The 'international community' is increasingly focused on peacebuilding as a significant priority, yet it often has a narrow focus, giving primary attention to roles of international actors. The peace process for Bougainville, Papua New Guinea (PNG), offers a case of a peace process that might well be regarded as an extreme example of an intervention supporting a locally initiated process, largely in accordance with agendas set by local actors, one where the international intervention has combined a lightness of touch with some sensitivity and creativity on the part of those involved in coordinating the intervention.

International intervention originated, under New Zealand leadership, as one intended to support the parties rather than an imposed intervention. Australia increasingly took a leadership role, but initially continued in the mode established by New Zealand, in part because of Australian sensitivity to the suspicion with which the Bougainvillean parties regarded it, given their history of supporting the PNG government economically and militarily.

This article focuses on two aspects of international support to the peace process. Firstly, the major external actors have sought to utilise funding to create incentives for parties to support the peace process or particular aspects of it. Their use of development funds, in particular, illuminates opportunities around the creative use of development funding, as well as some unintended consequences of deploying development funds as 'peace dividends.' The existence of a range of problems with the use of incentives even in a 'light touch' intervention highlights the difficulties involved in the use of such measures in interventions generally.

The second significant aspect described in this article relates to the internally generated incentives that have had a major impact in the Bougainville peace process. Major difficulties often arise with implementation of two key aspects of peace agreements – namely, disarming combatants and implementing agreed constitutional changes. In Bougainville, the parties found creative ways of sequencing and linking stages of implementation of each of these aspects to provide incentives to each side to implement what they had agreed.

Bougainville (population approx. 200,000) is an island east of mainland PNG, and was the site of a violent secessionist conflict that took place from 1988 to 1997, before a peace process led to the Bougainville Peace Agreement in August 2001. The secessionist Bougainville Revolutionary Army (BRA) engaged the PNG Police and Defence Force (PNGDF) in a guerrilla struggle. The conflict was precipitated largely by disagreements over distribution of revenues from a giant open cut copper and gold mine that operated from 1972 until its indefinite closure in 1989, which caused major fiscal problems for PNG.

When PNG forces withdrew from Bougainville in 1990, the leader of the BRA, Francis Ona, made a unilateral declaration of independence (never recognised by any other country). Local differences amongst Bougainvilleans contributed to the development of armed opposition to the BRA by groups later known as the Bougainville Resistance Forces (BRF). PNG forces soon began returning to various parts of Bougainville, usually at the request of groups threatened by localised conflict. The two main dimensions of conflict (PNGDF versus BRA, and intra-Bougainvillean) caused or contributed to several thousand deaths and massive levels of destruction of infrastructure and economic activity.

A peace process began in mid-1997 on the initiative of the opposing Bougainville factions. Although the BRA was in the military ascendancy, a majority of its leadership recognised that complete victory would probably take some years and be so divisive as to render it nugatory. There was strong pressure from the Bougainville parties, in particular, for substantial roles for the international community to facilitate and support the peace process. Parts of the PNG government were initially reluctant, seeing this as 'internationalisation' of the conflict, giving encouragement to secessionist aspirations and challenging PNG sovereignty. But a combination of strong support on all sides for doing whatever was needed to end the conflict, pressure from the BRA, and the election of a new PNG national government in mid-1997 contributed to agreement between the parties by early 1998 on the main elements of international intervention.

The peace process involved three main phases, each requiring different forms of support.

First phase

In the first phase, from mid-1997, the focus was on establishing the process and its institutional architecture, and building trust and confidence amongst the parties to a point where they could negotiate a settlement. The main forms of external support were facilitation, mediating divisive issues, providing security to enable negotiations, providing technical support to assist the parties to prepare for negotiations, and continuing the humanitarian and reconstruction assistance begun during the conflict period.

The first negotiations in New Zealand involved just the Bougainville factions with facilitation and some mediation from the hosts and some support from Australia. The Bougainville factions agreed to seek peaceful resolution of the conflict and called for a neutral peacekeeping force under UN auspices. Talks between PNG and the Bougainville factions began in New Zealand in October 1997 and produced the Burnham Truce, which would be monitored by the unarmed New Zealand-led regional Truce Monitoring Group (TMG). A key feature established with the first negotiations and continued thereafter was inclusion in BRA/BRF representation in the negotiations of local leadership of the community-based fighting units. This meant large numbers of people attending – almost 100 Bougainvilleans went to the first Burnham talks. New Zealand and Australia at times attempted to limit numbers to reduce costs and logistical pressures, but were persuaded by the Bougainville leaders that in Bougainville's political and cultural context inclusivity was vital.

