Types of foreign aid

Foreign aid is a broad, generic rubric that encompasses a wide range of specific types of external support, and equally diverse in purpose. It includes a wide variety of assistance and objectives: economic or development aid, military aid, humanitarian and disaster relief, and technical assistance. Different types of aid will have their own specific criteria, purposes or uses, let alone the specific motivations of donors (Bourguignon & Sundberg Reference Bourguignon and Sundberg2007). This is a critical point because of the debates regarding the ills and virtues of foreign aid in Africa and elsewhere. Not all foreign aid has socioeconomic development of the recipient as the primary objective or donor criteria and not all foreign aid to Africa has been for development. Notwithstanding the MCA's different approach, not all US foreign aid, to Africa or elsewhere, has been for development or democracy promotion. Its top aid recipients are all based on geopolitical objectives. The USA has provided financial and military aid to buttress some of the world's most repressive and corrupt governments in pursuit of its strategic or commercial interests.

As a sub-category, development aid, such as the MCA, also comprises a broad range of specific aid types: for example development financing, or what is formally referred to as official development aid (ODA), technical assistance, trade support and market access measures, cultural and educational supports, scientific exchanges and cooperation. In general, the intended purpose of development aid is to build up the recipient's productive and institutional capacity, spur growth, combat poverty and form human capital.

Importantly, there are different types of ODA, each with consequences for the recipient country as well as relations with donors. The first, concessional lending from bilateral and multilateral sources, has been, and still is, the largest category of development financing for Cabo Verde. A major source of its concessional financing, or soft loans, has been Portugal. These are low or no-interest loans with generous repayment terms such as long grace periods, extended repayment schedules and other lenient terms. For developing countries, low-cost concessional financing is a critical resource. However, since they are loans that must be repaid, they are debt-producing and their servicing can have negative fiscal and macroeconomic consequences. Today, especially after the 2008 global crisis, rising external debt is a serious problem for Cabo Verde. Roughly 70% of this external debt stock consists of concessional loans, the rest commercial. Servicing that debt puts a great deal of pressure on fiscal and social policy. The total external debt stock in 2017 was $169.5 billion escudos, just a shade under the estimated GDP of $173.4 billion escudos (BCV 2018). Total public debt is 126% of GDP (IMF 2018).

In contrast, aid in the form of grants does not have to be repaid, even if some may come with other ‘strings attached’. The MCA is grant aid. For a small, fragile and highly dependent economy like Cabo Verde, the significance of grant aid like the MCA, especially in such large amounts, extends beyond their potential contribution to growth and poverty reduction. Grant aid does not have a negative fiscal and macroeconomic impact.

Tied aid

Few aspects of aid have generated more controversy than so-called conditional, or tied, aid. The MCA is not tied aid. As discussed below, it is based on selectivity and ex ante eligibility criteria, rather than conditionality. Conditional aid, whether in the form of grants or loans, is given and disbursed contingent on specific promises recipients make regarding policy reforms, performance targets, institutional changes and/or how the aid is to be used and implemented. While conditions vary greatly, tied-aid is common practice by multilateral and bilateral donors, including the USA in its non-MCA aid. Conditional aid became widespread in the wake of the 1980s global debt crisis. Countries in Africa and Latin America especially were subjected to intrusive and extensive conditionalities included in structural adjustment programmes. Cabo Verde largely escaped a formal structural adjustment agreement, but tied aid is common in its experience.

Donors believe tied aid gives them more leverage in shaping aid effectiveness, but certain types of tied aid can have deleterious effects on recipients. Much research has shown that conditional aid does not work (Ohler et al. Reference Ohler, Nunnenkamp and Dreher2012). Tied aid is also controversial not just because of the excessive external influence it brings, but because it distorts the development policies and priorities of recipients. Tied aid restricts the recipient country's autonomy, may distort its development priorities, and can have damaging effects on its national private sector. Empirical research is needed, but observational evidence suggests that tied aid from Portugal – which often requires contracting Portuguese firms or sourcing labour, materials and equipment from Portugal – has had damaging effects on Cabo Verde's private sector, especially the construction sector. Aid from China has similar stipulations.

