The partition of Ireland in 1920 replaced one embattled minority with another. Unionists, descendents of settlers from Britain, became a majority in the newly formed Northern Ireland which remained part of the United Kingdom, while a new minority of Irish nationalists was left on the 'wrong' side of the border. Misrule on the part of the one-party unionist government eventually spawned both a non-violent civil rights campaign that sought equality for nationalists within Northern Ireland and an armed campaign utilising terrorist tactics that sought to force the reunification of the island. Through the bitter 'Troubles' that ensued, British security forces contained, but could not defeat, the Irish Republican Army (IRA). Towards the end of the 1980s, a 'peace process' crystallised under the sponsorship of the British and Irish governments. This was a dual project of both ending the violence and attaining the widest possible consensus on a new mode of governance for Northern Ireland. Multi-party negotiations produced the Belfast Agreement, finalised on the appropriately redemptive date of Belfast, 10 April 1998. However, in the years that followed, a number of outstanding issues, principally the decommissioning of paramilitary weapons, perpetually destabilised the fledgling institutions and poisoned relationships, until stable power-sharing was at last established in May 2007.

What role did incentives, sanctions and conditionality play in the Northern Ireland peace process? Former Ulster Unionist Party (UUP) leader David Trimble recently offered conditionality as Northern Ireland's major lesson for other peace processes, most notably Israel-Palestine (The Guardian, 25 October 2007). He argues (in contrast to Michael Ancram's contribution to this publication on exploratory dialogue without conditions) that the international community must set clear preconditions of recognition and non-violence before negotiating with groups using terrorist tactics. Critical words are tucked away at the end of Trimble's article, however: 'some flexibility is required.' Crucially, while the Northern Ireland case is replete with examples of conditions and sanctions, consistent and unambiguous conditions and sanctions are few and far between.

At the outset, three important contextual points must be made. The first is that the defining feature of the 1990s peace process was its attempt to include the paramilitary groups. The British and Irish governments had come to the realisation that a durable accommodation in Northern Ireland was impossible without the participation of the violent extremes. Other peace efforts in the 1970s and 1980s had sought to marginalise them and had failed. The second point to note is the peculiar status of the managers of the peace process, the British and Irish governments, vis-á-vis the conflict – not quite external actors, not quite participants. Both Britain and Ireland had territorial claims on Northern Ireland, yet the Anglo-Irish Agreement of 1985 formally expressed their desire to collaborate unselfishly for peace and stability. The third point is that this short article will focus primarily on how the republican movement – the IRA and its political wing Sinn Féin – was dealt with, rather than loyalist (pro-British) paramilitaries. Very similar efforts to include loyalist paramilitaries were made (a place in negotiations, prisoner releases) but the key difference from the IRA was that their allied political parties enjoyed very little public support.

In the late 1980s and early 1990s, the British and Irish governments and the non-violent nationalist Social Democratic and Labour Party (SDLP), were each engaged in separate and clandestine contacts with the republican movement aimed at making a case for bringing the IRA campaign to an end. That case included moral arguments, electoral considerations (Sinn Féin's militant associations had put a cap on its level of support) and, most importantly, the British conveyed that there would be a place for Sinn Féin at the negotiating table in the event of a permanent ceasefire. The 'Downing Street Declaration' issued jointly by London and Dublin in December 1993 was addressed mainly to republicans, setting parameters for future talks but holding out the opportunity to participate in them under certain conditions. The key parameter was the principle of consent: that the status of Northern Ireland would continue to be determined by the wishes of a majority there was not up for negotiation. The key condition was that parties seeking entry to the talks must 'establish a commitment to exclusively peaceful methods' and must have shown 'that they abide by the democratic process.' While Sinn Féin formally rejected the declaration, the IRA ceasefire on 31 August 1994 suggested that republicans were interested in the opportunities it might open up.

