SEARCH

SEARCH BY CITATION

Keywords:

  • deliberative polling;
  • deeply divided societies;
  • Northern Ireland;
  • education policy

Abstract

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information

Deeply divided societies would seem to be infertile ground for mass deliberation. ‘Enclave deliberation’, among people on the same side, may well occur. But people on opposing sides may not trust one another, they may not listen with an open mind, or they may regard the other side's arguments as insincere cover for sectional interests. Perhaps, though, we underestimate their deliberative capacities? This article examines a deliberative poll (DP) in the Omagh area of Northern Ireland, a society having only recently emerged from protracted violence, reflecting and reinforcing the deep divide between Catholics and Protestants. The topic – the future of the local schools – was one on which many of the issues were heavily impinged by the Catholic–Protestant divide. We examine the extent to which a representative sample, including both Catholics and Protestants, was able to deliberate constructively and how the experience changed their policy attitudes and their opinions of one another.

Deeply divided societies, like those of Bosnia, Lebanon and Northern Ireland – riven by ethnic, national, religious, linguistic or other divisions severe enough to threaten the very existence or nature of the state, often accompanied by civil violence – would seem to be infertile ground for mass deliberation. It may not even be possible to get people on opposing sides in the same room. If they do meet they may not really deliberate. They may not trust one another. They may not listen with an open mind. They may regard the other side's arguments as insincere cover for sectional interests. They may have an all too vivid sense of what the other side wants but not see how it could reasonably want it. In Rawlsian terms, they may have trouble viewing each other as reasonable people engaging in reasonable disagreements (Rawls, 1996, pp. 54–8). There may be ‘enclave deliberation’ among people on the same side. But enclave deliberation, apart from not spanning the society, may only deepen the lines of division (as Sunstein's [2009] work on polarisation suggests).

Instead, the most common prescriptions for deeply divided societies have centred on political elites and the institutions structuring their interactions (see O'Flynn, 2007; Dryzek, 2005). In the ‘integrative approach’, electoral engineering incentivises compromise by enabling moderates on each side to cooperate without getting punished by their own supporters at the polls for appearing weak (Horowitz, 1991; Reilly, 2001). In ‘consociational democracy’ (Lijphart, 1977; O'Leary, 2005), opposing elites share governmental power, make pragmatic bargains and seek to sell them to their supporters. These elite-centred strategies may increase stability and decrease violence. But their focus is on managing the conflict, rather than resolving it, and on facilitating ‘peaceful coexistence’, rather than moving towards a more unified democracy without deep division.

Is more possible? We believe that some hope may lie in mass deliberation after all – at least under the right conditions and by a manageably sized but representative subset of the whole population. Ordinary citizens may actually have less intractably opposing views than the elites who speak for them, and the experience of grappling together with policy issues may both help them to see this and bring their views still closer. It may also, perhaps still more importantly, reduce their levels of mutual hostility and distrust. In turn, an event demonstrating all this may encourage elite-level compromise – emboldening moderates while making it harder for hardliners to ‘play the ethnic card’.

This article examines a deliberative poll (DP) in the Omagh District Council area of Northern Ireland, a society that has only recently emerged from protracted, systemic violence, reflecting and reinforcing the deep divide between Catholics and Protestants.1 A random sample were interviewed, then invited to gather to discuss policy issues, under conditions facilitating a free and civil exchange of views (see below). The topic – one heavily impinged by the Catholic–Protestant divide – was the future of local schools.

The results suggest that a representative sample, including both Catholics and Protestants, can be assembled to discuss policy issues; that, once assembled, they have enough in common to permit meaningful and constructive deliberation; that they tend to emerge better informed about the policies under discussion; that their policy attitudes often change as they learn, think and talk about the issues; that at least some of those changes tend to be in the direction of policies involving greater religious mixing and otherwise plausibly serving the interests of the society as a whole; and that the opposing sides tend to grow more respectful and trusting of each other. But let us examine the DP, the context and the results more closely.

The Issues

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information

Beyond requiring all schools receiving state funding to deliver the same academic curriculum, the educational system in Northern Ireland has largely taken ‘segmental autonomy’ (Lijphart, 1977) as a given. In only a small minority of schools is there any significant mix of Catholic and Protestant pupils. More than 90 per cent of the schools are either state controlled (48 per cent) or Catholic maintained (43 per cent), and only 4.9 per cent of the pupils attending state-controlled schools are Catholic, while only 1 per cent of those attending Catholic-maintained schools are Protestant (Department of Education, 2007a). Fewer than 10 per cent of the schools are either formally integrated (4 per cent) or under other forms of management (5 per cent).2

Two practical constraints, however, are forcing schools to consider greater coordination or consolidation. The first is curricular. Recent legislation requires all post-primary schools to adopt the new ‘entitlement framework’, aiming to place greater emphasis on work skills, to offer a wider range of learning opportunities and to give pupils greater flexibility in choosing from that range. Schools must offer 24 subjects for pupils aged 14–16 and 27 for pupils aged 16–19 from which to choose. At least one-third of those subjects must be academic, and at least one-third technical or vocational, with the remaining third left to the discretion of the school (Department of Education, 2005b). Many schools will have trouble meeting these requirements on their own.

The second constraint is demographic. Declining birth rates have meant declining numbers of pupils entering the schools. The entering enrolment has been dropping by about 2 per cent per year. Many schools are short of pupils. Thus the government-commissioned Independent Strategic Review of Education recommended making ‘collaborative approaches to the sharing of facilities and resources … standard practice, while ensuring that the particular identity or ethos of an individual school is preserved wherever possible’ (Department of Education, 2007b, p. 116).

The Omagh Deliberative Poll

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information

The Omagh DP was held on 27 January 2007. Omagh, a district council area of nearly 48,000 inhabitants in County Tyrone, made sense for several reasons.3 It has a mixed population of Catholics and Protestants. Its primary and post-primary schools represent all the major school types. Some have mainly Catholic pupils, some have mainly Protestant pupils, and a few have pupils more evenly split between the two communities. The schools range in size from small to large (from roughly 200 to nearly 1,000 pupils at the post-primary level). As in Northern Ireland as a whole, the birth rate and thus the school-age population is declining. And finally, the availability of a large piece of public land, formerly used as a British Army barracks, made a shared campus a real possibility (Department of Education, 2007c).

The design was the usual deliberative polling design (see, e.g., Fishkin, 2009; Fishkin and Luskin, 2005; Luskin et al., 2002). A random sample of parents were interviewed, and then assembled to discuss the future of the local schools.4 Those parents agreeing to attend were sent balanced briefing documents conveying relevant factual information, outlining the policy options and sketching the arguments for and against them. The discussion took place in randomly assigned small groups, numbering about ten participants apiece. The small group sessions were alternated with plenary sessions, providing the participants with the opportunity to question panels of policy experts and policy makers.

The small group discussions were led by moderators trained to intervene only neutrally and as little as possible. There was no push towards (or away from) consensus; the participants were explicitly told that they need not agree on anything, and that they might come to agree either more or less over the course of the day (an element of the DP design predating but consistent with Karpowitz and Mansbridge, 2005 and Neisser, 2006). The combination of random assignment and random sampling meant that they found themselves talking to others of widely varying backgrounds and views – from other walks of life, as well as from the other community. They took no votes, made no decisions; at the end, they simply completed an augmented version of the same questionnaire as on first contact.

