10. Statebuilding that contained conflict but shelved specifics that fuelled conflict

Diagnosing specificities or templates for pillars?

In Chapter 9, we argued that a necessary ingredient for peace was to end gaming the state with compensation claims by replacing it with a rule of law. As always, the rule of law turned out to be a beautiful theory that ushered in some ugly practices by RAMSI and the post-conflict state. But it was a fairer rule of law than that which preceded it. When the bar has been progressively raised on some critical barriers to enduring peace—including ending the culture of gaming government compensation, the culture of corruption and harassment on the streets—perhaps we should count a peace operation as a success.

Solomon Islands is still a state riddled with corruption, but today it has an auditor-general and media that expose this publicly, an admirably multidimensional anti-corruption strategy and some of the very worst kleptocrats, even former prime ministers, have been convicted. Zero tolerance of corruption would mean no-one left standing to run the nation. Determination to prosecute the most seriously corrupt at every future point in the nation’s history, however mighty, can gradually reduce the heights of corruption to which one must jump before being sanctioned. Commitment to progressively lowering the bar to cut down the worst few cases can mean that future generations come to look back with shock at the corruption tolerated by their forebears.

Where there has not been continuous improvement is in bottom-up reconciliation and bottom-up development. A challenge for the Truth and Reconciliation Commission is to signal the need to reverse this. Certainly, there were short-term statebuilding imperatives for RAMSI centred on Honiara. But the retreat from AusAID’s previous village-up civil society development strategy should not have gone as far as it did. Primary and secondary education are in desperate need of restoration to a higher proportion of donor support for Solomon Islands. Inferior access to health centres on the Weather Coast is an injustice at the root of this conflict that can be fixed.

RAMSI support to keep combatants secure in prison could now be pruned right back. Indeed it might be time to consider the proposals of ex-MEF leaders for a Forgiveness Bill that might pardon most ex-combatants in 2012 (as discussed in Chapter 6). It might be time for Sir Peter Kenilorea’s agenda of upending policing, so that village constables become key personnel in preventing rural unrest from spreading to the capital in future (see Chapter 5).

Prosecution of foreign loggers who pay bribes to chiefs or cabinet ministers, who make sexual slaves of children in remote parts of the archipelago, who assassinate environmental activists, is overdue. Perhaps the most disappointing thing about RAMSI is that its law enforcement emphasis has not extended to international commercial interests who corrupted the Solomons tax system and its customs administration, who funded the buying of votes of no confidence in the Parliament and, worst of all, who destroyed regulation of the sustainability of the export industries—logging and fishing—that are the Solomons’ greatest hope. What a tragedy it is that Australian Prime Minister Howard missed the opportunity to insist in 2003 that independent law enforcement against those who corrupt the regulation of logging and fishing be a condition of going in to save the prime ministership, and possibly the life, of Sir Allan Kemakeza.

Sustainable trees and sustainable fish stocks are keys to the future of Solomon Islands. These are areas where the nation does not need more foreign investment, but less. Strong leaders are needed to transcend the shadow governments of foreigners to make these assets work for the villagers of all the nation’s islands. These indigenous leaders already exist; their formidable capacity can be drawn back into Solomon Islands institutions as expatriates exit.

Allen and Dinnen (2010) have discussed the ‘often neglected’ importance in the onset of political instability and the tensions of ‘the disruption to political patronage networks engendered by the combined impact, in the late 1990s, of declining demand for Solomons log exports due to the Asian financial crisis and the reform agenda of the Ulufa’alu government’ (see Hameiri 2007, 2009a). This government instability has been a key weakness of the state that has undermined effectiveness in nipping violence in the bud and has fostered profligate resort to state compensation payouts to criminals in desperate attempts to buy political survival. Political instability has led to even more desperate resort to playing the ethnic violence card in politics.

