Skip to navigation – Site map

The Challenge of Reconstructing "Failed" States

What lessons can be learned from the mistakes made by the international aid community in Afghanistan?
Le défi de la reconstruction des États «faillis» : Quelles leçons tirer des erreurs commises en Afghanistan ?
El desafío de la reconstrucción de estados fracasados
Serge Michailof


While international aid has significantly increased its support for the reconstruction of fragile countries and "failed" states emerging from severe crises or civil war,  we begin today to realize that the disaster in Afghanistan is largely due to mistakes made by the Western coalition, and in particular, to the disorganization of its development support efforts in this country. In this uncompromising article, Serge Michailof undertakes a systematic analysis of international donors' mistakes in Afghanistan between 2002 and 2010. His purpose is to draw lessons that can be applied more broadly and thus avoid repeating the same mistakes in other contexts.

After reminding the reader that aid efficiency in countries emerging from a crisis is a source of considerable disappointment, and that not only promises and flashy announcements but considerable financial resources are required to rebuild those countries, Serge Michailof next makes a candid analysis of donors’ poor performance in Afghanistan. He emphasizes the lack of a coherent strategy and clear goals to guide their actions. He stresses the fact that ensuring the safety of the people and providing a local administration is critical in these circumstances, although largely forgotten by donors. Since foreign forces cannot fulfill this need indefinitely, building reliable national security forces is essential. He then describes the consequences of the lack of coordination and strategic management of aid, which results in severe inconsistencies in resource allocation, and proposes a new aid model to address such deficiencies. From the Afghan case analysis, he draws general conclusions regarding the principles governing aid coordination in fragile countries and the inadequacy of the millennium development goals as a conceptual framework in such contexts. He stresses the dire consequences of approaches governed by donors' short-term concerns, which end up establishing a parallel donor-driven administration that can only weaken State institutions. While State building requires founding modern institutions, this issue is paradoxically, neglected by the international aid community, even though this is feasible in difficult contexts as demonstrated by a number of successful achievements in Afghanistan. He then criticizes the superficial formal type of democracy systematically imposed by the international community in such contexts and stresses the need for grassroots democracy. Drawing from this extremely in-depth analysis, he finally makes a series of general recommendations intended for both official development aid institutions and non-government organizations (NGOs) that are currently facing and will continue to face similar challenges in other areas of the world.

Top of page


Top of page

Editor's notes

Text of the presentation given at the Humanitarian Forum, 1 October 2010, Annecy, France1. This article is also posted on the Fondation pour les Etudes et Recherches sur le Développement International website (

Full text

Aid performance in 'failed' States is disappointing

1For more than forty years, developing countries have experienced a series of wars and armed conflicts. These have essentially been internal conflicts like in Afghanistan, Cambodia, Rwanda, and Colombia, although they have had a strong tendency to involve neighboring countries. Many such conflicts have affected countries without significantly upsetting the ability of the government to continue functioning. This is currently the case, for example, in the Ivory Coast, even though, paradoxically, this country’s public administration cannot measure up to what it was thirty years ago. In these cases, reconstruction is essentially an issue of political stabilization and 'good' policies facilitating private investment recovery. In these situations, the Official Development Assistance (ODA) plays its traditional role: its goal is to enable these countries, some of which, like the Ivory Coast, have lost several decades in terms of development, to make up for lost time.

2But very often, the conflicts are either the cause or the result of a progressive collapse of the State -- not only a financial bankruptcy, but an administrative, organizational and institutional collapse. In these countries, State institutions, like the police, justice system and army, and financial institutions, like the treasury, customs, taxes and central bank, are dysfunctional. Public health and educational services no longer function properly. The distribution of electricity or drinking water is almost nonexistent. The roads are largely impassable as soon as one leaves the capital, and the transport system is in ruins. The social situation is frightful. The economy has all but shut down. These are what we now tend to call ‘failed’ States, like Cambodia in 1979, the Democratic Republic of the Congo (DRC) and Afghanistan in 2001, and Haiti and Somalia still today. These situations represent regional and even global public challenges for the international community, which does not know how to address these problems and sometimes, by its own interventions, tends to do more harm than good.

  • 2  As illustrated by the terrifying description of Congo's colonization in: “King Leopold’s Ghost” Ad (...)

3The failed States usually collapse for a wide variety of reasons. Often, such as in the DRC, the government structure that implodes was already significantly corroded by its colonial history2, corruption and catastrophic policies. Civil war in these cases merely accelerates an advanced process of decay. The exceptional duration of conflicts that have sometimes lasted for more than a generation, like in Afghanistan, explains the destruction of institutions and the weakness of the government structure. Over these long periods, the disappearance of educational systems has had a particularly severe impact on human capital.

4Over the past dozen years, international aid has been widely mobilized to try and sustainably restore order in these countries, revive their economies and reduce the widespread misery of the population. But these reconstruction tasks clearly present a particular challenge. In spite of the considerable political, financial, economic, technical and military efforts that have been made, representing tens of billions of dollars to date, the results are on the whole very disappointing. Although the DRC is gradually stabilizing, it remains precariously fragile and unstable. But the most troubling case undoubtedly remains that of Afghanistan, which is once again sinking into a spiral of insecurity and war. Unfortunately, in Central Asia as well as in Africa and other regions of the world, other countries are following the same path as Afghanistan; and yet what is most troubling is that in these countries, as we will see later, aid seems to end up being part of the problem, instead of sustainably resolving challenges.

5I will illustrate my presentation with an analysis of a concrete example, which is the failure of the international community’s action in Afghanistan since it began at the end of 2001. I will successively examine the need for a more coherent intervention strategy; the urgency of initially restoring security; the lack of focus that accompanies aid objectives in this context; the lack of strategic management and strategic allocation of resources which are largely wasted; the inadequacy of the conceptual framework offered by the standard poverty reduction strategies developed in the context of the millennium development goals; the harmful effects of the donor community’s short term bias; the critical importance of reconstructing or constructing a modern government structure; and finally, the complexity of the political reforms needed to establish the legitimacy of authority. I will lastly attempt to draw general conclusions and propose more specific recommendations from this analysis for international NGOs.

Real financial resources, and not promises or media hype, are needed to reconstruct a failed State after a conflict

6Restoring sustainable peace in ‘failed’ countries and territories at the end of a civil war is a task that now commonly mobilizes efforts on the part of the international community. This issue has become a major subject of attention; the World Bank will make it the main theme of its World Development Report in 2011. For the NATO partners engaged in Afghanistan, this issue has become critical. Although considerable resources have been mobilized by now, the situation is still deteriorating in many regions or at best stabilizing. Despite progress made under the new leadership of General Petraeus who has at last, but very late in the process, been able to impose a comprehensive and coherent counterinsurgency approach, the failure of the Western coalition (as of early 2011) is now a serious possibility.

  • 3  On this subject, see the extraordinary publication by Ahmed Rashid: “Descent into Chaos - The US a (...)

7An analysis of the events that have occurred in Afghanistan since the end of 2001 shows that the deterioration is clearly related to factors linked to Afghan domestic policies and a mafia system that now corrupts its regime. It is also related to regional policies, in light of the Pakistan army’s ambitions in this country3. But the very serious mistakes that have been made not only by the Bush administration, but also by the entire international community since 2001, go a long way towards explaining the present disaster. These have been traditional mistakes made by a community of donors that have managed aid to Afghanistan according to its usual procedures and that have consequently mishandled considerable financial and human resources. All of us have therefore collectively wasted what was undoubtedly a unique opportunity. Of course, we must harbor no illusions. There will never be a standard magic bullet that, given enough money and technology, could easily put these countries on the road to peace and prosperity. In all cases, the task is challenging, for it is usually necessary to invent and experiment in real-life contexts. However, common characteristics also clearly emerge in these situations and it is important to identify them.

8The first question that should be asked is whether adequate financial resources have been mobilized. This question is important because, under the principles of aid selectivity, aid flows should be directed towards the most efficiently-managed countries. As a result, fragile countries and failed states are often neglected by donors. In most cases, the latter do not become interested until the situation is so degraded that the country reminds the international community of its existence through massacres, terrorism and the regional spread of insecurity, i.e. when it is already too late to easily undertake effective action.

9Another issue is linked to false promises. In the context of the well-publicized donor conferences, the promises of aid that are regularly made by donor countries usually involve the sum of highly disparate elements that finally do not add to much in terms of new money. This is for instance the case of the cancellations of debts that never had any chance of being paid, emergency aid, high cost technical assistance spending or previously-programed aid which was already in the disbursement channels. This means that it is very difficult to accurately determine, based on official announcements, the amount of new money that will effectively be available. This practice of promising aid, which in reality is a form of false aid, is commonly practiced by France, which thereby attempts to conceal the disappearance of its grant resources for its bilateral aid. This being said, the hype related to aid complicates the work of local authorities and increases local public scepticism regarding both aid and local leaders since the population obviously does not see ensuing concrete results from theoretical aid flows.  

  • 4  Note here that for General McChrystal, this military effort was greatly insufficient and has histo (...)

10In the case of Afghanistan, money has not been a major problem, at least … with regards to military spending. Indeed, from 2002 to 2009, the United States spent around $230 billion to conduct military operations. It is currently spending over $200-300 million per day for this purpose4. However financial support for the country’s development has apparently not been lacking either, since from 2002 until the end of 2007, the international community has mobilized some $50 billion, half to reconstruct local security forces (mainly the army) and the other half for development aid. To give an order of magnitude, the sum of $50 billion corresponds to about 10 years worth of World Bank aid for the whole of Sub-Saharan Africa -- while recently discussing these figures with a Congolese Minister, he was clearly stunned by such amounts. To the $50 billion, which are quite considerable, at least on the surface, were added the financing promises obtained in the 2008 donors' conference in Paris and 2010 conference in Kabul, i.e. approximately another $20 to 30 billion. Finally there has also been no shortage of technical support: around $3 billion were indeed spent on civilian technical assistance from 2002 to the end of 2009, and more than 2,000 non-governmental organizations (NGOs) at some time have been working there. For a country whose official GDP (excluding the drug economy) is around $10 billion, the scope of this international support is really impressive.

  • 5  The figures given are in current dollars.
  • 6  According to the OECD DAC in 2006, 44% of aid to Afghanistan was still tied (conditional to purcha (...)

11However, the gap between military spending and civilian development spending is immense. At end of 20075, when the security situation seriously deteriorated, only $14.7 billion in development aid had actually been disbursed out of the $25 billion that had been promised, around $10 billion remaining mere promises or lingering in payment channels. Basically the ratio between military spending and development spending was around 1 to 9, which puts the development aid effort into perspective. We should also note that the quality of this aid, much of which was tied 6 in the case of American aid, which was by far the largest share, left much to be desired. We will see later in this presentation that in the end, the amount of effective development spending was quite small and did not cover many of the most urgent needs.

But there also needs to be a strategy to drive a reconstruction effort

12The first mistake that was made by the international community in Afghanistan, or rather by the highest American leaders (G.W. Bush and D. Rumsfeld, against the urging of Colin Powell, the Secretary of State), was made immediately after the military victory that led to the fall of the Taliban regime in late 2001. This mistake was to de facto subcontract security responsibilities in the country to the Northern Alliance's back-up forces, to which the US had subcontracted land-based operations in 2001 along with ‘commanders’ and warlords who had sworn allegiance to the coalition. Absorbed in the preparation of their Iraq adventure, American leaders at the start of 2002 withdrew a large percentage of their contingent from Afghanistan even though this contingent was already quite small. They also initially limited the security mandate of the International Security Assistance Force (ISAF) to the city of Kabul.