The initial interview sample consisted of 568 parents of school-age children in the Omagh area, interviewed in early January 2007 (T1).5 All were then invited to attend the deliberations, and 127 eventually participated, of whom 121 completed the post-deliberation (T2) questionnaire. The interviewing and recruitment (of interviewees to attend the event) occupied a shorter than usual period, which may have had some effect on the participant sample, as we speculate below.

The briefing materials were drafted in consultation with representatives of all the organisations responsible for managing Omagh's schools, namely the Department of Education, the Council for Catholic Maintained Schools, the Northern Ireland Council for Integrated Education, the Western Education and Library Board (the local schools authority), Comhairle Na Gaelscolaíochta (Council for Irish Language Schools), the Trustees of the Christian Brothers and the Trustees of Loreto Convent. The expert panel consisted of representatives from the same management interests, plus the Community Relations Council of Northern Ireland.

For the briefing materials, these stakeholders only had to agree on a menu of competing policy options, along with the advantages claimed by each option's supporters and the disadvantages claimed by each option's opponents. Most, indeed, strongly opposed varying subsets of the options, and even achieving agreement on the overall menu and description of claimed advantages and disadvantages was a challenge. Similarly, getting representatives of all the stakeholder groups to participate in the expert panels was difficult, and there remained some suspense, up to the day of the event, as to whether they would all show up.6

Portions of the proceedings were recorded and broadcast by BBC Northern Ireland as a way of exposing the broader public to the discussion and inducing public officials to heed the results. The event was also covered, to varying degrees, by the local, national and international press.

Representativeness

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information

A preliminary question is the extent to which the participants can reasonably be taken to represent all parents of school-age children in the Omagh area. Two main sorts of comparison are possible: (1) of the participants directly with the population of Omagh, on those variables for which Northern Ireland Neighbourhood Information Service (NINIS) data are available, as they are for many socio-demographic characteristics; and (2) of the participants with the ‘non-participants’ (the interviewees who took the T1 questionnaire but did not attend the discussions), on those variables, including all beliefs and attitudes, for which census data are unavailable.7

The participant sample looks highly representative in most respects – perhaps most importantly with respect to religious background and national identification. Protestants constitute a slightly larger share (33.9 per cent), and Catholics a slightly smaller one (62.8 per cent) of the participants than of the population (29.7 per cent and 69.1 per cent, respectively), but these differences are small and statistically insignificant. National identification is not available from the NINIS, but the participants differ from the non-participants scarcely at all in that respect. Among the participants, 27.4 per cent considered themselves to be nationalist, 15.3 per cent considered themselves to be unionist, and 54.5 per cent said they were ‘neither’. The corresponding percentages for the non-participants were 25.7 per cent, 10.6 per cent and 57.9 per cent.8

Also importantly, the participant sample is highly representative with respect to both school policy and inter-community attitudes. The participants and non-participants differ significantly on only 4 of 17 school policy attitude items (described below) and 1 of 18 inter-community attitude items (also described below). While our participants were more pro- than anti-mixing and more positive than negative in their perceptions of the other side, the non-participants were too. This has also been the case for the samples in other surveys (of Northern Ireland as a whole, not just Omagh).9 These initial attitudes may surprise some readers, and we shall discuss them further below, but they do appear to be representative.

Two demographic biases do need discussing. The smaller of these concerns education. The participants were distinctly better educated than the population, with 19.5 per cent versus 14.4 per cent of the population at degree level or higher, and 22.0 per cent versus 44.2 per cent of the population having no formal qualification. These numbers, however, are not fully comparable. Given the trends in aggregate education levels, the population from which we sampled (parents of school-age children, almost entirely in their 20s, 30s or 40s) is better educated than the one to which the census data refer (‘household reference persons’ aged 16 to 74), making some though probably not all of the bias suggested by the numbers just reported illusory. The larger bias concerns gender. Far too many of the participants (75.8 per cent versus 49.9 per cent in the population) were women. Since gender roles in Northern Ireland remain largely traditional, the mothers in the sample may simply have been more interested than the fathers in being interviewed about school policy.10

There is no real evidence, however, that either bias made much difference. The participants with no formal qualification, with a secondary-level qualification and at degree level or higher all changed their views in the same ways (in the same direction, to the same degree). So did the male and female participants. The online Appendix S2 gives the details. To be sure, the over-representation of women and the well educated could still have affected the small group discussions in ways that made the attitude changes for both men and women or for all education levels larger (or smaller) than they would have been with fewer female or well-educated participants. But there is also scant evidence of that. The small groups in which at least 80 per cent of the participants were women changed in the same ways as those in which fewer than 80 per cent of the participants were women, and those in which at least 80 per cent of the participants had secondary-level qualification or higher changed in the same ways as those in which fewer than 80 per cent of the participants did so. Again the online Appendix S2 provides the details.

Questions and Hypotheses

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information

Without attempting too precise a definition, we take deliberation to be a form of elevated – serious, substantive, and open-minded – discussion (see Fishkin and Luskin, 2005; O'Flynn, 2006 for more). It could, in principle, concern matters of fact (as in jury deliberations), rather than courses of action, but in our view the most important sort of deliberation, and that with which we are concerned here, is part of ‘deliberative democracy’ – deliberation about policy or electoral choices. The goal is to enable the participants to refine their individual policy preferences and the reasoning behind them, not merely (or necessarily, though we do expect it) to transform their perceptions of others.11

The evidence from past DPs suggests that deliberation frequently changes policy attitudes, both at the individual level (‘gross change’) and in the aggregate (‘net change’); induces considerable learning, which has often been linked to policy attitude changes; and fosters ‘better citizenship’ – more participatory attitudes and behaviours, greater acceptance of political differences and the like (see, e.g., Fishkin and Luskin, 2005;Luskin and Fishkin, 2002; Luskin et al., 2002; Luskin, Fishkin and Hahn, 2007a; Luskin, Sood, Fishkin and Kim, 2009).

But these results were not from deeply divided societies, where mass deliberation is widely thought to be impracticable, if not impossible (e.g. McGarry and O'Leary, 2009; Mouffe, 1999; O'Leary, 2005). Does deliberation produce similar learning and policy attitude change in deeply divided societies? Is this even on issues closely related to a deep divide, like many educational issues in Northern Ireland? Can deliberation across deep divides induce some greater appreciation of the other side and its point of view? Can it (whatever in this vein it may require) create greater mutual trust and understanding?

We expect our data to answer in the affirmative. At least under the sort of safe, fair and facilitative conditions provided by a DP, we expect to find that ordinary people will have enough in common to engage in meaningful deliberation – even in this deeply divided society. This much is admittedly a matter of faith (no pun intended) more than reasoned expectation (although we obviously know the answer as we write).

But given that much, we further expect deliberation to have something like its usual effects, if possibly in somewhat weaker degree than usual. The participants should learn about the issues under discussion, and that learning, coupled with the exposure to other points of view and the occasion to think harder about the issues, should in many cases change their policy attitudes. Other expectations are less certain. Some concern the nature of any policy attitude change. Policies involving or implying more mixing of Catholic and Protestant pupils could either gain or lose support, to the extent that the deliberation either strengthens or weakens negative inter-group attitudes and perceptions or makes them seem either more or less relevant to the choice between policy alternatives. Proposals unrelated to the Catholic–Protestant divide may similarly, for reasons harder to anticipate, either gain or lose support. Other uncertainties concern the effects on the differences between Catholic and Protestant participants. Will they change in the same direction? Will the gap between them shrink or widen (will they converge or diverge)? Will they wind up on the same side of the issue? How far apart? We shall see.

