History, structure and action in the settlement of complex conflicts: the Northern Ireland case

Jennifer Todd – Joseph Ruane

2014

Introduction

History is important in all conflicts, but in different ways and to different degrees. Sometimes the legacy of history is a political arena where institutions are weakly embedded, power relations unstable, and strategy and interest more immediate determinants of conflict than ‘identity’ or ‘values’. This is the case in so-called ‘warlord’ conflicts or ‘new wars’ such as in the Democratic Republic of the Congo (Kaldor, 1999). Sometimes the legacy is a complex structure open to mobilisation on different cleavages depending on elite action and choice of repertoire, and conducive to different forms of conflict rather than just one. North India, where conflicts may take religious, caste or other forms, is a case in point (from different perspectives, see Kakar, 1996; Brass, 1998; Varshney, 2002). Sometimes early events produce tightly configured, deeply embedded structural patterns, interrelated in complex ways with symbolic systems, and generative of one primary axis of conflict. Northern Ireland is such a case.

In each of these three cases, the preconditions of settlement differ. Our interest lies in the third case, where the key to settlement lies in weakening the structurally generated tendencies to conflict and breaking the linkages that reproduce them. This simple point has implications for method and for substantive explanation. It makes identifying the long-term conflictgenerating structural relations and the basis of their reproduction critical to explaining conflict and identifying paths to settlement. This is not to disregard the role of proximate causes or of action. But their impact and effectiveness will depend on the longer-term structures.

The first section of the article sets out a general historical-structural model for explaining conflict and settlement and shows how it may be operationalized. The second section traces the historical patterns of conflict in Northern Ireland and shows how change in them opened the way to settlement. The third section looks more closely at the the interrelation of action and structure in the settlement process. It assesses the implications of the analysis for present tensions and future prospects in Northern Ireland.

History, structure and action in settlement processes

The importance of historically embedded structures has long been recognised in theories of ethno-national conflict and settlement (Coakley, 2012: 198-212; Wimmer, 2002: 85-195). Comparative research has tended to focus on synchronic and short-term factors – including the role of militants, mediators, negotiators, guarantors, spoilers, institutional design, disarmament, and confidence building measures (Hartzell and Hoddie, 2007).1 However, a recent trend in the literature brings attention back to the structural conditions of conflict, in particular to the role of ‘horizontal’ inequalities between culturally-defined groups (Stewart, 2008; Brown et al, 2012; Cederman et al, 2011; Cederman et al, 2010). This is a valuable development, but the structures need to be understood historically.

Historical structures vary widely in their temporality: all are enduring, but they range from those that are centuries (even millennia) old and have become embedded and interlinked with others, to ones that are middle-term, to others that are shorter again. In much of social life, and in complex conflicts, it is the longer-term structures that determine what factors are present, how they combine, how they operate, how susceptible they are to change, and how effective particular kinds of interventions may be (Pierson, 2004: 10-16). Long-term structural relations are reproduced in a variety of ways, and sometimes systemic linkages and feedback patterns link the structures into complex configurations with path-dependent properties of their own (for discussions, see Dodgshon,1998: 118-119, 123, 142; Pierson, 2004: 27; Ruane, 2003). Examples include mutually reinforcing feedback patterns between economic and political power, class and ethnicity, structurally-defined life prospects and identity. While the configurations are highly resistant to change, altering the linkages may have transformative effect.

In this article, we distinguish social and geopolitical structures. Long-term social structures include spatial relationships, land and settlement patterns, class structures, ethnic composition, relationships of power and inequality, culturally embedded divisions, linguistic structures, and the institutionalised structures of states and world religions. These social structures organise the convergent situations, expectations and responses of a multitude of individuals who in their own actions reproduce them (Bourdieu, 1992: 52-65). Geopolitical relationships are a distinctive kind of structure, and may also show continuity over the long term, as in the enduring rivalries of major states, the internal territorial management strategies of composite states, or the adaptive strategies of peripheral regions. They differ from other social structures in being composed of a smaller number of elements (cores and peripheries, states and international organisations, alliances) and are formed and maintained with strategic ends in view. Equally, they are open to strategic revision.

There is a reciprocal relationship between geopolitics and wider social structures. The latter provide the resources (human and material) for different geopolitical strategies and generate the interests (‘national’, ‘class’ or ‘ethnic’) which these strategies are designed to serve. For example, in the classic state-building process, the periphery is subject to an expanding core and develops in a way that is both complementary and subordinate to it (Rokkan & Urwin, 1983: 1-18); in empire-building, the process is more radical, with the colony reconstructed to accord with the geo-economic and geopolitical interests of the metropolis (Day, 2008:1-10; Ferro, 1997:1-23); in world-systems, the core powers do this at a global level (Wallerstein, 1974: 347-357; Shannon, 1992: 23-43).

From a historical-structural perspective, actors operate within long-term structures and configurations which define their interests, limit their possibilities, and constrain their action. In turn, their historical understanding lets them interpret, anticipate and respond to what they experience. Behaviour is not structurally determined: actors have the capacity to reflect on their structural context, to reflexively act upon it and to re-shape it, although they do not always understand the significance of their action, its results may be very different from their intent, and its outcome may be uncertain for some considerable time. Finally, not even the most radical transformation changes everything: some structures persist and continue to set parameters for the future. The relative importance of action and structure in processes of change, the opportunities for effective action and the ways it impacts, are key questions for contemporary research (see, for example, Mahoney and Rueschemeyer, 2003; Mahoney and Thelen, 2009; Della Porta, 2013).

These considerations are of direct relevance for the study of conflict. They provide a framework for explanation of the generation and regeneration of conflict, the identification of possibilities for and constraints on settlement, and the understanding of what any particular settlement did or did not change. Historically embedded structural relations and configurations lock in oppositions of ‘ethnic’ interest and limit political possibilities of change (Ruane and Todd, 2004). Whether these configurations should be upheld or changed is itself a source of conflict, and action to change them – whether in the form of social movements, violence or negotiation – itself produces new cycles of conflict and rolling processes of settlement (Darby and MacGinty, 2008: 1). But when the systemic linkages are weakened, the results may be transformative, opening new paths for action and strengthening processes of structural change.