There were both 'push' and 'pull' factors in why republicans were amenable to British overtures. The 'push' factor was the corrosive stalemate. As the British knew, republicans were war-weary and there were those within the leadership fully aware of the limits of militarism and eager to explore the potential of peaceful politics. The 'pull' factor was the alluring prospect of building a powerful 'pan-nationalist front' to pursue republican objectives, something impossible as long as terrorism continued. Such an alliance would comprise Sinn Féin, the SDLP, the Irish government and an increasingly nationalist-inclined United States. In the early days of the Clinton administration, powerful Irish-Americans from the political, media and corporate worlds lobbied the White House to take an interest in Ireland, arguing that the US could play a role in encouraging republicans towards peace. Hence the granting of a visa to Sinn Féin leader Gerry Adams to visit New York briefly in January 1994.

This affair neatly exemplifies incentives in peace processes in all their opportunity, risk and moral ambiguity. The granting of the visa showcased the benefits of halting violence: the potential to cultivate powerful alliances, an enlarged audience for republican concerns, prestige, and even simply freedom of movement. At the same time, the action antagonised the unionists and the British government who were outraged at what they interpreted as a conferral of legitimacy on terrorism. Certainly, the visa episode is now regarded by many as a catalyst in republicanism's move away from militarism. If this is the case, its success was in its relatively modest, symbolic nature. Apart from a temporary dip in relations with Britain, it cost little of substance to anyone. And, like similar instances during the peace process, the visa was an important 'self-esteem' boost for republicanism that compensated to some extent for the loss of morale entailed in winding up the armed struggle.

Subsequently, US involvement in the peace process became more even-handed, remaining low-key but important. The US contributed personnel (most notably talks chairman George Mitchell), opportunities for politicians to meet outside of Northern Ireland, and a platform for key speeches and events. Clinton was a constant encouragement, visiting three times and intervening by phone at crunch moments. Investment by US companies helped encourage a sense of the economic boons of peace.

A prolonged impasse followed the IRA ceasefire. The British government and unionists doubted that it was permanent. They accordingly demanded the decommissioning of weapons as a confidence-building measure. To republicans this was a new precondition, an artificial stumbling block thrown up to prevaricate on addressing the iniquities of the status quo. The issue was referred to an international commission, which recommended that some decommissioning occur voluntarily during talks, rather than prior to them as the British government desired. In the meantime, parties should sign up to six principles of non-violence (the 'Mitchell Principles') to gain entry to negotiations. But IRA patience had been exhausted and the ceasefire ended in February 1996 with a massive bomb attack in London. All-party talks went ahead without Sinn Féin in June 1996 but the governments continued to assure republicans that they could participate if violence stopped once again.

The governments had reached out to loyalists by deliberately using the proportional representation system in elections to determine the composition of talks delegations in order to ensure the inclusion of the small loyalist paramilitary-linked parties. When the ceasefire was renewed in July 1997, Sinn Féin promptly espoused the Mitchell principles and was at the talks table within weeks. While hopes for decommissioning during the negotiations remained, all sides knew this was a remote possibility. Sinn Féin was expelled from the talks in February 1998 after a number of IRA-linked murders, but it was clearly a token gesture and they returned after missing only six days of talks.

To many unionists, the softening of the British position on disarmament was pandering to the threat of violence: peace at any price. Yet the flexibility was predicated on two calculations. The first was the very real danger, evident in the breakdown of the ceasefire, of pushing republicans too hard and causing them to lose faith in their non-violent strategy. The second judgement was that republicans were locked into a process that made decommissioning inevitable in any case. Sinn Féin, rather than the IRA, was increasingly the vital organ of republicanism. Its electoral mandate was growing and it had given tacit acceptance to the consent principle and explicit acceptance to the Mitchell Principles. Sinn Féin was thought to have set a course of which the full and final removal of its violent associations was the inescapable destination.

The UUP had grudgingly accepted the shifts in the British position on disarmament. But once agreement was reached and it had gained an important means of leverage over republicans – the power to prevent the establishment of the new power-sharing Executive and Assembly – the UUP was determined to force the issue. Under the slogan 'no guns, no government,' the UUP argued that the agreement's non-violence principles required IRA disarmament prior to Sinn Féin entering government. In fact, the agreement fudged the issue of decommissioning and appeared to leave it as voluntary. Sinn Féin responded to the UUP by saying that, if it was going to happen at all, decommissioning would be on condition of substantial progress on the republican agenda – equality and police reform, demilitarisation and the setting up of the institutions. The deadlock lasted for eighteen months.