Finally, there are the possible effects on the participants' attitudes about the conflict, the other side and the value and possibility of peaceful accommodation. These could be negative. Heated disagreement could stoke hostility and reinforce negative stereotypes. But we expect them to be largely positive. Talking with others in a safe public space about issues related but not always reducible to the deep divide may increase perceptions of the other side's trustworthiness and openness to reason. Other DPs have produced effects of this general sort (Luskin and Fishkin, 2002), although deep divisions may pose a sterner test.

The School Policy Issues

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information

The Omagh school policy attitude items fall into four main clusters.12 Ordered by ease of extrication from the Catholic–Protestant divide, beginning with the most easily extricated, these are as follows.

Age Groupings

Four items asked about the ideas of having mostly schools for ages 11–18, with some schools for ages 11–16 (the status quo); of having most schools for ages 11–16, while converting one or two schools into 16–18 sixth-form colleges; of switching to a system of junior and senior high schools for ages 11–14 and 14–18, respectively; and of having schools combining primary and post-primary pupils (for ages 7–14, for example). Note that supporting the first idea (again, the status quo) implies opposing the other three, while supporting any of the others implies opposing the former.

School Types

Another three items concern possible sortings of schools by ability or subject matter. These ask about the ideas of having both academic schools and technical/vocational schools (essentially the status quo, since, de facto, grammar schools are academic, while comprehensive schools are vocational); of having a system of specialist schools (e.g. for language, science or technology); and of having a system of all-ability schools, all providing the same broad curriculum.

School Collaboration

Four items concern the ways in which existing schools might collaborate. They ask about the possibilities of schools sharing facilities like technology labs or sixth forms; of their sharing a campus while retaining distinct identities; of pupils travelling to neighbouring schools to be taught subjects unavailable at their own; and of teachers travelling to neighbouring schools to teach subjects unavailable there.

Religious Mixing

Finally, six items raise the possibility of mixing Catholic and Protestant pupils in the same classroom. Two raise it somewhat obliquely, asking about the ideas of retaining all the existing school types in the Omagh area (the status quo)13 and of establishing schools jointly managed by the Catholic Church and the Western Education and Library Board or Protestant churches. Nominally, these items concern school organisation rather than mixing, but both carry clear implications for mixing – negative in the first case, positive in the second. The remaining four items raise the issue more directly. They ask about the ideas of requiring schools that need to partner to deliver the new curriculum to partner with their closest neighbouring school, even if not of the same religious composition; of requiring unmixed schools to partner with a school containing children of the other religion; of at least sometimes teaching the pupils from partnering schools of differing religious composition in the same classroom; and of increasing the number of formally integrated schools, in which all parties responsible for managing education, including the churches and the Education and Library Board, have a right to play a role.

Inter-community Beliefs and Attitudes

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information

Here we have measures of each side's perceptions of the other's trustworthiness and openness to reason; how positively or negatively they feel towards one another; their agreement or disagreement that ‘mixed education promotes mutual respect and understanding’; and their views on the importance of children ‘attend[ing] school only with other children of their own religion’ versus ‘attend[ing] schools that have a balanced enrolment of Protestant and Catholic pupils’, the extent to which ‘changes in the Omagh area's education system can equally benefit’ both Catholic and Protestant children versus being inevitably zero-sum, and the extent to which ‘better relations between Protestants and Catholics will only come about through more mixing of the two communities’ versus ‘only com[ing] about through more separation’.

Results

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information

Knowledge

The questionnaire contained seven factual knowledge items relating to education in Northern Ireland generally and Omagh specifically. These asked about the percentage of schools having a majority of one religion but at least 10 per cent of the other in their enrolment (5–10 per cent); the percentage change over the past five years in the number of children entering Omagh schools (a 10 per cent decrease); the minimum number of subjects 14-year-old pupils must have the opportunity of choosing under the new entitlement framework (24); the proportion of the subjects offered that must be ‘applied’ under the entitlement framework (one-third); the proportion of grammar school pupils going on to university (roughly three-quarters); the relationship between age and funding (greater for older pupils); and the official employer for voluntary grammar school teachers (the school's board of governors).

Table 1 shows that the participants learned a great deal. Across the seven knowledge items, the percentage answering correctly soared from 21.8 per cent to 50.3 per cent, a gain of 28.6 per cent. On six of the seven items, the gain was statistically significant, and on some it was night and day. The percentage correctly identifying a menu of at least 24 subjects for all 14-year-old pupils as one of the new curricular requirements rose from 21.0 per cent to 75.0 per cent, and the percentage knowing that schools receive more funding for older than for younger pupils from 22.6 per cent to 79.0 per cent.

Table 1. Knowledge Gains, T1–T2
Item (identified by right answer)T1T2T2−T1p
  1. Note: n = 124.

5–10% of majority-Protestant/-Catholic schools have at least 10% of other religion0.2420.3550.1130.026
The number of children entering Omagh schools has decreased by 10%0.1850.4760.2900.000
The new entitlement framework requires that all 14-year-olds be provided a choice of at least 24 subjects0.2100.7500.5400.000
The new entitlement framework requires that of all the subjects offered be applied0.2900.6290.3390.000
About ¾ of grammar school pupils attend university after leaving school0.2900.4350.1450.002
School funding is currently greater for older pupils0.2260.7900.5650.000
The official employer for all teachers in voluntary grammar schools is the school's board of governors0.0810.0890.0080.783
Knowledge index (% correct)0.2180.5030.2860.000

Policy Attitudes and Attitude Change

We examine two broad questions about these school policy attitudes: (1) where they wound up; and (2) how they changed to get there. To maximise comparability, we linearly translate all our policy attitude items to the 0–1 interval, with 0 representing the strongest opposition, 1 the strongest support, and.5 neutrality.14 In these terms, (1) is a question of the T2 mean μ2 (>.5 for support, <.5 for opposition), while (2) is a question of the difference of the T2 and T1 means μ2 − μ1 (>0 when attitudes become more supportive, <0 when they become more opposed). The answers (1) paint a picture of what a more deliberative public might think and (2) suggest how deliberation has changed that picture from what it would otherwise have been (and thus also, albeit less certainly, how further deliberation might change it further).

Table 2 shows the results. Without exception, the mean opinion falls distinctly to one side or the other of each issue. With only one exception, that side represents support for whatever the item text proposed – in three cases the status quo, in the rest departures from it (a half dozen of them involving more religious mixing). This may suggest some acquiescence bias, but the strong support, even at T1, for all but one of the proposals for increased religious mixing is actually consistent with other survey results, which have routinely shown majorities wishing to send their children to mixed schools and wanting to see the number of such schools increase (Boyle and Hadden, 1994, p. 33; NILT, 2010; Taylor, 2001, p. 43). That our more concrete proposals in that direction also draw high levels of support suggests that these attitudes are real, and more than just vague aspiration.15

Table 2. School Policy Attitudes, T1–T2
 T1pT2pT2−T1p
  1. Note: The p values for T1 and T2 refer to the null hypothesis of the mean being 0.5. For T2−T1, they refer to the null hypothesis of no change. The ns range from 108 to 121.

