Subcontracting State-Building

Abstract

Contemporary development assistance often takes the form of subcontracted state-building. Foreign donors hire for-profit firms to provide services and to improve or create institutions in developing countries, particularly those experiencing internal conflict. This arrangement creates two counterproductive dynamics: first, it introduces agency problems between donors, recipient states, subcontractors, and citizens; and second, it undermines the long-run development of domestic bureaucratic capacity by creating disincentives for the host government to invest. These dynamics hinder, rather than foster, the legitimacy of state institutions. This paper summarizes trends in external support to state-building since the 1970s and illustrates subcontracted state-building with examples from Colombia.

Keywords:

• State-building
• civil wars
• development aid
• contractors
• legitimacy
• Colombia

1. Introduction

Insecurity within states is commonly expected to be one of the major international security challenges of the twenty-first century. Criminal networks, terrorist organizations, and rebel groups slow economic development locally, create negative externalities regionally, and can have global impact in rare cases.¹ Confronted with these challenges, the predominant diagnosis to date has been that states do not have enough capacity, or are too ‘fragile’ to contain these threats,² and the international community has largely turned to state-building as a solution. The idea is nicely captured in the World Bank’s 2011 World Development Report, which claims that ‘The central message of the Report is that strengthening legitimate institutions and governance to provide citizen security, justice, and jobs is crucial to break cycles of violence’.³ In some ways, the modern dilemma turns the Tillian model of state-building on its head: rather than external wars leading to (internal) state-building, internal wars and instability prompt state-building efforts by external actors.

The notion that basic state function must be improved to break the cycles of violence that hinder security and development is a laudable one in principle. Policy documents from major donors increasingly provide holistic guidance on how foreign support can be used to improve state capacity.⁴ Unfortunately, we have scant evidence that the strategies posed in the policy literature state-build effectively, or on the institutional structures currently used by international donors to execute the necessary tasks. Development assistance as commonly practiced today, particularly by the United States and major European donors, takes the form of subcontracted state-building, in which foreign donors hire firms to provide basic services and to improve or create institutions in developing countries.

We argue that subcontracted state-building undermines the efficacy of state-building by external actors and it is unable to foster state legitimacy. Social contract theories of the state posit a relationship between the governed and the government: citizens tacitly consent to be governed in exchange for protection by the state. Lake⁵ has said that this theory undergirds US state-building efforts in particular, since the invasion of Iraq and war in Afghanistan. Rather than seeking legitimacy through democracy promotion, the approach rests on establishing security and providing basic services to the population, which in theory will lead to legitimacy.⁶ But if the role of external state-builders poses challenges for triggering this virtuous cycle, then the addition of another player in the chain – the contracting organizations – complicates the process even further.

Specifically, we expect subcontracted state-building to create two counterproductive dynamics: first, it disrupts citizens’ ability to observe if their government is upholding its side of the social contract, and for the government to demonstrate that it is doing so. When the delivery of services is successful, can the government claim credit for this delivery? When there are problems, can the local or national government effectively intervene to hold the contracting firm accountable? The subcontracted firm is the agent of the donor country, not the host country or the citizens. And within that primary relationship, agency issues abound, especially in contexts like ongoing internal conflicts, where monitoring is difficult and it is not clear how to measure ‘success’: contracting firms take advantage of these environments to increase their profits, sometimes egregiously.

Second, subcontracted state-building undermines the long-run development of domestic bureaucratic capacity by creating disincentives for the host government to invest. This barrier is a frequently studied problem in the aid literature, which Moss, Pettersson and van de Walle call the ‘aid-institutions paradox’.⁷ Funding by donor countries can theoretically spur additional spending by the host government, but it often disincentivizes it.

The agency problems involved in subcontracting assistance to foreign governments might be addressed by reforming contracting structures to better align incentives between the interested parties, though whether this can be done in practice is questionable. The second problem is just as intractable. Unless the host-nation government in question is a pure social welfare maximizer, any foreign support for state-building will lead to lower host-nation expenditures on service provision and thus lower long-term capacity building. In the very best case, foreign support could be neutral in terms of host-nation investments, and external support for major state-building programs can reduce host-nation governments’ incentives to invest in the programs.

Subcontracted state-building is one policy lever among many that external actors can use to try to foster institutional outcomes within states. Others include economic sanctions, election monitors, and bilateral investments treaties.⁸ Krasner and Weinstein⁹ consider foreign assistance as falling generally under a ‘contracting’ arrangement, and contrast it and other contracting forms with coercion and imposition. Compared to the latter two, they find that contracting arrangements are more effective for generating improvements in democratic governance. We argue that the actual practice of providing the aid is consequential for how effective it is likely to be, and largely overlooked in the literature.

