Oxford University Press, 2008, 272 pp., $24.95
For the first two years of the Afghan reconstruction program I worked and fought with Ashraf Ghani while he served as Minister of Finance in the Afghan government and I was Administrator of USAID. The book he has written with Clare Lockhart might better be titled Frustrations of a Finance Minister because in it he vents against the international aid system he had to deal with during his tenure. He re-argues the endless conceptual and programmatic debates we had on development issues and tries to suggest an improved, alternative system to the current one.
That alternative derives from a description of the essential functions of the state, as understood in classical Western political theory, and the reformulation of the relationship between the aid system and the state to speed the re-establishment of those functions. Ghani and Lockhart believe that a better aid system along these lines, which they claim has worked in other circumstances, can save Afghanistan and a range of other failed states as well.
There is nothing unusual about venting old frustrations after the fact. Ghani and Lockhart might as well just get in line. Since the media tells us that Afghanistan is not going well right now, the record of the past seven years is being scrutinized with a mind to fixing blame on anyone, or everyone, other than oneself. So Donald Rumsfeld and his former staff blame their counterinsurgency campaign difficulties on the reconstruction effort they claim was mishandled by the State Department and USAID, the UN and World Bank. We in the aid community, in turn, blame the Defense Department and then NATO for not providing the security we needed to carry out the reconstruction effort, particularly in the south, and for interfering in an effort they understood but poorly. This is no small matter: 291 USAID-funded contractor staff have been killed by insurgents in Afghanistan during the reconstruction process, the highest death rates for the Agency since Vietnam.
Ghani’s complaints about the international aid system turn out to be a subset of these larger arguments between the Defense Department and USAID versions of what went both wrong and right. The White House decision to put the Pentagon in charge of the reconstruction effort in Iraq was based on dissatisfaction with the established aid system as it functioned in Afghanistan and elsewhere. Ghani became a favorite of Rumsfeld’s office in part because of the vitriol he directed against the international aid system. So Fixing Failed States has a political history and, as it turns out, a future political purpose as well.
The problem is that recriminations over the shortcomings of the reconstruction and development effort in Afghanistan are increasingly beside the point. Recent incident reports show that the west, east and north of Afghanistan are doing relatively well, but the Pashtun heartland, Hamid Karzai’s tribal area in the south, is not. The situation in Helmand, Nimruz and Kandahar provinces, however, has little to do with how reconstruction efforts have fared. Until someone deals effectively with the Pakistani tribes along the border that host Osama bin Laden and support the Afghan insurgency, the situation will remain difficult. The notion that a reformed reconstruction effort can fundamentally alter the calculus there is one way of several to misunderstand the problem.
More fundamentally, the idea that Western concepts of the state, and development experiences distilled from other quite dissimilar cases, really have relevance in Afghanistan is not persuasive. While theoretically about all failed states, Fixing Failed States is really more about Ghani’s experiences in Afghanistan and his aspirations for its future than anything else—and Afghanistan is not a typical state, failed or otherwise. So there is a mismatch between the book’s thesis and the Afghan case from which it is supposedly derived and to which it supposedly applies. Alas, it’s not the first time smart and well-meaning people have confused their own experiences for universal truths.
Ghani and Lockhart tell us, for example, that state building fails when outside parties—the ones with the money—lack a strategic plan written by local leaders. That’s right: If a society is not invested in a project designed to assist it, it is less likely to succeed. This is why the USAID business system until recently required a country strategy written by the local mission staff and the host government ministers (and approved in writing by them) before any U.S. government money could be spent. I suspended this requirement in Iraq and Afghanistan because the cloud of politics enveloping the Defense Department, the State Department, USAID and local Afghans and Iraqis made that kind of consensual planning impossible within a reasonable timeframe. Ghani’s experience thus encompasses a rare exception to the rule of how the international aid system works, but he takes the exception to be the rule.
Similarly, Ghani and Lockhart propose as a general reform of the international system that we transfer donor reconstruction funding over to the treasury of failed or recovering states (called budget support), work with them to write a national strategic plan, and focus on institution-building in the new ministries of these states. But that already is a general principle of policy, as a look at World Bank, UN, and bilateral aid agency strategy and planning documents for the past decade would illustrate. In what is now a vast literature on failed and fragile states compiled by scholars and practitioners alike, no one disagrees with the proposition that development and reconstruction are about building institutional capacity in host countries. We debate implementation strategies instead: How long does this need to take? How do we do it with some predictable success rate?