A 'road map' for steps towards a negotiated process was set out in the Lincoln Agreement, signed in New Zealand in January 1998. The main steps included: negotiating an 'irrevocable' ceasefire agreement; transforming the TMG into the Peace Monitoring Group (PMG), reporting to the parties through the Peace Process Consultative Committee (PPCC) chaired by a small UN observer mission and involving the PMG; establishing a 'reconciliation government' to unify the Bougainvillean factions in advance of negotiating a political settlement; and establishing the UN Political Office in Bougainville (UNPOB) with a mandate to monitor the peace process. In general the 'road map' was adhered to, although difficulties in establishing the reconciliation government contributed to splits in Bougainvillean groups and to a 12-month delay in beginning political negotiations.

Second phase

The second phase – negotiating a political settlement – began in June 1999 and continued until the signing of the Bougainville Peace Agreement (BPA) on 30 August 2001. Most aspects of the agreement were the result of two stages of negotiation: first between the opposing Bougainville factions that needed to reach compromise positions; then between the combined Bougainvillean groups and PNG. The agreement comprised three main elements: a constitutionally guaranteed referendum on Bougainville's independence, deferred for 10 to 15 years; a constitutionally guaranteed high level of autonomy for Bougainville; and the withdrawal of the PNGDF and police 'riot squads' and a multi-stage process for disarming of Bougainvillean combatants (referred to as the disposal of weapons).

During this second phase, the need for a buffer between parties and for creation of a secure environment reduced. Facilitation of talks now involved mainly negotiations occurring in Bougainville or elsewhere in PNG. There continued to be a focus on inclusion in such talks of numerous locally based BRA and BRF representatives. A number of extended negotiating sessions were held in the PNG capital, Port Moresby, entailing considerable expense. The only talks held outside PNG involved a major meeting on weapons disposal held in Townsville, Australia, in February 2001. Some mediation continued, by the UNPOB and Australia in relation to critical issues arising in negotiating the peace agreement. Funding continued in support of provision of technical advisers to the parties. Humanitarian assistance reduced as a donor priority as conditions slowly returned towards 'normality,' with aid increasingly directed to reconstruction and to restoration of economic activity, notably cocoa production.

Third phase

The third phase – implementing the political settlement – has proceeded more slowly than expected, but has involved a high level of cooperation between Bougainville groups and the PNG government. Constitutional laws to implement the agreement were developed jointly and then passed by the PNG parliament early in 2002. PNG forces withdrew from Bougainville from 2001 and the weapons disposal process was implemented in Bougainville with a reasonably high level of success, ending in 2005. The agreed autonomy arrangements were set in place through, first, a participatory process for the making of a constitution for an Autonomous Bougainville Government (ABG), and then the conduct of elections for the ABG, which began operating in June 2005.

The third phase has seen reduced concern about a secure environment in Bougainville, enabling first reduction in size of the PMG and its eventual withdrawal in July 2003, with a small Bougainville Transitional Team then provided by Australia for six months to help manage consequential uncertainty of some Bougainville parties. Assistance with implementation included technical advice on the constitutional laws and elections, support for the weapons disposal through UNPOB and PMG contributions to secure containment of weapons, and funding to provide incentives for disposing of weapons through 'projects' for communities where disposal occurred. Development aid continued, and from about 2003, building government capacity in Bougainville became a focus for the Australian Agency for International Development (AusAID).

While there has been considerable progress, there has not been a simple transition from conflict to absolute peace. Various tensions, divisions and difficulties have deepened or emerged during the peace process. There have sometimes been tensions between the main parties to the peace process and significant divisions also developed at various points amongst some of the Bougainvillean groups supporting the peace process. With difficulty, these differences were for the most part managed, and compromise positions emerged on most issues. There have also been significant dissident groups refusing to support the peace process, notably former BRA elements that followed original BRA leader Francis Ona, who until his death in July 2005 opposed but did not actively undermine the peace process.

In general, the diverse aspects of the intervention developed at the request of the parties to the process. Requests initiating various aspects of the intervention, or giving extended mandates to, say, the PMG or the UNPOB, were generally made by the parties through the negotiation process. Others involved direct requests by one party, as with those for technical support to the Bougainville Administration. At the same time, the ambassadors of Australia and New Zealand and the Director of UNPOB played significant roles in encouraging progress in the peace process. At times they applied considerable pressure to the parties to move in particular directions.

Whilst the best known roles of the Australian and New Zealand governments in support of the peace process include facilitation, monitoring and mediating divisive issues, they have also used funding to not only support the peace process, but to influence the way the process developed and create incentives for Bougainvilleans, in particular, to support the peace process.

