Alternative motivations for supporting choice and diversity of provision

For some, choice is of intrinsic value, something that parents value for its own sake. Others point to the extrinsic benefits it supposedly brings (Dowding and John, Reference Dowding and John2009). Parents can ‘vote with their feet’ and move their child from one school to another if they think doing so would enhance their household's utility (Tiebout, Reference Tiebout1956). Meanwhile, paid only for the pupils that they teach, schools have to compete with each other for pupils. The resulting competition between providers should encourage innovation and improve standards.Footnote ²

This approach to the provision of public services has been criticised as viewing the citizen from the overly narrow perspective of ‘homo economicus’: individual actors with myopic interests who rationally calculate cost/benefit trade-offs (Jordan, Reference Jordan2005, Reference Jordan2006; Clarke et al., Reference Clarke, Newman, Smith, Westmarland and Vidler2007). According to one such critique, the promotion of responsive public services ‘constructs the public interest as a series of specific and individualised encounters and interactions: each consumer consumes a particular bit of service. Collective consumption of public services is invisible’ (Clarke, Reference Clarke2004: 39).

In any event, the assumption that parents are autonomous consumers of educational services looking to satisfy particularistic household needs may overlook the importance of social context in shaping individual preferences. Alongside ethnicity, language, class and, more recently, gender and age, religion is a key influence on how people perceive themselves and their interests. Insights of social identity theory might help us understand how religious identity and context might be more important than belief in the merits of an individualised approach to public service delivery in shaping mass attitudes towards certain types of school choice.

Social identity theory describes a process of depersonalisation that helps individuals to reduce the complexity of social life (Tajfel, Reference Tajfel and Tajfel1978, Reference Tajfel1981; Turner et al., Reference Turner, Hogg, Oakes, Reicher and Wetherell1987; for a social policy perspective, see Taylor, Reference Taylor1998). Members of a social group tend to internalise their membership by thinking of themselves in collective rather than individual terms: for example, ‘I am Catholic’. To further simplify social life, members resort to self-categorisation (Turner et al., Reference Turner, Hogg, Oakes, Reicher and Wetherell1987) and assign social objects into us/them categories: us represents the ingroup (e.g., ‘Catholics’), where people feel they belong; them stands for the outgroup (e.g., ‘Protestants’), where non-members are located. To make these categories clear and coherent, members maximise their distance from outsiders through stereotyping. This process leads to feelings of intra-group favouritism, greater concern about group rather than individual interests, antipathy towards outgroup members and perceptions that the outgroup presents a threat to ingroup interests.

A key claim of social identity theory is that people's sense of identity depends on the context in which they find themselves. An individual may have multiple identities, such as religious, occupational, sexual and national, but which of these is activated at any one point in time varies (Lau, Reference Lau1989). Individuals are more likely to define themselves in terms of a religious category – and react accordingly – when faced with situations that have religious connotations. When religious ‘category salience’ is high, such as when a perceived antagonism between religious ingroup and outgroup is prominent, individuals tend to feel a stronger sense of attachment to their religious ingroup and its interests, and a greater tendency to be biased against any religious outgroup. However, when the context primes a different definition of the self (such as occupational membership), then the religious component of self-image becomes weaker and religious intergroup antagonism loses its salience.

This suggests the question of faith-based schooling could be a trigger that stimulates feelings of religious identity and of the group interests attached to this identity. That, for example, is certainly what is suggested by one recent study of the demographic basis of support for faith-based schools in Britain (Clements, Reference Clements2010). Faith-based schools may invoke images of ‘us’ and ‘them’ and a desire to defend ‘us’ from ‘them’. Thus, for example, Catholics in Britain may regard the provision of a Catholic school as a means of protecting the interests of their religious ingroup rather than as a mechanism for facilitating individual choice. Conversely, Protestants may view the provision of Catholic schools as a threat to the country's predominantly Protestant heritage (their own ingroup). Moreover, intergroup conflict of this kind is not necessarily confined to the pious. The non-religious may regard faith-based schools of any description as a threat emanating from a religious outgroup to the secular character of British society (their own – secular – ingroup). In short people's attitudes towards faith-based schools may have much more to do with their collective religious identity than with any demand to see individual preferences reflected in how public services are delivered.