Designing an architecture of politics to render it more stable is no simple challenge. We share Jon Fraenkel’s doubts about imported electoral engineering solutions from other parts of Melanesia that have not worked well there, or have been counterproductive (Fraenkel 2004b, 2006; Fraenkel et al. 2008). Indigenous innovation is needed to craft more stable versions of Solomon Islands political culture. That will best emerge from a rich, open, creative, plural discussion of options, as opposed to deals among extant elites, which risk elite gaming and positioning to preserve privilege. The Working Committee on Political Party Integrity Reform has shown some promise, as indicated by a Lowy Institute report (Hayward-Jones 2008b) on what seemed a constructive, imaginative dialogue. In common with Fraenkel et al. (2008), that dialogue makes a good case for the dissolution of Parliament when a motion of no confidence succeeds. Because such a large proportion of sitting members loses their seat in every Solomon Islands election, members might be reluctant in no confidence motions if their success would require dissolution of Parliament. The other remedy advanced in this book is targeted prosecution of the most serious political corruption cases. Passing plain envelopes stuffed with cash to buy votes to unseat a prime minister through a vote of no confidence is clearly the kind of serious corruption that would be caught by a policy of gradually lowering the prosecutorial bar on anti-corruption enforcement. Like Fraenkel et al. (2008:9), we suspect that ‘one or two convictions of would be lobbyists offering cash in the run up to a Prime Ministerial election’ would help greatly to chill the practice.

Other peace processes have been much more successful than the Solomons’ in securing the return of refugees chased from their homes and rebuilding those homes when they have been demolished. The nettle of improving land-dispute resolution has not really been grasped. While it is encouraging to see Guadalcanal Plains Palm Oil Limited and Gold Ridge return to becoming the huge employers and exporters they once were (more painfully in the case of Gold Ridge), and while it is hoped that one day Malaitans might work in some of the jobs they were chased from, it is disappointing that for now Malaitans are unwelcome to apply for the jobs and that areas around these key economic projects are no-go zones for Malaitans. A national capital in which most of the population are squatters is a risky upshot, especially with high-violence areas such as Burns Creek where governments discourage further squatting by refusing to provide basic services such as running water. Where can they go, these Malaitans with a long, proud history of labour migration from their island, with so few employment opportunities? One response to this knotty dilemma could be a formidable opening of Australian and New Zealand labour markets to Solomon Islanders.

There is little sense in which the planners of RAMSI in Canberra in 2003 designed a mission that was attuned to helping with the specific structural and proximate factors listed in Table 8.1 as implicated in the onset of this conflict (Allen and Dinnen 2010). There is also no sufficient sense in which RAMSI attuned its 2003 strategy, or subsequently retuned it, in light of the peacebuilding weaknesses specified in Table 8.1 and Chapter 9, most of which had already been identified by the National Peace Council and subsequently listed in its 2004–09 strategic plan (NPC 2004).

This is not to deny that RAMSI did not very quickly turn around some critical risk factors: guns, unsafe streets in the capital and fiscal insolvency. And we will list in the next section some ways that RAMSI did learn to craft peace by creating spaces in which locals could diagnose specificities of reform. Some of the challenges listed in Table 8.1 were viewed by RAMSI as matters of national sovereignty. Included in this category were the challenges of reconciliation that heals indignities felt by people, fixing the dysfunctional culture of state compensation, fixing land law and land administration, fixing forest and fisheries administration, repatriating refugees and reintegrating combatants, rebalancing uneven development, constitutional reform to improve linkages of governance down to and up from villages, Gold Ridge, youth unemployment and alcohol abuse, internal migration, building health centres on the Weather Coast and a good number of others. At the same time, improving all elements of the criminal justice system, tax and customs administration and various other state functions we have discussed, in which RAMSI was highly interventionist, were somehow not in the category of matters to be left to national sovereignty.

One interpretation is that instead of diagnosing Solomon Islands’ problems RAMSI might help fix, RAMSI went in with fairly standard World Bank good governance objectives and templates for fixing them. A slightly different interpretation is that RAMSI’s planners in 2003 did have an analysis of what were the root causes of the conflict that RAMSI should seek to remedy.[1] But it was an analysis that was short on specificity. What RAMSI’s planners hoped to do was re-establish the rule of law and rebuild a weak or failed state. Yes, the Solomons was a weak state and had suffered a collapse of rule of law. But not all aspects of state performance in critical areas were weak. Few countries have had a larger malaria problem than the Solomons and few countries have fared as well in programs to reduce it. Few developing countries have courts of law in their capital as professional and independent as the Solomons’ and few have as professional and independent a central bank.