  • 7  Which had been set up by the Soviets shortly before their departure.

13By doing so, they ended up leaving only 8,000 men to ensure the security of a mountainous country larger than France with 30 million inhabitants, which was recovering from more than 20 years of fratricidal war and whose infrastructure was devastated. The ratio of international forces to the population was therefore in the range of 1 per 3,750 inhabitants, knowing that this ratio is even misleading as the mandate of the forces was not to oversee the population’s safety but to pursue members of Al-Qaeda. I will simply point out that this ratio at the end of the Kosovo war was 1 to 40. At the end of the East Timor conflict it was 1 to 112, and at the end of the conflict in Bosnia it was 1 to 205. Such a situation created a dramatic security void across the country. Even worse, the Central Intelligence Agency (CIA), in its ‘war on terrorism’, financed and armed many ‘commanders’, in fact almost anybody who promised help in the pursuit of terrorists, with experience subsequently proving that these warlords mostly used American support and the Air Force's fire-power to settle old tribal disputes or local conflicts of power. Under these conditions, all manners of crooks, including minor and major warlords, rapidly recreated the type of mafia order that had already led the entire country to chaos in 1992 after the fall of the Najibullah regime7, a chaos that had facilitated the Taliban's conquest of the country. American policy, instead of helping to create a modern State, actually strengthened a Merovingian-like system of power that ultimately and very paradoxically corresponded, from an institutional viewpoint, to a regression from Taliban rule.

Ensuring personal safety and a minimum level of local governance" is undoubtedly the most urgent task

  • 8  The Bush administration’s institutional memory loss with regard to end-of-conflict procedures and (...)
  • 9  “Notre maison brûle au Sud, que peut faire l’aide au développement?”, S Michailof, A Bonnel, Fayar (...)

14The fact that no attempt was made in 2002 to establish an international peace force together with a provisional civilian administration and some financial resources in order to urgently launch an initial rural rehabilitation program is quite simply unbelievable. This negligence negated all the experience that had been acquired over the past thirty years in the area of post-conflict reconstruction8. In my most recent book9, I explained how international experience clearly shows that State failure usually starts with the degradation of security and the disappearance of local State authorities (justice, territorial administration, etc.) in outlying regions. I pointed out that this collapse of the structure of government rapidly leads to the establishment of militia that turn to mafias, raise taxes, replace the failing government structure and erode its legitimacy. The Taliban regime had at least widely crushed or controlled the mafias and ensured security and a degree of justice (albeit it expeditious) in the parts of the country that were under its control. In most of the south and east of Afghanistan, a similar development, paired with Western shortcomings in this area, confirmed the failure of the Afghan State, which was already badly off after a few years under the Taliban's obscurant regime.

  • 10  On this subject see: “Cambodia after the Khmer Rouges, Inside the Politics of Nation Building”, Ev (...)

15Here it is instructive to compare America’s negligence of these security, justice and territorial administration matters with the approach that was implemented by Vietnamese leaders to bring peace to Cambodia. The Vietnamese had captured Cambodia in 1979 to topple the Pol Pot regime, which for them had become a nuisance. Like the Americans in Afghanistan in 2001, the Vietnamese made short work of Pol Pot’s army. But unlike the Americans in Afghanistan, they were obsessed with the fear of seeing a Khmer Rouge rebellion resume and encroach deep in the countryside, with the rebels taking advantage of the security chaos to expand and make Vietnamese control impossible. They consequently imposed strict rule in Cambodia's rural areas through a strong military presence and the rapid reconstruction of a territorial administration under their rule. Owing to the lack of Cambodian civil servants (pro-Vietnamese Communists were massacred by the Pol Pot regime), they did not hesitate to call upon Khmer Rouge personnel under their close supervision to assume this leadership10.

  • 11  For a very interesting discussion about the error represented by the 'light footprint’ option, see (...)

16In fact, the 'light footprint'11 principle, i.e. a very minor foreign presence, which was initially promoted by both D. Rumsfeld and the United Nations in Afghanistan, turned out to be a grave error. This principle never kept international institutions' SUVs from congesting the streets of Kabul; but it did preclude the establishment of a regime that would ensure security, local justice and minimal administrative supervision in the field. Of course, the establishment of a large security force and civilian administrative resources would have been an extremely difficult task in this huge and mountainous country. But after the fall of a discredited Taliban regime, there was such demand for order and justice that even a strong foreign presence, if presented as temporary, would have been accepted.

17Over time, as security deteriorated across the country, on account of crime, rackets, vendettas, gangland killing, the violence of small warlords and the return of Taliban forces taking advantage of the chaos, the Western coalition attempted to implement a dual response: military on the one hand, and 'developmental military’ on the other hand. In the first few years, this military action, on which I will not dwell, was essentially undertaken by the American forces in ‘Operation Enduring Freedom’, which operated according to a strict military logic. Moreover, and surprisingly, these forces intervened without any cooperation whatsoever from other American forces and the ISAF allies or Afghan and foreign civilians. They also showed total disregard for Afghan cultural sensitivities. With the general degradation of security, the ISAF was then put back on centre stage and was henceforth directly engaged in the military effort. It is now commonly known that the collateral damage that was caused by this foreign military intervention succeeded in alienating the rural Afghan population.

International forces cannot sustainably meet the need for security and local administration  

  • 12  For example, five Médecins sans frontières (MSF) employees were shot down in 2004 by a local comma (...)

18The action I describe as ‘developmental military’ aimed to fill the security void and meet the most urgent reconstruction needs when the first problems arose in this area. It primarily relied on the establishment of ‘Provincial Reconstruction Teams’ (PRTs). These structures first ensure a military field presence intended to protect the activity of development agencies. Secondly, they directly undertake minor development actions financed with military budgets. The establishment of these PRTs (there are currently 25 PRTs run by 13 different countries) gave rise to lively debates and sharp criticism from the NGOs active in the field. The latter were first concerned that they would be confused with the military units and that the neutrality to which they are rightly attached would be affected, with the additional risk that their personnel would be targeted by the insurrection. They then criticized the non-lasting and non-‘appropriated’ (by the populations) character of the development actions undertaken by military officers, whose expertise in this area is naturally limited. These criticisms are understandable. But there are also few good solutions for undertaking development actions in an insecure environment. Experience, moreover, has shown that the clearly affirmed neutrality of NGOs has not protected them from deliberate attacks12.

19On the whole, contrary to what may be derived from official statements, this program's overall results have been, at least until very recently, extremely disappointing, to the point that when the leaders of some major aid agencies in Kabul are questioned in private, they consider that these programs should be radically reoriented or even shut down in areas where insurrection has not yet become a serious problem. But these operations also continue for reasons that involve the NATO countries' communication policies (what else are they supposed to show to famous visitors?). Perhaps the disappointment stems less from a radical conceptual error than from inadequately defined implementation terms, a tardy launch, long insufficient resources and especially working methods that have led these PRTs to prolong their presence and stand in for defaulting local authorities, while they should have been a temporary solution. So many mistakes have been made related to this type of approach, whose implementation could be extended to other countries and contexts, that I will present a brief inventory.

  • 13  On this subject, see the very perceptive analysis “A Mean to What Ends? Why PRTs are Peripheral to (...)

20Regarding the implementation terms, the first criticism that can be made of this program is its lack of coherence and ultimately its ‘amateurism’13. Lack of coherence as each allied contingent has designed and implemented its own program without paying much heed to the other detachments’ actions. In the PRT program there are therefore British, German Dutch and American approaches, for a total of virtually 13 different approaches. Although there is no French PRT, in the Kapisa valley there is a French approach combining security and development actions that has recently embraced a fairly similar philosophy. Of course, experiments were necessary to develop the most effective approaches in this novel field. But compartmentalization has meant that these experiences could not be shared to obtain a doctrine that would be common to the ISAF's various forces.

21The consequences of this lack of doctrine and the experimental character of the approaches have been worsened by the rapid turnover (generally every 6 to 12 months) of the military units. Each PRT leader, although initially very inexperienced in this area, has implemented his own approach, largely based on his own vision, placing the emphasis on military action, civilian development action, or the protection of his forces resulting in a ‘bunkerization’ logic. Even on the level of security actions, there is virtually no prospect of coherence as the various contingents are subject to different rules of engagement; for example, German forces are not authorized to use force except for self-defense.

22Lastly, these operations have entrusted military officers, who have little knowledge of development issues, with responsibilities for which they were and are inadequately prepared. Most PRTs were and still lack structured civilian technical teams, with rural works engineers, public works specialists, anthropologists, procurement experts and project managers capable of preparing appropriate development actions that could then be managed by the beneficiaries. This criticism should, however, now be qualified as the Americans, Canadian and British have recently been expanding their PRTs by recruiting solid civilian teams; learning from past mistakes, they are now developing a doctrine. Under the ‘civilian surge’ that has been promised by President Obama, the American PRTs should soon be properly equipped with technical personnel, even though there has been no inrush of volunteers.

23The second criticism is that this program was obviously launched much too late. The country literally should have been covered with PRTs from the start of 2002. And yet their gradual installation did not start until 2004-2005, and they still cover the country only partially in 2010. In the meantime, rising insecurity has caused them to interrupt their development actions in hostile regions where their personnel are practically confined to urban centers.

  • 14  In 2004, the budget devoted to American PRT development actions was only $52 million for a total a (...)

24The third criticism is that the PRTs were allotted extremely insufficient financial resources14 for too long, while American aid was largely wasted on large contracts. It is true that since 2007, American, British and Canadian PRTs have been adequately funded. But this is not the case for the PRTs of other nations. And unfortunately in this type of situation, it is not possible to make up for lost time.

  • 15  On this subject, see the remarkable study: “The Taliban’s Winning Strategy in Afghanistan”, Gilles (...)

25The fourth criticism is that the PRTs, by virtue of their nature, could not meet the population’s demands for better local governance. Constant complaints regarding the lack of justice, police violence and corrupted district chiefs were and are still ignored. Western military officers, who incidentally have little knowledge of these problems, cannot participate in their resolution when a fledgling local administration has been established and is supposed to handle them but does very little. This point is clearly acknowledged in General McChrystal's report, but there remains the question of how to resolve it. The situation can be contrasted with the main Taliban groups' surveillance over their parallel administration's behavior, which allows them to dismiss leaders who are too corrupt or display criminal behavior15.

26The fifth criticism underlines that the actions of the PRTs, and of many of the main donors, have taken place without meaningful supervision and coordination with Afghan institutions. Their actions have been developed outside of local budget procedures and without the integration of regional priorities as defined by local authorities. Faced with the frequent failure of Afghan institutions, the PRTs took their place and by doing so weakened the Afghan institutions even further.

27All things considered, we can see that the PRTs, which should have merely been a stop-gap measure, during the 2 to 3 years it would have taken to build capable Afghan institutions and initiate international aid programs, have turned out to be a permanent mechanism. Given the weak institutional progress that has been made in the countryside, it is hard to imagine how they could withdraw without many of their achievements collapsing in the process. The mission of the PRTs, which in reality has remained vague, should have privileged support for nascent Afghan institutions, and particularly the police and local law forces, which Western countries should have been prepared to support and finance. In particular, the PRTs should have supported technically and financed emerging provincial bodies: provincial development committees and especially the Community Development Councils created by the Ministry of Rural Rehabilitation. By widely ignoring these institutions, the approach has contributed to marginalizing them. Once again, these failings are being corrected, but it will be hard to make up for the time that has been lost and for the weakness of Afghan local institutions which remains appalling.

