I first heard Paul Collier speak at a UK Government Economic Service conference about four or five years ago. The city was Cambridge and the subject that morning was on "economics of conflict : reducing the global incidence of civil wars". It was an energetic lecture full of empiricism and conviction. Economists and sceptics flock together, but that morning nearly every lunch table exude praise at Collier's refreshing insights on a subject that very few economists have given much thought. Since then Collier has gone on to publish the widely acclaimed "The Bottom Billion", along the way solidifying his place among the leading development experts in the world. Judging from his latest effort Wars, Guns and Votes: Democracy in Dangerous Places, that wonderful enthusiasm remains undiminished.
The basic premise of Wars, Guns and Votes is straightforward: security and accountability are crucial to the economic and social development of national states. They foster economic growth and ensure more equitable distribution of wealth. More importantly, security is a "universal right" or, as Sen would put it, it is both an end and means of development. The problem is that security and accountability are fundamentally public goods whose benefits go beyond individual states and therefore failure to provide them has repercussions beyond international borders. Bringing it closer to home, failure in securing peace in the Democratic Republic of Congo (DRC) is not just costly to our Congolese brothers but also to our nation and other neighbouring states through regional risk contagion, reduced intra-regional trade, increase in refugees and so forth. Like many public goods, the "free-rider" nature of security and accountability leads to inevitable under-provision.
As national states develop they tend to develop systems that internalises these "external costs" of insecurity through internal systems of accountability and external cooperation with their neighbours or to be more precise, they begin to see the benefits of sharing their sovereignty to further their long term strategic objectives. Unfortunately, for many poor African nations the process of gradual improvement towards more secure and accountable systems is impeded by significant structural weaknesses, which stem from their respective geographic sizes. Many African countries are either too large to coordinate internally (e.g. DRC) or too small to supply accountability and security that benefits everyone or put differently, the country may be too small to enjoy the benefits of security and accountability alone (e.g. Rwanda).
According to Collier, these issues matter because without recognising the structural nature of the problems, the push for democratisation is likely to amplify rather than reduce instability and political violence. Indeed, experience shows that rather than enforcing accountability and security, the African "democratic wave" has in many cases led to relapse of violence and sometimes spawned a new generation of dictators under the cloak of "democracy". Collier soberly concludes "we have most surely underestimated the degree of difficulty and promoted the wrong features of democracy: the facade rather than the essential infrastructure". This of course is not surprising because real democracy emerges from deep structural changes to the incentives of those in leadership and would-be-pretenders to the throne. The problem, argues Collier, is that these changes are difficult to come from within. Indeed even where change can be sparked, sustained security and accountability requires some form of "external intervention" or the nation state sharing some of its national sovereignty with outsiders.
Collier's proposed solution follows naturally : in order to reduce the spectre of civil wars, the international community should adopt a strategy whereby a "small intervention" from the international community can harness the political violence internal to unstable African nations. This idea of course is not new and for many who have read Collier's earlier book "The Bottom Billion", much of the book is all too familiar and simply adds more detail.
To be sure, there's a much clearer explanation of ideas in Wars, Guns and Votes than previously presented. The proposal is to establish an international standard for elections linked to a "powerful carrot". African Governments would be invited to sign-up to these internationally mandated standards. Those that do would be guaranteed "international protection" against internal aggressors (real or imaginary), who might want to take power by force. Collier assumes these to be largely disaffected military leaders who believe they can do a better (or worse) job. If the country signs up it will then be monitored, rewarded and punished on a different scale from others (no detailed exposition on the nature of these standards is provided in the book). If the leaders decided to renege on their agreement to "good governance", international protection would then be removed. Collier gets muddled here with some parts of the book suggesting that the international community might then encourage a "coup" (for example, on Zimbabwe he is forthright in noting that toppling Mugabe would have been ideal) with other parts largely silent i.e. the international community would simply remove protection and perhaps "hope" that the dissatisfied government opponents instills discipline. If the country behaved itself it would enjoy all the privileges, including some financial rewards from donors. Incidentally Collier is equally silent on whether the country that reneges first time can then come back and "play the game" again. As one thinks about the problem it becomes obvious that in so far as it's not in the interest of the international community to allow the country to remain a pariah state, it's inevitable that such a country would get another shot! This obviously weakens the credibility of the system.
