Copyright holder: Tyndale University, 3377 Bayview Ave., Toronto, Ontario, Canada M2M 3S4 Att.: Library Director, J. William Horsey Library Copyright: This Work has been made available by the authority of the copyright owner solely for the purpose of private study and research and may not be copied or reproduced except as permitted by the copyright laws of Canada without the written authority from the copyright owner. Copyright license: Attribution-NonCommercial-NoDerivatives 4.0 International License Citation: Accepted Manuscript (AM) Citation: Green, E. H. (2018). The Contours for Researching Religion and School Choice. Journal of School Choice International Research and Reform, 12, (4), pp. 1-25. Version of Record (VOR) Citation: Green, E. H. (2018). The Contours for Researching Religion and School Choice. Journal of School Choice International Research and Reform, 12, (4), pp. 528-545. https://doi.org/10.1080/15582159.2018.1524229 ***** Begin Content ****** TYNDALE UNIVERSITY 3377 Bayview Avenue Toronto, ON M2M 3S4 TEL: 416.226.6620 www.tyndale.ca Note: This Work has been made available by the authority of the copyright owner solely for the purpose of private study and research and may not be copied or reproduced except as permitted by the copyright laws of Canada without the written authority from the copyright owner. Accepted Manuscript (AM) Citation: Green, E. H. (2018). The Contours for Researching Religion and School Choice. Journal of School Choice International Research and Reform, 12, (4), pp. 1-25. This is a pre-copyrighted, author-produced version of an article accepted for publication in Journal of School Choice International Research and Reform following peer review. The version of record is available at: Version of Record (VOR) Citation: Green, E. H. (2018). The Contours for Researching Religion and School Choice. Journal of School Choice International Research and Reform, 12, (4), pp. 528-545. https://doi.Org/10.1080/15582159.2018.1524229 Title: The contours for researching religion and school choice. Running Header: Contours for religion and school choice. Keywords: religion, school choice, social theory Author: Dr. Beth Green Affiliation: Program Director Education, Cardus Canada Contact Details: Email bgreen@cardus.ca Tel. 905.528.8866 x.135 Post. Cardus, 185 Young St., Hamilton ON L8N 1V9, Canada Abstract: This paper argues that contemporary conceptual frameworks for understanding religion may not be adequate for the arena of religion and school choice. Successive published research suggests that people are choosing religious schools for reasons other than religion. But what if the reasons why people might be religious and their motivation for choosing a religious school are more complex? This paper considers some of the latest theoretical conversations around religion and social science. In particular it discusses approaches developed by U.K. Catholic sociologist of education Gerald Grace. The paper proposes some new contours for researching religion and school choice in North America. Introduction One of the interesting phenomena in the school choice landscape is the rising popularity of religious independent schools. For some this seems curiously out of step with the secularization of Western culture and society. Successive published research, in this journal and in others, argues that people are choosing religious schools for reasons other than religion. Ball (2003) 1 argues this in the U.K. context; Campbell, Proctor, and Sherington (2009) argue it in Australia. Jeynes (2002) has deployed it in relation to independent religious schooling in America. The kinds of reasons given are white flight from the inner city, dissatisfaction with the common school, securing an educational advantage—all of which are very powerful explanations. Nevertheless, these explanations rest heavily on rational choice theory as an underlying conceptual framework. Pelz and den Dulk (2018) argue that it is a dominant theory in the school enrollment literature in North America. Rational choice theory assumes that actors make choices to maximize their rewards and that religion is a secondary motivation or option in realizing material benefits. But what if the reasons for choosing a religious school are more complex than this? Quite possibly it is not a case of either/or. This paper discusses the common theoretical approaches underpinning research into religion and school choice in North America. The literature reviewed comes from the Journal of School Choice: International Research and Reform. The journal has an impact factor of 0.42 (Scopes) and was chosen because it is the leading publisher of international scholarship on school choice and reform across multiple disciplines. A keyword search of titles and abstracts using the terms religion and school choice together with appropriate proxies such as faith resulted in 22 articles, 14 of which were directly relevant to this discussion. The point, however, is not the absence of rigorous peer-reviewed literature on the topic; it’s the type of conceptual frameworks deployed in the extant literature and whether they are sophisticated enough. The second section of the paper introduces some substantive critiques of contemporary frameworks for researching religion in the social sciences, drawing out the implications for religion and school choice research in particular. The third section of this paper suggests some new contours for researching religion and school choice. It presents the work of U.K. Catholic 2 sociologist of education Gerald Grace, whose work is largely credited with establishing a rigorous empirical subfield of Catholic education research in the United Kingdom (Green, 2016). This approach offers some very practical examples for future field-building in religion and school choice that rest on more nuanced conceptual frameworks. Contemporary Frameworks School choice policies cast parents as rational actors making informed choices to suit the educational needs of their children. Holmes Erickson (2017) writes that “many questions regarding school choice focus on whether parents