The construction of modern national security and modern local government institutions is essential

28As soon as the decision to intervene in Afghanistan was made, i.e. in mid-September 2001, the American administration should have seriously examined the issue of the country’s administration during the transitional period from the expulsion of the Taliban to the actual establishment of a new power. Let us remember that the American authorities had carefully planned the administration of Germany and Japan in 1945; they had even planned the administration of France in 1944 to the great indignation of General de Gaulle. Such an analysis would have revealed five facts: 1) Once the Taliban were ousted from power, it was necessary to meet local needs for public administration in rural areas. 2) As in all fragile countries at the end of a conflict, demand for security and local administration would be the principal demand from a population that had been traumatized by more than 20 years of war and violence. 3) The demand for security and local administration would require the rapid establishment of an army, a police and local law enforcement agency, a territorial administration and a well funded ministry of rural reconstruction. 4) Since the time needed to create such institutions was at least 3 years, a provisional PRT-type system would have to be established urgently to fill in the most serious gaps and would require the mobilization of substantial military and civilian resources. 5) But the provisional administrative system would then need to strengthen the nascent institutions and plan to self-dissolve and in no case replace failing institutions.

29At the beginning of 2011, more than 9 years after the fall of the Taliban regime, if we review the status of Afghan State institutions that contribute to personal safety and local administration – institutions that should have been operational in 2005/2006 – the situation is disastrous. One can even wonder where, aside from margins and fees for subcontractors, the $25 billion went that were in theory devoted to these tasks. Out of all of the institutions that were needed for the presence of an Afghan State in rural areas, only the Ministry of Rural Rehabilitation and Development (MRRD) was actually created with traditional aid funding. In less than a 3-year period, the MRRD, developed and managed by a minister demonstrating both charisma and managerial capacities, functioned as a modern and efficient institution and effectively covered the entire country. But its activities were under-financed, which was the case for even the remarkable program it quickly established, the ‘National Solidarity Program’, which still provides small grants to over 17,000 local communities.

  • 16  As do the Ministry of Finance, the Central Bank and the NDS, which were also established as modern (...)

30Yet this program is particularly important. In spite of its unfortunately modest financial contributions, it indeed facilitates the establishment of a grassroots democracy through the creation of development committees which manage these grants. The fact that this institution has functioned remarkably well16proves that the establishment of modern institutions in a short period is entirely possible even in very difficult environments. A structured Afghan machinery of government, including at a local level, could have been operational in 2005/2006, precisely at the time when security chaos started encouraging the return of the Taliban. But again, this goal should have been clearly stated and shared by the international community and Afghan political leaders. Instead, lacking alternatives as no international forces were available, both the CIA, which was a major player from 2002-2004, and President Karzai made the political choice to enter into an alliance with local warlords and ‘commanders’, an alliance that is entirely incompatible with the construction of a modern administration. The rest of the international community could not or did not wish to react.

The international community was unwilling to provide the needed long-term commitment to fund an army capable of meeting the expected challenges

31Aside from the MRRD, several other modern institutions were also rapidly established beginning in 2002. First the Ministry of Finance and the Central Bank, thanks to another energetic Minister, Mr Ashraf Ghani, and strong USAID support, as will be explained later. Second the army and the National Directorate of Security (NDS), which is the country’s intelligence institution. The NDS was created by an energetic leader, Amrullah Saleh, a former protégé of Commander Massoud who staffed this intelligence service by recruiting personnel based on merit. This directorate has regularly published highly perceptive analyses of the situation. A fairly well-structured Afghan army was also set up with strong American support; but for a long time, its size remained completely unsuitable compared to the challenges at hand. During the first few years, the forces of certain warlords such as ‘General’ Fahim (who was incidentally Defense Minister at the time) and ‘General’ Dostom, were numerically bigger and better equipped, including with heavy weapons. This army slowly grew starting in 2004 under the impetus of the new Defense Minister, Mr. Wardak. But it was initially limited to some 30,000 men due to the expected Afghan State’s budget constraints, for neither the United States nor other donors wanted to make a long-term commitment to fund it.

  • 17  After the 2003 departure of General Eikenberry who was initially in charge of coordinating US supp (...)
  • 18  At the same time, while the Taliban was installing the first components of its political infrastru (...)

32American support in this area was unequal and limited17. In 2005, Rumsfeld refused Wardak’s request to increase the army to 70,000 men18. At one point, the United States even requested that the Afghan budget cover its cost, which created a panic at the finance ministry. The army now officially has around 120,000 men, but in reality the actual operational force is smaller. Plans are currently to increase the force to 240,000 men as quickly as possible in the context of the new American strategy. But after all the time that has been lost, one can clearly question whether it is realistic to expect this ramp-up to occur in such a short time frame. Military morale has in fact been affected by the general security decline, rumors of American withdrawal, and Taliban threats against soldiers’ families, which means that few Pashtuns are now volunteering and the army is losing is role as an inter-ethnic melting pot. On the whole, the ISAF’s Western armies have had to make the bulk of the military effort, as the Taliban have penetrated the country. By doing so, in the eyes of the population, they have gradually turned into an occupying army, bombing and raiding the country's houses. Since an American GI in Afghanistan costs a minimum of $250,000 per year versus around $5,000 for an Afghan soldier, we understand that the international community’s refusal to initially agree to provide long-term funding for the creation of an army tailored to the country's size and security requirements was at the very least a poor cost-benefit calculation.

The division of responsibilities among coalition countries proved to be a disaster

  • 19  The “Economist intelligence unit” October 2008 report writes: “The state of the police is abysmal… (...)

33Let us proceed with our review of institutions. In the framework of the country's reconstruction, a system for the division of responsibilities among donor countries had been adopted. But the donor countries that were involved had grossly underestimated the responsibilities they were assuming in this process. For example, the reform of Afghanistan's justice system was entrusted to Italy. We can wonder what may have become of Italy's aid for the justice system which is still a disaster characterized by widespread incompetence and corruption. The establishment of a local justice system, which could have easily relied on traditional codes, obviously did not receive the international attention that was necessary. But it is above all the state of the police that is catastrophic19 and poses an extraordinary problem. Germany, which hosted the Bonn conference in 2001, was expected to take charge and restore an effective police force. And yet the police force is commonly considered to be the country's main racketeer and one of the main participants in drug trafficking.

  • 20  This refusal caused Minister Jalali to resign and go into exile in the United States in 2005.

34Here the entire international community is incredibly guilty. The division of responsibilities between members of the Western coalition proved to be particularly disastrous. Germany, poorly estimating the scope of the problem, did no more than provide some training and technical assistance to a force that needed to be completely rehabilitated, beginning with removal of its top leaders and the Ministry of Interior’s management structure. Starting in 2004, faced with an upsurge of complaints about the police's behavior and the weakness of German support, the United States started paying attention to the problem. But in fact, the United States has no federal guard or federal police force capable of providing this type of institutional support. The Americans therefore subcontracted this support task to a private security firm (Dyncorp) with no serious institutional development experience in a country like Afghanistan and whose concern was just to fulfill a technical assistance contract whose objectives were given in ‘outputs’ (number of trained police officers) and not in concrete results. Furthermore, various US administrations fought over supervision of this operation. In 2004, Rumsfeld did not support Interior Minister Jalali who requested to “clean the government, the Ministry of Interior and the administration”. He also did not want to confront President Karzai, eager not to jeopardize his alliances with local warlords and ‘commanders’20. It was not until 2008, six years too late, that this clean-up of the Augean stables was finally undertaken. But the recent removal of Minister Hanif Atmar, who had undertaken this exceptionally difficult task so late in the reconstruction process, and the lack of high-level political will in this area, raises serious doubts as to the likelihood that this task will be successfully completed. Once again, precious years have been lost.

Extremely weak local government structures are colliding with mafias and ‘strong-men’

  • 21  Rumours about the auctioning of district chief positions abound in Kabul.

35The local government structure has also suffered from a lack of financial and human resources, the disorder and corruption of the Ministry of Interior and the nepotism that determined the choice of its leaders. It was not partially brought back under control until 2007, when it was removed from the Ministry of Interior’s responsibility and entrusted to a serious leader. But the choice of local administrators is still far from relying on skills-based criteria21 and the overall human and material resources are extremely weak. In fact, now that some PRTs are properly structured, their actions are hampered by the deteriorating security, the dramatic weakness of most Afghan state institutions, the weakness and corruption of local government, and the still harmful role played by ‘strong-men’ at the local level. The latter are former warlords, drug traffickers or the famous ‘commanders’, leaders of more or less mafia-controlled groups with which the current regime enters into opportunistic alliances.

  • 22  “Afghanistan, au Coeur du Chaos”, Ariane Quentier, Denoël, 2009.

36At a local level, among these “commanders”, many are simply small thugs who terrorize and swindle the population like the famous Algerian ‘emirs’ from 1993 to 1998. At the top, some are real ‘godfathers’ who combine tradition and modernism in typical Francis Ford Coppola style. I will always remember the royal welcome that was reserved for my delegation by Mohamed Atta, who incidentally was appointed governor in his Northern fiefdom. He had two satellite phones and a lap top within reach, and we departed for lunch in his residence/palace in a column of Mercedes SUVs with tinted windows and pick-ups filled with armed men. I did not have the bad taste, given the quality of the reception, to ask him if it was his governor’s salary that permitted this lifestyle … As Ariane Quentier summarizes in a remarkable eyewitness book22: “by deciding to bomb from above while delegating military operations below to Afghan middlemen, by refusing to deploy troops, by arming warlords while encouraging anti-drug agencies to loosen their control, the United States played with fire and was over-confident … It also gave the new Afghanistan a terrible signal by placing trust in very disreputable individuals”.

  • 23  This figure incidentally corresponds to the numbers that were rapidly reached by the Algerian army (...)

37International experience has helped define ratios for establishing security in a country with a high risk of civil war resurgence, which was clearly the case of Afghanistan in 2001. This ratio is around 1 security agent (police officer, military officer or militiaman) per 50 inhabitants, which implies a need of about 600,000 men for a country like Afghanistan23. The new US administration has clearly understood the urgent need to quickly build up a strong local security force and the total number of police and army forces has increased from 120 000 in July 2007 to 260 000 by November 2010 which is quite an achievement, even if we car harbor serious doubts about the quality of the force. It is quite clear that the Afghan budget will never be able to maintain a 600,000 or even a 300,000 force unless oil is to be discovered! But the international community’s refusal to confront this type of problem on time and the need to fund the establishment of security and local administration machinery suited to the context, largely explains the situation we are currently faced with. The Obama administration and General Petraeus have of course rediscovered these hard facts at the end of 2009, just as the Nixon administration had discovered them in Vietnam after the 1968 Tet Offensive. But politically, in view of the prevailing weariness, the 4 to 5 years needed to beef up these institutions are lacking, while the Taliban in the meantime has become a force with such a high capacity for disruption that a successful outcome is in serious doubt.

  • 24  Note that the countries that supply the troops, and not the wretched soldiers, pocket the tidy sum (...)