More on the credibility problem later, the immediate question of course is whether African dictators (decent or otherwise) would sign-up. Collier is equivocal and demonstrates the logic through a simple "sequential game" between the country and the international community, which is solved through "backward induction". According to Collier, not only would international protection guarantees be attractive to a sitting government, but it would also be attractive to donors, which may translate in cash, allowing the reformed dictator to gain legitimacy in the eyes of his people. For their part the opposition parties would be equally amenable to seeking such an arrangement because it allows them to pre-commit to good governance, significantly enhancing their electoral appeal. As examples of where some version of the model has worked, Collier cites an old French guarantee in Francophone West Africa and the current NATO alliance. Collier's innovation is linking "security guarantees" with strong requirements on governance and electoral democracy.
That "solves" the security problem, what about accountability? Here, Collier advocates better enforcement of probity in public financing by allowing donors to impose greater "governance conditionalities" specifically with respect to budget processes. An interesting complementary proposal is for the development of new forms of budgetary processes that separates each of the government ministries into three: overall policy, allocation of money and actual supply of actions. The relevant ministry would only be responsible for overall policy and would not engage in any spending. This way it becomes easier for donors and the general public to monitor how money is being spent. Unfortunately, Collier does not discuss this idea in sufficient detail for it to deserve serious consideration by reformers. There are some musings about NGOs and charities being potential players in the actual supply of actions, but without concrete examination its difficult to judge how feasible the idea is.
Wars, Guns and Votes is certainly problematic, but not completely useless. There are valid points made throughout the book, particularly in the earlier chapters, backed by empirical force. For example, the observation that "democracy in the societies of the bottom billion has increased political violence instead of reducing it" certainly accords with my own observation that what we have in many African countries is not genuine democracy but rather electoral democracy that has not translated into genuine opening up of participation. Even in a peaceful nation such as ours and where regular elections are held, full democratisation and empowerment remains something to be hoped rather than grasped. Part of the problem is one alluded to by Acemoglu and Robinson, that is genuine democracy does not only require a shift in incentives as Collier identifies but it must be accompanied with significant shifts in power from the ruling elite to the masses. That can only occur through some redistribution of de-facto power and change in institutional arrangements (de-jure).
Collier is also right on the need for reducing aid to military spending. According Collier "on average 11 per cent of aid finds its way into military budget". Part of the problem as Collier discusses is that the move towards "country ownership" of aid through budgetary support rather than direct spend has had the unintended consequences of African governments diverting money intended for health and education towards military expenditure. This is worryingly because according to the evidence presented far from deterring violence, increased military spending especially in post conflict situations (where budget support is significantly high) provokes further violence. It appears aid financed military spending sends the signal to citizens that the political conditions may be unstable going forward which leads those opposed to government to rearm. The upshot of Collier's conclusion is that in post conflict situation better ways must be found for delivering aid e.g. through infrastructure spend and investment in human capital rather than direct budget support.
Unfortunately, as already noted there are also many areas of the book that are far from convincing. These largely fall into two camps -incomplete ideas; and, inadequate solutions. These are discussed in turn.
There are a number of instances where Wars, Guns and Votes fall significantly below the mark. I have already hinted on the muddle of the proposals and the unanswered questions related to implementation, but the problems go beyond that.
For example, the role of traditional leadership in African states in preventing corruption and violence is not fully explored. Collier dedicates a whole chapter to discussing the problem of ethnic politics and presents empirical evidence that shows that controlling for all other factors ethnic diversity is associated with greater instability and general economic retardation for the poorest nations nations. Whilst that has some empirical support , Collier's general treatment of this area is particularly lacking. For example, there's no formal discussion on the role of traditional institutions in restraining violence or perpetuating it. I remain convinced, the issue is not the absence of visionary leadership per se, but the extent to which democratic processes are fully embedded with traditional ones. African countries that have avoided civil conflicts have done so by managing these processes (e.g. Botswana whose path to tribal integration was a managed process rather than a growth driven occurrence). The issue is not the absence of visionary or charismatic leadership but development of simple and effective policies that are culturally consistent. The reason many African countries have relapsed into violence is not the absence of foreign intervention but too much promotion of foreign imposed ideas that do not recognise the unique processes that shape particular African settings.
