Citizen preferences about border arrangements in divided societies: Evidence from a conjoint experiment in Northern Ireland

Abstract

Border arrangements are often critical to the successful negotiation of peace settlements and the broader politics of post-conflict societies. However, developing an understanding of popular preferences about these arrangements is difficult using traditional surveys. To address this problem, we used a conjoint survey experiment to assess preferences about post-Brexit border arrangements in Northern Ireland. We mapped areas of convergence and divergence in the preferences about post-Brexit border arrangements of unionist and nationalist communities, simulated the degree of public support for politically plausible outcomes and identified the border arrangements that both communities could agree upon. In so doing, we outlined an empirical approach to understanding public preferences about border arrangements that could be used to understand the degree of support for similar institutional arrangements in other divided societies.

Keywords

Borders Northern Ireland survey experiment divided societies

Introduction

Border arrangements are among some of the most publicly recognised institutions in contemporary societies, governing both international relations and domestic politics (Diener and Hagen, 2012; Simmons, 2005). They are an essential component of the resolution of inter- and intrastate conflicts (Hensel, 2001; Holsti, 1991; Schultz, 2015) and often define coexistence of different communities in divided societies (Cederman et al., 2019). Yet, public preferences about border arrangements are rarely gauged. Even peace polls on territorial conflicts rarely focus on border issues (see for example Irwin, 2004; Kaymak et al., 2008; Shamir and Shamir, 1995). This is not only a gap in the academic literature, but more importantly for the context we consider here, a key missing element of information for policy makers and post-conflict negotiators. Elites may end up negotiating with each other in relative isolation from public opinion, and it can therefore be difficult for politicians to gauge the level of popular support for different peace or border settlement options (Lederach, 1997). The failure to ensure that there is popular support for agreements can lead to the rejection of peace settlements by referendums, as demonstrated in Colombia (2016) and Cyprus (2004), while public endorsement of settlements can help secure their stability (Darby and Mac Ginty, 2002; Guelke, 1999; McGarry and O’Leary, 2009). Because settling territorial disputes is crucial to peaceful inter- and intrastate relations (Owsiak, 2012) a fuller understanding of the preferences of affected communities is central to negotiating viable peace solutions.

Borders function as political institutions filtering and controlling the entry and exit of people and goods between territorially defined jurisdictions (Simmons, 2005; Simmons and Kenwick, 2019). As is the case with all political institutions, we can identify a range of dimensions along which they vary in ways that shape their effects on interstate relations, trade, security and cultural identity (Simmons and Kenwick, 2019). For example, border institutions differ in the intrusiveness of their inspections, the location of check points, the financial costs they impose on users and the legal jurisdiction into which they fall, which can encompass financial and security issues.

Public opinion concerning these differences in border arrangements cannot be captured accurately by standard single-item survey questions, which fail to identify the complexity of views about these institutions. In evaluating border provisions, respondents must consider the qualities of one border attribute traded-off against another. For example, decreasing the costs of crossing the border by reducing border checks may increase security concerns. In order to develop an accurate understanding of how citizens view border arrangements we must grasp how they think about these trade-offs.

The methodological innovation that we propose in this paper is to apply conjoint analysis to identify citizen preferences about borders. Conjoint analysis has been applied to a range of questions of interest to political scientists, including preferences about the attributes of political candidates (Teele et al., 2018), immigration (Hainmueller and Hopkins, 2015), welfare policy regimes (Häusermann et al., 2019) and peace settlements (Morgan-Jones et al., 2019; Tellez, 2019), among others. In conjoint analysis, respondents rank or rate two or more hypothetical choices with multiple attributes; the objective is to estimate the influence of each attribute on respondents’ choices or ratings (Hainmueller et al., 2014). The values of different attributes are randomised across respondents, enabling strong causal inferences to be drawn. This proves particularly useful in mapping community-based preferences in divided societies, where it enables explaining differences and exploring solutions acceptable to all sides. Using this approach, we offer a deeper understanding of how inter-community preferences diverge and how they could converge in a jointly supported solution.