More to the point, as Ghani knows quite well, building capacity requires broad public support, cultural values that are conducive to orderly government functions, and effective leadership in the host country, and we sometimes don’t have any of these. Many of Hamid Karzai’s cabinet ministers, now and in the past, were chosen for their loyalty during and after the post-9/11 Afghan war, their bravery as military commanders, and their leadership in ethnic and tribal groups that Karzai needed to consolidate his political authority. Some were warlords brought to Kabul and given government ministries so Karzai could watch and keep control of them. No more than a third were competent enough to run a ministry involved with reconstructing a war torn country. Many members of the Afghan civil service with whom we dealt lacked even an elemental understanding of planning and budgeting. Some were even illiterate, as Ghani himself once told me. To his credit, at one point Ghani froze all new hiring to avoid the continued padding of the public payrolls by our warlord allies.
Ghani was also well known in Kabul for his efforts to bring Western management and policy standards to his country in the first four years of the new government. In a now-celebrated meeting of the new cabinet, Marshal Muhammad Fahim, Karzai’s Vice President and the Tajik commander who led the redoubtable Northern Alliance, demanded that Ghani provide funds from the treasury to pay his militia. Ghani refused, knowing Fahim’s reputation for corruption and drug dealing, to which Karzai half-jokingly cautioned: “Ashraf, you’d better pay Marshall Fahim the money or he might kill you.” Nervous laughter ensued, because many ministers knew that Karzai’s humor disguised a real threat by Fahim to Ghani. Ghani replied that he would provide salaries if Fahim gave him a list of names and ranks of all troops to be paid. Fahim never did; Ghani rightly suspected that many of the troops Fahim claimed needed to be paid did not exist.
Ghani was in a difficult position, to put it mildly: He knew that Afghanistan lacked the kind of institutional capacity that would have made a demand like Fahim’s unthinkable, and that giving in to such demands would make building institutional capacity itself unthinkable. Ghani was a man of personal integrity (which says a lot given widespread corruption within the Karzai government), courage (as the Fahim story illustrates), and a leader on policy issues, even if he often came to very different conclusions than the aid community.
Ghani is reportedly running for president of Afghanistan in the next election; I presume that Fixing Failed States is his platform, at least as articulated to the international community at large, which, in the Afghan case, is a serious political player and will likely remain so for some years. I wish him well. If I were Afghan, I might vote for him. But I am not reviewing Ashraf Ghani; I am reviewing his book, and that, as I have already suggested, is another matter.
Aside from mistaking the recent Afghan development crucible for the methodological norm, Fixing Failed States surveys stable regions of the world in order to distill cases in which innovative and competent local leadership succeeded in driving the development process. There are such cases, but many are not applicable to failed and recovering states, and they are certainly not applicable to Afghanistan. Most of us who have worked in reconstruction efforts elsewhere (I’ve been involved in at least 12 efforts over the past 18 years) know that competent governance is the first requirement for success. But what happens when the competent elite has long since departed (if there ever was one, they were the first to leave) and we are left to deal with thugs and criminals, drug lords and money launderers, militia leaders and arms dealers, and too few courageous patriots?
If, rather than searching the world for inapplicable examples, Ghani and Lockhart had focused on conceptualizing and detailing new approaches to state building on the basis of what has worked in Afghanistan, they would have done a great service to the craft of fixing failed and failing states. They would have augmented the work Paul Collier has usefully done in applying empirical research to test different hypotheses about state failure and nation building.1
For example, in the first couple of years of the new Afghan government, USAID and the Karzai Administration undertook a quiet initiative to recruit 900 professionally trained Afghans from the diaspora to staff the new government ministries in senior positions. USAID put the Afghans on the payroll of its contractors (reviled by Ghani), professionals who were supposed to be later transferred to the Afghan public payroll. Some of these early officials later became cabinet ministers and deputy ministers. This effort represented an innovation that proved useful to institution building in Afghanistan and that might be adaptable to other failed states, but Ghani never mentions it—perhaps because it would show that the aid system sometimes does work. USAID has also appointed competent Afghan cabinet officials from line ministries to the USAID procurement committees that select contractors and NGOs in contract competitions to run programs. The purpose of this is to get Afghan buy-in and ownership on USAID contracts, a strategy that has worked well.