Age grouping
Keeping the traditional pattern of schools for ages 11–18 and 11–160.7340.0000.6150.000−0.1190.002
Having most schools 11–16 and converting one or two schools into 16–18 sixth-form colleges0.6390.0000.5320.243−0.1070.002
Switching to a system of junior and senior high schools for ages 11–14 and 14–180.5810.0080.6300.0000.0480.241
Having schools combining primary and post-primary pupils, e.g. ages 7–140.4310.0300.4050.003−0.0260.476
School types
Having both academic schools and technical/ vocational schools0.7920.0000.7580.000−0.0340.255
Having a system of specialist schools for, e.g., language, science or technology0.6300.0000.6190.000−0.0110.741
Having a system of all-ability schools, all providing the same broad curriculum0.7180.0000.6670.000−0.050.132
School collaboration
Getting schools to share facilities like a technology lab or a sixth form0.7250.0000.7580.0000.0320.331
Getting schools to share a campus while still retaining distinct identities0.6750.0000.5970.002−0.0780.018
Getting pupils to travel to nearby schools for subjects unavailable at their own0.6140.0010.6230.0000.0090.799
Getting teachers to travel to nearby schools to teach subjects unavailable there0.6580.0000.6430.000−0.0150.651
Religious mixing
Retaining all school types currently in the Omagh area0.7010.0000.6050.000−0.0960.004
Requiring partnering with the closest nearby school, regardless of religious composition0.6550.0000.7250.0000.0700.028
Requiring unmixed schools to partner with a school with children of a different religion0.6500.0000.6220.000−0.0280.399
At least sometimes teaching the pupils from partnering schools of differing religious composition in the same classroom0.7710.0000.7800.0000.0080.759
Establishing schools jointly managed by the Catholic Church and the Western Education and Library Board or Protestant churches0.6580.0000.6330.000−0.0250.477
Increasing the number of formal integrated schools0.7450.0000.7250.000−0.0200.476

On five of the seventeen items, there was statistically significant and substantively important attitude change.16 To help benchmark the numbers, note that if one-and-a-half times as many participants (say 24 per cent versus 16 per cent) move twice as far (two response categories versus one) in one direction as in the other on a classic five-point agree–disagree scale – quite a notable aggregate shift in the one direction – that corresponds to a net change of 0.08 on the 0–1 scale to which all these items have been translated (Luskin et al., 2006). In this light, two of these changes (on schools sharing a campus while retaining distinct identities and requiring partnering with the nearest neighbouring school, regardless of religious composition) were quite notable (0.070 and 0.078, ignoring the signs), and another three (on retaining all school types currently in the Omagh area, keeping the traditional pattern of schools for ages 11–18 and 11–16, and having most schools 11–16 while converting one or two into 16–18 sixth-form colleges) very big (between 0.096 and 0.119 in magnitude).

These changes are mainly of two overlapping sorts. Four of the five show either decreased support for the status quo or increased support for departures from it, probably reflecting increased recognition of the demographic and curricular constraints previously noted. The lone exception was the most ambitious version of school collaboration, the idea of getting schools to share a campus while still retaining distinct identities, support for which ebbed by 0.078. Our surmise is that this struck many respondents as an awkward and taxing arrangement.

More importantly, two of the five changes show increased support for policies entailing increased religious mixing or decreased support for policies not doing so. Two of the six religious mixing items thus changed in a pro-mixing direction. Support for requiring schools to partner with the closest neighbouring school, regardless of religious composition, increased, while support for retaining all existing school types, which would limit any increase in religious mixing, declined.

Catholic versus Protestant Policy Attitudes and Attitude Change

The same two broad questions – (1) where they wound up and (2) how they changed to get there – may be asked about the school policy attitudes within each community. But here three additional questions comparing the communities also arise: (3) how close or far apart they wound up; (4) whether their attitudes changed in the same or opposite directions; and (5) to what extent they converged or diverged. If the Catholic and Protestant means are μC1 and μP1 at T1 and μC2 and μP2 at T2, (3) is a question of |μC2 − μP2|, (4) a question of the sign of (μC2 − μC1)(μP2 − μP2) (positive when the groups change in the same direction, negative when they change in opposite directions, and 0 when either does not change), and (5) a question of the sign of |μC1 − μP1| − |μC2 − μP2| (positive when the groups converge, negative when they diverge, and 0 when they remain the same distance apart). Note that any combination of answers to (4) and (5) is possible. Changes in the same direction can produce either convergence or divergence depending on their relative magnitudes, as can changes in opposite directions, depending on the proximity of the starting points, the directions of the changes (towards or away from the other group's starting point) and their combined magnitude (see Figure 1).

figure

Figure 1. Directions of Change and Convergence/Divergence

Note: Ct and Pt are the Catholic and Protestant group means at time t (= 1, 2) (or T1 and T2, in the notation of the text). In all these diagrams, the order of the group means remains the same over t, namely P1 > C1 and P2 > C2. But all four combinations can also obtain when the order of the group means changes over t (say P1 > C1 but C2 > P2). We have not bothered to illustrate that case, given its rarity.

Download figure to PowerPoint

Table 3 addresses these questions, presenting the means for Catholics (C), Protestants (P) and the difference between them (P−C) at T1, at T2 and as they changed from T1 to T2. Its last column also presents the change in the absolute distance between the two groups, a measure of how much they diverged (when the entry is positive) or converged (when it is negative).

Table 3. School Policy Attitudes, Catholic vs. Protestant Participants, T1–T2
 T1T2T2−T1
CPP−CpCPP−CpCpPpP−Cp|P − C|
  1. Note: n's range from 68 to 78 for Catholics and from 36 to 40 for Protestants.

Age grouping
Keeping the traditional pattern of schools for ages 11–18 and 11–160.7370.728−0.0090.8700.6530.559−0.0940.169−0.0840.064−0.1690.014−0.0850.2870.085
Having most schools 11–16 and converting one or two schools into 16–18 sixth-form colleges0.6260.6450.0190.7360.5030.5780.0750.193−0.1230.006−0.0680.2530.0550.4480.056
Switching to a system of junior and senior high schools for ages 11–14 and 14–180.5850.566−0.0200.7580.6040.6550.0510.4870.0190.6980.0890.2440.0700.4380.031
Having schools combining primary and post-primary pupils, e.g. ages 7–140.4410.408−0.0330.6230.3940.4110.0160.810−0.0460.2980.0030.9690.0490.551−0.017
School types
Having both academic schools and technical/ vocational schools0.7800.8220.0420.3790.7800.703−0.0770.1500.0001.00−0.1190.007−0.1190.0410.035
Having a system of specialist schools for, e.g., language, science or technology0.6070.6700.0640.2590.6400.565−0.0750.1770.0330.434−0.1050.076−0.1390.0560.011
Having a system of all-ability schools, all providing the same broad curriculum0.7000.7390.0390.5290.6620.658−0.0030.962−0.0380.347−0.0810.217−0.0420.580−0.036
School collaboration
Getting schools to share facilities like a technology lab or a sixth form0.7350.695−0.0400.4810.7680.729−0.0390.4520.0330.4440.0340.5310.0010.990−0.001
Getting schools to share a campus while still retaining distinct identities0.6870.646−0.0420.4670.6010.595−0.0070.916−0.0860.046−0.0510.3510.0350.616−0.035
Getting pupils to travel to nearby schools for subjects unavailable at their own0.5680.6870.1190.0660.6280.605−0.0230.7160.0600.215−0.0820.166−0.1420.063−0.096
Getting teachers to travel to nearby schools to teach subjects unavailable there0.6300.7070.0770.2070.6470.633−0.0150.8000.0170.706−0.0750.131−0.0920.169−0.062
Religious mixing
Retaining all school types currently in the Omagh area0.7030.686−0.0160.7570.6550.522−0.1340.025−0.0470.233−0.1650.006−0.1170.0910.118
Requiring partnering with the closest nearby school, regardless of religious composition0.6610.612−0.0500.3960.7440.678−0.0660.2840.0820.0630.0660.124−0.0160.7850.016
Requiring unmixed schools to partner with a school with children of a different religion0.6350.6580.0230.7020.6180.6180.0001.00−0.0160.693−0.0390.493−0.0230.745−0.023
At least sometimes teaching the pupils from partnering schools of differing religious composition in the same classroom0.8010.705−0.0960.0520.7890.750−0.0390.453−0.0100.7780.0530.2910.0620.3000.006
Establishing schools jointly managed by the Catholic Church and the Western Education and Library Board or Protestant churches0.6940.597−0.0970.0990.6690.568−0.1010.090−0.0250.569−0.0290.645−0.0040.9630.004
Increasing the number of formal integrated schools0.7690.697−0.0720.1870.7230.7260.0030.955−0.0460.2380.0290.4520.0750.170−0.069