are rational decision-makers and select quality schools” (p. 491). These are the natural policy and research questions emanating from a conceptual framework in which education is positioned as a product and parents as consumers (Grace, 2002). This utilitarian approach to modeling the economic and social behaviors of actors is known as rational choice theory, and it assumes that humans generally make rational choices to maximize rewards and benefits. Stark and Finke (2000) are two of the leading proponents of rational choice theory as a framework to explain religious behavior. Put very simply, they propose that because rewards are limited and inaccessible to all, religious beliefs increase the human capacity to imagine or access the possibility of rewards in supernatural or transcendent contexts. Two explanations of the role of religion in school choice emanate from rational choice theory. Pelz and den Dulk (2018) label them the “social identity theory” and the “secular good theory” (p. 80). They argue that these two theories operate concurrently and are dominant in the North American school choice literature. The “social identity theory” conceptualizes school choice as a vehicle for parents to pursue religious distinctiveness by selecting private schools that transmit their belief systems. The “secular good theory” assumes that school choice makes religious 3 private schooling available to parents who value a whole host of other benefits associated with religious schooling, such as stronger academics, discipline, and safety. We can use the Pelz and den Dulk (2018) definitions to track the theoretical approaches adopted in this journal to assess whether these two theories are indeed dominant and whether there are any other emerging possibilities to consider from within the published literature itself. The articles group broadly into three themes: the legal and cultural implications of religious schooling; the role of religion as a factor in parental choice; and the impact of religious schools on educational outcomes. In keeping with Pelz and den Dulk’s observations the social identity theory and the secular good theory appear prevalent in these articles and if anything the secular good theory is most in evidence. Keller (2009), Komer (2009), and Burke and Steppman (2014) all consider the constitutional implications of funding private religious schools in the United States, where the Constitution preserves a legal separation between the church and state. These authors draw on the social identity and the secular good theory as they examine the religious freedoms of individuals and institutions in the United States. In 2009 the Arizona Supreme Court ruled against two voucher programs that gave the parents of special needs students the choice of educating their children in a private or public school (Keller, 2009). Cases like this test the constitutionality of public funding’s following students to private schools, particularly when they are religious schools. Arizona’s two Blaine Amendments prohibit the appropriation of public funds for churches, private or sectarian schools, or public service corporations. Keller’s (2009) argument is that private individuals should not be limited in their ability to decide where to use their public benefits. It rests on the assumption that parents are rational actors who are, as private individuals, entitled to enact choices that maximize rewards. In this analysis religious identity is arguably 4 secondary to other desirable outcomes, namely, the quality of special needs education. Komer (2009) and Burke and Steppman (2014) argue that state courts routinely misinterpret religion clauses, which should not be a barrier to the inclusion of religious schools in properly designed, religiously neutral, school choice programs. Both articles challenge, as anachronistic, the historical sectarian context in which the Blaine Amendments were enacted. These authors argue that mechanisms which fund parents not schools do not in practice contravene the amendments. This is a call for a legal interpretation more cognizant of contemporary religious diversity. This interpretation is underpinned by social identity theory, but the secular good theory is paramount in the mechanism of the legal arguments proffered. Holmes Erickson (2017) conducted a major review of the literature on how parents select schools when they are participating in private school choice programs in America. She investigated the incentives and motivations behind parental participation in schooling markets and the characteristics that parents consider when selecting schools. The criteria focused on parental experience and combined classic school choice studies with more recent research. Holmes Erickson (2017) specifically incorporated theoretical literature that focused on “how parents function as consumers in the market” (p. 493). She also included empirical literature on how parents gathered information and what information they valued when selecting a school. The criteria for the review and the questions it investigated were framed by secular good theory. Holmes Erickson (2017) found great consistency in the school characteristics that parents say they prefer. Her review confirms the classic argument made by Moe (1995) and Friedman (2002) that parents value academic quality. She did also find that parents make trade-offs among preferences when selecting a school. Holmes Erickson argues that parents may value other things besides academic quality. This creates a limited opening for social identity theory as a way of 5 conceptualizing the role of religion as a factor in choice. But this is not an explanation that Holmes Erickson utilizes a great deal in this review. Her analysis primarily draws on the secular good theory, and she argues that religion is typically factored in as a “desired outcome” (p. 497) rather than a causal explanation. It is important to note that Holmes Erickson is careful to avoid causal explanations in this review, preferring instead an interpretive analysis. The reason she gives for this is that experiments and other