Both AusAID and New Zealand Aid (NZAID) have funded a wide range of activities in Bougainville. AusAID's initial focus on humanitarian assistance and small-scale reconstruction gradually shifted to support for larger infrastructure projects, income generation, and building capacity in the Bougainville Administration, especially in the law and justice sector. The UNDP has also had significant roles in funding aspects of facilitation of the peace process, and the European Union has funded some significant development projects.

Some funding in support of the peace process was quite specifically targeted to encourage communities to support the process, or to encourage progress with particular aspects of the process (such as weapons disposal). It did not always achieve the solely positive impacts intended. Well-intentioned funding inputs have contributed to broader problems in Bougainville, notably the payments of allowances to take part in aspects of the process, 'peace dividends' projects, and the funds intended to encourage weapons disposal. These inputs have encouraged perceptions both of availability of an unlimited pot of funds for peace process activities and that involvement in the peace process was about money.

Facilitation interpreted as incentives

While the donors probably always made clear distinctions between, on the one hand, funding intended to facilitate the process (eg for transport to meetings, allowances for participants in meetings, paying costs of accommodation and venue hire) and on the other hand funding intended to provide incentives to participate in the peace process, that distinction was perhaps not always as clear to participants.

Significant funding for facilitation of negotiation sessions and meetings of the PPCC, for providing technical advisers to the parties in the negotiations, and for implementation of the 2001 peace agreement undoubtedly made significant contributions to progress in the process. But some facilitation funds had mixed impacts, in part because recipients interpreted the provision of funding as incentives for participation in the peace process.

Australia, New Zealand and UNPOB (through the UNDP) at various points offered allowances and other financial benefits to Bougainvilleans (and even PNG officials) attending peace process meetings and taking part in awareness activities, and also met some categories of costs associated with reconciliation ceremonies (eg awareness-raising and transport, but not for compensation payments made as part of reconciliation ceremonies). The funds were offered in good faith, largely in response to widespread lack of income. But in doing so, it quickly became the norm for many people to show reluctance to take part in peace process related activities without some form of recompense.

Incentive payments and peace dividends

Early in 2001, with little progress made towards agreement on a weapons disposal process, Australia offered incentives in the form of a AUD$5 million fund to provide small income-generating projects to former combatants groups from communities in which weapons disposal was proceeding. The process of allocation and distribution of the funds proved divisive and disruptive, and there has been very little evidence of any sustainable contribution to income-earning activity.

Australian and other donor funding was also directed to provision of humanitarian assistance and development aid, usually allocated with a view to encouraging support for the peace process. In the late 1990s, donors offered what were termed 'peace dividends' – mainly small-scale projects (eg classrooms, health centre buildings and small commercial projects) to communities supporting the peace process. While most projects made important contributions to improved services, the term 'peace dividends' was an unfortunate one in the circumstances, suggesting that peace was not necessarily something to be supported because of its inherent value, but rather on the basis of receipt of financial incentives. Even as late as 2005-06, the provision of funding for such small projects as classrooms and health facilities in areas where dissidents were well-supported have been perceived to be an enticement for opponents of the peace process to join it.

Development aid for reintegration

There has been some creative use of development aid to encourage economic development, with priority for projects that create economic opportunities for former combatants. Even major infrastructure projects are directed towards such peacebuilding goals. For example a major Australian-funded road project rehabilitating over 600 kilometres of the coastal trunk road has required the international contractor to develop numerous small construction businesses along the road route, each involving former combatants as much as possible, with a view to both their reintegration into communities and the spread of economic benefits more generally, as well as maximising community support for the project.

Attempting to build state capacity

Both AusAID and NZAID have long recognised the importance of building the capacity of the weak Bougainville Administration (the administrative arm of the old provincial government and more recently the ABG). Administrative capacity had been largely destroyed during the 1990s, so by the time the peace agreement was being implemented, a key concern was that the implementation of autonomy arrangements would be limited by administrative capacity, possibly undermining support for the peace process.

There were often complaints by Bougainville leaders about insufficient consultation by AusAID and NZAID on humanitarian assistance and development funding, though this often reflected the weak capacity of the Bougainville Administration. Australia has been proactive in initiatives to build the BA's capacity, but implementation problems have sometimes contributed to tensions between AusAID, in particular, and the Bougainville government and administration.