Brexit and the border

Northern Ireland represents an ideal case to examine border issues in a divided society, as the border is central to the politics of the peace process and has been a salient aspect in the politics of Northern Ireland since the early 1900s (Rankin, 2007). The partition of Ireland in 1920 split the island into two political units: the Irish Free State, which chose to leave the United Kingdom (UK), and Northern Ireland, which remained part of it. On the one hand, Irish nationalists who wanted a united Ireland became a minority in a region of the UK dominated by unionists. Unionists, on the other hand, were committed to remaining part of the UK and feared becoming a minority if Ireland were to re-unite. Thus, the imposition of the border was central to the definition of the political identities of both communities (Coakley, 2017).

The location of the border and arrangements for controlling it were contested for the rest of the 20th century and were important to the violent conflict that emerged in the 1960s. Following the signing of the Belfast/Good Friday Agreement in 1998, the importance of the border in the politics of the region was reduced. The Agreement reduced violent conflict and established power-sharing provisions, bringing stability. In addition, the integration of the UK into the European Union’s (hereafter EU) common regulatory framework saw the reduction of border infrastructure and an increasingly free flow of goods and people. During the past 20 years, the border has become invisible, and this has facilitated trade, with cross-border trade accounting for 61% of the total volume of exchanges between Northern Ireland and the EU. A further effect of EU integration has been to increase the security of the nationalist community in Northern Ireland, as they perceive their rights as more securely protected under EU law and their free movement across the border guaranteed (Guelke, 2017).

The UK vote to leave the EU in 2016 sharply increased the importance of the border and reawakened concerns about the stability of the peace process (Guelke, 2017; Phinnemore and Hayward, 2017). This was compounded by three further conditions, placing extra pressure on the border settlement. Firstly, during the Brexit referendum campaign, the main unionist and nationalist parties took opposing positions. The largest unionist party, the Democratic Unionist Party (DUP), campaigned for the UK to leave the EU, whilst Sinn Fein, the largest nationalist party, wanted it to remain (McCann and Hainsworth, 2017). Secondly, the UK’s 2017 Westminster general election resulted in a Theresa May-led conservative minority administration supported by the DUP. This meant the UK government was dependent on one of the region’s parties to sustain itself in office. Thirdly, the UK withdrawing from the EU’s single market and customs union, strengthens the practical implications of the fact that the Northern Irish land border would now be an EU external border. This could result in the need for extensive border checks to establish that goods and people moving from one jurisdiction to another meet the relevant regulations (Hayward et al., 2017). This possibility dramatically increased the political significance of the border, given the requirement for the UK government to negotiate a withdrawal agreement with the EU in order to provide for an orderly exit.

From December 2017 until January 2020 the UK Parliament was deadlocked and unable to approve the Withdrawal Agreement that Theresa May’s government had negotiated with the EU. At the time of data collection in May and June 2018, no withdrawal agreement outlining the status of Northern Ireland had been ratified by the UK Parliament. Only after the Westminster general election of December 2019, when Boris Johnson’s Conservative government secured a substantial majority, was the parliament able to ratify a renegotiated Withdrawal Agreement Bill in January 2020.¹

Despite the salience of the border issue in the politics of Northern Ireland and these negotiations, little attention has been paid to the preferences of residents of the region, with the exception of Garry et al. (2018). In particular, no one has gauged citizen preferences surrounding the trade-offs inherent in possible changes to border institutions. More open borders might smooth the economic transaction costs of crossing borders but raise security concerns. Harder to cross borders might provide more practical and symbolic support of security and identity concerns but entail higher transaction costs. This is exactly the kind of trade-off Northern Ireland is now facing as it evaluates the consequences of Brexit. To what extent should borders between Northern Ireland and the Irish Republic and/or Northern Ireland and Great Britain be regulated? Should Northern Ireland maintain close contact with the Republic of Ireland and the EU at the risk of more legal distance from the rest of the UK? Complicating this issue is the fact that the relative anticipated cost of Brexit to the Northern Irish economy is high. The UK government estimates that between 8% and 12% of the Northern Irish GDP could be lost depending on the precise Brexit outcome (Hughes and Hayward, 2018). Would compensation for these losses shape views on acceptable border arrangements?