rigorous econometric methods designed to understand causality are rare in the literature she reviewed because the topic is descriptive. In their analysis of religiosity and the impact of religious secondary schooling, Wadsworth and Walker (2017) cite a wide literature drawing on the secular good theory in relation to Catholic schooling and the labor market. Jeynes (2002) has proffered a similar explanation about the goods associated with private religious Protestant and Catholic education in America, suggesting that the exercise of parental choice in the private sector has to do with perceived higher academic outcomes and group benefits like better socialization into study skills and work habits. Wadsworth and Walker (2017) do also argue for the concurrent social identity theory, suggesting that many families choose Catholic schooling to instill religiosity and that this is effective, at least in the years immediately following graduation. The religious values of parents form the basis of Boerema’s (2009) analysis of school mission as a factor in parental choice and in Reichard’s (2012) empirical case study of parental religious values as a factor in school choice. These are two studies in which the social identity theory is more prevalent and in which the limits of secular good theory as an explanation of parental choice are explicitly tested. Boerema’s (2009) investigation of parental choice and private schooling suggested that, for parents in the Canadian province of British Columbia, mission may matter more in their decision-making than academic achievement. Boerema’s (2009) hypothesis 6 was that academic performance might be a by-product of mission. He argued that the existing school choice literature failed to treat with adequate complexity the differences within the private school sector because of the prevalence of systems-wide analysis. He writes that “one of the challenges in doing analysis based on student achievement is that schools and by extension parents have a variety of goals of which academic achievement captures only a portion” (p. 124). Reichard (2012) designed an empirical case study to examine whether parent religiosity was a statistically significant factor for parents of a private, urban, pre-K-12 religious private school participating in a voucher program. He argued that if parent religiosity were found to be a significant factor then it is likely to form a basis for decision-making. Parents are not simply motivated in their selection by the maximizing of educational outcomes or perceived benefits for students. Reichard (2012) proved his null hypothesis that there was no significant difference in religiosity between parents who paid private tuition and those who used vouchers. His findings suggest that parents who choose religious private schools for their children would do so even if they were financially able and not participating in the voucher program. Both Boerema (2009) and Reichard (2012) acknowledge the limits of their small case studies and their models. It may be that these approaches still don’t adequately address the kind of complexity they are seeking to identify. The main point is that the discussion in both articles centers on rational choice theory as the dominant conceptual framework for understanding religion in relation to parental choice. Both articles implicitly raise the question as to whether this is an adequate conceptual framework to account for the role of religion in school choice research. Holmes Erickson (2017) made a distinction between experimental or econometric research designs. She argued that experimental designs are more capable of establishing causes than interpretative cultural investigations, which are exploratory. In keeping with this distinction, the 7 studies of religion and school climate and/or mission in this journal tend to draw on broader theoretical lenses in their explorations. Although Basford (2010); Sikkink (2012); Beckman, Drexler, and Eames (2012); Van Brummelen and Koole (2012); and Candal and Glenn (2012) undoubtedly draw on the theory of social identity, they also conceptualize the role of religion using more personalist conceptual frameworks. Persons-in-community or personalism was utilized by Bryk, Lee, and Holland (1993) as a theoretical approach in their classic study of Catholic education. It attempts to keep within the conceptual frame interactions between religious beliefs and the particular structures of religious educational communities. These community structures would include effective authority, a strong sense of collective identity, and a pastoral ethic. Sikkink’s (2012) main critique of the school choice literature is that insufficient attention has been paid to within-sector differences in the religious school sector. In this Sikkink is in agreement with Boerema (2009). The study used descriptive analysis of national-level survey data from the Cardus Education Survey (CES) to show average differences across religious school sectors in North America on several measures of school climate and mission (Sikkink, 2012). It found a strong focus on relational goals in evangelical Protestant schools in North America, an emphasis on academic outcomes in Canadian and U.S. Catholic schools, and nearly uniformly positive evaluations of high school experience from private religious and nonreligious schools in North America. Wolf, Greene, Kleitz, and Thalhammer (2001) and Wolf (2007) have all investigated the relationship between religious schools and strong civic outcomes. It should be noted that Sikkink’s empirical research sits with this kind of work on school sector outcomes rather than parental choice. This is nevertheless highly relevant to the discussion of the role of religion in 8 school choice because so much of the school choice debate centers on educational outcomes, and how these outcomes can be adequately described, and compared within sectors and across educational jurisdictions. Sikkink (2012) acknowledges that characteristics typically associated with private religious schools such as effective authority structures and a strong collective