Chapter 6
Management Decentralization and Performance-Based Incentives: Theoretical Consideration for Schools

JANE HANNAWAY

The Urban Institute and Stanford University

The conventional wisdom is that schools in the United States are in trouble. Student performance overall is stagnant, at best, and performance disparities among students are large and persistent. Compared to other industrialized countries, student performance levels in the United States do not look good. Simple solutions—for example, increasing teacher salaries, reducing class size—have not yielded promising results, and most analysts agree that the problems of educational productivity require more complex remedies than simply increasing expenditure-related inputs. The system itself, they argue, must be changed: better-managed resources will yield better returns. Governance reforms such as school-based management, charter schools, and choice schemes are at the center of education policy discussions in the 1990s. They promise better managed resources. But how effective are governance of reforms in education? Are they the key to improved student performance?

This chapter complements the one by Summers and Johnson in this volume, who report on the results of a comprehensive review of published and unpublished studies on the most common type of governance reform—school-based management (SBM). After an exhaustive examination of the evidence, they conclude that it simply is not known whether SBM is worthwhile. Advocacy for increasing school-site discretion has been vigorous, and school districts across the country have embraced it in various forms, but research on its effects on student performance is remarkably sparse. Moreover, the little valid empirical evidence that is available suggests that at least this governance reform is not likely to lead to significantly higher levels of student achievement.

Too often we shy away from asking the hard questions about the effect of



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



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 97
--> Chapter 6 Management Decentralization and Performance-Based Incentives: Theoretical Consideration for Schools JANE HANNAWAY The Urban Institute and Stanford University The conventional wisdom is that schools in the United States are in trouble. Student performance overall is stagnant, at best, and performance disparities among students are large and persistent. Compared to other industrialized countries, student performance levels in the United States do not look good. Simple solutions—for example, increasing teacher salaries, reducing class size—have not yielded promising results, and most analysts agree that the problems of educational productivity require more complex remedies than simply increasing expenditure-related inputs. The system itself, they argue, must be changed: better-managed resources will yield better returns. Governance reforms such as school-based management, charter schools, and choice schemes are at the center of education policy discussions in the 1990s. They promise better managed resources. But how effective are governance of reforms in education? Are they the key to improved student performance? This chapter complements the one by Summers and Johnson in this volume, who report on the results of a comprehensive review of published and unpublished studies on the most common type of governance reform—school-based management (SBM). After an exhaustive examination of the evidence, they conclude that it simply is not known whether SBM is worthwhile. Advocacy for increasing school-site discretion has been vigorous, and school districts across the country have embraced it in various forms, but research on its effects on student performance is remarkably sparse. Moreover, the little valid empirical evidence that is available suggests that at least this governance reform is not likely to lead to significantly higher levels of student achievement. Too often we shy away from asking the hard questions about the effect of

OCR for page 97
--> education reforms on student achievement, sometimes for good reasons 1 and sometimes not. But clearly the value of any education reform effort is determined ultimately by its contribution to student learning. As Shanker (1994) recently wrote: Is student learning better in schools where there are democratic school councils or in schools where the principal runs the show?… The whole point of school reform is to have students learn more. If this doesn't happen, the experiment is a failure, no matter how happy the children, the parents and teachers—and the reformers are. Establishing the link between school-based management and student achievement is not, however, just an empirical problem; it is also a theoretical one. We need to ask why school-based management might be expected to promote student achievement? A long series of connections must be realized for SBM to affect student achievement. Only when we have some understanding of what it is about decentralization that affects or does not affect student learning can we develop appropriate models to estimate the effects of governance reforms on student achievement, and only then can we come to reasonable interpretations of results that are useful for policymaking. This chapter reviews common claims made about the presumed intermediate effects of school decentralization in terms of the available evidence. My analysis suggests that we should not be surprised by the findings of Summers and Johnson. There is little reason to expect decentralization alone to have significant beneficial effects on student performance. Indeed, there is reason to suspect that it might have negative consequences by increasing disparities (Hannaway, 1995). Others agree with the analysis here suggesting that the problem in education is not just who makes decisions. The problem, they claim, is the lack of incentives that tie decisions to performance (Hanushek, 1994). So, also reviewed are the likely consequences of instituting performance-based incentives. I claim its prospects alone, too, are problematic. But performance-based incentives, in tandem with decentralization, may have significant benefits, partly because the advantages of one reform may help balance the limitations of the other, as explained below. Reviewed first is the evidence on decentralization, followed by a discussion of the limitations of incentives. Decentralization is Insufficient Many claims have been made that decentralized governance arrangements have important efficiency advantages primarily because local actors are presumed to have better information about the problems and production possibilities of the system. These advantages are especially pronounced in situations where 1   See, for example, Koretz et. al. (1992), Koretz (in this volume), and Haertel (1986) for a discussion of some of the problematic issues involved in testing and assessment.