The short answer to (1) and, in large measure, to (2)–(5) is that Catholics and Protestants resembled each other (and thus the sample as a whole). On every single item, Protestants and Catholics both started and wound up on the same side of the scale (the supportive side, on every item but one). On six of the seventeen items, there is some perceptible gap at T2, but only one of those differences (on retaining all existing school types) is statistically significant (p < 0.05, two-tailed), and only one more (on establishing schools jointly managed by the Catholic Church and the Western Education and Library Board or Protestant churches) is relatively close (p = 0.090). The Protestants emerged as less supportive of retaining all existing school types – making them, in that respect, more pro-mixing – but also less supportive of establishing jointly managed schools – making them, in that respect, less pro-mixing. More generally, there is no particular pattern to the directions of these generally modest differences, with Protestants falling to the pro-status quo or pro-mixing side of Catholics on some but to the opposite side on others.

The T1 to T2 changes, too, were similar in both sign and magnitude. On none of the seventeen items did Catholics and Protestants change significantly in opposite directions. On every one of the six items on which at least one of the groups changed significantly, the other changed either not at all (in one case) or in the same direction (in the other five, significantly or almost significantly in two). On a seventh item, both groups changed almost significantly (p = 0.063 and 0.124) in the same direction. The degree of change differed significantly only on the idea of having both academic and technical/vocational schools, of which Protestants became much less supportive (moving away from the status quo), while Catholics stood pat. On three other items, the differences between the groups' attitude changes did approach significance (p < 0.10): Protestants again became much less supportive of having a system of specialist schools, while Catholics again stood pat; and both Protestants and Catholics became less supportive of both getting pupils to travel to nearby schools for subjects unavailable at their own and retaining all existing school types, but unequally so, with the decrease in support being greater among Catholics in the first case and greater among Protestants in the second.

Nor, finally, did these changes produce any routine convergence or divergence. The mean |P−C|, across the seventeen items, is only 0.004. Taking 0.080 as a cut-off suggests only two notable cases of divergence, and only one of convergence. The communities diverged on the ideas of keeping the traditional pattern of schools for ages 11–18 and 11–16 and of retaining all existing school types. In both cases, initially equal and very high levels of support declined in both communities, but more sharply among Protestants. The communities converged on the idea of getting pupils to travel to nearby schools for subjects unavailable at their own. Initially high levels of support declined in both communities, but more among the Protestants, who were initially more supportive.17 But these are exceptions. In the main, Catholics and Protestants changed their views (to the extent they did so) roughly in tandem.

Inter-community Beliefs and Attitudes

At least on the surface, the participants had mostly positive views of one another, even at the outset. The T1 means in Table 4a suggest that both Catholics and Protestants felt much more favourably than unfavourably towards the other community (0.768 and 0.654, respectively). Each also saw the other as somewhat more open to reason than not (Catholics at 0.543 regarding Protestants, Protestants at 0.560 regarding Catholics) and distinctly more trustworthy than not (Catholics at 0.621 regarding Protestants, Protestants at 0.642 regarding Catholics). Granted, these means may overstate the degree of positivity. Sizeable numbers of respondents gave midpoint responses (neither trustworthy/open to reason/favourable nor untrustworthy/not open to reason/unfavourable) in characterising the other side, and there is reason to suspect most of those responses of representing veiled antipathy.18 But positive responses far outnumbered overtly negative ones, by factors ranging from 1.3:1 to 8:1, and even treating, say, three-quarters of the midpoint responses as de facto negative would leave relatively even T1 splits between positive and negative perceptions – more positive distributions than might have been expected in a ‘deeply divided’ society.

Table 4. Inter-community Beliefs and Attitudes
 T1T2T2−T1
MeanpMeanpMeanp
  1. Note: In part a, n's = 68–74 for Catholics and 35–37 for Protestants. In parts b and c, n's = 67–76 for Catholics and 37–40 for Protestants.

a. Perceptions of and feelings towards the other community, T1–T2
How open to reason?
P re C0.5600.0530.5540.158−0.0060.885
C re P0.5430.2090.6600.0000.1170.003
(P re C) – (C re P)0.0170.701−0.1060.030−0.1230.027
How trustworthy?
P re C0.6420.0000.7110.0000.0690.113
C re P0.6210.0000.7090.0000.0870.003
(P re C) – (C re P)0.0200.6350.0030.958−0.0180.732
Feeling favourably vs. unfavourably
P re C0.6540.0000.6830.0000.0290.399
C re P0.7680.0000.7510.000−0.0160.556
(P re C) – (C re P)−0.1130.014−0.0680.1430.0450.304
b. Beliefs about inter-community relations
Important to attend schools with balanced enrolment
All0.7300.0000.7350.0000.0030.897
P0.7460.0000.7510.0000.0050.910
C0.7220.0000.7270.0000.0050.875
P – C0.0250.6750.0240.6800.0000.996
Changes in the education system can equally benefit children from both communities
All0.5740.0170.6420.0000.0730.048
P0.6270.0160.6700.0010.0430.437
C0.5450.2470.6270.0040.0820.101
P – C0.0820.1960.0430.4870.0390.601
More mixing vs. more separation will improve relations
All0.7230.0000.7370.0000.0130.617
P0.7700.0000.7970.0000.0280.476
C0.6990.0000.7050.0000.0070.859
P – C0.0710.1720.0920.073−0.0210.695
Mixed education promotes mutual respect & understanding
All0.7930.0000.7910.000−0.0020.933
P0.8250.0000.7880.000−0.0380.467
C0.7760.0000.7930.0000.0160.567
P – C0.0490.263−0.0050.9210.0540.360
c. Perceptions of the other side's beliefs
Important to attend schools with balanced enrolment
P re C0.5250.5110.5160.721−0.0090.855
C re P0.4250.0320.5250.4890.1000.009
(P re C) – (C re P)0.1000.052−0.0090.868−0.1090.086
Changes in the education system can equally benefit children from both communities
P re C0.5330.2800.5410.4080.0070.881
C re P0.5530.1310.4850.698−0.0670.162
(P re C) – (C re P)−0.0190.6730.0550.3690.0750.279
More mixing vs. more separation will improve relations
P re C0.5190.6070.5060.874−0.0120.798
C re P0.5080.7930.5130.6730.0050.906
(P re C) – (C re P)0.0110.825−0.0070.891−0.0170.785

Table 4b shows that both communities also tended to endorse empirical premises consistent with peaceful accommodation, agreeing overwhelmingly that ‘mixed education promotes mutual respect and understanding’ (Catholics at 0.776, Protestants at 0.825), that it is important for children to ‘attend schools that have a balanced enrolment of Protestant and Catholic pupils’ (Catholics at 0.722, Protestants at 0.746) and that ‘better relations between Protestants and Catholics will only come about through more mixing of the two communities’ (Catholics at 0.669, Protestants at 0.770). Only the proposition that ‘changes in the Omagh area's education system can equally benefit’ both Catholic and Protestant children drew somewhat less agreement – but still more agreement than disagreement (with Catholics at 0.545, Protestants at 0.627).