identity may positively influence nonreligious outcomes such as civic advantages. Like Boerema (2009), Sikkink’s explanation is that it is the religious mission itself that contributes to the distinctiveness of religious school outcomes, and he has consistently argued that detailed descriptive work of within-group differences is necessary to adequately theorize the religious school sector effect. This supplements social identity theory with thick descriptive accounts of religious personalism. Van Brummelen and Koole (2012) discussed the nature of cultural engagement and citizenship in their research into religious schooling. It is also their assumption that religious mission is a determining factor in the nature of school culture and civic outcomes, although the theoretical framework was more implicit than explicitly rendered in their analysis. Candal and Glenn (2012) deploy race, integration, and group-identity approaches in their study of race relations in urban religious high schools. This is another study that seeks to redress the lack of adequate empirical description of factors that might contribute to religious school culture and outcomes. Candal and Glenn (2012) carried out two case studies of private urban religious high schools in America, one an evangelical Protestant school and the other Roman Catholic. They found that students at the Catholic school placed a stronger emphasis on professional success in their careers and the students at the evangelical Protestant school were more focused on being a transformative influence on society. Their conceptual framework kept school mission and the functional school community in view even as they drew on social identity theory. They 9 have adopted a very similar approach to that undertaken by Boerema (2009) and Sikkink (2012). In addition there is strong theological thread in Candal and Glenn’s (2012) conceptual framework. Theological concepts that frame the belief, structure, and practices of the school community were treated as valid explanations of behavior, although they are not the only lens deployed within the analysis. This is in marked contrast to rational choice theory, which privileges economic and social explanations over religious and theological explanations. Studies of religious school mission and culture within the Journal of School Choice do suggest some other lines of fruitful conceptual inquiry in addition to the secular good theory and social identity theory, which are undoubtedly dominant. [Insert Summary Table here] Substantive Critiques: Religion, Society, and Education Religious schools are positioned within the rational choice framework as one offering among many that parents might pick as a means to achieving outcomes that can be explained in both religious and secular terms such as the conservation of religious identity, or peer socialization within a homogenous group. This is one way of accounting for the rise of religious and cultural pluralism, a dominant feature of late modernity (Smith, 2017). In her historical review of religious schooling in the United Kingdom, Chadwick (1997) makes the important point that this constitutes a substantive conceptual reframing of the mission and self-understanding of religious schools and their relationship to the state. Chadwick (1997) and Grace (2004) trace alternate ways of conceptualizing the distinctive purpose of religious schooling such as moral formation, the preferential option for the poor, and the reproduction of clerical and lay vocations. 10 Glenn (2002) and Berner (2017) compare and contrast the different histories of religious schooling in the Netherlands, the United States, and Canada particularly in the context of emigration to the New World. They argue that complex motivations and multiple educational goals may be of significance to both secular and religious actors today. They propose that this level of diversity is best accommodated within a robust framework of educational pluralism. Berner (2017) defines educational pluralism in this way: “changing the structure of public education so that state governments fund and hold accountable a wide variety of schools, including religious ones, but do not necessarily operate them” (p. 3). Compared to rational choice theory, the reasons people give for their own actions become more relevant in research and policy accounts framed by educational pluralism. Questions about how well social science understands and accounts for religion and religious motivation have emerged in the last two decades. These questions stem from significant critiques of the ability of social science to reflect well on its own ontology and practice, particularly its self-characterization as “distinctively secular” (Cannell, 2006, p. 1). This critique has implications for how religion and school choice have become widely characterized in research and policy. This paper summarizes arguments put forward by three authors, Fenella Cannell, Christian Smith, and Gerald Grace to serve as an entry point into the literature. They have been chosen because they work in different social science disciplines and international contexts. Cannell researches the anthropology of religion, Smith is a sociologist of religion, and Grace a sociologist of education. Cannell, is a secular humanist, and Smith and Grace are Roman Catholic. Readers who want to go straight to the source should read The anthropology of Christianity (Cannell, 2006), Religion: What it is and how it works (Smith, 2017), and Catholic schools: 11 markets and morality (Grace, 2002). All three studies question why it is that frameworks for the study of religion appear insufficiently complex and resistant to treating religion as a “real” or “true” causal explanation of human action. Smith (2017) proffers an alternative theoretical model for conceptualizing religion in late modernity. Grace (2002) maps out a policy and research agenda for religion and education. Cannell poses an important question to her own discipline of anthropology, asking why it “sometimes seems exaggeratedly resistant to the possibility of taking seriously the