The historical-structural approach outlined here involves four sets of claims about complex conflicts. First, the underlying conditions of both conflict and settlement are structural and geopolitical, often interconnected in more or less tightly configured systemic relations. Second, where the underlying conditions predispose to conflict, changing them is the condition of a lasting settlement. Third, changing them is likely to require focussed action at nodal points in the structural configuration. Fourth, change at those points gives incentives for actors in the conflict region to revise their strategies and move towards settlement.

A full assessment of the historical-structural approach, testing its usefulness relative to other approaches, assessing its assumptions, demonstrating its explanatory power and showing its range of applicability over different types of case, would require a series of historicallyinformed empirical case studies. In this article we take only one case, Northern Ireland and show how the approach can be operationalised.

The Northern Ireland case is significant because of the multiplicity of factors and processes at different socio-spatial levels that impacted on conflict and on settlement (Whyte, 1991; Cox, 2006). Almost every mechanism associated with settlement in the literature can be found in this case (see, for example, White, 2013). But the presence of multiple factors does not mean they are equally important.2 The challenge is to identify the key ones and how they connect. An historical-structural analysis shows in this case: that the most important cause of conflict is the underlying structural configuration; that changes in this open the way for settlement; that this provides a better explanation of why some settlement initiatives fail and others succeed than do actor oriented explanations or structural explanations alone; and that change can come from action at nodal points in the configuration.

Northern Ireland: changing the historical patterns of conflict

Patterns of conflict
In Ireland, historically deep patterns of conflict were generated by a complex structural configuration. Its origins lie in the establishment in the seventeenth century of a deeply divisive, crisis-ridden, but stubbornly persistent, system of structural and geopolitical relationships that reproduced itself over time. It interlinked structurally embedded distinctions between ethnic and religious populations and relations of dominance and inequality between them. It was underwritten by the English/British state which depended on the locally dominant population for stable governance and administration and whose practices of territorial management were adapted to this purpose (Ruane and Todd, 1996: 16-48).

The state’s role was crucial in locking in this configuration, and it persisted through change in the state’s geopolitical interests. The English state’s interest in colonisation and explicit support for the Protestant interest from the sixteenth to the eighteenth century became in the nineteenth century a British interest in imperial power, state security and regime maintenance, and in the twentieth a set of embedded institutional habits of governance and official understandings. At each stage, the outcome was Protestant dominance and Catholic subordination. Challenge to any aspect of this configuration quickly led to challenge to the others, and it quickly provoked a response from the state and/or from loyalists claiming to defend the wider Protestant population. A changing balance of power produced renewed conflict, but only change in the configuration could unlock the conflict tendencies. Finally in 1922, the new Irish state partially dismantled this configuration in its territory, although more by changing the demographic potential for challenge than by changing the structural basis of division (Ruane, 2012).

Structural change and its impact on settlement initiatives after 1968
The historic system of geopolitical relationships was reconstituted in Northern Ireland after 1921 and it survived effectively unchallenged until 1968. When challenge came, it took multiple forms, including popular mobilisation (the Civil Rights movement, marches, communal rioting), re-organised nationalist politics (the Social Democratic and Labour Party, later Sinn Féin) and violent insurgency (the Provisional IRA), and it provoked violence in response. In the decades that followed, paramilitary and state violence cost over 3,000 lives. We are concerned with the sequence of settlement initiatives from 1968 onwards.

Table 1 shows the sequence of major settlement initiatives, the actors involved, their provisions, and their relative success or failure. The initiatives which achieved significant interparty agreement were the 1973 ‘Sunningdale’ experiment and the 1998 ‘Good Friday’ Agreement (GFA). Each involved consociational government and an institutionalised Irish dimension. The first failed, brought down by a loyalist industrial strike within five months of being set up; the second has stayed in place for 15 years, with some minor revisions agreed at St Andrews in 2006.

What explains this difference in outcome? For many commentators Sunningdale and Good Friday are essentially the same agreements, and the success of the latter came about through a lowering of expectations and an increasing realism. As Seamus Mallon of the SDLP memorably declared, the GFA is ‘Sunningdale for slow learners’. This view underestimates the differences in institutional form between the two initiatives (Wolff, 2001; McGarry and O’Leary, 2004: 1-61, 97-131; 260-293) and, as we argue below, it overstates the change on each side (. Most of all, it ignores the structural change that took place in the interim: this had radical implications for the balance of power within Northern Ireland, and made for new constraints, options and possibilities for all parties. The change had two interrelated aspects – social structural (in particular demographic and economic changes which ensured substantive equality between Protestant and Catholic) and geopolitical (British state repositioning in alliances, procedures of territorial management, and legal and constitutional frames, which placed unionist and nationalist blocs on a much more equal political footing). There was indeed a learning process, but what was being learned was the shape of the newly emerging situation and how to deal with it.

At each stage, action and structural change were involved. The crisis of 1969-1972 was, most immediately, the product of political mobilisation – the civil rights movement and the launching of a new, more effective, IRA campaign. Successful mobilisation was made possible by slow-moving structural changes within Northern Ireland and between Northern Ireland and the British state in the decades after 1921 which reduced the relative power resources and capacity of control of the unionist bloc (Ruane and Todd, 1996: 125-126). However the traditional patterns still held through the 1970s. The British state continued its long-term practices of territorial management, upholding the local structures of power, using the locally dominant actors (the Protestant community and the unionist bloc) as its administrators and security forces, and insulating conflict from the British centre. Protestants retained a marked advantage in the social structure – demographically, economically, at the higher levels of the state and in the security forces (see Table 2). Their power, position and loyalty made them both an obvious and an essential ally for the British state as violence escalated.

Nationalist mobilisation posed a challenge to this alliance that could not be ignored. As it became clear that the unionist government could or would not accommodate nationalists, in 1972 the British government took the radical step of replacing unionist rule with direct rule from London, as a first step to putting in place a new devolved settlement that would have the support of nationalists as well as unionists. What emerged was the ‘Sunningdale’ initiative of 1973-4. Within five months, it was challenged by a province-wide loyalist-led and paramilitary- enforced industrial strike that brought the economic life of the province to a standstill. As the strike gained momentum, it won the political support of the mainstream unionist population. Throughout the period of the strike the British government did not intervene to save the power-sharing institutions (Kerr, 2005: 68-69).