38It should be noted that, today, the same unwillingness from the international community to confront hard facts is present in the DRC. This country’s stability is first threatened by its army’s total lack of discipline and organization, this army having been formed by the aggregation of multiple militias and gangs of thugs. The overall goal, by integrating them was to control them somewhat and by paying some salaries, to reduce their intrinsic harmfulness. Rather than properly fund and urgently train a “gendarmerie” type force and a modern disciplined Congolese army, this same international community prefers to pay more than a billion dollars per year to finance the 16,000 unmotivated mercenaries24 from MONUC, the United Nations force, whose discipline and ardor in combat are so mediocre that they are incapable of restoring order in Kivu. For want of resources and out of lassitude, the international community is going to reduce MONUC's size and may withdraw it before the problem of the Congolese army and police’s discipline and efficiency has been resolved. The same analysis today should bring the international community in the DRC, which is still focusing on poverty reduction and particularly on the social sectors, to radically revise the order of its priorities. Here again, the construction of a modern army and a modern national guard, a respectable justice system and of course a road network in a country where 10 of the 11 provincial capitals are connected to the capital by air only, should be given a higher priority than anything else. But do donors in these countries have a strategy based on common sense or political correctness?

In fragile countries, there is no pilot in control of international aid

  • 25  For a more detailed analysis of these issues see: “The Seven Capital Sins of the International Com (...)

39We are in fact beginning to understand, through these first examples, that the international community's activity in a country like Afghanistan (but the situation, although much less serious, is fairly similar in the DRC) scarcely responds to a clear strategic vision under the authority of a clearly identified coordinator25. First, cooperation between civilians and military forces is extremely limited, with military forces ‘waging war’ and with civilians not daring to become involved in institutional construction in areas that are not their usual spheres. In fact, they share no common objectives, as the former, at least until the arrival of the new American administration, have been busy killing or capturing ‘terrorists’ while the latter are busy fighting poverty.

40But beyond the thorny issue of civil-military cooperation and the lack of an overall strategic vision which would include both security and development issues and that only General Petraeus has apparently been able to forcefully put forward, there is also the singular problem of strategic vision and oversight in purely development-related areas. What, concretely, is actually happening? First, contrary to common opinion, aid is effective when it concentrates its actions and coordinates them properly in specific areas. It is not by chance that currently, more than 7 million Afghan children go to school and that progress is now noticeable in public health. It is also not by chance that the aid resources administered by the trust fund that was established by a group of donors and is managed by the World Bank (the ‘Afghan Reconstruction Trust Fund’ or ARTF) in accordance with revised Afghan procedures are producing satisfactory results in multiple areas. Aid is indeed a rare asset that should be used with the goal of maximizing its impact thanks to a careful selection of those sectors and zones that are critical to reach a set of objectives. Nonetheless, while laudable efforts have produced encouraging results in the education and health sectors, the fact that there is no overall strategic management of international aid resources is plain for all to see. In fact, neither its breakdown by sector nor its geographic breakdown appear to respond to what seem to be the most pressing priorities.

  • 26  This issue is now finally being corrected. During his conference at Sciences Po on Nov 23, 2010, G (...)

41Let us first examine the sectoral breakdown: Seventy percent of the Afghan population is rural but out of the almost $15 billion that were actually disbursed at the end of 2007, only around $500 million were disbursed in the agricultural sector26. It can indeed be said that to develop agriculture in a country like Afghanistan (but this is just as true in the DRC), the first measure to be taken before placing too much money in production support is to develop and maintain a network of rural roads, as the rural world's isolation is a problem that is as serious as the problem of security. But while donors in Afghanistan have effectively, by dint of millions of dollars, rebuilt the main strategic road network that was initially established by the Soviets in the 1950s and 1960s, rural roads have until recently been largely neglected. For example, out of the 50,000 km of rural roads that would essentially be necessary to connect all of the country’s villages to the provincial capitals, only 9,000 Km had been built or rehabilitated as of the end of 2007.

  • 27  See “Afghanistan: Economic Incentives and Development Initiatives to Reduce Opium Production”, C W (...)

42At the current rate of construction/rehabilitation and destruction due to a lack of maintenance, a rapid calculation performed by the World Bank in 2008 suggested that it would take one century to link all of the country's villages to its provincial capitals. At the present time, the military is making major efforts to build rural roads. But the selection criteria are related to military and not economic considerations; furthermore, no serious maintenance mechanisms are in place. All in all, the isolation of the Afghan rural world still remains dramatic in most mountainous regions, as isolated valleys cannot enjoy the benefits generated by a gradual development of basic education and health services. This isolation also prevents them from participating in the national economic recovery, from entering trade channels, and from benefiting from the recovery of the transit trade. There is, however, one exception, which is the opium economy. Opium’s high value in relation to its weight means that it is possible to bypass roads and simply use backpacks or small donkeys27.

There is no coherence in aid resource allocation

  • 28  On this subject, see the study: “Immediate Priority Needs of Vulnerable Farmers engaged in Opium P (...)
  • 29  Hajji Fazul Rahim, quoted by the New York Times, 28 November 2009.

43In addition to the lack of massive investment in the construction of rural roads in the 2002-2007 period, a similar lack of massive effort in the rehabilitation and extension of irrigation networks is just as outrageous. These networks have in fact been severely damaged by the Soviets under their scorched earth policy and by a lack of maintenance over the past 20 years. As a result, many Afghan valleys are confronted with a Malthusian crisis. The already staggering natural population growth has been aggravated by the return of most of the 5 million people who had sought refuge in Pakistan and Iran. This rural population has to cope with both a lack of land and its low potential, for want of irrigation and road investments that are necessary to bring in inputs and evacuate surpluses. Young people cannot find work on-site. They are forced to crowd into cities, where they increase the mass of the urban unemployed, or participate in the drug economy, or join the insurrection that offers attractive salaries. The oldest have to grow poppy, which alone ensures rural families a minimum level of monetary income and guarantees that they will have the capacity to buy cereals in the event of a food shortage28. An Afghan tribal chief was indeed recently quoted in the New York Times: “Most of the Taliban in my area are young men who need jobs. We just need to make them busy. If we give them work, we can weaken the Taliban”29.

44Another example of serious incoherence in the sectoral allocation of international aid is the under-investment in the energy sector. It is known that after agriculture, electricity is one of the necessary bases of economic development. And yet what is the current reality? Nine years after the fall of the Taliban regime, most of Afghanistan is still immersed in darkness. Even in Kabul, the Energy Minister, Ismail Khan, a major warlord (who, in fact, still controls the Herat region), was unable to keep the promise he made to supply 8 hours of electricity per day for the 2008 Ramadan period. It was not until 2010 that Kabul was finally provided with electricity. How could one possibly hope to attract investors without electricity? And how can the Western world claim to be seriously committed to helping this country if Kabul's poor neighborhoods, which cannot afford private generators, remain immersed in darkness every evening at 7 o'clock? What a sign of failure!

45What about the geographic allocation of aid? This allocation followed two types of logic. First, aid was heavily concentrated in Kabul where, in the 2007-2008 Afghan fiscal year, it reached the exceptional sum of $600 per inhabitant. Secondly, it adhered to combat zones. In this same year, it indeed reached $450 per inhabitant in the Nimroz and Helmand provinces, which were heavily affected by combat, and where on the pressing demand of the Western armies engaged in the field, it attempted to win over ‘hearts and minds’. In more peaceful provinces, it remained minor, with for example around $50 per inhabitant during this period in Wardak. The money has therefore been largely wasted on zones where fighting precluded any real development, and this policy did not focus on rebuilding regions at peace. It so happens that the ‘peaceful’ region of Wardack has since switched over to insurrection. Many villagers have noted that only insurrection in fact seems to attract aid.

46Last criticism: Most aid short-circuits the Afghan authorities. The problem is surely not simple. The billions of dollars that slowly circulate through the channels of international aid remind us of the challenge of rapidly and efficiently using donors’ dollars in a country where technical and institutional capacities have collapsed. Even when rapidly disbursed, aid can end up blocked in the channels of the Afghan administrative machinery, become lost in these channels, or take years to reach its recipients. In these circumstances, international aid needs to make a trade-off between the short and long term. For very short-term effectiveness, the American methods of contracts signed without calls for tenders, on the fly, with companies capable of implementing projects in tight time frames, are understandable. It is very expensive, there is a worrying lack of transparency, but in the end, it produces results, like large roads. The Chinese, incidentally, use a similar practice.

47But this approach has considerable disadvantages. It can be used for little other than major infrastructure work and, as we noted above, there are hidden costs related to this practice. The worst thing is that no local institutional capacities are created, particularly to ensure result sustainability. Who, for example, will perform maintenance work and with what resources? Furthermore, local authorities are completely short-circuited and powerlessly observe the passing of bulldozers. This on-the-fly approach, moreover, is of course totally unsuitable for the implementation of a complex rural road program or a decentralized irrigation program, actions which require thousands of detailed technical studies and hundreds of thousands of hours of discussions with villagers. Only local institutions, initially supported by NGOs, can manage such operations.

A “manager” is needed to try to manage ‘unruly’ donors

  • 30  See “Responding to Afghanistan’s Development Challenges”, William Byrd, World Bank, 2007.

48As far as international aid is concerned, there has been no shortage of commitment or individual talent. There also has been no shortage of high-quality analytical studies to help with decisions, and it is therefore not because of a lack of information that the right decisions have not been made30! The lack of logic in this sectoral and geographic allocation of aid resources stems from a total lack of strategic management. The problem is both institutional and political. It is clear that the needed oversight should be entrusted to a top leader in the Afghan government. But as a general rule, few governments in so-called “fragile countries” have the technical capacity not to mention the political authority needed to impose discipline and a coherent strategy on donors. In a country that has been as weakened as Afghanistan, this task was even more difficult since the governmental architecture (lack of a Prime Minister and weakness of the President's services) was problematic.

  • 31  Poverty Reduction Strategy Papers, prepared by governments of aid recipient countries to determine (...)

49It is true that the first Minister of Finance, Mr Ashraf Ghani, who was perfectly able to manage foreign aid in a coherent way, attempted to do so during his tenure. But he had no formal mandate and he antagonized many in his attempt to impose a minimum of order in this area. His departure in 2004 left the management of the aid system adrift. It is also true that teams of consultants prepared so many strategy papers that it is easy to lose track. But these were formal exercises intended for donors’ conferences (London in 2006, Paris in 2008, Kabul in 2010) and it is in this respect a pity to see that the bible in this area, the ‘Afghan National Development Strategy’, which was greatly inspired by the PRSP31 approach, contains some forty priorities and is a catalog that essentially takes care to use the politically correct vocabulary expected by donors… Only for the 2010 Kabul donor conference, the set of objectives was finally reduced to a more manageable number of 10 priorities.

  • 32  Millennium Development Goals. These eight broad goals were adopted by the international community (...)

50Under these conditions, aid institutions tend to behave like consumers in a supermarket. National strategy papers provide them with long lists of needs. Donors then choose the needs from this long list that best correspond to the wishes of their constituents and pressure groups. Their institutional requirements, and not the country's most pressing needs, determine their action. As their MDG’s32 determined conceptual framework is still poverty reduction, they focus on the social sectors of education and health, forgetting by the way the urgency of building rural roads, which nobody is interested in; the rehabilitation of agriculture and irrigation networks, both terribly complicated subjects; energy, a sector that has been ‘polluted’ in Afghanistan by the presence of a major warlord at its reins; not to mention the restructuring of the justice system, of the Ministry of Interior and the district and provincial administration, etc. These areas are indeed complex. But they are only insurmountable if each donor tries to address them individually instead of collectively, without having the political courage to raise the issue of some ministers and governor’s’ extraordinary incompetence or corruption. In the end, each donor does what it wants, and in the very words of an ambassador in Kabul, “international aid is as difficult to manage as a herd of cats!”