To explore how public opinion in the region evaluates these characteristics of border regulation, we analysed the results of a survey experiment administered to Northern Irish citizens in May and June 2018.

Experimental design and analysis

We gathered a sample of 759 respondents (age 18+)² from the Qualtrics Northern Irish Online Panel comprising the unionist and nationalist communities, as well as citizens who did not identify with either group.³ Respondents were presented with pairs of hypothetical border agreements and asked to choose one. Each agreement had five attributes⁴ mirroring the key dimensions of the future border arrangements to be agreed:

Location of border stations;

Characteristics of border checks;

Monitoring of border crossings;

Responsibility for the costs of maintaining border infrastructure;

Compensation for changes to border arrangements.

Each attribute had between two and five values, proposing alternative solutions. Table 1 reports the list of dimensions and corresponding values, and Figure 1 shows an example of paired choices. Overall, respondents saw four pairs of border settlements in separate screens and were asked to make a choice between the two options in each pair, for a total of eight potential agreements evaluated by each individual.⁵

Table 1.

Northern Irish border dimensions and values.

Location of border stations	At ports of exit from the island of Ireland
	At ports of entry to England, Wales and Scotland
	At the land border between Northern Ireland and Republic of Ireland
Characteristics of physical border checks	Border officers physically examine all goods and customs paperwork crossing border
	Pre-departure electronic customs registration of all goods crossing the border combined with remote electronic monitoring of vehicles crossing the border and random physical checks of goods at depots away from border
	Random physical checks of goods at depots away from border
	Pre-departure electronic customs registration of all goods crossing border combined with remote electronic monitoring of vehicles crossing border
	No checks on goods crossing border
Control of border crossings	Separate control and operation of border crossings by Republic of Ireland and UK governments with both sides working on their own
	Shared control and operation of border crossings by Republic of Ireland and UK governments including mixed UK/Republic of Ireland teams on both sides of the border working together
Responsible for costs of maintaining border infrastructure	Mainly UK government
	Business and individuals using the border
	Mainly government of Republic of Ireland
	Shared by governments of UK and Republic of Ireland
Compensation for changes to border arrangements	None
	Public spending in Northern Ireland increased by 5%
	Public Spending in Northern Ireland increased 10%

Figure 1.

Sample pair of border agreements.

The primary outcome of interest is the binary variable ‘border arrangement preferred’. This takes the value of 1 when respondents select the settlement and 0 otherwise. We estimated the marginal effects of the attributes’ values – coded as dummy variables – using a linear probability model following Hainmueller et al. (2014). We clustered the estimates’ standard errors by respondent to account for intra-subject correlation in Stata 15.

As we were primarily concerned with how preferences on border arrangements diverge or converge across communities, we present comparative results for unionists (331 individuals) and nationalists (242 individuals) in Figure 2. The figure gives clear evidence that location of the border is strongly defined by community identification. Across this particular dimension, preferences of nationalists and unionists were unsurprisingly divergent. Nationalists were strongly opposed to a land border, but indifferent to where a potential east–west border would be. Unionists strongly preferred a land border between the northern and southern parts of the island of Ireland (North–South border) over an east–west border separating Northern Ireland from Great Britain. Preferences about whether the border should be at ports of entry to the UK or ports of exit from the Republic were insignificantly different from each other.

Figure 2.

Effect of border arrangements on Northern Irish respondents by community.

Turning to border checks, we found that both communities ranked physical checks lowest, suggesting a strong preference for non-intrusive and time-saving forms of control. Nationalists preferred no checks over any physical or digital form of border control. Unionists were slightly more open to some form of check but generally favoured the least intrusive option: digital registration.

The two communities preferred shared over separate control. When it came to paying for the maintenance of the border infrastructure, the preferences of unionists were aligned with those of nationalists and both were insignificantly different from the baseline category. As for compensation, nationalists thought an increase in public spending was preferable to no compensation whatsoever, whilst unionists preferred compensation only in the order of a 10% increase in public spending.

All in all, the border location emerged as equally important to both communities and as the main source of divergence. Unionists were 15% more likely to support an arrangement with a land border, while nationalists were 15% more likely to reject such a scenario. However, nationalists displayed an even stronger likelihood of supporting an agreement with no checks (20%) compared with physical checks. This dimension was highly salient to unionists as well: compared with physical checks, provisions for electronic registration only increased support for an agreement containing such a feature by 12%.