Soederberg (Reference Soederberg2004) has argued that the MCA is merely a change in tactics, not substance, since its ‘pre-emptive development’ approach is part of the same coercive donor practice that imposes neoliberal globalisation by threatening to withhold funds unless recipients adopt its prescriptions. As discussed below, for good or ill, Cabo Verde had already adopted neoliberal economic reforms over a decade prior. For recipients, it makes a substantive difference if aid is not tied. Cabo Verde had latitude in determining how to use the MCA. Indeed, the MCA's rejection of conditionality was unexpected from a US administration that categorised countries as ‘with us’ or ‘against us’, and accompanied the programme with a major expansion of US aid to combat AIDS. Nonetheless, the programme's ostensible country ownership was not as simple in practice, especially since recipients have to navigate a two-level game of hard negotiations with the MCC and domestic actors over how to use the grant.

The Millennium Challenge Aid programme

The Millennium Challenge foreign aid programme was created in 2002 by the George W. Bush Administration. The MCA was a radically different approach adopted by the USA to select recipients and administer its foreign aid, previously all administered by the United States Aid Agency (USAID). The MCA was an entirely separate, parallel aid programme, independent of USAID. It was meant to supplement regular foreign aid provided by the USA, although there was little connection or coordination (Nowels Reference Nowels2006). In Cabo Verde's case, for instance, USAID had scaled back and eventually ceased its operation before the MCA. Despite the sizeable individual grants provided under the MCA, the new programme's roughly US$1 billion annual budget was dwarfed by other US foreign economic and military aid programmes.

From the standpoint of selection, implementation and donor practices, the MCA is a substantive departure from common aid practices in the international system, including traditional US practices. First, as noted, it was to be separately and independently managed. The Millennium Challenge Corporation (MCC) was created by Congress to administer the programme. Second, it aimed for fewer recipients but larger, more ‘impactful’ aid. It emphasised large-scale infusion of project-based aid to target growth and poverty reduction in a select few countries over a five-year time period. Rather than many, independently managed, small-scale annual aid projects, the idea was that the bigger size and longer time periods would be more impactful. This approach was premised on a widely shared idea – supported by some academic research – that increasing and sustaining aid beyond a certain threshold was necessary to boost growth (Sachs Reference Sachs2005). As such, the MCC prefers to invest a substantial amount of resources on one or two major projects.

The third feature that sets the MCA apart is the adoption of a rigorous, transparent, quantifiable approach to both recipient selection and project evaluation. Recipient selection is discussed below. After a country is selected as eligible, it must then submit feasible projects that quantitatively demonstrate the intended social and economic impact. Thus, mere eligibility does not guarantee aid; nor does mere submission of any proposal result in financing. Grant proposals must still pass rigorous empirical evaluations. MCA grants involve data-driven econometric analyses and economic rates of return (ERR) estimates, and continual monitoring and evaluation after implementation starts. Environmental impact assessments were also introduced.

Fourth, the MCA programme embraces a different approach to identifying recipients. The MCA is based on selectivity rather than conditionality (Carbone Reference Carbone2004; Dennis Reference Dennis2008; Molenaers et al. Reference Molenaers, Dellepiane and Faust2015). It is a competitive grant. Recipients are selected based on their actual socioeconomic and political performance as calculated by quantifiable, objective criteria based on internationally used indicators. That is, the MCA relies on an ex ante, largely objective screening to select recipients (Dennis Reference Dennis2008). Prospective countries do not ask to be chosen; nor are otherwise chosen based on hidden, subjective criteria, or commercial and strategic ties to the USA. If much foreign aid internationally is based on political or strategic ties, commercial interests, colonial histories or other subjective criteria, the MCA is different. Although the selection process is far from perfect and allows for discretion, it has thus far been faithful to its apolitical, objective selection of recipients (Nowels Reference Nowels2006; Owusu Reference Owusu2017). Provided it is not under US sanctions or other legal restrictions, a country's policy performance, quantitatively and publicly measured, is the only criterion. As such, candidate countries are competing against each other on policy performance, each with a scorecard containing specific metrics of policy performance in three areas: free politics, free markets and good governance. Cabo Verde has had ample experience with political selectivity in aid as a result of its special partnership agreement and other aid programmes from the European Union. The MCC is thus not alone in embracing selectivity. Even outside the European Union, ex ante political selectivity was becoming widespread in the 2000s (Molenaers et al. Reference Molenaers, Dellepiane and Faust2015).