religious experiences of others” (Cannell, 2006, p. 3). She traces the historical development of anthropology and sociology as relatively young disciplines birthed in the context of nineteenth and twentieth century modernism. One of the dominant trends of the modernist age was to seek rational and predictive accounts of the social world and human behaviors, with religion increasingly conceptualized as non-rational and the secular as rational (Cannell, 2006). Cannell is interested in asking what difference religion makes to how people in different contexts understand themselves and the world. These seem like important social science questions that researchers and policy makers ought to ask. The implication of Cannell’s case is that these questions are routinely neglected in favor of secular rationalist accounts of human behavior, which are regarded as more plausible; this is borne out in this author’s own survey of the field of education and religious school effects (Green & Cooling, 2009). Smith’s original contribution to the social sciences is the development of a critical-realist personalist theoretical account of personhood, morality, human motivation, and practice. He has done so in explicit opposition to the dominance of rational choice theory (Smith, 2014). Critical realism charts a middle route between the extremes of positivism and relativism. Positivism only recognizes as real or true that which is observable and can be measured; relativism insists that 12 truth cannot be absolute since it is always relative to cultural, social, and historical context (Pring, 2015). Critical realism insists that there is an external reality, independent of human beings, much of which is unknown and that our perception of reality and our judgments about it are mediated by our historical and social contexts (Smith, 2017). These contexts need to be part of our explanations of what we know and how and why we act (Smith, 2017). Personalism makes human personhood and identity the grounding of theories of human social life rather than social structures. It assumes that there are particularities associated with being human, grounded in our biology, so that personhood is not simply a social construction (Smith, 2010). Being religious would be an example of one of these particularities (Smith, 2017). This should recast some of our working assumptions about the future for religion and school choice. One of the implicit tendencies in the literature is to review demand for religious schooling as an outlier, something anachronistic or short-lived (Campbell et al., 2009). Smith’s theory of religion shifts this framing, arguing that we need to stop expecting one “dominant trend” to determine the fate of religion. According to Smith (2017) it is unlikely that people will stop being religious. He argues instead that practices of religion and secularism will shift and that they will be much more locally determined and interdependent. If Smith’s (2017) hypothesis is correct, then changing patterns of religious practice will continue to interact with school choice in ways that shape public institutions of education for all of us, including those who do not identify as religious. “We need a conceptual framework that offers more complexity, openness and contingency” (Smith, 2017, p. 252). Smith offers a theoretical model built on critical-realist ontology, personalism, and practice. This offers a number of advantages, three of which are particularly relevant for the discussion advanced in this paper. First, Smith’s model does not artificially separate belief from practice. 13 Second his model avoids the ontological extremes of both positivist and relativist accounts of reality and, third, it does not treat religion as a secondary cause explained by other, more fundamental factors like social class or income. Smith regards the practice of religion itself as capable of creating “new social features and powers that are able to influence people’s lives in the world” (Smith, 2017, p. 4). His argument, elsewhere, is that social science has paid too little attention to the ontology of positivism within which it was formed and that, therefore, it is insufficiently interested in human beings and the complex relationships between their identity, practice, and the social world (Smith, 2014). This paper briefly considers two implications for researching religion and school choice taken from Smith’s general theory of religion. The first implication has concrete methodological consequences for religion and school choice research. Smith’s theory necessitates a focus on practice rather than on beliefs and behaviors. As outlined earlier, the trend in school choice research is to measure educational outcomes in order to identify causes and impact within systems of choice and compare sectors. Thus, we are heavily reliant on survey method, experimental and quasi-experimental research designs. There have been some attempts to broaden the list of what counts as an educational outcome. The Cardus Education Survey (CES), for example, focuses on the religious, spiritual, and social outcomes for graduates, but it still predominantly measures beliefs and behaviors rather than practice (Green, Sikkema, Sikkink, Skiles, & Pennings, 2016). The temptation within the discipline is to redress the balance by emphasizing interpretative accounts of meaning as Taylor (2001) sought to do in her case study of religious schooling and the emergence of school choice policies in Alberta, Canada. While more nuanced than some studies, Taylor’s work ultimately subjected the interpretative, in this case the local story of religious practice, to a limited set of primary socioeconomic and 14 historical causes traditionally used to narrate parental demand for religious schooling. Perhaps this arose because the author did not regard her research as a study of religion per se but a study of school choice. This distinction between the political and economic as causal and religious as interpretative is a function of the same set of assumptions exemplified by Holmes Erikson (2017) noted above. The following divisions