Previous research on school choice

Previous research on how parents exercise school choice has certainly raised questions about whether they do conform to the individualistic rationale of utility maximisers seeking the best environment for their particular child. Much of the literature, largely focused on the US, emphasises the importance of attachment to a religious or other reference group in motivating parents’ choice of secondary school, especially when one's own group is perceived to be under attack or in the minority. For instance, Smith and Meier's (Reference Smith and Meier1995) study uses school enrolment data from Florida districts to demonstrate that the choice parents make is a function of a school's religious or racial composition rather than strictly academic criteria. According to their research, parents that leave the public school system for other options seek a homogenous educational environment, either in its religious or its racial composition. The finding is validated by a number of other studies that use both aggregate and individual level data (e.g., Wells and Crain, Reference Wells, Crain and Cookson1992; Whitty, Reference Whitty1997; Sander, Reference Sander2005; Cohen-Zada, Reference Cohen-Zada2006).

There has been some research in Britain that has examined whether the actual selection calculus and role adopted by users of public services is that of the citizen as individual consumer. Clarke and colleagues (Vidler and Clarke, Reference Vidler and Clarke2005; Newman and Vidler, Reference Newman and Vidler2006; Clarke et al., Reference Clarke, Newman, Smith, Westmarland and Vidler2007) have investigated whether citizens see themselves as customers and shoppers when accessing a range of public services. Using survey and focus group data from two English urban locations and looking across three sectors (health, policing and social care), they find no evidence that users think of themselves in that way. Instead, users tend to assume other identities: activists, citizens, members of the local community or patients.

Meanwhile, a number of studies in both England and Scotland have focused on the motivation behind parental school choice in particular (West et al., Reference West, Varlaam, Mortimore and Hargreaves1984; Stillman and Maychell, Reference Stillman and Maychell1986; Adler et al., Reference Adler, Petch and Tweedie1989; Echols et al., Reference Echols, McPherson and Willms1990; Hunter, Reference Hunter1991; West et al., Reference West, David, Hailes and Ribbens1995; Carroll and Walford, Reference Carroll and Walford1997; Burgess et al., Reference Burgess, Greaves, Vignoles and Wilson2009a, Reference Burgess, Greaves, Vignoles and Wilson2009b). These have asked parents themselves to state the reasons behind their choice of school, and in addition have often examined whether the selection criteria parents use vary by parental class or ethnicity. Their findings suggest parents often select a school based on considerations other than academic performance, including a school's reputation, discipline, proximity to home and the pupil's own preference. However, in contrast to much of the research in the US, this research suggests that a school's religious character is not commonly an important decision criterion.

The present analysis

In this analysis, we take a different approach from most previous studies. We focus on the attitudes of the public in general towards school choice and faith-based schools rather than how and why parents choose a school when presented with the opportunity to do so (for other studies of the views of the public in general see Clements, Reference Clements2010; Exley, Reference Exley2012). Our data come from a module of questions on attitudes towards public services that was administered in 2007 on surveys of the resident adult population in each of the four territories of the UK, with each survey conducted in the same or a functionally equivalent manner.

First, the module was included in the 2007 British Social Attitudes survey (Park et al., Reference Park, Curtice, Phillips and Clery2009) and administered to a probability sample of 2,022 respondents, of whom 1,735 were resident in England and whose answers provide our evidence on opinion in England. Second, the module formed part of the 2007 Scottish Social Attitudes survey (Curtice et al., Reference Curtice, McCrone, McEwen, Marsh and Ormston2009), which interviewed a probability sample of 1,508 respondents. Third, in Wales the questions were included in an ad hoc survey (known as the 2007 Wales Life and Times survey) administered to a probability sample of 884 respondents. All three of these surveys were administered face to face by NatCen Social Research. Finally, in Northern Ireland the module was included in the 2007 Northern Ireland Life and Time survey (http://www.ark.ac.uk/nilt/), which was conducted using much the same methodological approach as the three NatCen surveys, and a probability sample of 1,179 respondents were interviewed. To the best of our knowledge, this exercise provides the first-ever opportunity to compare the attitudes of the public in general towards faith-based schooling across all four parts of the UK. In each case, the data have been weighted to correct for known unequal probabilities of being selected for interview and for differential non-response.