The British decision has been explained by short-term considerations – the government was unprepared for such an eventuality, the party in power (Labour) had not negotiated the agreement and had little commitment to it, the prime minister did not pay attention to the deteriorating situation in Northern Ireland, the army lacked the necessary skills to replace the striking power workers – but the most important reason was the British belief that it could not risk a major confrontation with the unionist community (Rees, 1985: 90 ; Craig, 2010: 178- 180). Recent research suggests that there were choices for the state at the time (see Craig, 2010: 176-180; Kerr, 2005: 68-71; Patterson and Kaufmann, 2007: 166-167). But, as in the past, this belief led the government to act in ways that underwrote and further strengthened the unionist position. It is one example of the systemic feedback patterns that have historically kept the conflict in place.

Inequality also defined the character of the Sunningdale initiative. While power sharing and an Irish dimension made it an important political break with the past, there was no corresponding attempt to change the social structures that underpinned informal Protestant power (see Table 2). The higher levels of the civil service were still heavily unionist, the security forces were massively Protestant, there were marked communal inequalities in the economy, and the trade union movement was Protestant-dominated. Crucially, the British still regarded the support, or at least acquiescence, of the unionist community as the condition of political stability. This meant that the new role of nationalists in the power-sharing executive was of limited value to the Catholic/nationalist community as a whole, since any proposal that risked undermining the unionist position would be subject to an effective unionist veto. Sunningdale would have integrated nationalists into a unionist-dominated state apparatus and, once within it, their capacity to re-structure it or the wider social and political order would have been very limited. It is a measure of the weakness of the nationalist community that this appeared to so many to be a good deal.3

The collapse of Sunningdale renewed British dependence on the Protestant community for security and administration and the prospect of a new settlement radically diminished. Unionist self-confidence recovered as did its resistance to reform and to power sharing. The SDLP was pursuing allies in the Republic, the USA and later the EU. It would not accept anything less than the package agreed at Sunningdale. Successive settlement initiatives failed (see Table 1). Meanwhile, the IRA reorganised in response to intensified security measures, and proved impossible to defeat. Catholic political opinion was hardening and in the wake of the 1981 republican hunger strikes it became increasingly sympathetic to republicanism. This opened the prospect of an even more radical Northern nationalist mobilisation, one that risked destabilising North and South (Lillis, 2010). Partly at Irish instigation the British government began what would be the single most important political development of the 1980s: a shift from its alignment with unionists to a partnership with the Irish state and to a more even-handed stance in Northern Ireland. This was a shift of historic proportions, the breaking of the centuries-old geopolitical alliance that had been a key component of the historic conflict. It took two decades and considerable difficulty to complete (Todd, 2013), but it was made easier by the fact that for the British it had little to do with identity or empathy and that political stability demanded it.

The threshold moment was the Anglo-Irish Agreement (AIA) of 1985 which gave the Irish government a formal role, ‘less than executive but more than consultative’, in Northern Ireland’s affairs and established a permanent Irish presence in the Anglo-Irish Secretariat in Maryfield, just outside of Belfast. Interpretations of the meaning and impact of the AIA differed at the time and since (Aughey and Gormley-Heenan, 2011), but unionists certainly understood its significance – that the British government had fundamentally altered its way of dealing with Northern Ireland opening, as they saw it, the floodgates for change (Todd, 2011b). As greater attention was paid to nationalist opinion and to socio-economic reform there was an evident narrowing of the gap between Protestants and Catholics across a range of socio-economic variables, in particular demography, employment and education (see Table 2).

The Downing Street Declaration of 1993 confirmed the shift in the British stance and established the terms in which a new settlement would be conceived. This made the political conjuncture of the 1990s profoundly different from that of the early 1970s. If the union would remain as long as a majority of the population so wished, there was no longer any question of integrating nationalists into a unionist-dominated state.

Equality in the Sunningdale context meant sharing power in a state that remained unionistdominated. By 1998 there were fewer unionist-dominated institutions than in the past, and where they still existed they were to be radically redesigned. The most dramatic example was in policing. The Royal Ulster Constabulary had an almost wholly Protestant membership, unionists regarded it as a symbol of the state, and its members had suffered greatly in its defence. It was radically restructured as the Police Service of Northern Ireland, with different symbols, a different ethos, and a soon-to-be dramatically different Catholic-Protestant representation (IICP, 1999). Other changes followed, including much stricter controls of marches and reform of criminal justice. If nationalists were now (as unionists and dissenting republicans liked to remind them) ‘administering British rule’, it was rule in new institutions and resting on very different geopolitical assumptions and socio-structural foundations than its Sunningdale predecessor. Acceptance of the GFA depended on a lowering of earlier expectations among unionists and republicans alike. But to see it as ‘Sunningdale for slow learners’ is to miss the historic dimension of the changes – the undoing of structural relationships that had locked in conflict for centuries.

Action and structure: parties, states and populations

Action and structure in settlement processes
Most explanations of settlement success or failure in Northern Ireland are framed in terms of action: the failures in terms of government miscalculations, spoiling tactics by republicans and by loyalists, nationalist over-ambition and unionist recalcitrance and (latterly) the successes in terms of mediation, inclusion, and guarantors. While these factors are relevant to settlement failure in 1973-4 and success in 1998, our point is that they do not explain the difference in the outcomes. Very similar actor-oriented factors existed in the two periods, but they functioned differently in the different structural contexts. We take four examples to illustrate the point.

Unionist attitudes. For some analysts, Sunningdale was doomed from the start because it faced extensive popular unionist opposition even before the Council of Ireland was agreed (see Rees 1985: 43; Patterson and Kaufmann, 2007: 161-164). In fact, close to as many unionists had serious reservations about the GFA. Just under half of unionist voters, and just less than half of elected unionist representatives, supported the power-sharing initiative in the June 1973 election: counting Alliance Party support, we can estimate that over half of Protestant voters supported it (Patterson and Kaufmann, 2007: 161-163; Dixon, 2008: 139). In the February 1974 election Unionist support for power sharing was decimated and only 19% of voters supported the power-sharing pro-union parties. But the situation in 1998 and afterwards was not very different. The GFA was supported in the referendum by just over half of Protestants; by 2003 only 28% of Protestants still supported it. The pro-agreement unionist parties won a bare majority of unionist seats in the 1998 elections (Elliott, 1999), but had lost this majority by 2001. Initial unionist support was somewhat stronger, and the fall-off of both unionist and Protestant support less sudden and dramatic than in the earlier period, but the weakness in both respects was clear.