Even though the MCA was a product of the US's so-called global war on terror, the intent behind the MCA's creation was to separate aid from geostrategic, commercial or diplomatic considerations. The use of transparent international indicators was novel, even if the indicators selected favour the US's underlying ideological agenda or that some individual awards raise questions about the recipient's eligibility. The new programme is premised on the idea, backed by some research and corroborated by the Cabo Verde case, that aid is more effective in recipient countries that have good governance and a good macroeconomic policy environment (Burnside & Dollar Reference Burnside and Dollar2000; Acemoglu & Robinson Reference Acemoglu and Robinson2012; Adedokun Reference Adedokun2017).

The country selection process occurs in two stages, the first being a pre-selection of a pool of ‘candidates’ based solely on per capita income. In early 2004, the MCC identified 63 candidate countries, Cabo Verde among them, based on their per capita income (MCC 2004a). Twelve additional countries that satisfied the income threshold were disqualified because of existing US sanctions or other legal prohibitions. Both low income and lower middle-income countries are considered in this preliminary screening.

In the second phase, candidate countries on the list are then pared down to ‘eligible’ countries, based on a scorecard containing the new programme's indicators in the three categories of democratic governance, human development and economic freedom. In May 2004, the MCC released a list of 16 eligible countries, including Cabo Verde, who were invited to submit proposals for an MCA grant for fiscal year 2004 (MCC 2004b). In 2004, the scorecard consisted of 16 indicators. By 2018, these had expanded to 20. Among the specific indicators are: rule of law, civil liberties, political freedom, control of corruption, public spending on education, girls’ primary education, immunisation rates, fiscal and trade policies, cost of starting a business. These indicators have their limitations and ideological biases. They are equally weighted, which raises its own set of issues. Questions and criticisms have been raised regarding their appropriateness, internal coherence, deficiencies in calculation, their application and ideological biases as well as the presumed complementarity of the programme's goals of promoting democracy and development (Knoll & Zloczysti Reference Knoll and Zloczysti2011; Lebovic Reference Lebovic2014). To be eligible, countries must pass at least half of the indicators, scoring above the median for countries in the income group, though the MCC retains some leeway in assessing progress if countries fail on some indicators. Countries that are making progress but fail to pass the eligibility indicators can be awarded a Threshold Grant to finance improvements and reforms in the specific areas required to be eligible. Countries must also pass a separate corruption control indicator. The MCA thus favours countries that already have free market policies and political liberalisation, but also who have strong anti-corruption policies. As noted, eligibility does not mean countries automatically receive aid, since countries must then take the initiative, be sufficiently entrepreneurial in their diplomacy and governance structures to pursue the grant, and prepare and submit sound project proposals that the MCC can approve. Unsurprisingly, therefore, only five of 16 eligible countries signed a compact the first year.

The MCA, it is fair to say, rewards success. It favours political and economic reformers who are already making progress in clean government and socioeconomic performance but are still facing growth and poverty hurdles. Some research suggests aid is more effective in countries already on a growth path (Larsen Reference Larsen2016). Unlike the preferential support provided under the LDC category or multilateral agencies, the MCA adopts explicitly political, or rather policy performance, criteria for awarding aid. That is, for the MCA good governance and market-based policies are prerequisites. While all low- and lower middle-income countries are potential candidates, only those already making progress on these policies are eligible. As such, a valid concern with the MCA programme is that it does not necessarily include the countries that need aid the most – the poorest trapped in poverty, social instability, governance breakdown and economic stagnation. At its launch, there was a lot of concern African countries would be disadvantaged in the new programme – although half of the first group of 16 eligible countries in 2004 were in Sub-Saharan Africa.