A diagnosis that correctly yet sweepingly says that a large part of the problem was a weak state does not help. It does not help because all states have their weaknesses and many of those weaknesses are irrelevant to the onset of armed conflict. A mostly strong state that has specific weaknesses, such as Saddam Hussein’s Iraq and George W. Bush’s United States, can get itself into needlessly nasty conflicts because of those weaknesses. The question is which specific state weaknesses that drive conflict are fixable. Then the need is to work with locals to help to fix them. And one might add that there is a need to diagnose which specific non-state institutions encourage violence and to work with locals to help reform them—a whole-of-society as opposed to whole-of-government approach.

RAMSI was slow and limited in learning to conduct that kind of diagnosis. In its early years, it was not responsive to the specificities of the weaknesses it found—nor did it respond to the identification of strengths by building out from them. It did not follow a responsive strategy of ‘pick problems and fix them; pick strengths and expand them’ (Braithwaite 2008:115–26). First, it had a statebuilding mind-set, so that assisting with village development, for example, was not in its sights. Its diagnosis being that a weak state was the problem, it instead set out to strengthen what it saw as the core institutions of the state: the institutions of law and order, the finance ministry, state accountability institutions and the institutions of the Public Service (including administration of the Parliament). The names of the RAMSI pillars capture this core-of-the-state ethos: ‘law and justice’, ‘economic governance’ and ‘machinery of government’. Yet leaving the Solomons with a prison administration with much improved security professionalism is a less valuable legacy than helping to leave it with an improved land administration would have been. By this we mean a land administration that solves some of the problems of land injustice that fuelled the conflict or that resolves the insecurity of land tenure that shackles economic development (Fitzpatrick and Monson 2009). ‘Registration of customary land by tribes/clans’ has been on the list of the ‘Demands by the Bona Fide and Indigenous People of Guadalcanal’ from 1988 to this day (Parliamentary Inquiry 2009:198). Malaitans have different perspectives on the land law and administration solutions needed, but agree with their adversaries that this was a root case that must be addressed. It was not part of the RAMSI methodology to go back to the Bona Fide Demands to consider which of them should be picked up as peacebuilding priorities.[2]

At one level, it was reasonable of RAMSI planners to believe that they could not fix everything and that it was best to concentrate on a finite set of challenges and tackle them well. But a more attuned, limited set of challenges might have been the specificities with the best chance of preventing recurrence. For almost none of these challenges are internationals the most effective players (one exception being deployment of soldiers). The question is where internationals can best add value to local efforts to fix the problems that matter most. That is not the question the planners of RAMSI asked. It was also not the way Prime Minister Howard and Foreign Minister Downer of Australia framed their intervention strategy.

A positive development in terms of a return to the kind of specificity of peacebuilding analysis that we found the National Peace Council to be doing before it was closed has occurred with the recent Solomon Islands Parliamentary Inquiry (2009) into RAMSI. Its report focused in quite a systematic way—on the basis of an impressive body of testimony from across the archipelago—on what it concluded were root causes of the conflict that should be addressed. The Special Coordinator of RAMSI and the high commissioners of Australia and New Zealand all testified that they did not see it as their responsibility to address the root causes (Parliamentary Inquiry 2009:205), which the inquiry accepted. The members of the inquiry agreed that it was the responsibility of the Solomon Islands Government to do so. They concluded that the first step of the way forward is for the Truth and Reconciliation Commission to recommend an approach to tackling root causes and other conflict-prevention priorities. Then, the next step is for the Solomon Islands Government to assert leadership in its response to the commission’s findings. This does indeed seem a way forward at the time of writing. Once that Solomon Islands government leadership to tackle key risks and build on key peacebuilding strengths has been grasped, one might hope for a policy shift to supporting those Solomons priorities on the part of Australia and New Zealand. Persevering with core pillars of statebuilding as donor priorities would then surely be inferior to that responsiveness.

