Foreign aid - an analysis on its effectiveness and future application.

(1)

Amsterdam School of Economics Faculty of Economics and Business University of Amsterdam

Master Thesis

Foreign Aid - An Analysis on Its Effectiveness and Future Application

Adihasta Komala 0297232

Semester 1, Academic Year 2013/2014

December 27, 2013

MSc Economics, major: International Economics Primary supervisor: drs. Naomi Leefmans Second supervisor: dr. Kostas Mavromatis

(2)

2

Acknowledgements

I would like to thank God Almighty for the strength and blessings during the writing process of this master thesis. I am also grateful for the patience and understanding of drs. Naomi Leefmans. Moreover, the mental support given by drs. Ellen de Jong has been of great value too, and I thank her for that. Last but not least, I would like to thank my family and friends for their continuing support. Each and every one of you has contributed to the completion of this thesis. Terima kasih. I am humbled and thankful to be a part of your family.

(3)

3

1. Introduction

In its early days, foreign aid has been provided to rehabilitate post-war European economies, and to bring about humanitarian and economic development. The key objective of the West has always been to achieve long-run development in the Rest via the transfer of bilateral and multilateral resources, i.e. to promote growth by financing new investments in public goods. Aid money was provided to build roads, schools, electricity, ports, and other infrastructure which was meant to trigger the process of capital accumulation. Subsequently, if these investments are productive, the rate of growth should accelerate. The initial rationale for development aid in the 1950s was the assumption that the poorest countries are in a

poverty trap from which they cannot escape without an aid-financed Big Push (Easterly,

2007, p.33). Moreover, they also suffer from a financing gap to invest in infrastructure and human capital to propel development. Hence, with the help of the appropriate actions, self-sustained growth can be achieved and aid will no longer be needed. Unfortunately, history shows mixed results regarding aid effectiveness.

During the last couple of decades, vast amounts of development aid have been transferred to different regions of the world. See Table 1. The World Bank1 states that official development assistance was about $140 billion (current $) in 2011. About a third of this amount, $47.5 billion (current $), went to Sub-Saharan Africa. Which equals to $53.50 per capita. In contrast, East Asia & Pacific with their 1.9 billion people, which is more than twice as much compared to Sub-Saharan Africa, official development assistance per capita was only $3.54. According to Easterly (2007, p.4) the West has already spent as much as $2.3 trillion on development aid over the last five decades, but did not manage to bring significant progress in most of the developing world. Critics cite the widespread poverty in Africa and South Asia, and point to countries that have received significant amounts of aid but have disastrous growth records, i.e. Papua New Guinea, Sudan, and Haiti (Perkins et al., 2013, p.500). These critics see aid as a political tool that triggers corruption, distorts incentives, and supports corrupt dictators and elite business interests. In other words, according to the critics, aid has little effect on growth and often has done more harm than

1

(5)

5

good for the world’s poor. Hence, they favour for aid programs to be radically reformed, significantly curtailed, or abolished altogether.

Millions of US $ % of GNI Dollars per capita

East Asia & Pacific 7,848 0.04 3.54

Europe & Central

Asia 10,721 0.05 11.97

Latin America &

Caribbean 11,621 0.21 19.31

Middle East & North

Africa 15,556 0.5 40.09

South Asia 16,733 0.73 10.28

Sub-Saharan Africa 47,463 3.93 53.50

World 140,736 0.2 20.20

Table 1. Net Official Development Assistance in Current US $ by Region in 2011

Source: http://databank.worldbank.org “World Development Indicators”

Aid transfers came in different modalities, and five of the major ones are Project Aid, Program Aid, Sector-Wide Approaches, Humanitarian Assistance, and Technical Cooperation (Gibson et al., 2005, p.120). The results of these modalities are bundled in the aid effectiveness literature (AEL), and dozens of studies produced over the past fifty years have assessed the relationship between aid and growth through econometric analysis of cross-country data relating the two variables (Miller, 2012). Doucouliagos and Paldam (2009) surveyed this aid effectiveness literature – consisting of 97 econometric studies of three families – and concluded that after 40 years of development aid, the evidence indicates that aid has not been effective. In addition, Miller (2012) argues that the results of these macro-level studies turned up inconclusive or contradictory. Which means that there is a weak link between development aid and improved living conditions. At the micro-level, however, many funded projects report positive micro-level economic returns, but these are somehow

(6)

6

undetectable at the macro-level. This phenomenon is coined the micro-macro paradox (Miller, 2012, p.78).