Nationalist demands. Sunningdale, it is said, was brought down in large part by excessive nationalist demands (McGrattan, 2010: 71-88). In fact, those demands had strengthened by 1998. In 1973 the Social Democratic and Labour Party (SDLP) had accepted the principle of Irish unity only by consent and had sought power-sharing and a Council of Ireland. It did not seek further equality measures, and was willing to shelve its aims of security reform (Farren, 2010: 68-88; McLoughlin, 2010: 50-59). In 1998, it again endorsed the principle of unity only by consent and again sought power sharing and a North-South council.4 This time, it also insisted upon an Irish government role in a British-Irish Intergovernmental Conference, significant change towards equality and radical reform of policing and justice. As Brian Feeney noted, ‘the SDLP wasn’t going to sign up to working an administration in Northern Ireland unless it wasn’t Northern Ireland’.5 Moreover it was flanked by Sinn Féin, newly peaceful but even more assertive in its demands for equality and determined to work actively towards Irish unity.

Spoilers’ actions. Spoilers certainly played a role in the destruction of Sunningdale, which was opposed by unionists in the Democratic Unionist Party (DUP) and the Vanguard Unionist Party, some members of the Ulster Unionist Party (UUP), loyalist paramilitaries and republicans. Together, the unionist grouping brought down the executive through the Ulster Workers’ Council strike, its resolve strengthened by the continuing IRA campaign. The 1998 Agreement was opposed by the DUP and by members of the UUP who subsequently joined the DUP. Dissident republicans opposed it and maintained a violent campaign. The difference was not the presence of spoilers in the earlier period and their absence in the later. It lay in their capacity to spoil. In 1974 unionist spoilers were empowered by their networks in the civil service, security forces, trades unions, and – most importantly – by British inaction.

In 1998 their networks were weaker, British resolve to resist mass action was clear and in time spoilers transformed themselves into peace-makers: in 2006 at St Andrews, the DUP agreed to work a slightly amended version of the GFA.

Inclusion of the extremes. The 1998 agreement, it is said, worked because it included the extremes in negotiations and in the executive. This implies that inclusion caused the negotiations to succeed. But the causality lies elsewhere: the changed structural conditions in 1998 opened the prospect of an inclusive settlement, and allowed the inclusion of extremes whose participation in 1973 would have been opposed vehemently both by themselves and by the other parties.

These examples show that the difference in settlement outcomes between 1973-4 and 1998 had very little to do with the moderation of political opinion; still less did it represent a shift in the attitudes of the communities toward one another (see Bloomfield, 2007: 234, 40-49). If parties and populations lowered their immediate expectations, they did not change their aims, and if anything the mutual antipathies — even between the UUP and the SDLP – had increased since 1973. The key changes were in situation and strategy: the end of the republican armed struggle, the entry of Sinn Féin into the political process, and the willingness of other parties to negotiate with them, and finally to reach agreement.

Why was Sinn Féin willing to enter negotiations, reach agreement, and participate in devolved government? It is sometimes argued that Sinn Féin changed its strategy because of the (impending) defeat of the IRA.6 While a full discussion is not possible in this article, we believe that there is little evidence to bear out this claim. Nor – even were it possible – was it in British interests to push forward to such a defeat which would have ensured republican resurgence in the future. The better option from a British standpoint was to bring Sinn Féin fully into the political system in return for a comprehensive and secure de-militarisation.7 But this still leaves the question why unionists were willing to accept the entry into the political system of those they had so long tried to defeat.

As presented in the last section, what made settlement possible was the change in the longterm structural configuration. At the social-structural level, there was a major change in the structures of ‘horizontal inequality’. The reduction in the extent of demographic, socioeconomic and political inequality between Catholic and Protestant was becoming evident in the early 1990s. Table 2 shows the shifts between 1971 and 2011, and while the measures are not fully comparable, the trend is clear. It also shows that the improvement in the structural position of Catholics was slow, visible from the early 1990s particularly in demography, education, and some spheres of employment but only generalised to security, justice, and (unevenly) to public culture in the 2000s (Ruane and Todd, 2012). Socio-structural equalisation followed settlement as much as it delivered it.

This makes the crucial shift the geopolitical one, and it marked a clear departure from the past. It involved a phased process of re-positioning by the British state, changing not just its geopolitical alignment with the unionist bloc (and indirectly the Protestant community) but also the norms, entrenched habits of territorial management and understanding of sovereign boundaries that underpinned it, and thereby moving to a more neutral position. This expressed itself practically in active British intervention to create a more equal society in the face of organised unionist opposition. The presence of international guarantors, in particular US President Bill Clinton, willing to underwrite the principles of agreement and to stand against British policy if necessary, gave credibility to the repositioning (Dumbrell, 2000: 214- 222). That this was a clear breach in the historical pattern was evident to all. It had radical implications for the calculations of the parties: it meant that the partial equalisation achieved in the 1990s took on added significance, opening the prospect of much more radical structural change in the future (see Ruane and Todd, 2007).

Equalisation at the geopolitical level and the anticipation of fuller equalisation at the socialstructural level changed the options for unionists, nationalists and republicans. Unionists were intensely conscious that they had lost ground and took the opportunity of negotiating a place in government that would give them a renewed, if now shared, say over the future (Aughey, 2001). Nationalists were conscious that they were negotiating from a new position of strength and had the possibility of negotiating a power-sharing administration in which equality and an effective voice would be guaranteed, without sacrificing their ultimate objective of Irish unity. Republicans were offered an escape from an armed campaign that could not achieve its goal and a means of pursuing Irish unity through political means (McLoughlin, 2014). It was because these options opened that party strategies changed and mechanisms of negotiation and mediation could be effective in delivering a settlement. Both process and result differed from the earlier period, in the inclusion not just of republican negotiators but of nationalist and republican perspectives, and in an agreement that gave a new emphasis to equality.