Learning the craft of peace

RAMSI started work in 2003 with a clarity of focus on three core pillars of the state that many modern statebuilders would find admirable. RAMSI was not monolithic, however. Many within the mission lost sight of that clarity of focus. In the jargon of the field, those who lauded the initial RAMSI template of statebuilding pillars would say ‘mission creep’ set in. We suggest it was mission learning that set in among those who broke ranks with statebuilding templates and that RAMSI’s non-monolithic qualities were its saving grace. Actually, one of the things many RAMSI personnel learnt was how to constrict the mission’s involvement in certain areas and allow local actors to expand their peacebuilding ambitions.

A key example is the establishment of the Truth and Reconciliation Commission in 2009. For years this had been government policy, and even before RAMSI it had been a bottom-up priority of the church and civil society networks. Before 2009, RAMSI put considerable energy into resisting a Truth and Reconciliation Commission so as to maintain focus on building its three pillars. Australian thinking was that a Truth and Reconciliation Commission would give ‘mixed messages’ concerning the core criminal enforcement, rule-of-law focus of the mission. An officer of the Commonwealth Secretariat recalled that ‘we received repeated requests, starting around 2003 [from the Solomon Islands Government], for advice on comparative models of truth and reconciliation processes from other Commonwealth countries’. When the secretariat dropped into Canberra on the way to offer such assistance,

Canberra had expressed alarm at the idea of a formal truth commission…I was left in no uncertain mind by Canberra and the High Commissioner in Honiara, and the head of RAMSI, that I should let the request from Solomons for comparative TRC ideas just bubble along. This was considerable pressure…given Australia is an important Commonwealth country too.

By 2009, Australia and RAMSI had stopped all resistance to a Truth and Reconciliation Commission and were even offering support. Does this mean RAMSI had lost its way, drifting dangerously into mission creep? We would rather say that RAMSI had learned to listen. The learning diffused RAMSI’s focus little because locals asserted the leadership over the Truth and Reconciliation Commission model they wanted and called in most of the outside help from the UNDP and the International Centre for Transitional Justice, not from RAMSI. Other examples from the pages of our book of how RAMSI learned to be responsive to peacebuilding specificities articulated by local voices include the following.

• RAMSI withdrew from the police posts where local voices said they were least needed. It did so in a way that left behind the physical assets that local police said they needed to remain effective, also learning lessons from policing failures that RAMSI police leaders had experienced in Timor-Leste on this matter.
• RAMSI had a defensible point of view on the dangers of federalism reform as something that might fragment the cost-effectiveness and accountability of a small state, and compound opportunities for corruption. Belatedly, RAMSI leaders recognised that this might be their view, but they could not keep the lid on a federalism debate that most Solomon Islands leaders wanted. So it eased its resistance to that federalism conversation.
• RAMSI listened to the community critique that its law enforcement had been excessively focused on visible militants and insufficiently focused on shadowy ‘big fish’. Some quite big fish were put behind bars after 2006 and this critique diminished somewhat.
• RAMSI listened to the critique of Sir Peter Kenilorea, and indeed of the nation’s two most recent prime ministers, that policing was becoming too Western and too Honiara centred, supporting the development of a part-time village constable program with advice from a Bougainvillean with experience of their village auxiliary police program. Or at least some key players in RAMSI listened.
• When something positive was accomplished in policing, RAMSI public relations shifted away from publicising RAMSI’s contribution to praising the Solomon Islands Police Force in order to build community confidence and trust in their own policing institution (as opposed to public relations for RAMSI).
• AusAID staff within RAMSI have supported a renewed shift to village-based development assistance informed by village voices.
• RAMSI police—or some of them—learned that they had mismanaged the riots outside Parliament and on the streets of Honiara in 2006. They realised there were things they needed to learn from the wisdom of indigenous policing practices that had many times in the past prevented that kind of rioting from spinning out of control. They also learned lessons from formed police units in nations with wider experience of riot control than Australia and New Zealand.