are problematic for the discipline: qualitative versus quantitative, causal versus interpretative, and religious versus secular. Smith’s work challenges these false dichotomies because he considers peoples’ reasons to be definite causes in human life. Smith (2017) writes that when studying religion, researchers “need to be both culturalist and causal” (p. 163). Sikkink (2012) and Candall and Glenn (2012) illustrate this approach as they discuss both the cultural and the causal in their research accounts. Smith (2017) argues that we are limited in our ability to identify many causal factors, that our accounts often overemphasize our internal capacities and underestimate how belonging to a social group generates a group- serving bias. He argues that there must remain space for diverse interpretations of causal responsibilities particularly when religious people themselves attribute causal influences to superhuman powers. The third implication of Smith’s work is the grounding of the human person as primary agents of causal influence in human social life. This stands in marked contrast to the lens of rational choice theory or secular good theory (Pelz & den Dulk, 2018). Secular good theory elevates structural explanations over personal ones because it presupposes that human beings always act toward minimizing cost and maximizing reward (Smith, 2017). Smith is arguing that being religious is part of being human; it stands on its own rational ontological grounds. People are not religious merely because they have been socialized that way or for strategic reasons to gain advantage in a cost-benefit analysis. Rather “the prevalence of religious practice is explicable in the light of 15 what we know about human persons”; “it makes sense to say that religion is needed for humans, given the kind of animals we are” (Smith, 2017, p. 232). New Contours for Researching Religion and School Choice In moving toward a more nuanced framework for the study of religion and school choice, we are fortunate to have an exemplar in the scholarship of Gerald Grace. Grace is a sociologist of education who has developed a rigorous practice-based theory of urban education in the United Kingdom and who has made it his life’s work to map a similarly rigorous research agenda for U.K. Catholic schooling (Green, 2016). Routledge recently published Grace’s selected works as part of their World Library of Educationalist series. This series compiles the finest works of international experts and being included is a testimony to the scholarly impact of his work. Grace’s (2002) monumental study of Catholic schooling in England has been cited 370 times (Google Scholar). Grace has no prior ideological commitment to what he would regard as neoliberal school choice research and policy. Catholic schools: Mission, markets and morality, is a critique of privatization and a radical call to Catholic education to retain its distinctiveness even as it is complicit in and reshaped by Western consumer capitalism (Grace, 2002). Two elements make Grace’s work valuable to this discussion. First, Grace calls for policy scholarship better able to take account of religion in education. Second, as Grace models what this scholarship looks like in the context of Catholic education, he lays out a practical research agenda, some of which could be emulated in religion and school choice research. Grace (2002) uses the term “policy scholarship” to denote that there can be no significant understanding of contemporary religion and schooling that does not involve some form of engagement with theological, historical and social rationale (p. 4). He contends elsewhere that the failure to take religion seriously in education research has resulted in a conspicuous silence 16 on the existence of faith-based education and its relevance to policy-making, modern cultural life, and religious practice (Grace, 2004). There are a number of features in Grace’s attempts to delineate the field of Catholic schooling in Britain that can be applied in order to nuance research into religion and school choice. These features are his notion of policy scholarship, his development of analytical questions, and his insistence on the need for multiple forms of research accounts. Grace’s (2002) notion of policy scholarship is a helpful one for grappling with the complexity of religion and school choice research. It would more adequately respond to Smith’s (2017) challenge that social science research should eschew the binaries into which it restricts itself, for example, qualitative versus quantitative, belief versus practice, causal versus interpretative, or secular versus religious. A policy scholarship approach potentially broadens the methodological base of religion and school choice research, offering the possibility that debates could be less reliant on economic and structural accounts of belief and behaviors. At the conceptual and methodological level, policy scholarship requires the development of strong analytical questions and the creation of multiple research accounts, which will be discussed in a moment; but some practical levers have to operate at the meta-level for policy scholarship to happen. Grace (2015) would argue that intentionally creating conversations in peer-reviewed academic journals like this one is a very practical lever, hence his establishment of the journal International Studies in Catholic Education. The deliberate creation of research initiatives and networks with a sustained focus on religion and school choice would also support policy scholarship by re-educating funders and policy-makers and by building research capacity. Grace (2015) suggests that it is important that research initiatives draw together academic and practitioner audiences, or at least disseminate findings across mixed networks so that multiple 17 accounts, including those that are practice-based rather than structurally driven, are included in the policy discussions. Grace (2004) argues that policy scholarship also requires a deeper set of analytical questions that take seriously the practice of the religious actor and their