Fifth, the MCA emphasised country ownership. To be sure, all donors tout this idea, though usually empty of substance. The MCA is not tied aid. Recipients choose how to use the aid based on their priorities. As elaborated below, country ownership has substantive meaning under the MCA, but the concept in practice is far more ambiguous and complex. Finally, the process of determining the amount of MCA financing is ambiguous. As noted, the programme's philosophy is ‘impactful’ aid, but what is it or what is the appropriate amount to be impactful? The programme does not have a funding criterion. Presumably, determination of the amount of aid is based on technical assessment of the country's compact proposal. However, given its own annual budget constraints, it is likely the MCC sets an arbitrary limit on financing per recipient even before its compact is submitted. Moreover, the MCA's overall budget appropriated by Congress was much lower than anticipated, forcing the MCC either to reduce the number of recipients or scale back the size of financing for projects.

Preparing the MCA compact

The preparation and drafting of an MCA compact proposal unfold in a number of steps. The compact process begins with the eligible country submitting a concept paper, which outlines and justifies its development priorities that need MCA support. In practice, of course, these concept papers are more of a wish-list than a strategic document. In 2011, the team went a step further, preparing individual concept papers for each of the desired areas of intervention.

A second critical step is to establish a team responsible for leading and managing the entire process, from preparing the proposal to organising the national consultation process. Eligible countries can designate the planning ministry, an ad hoc inter-ministerial team, hire international consultants, or outsource a team from national experts outside the public administration. For each compact, Cabo Verde relied on the latter, an independent team of national experts under the supervision and coordination of the prime minister's office and the planning department respectively. This model allowed much greater efficiency and coherence, while shielding the process from bureaucratic infighting, political interference, special interest lobbying and delays. The core team was responsible for both the sector consultation process as well as preparing all the analyses and documentation involved, including the final compact. Putting together a team to lead the MCA preparation work seems trivial. Yet not all governments have sufficient organisational and human resources, institutional adaptability for the required inter-ministerial coordination, high level leadership to manage both the process and the unavoidable organisational frictions, and the national statistical capacity necessary, let alone the enabling national political climate for consensus building.

A major step in compact preparation is national consultation. The MCA requires extensive national consultation in order to garner input from civil society and business groups. Its purpose is to promote consensus-building among the government, business groups, civil society organisations and other stakeholders. It also raises public awareness and familiarity. This aspect of the MCA process came naturally to the government which had organised a number of national forums on its ‘transformation agenda’ since coming to power in 2001. The goal of national consultation is to identify development priorities, brainstorm project ideas, and to ensure that projects accurately reflect national priorities, country ownership and buy-in from different groups. Even before drafting the project proposals, the government in Cabo Verde organised a national forum for each grant, bringing together public officials, private sector leaders and organisations, civil society and non-governmental groups. These large national forums were supplemented by smaller sector-specific workshops and stakeholder meetings.

For both compacts, there was genuine engagement of civil society and the private sector, with extensive participation of non-government and community development organisations. The day-long forums were not mere formalities, but involved thoughtful reflections and assessments of the country's needs, criticisms of government policies, heated debates and disagreements about priorities, needs and root causes, and the expected wish list of possible projects. On both occasions, the forums were broken down into smaller sector-focused working groups whose deliberations informed the proposals sent to the MCC. In many cases, members of these groups also participated in smaller project teams during the project definition phase. Given the diversity of stakeholders and their preferences, it was a challenge for the team to distil their input and bring coherence to the process. Invariably, this engagement of broad sections of society results in both a long wish list of projects and delicate politics as various groups and public sector entities vie for the limited resources. Nonetheless, given the participation of so many experienced and expert professionals from the private sector, civil society groups and the public sector, this input, even when divergent, was valuable to the team preparing the compact.