How did this learning happen? RAMSI leaders could be as defensive as those of any peace operation. And many were so defensive that they learnt little from their own mistakes, blaming them on local incompetence. Yet a strength from quite early on was that RAMSI built in a commitment to evaluation modalities that would allow it to learn from its internal and external critics. The result was a culture of grudging, if sometimes slow, responsiveness to critique. This was evident at a conference organised by The Australian National University in Canberra in 2006. Mary-Louise O’Callaghan, an Australian journalist married to a Solomon Islander who had lived in the country for many years, delivered a scathing account of RAMSI’s unresponsiveness to local voices at the conference. RAMSI police leaders were furious after her presentation, asking pointed questions about who had the bright idea of inviting such an unfair critic. Within days of follow-up conversations with those police leaders and then others in the Australian Government, RAMSI had made one of its best personnel decisions: hiring O’Callaghan as a communications adviser. From that office, she helped push RAMSI into countless community relations listening projects touring the villages over the next four years.

One of the criticisms O’Callaghan made of RAMSI police that day was of inadequate training in the culture and languages of the countries to which Australian police were deployed. That situation has greatly improved since 2006, with Pacific islander trainers working at the Majura International Deployment Group headquarters of the Australian Federal Police, more apt scenario training, improved post-rotation debriefing and sharing of lessons with the next rotation of personnel. Past antagonisms and mutual misunderstandings between AusAID, police, military and non-Australian components of RAMSI have been partially remedied by extended placements of AusAID, military and foreign trainers at Majura and placements of AFP officers in AusAID and at Australia’s military peacekeeper training centre. Both high-level and low-level committee structures within RAMSI on the ground have also been adapted to foster cross-cultural learning not only between foreigners and Solomon Islanders but also to foster police responsiveness to the culture of development agencies and vice versa. Many Australian RAMSI officers do not learn from Pacific island and New Zealand contingents, but most return from RAMSI enriched by what they have learnt not only from Solomon Islanders but also from the inhabitants of other Pacific islands.

This learning has also been institutionalised at a higher level by continuous Pacific Islands Forum reviews of RAMSI.[3] Even more important have been reviews conducted by the Parliament of the Solomon Islands (especially the one in 2009), which incorporated televised engagements with citizens on all the main islands of the nation. The People’s Surveys (McMurray 2008) that were part of RAMSI’s performance indicator reviews also generated more systematic public opinion feedback on where citizens were more and less disappointed with what RAMSI was doing. The lively local mass media also pushed productively for greater RAMSI responsiveness. We hypothesise that it has been RAMSI’s learning to be more responsive that has maintained the high public support for it in those surveys.

In sum, a feedback and responsiveness culture was gradually institutionalised within RAMSI and by the Solomon Islands state and civil society. On all fronts, the most important way responsiveness took more steps forward than backwards was by creating spaces in which RAMSI leaders could leave their Honiara compound to enter dialogue in villages and towns. There were, however, certainly backward steps, such as cutting funding from the National Peace Council. The result of this dialogue is the ‘peacebuilding creep’ we are describing in this conclusion as something different from ‘mission creep’. In these spaces where dialogue was enabled, RAMSI became responsive to suggestions as to who would be best able to do what. Perhaps it was the church rather than RAMSI which should advance this agenda; perhaps it was chiefs who should be given modest resources to support that agenda rather than national or foreign elites in the capital; perhaps a Truth and Reconciliation Commission supported by the UNDP could take on another, and so on. There were also timing issues in such collaborative dialogue: ‘Last year [a particular group] was my most appropriate partner on this challenge, next year you will be the better partner; can we work together as partners on this next year?’ Rather than mission creep, this was mission creativity. In part, it allowed societal creep onto RAMSI’s agenda and far beyond, including onto some of the neglected specificities and needed diagnostics of contextual peacebuilding.

Figure 10.1 RAMSI Special Coordinator James Batley receives a gift of betel-nut from a community elder of Chief Moro’s village on the Weather Coast in August 2005 as other Solomon Islands and foreign dignitaries look on

Photo: David Jones

Responsiveness to local context and local voices was a slow, sometimes backtracking process for RAMSI. We think there might be some inevitability about this. We wonder if the reason policing of the colonial era seemed more responsive to the realities of Solomon Islands societies than policing of the RAMSI era was simply that colonial policing had been making mistakes, evolving and learning from local context for longer. We do not have historical data from the evolution of colonial policing in Solomon Islands to support this inference. We simply find it an interesting hypothesis to support our future empirical work on this question.