In general, in their dealings with republicans, unionists opted for strict conditions and sanctions while the British government favoured enticement. The British proceeded with prisoner releases, demilitarisation and police reform without any weapons having been given up. This was intended to prove to republicans the advantages of the democratic process, but profoundly demoralised unionists who felt that to persuade rather than compel terrorists to disarm was morally repugnant. To shore up unionist support, the British handed them a couple of significant victories, limiting police reform and decreeing, against nationalist wishes, that the UK flag should continue to be flown from public buildings.

However, UUP leader Trimble did believe that those at the helm of Sinn Féin were genuine in their commitment to steering republicanism away from violence and that it was just a matter of time before decommissioning materialised. He was also wary of how excessive compulsion tactics and preconditions could have an adverse effect on Sinn Féin's ability to manage its grassroots. Accordingly, to the disquiet of much of his party, Trimble moderated his stance by participating in a number of sequencing experiments in which the UUP agreed to share power on condition that decommissioning follow soon after. On the first two occasions, there was no movement on weapons and power-sharing collapsed. The British government supported Trimble by suspending the institutions, at once encouraging unionists and sanctioning republicans. Another penalty employed by Trimble was to block Sinn Féin ministers from attending meetings of the all-Ireland ministerial council, one of republicans' favourite aspects of the agreement. Eventually, limited acts of decommissioning did happen, which Trimble, at least, put down to his measured applications of pressure aimed at showing the IRA that unionists meant business, but carefully calibrated not to destroy the process altogether.

In November 2003, to the dismay of many peace process supporters, the UUP was eclipsed electorally by the harder-line Democratic Unionist Party (DUP), which had opposed the agreement. But the DUP's approach proved to be very similar to that of the UUP – an imperious appearance was created but the door to accommodation left slightly ajar. Just as Sinn Féin had realised in the 1990s that it had no other option than to work within the parameters of the British-Irish sponsored peace process, the DUP was aware that to continue to reject the agreement, much of which had been implemented irreversibly anyway, would not be in unionism's long-term interest. The DUP preached its conditions to Sinn Féin, yet right up virtually until the party agreed to share power with Sinn Féin in May 2007, the DUP was vague about the exact extent of change in republicanism it was demanding. When what was billed as full and final decommissioning occurred in September 2005, and Sinn Féin gave its support to the police service in early 2007, the DUP was publicly sceptical, adding more demands, but quietly accepted the actions.

It is of course difficult or even impossible to prove a causal link between an incentive or sanction and a subsequent action, especially when dealing with highly secretive organisations. For instance, the first act of IRA decommissioning in October 2001 was credited by many to the new global anti-terror climate in the wake of 9/11 rather than unionist conditionality. Nonetheless, the Northern Ireland case does seem to demonstrate the utility of such measures. Three main conclusions can be drawn. Firstly, this case study underlines the importance of understanding what motivates a party. What incentives might prove tempting? What degree of sanctioning will it take before buckling? Secret dialogue between the parties prior to public engagement helped build this knowledge. Secondly, balance must be reached between tough sanctions and conditions on the one hand and flexibility – room to back off from or modify those conditions – on the other. This is the classic conundrum of how to deal with enemies while maintaining supporters. A hard-line appearance must be created to steady the nerves of supporters and pressure opponents, but room for accommodation must be maintained. To this end, in Northern Ireland, policies and statements frequently contained heavy doses of constructive ambiguity. Thirdly, the case displays how incentives, sanctions and conditions work best when the peace process is well-managed, stable and coherent. The British and Irish governments provided these qualities, as did other expert personnel whom the governments appointed, most notably George Mitchell and Canadian General John De Chastelain who headed the international decommissioning commission after the agreement. They had a monopoly on peacemaking efforts; as the cliché went, the peace process was 'the only show in town.' When a sanction or an incentive annoyed one or other of the sides, they knew that to resist the governments' agenda outright would be futile and that their goals could only be advanced through adapting to the parameters set.