So beyond mistaking the aid community’s approach to Afghanistan as the international norm, Ghani and Lockhart have mistaken international development success stories for what might work in special failed-state cases like Afghanistan. This is not a trivial error, for drawing distinctions among classes of cases and among individual challenges within classes is really the beginning of wisdom in the development business. Institution building in development is not a science. We have no quantum theory of institution building, no set of interventions that, if scrupulously followed, will always yield functional institutions. At best, we have developed a set of approaches to institution building that work in some countries, sometimes. They do not work everywhere, with scientific predictability, because local cultures and the political and tribal dynamics that arise from them, the state-collapsing trauma the country has been through, and constraints imposed by the neighborhood invariably limit what outsiders can do to influence events. There can be no unified field theory of development, for development turns out to be many things in many circumstances. Anything one can say that is generally true about the challenge is bound to be too general to be of any use.
We do know that successful institution building, once we figure out how to make it work on a case-by-case basis, takes 10–15 years at a minimum, and a consistent investment of money over the entire period.2 Certainly, we would have been much more successful in Afghanistan by now had we been given adequate funding in 2002 and 2003. We did not receive that funding because of the dysfunctions of the U.S. Federal budgeting process. Ghani and Lockhart argue instead that we can build institutions easily and quickly if efforts are aligned in such a way as to engage local talent, incentives and pride, so that the transfer of control from donors to local government should take place as soon as possible. They contend that “fundamental institutional change can occur within relatively short periods of time if the political and social imagination of leaders and the public rise to the challenge. . . . State architecture has shown that it can be flexible and dynamic: the architecture of international organizations has been rigid, bureaucratic, and at times downright dysfunctional.” But they offer no empirical evidence to support the argument that quick fix institution building forms of development assistance can work in Afghanistan, because there isn’t any. Building institutions virtually from scratch is difficult and time-consuming no matter who tries to do it.
Perhaps Ghani and Lockhart confuse the triage work of outside aid organizations with sustainable indigenous institutional development. It is true that the international aid apparatus knows how to do triage, and do it quickly. It amounts, frankly, to a neo-colonial model in which the international aid system takes over the functions of government at the early stages of reconstruction in order to provide public services. We know how to lower child and maternal mortality rates, how to build roads and bridges, how to increase agricultural productivity, and how to raise literacy rates by using sometimes expensive outside contractors and international non-governmental organizations, which is what we successfully did in Afghanistan in the first five years of reconstruction, insofar as security allowed. Much of what Ghani and Lockhart attack the international aid system for failing to do in the first four years was done gradually over time as the Afghan government showed evidence it could handle discrete elements of the reconstruction effort itself. The confuse the birth pangs of the effort in those chaotic early years with what a mature reconstruction program should, and more commonly, does look like in later years.
Yet it is this triage system that Ghani attacks most vehemently—as a hindrance to institution building. This is simply unfair. The aid community faces considerable pressure to deliver services in the wake of war and disaster and ignore institution building until later. Some of those pressures well up from the local population, which demands public services its own new leaders cannot deliver, even as the international news media criticizes the slow pace of reconstruction and broadcasts the suffering of the local population to the world.
Moreover, if the U.S. military is present, it wishes to win local hearts and minds for entirely valid if parochial reasons by delivering high-profile services (nearly always construction projects not approved or integrated into ministry planning, and with no follow-on maintenance or support) as quickly as possible. The greatest impediment to genuine institution building in the early stages of both Iraq and Afghanistan was Donald Rumsfeld’s Pentagon. USAID would add capacity building into reconstruction budgets, and they would delete it as a luxury we could not afford. Later, in Iraq, Rumsfeld intervened several times to restore cuts by the Office of Management and Budget to USAID institution- and capacity-building budgets (to our astonishment, given his previous opposition). We have all learned some lessons about reconstruction, including the Pentagon.
Oddly enough, while Ghani and Lockhart insist that even the early focus of aid efforts be on institution building, they trash one of several means of achieving it: technical assistance. They call it a waste of money. I say “oddly” because Lockhart spent her time in Afghanistan providing technical assistance to Ghani in building the capacity of the Afghan government. After Ghani left Afghanistan to write his book, he provided technical assistance to the new regional government of Southern Sudan under a World Bank contract.
Clearly, some technical assistance does not work, since it is dependent on the motivation of the receiving government. The international state-building machine is sometimes inefficient, poorly coordinated, too slow and usually expensive, but it has had more successes than failures. But what he proposes—that donor aid funding be immediately transferred to the Afghan government’s treasury, which would supposedly spend it on an Afghan-written reconstruction and development plan—would have made things worse. It would have dramatically increased the already high level of corruption in the Afghan government, and slowed implementation even further because of the absence of capacity in the ministries.