The remainder of this paper proceeds as follows. Section 1 summarizes the trends in external support to state-building since 1970s, both in terms of the targets of aid and method of execution. Importantly, we show that contracting through for-profit firms has consistently been a much greater share of US foreign assistance than direct support to foreign governments in conflict zones but not elsewhere, the opposite of what we might want if that aid is intended to bolster the supported country’s governance capacity. Section 2 provides an argument for why aid undermines bureaucratic capacity in recipient states. Section 3 presents brief overviews of state-building efforts by USAID in Colombia to illustrate the arguments. Section 4 concludes.

2. Trends in assistance to state-building

Over the last three decades, a great deal of international aid has been directed toward state-building in conflict-affected or weak states. Between 20 and 40% of global overseas development assistance (ODA) has gone to countries in conflict, with the US and UK contributing slightly more to conflict-affected countries than the European Commission.¹⁰ Overall levels of civil conflict have declined over the past two decades while aid allocation to conflict zones remain high, meaning the concentration of aid to a few conflict affected countries has gone up since the 1990s.

Over time, the method of actually providing this aid has changed. In the last decade, reliance on NGOs and for-profit subcontracting has increased. As we describe below, this is troubling as it is exactly those places where conflict has broken out that the agency problems and perverse incentives of subcontracted state-builiding are most pernicious. More generally, Dietrich¹¹ shows that subcontracting is more frequent when aid is delivered to states that have the least capacity, even for donor states that provide more aid on average to government agencies. This section examines these trends in more detail and provides background on how they emerged, focusing on USAID.

2.1. USAID goals and execution over time

For roughly the first 10 years of its existence, USAID invested in large-scale project involving infrastructure and road construction. These projects were largely executed directly by USAID itself, occasionally contracting academic experts for input. The move toward contracting out project execution originated with the round of reforms under the Reagan administration’s Private Enterprise Initiative in 1981. Its goals were to increase the use of the private sector to deliver services, to encourage developing countries to open their economics to boost private enterprise, and to pursue policy reform reflecting free market principles in developing countries.¹² At its peak during the Vietnam War, the agency had roughly 15,000 staffers, which dropped to less than 2000 in 1992.¹³ USAID was narrowly saved from elimination altogether when the Clinton administration shifted it to the State Department. Shortly after, tens of missions were closed, and the staff was reduced by a further 37% over the course of the Clinton administration through a hiring freeze and reduction-in-force (layoffs).¹⁴ Between 1995 and 2002, USAID lost a further $50 million for staff hiring.¹⁵ The reduction in staffing coincided with an ongoing evolution ‘from an agency in which US Foreign Service officers directly implemented development projects to an agency where a declining number of full-time staff struggle to oversee contractors and grantees who carry out the day-to-day development work’.

Apart from hoped-for indirect effects during the cold war, development aid was largely divorced from security concerns. It was not until after 9/11 that development assistance became explicitly tied to security concerns abroad. As terrorism became a prime security concern, weak states came to be seen as potential breeding grounds or safe harbors for terrorist organizations. Major international aid agencies began to invest more in building state institutions and services. The US invasions of Afghanistan and Iraq led policy-makers to adopt (eventually) counterinsurgency approaches that featured state-building elements. By 2004, USAID had executed $1 billion in reconstruction contracts in Afghanistan, and had awarded 11 contracts for $3.3 billion by 2004 in Iraq,¹⁶ numbers that continued to increase in subsequent years. Given its new role in counterinsurgency efforts, USAID’s budget doubled between 2002 and 2003; though the operational budget which paid for officers to supervise that increase only went up by 11%.¹⁷ The agency also opened three new missions in Southeast Asia, Yemen, and Pakistan, in addition to the ones in Afghanistan and Iraq. This growth was challenging for the agency after a decade of decline and the shift from program management to contract oversight. Incredibly, in Iraq USAID had to ‘outsource even the oversight of the contracts’.¹⁸ (It hired the Army Corps of Engineers to oversee the largest contract, to Bechtel Corp, and Management Systems International to oversee the rest.) Norris writes¹⁹:

The 10 largest USAID contractors received more than $3.19 billion in 2011, and more than 27 percent of the agency’s overall funding was directed to American for-profit firms [in 2011]. To put this in perspective, if the for-profit contractor Chemonics were a country, it would have been the third-largest recipient of USAID funding in the world in 2011, behind only Afghanistan and Haiti.