relationship to institutional structures. This would seem to have particular application to religion and school choice research. Grace’s research in Catholic education provides an example of how to use such questions to trace the interaction between the different religious practices of clergy, lay leaders, parents, and pupils in school communities and among different kinds of institutional powers such as religious orders, education policy-makers, and historic ways of understanding Catholic school ethos or catholicity. Green (2012) extended Grace’s (2004) use of analytical questioning to her study of school culture in a religious City Technology College (CTC) in England. CTCs were the forerunner to the academy policy in England and like academies were designated independent schools although they received per capita funding from the state. Although this was not a school choice study, the conceptual approach could be applied to religion and school choice research. Green (2012) developed overarching categories for multiple sequences of analytical questions in order to account for the way that the religious experience of actors worked in the logic of practice surrounding the religious school sector. These categories included the descriptive, structural, and theological. Descriptive questions map context, and they might, for example, count instances of things such as student enrollment data. Structural questions probe the significance of power and location. Examples of structural questions would be: Are strategic decisions and institutional structures entirely new or, how is historical location and historical context significant and are they 18 codependent? (Green, 2014, pp. 298-299). Questions in the personal category are designed to integrate and, in some instances, to prioritize the practice and explanations that are meaningful to religious actors. For example, Green posed some of the following questions in her personal/theological analysis: Do leaders (in religious schools) speak in terms of faith formation and, what theological wells do actors draw from? (Green, 2014, pp. 298-299). This approach deliberately moves away from deploying theory as one, total, overarching causal explanation and uses it instead as a mechanism for organizing and modeling what Smith (2017) describes as “differentiated” and “stratified” levels of reality (p. 150). Finally, Grace (2002) argues that literary image and personal reflective accounts should supplement social theory and empirical accounts. His argument rests on the belief that academic research studies will only provide one form of account of the culture of Catholic schooling and require triangulating with other perspectives. He is making an argument here about how to build a fuller picture of educational culture and how a new subfield interacts with established research perspectives. Grace (2002) would argue that literary image accounts disseminate and amplify particular versions of Catholic education and can comment on academic research assumptions, and vice versa. Grace (2002) also proposes that the individual reflections of graduates looking back on their educational experience allows for another perspective on the positive and negative features of educational formation. It would be beneficial to reflect on how religion and school choice is portrayed in film and media, especially social media. We are at a point in time where graduates of school choice programs are reaching early adulthood; it is important to capture their personal reflections in addition to the outcomes data and the sector-level analysis of academic and employment trajectories. Finally, as Maranto and Redding (2009) have argued it is important to 19 note the ways that prior religious and ideological commitments influence the questions posed by researchers and the phenomena they choose to research. Such commitments are not necessarily a problem providing they are acknowledged and, extending their range would be a useful aid to reflection. As Grace (2002) states, “these accounts are not informed by particular theoretical stances but they provide rich sources of insight” into the long-term effects of education policy. Conclusion Following Smith (2017), this paper argues that we hamper our attempts to understand religion and the motivation of religious actors when we lean heavily on structural or econometric accounts, which often treat religion as a secondary cause. A review of the literature published in this journal shows that rational choice theory is commonly used in religion and school choice research. The paper discussed the substantive critiques arising from within the social sciences themselves of existing theoretical approaches to the study of religion by drawing on the work of Cannell (2006), Smith (2017), and Grace (2002). Smith’s (2017) model of a general theory of religion offers a way of transcending some of the problematic divides between structure and person, cause and interpretation, belief and practice. The paper proposes, following Grace (2002), that we could strengthen religion and school choice research by adopting the notion of policy scholarship, developing a broader range of analytical questions, and promoting multiple research accounts. In conclusion, three overarching questions may prompt greater reflection for religion and school choice research along the conceptual lines argued for in this paper. • What kind of school choice research accounts might help different communities to understand the place of distinct school sectors, especially religious ones? 20 • What are the occasions where school choice researchers would be prepared to change the order of research questions and lead with more description and interpretation rather than explanation and prediction? • How well do the concurrent theories of social identity and secular good serve our conceptual analysis, particularly in relation to parental motivation? Are there ways to introduce greater nuance as researchers generate analytical questions, design methodology, and frame discussion? Considering these questions could help to nuance the conceptual