51But aid also lost interest in some sectors when trade-offs had to be made. Indeed, out of the approximately $15 billion that were actually disbursed between 2002 and mid-2007, if we subtract the shameless margins of some companies that managed a large part of American aid, and technical assistance spending, less than about $10 billion probably remained to meet the needs of a country the size of France, where absolutely everything needed to be rebuilt. In fact, contrary to what might be thought, this is quite a small sum. It is in any case very small compared with the amounts that were spent on military operations during this period, the real ratio being closer to 1 to 13 than the 1 to 9 initially mentioned. Clearly another balance between military spending and civilian development spending would have been warranted.

The generous principles defined in the Paris and Accra aid effectiveness conferences are inoperative

52Under these conditions, the well-known principles of the Paris 2005 and Accra 2008 conferences on aid effectiveness are largely inoperative and almost naive. In fragile countries, aid needs in fact to be able, if needed, to manage itself. But when aid mobilizes powerful players like the United States Department of Defense, USAID, NATO, the World Bank, and the Asian Development Bank, traditional coordination mechanisms are unable to define and impose priorities that everyone will respect. There is no lead manager who can take charge. Although the United Nations may have the legitimacy to do so, and UNDP always attempts to take the lead, neither the UN nor its UNDP arm have the technical capacity and the political clout to do so. Finally the United States, which could have assumed this responsibility, was unavailable due to its mobilization in Iraq.

53On the whole, this drastic inability of international aid in Afghanistan to develop a mechanism allowing a clear strategy to drive its overall action now means that radical reforms need to be undertaken for future interventions in other difficult regions. There obviously needs to be a pilot and a clear strategy that all donors must follow. America’s first ambassador in Kabul of Afghan origin, Zalmay Khalilzad, who was perfectly aware that the country’s reconstruction required the establishment of strong State institutions, was certainly capable of exercising this type of responsibility, on account of his authority and knowledge of the country. But the Bush administration sent him as ambassador to Baghdad when the security situation in Iraq collapsed in 2005, clearly showing where the US placed its priorities. In 2007, the British, deeply worried by the disorder in Western aid efforts in Afghanistan, proposed that the former United Nations High Representative for Bosnia, Lord Ashdown, a former high ranking military officer with a strong personality, coordinate all military and civilian aid to Afghanistan, thus providing the needed strategic guidance. But President Karzai refused, seeing Lord Ashdown as a kind of colonial governor who was being imposed on him. This issue is clearly sensitive.

  • 33  Maybe today, no one wants to endorse the responsibility of the coming failure…..

54And yet it is obvious that managing huge aid flows in such countries requires a high-level coordinator with the requisite authority. Apparently, only a high representative of the United Nations Secretary-General can have the legitimacy to ensure such coordination. But for him to impose his authority, in addition to a strong personality, he would need donors to reach a consensus about the principle of having a clear common overall strategy driving the global aid effort, which for the moment is not at all the case; he would also require a very solid technical team to assist him, a team that in all likelihood, only the World Bank could easily mobilize. The implementation of such a duopoly between the high representative of the United Nations Secretary-General and a technical team from the World Bank seems unavoidable in this type of situation, as I suggested to the World Bank President, Mr. Robert Zoellick and his management team. I admit that I am confounded by the fact that such a solution has not been adopted in Afghanistan33. The high representative of the United Nations Secretary-General himself should report to the high representatives of the various countries involved in this support, drawing from the model introduced by the Obama administration.

  • 34  As such, a project that falls under one of the jointly defined priorities and is financed by aid c (...)

55The simple presence of an aid coordinator enjoying extensive power would certainly force the government to reorganize itself to gradually take over control and strategic management of aid resources that finances all public investment and most State current expenditures. Such coordination would require imposing priorities and discipline upon donors and would imply the allocation of most international aid to budget instruments like the ‘Afghan Reconstruction Trust Fund’, which was mentioned above. For project aid, it would be possible to rely on incentive tools, such as multilateral trust funds to top up the financing of donors that effectively adhere to the set priorities34.

The conceptual framework offered by the MDGs and the standard PRSP approach is not suited to the present context

56This approach would also entail the preparation of a common strategy shared by all donors and the definition of joint priorities together with local authorities. This type of approach would thereby force them to clearly specify their actual priorities, instead of contenting themselves with mixed-bag lists. In the current context, each donor individually prepares its own strategy in a conceptual framework that is still governed by the MDGs and the PRSP poverty reduction logic. However, in a country like Afghanistan, this laudable objective can only be the consequence of achieving a much more immediate goal, which is the reconstruction of a functioning government machinery. The ideology which, since the Millennium Development Goals, has treated aid as charity work, has particularly clouded the judgment of donors in Afghanistan, whereas an approach based on realpolitik – when faced with a failed State it is first necessary to rebuild the machinery of government - would have been much more realistic.

57In 2007 in the DRC, the World Bank prepared, in conjunction with other funding agencies, and particularly the European Union and the United Nations institutions, a common assistance strategy to which almost all the country’s donors gradually adhered. This exercise is therefore entirely possible and necessarily summons local authorities who need to organize themselves accordingly to respond. But in the DRC, the World Bank was solely able to act as honest broker between donors whose objectives were still determined by their own preferences and not by the country's objective needs. In the end, this exercise, which was useful on account of the discussions it sparked among donors and between donors and the government, did not result in real discipline capable of directing aid flows to sectors that had been neglected by donors, particularly in the area of basic economic infrastructures such as roads. This inability from Western donors to extricate themselves from the MDG philosophy and their now favorite sectors, which are health and education, gave Chinese business in the DRC an ideal opportunity to propose massive (and very expensive) road financing that at one point almost jeopardized both the support and reform program that was then under negotiation with the IMF and the reaching of the HIPC completion point.

The creation of parallel administrations by donors ends in disaster

  • 35  The context is similar in Cambodia and DRC.

58We just saw that when common objectives and strategic aid management are lacking, aid is dispersed in areas and sectors that do not correspond to the most obvious priorities. Apart from the lack of strategic management of aid resources, traditional intervention methods also contribute to aid ineffectiveness in very fragile countries. We know in fact that in these countries, the ineffectiveness of public policies and therefore of international aid stems from the dysfunction and corruption in local institutions. In failed states like Afghanistan in 2001, the problem was even more basic since not only had its institutions disappeared, but after years of war and civil strife, so had its qualified personnel35.

  • 36  PIUs, a term that is systematically used in donors’ technical documents.

59In a context where there is a severe shortage of qualified personnel, donors, whether multilateral and bilateral agencies or NGOs, have therefore proceeded in Afghanistan as they usually do in this type of situation, i.e. they have established what are called ‘Project Implementation Units’36. For each project that they fund, they recruited the few available Afghan technical personnel from exile in Pakistan or Iran. Lacking budget constraints, and faced with a shortage of such professionals, the salaries rapidly climbed and ended up following the United Nations salary structure (the UN, incidentally, has become one of the country's leading employers). The scope of aid financing was such that starting in 2003-2004, a de facto parallel administration financed by donors was therefore established. This parallel administration, which numbers about 120 000 (the Afghans speak of the ‘second civil service’), manages hundreds or even thousands of miscellaneous operations when NGOs' projects are taken into account. In Afghanistan, this system’s flaws, which are visible in all fragile countries, have become exceptional. The salary differences between the aid-funded project administration and the public administration that is simultaneously being built are at least 1 to 10 and can reach 1 to 40. Technical project managers who would earn $150 per month in the public administration are commonly paid $800 to 6,000 by donors.

60This approach has considerable drawbacks. Not only is the State's bureaucracy having a difficult time modernizing, but this system drains the public service of its residual human capacities as well as those it is trying to train. Indeed, as soon as technical personnel are trained, they flee to this parallel administration to benefit from its exceptional salaries. And yet the project implementation units are not long-term structures. They disappear with project completion and the end of corresponding foreign funding. This parallel administration is therefore in a constant cycle of capacity creation and destruction. No lasting institutional capacity is built. For want of an alternative, aid coordination and the definition of sectoral policies, which are the responsibilities of central administrations, also have to be entrusted to donor-funded project implementation units. When we travel up the hierarchical chain we see that many high Afghan leaders, even at the ministerial level, also belong to such project implementation units to collect the corresponding salaries. The system is totally unstable and can only last as long as international aid agrees to pay.

Building modern institutions and ultimately a modern government structure is a necessary condition to establish lasting peace

61In most central administrations, where there is a minimal level of essential work to be done, the leaders do not have available Afghan technical personnel. Since foreign technical assistants are perceived as a free good, the easiest approach is to use technical assistance to make up for the lack of competent local staff. Therefore, taking into account the high costs of international consulting firms and security fees, government agencies and ministries commonly pay $30 to 40,000 per month to consulting firms providing experts in charge of work that indeed cannot be done by local staff paid $150, but would be advisable to entrust to Afghan professionals paid $500 to 800. These foreign experts work in a country where they do not understand the languages and the culture, where they do not have local counterparts, where local managers cannot define or monitor properly their work, and where their loyalty to the Afghan authorities is uncertain since their careers depend on the international consulting firms that employ them.

  • 37  Review of Technical Assistance and Capacity Building in Afghanistan, discussion paper for the Afgh (...)
  • 38  “Rethinking technical Cooperation, Reforms for Capacity Buiding ” Elliott Berg coordinator, UNDP, (...)

62This technical assistance is presently posing a dual problem: A financial problem since its cost, which reaches now around $500 million per year37, represents about twice the cost of civil service salaries; and a political problem, since its overbearing presence is increasingly being resented. Furthermore, this technical assistance cannot train local professionals to replace it since Afghan staff leave as soon as they are trained. The result then is an impasse, with criticism being made of the technical assistance whose quality is deteriorating as security conditions themselves are worsening. But nobody, including donors and high Afghan leaders, knows what else to do. This problem, which has become critical in Afghanistan, is not specific to this country. In the DRC, government leaders are currently facing a similar problem as they attempt to form the technical teams that they urgently need. I was recently on a mission for the Cambodian government, which was faced with the same type of challenge. It is astonishing that far from being new, these issues were already present in Africa in the 1980s and that appropriate solutions had already been devised in the early 90s38.

  • 39  Paid by the British in the 19th century, the Soviets and Americans from 1945 to the 1970s, then on (...)

63One solution could entail trying to rapidly increase the public sector’s salaries. But the weakness of the tax ratio in a country that paradoxically has always lived off its geopolitical income39 means that the necessary fiscal resources cannot be mobilized. In this respect, the government depends on the good will of donors who nonetheless are reluctant to make long-term commitments, which are essential in this field. A hasty salary increase would also feed significant inflation which is already being provoked by the injection of large amounts of aid, already fueling a standard ‘Dutch disease’ syndrome. Furthermore, while salary increases are essential to attract and retain valuable professionals, they are far from being sufficient to ensure this administration’s quality.

64In addition to acceptable wage conditions, in order to build an effective government structure, at least five other conditions indeed need to be met: 1) recruitment of leaders must occur on a merit basis, 2) promotions must not be separated from performance, 3) the organization of State institutions must follow basic rules of functionality, 4) work must be governed by clear rules and procedures, and 5) the personalities at the head of these administrations must demonstrate qualities of leadership, organization and integrity. It is therefore necessary to undertake a comprehensive reform of the public administration in order to rebuild the government structure. This task is extremely ambitious and difficult since it needs to be implemented rapidly and on a large scale. It is also politically challenging.