Breaking patterns of conflict: the role of state elites
The actors who carried through the key changes were British state elites, although the initial moves were triggered by their Irish equivalents and later encouraged by them. Irish elites had a clear sense from the 1970s that change in British positioning was necessary to stabilise the situation in Northern Ireland and to lessen its dangers for the stability of the Irish state.8 It was only in the 1980s, spurred into action by republican mobilisation in the North, that they found a mechanism for achieving this through the insertion of an Irish institutional presence into British policy-making (Lillis, 2010). Seventy interviews and four witness seminars with members of the British and Irish political elite give insight into how they understood the process and how far they anticipated its outcome.9

The British elites differed in their understanding of the process of change, depending on their period of involvement in it. The officials who negotiated the AIA were articulate on its wider historical context and significance, which they tended to frame within a long British history of varying relations between the ‘cousins’ of Britain and Ireland (Goodall, 2010). Those involved in the negotiations of the 1990s were slower to discuss the wider historical significance of what were by any standards pattern-breaking changes – opening the political agenda, broadening repertoires of action, focussing attention and resources, challenging veto players, and ultimately including Sinn Féin in a form of devolution without precedent in British history. For the most part the British actors involved in the 1990s and 2000s discussed their decisions, plans and implementation within a pragmatic and short-term frame, speaking in terms of ‘making things better’, or ‘getting the others to agree’ or taking a ‘flexible’ attitude.10

The changes that we have described as ‘repositioning’ came about in part by a combination of prime ministerial action, and the unintended incremental consequences of that action. It is clear that British prime-ministerial action was a necessary condition of change: Mrs Thatcher’s signing of the AIA of 1985; John Major’s focussed attention to Northern Ireland after 1991; and Tony Blair’s determination to reach an agreement and to tackle veto-players on the way.11 It is equally clear that each of these choices produced consequences that spiralled far beyond the initial intent. Mrs Thatcher saw the AIA as a way to improve security cooperation with the Irish state, and the Irish side saw it as a way to change British policy and improve the conditions for nationalists in Northern Ireland.12 In the event, it did neither. Both British and Irish respondents pointed out that the Anglo-Irish Intergovernmental Conference (AIIGC) never became a decision making body.13 Issues discussed at great length in the Conference, such as policing and collusion, would – say the British respondents – have been dealt with by the same form of inquiries even had the Conference not existed.14

In fact, the AIA had a multiplicity of unintended consequences. Almost all the Irish elite pointed to 1985 as a threshold point in two respects. It was the first time that the British withstood mass unionist protest. It was also the first time that nationalist perspectives and concepts found a place in the British policy debate.15 The British respondents also point to unintended effects. Some emphasised that it opened up policy choices that until then had been kept off the agenda: ‘It put some things on the table clearly’.16 Some emphasised the manner in which seemingly interminable meetings of the Anglo-Irish Intergovernmental Conference allowed British and Irish ministers time to assess each other, and later to form cooperative partnerships.17 Most important of all, the interviews show the change that had taken place in British discourse and agenda by the 1990s. The Irish imperatives of the 1980s – reform of security, remedying inequality, taking nationalist perspectives seriously – had diffused into British official common sense in the 1990s: ‘What we were trying to do was, I think, two things. One was to manage the dimensions of the conflict downwards. A lot of that was … the way that you handled security issues but also … very important things were done about underlying social and economic issues. Trying to make it a smaller and less heated conflict. Then trying to find a political settlement to which you could attract the main players.’18

Both British and Irish elites agreed that British-Irish cooperation helped achieve settlement. They disagreed on their respective roles and the sequencing of change. Without exception, the British elite emphasised that the changes made after 1985 were a matter of British choice and decision, while acknowledging the helpfulness – on occasion – of Irish and US advice: as one senior official put it, ‘I think one of the things the Irish officials did for us was they were kind of our unpaid consultants on nationalist sensibility’.19 The Irish elite focussed on their role in maintaining pressure for change, and the British habit of ignoring such pressure. They describe cooperation beginning in the negotiations leading to the Anglo-Irish Agreement of 1985, but only at the highest levels (Cabinet Secretaries and Prime Ministers); it only slowly percolated down to lower levels.20 Even after 1998, according to Irish respondents, constructive cooperation occurred only where clear agreement had been reached; significant British-Irish disagreement continued in areas of security, policing and justice well into the 2000s.21

The interviews show that the transformation of the conditions of conflict in Northern Ireland began with small steps with different ends in view – for Mrs Thatcher, primarily security. While some members of the elite had a long strategic vision, the actual process of change in state practices was as much incremental and unintended as it was top-down and planned. So too was its impact on Northern Ireland. As relatively unanticipated prospects of peace and settlement opened in the 1990s, politicians and officials focussed upon the practicalities of getting a framework for and later agreement on settlement: larger historical considerations were put aside and the shifts in British state practices in Northern Ireland were driven more by the need to achieve and sustain a settlement than by ideology or analysis (Powell, 2008) although a focus on the consent principle was constant. A continued focus on a fuller range of principles of settlement – equality, openness of borders, consent, self-determination – was evident in most of the Irish interviews.

If the British elite did not emphasise historical transformation as much as did the Irish, it is not that they were unaware of the historical significance of their actions, or incapable of going beyond immediate pragmatism. It is rather that they consistently reframed their discussions in a repertoire of continuity and gradual change, the same discursive repertoire that frames official British discourse about sovereignty (see Meehan, 2014). The Irish elite made clear that this was far from a continuous or easy process, but one that involved very sustained political dialogue and pressure, and one that was pushed along by recurrent crises. From the perspective of the state actors, the repositioning was nuanced. But these nuanced changes within the state broke the structural configuration that underpinned conflict, changed the British stance in Northern Ireland, produced significant social structural changes there, and transformed the political arena in a way that allowed for settlement.

Current tensions and future prospects

The GFA was designed to achieve an egalitarian and stable political settlement in Northern Ireland that could in the future, with agreement, lead to national or constitutional change. It incentivised the unionist and nationalist/republican parties to change their strategies and eventually to cooperate in stable government. It took close to a decade for the provisions of the GFA to be implemented but once this was achieved it seemed to augur well for the future. The peace and stability that appeared to have been achieved were shaken by continued communal tensions that spiralled into a ‘flags’ protest in 2012-3, intensified disputes over marches, and continuing loyalist unrest (Nolan, 2013: 160-165). It is still unclear what these tensions signify for the longer term. The historical-structural approach gives insight into their underlying causes.