As noted above, these results are consistent with the single-time snapshots of similarly general mixing-related attitudes in other surveys, as well as with the more concrete mixing-friendly T1 school policy attitudes reported in Table 2. There is some opposition towards mixing and the other side, but much less than ‘deeply divided’ suggests. This looks like division, but not one that is all that deep. Is Northern Ireland actually characterised by deep division? Yes, we believe – but among elites and activists, we suspect, much more than among ordinary citizens. More of that presently.

Although these high initial means limited the room for upward movement, several of these attitudes did become more positive. In Table 4a, Catholics came to see Protestants as much more open to reason than they had before deliberating (increasing their mean rating by 0.117, from 0.543 to 0.660), and each community came to see the other as more trustworthy than they had before deliberating (with Protestants increasing their mean rating of Catholics by 0.069, from 0.642 to 0.711, and Catholics increasing their mean rating of Protestants by 0.087, from 0.621 to 0.709). In Table 4b, the belief commanding the weakest initial agreement – that changes in the education system can equally benefit children from both communities – drew significantly more agreement, especially on the part of Catholics, who had more room to agree more.

Granted, each side saw the other's attitudes as less positive than they actually were and, with one exception, did not change those perceptions much. Table 4c shows the six mean perceptions (of the holding of three beliefs by each of the two sides) for which we have these measures. Five of the six start insignificantly different from neutrality, and the sixth (of Catholics' perceptions of the importance Protestants place on children's attending a school with balanced enrolment) starts below it. At least, that was the one significant change. After deliberating, Catholics saw Protestants as much more interested in balanced enrolments than they did initially – just as much in favour as the Protestants saw them as being.

Discussion

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information

Mass deliberation in deeply divided societies has often been dismissed as impossible or undesirable. As Brendan O'Leary (2005, p. 10) puts it, ‘those who embrace a politics of deliberative democracy as the prescription for conflict need reminding that deliberation takes place in languages, dialects, accents and ethnically toned voices and that it is not possible to create “ideal speech situations” ’. Even mundane policy issues may be inextricably related to the larger struggle for the state, leading even those who ‘express tolerant preferences’ to ‘practice suspicion’ (McGarry and O'Leary, 2009, p. 68). Thus ordinary people deliberating across deep divides may either fail to deliberate or deliberate but divide further.

The Omagh DP was only one deliberative forum, in only one district council area, on only one topic, but its results give grounds for believing this to be too gloomy a view. In the first place, we were able to assemble a more or less representative microcosm of Omagh, itself something of a microcosm of Northern Ireland. Once assembled, moreover, the participants did deliberate. They exchanged views. They learned about the issues. This probably did require some threshold levels of mutual trust, respect and understanding, but our T1 results, consistent with those from surveys of Northern Ireland as a whole, suggest that many ordinary Catholics and Protestants are past that threshold, holding relatively pro-mixing attitudes and neutral-to-positive perceptions of the other side.

Does that mean that Northern Ireland is not actually ‘deeply divided'? Not necessarily. It may be becoming less deeply divided, as it becomes more peaceful. But these results also suggest that the nature of ‘deep divisions’ may need reconsidering. They may not always or even usually be particularly great among ordinary citizens, focused as they tend to be on getting on with their everyday lives. It only takes a modest percentage of the population to engage in the sort of violence characterising ‘the Troubles’. Are the attitudes and perceptions held by elites or activists as conducive to deliberation (as distinct from bargaining) as those held by ordinary citizens? A good deal of work in psychology leads us to suspect not. As a rule, attitude extremity is an increasing function of interest, thought and knowledge (all traits on which elites and activists are characteristically high), provided that the perceived attributes of the attitude object (in this case, the other side and anything associated with it) are correlated, as is surely the case in Northern Ireland. (For the psychological theory and evidence, see, e.g., Judd and Lusk, 1984; Tesser et al., 1995) At the very least, our results suggest the desirability of parallel studies of ordinary citizens, activists and elites to see where, in a given society, the divisions are deeper versus shallower.

The participants' policy attitudes confirmed a serious coming to terms with the new demographic and curricular constraints. Both before and after deliberating, the participants mostly supported the proposals for school policy changes – some entailing more mixing – and there is reason to regard this support as genuine, if in some cases merely pragmatic. It was clear from the demographic and curricular constraints described above that changes were imminent, the issues of what those changes should be had received wide public discussion before the DP, and these were concrete proposals being considered by the policy makers answering questions in the plenary sessions. Granted, two more general status quo proposals (of keeping the traditional pattern of schools for ages 11–18 and 11–16 and of retaining all school types currently in the Omagh area) also drew distinctly more support than opposition, suggesting some mix of acquiescence bias, limited recognition of apparent inconsistencies, and implicit nuance. But the support for these status quo proposals sharply declined, suggesting that the high – and, by contrast, steady or increasing – support for the proposals involving change was real.

More importantly, there was also some increased support for religious mixing – implicitly in the reduced support for retaining all school types currently in the Omagh area and explicitly in the increased support for getting schools to partner with the closest neighbouring school, regardless of religious composition, if necessary to deliver the curriculum. The support for the other proposals relating to religious mixing, while not increasing, started and ended high. Deliberating did not lead Catholics and Protestants to agree (or disagree) more about these policies, though largely because their views were already (and remained) surprisingly similar.

Similarly, some of the participants' perceptions of the other community became still more positive, as did some of their beliefs regarding inter-community relations. Protestants came to see Catholics as more trustworthy than they had previously, while Catholics came to see Protestants as both more trustworthy and more open to reason than they had previously. The sample as a whole came to agree more with the proposition that changes in the educational system can benefit children from both communities equally (although the movement was concentrated among Catholics).

In context, these results are encouraging. The Omagh DP was a relatively modest intervention, consisting of just one day of organised, balanced deliberation, preceded by just a few days to a few weeks of at-home, mostly imbalanced learning and discussion set in motion by the invitation to participate. Not much, considering. Northern Ireland has only recently begun to emerge from decades of hostility, punctuated by serious violence, and its schools have arguably been one of the vehicles for reproducing the opposing cultures. On the assumptions of those dismissing the possibility of constructive mass deliberation in such settings, the Omagh DP might never have happened. Given that it happened, it could have been expected to have negligible or divisive effects on the participants' attitudes towards school policy issues and one another. In fact, however, there were some sizeable effects – and these in the direction of attitudes tending to make the society less deeply divided.