Lack of donor interest and political logic combine to avoid addressing these issues

65The construction of modern government institutions in Afghanistan as in all fragile countries is therefore essential. Leaving the responsibility of hundreds of projects to international donors or to PRTs cannot be a reconstruction policy. Neither is subcontracting the development of rural areas to NGOs that are not supervised by coherent national programs. The importance of reconstructing the government structure was initially recognized both by Afghan authorities and some funding agencies. We already noticed that Mr. Ashraf Ghani, who was Finance Minster in the early 2000s, led a complete reorganization of the Ministry of Finance and the Central Bank thanks to significant USAID funding. To do so, he recruited teams of Afghan professionals on consulting contracts who were supervised mostly by US technical assistants. He was successful in restructuring, customs, taxes and the treasury; he secured the payment chain of government expenses, entrusted government procurement responsibilities to a foreign entity, blocked corruption at this level, started using the budget as an economic policy instrument, and implemented a monetary policy worthy of the name. Dismissed in 2004 because he was becoming an obstacle to several powerful players, his actions to build an institution nevertheless weathered the course of time. The system remains fragile, as the status of the Afghan consultants that make up the ministry's backbone remains uncertain and the qualities of his successors may vary. If, however, Afghanistan had been able to both benefit from two dozen ministers of this caliber as early as 2002 and mobilize the billion dollars needed from donors over a 3- to 4-year period to rebuild the main State institutions, aid would have been used infinitely more effectively. The country would have had a serious government structure by 2005/2006 and would not be in the dramatic situation it is in now.

66In addition to the headache caused by the existence of two parallel administrations with incompatible wage structures, the official government administration and the project implementation units financed by donors, the five aforementioned conditions are met only in the few modern institutions that I have already mentioned and have been created since 2003: the Ministry of Rural Rehabilitation which was a showcase example, the National Army (ANA), the National Directorate of Security, the Ministry of Finance, the Central Bank and gradually the Education and Health Ministries. The Afghan drama is that the other key institutions critical for stabilizing the country, particularly the police, justice system and local government, have remained traditional administrations of a failed state governed by nepotism and corruption.

  • 40  The corresponding study is comprehensive and remarkable. It corresponds to the equivalent of sever (...)

67There are three main reasons for this dismal situation: 1) The lack of donor interest in institution development for want of a coherent strategy in spite of the availability of high-quality analytical studies40. 2) The resistance of traditional power structures, taking into account that in Afghanistan, like in most fragile States, institutions are distributed to political “friends” as booty in order to consolidate the political alliance process. 3) The political alliances encouraged by the Bush administration on which President Karzai's power is based.

68Under this logic, political-ethnic networks are naturally formed within each institution, and membership in these networks, and not merit, determines recruitment and promotion. This logic is very strong and contradicts any goal of effectiveness. Ultimately, in most cases, ministers with sometimes only very basic training blindly manage institutions whose missions are vague, whose organization is defective, whose internal procedures are nonexistent, and whose supervisory personnel are chosen based on ethnical and political criteria. This system’s obvious flaws increase the distrust of donors that are terrorized by corruption risks. They confirm their belief that only the project implementation units that they select, and whose personnel they pay, can implement the projects and programs they fund. The circle is now complete.

69What is remarkable is that in spite of these gigantic hindrances, some Afghan institutions have successfully organized themselves, recruited a nucleus of valuable professional staff, implemented policies that are gradually becoming coherent, and managed relatively effective action programs. To so do, they were assisted by various aid programs to build the capacities of Afghan administrations. Unfortunately, these programs were established in a way that was too limited and with too few resources.

70Given the political conditions at the time, at least until the 2004 presidential election, it was undoubtedly necessary to appoint a number of men with dubious track records at the head of certain ministries, for this was where they could do the least harm. The number of men with the authority and experience to oversee an ambitious modernization process was also certainly limited. But at least donors should have defined a clear strategy in this area, proposed it to President Karzai, released the billion dollars required to rebuild the government structure and offered the Afghan authorities a clear deal on the following basis: “find men and women with the necessary leadership qualities and we will give you the dollars you need to rebuild your institutions”. This deal was not offered. Instead, a program with an insufficient reach and a fragile design was implemented, the Priority Reform and Restructuring Program (PRR).  

This reconstruction of State institutions is possible even in very difficult environments

71This approach was developed on the basis of the Independent Administrative Reform and Civil Service Commission (IARCSC), which was created under the Bonn Agreement in 2001. The goal of this institution is precisely to facilitate the creation of a modern administration by selecting civil servants on the basis of merit only. This institution, which was established with the help of several donors, focused its efforts, for want of resources, on a limited number of key departments in the ministries deemed the most important. On this occasion, it introduced bonuses in order to offer incentive salaries based on the job market’s conditions. The approach was intended to be temporary, pending a new salary scale whose budget cost obviously presented a serious fiscal problem.

  • 41  From 2003 until 2006 for all of the approximately 30 ministries and concerned institutions combine (...)

72But this initial approach rapidly became blocked by multiple factors: Unequal salaries in the same institution between “PRR” and “non-PRR departments’ weakened motivation in non-beneficiary departments and created unmanageable tension. Technical and financial support from funding agencies remained too feeble41. Merit-based selection directly clashed with traditional practices based on the distribution of jobs to ‘friends’. Numerous ministers viewed it as a fundamental hindrance to their power. In the first few years, the IARCSC selection committee made a number of blunders and selection errors. And last but not least, apart from a small pilot project, the program was not extended to the local government which should have been its main beneficiary.

73In the face of mounting criticism, the government chose to forge ahead in 2005, extending the PRR to all the country’s ministries and departments. But not surprisingly, the lack of resources blocked the experiment which now only concerns about 31,000 staff out of a total of 350,000 civil servants. Note, to avoid ambiguity, that the cost of salary surpluses for an ambitious reform program would have represented only a fraction of the annual cost of technical assistance.

74The approach then focused on a general revision of the civil service’s salary scale and a reclassification of personnel thanks to a ‘Pay and Grading’ program which again ran into resource constraints. This program now benefits about 16,000 staff. Since then, other specific ad hoc schemes funded by different donors have proliferated and provide bonuses and top ups to civil servants and fund local consultants in what has now become a completely chaotic system. I have identified at least 9 different schemes providing such bonuses which may reach up to $12,000 per month for presidential appointees.

75Altogether an unspecified number of civil servants and local consultants benefit from such schemes lavishly paid by different donors. Since there is no coherence in the system because the different schemes are funded by different donors in an uncoordinated way, the wide differences in treatment now fuel resentment and jealousy among beneficiaries. Note that this approach does not settle the problem of the government’s overall effectiveness unless the other 5 points listed above are addressed. The key problem that should have been raised initially is the dual lack of Afghan leadership and strategic oversight of international aid, which never devoted the required attention and money to this major topic.

76Conversely, the MRDD sets a good example of what should be done. First it is necessary to systematically define the institution’s missions and objectives, carefully select a small team of high-quality senior professionals, and offer them compensation that corresponds to the job market. Develop the institution on this basis, by recruiting properly paid, high-quality personnel at each level. Define the organization, responsibilities, procedures, and working methods. Then mobilize financing, insofar as possible, in the form of national programs, where donor money melts into a common pot managed by the ministry. Prepare projects that are then, in the framework of these national programs, widely implemented in the field by closely supervised NGOs. And lastly, discipline international aid so that it complies with the basic principles of the Paris and Accra conferences on aid harmonization.

  • 42  Or rather, unfortunately, that “worked”, for recent management changes seem to have seriously weak (...)
  • 43  On this subject, see the fierce assertion by one of America’s best experts on Afghanistan on the n (...)

77The Afghan leaders and the international community had and still have an administrative reconstruction model that works42 right before their eyes. But there was lacking, and there still is lacking, will and real leadership on the part of the Afghans’ highest authorities. First is needed a political decision to exit opportunistic alliances with ‘power brokers’, ‘strong-men’, former warlords, clan leaders, drug traffickers and mafiosi, in order to build a modern State43. On the basis of such a political choice that can scarcely be imposed from the outside, all that would have been needed was about a hundred capable leaders with free hands who, in key posts, would have made a difference and cleared inevitable blockages and obstacles.

  • 44  On this subject, see: “Sortir du Piège Afghan”, Serge Michailof, Commentaire No 126, Summer 2009.

78Indeed, under the stress of war, some leaders of this type are now being appointed to important posts. Now is the last time to confirm this experiment which still remains too limited. There is also lacking, from donors who have also now largely lost their faith, a real will to depart from business as usual, to leave aside the habitual concern for party loyalty and accept a discipline where aid can be managed according to clear strategic objectives. A main leader on this end will therefore be needed to provide the missing coherence. In the present mess, I tend to believe that only the US can provide such leadership as they de facto do in the military area. In addition, more money will certainly be required. But even more importantly, time is required to make up for all the time that has been lost44Basically, from 2002 to 2005, Afghanistan could have made the choice to modernize its State. The Karzai regime and its Western backers ended up refusing this modernization. This unfortunate choice having been made, the country was returned to its demons.

The sustainable construction of democracy requires a bottom-up approach

  • 45  See “Notre maison Brûle au Sud”, cited above.
  • 46  “Afghanistan’s 2009 elections” William Maley (Asia Pacific College of Diplomacy), Afghanistan info (...)

79I would now like to address a subject that is outside my direct area of expertise and that goes far beyond the Afghan framework, but that cannot be kept silent -- and that is the democratic model that the international community is currently and systematically imposing on these post-conflict countries. I have already expressed my skepticism45 with regards to the ‘appeasing’ character of a top-down democratic process that is limited to presidential and legislative elections imposed by the international community upon unprepared social structures. In fact, while this approach undoubtedly puts an end to a political impasse with the first election, there is a strong risk that things will go wrong in the next elections, 5 or 6 years later, when the team in power that now controls the revenues and rents associated with its position inevitably refuses to surrender its place. The disastrous presidential election at the end of 2009 in Afghanistan, like the legislative elections in 2010, highlighted this type of problem, which also created a dangerous split in the political alliance that was agreed to in 2001 between the former Northern Alliance and some Pashtun groups. As noted by a well-known expert on Afghanistan, “The political process is corrupt. The only way to obtain a seat at the table now is not to try to get elected but to take up arms”46.  

  • 47  Ariane Quentier, in her book that was cited above (Afghanistan, Au Coeur du Chaos), points out tha (...)

80These issues are generally misunderstood by the international community, which naively believes that democracy automatically tones down political disputes and persists in imposing such democratic structures in accordance with a top-down approach with time frames that are much too tight47. The organization of the 2005 legislative elections in Afghanistan was severely defective; in fact, contrary to its mandate and the 2001 Bonn Agreement, the electoral commission systematically validated the candidacies of powerful “commanders” of armed militias, including those of former war criminals. The Afghan parliament, which undeniably has some remarkable personalities, including women with exceptional courage and talent, is therefore partly comprised of mafiosi and gang leaders who bought their votes and consequently immunity, which they hurried to reinforce by enacting an amnesty law. Here we have a caricature of a parliament that will heavily influence the country’s fate.