OCR for page 97
--> the work is complex, nonroutine, and not easily monitored by centralized mechanisms. On the surface, education appears to be an area where decentralized management arrangements are likely to be particularly advantageous. Consider the following, however. First, education is inherently a highly decentralized operation. Teachers work relatively independently in their classrooms. In fact, many observers use the term "isolated" to describe how teachers function in the education system. Evidence of wide variations from classroom to classroom in what gets taught and how it is taught provides fairly clear prima facie support for already high levels of teacher discretion.2 Claims that teachers are overly constrained by central policies and that freedom from these constraints would unleash creative energies and more productive teacher behavior are, at best, overstated.3 Indeed, from the perspective of individual teachers, SBM centralizes decision making in the school and rests operating and performance responsibility with agents who are closer to teachers. So any positive benefits of SBM may actually come from decreasing rather than increasing the discretion of individual teachers! Recognizing this possibility is important not only for the effective design of SBM efforts, but also because it allows us to then compare the relative merits of SBM with other means of influencing teaching efforts. Second, SBM proponents presume that locating more decision-making authority at the school level would involve actors with interests or information different from actors at the center. As a consequence, different and presumably better decisions will flow from the process. Case studies of decentralization efforts, however, question this effect. Malen and Ogawa (1988), for example, found that even when local school councils were given considerable authority and training, teachers and parents exercised little influence on significant decision-making areas. Weiss and Cambone (1994) found that while SBM opened opportunities for teacher involvement in decision making, relatively few teachers became engaged. Similarly, Easton and Storey (1994) examined rates of participation in the local school councils (LSCs) that were established as part of the Chicago School Reform Act of 1988 and found that while principals participated in council deliberations at high rates, teachers, parents, and community representatives were much less active. Teachers tended to become more involved over time, though their focus on improving instruction remained weak (Hess, 1993). A related and centrally important issue is the presumption in many decentralization reforms, including SBM, that the skills and knowledge necessary to improve student performance already exist at the school level and need only be 2   See, for example, Porter (1989). 3   Many centrally determined policies, with regard to mainstreaming, textbook selection, and some categorical programs, for example, undeniably have significant effects on teachers' classroom behavior; but they are a consequence of state and federal program regulations or court rulings and are unlikely to change significantly under most school-based management schema. For the most part, these regulations focus on social issues; that is, local discretion is constrained by policies determined by wider social values.