The third critical step in the process involves elaborating the analytical support for the compact proposal, the Constraints Analysis (MFP 2010). The grant proposal must be supported by a rigorous analytical foundation that identifies the binding impediments to growth and poverty reduction, and shows empirically why and how the proposed areas for intervention will overcome these impediments. With its origins in the academic field of development studies, the Constraints Analysis is more of an academic research document or econometric study than a typical aid project document or consultant report (Hausmann et al. Reference Hausmann, Rodrik and Velasco2005). It is theoretically informed, data-driven and substantiated by quantitative analyses. Based on the Constraints Analysis, seven project areas were initially identified for the second compact: water and sanitation; agriculture value chain; property rights and land titling; energy, specifically renewable energy; private sector financing; transportation; human capital capacity building, focusing on higher education and technical training. Separately, in-depth proposals were prepared for all seven. The first compact proposal similarly contemplated multiple projects in many of the same areas, but only four were financed. For the second compact, as noted, the list of projects was reduced to two – water and sanitation and land titling – following cabinet-level deliberations inside the government based on the Constraints Analysis, in addition to negotiations between the government and MCC. Throughout the Constraints Analysis and initial project concept phases, the country's team and the MCC were in continual discussion, with comments and feedback on drafts, and heated debates.

Country ownership with an asterisk

The MCA departs from common donor practices in foreign aid. Nonetheless, Cabo Verde's MCA compacts had their share of sceptics and critics. Specifically, some people have questioned US motives. The claim that the USA uses the MCA as an instrument to extend its influence in international affairs, while true, is insignificant. All forms of foreign aid are about influence. Aid is a foreign policy tool to achieve policy objectives. Aid has never been motivated by pure altruism (Feldman Reference Feldman1967). Since World War II, as exemplified by the Marshall Plan, the US has used its foreign aid as an instrument to pursue its national security and commercial interests in specific countries as well as promote general world order interests. The MCA is no exception. It embodies the increasing focus on security in aid policy. With the MCA, the USA pursues general world order objectives, such as promoting certain ideas, stability, open markets, pro-US attitudes, and pre-empting what it defines as radicalism and terrorism. Shaping the international order in ways more amenable to their interests has always been an objective of all great powers.

First, in the specific case of Cabo Verde, it was the one that pursued the MCA and aggressively lobbied for the grant despite not meeting all the eligibility criteria or missing important data. Second, whether for Cabo Verde or other recipients, the pattern in MCA aid distribution appears driven more by these general world order interests, namely the global anti-terrorism concern and open markets, rather than by direct strategic or commercial interests in each specific recipient. Globally, what stands out among all MCA recipients in the 16 years since its launch is the wide diversity and dispersal of the 49 recipients. Many MCA recipients have little or no specific, bilateral commercial or strategic importance to the USA. The recipients are diverse, not only in terms of geographic region or levels of development, but in terms of the diversity of diplomatic, economic and strategic relations with the USA or linkages to US global policy. While individual country recipients, such as Georgia, the Philippines, Kyrgyzstan, Ukraine, Kenya and Niger may make a strong case for direct commercial or geostrategic interests, the overall programme appears driven by its commitment to a competitive process and good governance criteria. For example, the compacts were terminated after coups d’état in Mali and Madagascar.

Through the MCA, the USA promotes its free market ideology as well as using it as a non-military instrument for its post-2001 global war on terror. The programme likewise exemplified the unilateralism prominent in US foreign affairs at the time (Carbone Reference Carbone2004; Woods Reference Woods2005). The MCA advances an ideological agenda, pursued with fervour domestically and internationally since the 1980s, including using its clout and veto power in the major international financial institutions to condition development financing on the adoption of these policies. In the post-2001 world, the USA devised the MCA as a non-military instrument to combat terrorism – premised on a presumed link between poverty and terrorism. This is an old idea in US foreign policy thinking, dating to the 1950s, that free market-driven economic growth and socioeconomic development are the best antidote to revolution, political instability, radicalism, and the spread of communism or other hostile ideologies. This old idea has been updated to include terrorism. The notable concentration of MCA grant recipients in West Africa, which today has emerged as an active battlefront in international terrorism, may be interpreted in this light, even though most of the MCA grants were awarded long before or awarded to countries unaffected by terrorism. Indeed, this aspect of the MCA – tying terrorism and poverty – exposes the programme's contradiction (Woods Reference Woods2005; Owusu Reference Owusu2017). As a merit-based aid based on apolitical criteria, it precludes targeted, politically driven selection of countries where concern over poverty or terrorism is greatest.