The module contained three attitudinal questions central to our inquiry. First, respondents were asked their views about the general principle of providing parents with the ability to choose a school. It ran:

Respondents were invited to respond using a four-point scale ranging from ‘a great deal’ to ‘none at all’. In addition, they were asked their views about one of the new forms of school provision introduced by the UK government in England, namely specialist schools. It read as follows:

In this case respondents were asked to reply using a five-point scale that ranged from ‘strongly support’ to ‘strongly oppose’.

Finally, the module included a question designed to tap attitudes towards the provision of faith-based schools. The wording of this item was as follows:

It should be noted that although respondents were being asked to express their views about faith-based schools in general, the question made reference to one Christian denomination in particular. This was because it was felt that in the absence of such a concrete example respondents may not necessarily fully appreciate the kind of provision – and its connotations – that we wished them to consider. It is recognised that the reference to Catholic schools in particular may well have coloured the views expressed by some respondents, though as will be seen this is, if anything, beneficial to the analysis we undertake. The scale of possible responses to this question was the same as that for the item on specialist schools, which means that the pattern of responses to the two questions on the two different forms of diverse provision is directly comparable.

Meanwhile we should also note that all four surveys carried the same questions on religious affiliation. Respondents were handed a list of religions and denominations and asked:

The only difference between the four surveys was that the list of denominational names offered to respondents reflected the organisational structure of religious denominations in their particular part of the UK.

Equipped with the responses to these four questions, the logic of our analysis is as follows. If the provision of faith-based schools is regarded as part of much the same individualistic agenda on school choice as the introduction of specialist schools, we should find the two forms of provision are similar in their levels of popularity and that both are positively correlated (to a similar degree) with attitudes towards school choice in general. Moreover, there is no particular reason why we should expect attitudes towards faith-based schools to be associated with religious identity. If, on the other hand, faith-based schools are considered to be a means of protecting and developing the collective identity and interests of a particular religion, we would expect to find a clear link between religious identity and attitudes towards the provision of such schools, while there may be little or no association with attitudes towards school choice in general or specialist schools in particular.

However, the link between religious identity and attitudes towards school choice may not be the same in all four parts of the UK. Given that the salience of social identities depends on context, and given that the structure of the religious economy and educational provision varies across the four parts of the UK, there is good reason not to expect this to be the case. First, social identity theory anticipates that religious identities will have a greater impact on people's attitudes where antagonism between religious groups is greater. Such antagonism – between Catholics and Protestants – has historically been greater in Scotland and (especially) Northern Ireland than in England and Wales (Bruce, Reference Bruce1985, Reference Bruce1986; Davie, Reference Davie1994). So perhaps in England and Wales attitudes towards faith-based schools are linked more to attitudes towards school choice in general and less to religious identity than in Scotland or Northern Ireland.

Equally, attitudes towards faith-based schools are likely to be influenced by the pattern of educational provision. A religious group that has access to such schools is more likely to be favourable to their existence than one that does not. So where provision is more or less confined to one particular group, faith-based schools are more likely to be regarded as a means of maintaining the values of that particular group and be met with suspicion by those who do not identify with that group. In both Scotland and Northern Ireland, faith-based schools are almost all Catholic schools. For example, out of 377 maintained faith-based schools in Scotland, 373 are associated with the Catholic Church. In contrast, in England, not only are the majority (67 per cent of 6,834) of faith-based schools Church of England (Anglican) schools, but also these co-exist with a substantial provision (29 per cent) of Catholic schools. Footnote ⁴ The position in Wales is not dissimilar. Thus once again we have reason to anticipate a stronger link between religious identity and attitudes towards faith-based schools in Scotland and Northern Ireland than in England and Wales.