The 1998 Agreement was achieved by changing the geopolitical root of the historic conflict, the British role in Northern Ireland. We have described its transformative structural and political impact. But however far-reaching, the process left almost untouched the sources of division at the structural level: the legacy of inequality, the depth of the cultural oppositions and the conflicts of identity embedded institutionally and symbolically (Ruane, 2012). These aspects of structure are disaggregated; they are reproduced at the local level, even while embedded in institutions and symbolic repertoires that stretch far beyond Northern Ireland, and are much more difficult to tackle.

The GFA provisions touched on such problems. Oppositional identities were to be granted mutual respect and parity of esteem; embedded inequalities were to be tackled by strengthening the rules and procedures for ensuring equality; boundaries were to be made more permeable, less exclusive, by encouraging cross-community and reconciliatory initiatives.22 Yet neither the principles nor the practice have been effective. There were clear criteria with which to tackle gross inequalities of employment and housing provision. But there are no agreed criteria of equality in social practices like the flying of flags, marches and commemorations, and there is now major disagreement as to whether inequality still exists, what form it takes, who suffers from it, how it should be addressed, and what the Agreement implies in respect of it. Moreover these are issues touching on ‘relative group worth’ (Horowitz, 2000: 166-181) which produce strong feeling and can quickly lead to confrontation.

The problem is made more difficult by the fact that the communal power balance is far from stable and issues of equality impact on it. For example, unionists see Sinn Féin as engaged in much more than a demand for equality: this is a ‘culture war’ whose intent is to undermine the identity, morale and will of the unionist population as a first step to achieving Irish reunification. Currently it is loyalists who feel under most threat from Sinn Féin’s ‘equality agenda’ and they have a long tradition of responding with violence to erosion of their position.

For the longer term, the dangers include a further shift in the demographic balance and in the composition of the Executive, the Assembly, and the local councils. This would put more power in the hands of nationalists and allow them to extend their understanding of what equality implies into more and more areas of social life. At that point, and particularly if it seemed to open the way to unity with an unchanged Irish state, unionists and loyalists might begin to reconsider their interest in political cooperation. In such circumstances, the reemergence of loyalist violence is highly likely.

There is no simple solution to this. The need to tackle the structural underpinning of division is clear, but the parties in government cannot easily do this without affecting their traditional bases of support. The British and Irish governments – with US mediation – are containing the tensions, supporting attempts at reconciliation, and hoping that this will start incremental or generational shifts in attitude. An historical-structural approach suggests that this will be effective only if there is intervention at key nodes in the deeper structures of division: the forms of British nationalism and unity, now under pressure from Scotland; the structures of religion and education, at once centralised and impacting locally and in everyday life; and the Irish state, where religio-national divisions parallel to those in Northern Ireland, if now weaker and more compartmentalised, have persisted.

Conclusion

This article has argued that analyses of conflict and settlement should take account of the long historical patterns that generate conflict tendencies. It has shown that in the Northern Ireland case change in these explains the difference between failed and successful settlement initiatives. It motivated changes in party, popular and paramilitary strategies, and it gave political significance to the still incomplete structural changes towards equality. The historical perspective interlinks changing actor perceptions and changing social structure, showing why structural change makes possible a change of strategy, and what structural changes are likely to have such an effect.

The article focussed on the role of long-term, slow moving structures in limiting the impact of action at different social levels. It showed that popular action is of key importance in precipitating power shifts and precluding stabilisation. It showed how focussed action within the state – in part precipated by the popular movements – nudged change in entrenched geopolitical practices, promoting initially unintended but finally important structural shifts in the conflict region itself. This in turn changed actor opportunities and strategies, permitting new political institutions to be agreed.

The historical-structural approach does not replace but complements analysis of actors and accords, situating the drama of conflict and negotiation within its structural context and showing how structural constraints may themselves be transformed. Mechanisms of peacebuilding, contact, negotiation and institution-building are essential to sustain peace, as the other articles in this volume discuss, but their impact depends on the longer-term structural configuration. We argued that a further phase of action, focussed now on the structures that confirm division and opposition, is necessary to unlock the popular oppositions and tensions that may endanger the hard-won settlement. Small shifts in these structures may have radical effects, opening new paths of action and strengthening processes of conflict transformation.