We hesitate to suggest that deliberative public fora like DPs can be the key to a more conciliatory politics but, to the extent that political elites pay them heed, they may be capable of playing a part. Results resembling those of the Omagh DP can make it easier for politicians to work together, undercutting the positions of hardliners decrying compromise as a sell-out (compare Barry, 1975 p. 505; O'Flynn, 2010). They show that civil, constructive discussion between communities is not only possible but fruitful – and an aid to mutual understanding. Even in deeply divided societies, it seems, mass deliberation, structured in this fashion, can be helpful.

Notes

Earlier versions of this article were presented at the biennial general conference of the European Consortium for Political Research, Potsdam, Germany, 9–12 September 2009, and at the biennial general conference of the Participatory and Deliberative Democracy Specialist Group of the Political Studies Association, Glasgow, UK, 23–25 June 2010. We thank Jennifer McGrady, Alice Siu and Gaurav Sood for research assistance and Tony Gallagher, Padraic Quirk and again Gaurav Sood for helpful advice and comments. We gratefully acknowledge the financial support of The Atlantic Philanthropies and the Renee B. Fisher Foundation.

  1. 1

    Or, more precisely, between two national groups, British unionists and Irish nationalists. While most unionists are Protestant, and most nationalists Catholic, the conflict is more national than religious: it is, at bottom, a border dispute about whether Northern Ireland should remain part of the United Kingdom or become part of a united Ireland. But since educational issues generally involve questions of religious ethos, and roughly half of the schools define themselves in religious terms, we use the popular shorthands ‘Protestant’ and ‘Catholic’.

  2. 2

    This last category includes special needs and Irish medium schools. See Department of Education, 2005a.

  3. 3

    The district council area includes the town of Omagh and the surrounding hinterland. The most recent census was in 2001.

  4. 4

    The use of random sampling is one of the design elements distinguishing deliberative polling from the ‘dialogue’ processes described in Hickey and Mohan (2004), in which the samples are non-randomly invited or self-selected.

  5. 5

    Addresses were randomly sampled from a postal address file. Within each household, the next-birthday method (approximating random selection) decided the parent to be interviewed.

  6. 6

    The stakeholder participants in the construction of the briefing materials and in the expert panels were limited to representatives of different school types. We suspect that these processes might have been still more difficult had they included politicians or policy makers.

  7. 7

    For greater details, see the online Appendix S1).

  8. 8

    The remaining 2.5 per cent of participants and 5.8 per cent of non-participants gave some other response or declined to answer. Clearly, many unionists and nationalists were effectively declining to answer by saying ‘neither’, a sign of the sensitivity of this question (see Boyle and Hadden, 1994, p. 33; Taylor, 2001, p. 43, on this ‘disclosure’ phenomenon). But the distribution of responses, including the ‘neithers’, was essentially the same for those interviewees who attended the event and those who did not.

  9. 9

    The most recent Northern Ireland Life and Times Survey suggests that ordinary people are becoming even more favourably disposed towards one another, as should be expected, as the conflict continues to recede (NILT, 2010).

  10. 10

    Both biases are imperfections of the interview sample, rather than the recruitment from it. Most likely, this was the result of the relatively compressed fieldwork, which limited the number and spacing of call-backs. Those needing more persistence to reach and interview – a description fitting both men and the less well educated – tended to go disproportionately un-interviewed. For more detail, see the online Appendix S1.

  11. 11

    At least in this focus on policy or electoral choices, deliberation differs importantly from ‘dialogue’ (as construed in, e.g., Hickey and Mohan, 2004).

  12. 12

    The texts for these and all other items can be found in the T2 questionnaire in Appendix S3.

  13. 13

    Meaning, as the question text specifies, Controlled, Maintained, Voluntary, Special and Irish Medium.

  14. 14

    The original scorings can be found in the questionnaire, in Appendix S3.

  15. 15

    Granted, not that many parents actually send their children to mixed schools. But it is also true that there are simply not enough mixed schools to go around (O'Flynn, 2009, pp. 277–8).

  16. 16

    This is a lower proportion than in most DPs (Luskin, Hahn and Fishkin, 2007b), but still impressive, given the context. Where one might have expected intransigence, there was sometimes change.

  17. 17

    Drawing the threshold lower, at 0.060, adds only two weaker instances of convergence (making a total of three) and none of divergence (leaving the total at two). On the idea of getting teachers to travel to nearby schools to teach subjects unavailable there, Protestants became distinctly less supportive, bringing them closer to Catholics, whose initially less supportive views did not change much. On the idea of increasing the number of formally integrated schools, Catholics and Protestants changed their views in opposite directions, each moving (insignificantly) towards where the other started.

  18. 18

    The means of other attitudinal variables related to the divide among those giving midpoint responses to these questions are generally much closer to the means among those giving negative responses than to the means among those giving positive ones. The n's are small, but the pattern suggests some veiled antipathy.

