Choice, integration, and educational opportunity: evidence on competitive incentives for student sorting in charter schools.

I. Introduction

Perhaps the most discussed issue in the debate about charter schools, aside from student achievement, is student enrollment patterns. While charters are not confined to traditional attendance zones, and therefore have the potential to aid in the integration of American education, concerns about charter schools "skimming" or "creaming" higher achieving students continue to plague this reform. Although charters serve a disproportionate number of minority students as compared to district-run schools, such information does little to help us understand--and in fact may obscure-causal factors that shape patterns of enrollment. Yet this issue is critical if we are to appreciate the potential of charters to contribute to integration of, and equitable access to, quality schools.

Although sorting and segregation lend themselves to important normative inquiries, they are also empirical questions that are squarely situated in theoretical perspectives of how schools should operate in response to different incentive structures. While charter schools are encouraged to develop alternative approaches, often for under-served populations, they are typically prevented from excluding students based on race, grades, test scores, or ability to pay, as is common in the private school sector. These prohibitions are intended to increase quality opportunities for disadvantaged students, thereby leveling the playing field. However, the laws and individual charter agreements that authorize these schools typically stipulate academic performance as a criterion for continued operation. Furthermore, these schools often operate within competitive environments. In order to survive, charters need to attract students and hold down costs. These constraints force them to pursue educational aspirations, measurable outcomes, and competitive advantages that balance social goals or regulations with market-style incentives. Consequently, we need to know how charters are mediating these demands. In this examination of incentives, we argue that researchers must consider multiple factors of race and economic advantage, in-school mixing, and enrollment patterns within different local contexts in order to appropriately address the issue of how charter schools are serving different communities.

In the next part of this paper we outline the conflicting theories that predict how charter schools might serve or exclude different types of students, focusing largely on the logic of two competing arguments. Then we offer a synthesis of the research on charter school enrollment. Although ''charter school" means different things in different states, and therefore research findings vary, we find a modest consensus that charter schools are not promoting greater integration in American education, not only in terms of race, but also in terms of socioeconomic status (SES), academic ability, and special needs. That review is followed by a summary of research--mostly our own--on supply-side dynamics, which illustrates some of the mechanisms that affect enrollment patterns around charter schools. In our summary, we highlight incentives driving schools to pursue "better" students while simultaneously striving to meet social goals such as integration. The concluding discussion then deconstructs the assumptions on which much of the rational-choice model for market-driven schooling is based.

II. Conflicting Expectations for Charter Schools

The issue of student enrollment patterns in charter schools is a significant concern. Notions of equity and social cohesion are basic values in American education. When combined with the ability to choose among options, these are fundamental elements on which choice-based education reforms should be considered. (1)

Enrollment patterns are important because they are associated with important and widely-valued school outcomes. (2) There are indications that more integrated school environments can improve relations among racial and ethnic student groups and increase the social capital of minority students. (3) However, there are also voluminous amounts of evidence to indicate that the "mix" of social characteristics in a school can impact student achievement--both positively and negatively--largely because of the effect of one's peers on one's own academic aspirations and attainment. (4) So how students are assigned to, or choose to attend, a different school has serious implications individually and in the aggregate for evaluating reforms in American education.