References

Aughey, A. (2001) Learning from the Leopard’, in: R. Wilford (Ed.) Aspects of the Belfast Agreement, pp. 184-201 (Oxford: Oxford University Press).
Bew, J., Frampton, M. & Gurruchaga, I. (2009) Talking to Terrorists: Making peace in Northern Ireland and the Basque country (New York: Columbia University Press).
Bloomfield, K. (2007) A Tragedy of Errors: The Government and Misgovernment of Northern Ireland. (Liverpool: Liverpool University Press).
Bourdieu, P. (1990) The Logic of Practice (tr. R. Nice) (Cambridge: Polity Press)
Brass, P. R. (1998) Theft of an Idol: Text and Context in the Representation of Collective Violence (Princeton: Princeton University Press).
Brown, G., Langer, A. & Stewart, F. (Eds) (2012) Debating Affirmative Action (Houndmills, Basingstoke: Palgrave MacMillan).
Brown, M.E. (Ed.) (1996) The International Dimension of Internal Conflict (Massachusetts: MIT Press).
Cederman, L-E, Weidman, N. B., & Gleditsch, K. S. (2011) Horizontal inequalities and ethnonationalist civil war: a global comparison, American Political Science Review, 105(3), pp. 478-495.
Cederman, L-E., Wimmer, A. & Min, B. (2010) Why do ethnic groups rebel? World Politics, 62(1): 87- 119.
Coakley, J. (2012) Nationalism, Ethnicity and the State (London: Sage).
Coakley, J., and Todd, J. (2014) Breaking patterns of conflict in Northern Ireland: new perspectives, Irish Political Studies, 29 (1) xx-xx [pages to be supplied] Cox, M., Guelke, A. & Stephens, F. (Eds) (2006) A Farewell to Arms? Beyond the Good Friday Agreement. Second edition. (Manchester: Manchester University Press). Craig, A. (2010) Crisis of Confidence: Anglo-Irish Relations in the Early Troubles (Dublin: Irish Academic Press).
Darby, J. & MacGinty, R. (Eds) (2008) Contemporary Peacemaking: Conflict, Peace Processes and Post-war Reconstruction (Houndmills, Basingstoke: Palgrave Macmillan). Day, D. (2008) Conquest: How Societies Overwhelm Others (Oxford: Oxford University Press).
Dixon, P. (2008). Northern Ireland: the Politics of War and Peace. Second Edition. (Houndsmills, Basingstoke: Palgrave Macmillan).
Dodgshon, R. A. (1998) Society in Time and Space: A Geographical Perspective on Change (Cambridge: Cambridge University Press).
Dumbrell, J. (2000) ’Hope and History‘: The US and peace in Northern Ireland, in: M. Cox, A. Guelke & F. Stephens (Eds) From Long War to Long Peace in Northern Ireland, pp. 214-222 (Manchester: Manchester University Press).
Elliott, S. (1999) The referendum and assembly elections in Northern Ireland, Irish Political Studies, 14, pp. 138-149. Fair Employment Agency (1983) Report of an Investigation by the Fair Employment Agency for Northern Ireland in to the Non-Industrial Northern Ireland Civil Service (Belfast: FEA).
Farren, S. (2010) The SDLP: The Struggle for Agreement in Northern Ireland 1970-2000 (Dublin: Four Courts Press).
Farrington, C. (2006) Ulster Unionism and the Peace Process in Northern Ireland. (Houndmills, Basingstoke: Palgrave Macmillan).
Ferro, M. (1997) Colonization: A Global History (London: Routledge).
Gallagher, A. M., Osborne, R. D. & Cormack , R. J. (1995) Fair Shares: Employment, Unemployment and Economic Status. Religion and the 1991 Population Census (Belfast: Fair Employment Commission).
Goodall, D. (2010) Edging towards peace: emerging from despair in Anglo-Irish relations, Dublin Review of Books, 13 (Spring), www.drb.ie/essays/edging-towards-peace Harbison, J.J.M. & Hodges, W. J. (1991) Equal opportunities in the Northern Ireland Civil Service’, in: R. J. Cormack and R.D. Osborne (Eds) Discrimination and Public Policy in Northern Ireland, pp. 177-198 (Oxford: Oxford University Press).
Hartzell, C. & Hoddie, M. (2007) Crafting Peace: Power-Sharing Institutions and the Negotiated Settlement of Civil Wars (University Park: Pennsylvania State University Press).
Horowitz, D. H. (2000) Ethnic Groups in Conflict. Second Edition. (Berkeley: University of California Press).
Independent International Commission on Policing (IICP) (1999) A New Beginning: Policing in Northern
Ireland. The Report of the Independent Commission on Policing (Norwich: Crown Copyright).
Kakar, S. (1996) The Colors of Violence: Cultural Identities, Religion and Conflict (Chicago: Chicago University Press).
Kaldor, M. (1999) New and Old Wars: Organized Violence in a Global Age (Cambridge: Polity Press).
Kerr, M. (2006) Imposing Power-sharing: Conflict and Coexistence in Northern Ireland and Lebanon (Dublin: Irish Academic Press).
Lillis, M. (2010) Edging towards peace: Emerging from despair in Anglo-Irish relations, Dublin Review of Books 13 (Spring), www.drb.ie/essays/edging-towards-peace.
McGarry, J. & O’Leary, B. (1995) Explaining Northern Ireland: Broken Images (Oxford: Basil Blackwell).
McGarry, J. & O’Leary, B. (2004) Northern Ireland: Consociational Engagements (Oxford: Oxford University Press).
McGrattan, C. (2010) Northern Ireland: 1968-2008: The Politics of Entrenchment (Houndmills, Basingstoke: Palgrave Macmillan).
McLoughlin, P. (2010) John Hume and the Revision of Irish Nationalism (Manchester: Manchester University Press).
McLoughlin, P. (2014) ‘The first major step in the peace process’? Exploring the impact of the Anglo-Irish Agreement on republican thinking. Irish Political Studies 29 (4) xx-xx [pages to be supplied]
Mahoney, J. & Rueschemeyer, D. (Eds) (2003) Comparative Historical Analysis in the Social Sciences (Cambridge: Cambridge University Press).
Mahoney, J. & Thelen, K. (Eds) (2009) Explaining Institutional Change: Ambiguity, Agency and Power. (Cambridge: Cambridge University Press).
Meehan, E. (2014) The changing British-Irish relationship: the sovereignty dimension, Irish Political Studies, 29 (1) xx-xx [pages to be supplied]
Moloney, E. (2002) A Secret History of the IRA (London: Penguin).
Nolan, P. (2013) Northern Ireland Peace Monitoring Report. Number Two (Belfast: Community Relations Council).
Osborne, B & Shuttleworth, I. (Eds) (2004) Fair Employment in Northern Ireland: A Generation On.(Belfast: Blackstaff).
Patterson, H. & Kaufmann, E. (2007) Unionism and Orangeism in Northern Ireland: The Decline of the Loyal Family (Manchester: Manchester University Press).
Pierson, P. (2004) Politics in Time: History, Institutions, and Social Analysis (Princeton: Princeton University Press).
Powell, J. (2008) Great Hatred, Little Room: Making Peace in Northern Ireland (London: Bodley Head).
Rees, M. (1985) Northern Ireland: A Personal Perspective (London: Methuen).
Rokkan, S. & Urwin, D. (1983) Economy, Territory, Identity: Politics of West European Peripheries (London: Sage)
Rose, R. (1971) Governing without Consensus (London: Faber and Faber).
Ruane, J. (2003) Theorising the transition: longue durée and current conjuncture in centre-periphery relations in Britain, France and Spain, in: J. Ruane, J. Todd & A. Mandeville (Eds), Europe’s Old
States in the New World Order: The Politics of Transition in Britain, France and Spain, pp. 16-37 (Dublin: University College Dublin Press).
Ruane, J. (2012) Ireland’s ethno-religious conflicts: path dependence and its legacies, In: P. J. Duffy & W. Nolan (Eds) At the Anvil: Essays in Honour of William J. Smyth, pp. 631-652 (Dublin: Irish Geography Publications).
Ruane, J. and Todd, J. (1996) Dynamics of Conflict in Northern Ireland: Power, Conflict and Emancipation (Cambridge: Cambridge University Press).
Ruane, J. and Todd, J. (2004) The roots of intense ethnic conflict may not in fact be ethnic: categories, communities and path dependence, Archives Européennes de Sociologie/European Journal of Sociology, 45(2): pp. 209-232.
Ruane, J. and Todd, J. (2007) Path dependence in settlement processes: explaining settlement in Northern Ireland, Political Studies, 55(2), pp. 442-458.
Ruane, J. and Todd, J. (2012) Beyond inequality: assessing the impact of fair employment, affirmative action and equality measures on Northern Ireland, in: G. Brown, A. Langer & F. Stewart. (Eds) (2012) Debating Affirmative Action, pp. 182-208 (Houndmills, Basingstoke: Palgrave Macmillan).
Shannon, T. R. (1992) An Introduction to the World-System Perspective (second edition) (Boulder, Colorado: Westview Press).
Shirlow, P., Tonge, J., McAuley, J. & McGlynn, M. (2010) Abandoning Historical Conflict? Former Political Prisoners and Reconciliation in Northern Ireland (Manchester: Manchester University Press).
Stewart, F. (Ed.) (2008) Horizontal Inequalities and Conflict: Understanding Group Violence in Multiethnic Societies (Houndmills, Basingstoke: Palgrave Macmillan).
Todd, J. (2011a) Institutional change and conflict regulation: the Anglo-Irish Agreement (1985) and the mechanisms of change in Northern Ireland, West European Politics, 34(4), pp. 838-858.
Todd, J. (2011b) Public opinion and persuasion , in: A. Aughey and C. Gormley-Heenan (Eds) The Anglo-Irish Agreement: Rethinking its Legacy, pp. 50-63 (Manchester: Manchester University Press).
Todd, J. (2013) Thresholds of state change: changing British state institutions and practices in Northern Ireland after direct rule, Political Studies, doi: 10.1111/1467-9248.12082. online publication October.
Varshney, A. (2002) Ethnic Conflict and Civic Life: Hindus and Muslims in India (New Haven, Ct: Yale Press).
Wallerstein, I. 1974) The Modern World-System I: Capitalist Agriculture and the Origins of the European
World-Economy in the Sixteenth Century (New York: Academic Press).
White, T. J. (Ed.) (2013) Lessons from the Northern Ireland Peace Process (Madison: University of Wisconsin Press).
Wimmer, A. (2002) Nationalist Exclusion and Ethnic Conflict (Cambridge: Cambridge University Press).
Wolff, S. (2001) Context and content: Sunningdale and Belfast compared in R. Wilford (Ed.) Aspect of the Belfast Agreement, pp. 11-27 (Oxford: Oxford University Press.