An example of subcontracted state-building in the context of an ongoing conflict in USAID’s assistance to Yemen. USAID’s strategy there is to ‘increase stability trhough targeted interventions in highly vulnerable areas;’ these interventions focus on improving livelihoods and governance. Stability is ‘characterized as reduced support for violent means of resolving conflicts and grievances and increased support for the central government to prevent fragmentation of the state and continued conflict’.²⁰ The USAID Mission (staffed by only 18 people) has three substantive project areas: the Community Livelihoods Project (CLP), the Responsive Governance Project (RGP), and the Youth Stabilization Initiative (YSI). Creative Associates won the three-year, $125M contract to implement CLP. Creative subcontracts ‘rapid response’ community projects to local NGOs in five areas: (1) agriculture and water; (2) economic empowerment; (3) education; (4) governance; (5) and health. A promotional video summarizes the impact of the governance work: ‘Government officials better understand the needs of rural and urban communities and are partnering with USAID to find relevant solutions for pressing issues’.²¹ The RGP is

designed to strengthen the capacity of civil society organizations and increase citizen participation during Yemen’s political transition and work with individual ministries to plan and improve their capacity to better respond to public needs during and after the transition period and National Dialogue process,

according to its implementer, Counterpart International, which won a $43M, three-year contract. The highlighted programs on its website focus on maternal health and empowering women.²² Whether or not these companies and USAID will be able to (1) effectively carry out these programs; (2) spur institutional capacity in Yemen, or (3) transfer credit to the Yemeni state is debatable, given the limited evidence we have from the impact of such aid in other contexts.

3. The problems with subcontracted state-building

3.1. Agency, accountability, and attribution

The problem of subcontracting are particularly concerning when it comes to state-building efforts. In the contemporary era of human rights and democracy, US-supported state-building is an effort to gain citizens’ consent where it does not already exist.²³ At the same time, this effort is not entirely without coercion, particularly in the context of ongoing insurgencies, where development assistance accompanies military operations.

Focusing on the civilian side of state-building, it becomes clear why international involvement can complicate the process. When international actors, such as USAID, act on behalf of a state in order to build that state, three problems arise. First, it is not clear from civilians’ perspective which actor deserves their allegiance: the state, or the foreign power propping it up. In Afghanistan, at least, those two actors were often far from coordinated. Even so, attribution might not be so concerning: Dietrich, Mahmud, and Winters²⁴ find in a recent paper that citizens in Bangladesh rate their local governments higher when they are informed that USAID is supporting local projects. The question is to what extent these findings would also travel to a context in which many non-state actors are competing for allegiance.

Second, it is not clear how citizens can hold foreign actors accountable, breaking the link that the social contract ideally establishes. Donor states often struggle to oversee the implementation of projects by the firms they contract. In the context of countries experiencing internal conflict, and those which need state-building support, it is highly unlikely that domestic governments at the local or national levels will be able to oversee these arrangements. Additionally, the implementing organizations are typically not directly accountable to either the host government, or the citizens in areas where they are operating. Even if the donor government makes such accountability a feature of the contract they issue to the implementing organization, the channels through which citizens and the host government can express discontent are often informal or unclear, and as mentioned, the ability of the donor government to enforce accountability mechanisms is often in question.

Third, foreign development agencies are rarely operating as the agent of the host-nation government. This is particularly true in cases where the government is weak and cannot effectively oversee their actions. Contracting the delivery of services in the pursuit of state-building creates another layer to each of these complications. Even in rare cases where the host-nation government is relatively strong, it often holds little sway over contractors beholden to foreign government officials. Those additional agency problems are most pernicious in places with ongoing conflict where aid is intended to bolster state capacity. Unfortunately, because governments in conflict zones often have severe capacity constraints and because security costs inflate the price of even the simplest projects in war zones, the proportion of aid delivered through for-profit contractors in conflict zones often vastly exceeds that given to governments. Additionally, again, Dietrich²⁵ finds evidence that donors are more likely to ‘bypass’ badly governed host states, and issue contracts to third parties.

To illustrate this difference, Figure 1 shows the proportion of USAID spending going to for-profit firms vs. foreign governments by year.²⁶ We separate out the proportion in countries at peace (upper panel) from those that have an ongoing civil war (lower panel). While the difference is negligible for countries that did not experience a civil war during the period, for-profit contractors took a much larger proportion of USAID disbursements in conflict zones. The difference is especially dramatic during the peak of the war in Iraq. As we described, this pattern is troubling as it is exactly those places where conflict has broken out that the agency problems and perverse incentives of subcontracted state-building are most pernicious.

Figure 1. Average proportion of USAID disbursements to for-profit firms vs. foreign governments. Disbursements through non-profits excluded. Data from Federal Procurement Data System (FPDS), and countries classified as being in conflict if they experienced 25 or more battle deaths in a given year according to the Uppsala Conflict Data Program (UCDP).