OCR for page 97
--> released. Studies of SBM suggest this assumption is a risky one to make. It is unclear that school-level actors have any better idea of how to improve teaching and learning in schools than central actors. A common problem identified in the studies of SBM, for example, is lack of training, suggesting that school-level councils, albeit empowered, are unsure how to proceed and are looking for direction. Although it is too early to judge the effect of Chicago's complex reform effort, preliminary reports suggest that the LSCs focus much of their effort on noninstructional areas such as facilities improvement and student discipline. Some of these efforts may contribute to an environment in which learning is more likely to take place, but observers claim that less than a quarter of the schools have developed plans that are likely to have any significant direct effect on instruction (Hess, 1993).4 No doubt part of the reason for a lack of focus on teaching and learning is that it is far from obvious how to proceed. If certain knowledge about how to increase student performance were readily available, it would have been put into practice long ago. Teachers themselves appear to be an unlikely source of instructionally oriented reform in schools. In the Chicago experiment, teachers see little need for changes in classroom practice. In a recent survey, teachers reported that they felt they were already performing competently in the classroom and that change on the part of the students, not teachers, is necessary for school performance to improve (Hess, 1993). Findings by Weiss and Cambone also suggest that teachers are unlikely to generate instructional reform. She found that teachers in the SBM schools she observed tended to focus their energies on areas not directly related to teaching and learning, such as hall behavior or management of the copying machine. Overall, Weiss and Cambone found little difference in the types of decisions reported in schools with and without SBM. This is not to suggest that allowing schools to determine their own priorities has no value, but experience to date suggests that the changes that take place are likely to be minor and, for the most part, not directly focused on teaching and learning. In short, there is no assurance that schools, left their own devices, will direct their efforts to a much greater extent than they already do to matters that contribute to higher student achievement. Local accountability is a third issue. One of the premises of SBM is that the local actors—principals, teachers, and, particularly, parents—precisely because they are local will be able to hold school operations more accountable than would a remote authority. They are supposedly a source of close-up quality control. It is known, however, from both survey and ethnographic studies that some schools, primarily higher-socioeconomic-status (SES) suburban schools, operate with high 4   Although instructionally relevant change was introduced in only a minority of schools, generating any changes that better daily life in these schools should still be considered an important accomplishment.

OCR for page 97
--> levels of parental control, even when it is neither explicitly encouraged or formally constituted, than schools in lower-SES areas. Higher-SES parents, for example, are far more likely to actively monitor academic assignments, placement decisions, classroom practices, and their child's progress. They are more likely than lower-SES parents to think that getting involved in core academic issues is both their right and their responsibility (Lareau, 1987, 1989). In short, higher-SES parents are already behaving in ways that effectively constitute an important management role in schools. Differences in family-school relations by social class are long standing and well known. Indeed, every study of the intersection of families and schools over the past two decades has underscored significant and important differences by social class (e.g., Ogbu, 1974; Stevenson and Baker, 1987; Heyns, 1978; Lareau, 1987, 1989). Higher levels of education, confidence, knowledge, and status facilitate the involvement of higher-SES parents. Recent survey results paint a similar picture of the differential involvement of higher- and lower-SES parents in their children's schooling. According to the National Education Household Survey, only 7 percent of parents with less than a high school education reported high levels of involvement in school matters, while 42 percent of parents with at least a college degree were highly involved (Zill and Nort, 1994). The existing patterns of parental involvement suggest that decentralization reforms might have one of three results. First, nothing may change. That is, higher- and lower-SES parents might involve themselves at about the same rates as they have in the past despite new opportunities for participation. As it is now, higher-SES parents apparently see few limitations on their involvement in school matters, so the provision of formal mechanisms may have little effect on their behavior. And even with more formal participatory mechanisms, lower-SES parents may still be reluctant to participate for some of the same reasons their participation is low now. A second plausible scenario is that participation rates of higher-and lower-SES parents might converge. That is, while the participation rates of higher-SES parents may already be near their limit, the establishment of formal structures might encourage greater participation by lower-SES parents. The granting of formal "participation rights" may overcome whatever hesitation lower-SES parents otherwise have about getting involved in school affairs. Of course, the involvement of lower-SES parents may or may not be effective in managing school operations and influencing policy decisions in ways that make the schools more productive. A third possibility is that the "take-up rate" on new opportunities for involvement may be greater for parents in higher-SES schools than it is for parents in lower-SES schools. Higher-SES parents already exhibit a greater propensity to get involved, and opening more channels may encourage further participation. In addition, if higher-SES parents also possess the skills, knowledge, and confidence to be involved more effectively than lower-SES parents, decentralization