Notes

Acknowledgement. We acknowledge funding from the Irish Research Council for the Humanities and Social Sciences for the research project on ‘Patterns of Conflict Resolution’, 2009-10 and the research project ‘Breaking Patterns of Conflict’, 2006-10. The interviews and witness seminars referred to in this article are the product of the ‘Breaking Patterns of Conflict’ research project in the Institute for British-Irish Studies, University College Dublin. Since these are still under embargo, quotations are unattributed; see Coakley and Todd, 2014, for a description of this collection. We thank Paul Arthur, Lorenzo Bosi, John Coakley, John McGarry, anonymous referees and participants at the European Consortium for Political Research panel on Violence and Politics, September 2013 and the Political Studies Association of Ireland panel on Breaking Patterns of Conflict, 19 October 2013.

1. Brown (1996: 22-3) pointed out that much more had been written on the structural than on the proximate causes of conflict, but since then much of the quantitative literature has focussed on the proximate causes. The same is the case for studies of settlement; see for example Hartzell and Hoddie, 2007.
2. The point is well made by McGarry and O’Leary, 1995, although we do not fully agree with their conclusion.
3. One can imagine a different possible past where British firmness in support of the executive was combined with a rolling process of reform (in security and economy) and led to changed popular and party-political options and strategies (see Bloomfield, 2007). It is a measure of the power balance that neither unionists nor the British considered such options realistic.
4. This time the Council was simpler, without a parliamentary tier but with more institutional stability, equal powerlessness, and lesser remit than the 1973-4 version.
5. Irish News, 22 August 2001.
6. The argument is made by Moloney, 2002, and implied in Bew et al, 2009. It was the view of the loyalist paramilitaries (Shirlow et al, 2010). But of the many members of the British and Irish elite whom we interviewed only one affirmed this view and many disagreed.
7. This was not, of course, unionists’ preferred path (Farrington, 2006, 120-149) but they were given little choice.
8. Witness seminar, 07 September, 2005.
9. Interviews and witness seminars were conducted as part of the ‘Breaking Patterns of Conflict’ project at UCD, Institute for British Irish Studies. For further details see the Introduction to this volume (Coakley and Todd, 2014).
10. Interviews with British politicians, 18 June 2009; 09 June, 2010; 29 September, 2010; 12 January, 2011; 18 July, 2011 and officials 21 September, 2010; 25 July 2011.
11. Witness seminar, 11 December 2006. Interviews with senior British politicians 18 June 2009; 29 September 2010, and officials 21 September 2010; 25 July 2011; Interviews with Irish officials,
15 July 2010; 16 January 2009 .
12. Witness seminar, December 11, 2006. Interview with Irish politician, 01 November 2007.
13. All of the Irish respondents working in Anglo-Irish relations at the time were disappointed at the lack of policy change. WS, December 11, 2006; Interviews with Irish officials, 3 December 2008; 8 January 2009; 23 September 2010. Senior British politicians too believed there was little attempt to make the AIIGC work, Interview 18 June, 2009.
14. Interview with British politician, 18 June 2009.
15. Interview with Irish official, 16 January 2009.
16. Interview with British politician, 07 May 2010.
17. Interview with British politician, 18 June 2009.
18. Interview with British official, 25 July, 2011.
19. Interview with British official, 21 September 2010.
20. Interviews with Irish officials, 13 December 2008; 08 January 2009.
21. Interviews with Irish officials, 19 September 2008; 15 July 2010; 23 September 2010.
22. The relevant sections of the Agreement are, respectively, (1) Declaration of Support, 3; Rights, Safeguards and Equality of Opportunity, 4; Economic, Social and Cultural Issues, 3, 5; (2) Rights, Safeguards and Equality of Opportunity, 4; Economic, Social and Cultural Issues, 2.iii; (3) Declaration of Support, 2; Rights, Safeguards and Equality of Opportunity, 13).

Source: Irish Political Studies – University College Dublin