In the Appendix, we report our robustness tests, including our models controlling for socio-demographic characteristics and sample weighting.

Simulation of alternative border arrangements

The picture presented above is one of relative agreement across the two communities along multiple dimensions. In other words, there is scope for bilateral support for a solution, despite the expected cross-community divergence on where the border should be located. To identify what solutions would secure support – overall and by community – we simulated different combinations of border arrangements and estimated the support they would get by community and overall. In Table 2, we present the results of simulations of a range of potential arrangements that might be considered politically realistic. This gave a feel for the practical consequences and the level of popular support when border attributes are varied.

Table 2.

Simulations of support for border arrangements.

	1	2	3	4
	No deal	North–south, mild checks	East–westmild checks	East–westintrusive checks and compensation
Location	Land border	Land border	Entry	Entry
Checks	Physical	Electronic only	Electronic only	Electronic+physical random
Control	Separate	Mixed	Mixed	Mixed
Maintain	UK	Shared	Shared	Shared
Compensation	None	None	None	10%
Overall support	42%[37%–46%]	53%[49%–58%]	54%[50%–59%]	65%[60%–69%]
Unionists	51%[44%–58%]	65%[58%–72%]	50%[43%–57%]	64%[57%–70%]
Nationalists	28%[21%–36%]	40%[32%–48%]	55%[46%–63%]	67%[59%–74%]

Simulation 1 explored the likelihood of support for a No Deal scenario in the event that a withdrawal agreement between the European Union and the UK could not be reached by 31 January 2020. In such a case, the border would be north–south, with the most intrusive form of checks (physical) control, operated separately, and the UK responsible for maintaining the border. This was the least preferred option overall (only 42% support) and by community (unionists 51%, nationalists 28%). Whilst low popularity among nationalists was to be expected, a predicted support of merely 51% among unionists suggests that a No Deal outcome would not satisfy the majority of the unionist community either. Simulation 2 explored the north–south border with less intrusive checks (electronic only), shared control and maintenance of the border and no compensation. This was supported by 53% of all citizens, but there was a dramatic community split: it received more unionist support (65%) but persuaded less than half of the nationalists (40%). Both solutions would therefore pose challenges to the peace process, as their legitimacy among nationalists would be weak.

Simulation 3 was an east–west border, with checks performed at ports of entry to mainland UK. Here, checks were electronic only; there was shared control and maintenance of the border and no compensation. Under this scenario, nationalists would be more satisfied than unionists – as there would be no barriers between the north and south of the island of Ireland – but the confidence intervals around the estimates of both communities fell below the 50% line, indicating that such a solution may not fully satisfy either or both communities.

Simulation 4 revised this scenario with the addition of compensation (+10% increase in public spending). This boosted the support for this type of arrangement, with a steep increase (+12%) in the likelihood of nationalists supporting it. Unionists would also welcome an increase in public spending and would be 14% more likely to support this arrangement if the increment were part of the package. The overall support of this scenario was 64%, indicating that including an increased public spending provision would be beneficial to reaching a shared solution. The provision of a combination of electronic and random physical checks (preferred by unionists), together with an increase in public spending in the region moved the lower confidence intervals of all the estimates safely above the 50% bar for each community as well as overall.

These simulations show the scope to design border options that would secure societal agreement and attract the overall support of both groups. Crucially, a No Deal arrangement would be very unpopular with the whole of the Northern Irish public and is likely to undermine the legitimacy of the border.

Conclusion

This study of Northern Ireland is the first to apply conjoint analysis to the question of citizens’ preferences about territorial borders. In comparison to traditional surveys, conjoint analysis provides respondents with a realistic decision-making environment that enables them to make choices across packages of options. It offers the analyst simple and concise visual maps of public support for the components and packages being discussed.⁶ With its application, we identified a number of elements relevant to the work of public representative, civil servants and negotiators. Firstly, preferences of unionist and nationalist citizens for post-Brexit border arrangements were much more convergent than was apparent at the political party elite and governmental levels during negotiations. Secondly, Northern Irish citizens were very concerned about keeping the economic cost of crossing the border low. Therefore, both unionists and nationalists were prepared to compromise on solutions with low border crossing costs. Third, this study has repercussions on the current post-Brexit border arrangements, that are likely to be similarly granular and contextual in nature. Public support will depend on how an east–west border in the Irish Sea is mitigated in practice; future research could explore how preferences change in response to the experience of these arrangements.