3.2. Counter productive for long-run

The basic problem of foreign support to state-building is well studied.²⁷ There are two central issues that create a moral hazard problem between recipient and donor. First, in order to keep receiving outside support, the government chooses to be inefficient and may even strategically tolerate some level of insurgent activity.²⁸ Second, because they get outside support, the government spends its money on things other than state-building. We focus mostly on the latter here.

We could summarize the challenges created by the dual moral hazard in an analogous way to the poverty trap, in which there are two equilibria of state-building: (1) state-building in which both the government and people invest together; and (2) no state-building, in which neither invest much because they do not trust each other.²⁹ In either case, both actors are rational. To push from the undesirable outcome (2) to the ideal (1), outsiders give aid to the local government, or in our case, directly execute projects on the government’s behalf. However, this assistance cannot alleviate the state-building trap unless the state not only claims credit but also begins to invest on its own.

Other theoretical perspectives highlight variants on this basic theme and have found rich empirical support. Chatterjee, Giuliano and Kaya,³⁰ for example, find that foreign aid targeted for public investment crowds out 80–90% of government spending. Werker, Ahmed, and Cohen³¹ find that foreign aid is overwhelmingly spent on imported non-capital goods, leading to no net long-run development gains. Ahmed³² shows that foreign aid can even help authoritarian governments to hold onto power longer by channeling money from public goods provision into patronage. Amidst such findings it should be little surprise that the aggregate impact of aid on economic growth is disputed at best.³³ Based on our discussion of how state-building assistance is managed and spent, we expect that aid should have a similarly null impact on institutional effectiveness.

4. Subcontracted state-building in Colombia

Colombia is an interesting case to explore the implications of subcontracted state-building because if it should work anywhere, it should be in Colombia, for several reasons. First, state-builders in Colombia do not confront a need to ‘nation-build’ before or simultaneous to institution building.³⁴ Colombians already identify with the nation.³⁵ Second, the Colombian state is also relatively bureaucratic and has some amount of presence in most territories. If institutions are a prerequisite for aid to be effective, then Colombia is in relatively better shape than most countries in conflict.³⁶ Relatedly, shoring up and improving institutional capacity in Colombia should not have to overcome a strong legitimacy hurdle. The constitution adopted in 1991 was not seriously challenged by competing armed groups, and though the main rebel group proposed a new constitutional convention during recent peace talks with the government, the talks concluded with no such plans. On the local level, though armed groups did establish parallel governance in many communities, they often co-opted existing state institutions to do so.³⁷ If legitimacy emerges from norms, practices, and beliefs as described in Duyvesteyn’s introduction to this special issue, then the shift from loyalty to armed groups – to the extent that it exists – should not be so vast as it appears in many other contexts. Third, and finally, the Colombian Government coupled its most recent state-building effort with a strong counterinsurgency push into territories identified as outside of the state’s control. This coercive leverage, according to Lake,³⁸ is necessary for institutions and legitimacy to develop. If we find that even under these favorable conditions, subcontracted state-building falters, then it should make us very skeptical that it can work anywhere.

4.1. Colombia’s consolidation program

In Colombia, the effort to establish the state’s presence throughout its territory has spanned several decades in response to ongoing violent conflict. The contemporary Colombian civil war began in the 1960s with the formation of several leftist insurgent groups. The largest among them, the Revolutionary Armed Forces of Colombia (FARC), is the longest-running insurgency in Latin America. (It recently reached an agreement with the Colombian Government and began to demobilize.) The insurgent organizations formed strong ties with peripheral communities over long periods of time, and in some cases even settled regions with civilians.³⁹ In the early 1980s, regional counterinsurgent groups (known as ‘self-defense’ forces) formed in reaction to the insurgents’ increasing presence, and in the absence of an effective state response.⁴⁰ The FARC and the ELN increased their capacity and territorial presence in the 1990s, as the regional ‘self-defense’ forces gained strength as well and eventually formed an umbrella organization in 1997 known as the United Self-Defense Forces of Colombia (AUC), a paramilitary organization that aimed to bring the fight to insurgent groups across regions. As the illegal armed groups gained in strength and expanded their presence, the war’s intensity increased.⁴¹ When the Pastrana administration entered office in 1998, it initiated peace talks with the FARC by ceding a vast territory as a demilitarized zone roughly 400km south of Bogotá. Talks failed in 2002 after FARC’s ongoing armed actions, and the unchecked expansion of the paramilitary groups. The same year, Álvaro Uribe was elected president on a hardline platform to defeat the guerrillas militarily.