We do not know of examples of weak states progressing in a linear fashion to become strong states in matters of years, as opposed to decades or centuries. The durations of international peace operations are measured not in decades or centuries, sometimes not even years, but in months. So we might simply say that the experience of history at this stage of Peacebuilding Compared is that learning how to be a strong state is always a slow dawn. So we must be humble in what we expect in a few years. Some of the kinds of responsiveness on which we have lamented slowness could be challenges we might never expect a peace operation to meet. Consider responsiveness to the criticism of locking up lots of bit players while letting the ‘big fish’ off. Even the most commercially sophisticated economies that have extremely strong states—the United States, the United Kingdom, Japan and Germany—take long years to build cases for criminal prosecutions against ‘big fish’. And then their prosecutions frequently fail. Hence, it seems unrealistic to expect a struggling transitional justice system with limited technical support from donors to have this capacity during the short life of a peace operation. RAMSI was an unusual mission in lasting more than a few years; and it was only after a few years that it locked up some really big fish. Hence, even though most of the big fish about which Solomon Islanders remain concerned roam free, we might still regard RAMSI as extraordinarily successful in its responsiveness on this matter. Certainly, RAMSI achieved some prosecutorial results against major targets more quickly than the International Criminal Court has done, or could do.

The kinds of learning we have been describing are inevitably slow and perhaps would be possible only with a persistent and well-resourced intervention such as RAMSI. For a less well-resourced peace operation of shorter duration than RAMSI, the ‘slow-food’ (Boege 2006; Bowden et al. 2009) approach of the ‘light intervention’ in Bougainville (Regan 2010) has considerable advantages in its different approach to learning. Bougainville saw countless waves of patient bottom-up reconciliation built on previous waves, expanding the geographical reach of the peace and the breadth and depth of renewal across society. The architecture of the top-down peace settlement was also sequenced, with linkages that required one side to meet a commitment before the other side would deliver their next undertaking in an agreed sequence (Regan 2008; Wolfers 2006). This sequenced architecture was a slow creation of Bougainvilleans themselves, with drafting assistance from a tiny clique of foreign advisors. It only began as a successful peace process on the ashes of a number of previous peace agreements that had failed. The contribution of the international peacekeepers was a small fraction of RAMSI in numbers, duration, budget and scope of operations (see Braithwaite et al. 2010b). There was in Bougainville nothing approaching the ambition John Howard and Alexander Downer took into Solomon Islands—as revealed in our interviews—of staying until a credible state was built to replace a failed state.

The role of peacekeepers was nevertheless extremely important in enabling the slow cooking of peace by and for the people of Bougainville. It became slow food that filled their stomachs with wan bel (‘one belly’, actually meaning harmonised relationships) that they savoured and secured because they, rather than the foreign peacemakers, owned it. The peacekeepers gave confidence to combatants who were afraid to meet to reconcile before they arrived; they created safe spaces at crucial times in which the peace could mature.

We do see one choice for internationals, then, as a slow-food approach, in which peacekeepers keep guns and spoilers out of the kitchen. And another choice is a slow-learning approach in which a multidimensional peace operation gradually comes to terms with the challenges of statebuilding over many years. When the latter path is taken, our analysis internal to the Solomon Islands case supports the comparative conclusion of Howard (2008) that the responsiveness and mission learning that we call ‘peacebuilding creep’ is the key factor that turns failing aspects of the mission into more successful ones. But we also conclude RAMSI could have adapted to the contextual challenges faster had it gone in with a radically different initial framing of the mission.

Reframing for faster, fuller responsiveness

We have argued that a strength of RAMSI is that it created spaces for dialogue where it could learn to be responsive to local agenda setting. It falteringly, yet increasingly, became a learning organisation, partly because of some good early design work on continuous reviews and performance evaluation that was responsive to feedback from Solomon Islanders.