OCR for page 97
--> reforms may lead to greater inequalities in schooling, as parents with different backgrounds make different levels and types of demands on schools. It is not difficult to imagine SBM schools in low-income neighborhoods functioning pretty much on their own, only weakly managed by parents, and SBM schools in more advantaged neighborhoods operating with parents exercising high levels of quality control. Differences in the focus of demands might also vary by school. Of course, the discretion to identify different objectives is one of the major advantages of decentralization. But one result may be that schools in some neighborhoods focus more heavily and more directly on academic excellence because parents have a better understanding of its long-term benefits, while schools in other neighborhoods are diverted to other objectives. Consistent with the last scenario, preliminary results from the Chicago reform effort suggest that the effect on student performance of shifting governance to the school level is critically linked to the characteristics of a school's students. In particular, schools with large numbers of low-income children (i.e., students eligible for subsidized school lunches) appear to be the least likely to improve student performance as a consequence of decentralization (Downes and Horowitz, 1994). 5 Negative distributional consequences may result from decentralization not only because of differences in the types and levels of parental demands on schools, but also because decentralization may result in a less-than-optimal allocation of teacher effort. School-based management, by definition, increases teacher involvement in management tasks, and there is some indication that involvement in management diverts time from teaching and classroom-related activities (Weiss and Cambone, 1994; Hannaway, 1993). A critical question for productivity in education is how teachers should allocate their time among the three basic activities—classroom teaching, management, and professional development—that consume most of their time. The trade-offs may be important for student achievement. It is probably reasonable to expect greater complementarities between instructionally oriented professional development activities and classroom instruction than between management and instruction. And there is evidence that these complementarities may be particularly important for teachers of at-risk students, who typically encounter severe instructional problems in the classroom (Hannaway and Chaplin, 1994). The clear implication, at least for teachers of at-risk students, is that teachers' nonteaching time might be better spent in professional development and support activities directly related to their teaching responsibilities than in management duties that relate only indirectly to student learning. 5   Changes in district policies in retention and inclusion of students with limited English proficiency in the testing program, however, may affect patterns of student performance results in the Chicago experiment (Bryk et. al., 1994a).

OCR for page 97
--> In summary, the traditional arguments for why SBM increases productivity in education are, at best, suspect. A careful look at what actually goes on in the education process suggests that teachers may wind up with less discretion, new high-quality information and influence channels may not open, and teacher attention may be diverted from the classroom. But decentralized management is not the only reform whose benefits have likely been overstated; this is also true of performance-based incentives. Incentives: A Solution and a Problem The Panel on the Economics of Educational Reform's6 proposal for education reform persuasively argues that, without performance-based incentives, decentralization is unlikely to yield expected improvements in performance. The above discussion and review of evidence support this view; decentralization alone is not likely to be sufficient to promote higher performance in education. Indeed, decentralization alone may do more harm than good by increasing disparities in performance more than it increases average performance. Lack of performance incentives is no doubt part of the problem. The difficulty of solving the incentives problem in any comprehensive way, however, should not be underestimated. Performance incentives create their own problems that are significant enough to call into question the easy efficacy of performance-based reforms in education. Performance-based incentives pose three major problems in education. First, to reward performance, the desired output must be articulated. This is not a trivial task. Education has numerous valid goals, some of which may not even be complementary. Parents, for example, expect schools to not only promote the academic performance of students but also foster creativity, curiosity, self-esteem, tolerance, good citizenship, athletic performance, and a host of other objectives. Even if we limit our discussion to academic objectives, there are still problems. Most Americans support higher academic standards in schools, but there is little agreement about what those standards should be (Ravitch, 1994). Efforts, for example, to develop national standards in history have been particularly contentious, but there have also been skirmishes surrounding math, science, and English standards. Second, there are measurement problems. Defining and agreeing on goals in education is one problem; measuring progress toward those goals is even more difficult, even if we again limit ourselves to only the academic ones.7 Although student performance in such basic skills as math and reading can be measured with reasonable and recognized standards, measuring performance in other aca 6   See Hanushek (1994). 7   See, for example, Koretz in this volume.