As the Uribe administration’s Seguridad Democrática policy expanded the presence of the Colombian Armed Forces (CAF) substantially beginning in 2002, the Ministry of Defense advocated for an approach that would secure the state’s territorial presence beyond the initial military incursion. Foreign support was critical to this effort in two respects. First, Plan Colombia, started in 2000 under the Pastrana and Clinton administrations, greatly ramped up US assistance to the Colombian military. Specifically, by purchasing mobility assets and providing for their maintenance, Plan Colombia allowed the government to project power into new areas, particularly in the southern part of the country. (Critics of the arrangement also pointed out that paramilitary forces arrived in the south first, and set up camps near the main military bases in the region.⁴² Though the military could readily take territory from the FARC, it lacked a bureaucratic mechanism to manage efforts across the government in newly recaptured territory. In 2004, the US military at Southcom proposed the creation of an inter-institutional agency, the CCAI, that would allow for the coordination of military and civil affairs in newly recuperated territories.

After three years, frustrations with the CCAI’s inability to forge coordination among agencies led the Ministry of Defense to propose the Política Nacional de Consolidación y Reconstrucción Territorial (The National Policy for Territorial Consolidation and Reconstruction – PNCRT) early in the Uribe administration’s second term. This program, led by the government of Colombia (GoC) and supported early on by USAID’s Office of Transition Initiatives in Bogotá, has three broad goals: to extend institutional presence to consolidation zones, to foster citizen participation and good governance, and to improve regional integration. The program began in the Macarena region of Colombia in 2009 and was primarily supported by USAID and its subcontractors for its first three years. One of the programs that it built on was an alternative development project launched several years earlier.

In 2005, USAID awarded a contract for the Areas for Municipal-level Alternative Development (ADAM) and More Investment for Sustainable Alternative Development (MIDAS) programs in regions where illicit crops were particularly prevalent (and by extension, where state presence is weak). Though reducing illicit crop planting and production was the primary goal of the program, it also sought to ‘promote and enhance the viability of the overall level of political, economic, and social development’.⁴³ The counter-narcotics efforts were directly linked to these goals and the broader war effort, because the revenue from the illicit crop trade was crucial for funding the illegal armed groups in Colombia. MIDAS additionally sought to improve the local administrations and access to justice, to protect human rights, and to reduce corruption.

The USAID contractor, Associates for Rural Development (now known as Tetra Tech), states:

ADAM works directly with municipal government and community-based organizations in eleven Departments of Colombia to promote and implement agricultural economic and infrastructure initiatives. The ADAM development model empowers communities to improve their own lives through economic and political partnerships at every level of Colombian Government. The ADAM model is the first development initiative in Colombia that integrates (1) improving agricultural production and marketing; (2) strengthening local government; and (3) improving the social and productive infrastructure.⁴⁴

However, despite ADAM pouring $200M into 73 municipalities and MIDAS $166M into 545 municipalities in alternative livelihood and development programs, these programs did little to encourage state capacity. According to the Office of Inspector General’s audit of ADAM and MIDAS, the program succeeded in generating some alternative livelihoods but failed to generate any reduction in illicit crop growth.⁴⁵ Further, it found that the livelihoods generated were rarely sustainable, implying that the program did not generate investment on the part of the government.

Experience with ADAM and MIDAS highlights challenges of subcontracting state-building. In the first place, it did little to catalyze effort by the Colombian Government. In the second place, it laid bare the challenges of contracting with for-profit firms to achieve outcomes whose quality is as hard to measure as ‘development’. (Indeed, one of the main recommendations from the audit was to improve monitoring and oversight of the implementing partner’s data collection methods.) Essentially, USAID faced an agency problem. If it could not assess how well the money was invested, the government of Colombia certainly could not. Further, there is no reliable information on how civilians perceived the program over time, compared with civilians who never received such investments or contacts with ARD.

The Colombian Government adapted its PNCRT plan in ways that in theory could have addressed the shortcomings of MIDAS and ADAM, as well as others more generally linked to the challenges we have highlighted. First, PNCRT explicitly aims to leverage more spending by departmental and local governments. The architects of the PNCRT are aware that relying on US-based companies can undermine state-building, particularly challenging legitimacy, if citizens attribute interventions and any benefits to foreign, non-state actors. For this reason, the GoC requires USAID subcontractors to present themselves under the brand ‘Colombia Responde’ to prevent people from attributing any positive outcomes to international assistance, the contractors, or NGOs rather than to the Colombian state.⁴⁶ Further, in practice, operators cannot implement any programming without the approval of the regional consolidation manager in terms of location and content, potentially resolving one of the key agency problems we describe above. According to Álvaro Balcázar, one of the architects of PNCRT, this was a purposeful effort to avoid confusing civilians. The director of regional programs in Latin America for one of the winning grantees, Chemonics, told us that he was unaware of any other program that Chemonics worked on that did not include a USAID or Chemonics logo in its messaging.⁴⁷