This case study of Northern Irish citizens’ preferences about border arrangements demonstrates the usefulness of conjoint analysis to studying public opinion in similarly divided societies. By pinpointing the relative importance of different elements of potential settlements, it enables identification of solutions that could mitigate disagreements over individual policy issues and ease reaching solutions. Securing cross-community support is vital to any form of settlement in post-conflict society. This study has shown how conjoint experiments provide rich insights into which solutions are likely to secure such support.

Footnotes

Acknowledgements

The authors gratefully acknowledge the comments and advice of Jack Bridgewater, Hannah Denecke, Erik Gahner Larsen, Conor Heaney, Timothy Hellwig, Daniel Kirkpatrick, Elizabeth King, Robert Nagel, Jane O’Mahony and Josh Townsley as well as participants in panels at the EPSA Conference 2018 and the APSA Annual Meeting 2018 in the preparation of this paper.

Correction (June 2025):

The article has been updated with correct dataverse link in the supplementary material section. For more details, please see the correction notice .

Author contributions

All authors contributed to the conceptualisation, design, interpretation and writing up of this study, EM-J and LS carried out the statistical analysis.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: the authors gratefully acknowledge the financial assistance of the University of Kent Faculty of Social Science Research Support Fund and the of the United States Institute for Peace (Grant no. 350 2261). The opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the funders.

Ethical approval

Ethical approval for this study was granted under the University of Kent’s ethics governance protocols on 24 August 2018.

ORCID iD

Edward Morgan-Jones

Supplemental materials

The supplementary files are available at http://journals.sagepub.com/doi/suppl/10.1177/2053168020929927. The replication files can be found at

Notes

References

Cederman

L-E

Rueggery

Schvitz

(2019) Redemption through rebellion: Border change, lost unity and nationalist conflict. Paper presented at the Juan March Institute, Madrid, Spain, 15 November 2019. Available at: https://ic3jm.es/wp-content/uploads/2019/11/Fractionalization_JuanMarch2019.pdf (accessed 30 January 2020).

Coakley

(2017) Resolving international border disputes: The Irish experience. Cooperation and Conflict 52(3): 377–398.

Curtis

(2020) Brexit and the Northern Ireland Border. House of Commons Library, 14 January. Available at: https://commonslibrary.parliament.uk/brexit/policy/brexit-and-the-northern-ireland-border/ (accessed 20 February 2020).

Darby

Mac Ginty

(eds) (2002) Contemporary Peacemaking: Conflict, Violence and Peace Processes. London: Palgrave Macmillan.

Diener

Hagen

(2012) Borders: A Very Short Introduction. Oxford: Oxford University Press.

Garry

McNicholl

O’Leary

, et al. (2018) Northern Ireland and the UK’s exit from the EU. What do people think? Report, Economic & Social Research Council, UK, May.

Guelke

(1999) South Africa in Transition: The Misunderstood Miracle. London: I.B. Tauris.

Guelke

(2017) Britain after Brexit: The risk to Northern Ireland. Journal of Democracy 28(1): 42–52.

Hainmueller

Hopkins

Yamamoto

(2014) Causal inference in conjoint analysis: Understanding multidimensional choices via stated preference experiments. Political Analysis 22(1): 1–30.

10.

Hainmueller

Hopkins

(2015) The hidden American immigration consensus: A conjoint analysis of attitudes toward immigrants. American Journal of Political Science 59(3): 529–548.

11.

Häusermann

Kurer

Traber

(2019) The politics of trade-offs: Studying the dynamics of welfare state reform with conjoint experiments. Comparative Political Studies 52(7): 1059–1095.

12.

Hayward

Campbell

Murphy

(2017) The Irish border as a customs frontier after Brexit. CEPS Commentary, 11 July 2017.