81There is, of course, no miracle solution in this area; but it seems that in such circumstances more time should be spent under a provisional UN or other administrative mechanism before rushing to national elections. The goal should be to facilitate the search for political compromise in order to establish monitoring mechanisms and checks and balances, so as to keep the election winner from hoarding excessive power and revenues. The ‘winner takes all’ system which becomes the norm in such countries is a recipe for a bleak future as demonstrated by recent events in the Ivory Coast. It is especially important to take time to create a grassroots democracy at the village level. This is where democracy is most urgently felt and demanded by populations. Villagers in most ‘failed’ States suffer either from the lack of a State, the State's atrocities, or the atrocities of local ‘strong-men’. The thirst for grassroots democracy is therefore immense.

The famous ‘compact’ between the international donor community and the Karzai regime should have been much simpler and much tougher  

82A constitutional reform is undoubtedly necessary in Afghanistan, in order to establish monitoring and power-sharing mechanisms. But at the same time, it will undeniably be essential to construct, if it is still possible, a real grassroots democracy, probably starting with Community Development Councils or similar structures, and by giving real substance to the various provincial bodies that have been created on paper but currently have no resources or authority. This is an ambitious project that should move this centralized Afghan State, which exists on paper only, to a more decentralized form. This approach, which aims to construct a grassroots democracy according to a bottom-up process, incidentally corresponds to the vision of Hanif Atmar, who was ‘blocked’ by President Karzai when he tried to pitch this idea. Contrary to common opinion, this type of approach can be implemented fairly rapidly in tandem with a local development program like the ‘National Solidarity Program’. But once again, is it too late in Afghanistan? That is, unfortunately probably the case in the southern and eastern parts of the country. The year 2011 will tell us whether it is also too late for the rest of the country.

83It is clear that the Afghan authorities and the international community now have their backs to the wall. When I questioned my driver in Kabul in 2008, a former civil servant who had quit his government job that did not even allow him to pay his rent, he made the following frightful comment: “out there, in the countryside, it's better with the Taliban who are honest at least; and in the city, it was better with the Communists who were less corrupt”.

84It is easy to see that under these conditions, sending tens of thousands of additional soldiers, which is what President Obama recently did, or mobilizing a few billion additional dollars in aid, which are partly fictive or largely wasted, makes little sense and is likely to end in failure. Difficult decisions now need to be made. The real compact between the international community and President Karzai should have been infinitely more stringent than the one that was presented in the London and Paris conferences. This ‘compact’ should have been summarized in 2 paragraphs:

Aid shall first and urgently address the reconstruction of the Afghan government structure. For this to happen, about a hundred Afghan leaders with the requisite qualities of integrity, leadership and organization shall be appointed to head key institutions and positions. The regime’s current alliance with despised warlords involved in drug trafficking shall be re-examined.

The international community shall contribute a few billion dollars in additional aid. But aid should be paired with a clear strategy for its use under the authority of a powerful ‘aid manager’ able to impose discipline to donors.

85The terms of this ‘compact’ would certainly have been deemed too restrictive by the Karzai regime. But if difficult decisions cannot be made, where are we headed? Perhaps it is best to limit our losses, leave a failed State with its failure and try to rapidly negotiate an honorable exit strategy with the Pakistani ISI and the Taliban, to at least keep Ben Laden from noisily settling back in Kandahar...

The international community's approaches in these countries need to be reassessed

86This rather harsh assessment of what must be called the failure of international aid in Afghanistan is offering as counterpoint not a model, which would be valid in all places and at all times, but the major principles that international aid should obey to tackle the new challenges posed by fragile post conflict countries' reconstruction needs.

More resources are needed and not just promises or media gestures, and they are needed urgently

87In this area, the ten or so billion dollars that were actually disbursed in Afghanistan from 2002 to the start of 2007 for development assistance proved to be highly insufficient. It is clear that the ratio between military spending and civilian spending should have been more evenly balanced. We also noted the importance of this aid’s quality so that it does not become lost in cascading subcontractor margins. Large sums need to be available rapidly. This poses thorny problems to countries heavily in arrears with major funding agencies. Such large sums should also be made available over the long term, which contradicts the traditional principles of performance-based aid, under which the exceptional amounts mobilized for post-conflict countries are rapidly reduced after a few years.

The objectives that the international community will set need to be perfectly clear

88In this respect, the charitable objectives that arose from the MDGs which serve as the conceptual framework for development agencies' actions are unsuited to this type of situation. It seems clear that objectives to rapidly restore personal safety and improve local governance should become top priorities. International forces cannot sustainably meet this type of need. In fact, the reconstruction (or ex nihilo construction) of the State machinery of government is fundamental and this objective should mobilize a large part of the international community's efforts. The army, the police, the local government and administration, the local justice system, the Ministry of Finance and other financial institutions should be straightened out as a priority. Other state institutions, particularly those that will facilitate the revival of the economy (public works, energy, agriculture, and irrigation) should also follow as quickly as possible. In short, the comprehensive reconstruction of the government machinery and the public administration reform should be high priorities. Such institutional reconstruction requires sustained long-term efforts and sometimes long term funding of these institutions' operating costs by the international community.

International action needs to be coordinated much more seriously

89In this respect, the major principles, which are certainly laudable, for aid coordination that were defined in the 2005 and 2008 Paris and Accra conferences on aid effectiveness, are mostly inoperative and even quite naïve in such environments. The issue of who manages and who can manage aid should be clarified. I had the occasion here to make a few concrete proposals that for the moment have been rejected by the main aid agencies and that, to be implemented, would require a top-level political decision. This is a subject that should now be examined by the G20. Aid resources must indeed be managed with a strategic vision. We know that fragile or failed states cannot perform such strategic management, at least during the initial period. The question of who can perform this strategic management of aid resources is still unresolved. This question triggers emotional responses. Experience shows that major donors are also unable to seriously coordinate their strategies and actions beyond main policy statements and that conversely, they engage in power and precedence struggles. The United Nations institutions that would have the legitimacy to coordinate aid do not have the capacity to do so. The imposition of United Nations ‘mandates’ is no longer a politically acceptable option, nor is the use of expatriate quasi-governors. There is obviously no easy answer to this question which is nonetheless central to aid effectiveness in these situations. My suggestion is to start by having major donors prepare joint country assistance strategies and to do so fairly rapidly (the first year in which aid interventions resume). In any case, it is important to build local aid management capacities which may entail modifications to the government architecture, capacity-building requiring salary innovations, and local political will. Respect of discipline, which is essential for such strategic coordination of aid, also requires the use of multi-donor trust funds and the development of specific tools such as matching funds financed by major funding agencies.

The aid strategy should ensure a balance between urgent operations likely to rapidly deliver visible results (the famous peace dividends) and medium- to long-term actions whose purpose should be to construct modern, viable institutions in key sectors

90In this area, there is a fundamental contradiction between traditional project approaches whose short-term effectiveness relies on project teams (the famous PIUs) who are paid high salaries, and the need to build lasting institutions on salary bases that can be borne by local public finances. This argues in favor of a rapid launch of comprehensive civil service reforms dealing with recruitment procedures, wage structures, human resource management methods, and the rules for selecting able leaders. Such reforms should also cover the extent of state responsibilities and the possibilities of developing service delegations in some key sectors such as education and health, public works and local development, where NGOs and some private companies can play an important role. But these reforms -- let us not fool ourselves -- are technically difficult, and require the mobilization of resources over long periods, which makes donors reluctant. They are also likely to collide with nepotism and with the logic governing the political distribution of spoils of power which is so common in post-conflict situations.

Technical assistance in this type of situation should also be strategically managed

  • 48  “Rethinking technical Cooperation”, Elliot Berg, UNDP 1993.

91Technical assistance should no longer be randomly and sporadically peppered around to meet emergencies. It amounts to huge financial sums, i.e. around 30% of ODA in some circumstances. As Elliot Berg pointed out nearly twenty years ago48, it will not be truly effective unless it is part of coherent institutional construction strategies.

The support given by the international community to the reconstruction of the machinery of government raises the issue of the supported regime's legitimacy and the limits of democracy in this type of context

  • 49  “War Guns and Votes, Democracy in Dangerous Places”, Paul Collier, Harper Collins 2009.
  • 50  See Côte d'Ivoire

92While general elections under the supervision of international observers are part of the standard compact with the international community and undoubtedly a necessary first step, their limits are also now obvious. In a recent publication49, Paul Collier thoroughly describes the caricature of democracies that are commonly being established in fragile post-conflict countries. Paul Collier reminds us that a democracy that (i) is limited to periodic elections, (ii) does not allow any change in government, (iii) allows the winner to carry off all of the political and economic power, and (iv) whose logic is largely ethnic, ultimately leads to a rise in political tension and not a resolution of conflict. In some cases where it is obvious that the team in power will use all the devices at its disposal, which are numerous, to stay in power whatever the cost50, the appropriateness of some elections can be called into question. If we take the case of the DRC, if for example Bemba were to be cleared by the international criminal court and allowed to run for the next presidential election, simple ethnic arithmetic and the powerful means that are at the disposal of the current power suggest that Bemba's only chance of rising to power would be by fueling turmoil during the election period. Does all of this justify that the international community should spend nearly $1 billion to organize elections whose sole result will be to aggravate tension that is already intense at the local level? But that being said, what is the alternative?

The challenge of reconstructing state institutions is considerable for technical but especially political reasons

93We already noticed that the required approach conflicts both with nepotism and the associated policies based on the distribution of the booty which these institutions potentially represent. In order to rebuild or build modern institutions, it is necessary to select leaders who have the leadership, integrity and organizational qualities required for this task. The latter, who can always be found provided that there is real will, need to be able to recruit their management teams on the basis of merit and be able to compensate them according to local job market conditions and not obsolete salary scales. This approach’s success, as we saw above, requires the financial, technical and political support of international aid, which needs to take care not to thwart it with short-term approaches. This clearly comes up against the obvious contradiction between concerns of effectiveness in state institutions, and the very strong logic of the redistribution of spoils of power between political partners that is specific to most post-conflict contexts. Such being the case, this contradiction needs to be tackled by development partners head on.

94All in all, these principles of common sense, which reflect the experience that I have acquired over time in many difficult contexts, still conflict with the current practices of aid, which is still too ideologically marked by a purely charity-oriented approach and whose traditional poverty reduction objective is unsuited to these situations. All the same, ‘failed’ States are going to multiply. As shown in my recent book, they represent regional public challenges, and some could become global public challenges. While international aid is finally one of the very few available ‘instruments’ in the Western tool kit that could facilitate the reconstruction of failed states, it still needs to clearly define this type of objective and better organize itself to do so. The culture shock that it will face to adapt to this task must not be underestimated. But the Afghan disaster clearly shows that ‘Business as usual’ in this area inevitably ends in failure.

What lessons can be learned from this analysis by international solidarity organizations and NGOs?

95Be clearheaded and realistic: NGOs need to understand that in such situations, the simple traditional compassionate approach to aid leads to an impasse and that the immediate objective in a failed State consists in building or rebuilding a relatively modern State and its governing institutions. In this context, it is important not to do anything that will hinder the achievement of this objective.