We have also concluded that RAMSI’s fundamental weakness was that at first it framed the intervention with the wrong question. This was: how do we build the fundamental pillars of a state that has failed? Alexander Downer framed it even more counterproductively in our interview as staying until the intervention had ‘re-engineered’ the Solomon Islands state. Our argument has been that there were many dimensions to this error. We also note as an aside that it follows from our analysis that it is possible for a mission to be motivated by an even more counterproductive framing question than the one RAMSI had—for example: ‘How do we achieve regime change?’ The most important two errors of the RAMSI framing were, first, that it was blind to the strengths of the Solomon Islands state and society as manifest in village institutions, churches, women’s networks, sporting organisations and much more that delivered human security to most villages even at the height of the conflict. Second, it did not take as a starting point a diagnosis of the specificities of the root causes of the conflict. So what might have been a sensible framing? We suggest three questions that might have formed a better starting frame for a collaborative peace process.

1. What is currently working to provide most people in Solomon Islands a high level of community safety and human security, and how might these strengths be supported?
2. What were structural and proximate factors that contributed to the conflict (as in Table 8.1) and how might these be remedied?
3. What new risks to the peace do locally knowledgeable people fear (such as revenue collapse as a result of ending log exports)? What can be done to hedge these risks?

Assiduously and collaboratively working at answering these questions cannot secure fast and full responsiveness in peacebuilding. Yet we contend they are simpler questions that could have delivered quicker specificity of targeting than the more abstract ambition of building core pillars of a state stigmatised as having failed. Or worse, regime change followed by who knows what?

We construe successful peacebuiding as a craft. A craftsperson does not follow a template. That is what factories do when they make a piece of furniture, for example. When a craftsperson makes furniture, there is no recipe that says do this much planing, this much sawing, gluing, sanding. The craft is a creative, evolutionary practice rather than a templated one. Furniture making as craft is about looking at how the piece is evolving in the process of the craft, how to go with the grain by now planing some more here but not there. As Sparrow (2000:201), from whom we have adapted the furniture metaphor, puts it, being a craft means resisting the idea of work being organised around tools; tools are organised around work. Furniture making is a solitary craft, though one learnt in communities of practice, and through apprenticeship, just as peacebuilding is. Peacebuilding is, however, a collaborative rather than a solitary craft. A richer analogy might be to the craft of the musician in an orchestra. The first violin is collaborating up to the conductor, down to the other violins, laterally to the woodwind section and last, but far from least, she is profoundly attuned to the emotional effects of the music in causing the audience to gasp, laugh, cry or lapse to serenity. A collaborative craft is not about following a fixed mandate because creative success turns on hearing and seeing opportunities to connect creation to positive human emotions. ‘We would not have great symphony orchestras if conductors focused only on keeping musicians from playing out of tune’ (Heimer 1997).

But is the sheet of music played by a musician a template? Well, yes; neither the craft of peacebuilding nor the musical craft is a template-free zone. In this volume, we have referred to many templates: the software in the Solomon Islands Customs Division that makes it impossible to change certain fields on the customs valuation without leaving behind your electronic signature, and completely impossible to change other fields, is a nice example of a sharply prescriptive template. What our conclusion resists is an international template that is an overarching constraint on where a mission cannot go and must go on questions of governance development. A mission needs a mandate that empowers and sets limits on how and why peacekeepers can arrest people, use different levels of force and take over certain functions from a sovereign government for an agreed period. It need not be a mandate that rules out forms of development assistance agreed in local dialogues and falling within the competence and budgetary capability of the mission to support with other partners. Accountability of a peace operation is not best assessed as compliance with the letter of a prescriptive mandate, but more in the way that some accountability was achieved for RAMSI, with Pacific Islands Forum and Solomon Islands Parliament hearings and reports, community surveys, measurement and reporting of good and bad outcomes and robust debate in local media. So perhaps we learn to see elements of the peacebuilding craft from the solitary furniture maker, and other elements from the symphony metaphor, but we get closer to the craft mentality by pondering jazz improvisation. Musical scripts are used and adapted continuously, but in ways generated by interactive learning from the moves of other musicians in an ensemble and from audience reactions. That is not to say there are no limits to the mandate of the jazz musician; however good you are, you cannot jump on the stage and take over someone else’s gig!