Nevertheless, the selection process, despite its novel approach and general fidelity to transparent indicators, is inexact and open to manipulation by allowing room for discretion in cases where data are missing or candidates are deemed to be ‘making progress’. The MCC has used its discretion to qualify and disqualify countries (Lebovic Reference Lebovic2014). In both rounds of awards in 2005 and 2011, some individual cases like Kenya and Indonesia were selected as eligible despite poor indicators, including extremely low scores on what is purportedly a disqualifying corruption indicator.

Cabo Verde may be a case of fortuitous convergence of general interests and generic US geostrategic concerns in the West Africa region, primarily in terms of international terrorism, money laundering and drug trafficking. Commercial interests are non-existent. Military-security interests directly tied to Cabo Verde are likewise secondary, relevant only insofar as its location in the sub-region and its increasing use as a major transit point for the drug trade. Growing security and law enforcement cooperation began to characterize bilateral relations after MCA-I, including maritime security and intelligence gathering. It was at this time too, four years after the first compact, that the USA and Cabo Verde first began discussions about SOFA. There is no evidence of a link or quid pro quo between the two. Indeed, globally, there is no overlap between the more than 100 SOFA accords signed and MCA recipients. Several countries have MCA but no SOFA, and vice versa, and many have SOFA long before the MCA. Nevertheless, after the first compact we see a marked increase in bilateral military-security cooperation.

As for the actual project proposals submitted to the MCC, there has been a good deal of controversy as to whether these reflect country ownership, or are imposed by the MCC. Others argue the MCC has a one-size-fits-all approach (Nowels Reference Nowels2006; Cardoso Reference Cardoso, Delgado, Varela and Costa2014). In practice, the process is neither. Country ownership is a complex concept. Presumably it means recipients have a final say, are free to choose among alternatives; that society, elites and all stakeholders participate in the process, and have unconstrained decision over which projects to fund. Formally, eligible countries identify their own development priorities and design projects accordingly. For MCA-II, the team freely and independently conceived of the projects, based on national consultation, sector input and the Constraints Analysis; at no point was there pressure by the MCC to impose its preferred projects. The compact process, however, is an iterative negotiations process. At a minimum, the MCC had its technical standards for project feasibility and was stubborn in its position of impactful, single-project aid. Upon completing the Constraints Analysis and drafting the project proposals, both the recipient country and the MCC negotiate a way forward. As anyone who participates in the compact preparation and negotiations can attest, it is not a politicised process or one where the MCC imposes its will. Nevertheless, the negotiation is between unequals; between the MCC team with lot to say about which projects get funded and a country with high aid dependency. Both compacts entailed hard negotiations, with several country missions back and forth; the MCC insisting on a single project and the country team contending multiple projects get funding. During the second compact, the MCC pushed repeatedly to drop the land titling project, but Cabo Verde insisted. A few additional observations are warranted.

Cabo Verde identified, conceived and implemented its own projects. For better or worse, it decided on what and how to use the MCA. While the principle of country ownership is not absolute, as discussed below, the fact that recipients identify their priorities, and design their own projects accordingly, distinguishes the MCA programme from common international aid practice. The prevailing practice, even by the USA's own USAID, is for aid agencies to conduct their own assessment of recipient needs, define their preferred programming for aid use and implementation method, and determine outcomes they want to achieve. In consequence, recipients often implement projects that may not be consistent with their development priorities nor contemplated in their planning – while assuming any debt and other obligations that may come with such aid.

Nevertheless, while the MCA's country ownership principle has substantive content, it is not an absolute. There are nuances in the process, and specific elements in the MCC guidelines and programme requirements that constrain the recipient's freedom of choice. Bilateral negotiations can get heated, as noted. The final negotiations, involving which and how many projects are financed, are hard and entail opposing viewpoints. First, the operational philosophy of MCA financing is impactful aid, i.e. substantial aid on one big project. Naturally, recipients have multiple priority needs, and will prepare proposals identifying multiple projects for financing. For each compact, Cabo Verde had initially identified as many as eight specific project areas across a number of sectors. The MCC's preference on both occasions was to finance only one project, but Cabo Verde insisted on more. After bilateral negotiations, the first compact in 2004 financed four of the eight projects initially proposed. The second compact financed only two of seven. In other words, this hard negotiation forces recipients to reorder and rethink their priorities. Additionally, even with multiple projects, financing is typically skewed toward one big project. Notable in both compacts is that financing was skewed among the projects approved; most of the grant financed one big project, 75% for ports and road construction and 60% for water and sanitation in the respective compacts.