96Do not multiply actions that are not coordinated or integrated into coherent sectoral strategies either with the local government or with major funding agencies. This type of uncoordinated charity action undoubtedly brings occasional relief to abominable misery. But it also fuels the general disorder and contributes to the construction of a parallel administration, thus undermining attempts to rebuild a modern administration. There are presently more than 300 health projects in DRC, out of which 90% cost less than $1 million. Is this making much sense? The attitude that consists in preparing and managing one's ‘own’ projects and implementing them with one's ‘own’ personnel paid and selected according to one's ‘own criteria’ without referring to anyone, because the local government is considered incompetent and corrupt and major funding agencies are also viewed as incompetent, ends up multiplying enclave operations whose sustainability is in any case highly dubious and whose ultimate impact will be negligible. The government and funding agencies are perhaps incompetent but it will be necessary to ‘make do’ and try to improve their performance rather than trying to bypass them.  

97Indeed, it seems important for international NGOs to actively engage major international donor agencies to allow them to benefit from the field experience of NGOs. To do so, NGOs need to do what they are not used to doing, which is to participate in the development of major donor country and sector assistance strategies. Their objective should be to clarify their own positioning and their future role. They should particularly insist on policies relying on delegation of responsibilities.Such delegation of responsibilities should aim to limit the implementing role of purely state institutions whose capacities in fragile states will clearly remain limited for a long period. It is, for instance, unrealistic to expect a failed State to efficiently operate a decentralized public health system in a rural environment. Specialized NGOs can do a much better job, provided their action can take place in the context of coherent health strategies and not add to the current anarchy.

98NGOs should alsoargue for greater coordination among funding agencies at global and sectoral levels and not hesitate to publicize blatant coordination failures of donors to the international public. They might think it is impossible to make themselves heard by major donors who have an easy tendency to autism. I do believe that a solid critical article in the ‘Financial Times’ opens many doors in Washington, London and even Paris.

99It is also important for them to deliberately seek to support the governance and local government institutions that are under construction and take care not to replace them as is often the case.

100Finally in degraded security contexts NGOs need to avoid letting their international and local personnel take foolhardy risks. On several occasions, I have been appalled by the completely thoughtless risks taken by NGO personnel in Afghanistan, which I imagine they took with their management's approval. Here NGO managers need to recruit technically experienced personnel with the requisite maturity and judgment to be able to work in this type of environment. That is not always the case.

101We know that the task is immense. Some conflicts have now lasted for a whole generation! But while the international community's errors in Afghanistan since 2002 are alarming, the analysis of these errors also offers new ideas for consideration and concrete solutions that are perfectly applicable in other contexts. This is undoubtedly a meager consolation, but it is important to at least learn from the many mistakes that have been made in this unfortunate country.

Top of page


Serge Michailof was until recently the executive director for operation of the French Development Agency (AFD) and was previously a director and senior advisor at the World Bank. His experience covers more than sixty different countries. He is professor at the Institut d’Etudes Politiques (“Sciences Po”) in Paris, and a regular consultant for the World Bank and other international organizations. He has published or edited several books.

Top of page


1  Text of the presentation given at the Humanitarian Forum, 1 October 2010, Annecy, France. This presentation was inspired by various publications I have written since 2007, and particularly chapter 6 of the book. I published (with Alexis Bonnel): “Notre Maison Brûle au Sud, que peut faire l’Aide au Développement” Fayard/Commentaires 2010, the chapter “Le Défi de la reconstruction de l’Administration en Afghanistan” that I wrote in the collective publication “Etats and Sociétés fragiles” JM Chataigner and H Magro, Karthala 2007, and various articles that are cited in these pages.

2  As illustrated by the terrifying description of Congo's colonization in: “King Leopold’s Ghost” Adam Hoschild, Mariner Books, New York, 1998.

3  On this subject, see the extraordinary publication by Ahmed Rashid: “Descent into Chaos - The US and the Failure of Nation Building in Pakistan, Afghanistan and Central Asia”, Ahmed Rashid, Viking, 2008.

4  Note here that for General McChrystal, this military effort was greatly insufficient and has historically been under-financed. It is true that the annual cost of the War in Iraq in 2008 ($140 billion) totalled more than all military spending in Afghanistan combined from 2002 to 2007.

5  The figures given are in current dollars.

6  According to the OECD DAC in 2006, 44% of aid to Afghanistan was still tied (conditional to purchases in the donor country).

7  Which had been set up by the Soviets shortly before their departure.

8  The Bush administration’s institutional memory loss with regard to end-of-conflict procedures and stabilization and reconstruction actions has been denounced by Francis Fukuyama: “Nation Building, Beyond Afghanistan and Irak”, John Hopkins, 2006. See also Ashraf Ghani and Clare Lockhart: “Fixing Failed States, a framework for rebuilding a fractured world”, Oxford University Press, 2008.

9  “Notre maison brûle au Sud, que peut faire l’aide au développement?”, S Michailof, A Bonnel, Fayard 2010.

10  On this subject see: “Cambodia after the Khmer Rouges, Inside the Politics of Nation Building”, Evan Gottseman, Silkworm Books 2004, and “Cambodia, Report from a Striken Land”, Henry Kamm, Arcade Publishing, 1998.

11  For a very interesting discussion about the error represented by the 'light footprint’ option, see Seth Jones' fascinating publication: “In the Graveyard of Empires- America’s war in Afghanistan”, Norton & Cie, 2010.

12  For example, five Médecins sans frontières (MSF) employees were shot down in 2004 by a local commander who, according to the subsequent investigation, merely wanted to affirm his power over the region to the district police chief. This assassination and the government’s lack of response caused MSF to stop its activities, with the result that its 1,400 local employees found themselves jobless.  In 2008, more than thirty NGO employees were killed and around a hundred were kidnapped.

13  On this subject, see the very perceptive analysis “A Mean to What Ends? Why PRTs are Peripheral to the Bigger Political Challenges in Afghanistan”, Barbara J. Stapleton, Journal of Military and Strategic Studies, Fall 2007.  

14  In 2004, the budget devoted to American PRT development actions was only $52 million for a total authorised US aid budget of more than $2 billion!

15  On this subject, see the remarkable study: “The Taliban’s Winning Strategy in Afghanistan”, Gilles Dorronsoro, Carnegie Endowment, 2009.  

16  As do the Ministry of Finance, the Central Bank and the NDS, which were also established as modern institutions at the same time. But it should be noted that recent management changes made at the MRRD seem to have weakened this institution.

17  After the 2003 departure of General Eikenberry who was initially in charge of coordinating US support for the army, American Air Force generals with no expertise in this area took the reins.

18  At the same time, while the Taliban was installing the first components of its political infrastructure in the southern part of the country, Rumsfeld transferred yet another thousand American soldiers to Iraq.

19  The “Economist intelligence unit” October 2008 report writes: “The state of the police is abysmal…”

20  This refusal caused Minister Jalali to resign and go into exile in the United States in 2005.

21  Rumours about the auctioning of district chief positions abound in Kabul.

22  “Afghanistan, au Coeur du Chaos”, Ariane Quentier, Denoël, 2009.

23  This figure incidentally corresponds to the numbers that were rapidly reached by the Algerian army and police during the civil war that started in 1993, as the Algerian population is approximately the same size as the Afghan population. Note that the IMF’s contribution in the framework of a macroeconomic adjustment programme and then oil revenues helped finance the Algerian military effort.  These figures also correspond to Iraq’s current security services.      

24  Note that the countries that supply the troops, and not the wretched soldiers, pocket the tidy sums paid by the UN.

25  For a more detailed analysis of these issues see: “The Seven Capital Sins of the International Community in Afghanistan, What Went Wrong in the Security Development Nexus and What should be the Lessons for US Foreign Assistance”, Serge Michailof, The German Marshall Fund of the United States, Nov 2008.

26  This issue is now finally being corrected. During his conference at Sciences Po on Nov 23, 2010, General Petraeus whom I challenged on this issue considers that about $800 million are now spent on agriculture on a yearly basis. But again so much time has been lost!

27  See “Afghanistan: Economic Incentives and Development Initiatives to Reduce Opium Production”, C Ward, D Mansfield, P Oldham. William Byrd, World Bank/DFID, Feb 2008.

28  On this subject, see the study: “Immediate Priority Needs of Vulnerable Farmers engaged in Opium Poppy Cultivation”, Urgence Réhabilitation Développement” (URD), Jan 2008.

29  Hajji Fazul Rahim, quoted by the New York Times, 28 November 2009.

30  See “Responding to Afghanistan’s Development Challenges”, William Byrd, World Bank, 2007.

31  Poverty Reduction Strategy Papers, prepared by governments of aid recipient countries to determine each country strategy to fight poverty and make the best use of aid resources.

32  Millennium Development Goals. These eight broad goals were adopted by the international community in September 2000 and today largely determine the specific objectives of donors in aid recipient countries.

33  Maybe today, no one wants to endorse the responsibility of the coming failure…..

34  As such, a project that falls under one of the jointly defined priorities and is financed by aid could for instance see its financing topped up by an allowance from the multilateral trust fund.

35  The context is similar in Cambodia and DRC.

36  PIUs, a term that is systematically used in donors’ technical documents.

37  Review of Technical Assistance and Capacity Building in Afghanistan, discussion paper for the Afghan Development Forum”, Serge Michailof, World Bank, April 2007.

38  “Rethinking technical Cooperation, Reforms for Capacity Buiding ” Elliott Berg coordinator, UNDP, 1993.

39  Paid by the British in the 19th century, the Soviets and Americans from 1945 to the 1970s, then only by the Soviets until 1990, by Pakistan in the Taliban period and by the US and international aid since 2002

40  The corresponding study is comprehensive and remarkable. It corresponds to the equivalent of several years of expert work: “Afghanistan, Building an Effective State, Priorities for Public Administration reform”, World Bank 2008.

41  From 2003 until 2006 for all of the approximately 30 ministries and concerned institutions combined, capacity building and related institution building technical assistance contributions accounted for less than one-third of what USAID had given to the Ministry of Finance alone. It was not until 2007 that a group of donors finally organised significant financial relief to support this process. Too late once again...

42  Or rather, unfortunately, that “worked”, for recent management changes seem to have seriously weakened this institution, but this does not challenge the method.

43  On this subject, see the fierce assertion by one of America’s best experts on Afghanistan on the need to put an end to the power and impunity of these ‘Power Brokers’: “Afghanistan’s Uncertain Transition from Turmoil to Normalcy”, Barnett R. Rubin. Council on Foreign Relations.March 2006.

44  On this subject, see: “Sortir du Piège Afghan”, Serge Michailof, Commentaire No 126, Summer 2009.

45  See “Notre maison Brûle au Sud”, cited above.

46  “Afghanistan’s 2009 elections” William Maley (Asia Pacific College of Diplomacy), Afghanistan info No 65, October 2009.

47  Ariane Quentier, in her book that was cited above (Afghanistan, Au Coeur du Chaos), points out that the Afghan election was above all a domestic American political issue “organised for the Americans” as noted by Younus Quanouni, one of President Karzai’s historical opponents.  

48  “Rethinking technical Cooperation”, Elliot Berg, UNDP 1993.

49  “War Guns and Votes, Democracy in Dangerous Places”, Paul Collier, Harper Collins 2009.

50  See Côte d'Ivoire

Top of page


Electronic reference

Serge Michailof, « The Challenge of Reconstructing "Failed" States », Field Actions Science Reports [Online], Commentaries, Online since 23 December 2000, connection on 14 November 2019. URL :

Top of page

About the author

Serge Michailof

Professor, Institut d’Etudes Politiques (“Sciences Po”), Paris, Regular consultant for the World Bank and other international organizations, E-mail:, Website:

Top of page


Creative Commons Attribution 3.0 License

Top of page
  • OpenEdition Journals