Expanding Choice in Elementary and Secondary Education: A Report on Rethinking the Federal Role in Education

February 2, 2010

Executive Summary

Education choice exercises a powerful pull on parents of school children: Twenty-four percent report that they moved to their current neighborhood so their children could attend their current school; 15 percent of public school students attend parent-selected rather than district-assigned schools; the charter school and homeschooling sectors have grown from nothing to 2.6 percent and 3 percent of total enrollment respectively; private schools capture 11 percent of enrollment; and virtual schooling is poised for explosive growth. Consistent with these behavioral manifestations of the desire of parents to choose their children’s schools, schools of choice consistently generate more positive evaluations from parents than assigned schools.

Arguments for school choice include improving school quality and efficiency through competition among schools for students; enhancing opportunity for students from disadvantaged families who may otherwise be trapped in ineffective schools; and spurring innovation through the greater administrative autonomy likely to exist in schools of choice. Opponents of choice theorize that it will stratify students by family background, result in niche schools that do not convey the nation’s common heritage, provide taxpayer support for religious instruction, and nullify the advantages of standardization in curriculum, teacher preparation, and management that accrue when schooling systems are designed to deliver a common educational experience across a universe of schools. Opponents of choice also argue that many traditional public schools perform superbly and that those that do not can be improved through better resource allocation and management.

Advocates and opponents of choice typically lock horns over idealized systems of schooling that do not presently exist in the U.S. Thus choice advocates frequently espouse voucher systems that would be similar to federal Pell grants at the postsecondary level. Parents would be able to choose any school they wished for their child, public or private, with government writing the check. In contrast, advocates for traditional schooling envision a system in which every school is good enough to ensure that families’ place of residence and income no longer correlate with the quality of the schools to which their children have access.

It is important to note that both the hopes of the advocates of idealized versions of choice and the fears of the detractors diverge from empirical reality. Charter schools and voucher programs are strongly favored by advocates of choice, but studies of the effects of charter schools on student achievement tend to show that on average charters nationally are performing in the same ballpark as traditional public schools, notwithstanding demonstrations that oversubscribed charter schools in Boston and New York City have generated above average academic gains. Studies of voucher programs, including those in Milwaukee, New York City, Dayton, and the District of Columbia, have found some positive effects, but the differences are not large or across the board. At the same time, concerns that voucher programs or charter schools would deplete the budgets of traditional schools, or result in skimming of the most qualified students, or destroy cultural cohesion or learning of common academic content have been unrealized.

The corresponding reality of public schooling is that the quality of schools is substantially correlated with geography and parental income and likely to remain so in the foreseeable future. While there have been improvements in performance in some large urban school districts and prospects for more, not even the strongest advocate of traditional public schools can maintain that we are close to a point at which a parent living in a low-income area can consign her child to the closest neighborhood school with confidence that the school will be as good, on average, as any other school within a reasonable geographic radius of her home, much less good enough to secure her child’s educational future.

We think the situation on the ground with respect to choice is so different from the idealizations that it warrants a new and different perspective on policy. Choice is most frequently realized within the public sector using the mechanisms of residence, magnet schools, and open enrollment systems, whereas the voucher-like systems applauded by choice advocates and feared by opponents are extremely rare. Further, the charter sector is neither large enough nor sufficiently prepared to go to scale to represent a threat to the traditional system of public schools.

Our policy recommendations are framed within the realities of large variation in the quality of public schools, widespread selection of schools by choice of place of residence, and choice being exercised predominantly within the public sector. These realities offer opportunities for common ground between advocates for choice and advocates for public schools. The goals these communities can share are providing more educational opportunity for children from disadvantaged backgrounds and reducing the number of low performing schools. The mechanisms they can share are: a) a system that affords parents as much choice as possible within the universe of taxpayer supported students and schools, b) portals by which parents can readily access rich information on the performance of schools that is framed to be useful in exercising choice, and c) a funding system that supports the growth of parentally preferred schools and school systems, including virtual education programs.