OCR for page 97
--> demic areas, such as those associated with higher-order skills, still pose serious problems despite considerable recent development efforts. Establishing value-added measures in these areas greatly compounds the difficulty. A third problem is distortion. The most serious distortion problem is that incentives focus effort on those objectives most amenable to measurement. Whatever is not measured and not rewarded is likely to get less serious attention than it might otherwise merit. Objectives associated with the development of students' problem-solving and higher-order skills, which present difficult measurement problems, are likely to suffer. History suggests that a second important distortion also might result: the performance contracting experiments of the 1970s showed that teachers tended to focus their efforts on students in the middle of the performance distribution, ignoring both high achievers and low achievers, because it was middle students who were most likely to make the biggest performance gains and, therefore, reap the biggest rewards for the teachers (Gramlich and Koshel, 1975). Similar results were reported from attempts in England in the nineteenth century to build performance incentives into education (Rapple, 1990). Drawing on basic research by Holmstrom and Milgrom (1992), using a principal-agent model, I have suggested elsewhere that one solution to the distortion problem created by incentives in situations where there are multiple goals may be in structuring tasks according to different objectives (Hannaway, 1993). For example, the teaching force might be divided into basic skills specialists and higher-order skills specialists, or the school day might be so divided. Incentives could be applied to those personnel or to those tasks for which they are more appropriate, and the structural boundaries around the other tasks would provide protection from distortion. Such a solution, however, would not be easy to implement and might create other problems (Hannaway, 1993). For example, a better-crafted incentives scheme might reduce the likelihood that teachers would devote their efforts disproportionately to some students at the expense of others, but it would also likely complicate its design and implementation. The advantages and distortions associated with performance-based incentives in education have been well known for some time, and, perhaps not surprisingly, experiments with them have generally been short lived (Cohen and Murnane, 1985; Martin et al., 1976). Indeed, in the face of likely distortions it is unclear that such incentives, on balance, are beneficial. That is, in situations where agents (teachers) pursue multiple objectives and where supervision and monitoring systems designed to ensure that all objectives are being adequately pursued are imperfect or costly, the overall system may perform better without explicit performance incentives (Holmstrom and Milgrom, 1992). At a minimum, incentives alone-like decentralization alone-may do more harm than good in education. Incentives alone are likely to lead to unacceptable distortions; decentralization alone is likely, at best, to have little effect and, at worst, to lead to unacceptable levels of disparities. Below it is suggested that

OCR for page 97
--> there may be real advantages to a system that combines a performance incentive scheme with a decentralized governance arrangement. Incentives and Decentralization Although this chapter has emphasized the limitations and problems associated with both decentralization and incentives, each has important potential to improve student performance, and this potential should not be easily dismissed. Moreover, the advantages of one may at least partially offset the drawbacks of the other. In fact, it could be argued that if a well-designed incentives scheme could be developed, there would be little need for decentralized management arrangements. The primary advantage of incentives is that they direct behavior. One of the major criticisms of education in the United States in recent years is that it has lost its central focus. What goes on across schools and classrooms, even in the same school district, is often highly variable and too little directed at challenging academic objectives. If incentives are developed to reward school-level actors for promoting defined areas of student achievement, almost certainly school-level actors will focus more heavily on student achievement, at least in those areas that count in the incentives scheme. Performance incentives are particularly appropriate because of the complexity of the work of teaching; standard rules and regulations governing the process are largely inappropriate. The behavior-directing power of incentives addresses a central weakness of decentralization. A primary drawback of decentralization policies is that many local schools, especially those in disadvantaged areas, may not have the capacity to guide school behavior effectively. Incentives, in contrast, can be counted on to direct some significant fraction of school behavior to valued educational objectives even if the objectives are limited. In effect, incentives ensure some base level of effort to some areas of student achievement and, as a consequence, schools with low capacity will not be left to flounder on their own. The major drawback of incentives is distortion. An advantage of decentralization is that it provides a way to tune the system relatively easily when it becomes distorted. It allows close-up monitoring and adjustments unencumbered by bureaucratic red tape, procedures, and hierarchy. If parents observe, for example, that teachers are overly concentrating on basic skills, they are able to move with little hindrance to correct the behavior, perhaps with the structural remedies described above. Local school actors would not be starting from scratch because the incentives system would have part of the school's work already under control. The job of local actors would be more narrowly defined—to monitor, adjust, and fine-tune the allocation of effort at school. Parents have natural incentives with respect to the operation of schools—the welfare of their children. They simply need a system that makes their job realistically manageable and limits the risk of poor management. Incentives provide a check on decentralization where the risk is that local