Second, the PNCRT also tried to mitigate the moral hazard risk. It emphasizes that the role of USAID operators is to ‘catalyze the arrival of the state’, not to perform the state’s functions, and not to serve as a traditional economic development program.⁴⁸ Though a member of USAID’s Office of Transition Initiatives (OTI) was highly involved in the concept and design of the program, from the beginning there was strong ownership by officials of the Colombian Government, namely Sergio Jaramillo, a Vice Minister of Defense in the Uribe administration and later National Security Adviser to Uribe’s successor, Juan Manuel Santos, and Balcázar, a bureaucrat in the Department of Social Prosperity. While the pilot project was primarily funded by USAID, Jaramillo and Balcázar worked to create a sustainable institution that would oversee Consolidation on the part of the government. In November 2011, the Santos administration unveiled its reorganization of cabinet-level institutions and introduced the Special Unit for Territorial Consolidation (UACT). Balcázar became the director of the UACT. However, by 2013, Balcázar had been replaced and USAID officials expressed concern that the Santos administration had turned its attention to the peace talks. It is possible that the GoC relied on USAID to fill the gap of institutional development rather than dedicated its own resources to invest in state-building.

Third, the approach of PNCRT features traditional development interventions once a community is judged to be sufficiently stable. In theory, once communities have reached this threshold, regional managers of the Consolidation program organize an assembly of community members and explicitly present a social contract to them. During this assembly, the government representative explains that it seeks to engage the citizens of the community to fulfill its obligations to them to protect their rights. At the same time, citizens have duties that include respecting the law and not growing illicit crops.⁴⁹ State officials then offer, following USAID/OTI influence, ‘rapid response’ assistance. Rapid response includes humanitarian assistance in the form of emergency food aid, and small grants to be used for infrastructure improvements in the communities (on bridges, health posts, or schools, primarily). The Consolidation program contacts community leaders, usually the presidents of the Juntas de Acción Comunal (JAC), and asks them to poll their communities to collectively prioritize what they will use the small grant for. The idea is not only to provide needed assistance and investment, but also to gain residents’ trust by delivering both private and public goods. We witnessed one meeting in which the government representative apologized to community leaders for a two-year delay in the arrival of their assistance, and introduced them to a representative from an international organization that would donate the aid.⁵⁰ Though the government claimed that this incident was unusual, it was clear that the leaders who were convened were disgruntled and did not trust the official.⁵¹

Longer-term assistance involves identifying or helping to mobilize producers associations, providing training in licit crop cultivation, and linking the producers to a value chain, or market. In the four regions where USAID operates, these activities are primarily undertaken by USAID contractors, though they are explicitly urged to identify government programs that could be mobilized to provide particular services to the populations. Institutional development is another aspect of this component: municipal governance is targeted for improvements as well. USAID contractors work with municipal officials on budgeting and policy planning, particularly with the Planes de Desarrollo and Planes de Ordenamiento Territorial. However, meetings with the contractors in December 2013 indicated that the bulk of their investments are in traditional development activities.

USAID work in Colombia across MIDAS, ADAM, and PNCRT is emblematic of the push toward incorporating state-building goals into development programming, and subcontracting the programming to for-profit firms. As we described in the previous section, this approach is an inefficient way to generate political order and virtuous institutions, because rather than the state forming a ‘contract’ with citizens, a foreign-based company attempts to do so. Secondly, foreign companies’ investment in governance allows the state to invest its resources elsewhere, reducing the likelihood that state-building will be sustainable over the long term. The Colombia case also raises a final, important complication: even if agency issues and long-run counterproductivity were not concerns, it is unclear that subcontractors’ approach to state-building would generate the governance and stability outcomes that donors seek. Rather, the actual programming seems to be typical development assistance geared toward improving livelihoods and empowerment, and the connection to state-building is tenuous at best. This problem stems from two issues: first, a lack of ‘know-how’ in terms of institution building among both donor agencies and contractors; and second, a confusion between ends and means. It is unclear that operating within and attempting to foster democratic norms and human rights – as Western donors attempt – produces effective states, or should do.

5. Conclusion

This article has described a common yet overlooked practice in international development: subcontracted state-building. We show that it is particularly important feature of assistance to states that are internally unstable or in conflict. Further, it has identified two crucial flaws in the approach: (1) it creates agency problems, and (2) it is counterproductive. For- profit firms are not the appropriate actor for establishing links between citizens and states. Indeed, they turn out to substitute for the state rather than prompt its development. State-building by external actors in response to internal wars, particularly through subcontracted development firms, is likely to be ineffective.

