tivity nor detailed enough to provide much guidance for the design of incentive programs in the education sector (Kohn, 1993; Nalbantian and Schotter, 1997).

A second source of evidence is studies of two specific accountability programs. To our knowledge, these are the only systematic studies of the effect of the programs on achievement. The first study examines the Dallas accountability program (Ladd, 1999), and the other examines the five-year experience of Charlotte, North Carolina, with its Benchmark Goals Program (Smith and Mickelson, forthcoming). The Dallas study examines the impact of that program on student performance on the Texas Assessment of Academic Skills (TAAS), a test that is linked to the state's curriculum and that serves as the basis for the statewide accountability system but is only one of two tests used in the Dallas system. The study design involves comparing the paths of student outcomes in Dallas schools to those in five other big Texas cities during the period that included the year before the Dallas program was implemented (1991) and the following four years (1992–95). This study finds evidence of gains in student achievement for whites and Hispanics but not for black students. Other positive effects included greater declines in the dropout rates and greater gains in attendance rates in Dallas than in the other big Texas cities. The study of Charlotte's program is less encouraging, in that it finds few or no gains from the incentive system (Smith and Mickelson, forthcoming).

One possible explanation for this mixed evidence of gains in achievement is that neither the Dallas nor the Charlotte program was embedded in an overall program of education reform. The need to embed such accountability programs in larger overall reform programs that include, for example, the development of professional capacity emerges as an important lesson from the experience with such programs to date. Evidence for this conclusion comes from the work of researchers under the auspices of the Consortium on Policy Research in Education (CPRE) designed to examine various theories of teacher motivation (Heneman, 1997; Kelley, 1997). They found that teachers within a school were more motivated to exert effort when the school met a variety of enabling conditions, including having a curriculum aligned with the state assessments, adequate revenues, strong districts and principal leadership, and adequate professional development for all teachers in the school. Further suggestive evidence for this conclusion emerges from the rapid gains in North Carolina test scores that were mentioned earlier. The combination of that state's accountability system and professional development strategy appears to be having a positive impact on student achievement.


The fourth generic finance strategy is to give more power to schools to spend as they wish or to parents who can use it to pay for the school of choice for their

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