisions would be based on objective measures of output, there would be no need for restrictive input and procedural requirements. While the current system largely prescribes what inputs colleges should use and how they should be used, under the new system, colleges would be free to use whatever means they deemed necessary, so long as they pro- duced results.
Along similar lines, some types of innovation would be encouraged. Specifically, existing and new col- leges would be much freer to experiment and innovate with new ways of accomplishing traditional things. So while the outcome of teaching students calculus will not change, colleges would have the incen- tive and capability to try new ways of achieving that goal, for instance by altering the curriculum or teach-