OCR for page 97
--> actors may not direct the school's instructional program well. Decentralization provides a check on the distortions likely to emerge in an incentives-only system. The major hitch is that local monitors, usually parents, must be able to detect distortions and be able to make judgments about both the seriousness of the distortions and ways to correct them. This is largely an information problem. Ultimately, for an SBM system with incentives to work most effectively, central authorities should provide information and technical assistance to school-level actors. The content of the information and the way it is provided are consequential. The information contained in ''report cards" on schools, for example, has some of the same problems that incentives do, albeit less severe ones. It can divert attention to those things easily measured. A second type of information, though more costly to provide, is likely to be more valuable. This is information to enhance local actors' understanding of the range of production possibilities in education. It might be acquired by visiting other schools to learn what they are doing; joining a network of schools where technical advice is shared, or bringing in consultants on particular issues. This type of information would help local school actors identify possible shortcomings in their school's instructional program as well as likely solutions. In short, performance incentive schemes in education and decentralized management arrangements each have advantages and disadvantages. To some extent, however, they balance each other out. Decentralized management systems are important because local monitoring can correct some of the distortions created by imperfect incentive systems and incentives are important because of the weaknesses in the management capacity of many local schools. Even with both decentralization and performance incentives, quality performance and process information is needed to ensure success. Research Considerations As Summers and Johnson observe in their chapter in this volume, information on the contribution of SBM and other education reforms to student performance is meager. But we should not be overly critical of education research for the lack of good evaluations. A number of research problems make it a difficult area in which to obtain clear results. One problem noted earlier is the absence of agreed-upon goals. The extent to which all schools are pursuing the same objectives is unclear. Even if a common set of agreed-upon objectives could be defined, there would still be serious measurement issues. If it were easy to identify and measure value-added in education, it would be easy to design incentive schemes to elicit the desired behavior; management issues would be of little concern. Management in education is important, but difficult to evaluate and prescribe precisely, because identifying and measuring agreed-upon value-added is so difficult. Second, any analysis of the value of governance reforms on student achieve-

OCR for page 97
--> ment has to deal with problems of endogeneity. To a large extent management approaches are shaped by the context in which they operate, as a large literature in organization theory attests. In education the degree to which authority is distributed to lower levels is significantly shaped by the social and political environment in which a school district operates. The more conflicting pressure in the political environment of the district, the more likely it is that the school system will be run in a centralized way. External political pressures, for example, as measured by union strength, urban location, or share of revenue received from federal and state sources, have been shown to be negatively correlated, on average, with student performance (Hannaway, 1993). Evidence in the general organization literature also shows that poor performance itself leads to more centralized operations. The implication of this is that simple correlations between governance and performance can be misleading. A third issue is the implicit assumption in much of the reform discussion of a tight connection between management and education productivity. While reformers are likely to overstate the tightness of the connection, some casual link between the management of education organizations and the academic achievement of students is reasonable. We should not, however, underestimate the difficulty of establishing the connection with empirical research. Earlier research has consistently shown that out-of-school factors, especially family background characteristics, are the dominant determinants of student performance. If management has an effect on student performance, it is not likely to be large and therefore requires careful research and precise measurement to capture. Unfortunately, accurate measures of output and process are difficult in education. The small differences in the performance of public and private school students, where management differences are large, support the likely difficulty of capturing clear links between management changes and performance in education. Fourth, decentralization may have different effects in different settings. If it is not explicitly taken into account that SBM may have a greater beneficial effect in some schools than in others, there is a risk of lumping all schools together and coming to faulty conclusions about effects. In some school districts, for example, otherwise alienated families and schools might be bound together in ways that reduce dropout rates later, even though they might have little effect on measured student learning in the short run. The interaction of technical assistance and decentralization also might vary by setting. For example, technical assistance might be an especially critical catalyst for effective decentralized management in disadvantaged areas. Realizing there are limitations to what we can learn from systematic evaluations of governance reforms, however, is no excuse for not attempting them. Indeed, the field is in dire need of more careful analyses of effects in order to guide policy. Finally, with organizational reforms of any kind the "devil is in the details." And in the case of governance reforms in education, we are not sure what the

