Collier’s diagnosis of the problems of the “bottom billion” is brilliant. Here is what I like about it (three S’s):

Sovereignty. Collier recognizes that it is time to deep-six the nintheenth-century swoon over government sovereignty in favor of a 21st-century concept of the people’s sovereignty. Kofi Annan described it when he called for responsible intervention when a government violates its people’s rights, most obviously in the case of genocide. The Chinese government officially denies people’s sovereignty, to the consternation of Western donors who decry Chinese support for Sudan and Mugabe. The Europeans are struggling to “share” it, which is one way to undo its excesses. In a holdover from the G-77 days when the non-aligned South, lacking global economic heft, had little but ideas with which to challenge the North, the post-apartheid South African government retains an outdated view of sovereignty, siding with tyrants over people in Zimbabwe and Burma. It is time to insinuate the idea of the people’s sovereignty into the international discourse on development.

Size (and scale and scope, too). Nigeria’s economy is just one-fifth the size of that of the Netherlands. The economy of sub-Saharan Africa—including Nigeria and South Africa—is smaller than the economy of New York City. Collier emphasizes that Africa’s tiny economic states are too small to be viable; they cannot benefit from economies of scale, particularly in the provision of public goods. Of course it is not just diseconomies in provision of schools, roads, and security that are the problem. Private producers, particularly of manufactured goods, face diseconomies that reduce the region’s growth potential.

Security, as in people feeling safe in their persons and property. For decades the development community and aid bureaucracies have turned a blind eye to the insecurities poor people face—from children kidnapped for sex trafficking in Thailand to women raped in Darfur and Bosnia to the would-be entrepreneur in South India who dares not risk a high-return investment where elites control the police and the courts—when the state fails to deliver the most fundamental public good. Collier reminds developmentistas that, without internal security, economic development is inevitably stalled. The Marshall Plan, he notes, was complemented in postwar Europe by the creation of NATO (though the analogy is somewhat faulty—U.S. troops were in Western Europe to protect it from an external threat, not internal ones).

I am less confident, however, about Collier’s endorsement of over-the-horizon military guarantees. I would recommend other interventions, less exciting but better grounded in experience and evidence.

But first, speaking of evidence, there has been plenty of grousing about Collier’s reliance on econometric results to “prove” that one thing (peacekeeping, democracy) causes another (growth, increased violence). I am inclined to be more tolerant. He is putting big issues on the table, and in reality the issues get more attention because he is purporting to show they matter. Would there be a Collier TED talk and an essay in Boston Review if he had never published a “scholarly” article?

What are the less book-able (as in publications, not bets) options for interventions? Here are three ideas:

First, find ways to foster sovereignty of the people instead of the incumbent government. Mo Ibrahim—the Sudanese founder of Celtel, the mobile phone service provider that has swept Africa—gives an annual prize to democratically elected African heads of state who step down (such as Mozambique’s Joaquim Chissano) when their terms end. Nicolas van de Walle, an Africa expert at Cornell University, recommends that donors make clear that they will halt aid where heads of state hang on beyond twelve years; that could apply today to Belarus, Burkina Faso, Cameroon, Egypt, Ethiopia, Uganda, Zimbabwe, and Uzbekistan. Raghuram Rajan, while Chief Economist at the IMF, suggested that in post-conflict countries, voters might like the option of electing a non-national for one transitional term. For example were Somalia to settle down, Somalis would almost surely elect Nelson Mandela or Kofi Annan over any one of their warlords. And, in low-income, democratically immature, resource-rich states, the people might choose international monitoring of the per-capita distribution of income from their resource patrimony, despite the apparent incursion on traditionally defined sovereignty. With veteran development economist Arvind Subramanian, I proposed in 2004 that such a provision be written into the Iraqi constitution; recent Kurdish-Arab turmoil in oil-rich Kirkuk suggests we will regret that it was not. In Chad and Azerbaijan, the same approach—imitating programs in Alaska and Alberta, Canada that guarantee per-capita distribution of oil-fund income, but with the addition of international monitoring—would be less draconian, more people-friendly, and more likely to help build democratic habits than would be Collier’s over-the-horizon military guarantees.

Second, to deal with the lack of scale and size, Africa needs donor buy-in for big infrastructure investments. Bureaucrats in the U.K. Department for International Development, the U.S. Agency for International Development, the World Bank, the African Development Bank, and elsewhere do not have sufficient incentives to finance regional infrastructure. Why? Because as the number of prospective recipient governments increases linearly, the work, risk, and elapsed time between a proposal and disbursement increase exponentially. There are other complications, too. A road network linking Swaziland, South Africa, Mozambique, Botswana, and Zambia has to go through Zimbabwe, and until recently donors had good reason to stay out of Zimbabwe. Tensions arise whenever multiple jurisdictions “cooperate.” Even the Washington, D.C. metro system and the Port Authority of New York and New Jersey have trouble with the politics of who pays and who benefits. Still, regional integration (getting trade agreements disentangled as well as investing in regional power grids and transportation networks) might benefit sooner and more directly the people of Mali, the Gambia, and Guinea-Bissau than would guarantees against coups that overthrow their elected heads of state.

Third, to obtain better security, donors should support more police training. The people of a poor state might choose first and foremost an honest, capable, and responsible police force, but traditional donors fund army training far more than police training. Donors also measure their contributions to health and education, but not to peacekeeping or crime-control, as Collier notes. The Center for Global Development tracks support for peacekeeping when ranking rich countries on their commitment to development; Ireland, Norway, and New Zealand score well, while the United States is dead last in contributions to UN peacekeeping efforts as a share of GDP. (It does better when non-UN programs such as the NATO presence in Afghanistan are included.) Finally, donors have costly safeguards against economic abuse of their resources, but barely notice the criminal pre-election political payoffs that, as former Nigerian Minister of Finance Ngozi Okonjo-Iweala has described, make officials, once elected, pathetically beholden to their patrons’ interests. With greater resources and training, police could prevent this kind of corruption, which ultimately costs the new democracies far more than it would Western donors funding those police.

The kind of internal security guarantees Collier suggests need not be off the table. Even if the empirical work behind his assertions is flawed, the story he constructs starts from first principles and is compellingly consistent with history and current experience. You never know where and under what conditions over-the-horizon military guarantees would be the best solution. But there are other proposals the development community should advocate as well, and first. Even though they do not involve guns and war, we ought not ignore and underfund them.