frameworks we rely on in religion and school choice research. One of the strengths of the field is its inter-disciplinary and international range. It is hoped that the discussion in this paper will encourage school choice researchers to draw from the considerable resources available for conceptualizing religion leading to research better able to account for the complexity of religious practice and motivation. References Ball, S. J. (2003). Class strategies and the education market: The middle classes and social advantage. London: Routledge. Basford, L. (2010). From mainstream to East African charter: Cultural and religious experiences of Somali youth in U.S. schools. Journal of School Choice, 4, 485-509. Beckman, J. E., Drexler, J. L., & Eames, K. J. (2012). “Faithful presence”: The Christian school head, personhood, relationships, and outcomes. Journal of School Choice, 6, 104-127. Berner, A. (2017). No one way to school: Pluralism and American public education. New York: Palgrave Macmillan. 21 Boerema, A. J. (2009). Does mission matter? An analysis of private school achievement differences. Journal of School Choice, 3, 112-137. Bryk, A. S., Lee, V.E., & Holland, P.B. (1993). Catholic schools and the common good. Cambridge, Mass.: Harvard University Press. Burke, L. & Steppman, J. (2014). Breaking down Blaine amendments’ indefensible barrier to education choice. Journal of School Choice, 8, 637-654. Campbell, C., Proctor, H., & Sherington, G. (2009). School choice: How parents negotiate the new school market in Australia. Crows Nest, N.S.W.: Allen & Unwin. Candal, C. S., & Glenn, C. L. (2012). Race relations in an evangelical and a Catholic urban high school. Journal of School Choice, 6, 82-103. Cannell, F. (2006). Introduction: The anthropology of Christianity. In F. Cannell (Ed.), The anthropology of Christianity (pp. 1-50). Durham, N.C.: Duke University Press. Chadwick, P. 1997. Shifting Alliances: Church & state in English education. London: Cassell. Friedman, M. (2002). Capitalism and freedom (fortieth anniversary ed.). Chicago: University of Chicago Press. Glenn, C. (2002). The ambiguous embrace: Government and faith-based schools and social agencies. Princeton: Princeton University Press. Grace, G. (2002). Catholic schools: Mission, markets and morality. London: Routledge Falmer. Grace, G. (2004). Making connections for future directions: Taking religion seriously in the sociology of education. International Studies in Sociology of Education, 14, 47-56. 22 Grace, G. (2015). Faith, mission and challenge in Catholic education: The selected works of Gerald Grace. London: Routledge. Green, E. H. (2012). Analysing religion and education in faith-based academies. British Journal of Sociology of Education, 33, 391-407. Green, E. H. (2014). The negotiation and articulation of identity, position and ethos in joint church academies. Research Papers in Education, 29, 285-299. Green, E. H. (2016). Faith, mission and challenge in Catholic education: The selected works of Gerald Grace. International Journal of Christianity and Education, 20, 267-268. Green, B., Sikkema, D., Sikkink, D., Skiles, S., & Pennings, R. (2016). Cardus Education Survey: Educating to love your neighbour: The full picture of Canadian graduates. Hamilton, Ont.: Cardus. Green, E. H., & Cooling, T. (2009). Mapping the Field. London: Theos. Holmes Erickson, H. (2017). How do parents choose schools, and what schools do they choose? A literature review of private school choice programs in the United States. Journal of School Choice, 11, 491-506. Jeynes, W. H. (2002). A meta-analysis of the effects of attending religious schools and religiosity on Black and Hispanic academic achievement. Education and Urban Society, 35, 27-49. Keller, T. (2009). The national implications of Cain V Horne. Journal of School Choice, 3, 353- 367. Komer, R. D. (2009). School choice and state constitutions’ religion clauses. Journal of School Choice, 3, 331-352. 23 Moe, T. (1995). Private vouchers. Stanford, CA: Hoover Institution Press. Pelz, M. L., & den Dulk, K. R. (2018). Looking within or reaching out? The effects of religion on private school enrollments in an era of school choice. Politics and Religion, 11, 79- 115. Pring, R. (2015). Philosophy of educational research (3rd ed.). London: Bloomsbury Reichard, J. D. (2012). Religious values and tuition vouchers: An empirical case study of parent religiosity as a factor of school choice. Journal of School Choice, 6, 465-482. Sikkink, D. (2012). Religious differences in school climate and academic mission: A descriptive overview of school organization and student outcomes. Journal of School Choice, 6, 20- 39. Smith, C. (2010). What is a person? Rethinking humanity, social life, and the moral good from the person up. Chicago: University of Chicago Press. Smith, C. (2014). The sacred project of American sociology. Oxford: Oxford University Press. Smith, C. (2017). Religion: What it is, how it works and why it matters. Princeton: Princeton University Press. Stark, R. & Fink, R. (2000). Acts of faith: Explaining the human side of religion. California: University of California Press. Taylor, A. (2001). “Fellow travelers” and “true believers”: A case study of religion and politics in Alberta schools. Journal of Education Policy, 16, 15-37. 24 Van Brummelen, H. & Koole, R. (2012). “Have they learned to swim around in a big puddle?”: Cultural awareness and engagement of Christian high school students. Journal of School Choice, 6, 40-61. Wadsworth, A., & Walker, J. K. (2017). Religiosity and the impact of religious secondary schooling. Journal of School Choice, 11, 131-147. Wolf, P. J., Greene, J. P., Kleitz, B., & Thalhammer, K. (2001). Private schooling and political tolerance: In P. E. Peterson (ed.), Charters, vouchers and public education (pp. 268- 289). Washington, D.C.: Brookings Institution. Wolf, P. J. (2007). Civics exam: Schools of choice boost civic values. Education Next, 7, 67-72. 25 ***** This is the end of the e-text. This e-text was brought to you by Tyndale University, J. William Horsey Library - Tyndale Digital Collections *****