Future research would test these claims. Iyengar and Shapiro (2017), for example, assess the impact of a range of stabilization programs in Afghanistan from 2004–2014 – many of which constitute the kind of subcontracted state building described here – and find some positive short-run impacts but no evidence of long-run changes. And, we are in the process of evaluating the impact of the PNCRT program in Colombia since 2010. Our research design will allow us to compare how regions overseen only by the Colombian Government compare with regions where USAID-contracted firms are operating. To the extent that civilians report greater confidence in local and national governance in the state-run regions, it could be evidence that subcontractors’ presence indeed does disrupt the social contract process.

The theory, as well, could also be profitably extended by considering explicitly how subcontracting is likely to differ in civil war and non-conflict settings. One key way could be that civil war environments feature numerous non-state actors vying for legitimacy, such as insurgents and warlords.⁵² We would expect that foreign firms are even less well equipped to be able to foster state legitimacy in such contexts. Another important potential consideration is the degree to which host states have engaged in forms of ‘wartime political orders’⁵³ with these non-state actors. Where states engage in cooperation or integration of these groups, the role of external actors is likely to be complicated even further.

These issues raise a dilemma for international donors: if aid fails to push states out of a ‘state-building trap’, what is the best alternative? Should donors continue to give in spite of the risks? And if so, should they continue to employ contractors? In the roughly 15 years since the newly invigorated demands on USAID, the agency has struggled to adapt to those demands given its staffing and funding constraints. However, the problems the agency faces and perpetuates with its contracting system have not gone unnoticed. The USAID Administrator under Obama, Rajiv Shah, gave a speech in which he said that USAID was ‘no longer satisfied with writing big checks to big contractors and calling it development’ and argued that development firms had become more interested in maintaining their business than in helping recipients to the point where they no longer needed aid.⁵⁴ Once again, the agency is undergoing procurement reform in an effort to direct a larger share of its budget to institutions abroad. The contractors formed the Coalition of International Development Companies to resist the reforms, arguing that providing funds directly to foreign states would lead to inefficiency and waste.⁵⁵

We argue that subcontracted state-building is an example of doing worse by doing something, and that donors and recipient states would be better off redirecting assistance into alternative programming that would directly assist government agencies rather than replace them. Of course, this approach raises distinct challenges, but we think that on the whole, they are less likely to impede state-building to the extent that subcontracting does.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

This work was supported by the Air Force Office of Scientific Research (AFOSR) [FA9550-09-1-0314]; and the Office of Naval Research [N000141310097].

Acknowledgements

The authors thank Eli Berman, Mike Callen, Simone Dietrich, Joe Felter, Daniel Flechas, Isabelle Duyvesteyn, Christopher Paik, Paul Staniland, and Seiki Tanaka for many helpful conversations. Mica Meintel-Wade provided invaluable expertise on U.S. government contracting and research assistance.

Notes

1. Clunan and Trinkunas, Ungoverned Spaces; and Department for International Development, “Building Peaceful States”.

2. e.g. USAID, “Fragile States Strategy”.

3. Cliffe and Roberts, World Development Report 2011.

4. e.g. USAID, “State Building in Situations of Fragility”.

5. Lake, “The Practice and Theory of US Statebuilding”.

6. Lake, “Building Legitimate States”; and “The Practice and Theory of US Statebuilding”.

7. Moss, Pettersson, and van de Walle, “An Aid-Institutions Paradox?”.

8. Krasner and Weinstein, “Improving Governance from the Outside In”.

9. Ibid.

10. Dietrich, “Donor Political Economies”, finds that donor countries’ preferred mechanism for providing aid relates to their political economy: ‘neoliberal’ US and UK subcontract most often, while ‘statist’ France and Germany provide more of their aid through foreign government institutions. Here, we focus on USAID because it is by far the largest provider of ODA, followed by DFID, the UK’s aid agency. The Japan International Cooperation Agency (JICA), not listed here, is one of the largest donors to war-affected states, and often subcontracts to private companies as well. However, it has started to shift towards funding multi-lateral agencies such as the UNDP instead (Tanaka and Yoshikawa, “Establishing Good Governance”), consistent with Dietrich’s argument.