Specifically, to support the expansion of choice we recommend that:

  • choice be exercised through systems in which parents have more options than at present (with the expansion of virtual education programs being a promising means to that end);
  • admission into particular schools within choice systems be open;
  • selection into oversubscribed schools and programs be determined by lottery (which could be conducted using weights to enhance socioeconomic or geographic balance when that is a desired goal);
  • choice systems not include a default (all parents would have to choose);
  • all schools supported with public funds within choice systems be subject to the same standards and assessment regimen under which traditional public schools within a state are required to operate in order to provide transparency for choice;
  • the popularity of schools as revealed through parental preferences be reflected in funding formulas so that more popular schools garner additional resources to meet enrollment demand; and
  • substantially undersubscribed schools be restructured or closed. In order to ground the exercise of choice in valid and easily used information on the characteristics and performance of education programs, we further recommend that:
  • school systems be required to provide timely and relevant information to parents to support choice; • one or more choice navigation websites be developed with the support of federal funds that would be independent of education providers; and
  • school systems be incentivized to link these choice navigation websites to their parental choice systems.

The choice navigation sites would provide substantially more information on the performance of individual education programs than is presently available to parents (via expanded data collections and enhanced investment in an information infrastructure by the federal government); allow parents to create rankings of programs based on the parents’ own dimensions of preference; and give parents access to decision support tools that would aid in considering dimensions of the performance of schools and education programs that have been linked empirically to better student outcomes.

We recognize that meaningful choice and competition can be constrained even when nominal choice is available, for example because all the schools in a district are low performing, or because transportation to higher performing schools is unavailable, or because all schools are homogeneous. We also recognize that both nominal and meaningful choice are constrained in school districts with small populations, many of which are rural. We suggest means for enhancing meaningful choice, for example, by having multiple operators of schools within urban areas, expanding inter-district choice, subsidizing transportation costs when parents choose schools out of the neighborhood, stimulating the formation of quality charter schools, and fostering virtual education by a variety of operators, including nationally chartered providers.

To support the enhancement of meaningful school choice, we recommend:

  • the development of a metric of the extent of choice at the school district level that would be available to the public and policymakers; and that
  • school districts with both low levels of choice and low levels of performance be especially encouraged at the federal level to increase their levels of choice.

Our recommendations do not represent advocacy for any particular type of education institution or program. Rather, school choice should be a democratic process that benefits from the informed participation of parents. Our recommendations are suitable to a range of schooling designs, from a school district in which there are no choices other than district-run public schools, to a system of charter schools, to a division of courses between traditional and virtual schools, to a voucher-based open market in which all providers are on an equal footing, and to many variations in between.

A traditional school district could follow our recommendations by instituting an open enrollment plan at all of its schools, giving additional funding for expansion to oversubscribed schools, closing manifestly unpopular schools, providing transportation to students so that residence does not prevent the exercise of choice, making accredited virtual courses fully count towards graduation, and linking the choice system to a high-quality choice navigation website. Our recommendations are equally applicable to an open market in which public, private, charter, and virtual schools compete on an equal footing for students and the tax revenues that are attached to them.

Our position is that whatever the education delivery design the public has chosen to put in place in a particular school jurisdiction, parents should be afforded the maximum degree of choice, provided with valid information on the performance of the education programs that are available, and have their preferences for education programs reflected in the funding of those programs.

We believe the best evidence suggests that a) parents, including those with low levels of education, can make choices of schools for their children that are sensitive to school performance; b) students from low-income backgrounds benefit from their parents’ decision to send them to higher performing schools; c) the form in which information is presented to parents has important effects on their choice of schools; and d) parental choice can create a competitive market for better schools if the growth of preferred schools and the closure or restructuring of unpopular schools is provided for.

Evidence also suggests that there will be substantial variation in the impact of choice systems on parental behavior, student outcomes, and competition among schools depending on the design of the choice systems and the education options that are available. Poorly designed systems may create greater stratification of schools, reduce educational opportunity for disadvantaged students, and have no systemic competitive effects. Thus, the power of choice to increase educational achievement and opportunity is very much in the details of the design and implementation of choice systems. Because the knowledge base on which to construct school choice systems is far from mature, our final recommendation is that:

the federal role in advancing choice be carried out in a learning context — thoughtful variation in the design of choice systems should be encouraged, systematic data on effects should be collected, and redesign should follow naturally from what has been learned.