Another example is Collier's undefended presumption of the desirability of electoral competition. Collier does not pay sufficient attention to defining the case for electoral competition. This may sound surprising, but we need to be clear on whether our goal is more electoral competition or simply deeper scrutiny. There are many countries in Africa where greater electoral competition has not translated into sustained economic growth divorced from issues related to rule of law. Conversely, some countries have little internal electoral competition but have developed checks and balances that foster growth (e.g. Botswana).
These are just two examples, which at the surface may even appear non-binding, but they hint at a fundamental problem at the heart of the Wars, Gun and Votes thesis. It does not pay sufficient attention to other forms of literature, which impacts greatly on the robustness of Collier's proposed solutions.
The golden rule is simple: if the diagnosis is partial, the solution will similarly be so, or worse generate unintended consequences. It therefore comes as no surprise that the weakest part of the Collier thesis is the proposed solution of "limited external intervention" as a mechanism to incentivise good governance. This assessment is informed by a number of observations :
a) The Collier proposal is predicated on a credible "intervention from outside". In order for any sitting dictator and the opposition to sign-up they have to believe that the relevant foreign power would be able to intervene and restore order. In practice this is unlikely. There are a number of conflicts where international guarantees would not work to prevent coups precisely because these countries are too diverse and too difficult to manage e.g. the DRC. Collier gives the British intervention in Sierra Leone as the model for intervention, but many African countries are much larger than Sierra Leone.
b) The proposal assumes unrivalled confidence and trust in western institutions. The central proposal is for the United Nations to take charge of framework for "international guarantees" and undertake the appropriate military interventions. Aside from the poor record of the United Nations in external interventions, there's a question of who punishes the United Nations if it abrogates from intervening in countries because specific sponsoring nations withdraw their support. There may well be circumstances in which the UN might not live up to its obligations and failure to which would render such charters meaningless. This is the classic problem of who polices the police. Surprisingly, Collier just glides over it without discussing how such pre-commitment problems would be resolved.
c)It suffers from a poor diagnosis of the dictatorial nature of many regimes. In Collier's simple sequential game , the threat to a sitting dictator emerges largely from a military coup or rebel insurrection. But if the Zimbabwe predicament taught us anything it is that often it is difficult to distinguish between the government and the military. The threat to Mugabe was not the military because Zimbabwe was run as junta. Collier says "Zimbabwe needed a coup" and that his system would have ensured Mugabe behaved accordingly and signed up to the governance charter. It is quite obvious to many observers that even if such a charter was presented to President Mugabe he would not have signed up because his position was ensured by the military. For Collier's incentives to work they has to be a force that threatens a sitting government sufficiently from within. In many countries such forces do not exist.
d) Collier overestimates the incentives of African government to sign-up to the voluntary charter. In Collier's model there are two things that induce government to sign-up to the voluntary charter : the protection from coups; and, the prospect of financial reward for the country. As Collier notes, but avoids deeper analysis, many nations in Africa have significant natural resources that push the incentives the other way. Why would a leader with significant oil deposits that he could sell waste time with a voluntary international charter, when they can perpetually buy-off their opponents? There's also the ever-present issue of China's role on the continent and its non-interventionist foreign policy. Again, Collier does not even mention these issues and the extent to which they may shape incentives facing African states.
I started the review by stating that Collier brings passion and expertise to the challenging issues facing developing countries. I conclude that unfortunately that is not always enough. Although Wars, Guns and Votes is steeped in empirical data it does not pay sufficient attention to the complexities of the African experience or take on board other areas of literature on the emergence and sustenance of dictatorial regimes. This has led Collier to propose, in my view, insufficient policy recommendations. That is not too say this is a wasted effort. On the contrary, Collier certainly adds rather than subtracts to the debate because the empirical evidence presented in many parts of the book provides a good platform for other economists to build on and that is no mean feat in an area that many economists have so far feared to tread.