The craft of peacebuilding is one of collaborative learning, of getting the timing right, deciding which different sets of collaborators should be assembled at different times to work sequentially on new problems. It is the craft of not playing the grand finale to the concert before the audience has begun to warm to the music, of not being mandated to stage an election until citizens have begun to believe in democracy. It is an uncharted social process engaging many actors. We must expect it to be slow.

Pillars and shadows

The pillars of great public buildings in capitals of the West symbolise the solidity of the state. It is an odd kind of learning from ancient Greek and Roman statebuilding that the West sent architects to Athens and Rome to draw templates of the only residues of those republics that stood: the pillars that once held up the rubble around them. Of course, the pillars were a misleading symbol of how the architectures of ancient states and more modern ones were gradually cobbled together over centuries.

Part of our analysis in this book has been that developing states need to pay attention not just to building neat pillars but also to the messy business of managing shadow states (Reno 1995) that fracture and corrupt those pillars. In the Solomons case, the shadow economy of logging and a complex of relationships among foreign business and diplomatic interests and local politicians and militants have shadowed and recurrently opened cracks in the architecture of the state. We have concluded that RAMSI’s contribution to improving that kind of management has been modest.

The particular way RAMSI set about repairing and constructing pillars of the Solomons state was by shadowing each key functionary of the pillar with an expatriate—a dual-desk approach. At first, RAMSI officers directly operated core organs of government, especially the security and finance sectors. This was necessary and successful in stabilising the society. Then the strategy was for these in-line expatriates gradually to build the capacity of their counterpart at their dual desk until they were ready to take over completely. We have concluded this was mostly not successful because outstanding operational police and operational finance bureaucrats were not always good trainers of their counterparts. They were often prone to frustration at the inferior education or experience of those they shadowed and to taking over again as soon as something was done badly. RAMSI frequently did not apply the basic lesson of peacebuilding that something done tolerably well by locals is always better than something done extremely well by outsiders.

In any case, as Bayley and Perito (2010) and Wilson (2010) have argued in relation to police building, it is better for operational peacekeepers to be replaced by management consultants experienced in building police institutions and police academy experts who set up training for local trainers who do the in-line training in the language, and using the cultural scripts, familiar to those who are trained. It is an odd theory of how to build an institution to train all members of the institution one by one before there is an architecture those individuals can inhabit. Much more than training is needed in a police force to build logistical capability, personnel and promotion systems, transport and communication systems, data processing, police accommodation, secure armouries and many other management systems. So we suggest it was a flawed conception of how to build the pillars of the state to put outsiders in to run them in their way and then to rely so heavily on those same individuals to remake themselves as trainers of counterparts to run those pillars in ways that locals felt would be sustainable. As Bu Wilson has said to us, quoting Ed Rees, ‘[w]e would not think it sensible to build a health system by flying a thousand doctors into a country, so why would we think that was the way to build a police service?’ (see further Wilson 2010:185).

The Solomons case reinforces the lesson from previous research that a boom-and-bust economy is a threat to peace (N. Ferguson 2006) and increases the risk of crime waves (Fischer 1999:Appendix N). Extreme poverty is a structural driver of armed conflict (Collier 2007), but even rich countries suffer more wars during periods of history punctuated by extreme boom and bust (like the period from the 1890s to 1939). Boom and bust in an economy in which the dominant exports—logs and fish—have been more in the shadow economy than in the formal economy of taxpaying have cashed up both foreign and indigenous corruption, contributing to unusually unstable government coalitions.

The current boom that took off after 2003 is not just a logging boom; it is also a boom in donor funding triggered by RAMSI. What a double tragedy it will be if, just as RAMSI created the security that amplified the 2004–10 logging boom, it amplifies the bust of a final collapse of logging stocks when that bust coincides with RAMSI’s exit. RAMSI succeeded with locals in getting guns off the street and in securing macroeconomic stability in the short term. But prospects for long-run peace and political stability are slim if there is not long-run macroeconomic stability, no matter how well statebuilding and incarceration of militants are executed. Shadow governments could unravel the state again if the specific grievances that led to conflict continue to be swept under the carpet.