Concerns have been raised that the MCC favours certain sectors and projects, giving rise to the one-size-fits-all approach criticism (Nowels Reference Nowels2006). As noted, Cabo Verde independently determined the project areas and priority. However, this criticism cannot be refuted entirely. Based on a cursory scan of all compacts financed under the MCA around the world, commonalities across projects appear. Specifically, infrastructure projects dominate financing in MCA compacts. Excluding threshold agreements or compacts yet to be implemented, the same broad sectors appear among the 30 countries currently implementing MCA compacts – infrastructure (roads and ports), energy, agriculture, land titling and property rights, water and sanitation, health and education. Among these, however, the bulk of MCA financing among all these countries is for transportation infrastructure, specifically road networks and ports, agriculture, with a big focus on water resources and land titling, and energy resources. However, the strong consistency among compacts may or may not reveal MCC preferences in negotiations involving unequal bargaining power. The predominance of certain sectors and types of projects may just as easily reflect similarity in development needs among the recipients. Developing countries face similar structural impediments to growth, such as weak and limited basic economic infrastructure, energy resources and low agricultural productivity. Yet low human capital, poor health systems and even fragile administrative capacity are also major development obstacles for these countries.

As the MCA-II team discovered, the MCA process and guidelines result in an indirect bias toward certain types of projects. These biases are the result of the heavily quantitative methodology and guidelines adopted. In other words, the quantitative metrics used to judge project viability and economic returns discriminate against certain sectors and project areas that may be harder to quantify or where data are scant. Developing countries also face binding constraints in the form of low human capital. They likewise face enormous challenges in terms of public health as a result of weak or insufficient health infrastructure and preventable diseases that translate into low productivity and a drag on growth. However, these types of MCA-funded projects are few in number. For a programme that espouses good governance in selecting recipients, government effectiveness has not been an area financed (aside from some threshold grants or only indirectly if one counts land titling and property rights). If good governance affects growth, then financing projects that strengthen the rule of law, civil society organisation, or the technical, administrative and planning capacity of the public sector makes sense. Indeed, during both compact negotiations, Cabo Verde proposed education projects which were met with strong push back from the MCC because of the greater imprecision in showing quantitatively their impact on growth and poverty. These sectors are notorious for their inherent difficulty to apply quantitative methodology.

Finally, another reason country ownership is not straightforward in terms of determining which projects get funded is that the hard negotiations are not limited to the recipient and the MCC. The MCA is thus a two-level game, as Putnam (Reference Putnam1988) famously put it. Both compacts entailed complex, often difficult international negotiations between the government and the MCC but also simultaneous delicate consultations, political navigation and negotiations among the many domestic actors. That is, defining which or how many projects to finance is not simple even for the recipient country, especially for politically sensitive or fragmented sectors. As noted, the water sector may end up being the most difficult, if not the least successful, of all the MCA projects. The water project involved breaking up existing structures and creating a new water agency. The existing municipal water systems opposed it, the power company favoured it, while many politicians worried about possible job losses from the restructuring. Even during its implementation, bureaucratic resistance, delays and political concerns over job losses distorted the original project.

The MCA grant engenders a number of other issues, challenges and adverse consequences.