References

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information
  • Barry, B. (1975) ‘Political Accommodation and Consociational Democracy’, British Journal of Political Science, 5 (4), 477505.
  • Boyle, K. and Hadden, T. (1994) Northern Ireland: The Choice. Harmondsworth: Penguin.
  • Department of Education (2005a) Northern Ireland Schools. Available from: http://www.deni.gov.uk/index/85-schools/10-types_of_school-nischools_pg.htm [Accessed 22 March 2009].
  • Department of Education (2005b) Entitled to Succeed. Available from: http://www.deni.gov.uk/index/22-new-arrangements.htm [Accessed 22 March 2009].
  • Department of Education (2007a) 2007 Northern Ireland Schools Census. Available from: http://www.deni.gov.uk [Accessed 22 March 2009].
  • Department of Education (2007b) Schools for the Future: Funding, Strategy, Sharing: Report of the Independent Strategic Review of Education. Available from: http://www.deni.gov.uk/review_of_education.pdf [Accessed 22 March 2009].
  • Department of Education (2007c) Chief Inspector Calls for Increased Collaboration in Education. Available from: http://www.northernireland.gov.uk/news/news-de/news-de-130907-chief-inspector-calls.htm [Accessed 22 March 2009].
  • Dryzek, J. S. (2005) ‘Deliberative Democracy in Divided Societies: Alternatives to Agonism and Analgesia’, Political Theory, 33 (2), 218242.
  • Fishkin, J. S. (2009) When the People Speak: Deliberative Democracy and Public Consultation. Oxford: Oxford University Press.
  • Fishkin, J. S. and Luskin, R. C. (2005) ‘Experimenting with a Democratic Ideal: Deliberative Polling and Public Opinion’, Acta Politica, 40 (3), 284298.
  • Hickey, S. and Mohan, G. (eds) (2004) Participation: From Tyranny to Transformation? Exploring New Approaches to Participation. London: Zed Books.
  • Horowitz, D. L. (1991) A Democratic South Africa? Constitutional Engineering in a Divided Society. Berkeley CA: University of California Press.
  • Judd, C. M. and Lusk, C. M. (1984) ‘Knowledge Structure and Evaluative Judgments: Effect of Structural Variables on Judgmental Extremity’, Journal of Personality and Social Psychology, 46 (6), 11931207.
  • Karpowitz, C. and Mansbridge, J. (2005) ‘Disagreement and Consensus: The Need for Dynamic Updating in Public Deliberation’, Journal of Public Deliberation, 1 (1), 348364.
  • Lijphart, A. (1977) Democracy in Plural Societies: A Comparative Exploration. New Haven CT: Yale University Press.
  • Luskin, R. C. and Fishkin, J. S. (2002) ‘Deliberation and “Better Citizens” ’. Available from http://cdd.stanford.edu/research/papers/2002/bettercitizens.pdf [Accessed 3 December 2012].
  • Luskin, R. C., Fishkin, J. S. and Jowell, R. (2002) ‘Considered Opinions: Deliberative Polling in Britain’, British Journal of Political Science, 32 (3), 455487.
  • Luskin, R. C., Fishkin, J. S. and Iyengar, S. (2006) ‘Considered Opinions on US Foreign Policy: Face-to-Face versus Online Deliberative Polling’. Available from http://cdd.stanford.edu/research/papers/2006/foreign-policy.pdf [Accessed 22 March 2009].
  • Luskin, R. C., Fishkin, J. S. and Hahn, K. S. (2007a) ‘Consensus and Polarization in Small Group Deliberations’. Paper presented at the Annual Meeting of the American Political Science Association, Chicago, August 30–September 2.
  • Luskin, R. C., Fishkin, J. S. and Hahn, K. S. (2007b) ‘Deliberation and Net Attitude Change’. Paper presented at the biennial General Conference of the European Consortium for Political Research, Pisa, September 6–8.
  • Luskin, R. C., Sood, G., Fishkin, J. S. and Kim, N. (2009) ‘Deliberation and Learning: Evidence from the Deliberative Polls’. Paper presented at the Annual Meeting of the American Political Science Association, Toronto, September 3–6.
  • McGarry, J. and O'Leary, B. (2009) ‘Power Shared after Deaths of Thousands’, in R. Taylor (ed.), Consociational Theory: McGarry and O'Leary and the Northern Ireland Conflict. London: Routledge, pp. 1584.
  • Mouffe, C. (1999) ‘Deliberative Democracy or Agonistic Pluralism?’, Social Research, 66 (3), 745758.
  • Neisser, P. T. (2006) ‘Political Polarisation as Disagreement Failure’, Journal of Public Deliberation, 2 (1), 132.
  • NILT (2010) Northern Ireland Life and Times Survey. Available from http://www.ark.ac.uk/nilt/2010/ [Accessed 3 December 2012].
  • O'Flynn, I. (2006) Deliberative Democracy and Divided Societies. Edinburgh: Edinburgh University Press.
  • O'Flynn, I. (2007) ‘Divided Societies and Deliberative Democracy’, British Journal of Political Science, 37 (4), 731751.
  • O'Flynn, I. (2009) ‘Progressive Integration (and Accommodation, Too)’, in R. Taylor (ed.), Consociational Theory: McGarry and O'Leary and the Northern Ireland Conflict. London: Routledge, pp. 264278.
  • O'Flynn, I. (2010) ‘Deliberative Democracy, the Public Interest and the Consociational Model’, Political Studies, 58 (3), 572589.
  • O'Leary, R. (2005) ‘Debating Consociational Politics: Normative and Explanatory Arguments’, in S. Noel (ed.), From Power Sharing to Democracy: Post-conflict Institutions in Ethnically Divided Societies. Montreal: McGill-Queen's University Press, pp. 343.
  • Rawls, J. (1996) Political Liberalism, second edition. New York: Columbia University Press.
  • Reilly, B. (2001) Democracy in Divided Societies: Electoral Engineering for Conflict Management. Cambridge: Cambridge University Press.
  • Sunstein, C. R. (2009) Going to Extremes: How Like Minds Unite and Divide. Oxford: Oxford University Press.
  • Taylor, R. (2001) ‘Northern Ireland: Consociation or Social Transformation?’, in J. McGarry (ed.), Northern Ireland and the Divided World. Oxford: Oxford University Press, pp. 3753.
  • Tesser, A., Martin, L. and Mendolia, M. (1995) ‘The Impact of Thought on Attitude Extremity and Attitude–Behavior Consistency’, in R. E. Petty and J. A. Krosnick (eds), Attitude Strength: Antecedents and Consequences. Mahwah NJ: Lawrence Erlbaum, pp. 7392.

Biographies

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information
  • Robert C. Luskin is Associate Professor of Government at the University of Texas at Austin and Research Adviser at the Center for Deliberative Democracy at Stanford University. Among other past affiliations, he has also been a Chercheur Associé at Sciences Po-Paris and a Fellow of the Center for Advanced Study in the Behavioral Sciences at Stanford University. He helped design deliberative polling and his work on deliberation, political knowledge and other topics concerning mass politics and statistical analysis has appeared in the American Political Science Review, the American Journal of Political Science, the Journal of Politics and Political Analysis, among other scholarly journals. Robert C. Luskin, Department of Government, University of Texas at Austin, Austin, TX 78712, USA; email: rluskin@mail.utexas.edu

  • Ian O'Flynn is Senior Lecturer in Political Theory at Newcastle University. He has been a Visiting Scholar at Harvard University and a Summer Institute Fellow at the University of Pennsylvania. His current research focuses on questions concerning the nature and requirements of democracy in multicultural and multinational societies. He is the author of Deliberative Democracy and Divided Societies (Edinburgh, 2006) as well as articles in scholarly journals such as British Journal of Political Science and Political Studies. Ian O'Flynn, School of Geography, Politics and Sociology, Newcastle University, Newcastle NE1 7RU, UK; email: ian.o'flynn@ncl.ac.uk

  • James S. Fishkin holds the Janet M. Peck Chair in International Communication at Stanford University where he teaches Communication and Political Science. He is the author of When the People Speak: Deliberative Democracy and Public Consultation (Oxford, 2009) as well as The Voice of the People (Yale, 1995), Democracy and Deliberation (Yale, 1991) and is co-author with Bruce Ackerman of Deliberation Day (Yale, 2004). He has been a Guggenheim Fellow and a Fellow of the Center for Advanced Study in the Behavioral Sciences at Stanford. He is Director of the Center for Deliberative Democracy at Stanford University. James S. Fishkin, Department of Communication, Stanford University, Stanford, CA 94305-2050, USA; email: jfishkin@stanford.edu

  • David Russell is Deputy Director of the Northern Ireland Human Rights Commission. He is also a board member of the Northern Ireland Community Relations Council. Formerly he worked as Policy Officer for the Northern Ireland Council for Integration. He holds a PhD in Politics from the University of York. His books include (with Ian O'Flynn) Power-Sharing: New Challenges for Divided Societies (Pluto, 2005) and his articles have appeared in scholarly journals such as Ethnopolitics and European Human Rights Law Review. David Russell, NIHRC, 39 North Street, Belfast BT1 1NA, Northern Ireland; email: david.Russell@nihrc.org

Supporting Information

  1. Top of page
  2. Abstract
  3. The Issues
  4. The Omagh Deliberative Poll
  5. Representativeness
  6. Questions and Hypotheses
  7. The School Policy Issues
  8. Inter-community Beliefs and Attitudes
  9. Results
  10. Discussion
  11. References
  12. Biographies
  13. Supporting Information
FilenameFormatSizeDescription
post1005-sup-0001-si.zip77K

Appendix S1. Representativeness

Appendix S2. Belief and Attitude Change by Education and Gender

Appendix S3. The T2 Participant Questionnaire

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.