OCR for page 97
--> critical details are. We need more than simple models, for example, of the ways in which the various elements of governance reforms, such as SBM with incentives, interact with different contexts to affect student performance. There are, indeed, too many studies that focus only on process when the goal is to affect output; but to have any confidence that reforms are having an effect, we have to understand the process. Furthermore, with a better understanding of process we may be able to design alternative, perhaps more efficient, reforms that produce the same process. In short, along with systematic evaluations on the effects of governance reform efforts on student performance, more studies are needed on the nature of education governance itself. References Bryk, A. S., P. E. Deabster, J. Q. Easton, S. Luppescu, and Y. M. Thum. 1994a. "Measuring achievement gains in the Chicago public schools." Education and Urban Society 26:306–319. Downes, T., and J. L. Horowitz. 1994. An analysis of the effect of Chicago reform on student performance. Mimeo, Tufts University, Medford, Mass. Easton, J. Q., and S. L. Storey. 1994. "The development of local schools councils." Education and Urban Society 26:220–237. Haertel, E. 1986. "The valid use of student performance measures for teacher evaluation." Educational Research and Evaluation 8:45–60. Hannaway, J. 1993. "Decentralizing in two districts: challenging the standard paradigm." Pp. 135–162 in Decentralization and School Improvement: Can We Fulfill the Promise? J. Hannaway and M. Carnou, eds. San Francisco: Jossey-Bass. Hannaway, J. 1995. "Decentralization and school reform: a demand perspective." Pp. 203–232 in Advances in Research and Theories of School Management and Educational Policy. R. Ogawa, Ed. Greenwich, Conn.: JAI Press. Hannaway, J., and D. Chaplin. 1994. Breaking the Cycle: Instructional Efficacy and Teachers of "At-Risk" Students. Washington, D.C.:The Urban Institute. Hanushek, Eric. 1994. Making Schools Work: Improving Performance and Controlling Costs. Washington, D.C.: The Brookings Institution. Hess, G. A., Jr. 1993. "Race and the liberal perspective in Chicago school reform." In The New Politics of Race and Gender, C. Marshall, ed. New York: The Falmer Press. Holmstrom, B., and P. Milgrom. 1992. "Multi-task principal-agent analyses: incentive contracts, asset ownership, and job design." Journal of Law, Economics and Organization. Koretz, D. M., G. E. Madaus, E. Haertel, and A. E. Beaton. 1992. National Standards and Testing: A Response to the Recommendations of the National Council on Education Standards and Testing. Washington, D.C.: Rand Corp. Lareau, A. 1987. "Social class and family-school relationships: the importance of cultural capital." Sociology of Education 56:73–85. Lareau, A. 1989. Home Advantage. New York: The Falmer Press. Malen, B., and R. Ogawa. 1988. "Professional-patron influences on site-based governance councils: a confounding case study." Educational Evaluation and Policy Analysis 10:251–270. Martin, D. T., G. E. Overholt, and W. J. Urban. 1976. Accountability in American Education: A Critique. Princeton, N.J.: Princeton Book Co. Ogbu, J. 1974. The Next Generation. New York. Porter, A. 1989. "A curriculum out of balance: elementary school mathematics." Educational Researcher 18:9–15.

OCR for page 97
--> Rapple, B. 1990. Payment by Educational Results: An Idea Whose Time Has Gone? The MacArthur/Spender Special Series on Illinois School Finance. Normal: Illinois State University. Shanker, A. 1994. "Where we stand: the Chicago reform." New York Times, Jan. 3, p. E7. Stevenson, D. L., and D. P. Baker. 1987. "The family-school relation and the child's school performance." Child Development 58:1348–1357. Weiss, C., and J. Cambone. 1994. "Principals, shared decision making, and school reform." Educational Evaluation and Policy Analysis 16:287–301. Zill, N., and C. Windquist Nort. 1994. Running in Place: How American Families are Faring in a Changing Economy and an Individualistic Society. Washington, D.C.: Child Trends. Inc.

OCR for page 97
This page in the original is blank.