11. Dietrich, “Bypass or Engage?”.

12. Berrıos, Contracting for Development, 12.

13. Zeller, “On the Work Force Roller Coaster”, 33.

14. General Accounting Office, “Strategic Workforce Planning”.

15. For a detailed discussion of the staffing changes in the 1990s, see Zeller, “On the Work Force Roller Coaster”.

16. Committee on Foreign Relations, “USAID Contracting Policies”.

17. Zeller, “On the Work Force Roller Coaster”, 38.

18. Ibid., 37.

19. Norris, “Hired Gun Fight”.

20. USAID/Yemen, “2010–2012 Yemen Country Strategy”.

21. http://vimeo.com/49834542.

22. http://www.counterpart.org/news/counterpart-in-yemen-building-responsive-governance.

23. The pursuit of explicit approval is consistent with Hume’s critique of social contract theory; he argued that the notion of citizens’ tacit approval undermines the claim that the contract is legitimate, and that consent must be sought by the government and volunteered by the population, rather than assumed.

24. Dietrich, Mahmud, and Winters, “Foreign Aid, Foreign Policy”.

25. Dietrich, “Bypass or Engage?”.

26. Appendix A1 repeats the excercise for 10 countries with ongoing conflicts from 2001 on.

27. See the excellent review by CGD (Moss, Pettersson, and van de Walle, “An Aid-Institutions Paradox?”).

28. Bapat, “Transnational Terrorism, US Military Aid”; and Felter, “Aligning Incentives to Combat Terrorism”.

29. We thank Seiki Tanaka for making this point.

30. Chatterjee, Giuliano, and Kaya, “Where has All the Money Gone?”.

31. Werker, Ahmed, and Cohen, “How is Foreign Aid Spent?”.

32. Ahmed, “The Perils of Unearned Foreign Income”.

33. see Rajan and Subramanian, “Aid and Growth”, among others.

34. Darden and Mylonas, “The Prometean Dilemma”.

35. There are of course exceptions and some regional variation. But in general, compared to many other countries experiencing internal conflict, Colombians share a national identity.

36. Moss, Pettersson, and van de Walle, “An Aid-Institutions Paradox?”.

37. Arjona, Rebelocracy; and Ch et al., “Endogenous Taxation in Ongoing Internal Conflict”; and Eaton, “The Downside of Decentralization”.

38. Lake, “Building Legitimate States”.

39. Ramírez Tobón, “Colonizacion Armada”.

40. Romero, Paramilitares y Autodefensas.

41. Echandía, Dos Decadas del Conflict Interno en Colombia.

42. Dube and Naidu, “Bases, Bullets and Ballots”.

43. Office of the Inspector General, “Audit of USAID/Colombia’s Alternative Development Program”.

44. Tech, “Colombia”.

45. Office of the Inspector General, “Audit of USAID/Colombia’s Alternative Development Program”.

46. The PNCRT program was contracted under the name ‘Consolidation and Enhanced Livelihoods Initiative’ (CELI) for two different regions: Central, and North/South. Associates for Rural Development (ARD; since renamed Tetra Tech ARD) won the 5-year contract for the Central region, valued at $114,953,981 and awarded 20 April 2011 (https://www.fbo.gov/?s=opportunity&mode=form&id%20=%2078349e99a9002969c2154b37a9a35516&tab%20=%20core&_cview%20=%201). The Central region includes municipalities in the departments of Caqueta, Tolima, and Meta. Six months later, Chemonics International, Inc won the bid for the $89,570,250 contract to cover the North and South regions (https://www.fbo.gov/index?s%20=%20opportunity&mode%20=%20form&id%20=%2094bd6a2be9168149cde328733b4a59b9&tab%20=%20core&_cview%20=%201). The large, lone municipality in the South region is Tumaco; in the North, the contractor is expected to work in the Bajo Cauca of Antioquia and southern C ́ordoba. Finally, USAID also awarded a $32 million grant to CHF Inter- national (now Global Communities) in early 2010 to execute Consolidation in the four municipalities of the Montes de María region.

47. Interview, October 2010, Bogotá.

48. Interview with Álvaro Balcázar, Bogotá, 27 May 2011.

49. Interview with Carlos Avila, Bogotá, 5 November 2010.

50. January 2011.

51. Focus group meeting, Vista Hermosa, January 2011.

52. Arjona, Rebelocracy; and Blair and Kalmanovitz, “On the Rights of Warlords”; and Malecaq, this issue.

53. Staniland, “States, Insurgents and Wartime Political Orders”; and Staniland, “Armed Politics and the Study of Intrastate Conflict”.

54. Norris, “Hired Gun Fight”.

55. Kenny, “It is Time to Reform USAID”; and Norris, “Hired Gun Fight”.

Appendix

Figure A1. Average proportion of USAID disbursements to for-profit firms vs. foreign governments by country. Data from the Federal Procurement Data System (FPDS) and the UCDP. Proportion funded through non-profit organizations excluded.