Thus Ghani criticizes the ineptitude of outsider contractors and NGOs and then describes the three great successes of the new Afghan government: the network of health clinics run by the Ministry of Health; his reforms of revenue collection in the Ministry of Finance, which provide the new government with a local source of income to administer public services; and the National Solidarity Program, which allows local communities to design and build their own small scale village development projects (which spent $300 million through 2007). Yes, these three programs have been successes (e.g., child mortality is down 25 percent and revenues are up 158 percent), but what Ghani does not mention is that all of them were led by the few competent Afghan ministers, all involved substantial technical assistance by the international community, and the implementation of all (to varying degrees) were managed or facilitated by USAID and other donors, World Bank contractors and NGOs. In the National Solidarity Program, for example, the ministry itself hired an American contractor, DAI, to help facilitate the program. Virtually the entire professional staff of Ghani’s finance ministry were expatriate technicians on U.S., UK and World Bank contracts, because there was little Afghan human capital capable of running the Ministry (if there had been, why didn’t Ghani hire them?) in the early years.
Since politics of several sorts are at work in Ghani and Lockhart’s writing of Fixing Failed States, I looked for some acknowledgement of the larger role politics inevitably plays in decisions about reconstruction in any failed state. Despite a chapter entitled “Failed Politics”, I could not find it. Yet politics clearly drives the decisions of leaders in failed states and functional states alike, not abstract theories of public policy, mathematical formulas of quantitative analysis, or the national strategic planning processes Ghani proposes.
Afghan examples, all of which Ghani knows well, are not hard to find. When Hamid Karzai visited Washington in 2002, he asked President Bush to reconstruct the shambles of the 950-kilometer highway from Kabul in the east to Herat in the west, the bottom half of the ring road that ties Afghanistan together. (The top half was to be done later by the World Bank and the EU.) He told the President that he needed to show the Afghan people, particularly his own tribe, the Pushtuns (from which the Taliban drew its support), that he could deliver visible aid and services from the Americans. So President Bush ordered USAID to build 715 kilometers of the highway, while the Saudis and Japanese were to build the remainder.
We were given 13 months by the two Presidents to complete the first segment through the Taliban heartland from Kabul to Kandahar. It would have been useful to first build up the capacity of the Afghan Ministry of Transportation to accomplish this task, but that is not what the politics of the circumstance demanded; the Americans had to be seen doing it, just the opposite of what one might have expected, and the opposite of what Ghani insists upon throughout the book. Louis Berger Co., a construction management company, won the contract competitively and worked with us to get the project done on time and under budget. Five non-American sub-contractors—three Turkish, one Indian, and one Afghan-Turkish joint venture—won sub-contracts to do the actual construction. The highway cut through enemy territory, where, during construction, sporadic attacks by Taliban claimed the lives of twenty workers. But finally, with 13 television cameras filming, President Karzai cut the ribbon inaugurating the new highway.
Sometimes political decisions came into conflict with previous ones. When it became obvious that Louis Berger’s highway-building labors were being slowed down by efforts to build a thousand schools (decided and announced by the two Presidents on a later visit to Washington by Karzai), I decided to postpone the school project, since Karzai had told us the first priority was the highway. We ended up building 680 schools, which I insisted should be designed to resist the kinds of earthquakes that later killed 20,000 Pakistani in 2005 and 9,000 Chinese school children in 2008. Karzai’s ministers criticized us for that decision: For political reasons, they wanted more schools faster at lower cost.
During the school construction, meanwhile, Congress passed a new law requiring USAID to ensure that the schools and health clinics were handicapped accessible: more politics. Obviously, building to earthquake and handicapped-accessible standards increased costs substantially and reduced the number of schools we could build. If that were not enough politics, Ghani adds more, claiming that some of these USAID-built schools later collapsed. Not so: Some schools hastily built by the Defense Department, not USAID, did collapse after they were built. Some USAID schools had roof trusses that twisted when particularly heavy winter snows weighted them down, but we discovered the weakness ourselves and the contractor corrected them without one school roof collapsing.
Both donor and host-country politics will always be more powerful than antiseptic planning processes in fixing failed states, particularly when U.S. troops are on the ground and the national security of the United States is at stake. But Ashraf Ghani should already know that, or at any rate he will find it out soon, if he is elected President of Afghanistan. If he doesn’t, the many Marshal Fahims of Afghanistan will remind him.
Collier, The Bottom Billion (Oxford University Press, 2007), and “Survival of the Fattest”, The American Interest (May/June 2007).
See James Dobbins, Seth G. Jones, Keith Crane and Beth Cole DeGrasse, The Beginner’s Guide to Nation-Building (RAND, 2007).