First, as noted, the MCA is based on a selectivity approach, not conditionality. However, there are some ex post ‘conditions’ or, more accurately, contractual requirements, that recipients must agree to. These requirements pertain to project implementation rather than policy. An overriding preoccupation of the MCA programme is to limit corruption and political interference in aid implementation. Specifically, the MCA requires countries to engage in institutional adaptations that may present certain challenges. Namely, recipients must create a separate national agency to implement the grants, an in-country MCA agency. For each MCA compact, Cabo Verde established an MCA–Cabo Verde agency, both staffed by competitively selected professionals and sector specialists. (A separate in-country MCC staff office was also established.) These autonomous, independently managed agencies, staffed exclusively by nationals, are responsible for project implementation, oversight and monitoring, with oversight and periodic review by the MCC. The agency's staffing and operations are financed through the grant, thus it does not impose any additional fiscal burden on the public budget. Professional and autonomous, the agencies operated free of political interference in Cabo Verde. (It should be noted that the head of the first MCA–Cabo Verde agency was the son of the then ambassador, although no one questioned his professional or management abilities.) The mission, jurisdiction and responsibility of the agency are limited strictly to the projects and their financing. The agency is not subordinate to, nor is directed by, ministries in the respective project areas. It thus shields both the MCA grant and the projects from political interference, clientelism, corruption and public sector inefficiencies. With the grant and projects professionally managed and executed by a dedicated entity, the rate of completion and aid effectiveness are high.

On the other hand, this autonomy introduces an element of institutional fragmentation and coordination difficulties. Countries are typically implementing multiple projects in a specific sector, some financed by the MCA, others not, thus putting pressure on coordination, resources and planning. In practice, during both compacts, continual discussion, coordination and information exchange took place between the agencies and sector ministries, with all the expected difficulties, tensions and deficiencies. Whether at the stage of project design or implementation, the MCA process depended heavily on ministries – for data and information, technical support, field logistics, project execution – but bureaucratic resistance and inertia, implementation weakness and inefficient coordination made the entire process difficult.

In Cabo Verde, the main problem was not resistance but inertia, lack of coordination and limited implementation capacity due to technical and organisational weaknesses in the public sector. As such, the MCA introduces an extra coordination challenge in an already fragile context. Some projects require additional institutional and policy reforms which may or may not occur. Moreover, the MCA may engender duplication or competition for the same resources. MCA projects can pull human and material resources from other areas or compete with other projects. Other agencies and ministries may be implementing projects at the same time. Like the port modernisation project, MCA projects are large scale and require vast quantities of materials and other inputs. In a micro-sized, import-dependent economy like Cabo Verde, this could put pressure on supply that may delay or crowd out other projects and businesses.

Another consequence of the MCA is that its projects and financing occur outside the normal planning process. This has been a complaint by some officials. Ideally, countries identify their priorities and design their MCA projects within their broader strategic plan and development planning framework. In practice, planning and budgeting are far from this ideal. MCA projects tend to be multi-sectoral by nature, and may often require follow-on interventions and complementary projects. Yet being outside the normal planning framework, countries may not as easily manage these moving pieces and bring coherence to all development planning. Likewise, in a micro economy susceptible to external shocks, conditions may adversely change rapidly, requiring countries to shift resources and priorities, but they are contractually fixed to MCA implementation and cannot shift its financing elsewhere.

Glaringly, the MCA did not induce nor explicitly involve the US-based diaspora in an economic or investment role in the homeland. Paradoxically, the diaspora lobby was critical to Cabo Verde's eligibility in the first compact. The MCA did not induce a nostalgic, or ethnic, trade and commercial activity. The MCA did not explicitly involve the diaspora as investors or contractors, nor otherwise find creative ways to harness and incubate its role in homeland development. Despite its overall success, the MCA may have represented a major missed opportunity from the standpoint of the potential economic transnationalism in bilateral relations.

Thus, another consequence of the MCA is to limit the recipient country's freedom of choice to use this development financing creatively to spur domestic production or foster national producers and entrepreneurs. As noted, the MCA's design and implementation guidelines seek to limit corruption, funds diversion and interference. As such, all contracts for project implementation are based on transparent internationally competitive bidding. In a small market like Cabo Verde, with a small and fragile private sector, most domestic firms and suppliers will be in competitive disadvantage in a bidding process involving foreign companies. Consequently, for both MCA programmes, Portuguese firms and suppliers were the primary beneficiaries, not Cabo Verdean firms. Thus, just as the MCA was a missed opportunity to involve diaspora investors, it likewise prevented Cabo Verde from using these large-scale public investment projects to spur ancillary economic activity and stimulate local producers and firms. Governments all over the world use large-scale public investments as a stimulus for domestic production and to foster local businesses.