Capacity-building support has included: providing advisers to the Bougainville Administration – particularly to 'strategically' important parts of it like the police; a project to train and fund part-time community-based police officers; funding a study and strategic plan for restructuring the administration; and AusAID's innovative effort to provide incentives for improving the administration's weak capacity, with funding directed to implementation of the agreed autonomy arrangements and encouragement of good government. Under this Governance and Implementation Fund (GIF), PGK 6 million per year goes to the Bougainville government (with some also to the PNG government) for projects and programmes selected jointly by Bougainville, AusAID and PNG. The GIF envisages increased funding and its transfer to full Bougainville control, but only if ABG financial management, planning and budgeting capacity improves.

A focus on the roles of the international intervention and the way it influenced the peace process through incentives and (to a lesser degree) conditionalities can result in a tendency to lose sight of two important and closely related factors that can be vital contributors to the success or failure of a process. The first is whether or not the parties to a peace process have what we might call locally generated incentives for engaging in the process and reaching a peace agreement. The second is whether such incentives can be creatively utilised in the peace process, and even translated into mutually reinforcing incentives that operate to improve the prospects of the peace agreement being honoured and implemented.

In the Bougainville case, there were incentives for the parties to engage in the peace process and reach a peace agreement. For PNG, they included such factors as the contribution of the conflict to internal divisions that were destabilising the country, severely adverse economic impacts, loss of morale in the security forces, and loss of international reputation in relation to human rights. For Bougainville they included fears of the long-term impacts of growing internal divisions, the deaths, injuries and trauma being suffered by so many, destruction of the economic base, and so on. Throughout the peace process, such considerations encouraged parties to support the process, even at times of extreme difficulty.

It was in large part because of the continued awareness of such considerations that the parties were able to agree to embed in the BPA their own system of incentives for its implementation. These incentives were provided through creative linkages between agreed-upon steps for implementation of key arrangements that in many other peace processes have been stumbling blocks to implementation of peace agreements. Sometimes implementation failures contribute to breakdowns in peace processes, and even to renewed conflict.

In this case the main linkages were between the provisions, on the one hand, for disposal of weapons by Bougainvillean factions and, on the other hand, the constitutionalising and implementation of the referendum, autonomy and other agreed-upon arrangements. In particular, the obligation on Bougainville's ex-combatant groups to move weapons to secure storage arose only on PNG both making the constitutional amendments implementing the agreement and withdrawing its forces from Bougainville. However, once passed by Parliament, the constitutional amendments did not come into operation until the UN mission verified completion of stage 2 of weapons disposal (secure containment). Subsequently, lack of substantial compliance with the agreed-upon weapons disposal process could have resulted in the UN mission delaying elections for the ABG (any party to the Agreement could call on the UNOMB to verify and certify substantial compliance in weapons disposal, and whether the level of security for the weapons was conducive to the holding of elections). All factions were thereby given incentives to honour obligations with which they might otherwise have been reluctant to comply.

While weapons disposal has removed a large proportion of weapons from the hands of the BRA and the BRF, the process has by no means removed all weapons from the community, and at the end of 2007 the ABG is attempting to develop a new weapons disposal process. Nevertheless, the linkages just outlined worked well in practice, and undoubtedly contributed to the reasonable degree of success achieved in weapons disposal.

There is an additional linkage yet to come into operation, but which is intended to operate to provide ongoing incentives for continued efforts in relation to weapons disposal. It involves setting the date on which the referendum is held within the period of 10 to 15 years after the ABG is established. Amongst the conditions to be taken into account when setting that date is the progress in relation to weapons disposal.

In essence, through this system of linkages and sequencing, each side agreed to implement an agreed step that was to some degree against its own interests, provided that the other side also implemented an agreed step that the first party regarded as being in its own interests. Similar links between decommissioning weapons and constitutional or electoral steps could perhaps have been helpful in the Northern Ireland peace process. But such arrangements could also be adapted to meet quite distinct goals. For example, a national government concerned that a new autonomous regime may be dominated by, say, an anti-democratic group, or may ignore human rights norms, could negotiate for specific benchmarks for implementation of democratisation or for human rights protections that must be achieved by the autonomous regime in order to trigger obligations on the national government to implement agreed steps which the autonomous regime sees as vital to its interests.

The various problems and dilemmas around international interventions described elsewhere in this publication are clearly more intense in situations where the intervention has been made largely at the initiative of international actors, and where the intervention is seeking to create the space for a peace process rather than supporting a locally-initiated process. By contrast, the Bougainville experience suggests that the focus of attention on externally developed and delivered incentives, sanctions and conditionality should not be at the expense of attention to internally generated incentives in peace processes, and the possible ways in which the latter might be integrated with the former.