13.

Hensel

(2001) Contentious issues and world politics: The management of territorial claims in the Americas, 1816–1992. International Studies Quarterly 45(1): 81–109.

14.

Holsti

(1991) Peace and War: Armed Conflicts and International Order, 1648–1989 (Vol. 14). Cambridge: Cambridge University Press.

15.

House of Commons Northern Ireland Affairs Committee (2016) Northern Ireland and the EU Referendum. Available at: https://publications.parliament.uk/pa/cm201617/cmselect/cmniaf/48/48.pdf (accessed 28 May 2018).

16.

Hughes

Hayward

(2018) Brexit, Northern Ireland and Scotland: Comparing political dynamics and prospects in the two ‘Remain’ areas. Scottish Centre on European Relations. Available at: https://pure.qub.ac.uk/en/publications/brexit-northern-ireland-and-scotland-comparing-political-dynamics (accessed 5 May 2018).

17.

Irwin

(2004) Using public opinion polls to support peace processes: Practical lessons from Northern Ireland, Macedonia, Cyprus, Israel and Palestine. In: Guelke

(ed) Democracy and Ethnic Conflict. London: Palgrave Macmillan, 139–167.

18.

Kaymak

Lordos

Tocci (2008) Building Confidence in Peace: Public Opinion and the Cyprus Peace Process. Brussels: Centre for European Policy Studies.

19.

Lederach

(1997) Building Peace: Sustainable Reconciliation in Divided Societies. Washington D.C.: United States Institute of Peace.

20.

McCann

Hainsworth

(2017) Brexit and Northern Ireland: The 2016 referendum on the United Kingdom’s membership of the European Union. Irish Political Studies 32(2): 327–342.

21.

McGarry

O’Leary

(2009) Power shared after the deaths of thousands. In: Taylor

(ed) Consociational Theory: McGarry and O’Leary and the Northern Ireland Conflict. London, UK: Routledge, 333–388.

22.

Morgan-Jones

Loizides

Sudulich

, et al. (2019) Can peace settlements secure the support of citizens in post-conflict states?: Evidence from a conjoint experiment in Northern Ireland. In: American Political Science Association annual meeting, Washington D.C., United States, 29 August–1 September 2019. Washington D.C.: American Political Science Association.

23.

Morgan-Jones

Sudulich

Cochrane

, et al. (2018) What are Northern Irish citizens’ preferences about post-Brexit border arrangements. In: American Political Science Association annual meeting, Boston, United States, 30 August–2 September 2018. Washington D.C.: American Political Science Association.

24.

Owsiak

(2012) Signing up for peace: International boundary agreements, democracy, and militarized interstate conflict. International Studies Quarterly 56(1): 51–66.

25.

Phinnemore

Hayward

(2017) UK Withdrawal (‘Brexit’) and the Good Friday Agreement. European Parliament Report. Brussels, Belgium: Policy Department for Citizens’ Rights and Constitutional Affairs. Available at: https://www.europarl.europa.eu/RegData/etudes/STUD/2017/596826/IPOL_STU(2017)596826_EN.pdf (accessed 20 September 2018).

26.

Rankin

(2007) Deducing rationales and political tactics in the partitioning of Ireland, 1912–1925. Political geography 26(8): 909–933.

27.

Schultz

(2015) Borders, conflict, and trade. Annual Review of Political Science 18: 125–145.

28.

Shamir

(1995) Competing values in public opinion: A conjoint analysis. Political Behavior 17(1): 107–133.

29.

Simmons

(2005) Rules over real estate: Trade, territorial conflict, and international borders as institution. Journal of Conflict Resolution 49(6): 823–848.

30.

Simmons

Kenwick

(2019) Border orientation in a globalizing world: Concept and measurement. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3326773 (accessed 20 January 2019).

31.

Teele

Kalla

Rosenbluth

(2018) The ties that double bind: Social roles and women’s underrepresentation in politics. American Political Science Review 112(3): 525–541.

32.

Tellez

(2019) Peace agreement design and public support for peace: Evidence from Colombia. Journal of Peace Research 56(6): 827–844.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.31 MB