Due to the contradictory results between macro- and micro-level studies, many scholars, policymakers, and aid practitioners have called into question the effectiveness of development aid to alleviate poverty and increase economic growth. Notably, Moyo (2009) and Easterly (2007) have argued that not only does aid fail to achieve its developmental aims, it also creates dependencies that keep countries poor and worsen the burden of poverty (Gulrajani, 2011, p.201). In addition, even without the dependency on aid, development, growth, and poverty reduction do take place (Riddell, 2007, p.255). On the other side, aid supporters such as Jeffrey Sachs and Joseph Stiglitz argue that although aid has not always worked well, its overall record is positive. It has been important in poverty reduction, accelerating economic growth, and achieving other development objectives in low-income countries such as health and education (Perkins et al., 2013, p.500).

It is clear that much energy is devoted in trying to prove whether or not official development aid does work. The empirical evidence on aid effectiveness is decidedly mixed, as various studies present marginal or no relationship between aid and development, while others show a positive impact. On balance, the evidence suggests that on average aid has had a modest positive impact on development outcomes, but with wide variation (Perkins et al., 2013, p.501). Instead of debating whether aid is effective or not, efforts have to be focused on understanding how to improve its impact (Riddell, 2007, p.410). Hence, this thesis will investigate the possibilities to improve aid effectiveness. In other words, how can we make aid work better? To answer this research question, an extensive literature survey has been conducted. This thesis will proceed as follows. Chapter two discusses the results of forty years of aid. Chapter three then looks at the harmful foundations of aid at the donor and recipient end. Subsequently, chapter four will put forward policy suggestions that attempt to improve aid’s impact. Finally, chapter five concludes.

2. How Effective has Aid been so Far?

Since the birth of foreign aid during the aftermath of World War II, aid has become an important form of international capital flow from the richest to poorest countries. Aid

(7)

7

effectiveness, however, has been a topic of much debate. The degree of aid effectiveness can be defined as aid that has achieved an objective, and the objective that currently dominates the aid effectiveness literature is that of economic growth for the recipient country (Miller, 2012). Perkins et al. (2013, p.518) also concur that economic growth has been the main benchmark used to assess aid effectiveness. However, as the following sections will show, objectives can take other forms as well. E.g. Perkins et al. (2013, p.501) argue that aid has been found to be successful in triggering development in some countries, and especially in areas such as health and agriculture. In other countries, however, it had little effect. Aid may even have held back development in cases where donors have given aid for either political purposes, or to political allies with corrupt and ineffective governments that had little or no interest in economic development.

This chapter will discuss four topics. First, the results of three generations of models relating aid to growth will be discussed in the aid effectiveness literature (AEL). Followed by the classification of those models in three views relating aid to economic growth. Third, a discussion surrounding macro-level studies will show that it is difficult to draw firm conclusions about the relationship between aid and growth on a macro level, whereas it is much easier to do so on a micro level. But first, a short elaboration on foreign aid will be presented.

2.1 Foreign Aid

Foreign aid entails the transfer of services, capital, or goods for development purposes, and it can be disbursed by governments, foundations, charities, businesses, or individuals. In order to be classified as aid, these transfers have to meet two criteria. First, they must promote economic development and welfare, which excludes aid for military or other non-development purposes. Second, they must be provided as either a grant or a subsidized loan (Perkins et al., 2013, p.502). Foreign aid transfers that are classified under the Official development Assistance (ODA), are flows to low- and middle-income countries and territories on the Development Assistance Committee2 (DAC) List of ODA recipients.

2_{The committee of the OECD which deals with development matters constitutes of 28 members: G8 (excl.}

Russia), Australia, Austria, Belgium, Czech Republic, Denmark, Finland, Greece, Iceland, Ireland, Luxembourg, the Netherlands, New Zealand, Norway, Poland, Portugal, Slovak Republic, South Korea, Spain, Sweden, Switzerland, and the European Union.

(8)

8

Besides the above mentioned criteria, most foreign aid is designed to meet one or more of the following four broad economic and development objectives (Perkins et al., 2013, p.518). First, stimulate economic growth through the building of infrastructure and institution, while supporting productive sectors such as agriculture, and introducing new technologies. Second, promote development objectives in improving education, the health sector, or political system. Third, support basic subsistence of food and shelter in emergency situations after humanitarian crises or natural disasters. Fourth, help stabilize an economy after economic shocks.

It is clear that aid transfers have a list of requirements before it can be called foreign aid. The following sections will discuss four topics that relate the aid to growth relationship. It will become evident that despite the thoughtful framework of foreign aid, over the years, aid has gone through trial and error. This aggravates the difficulty to make a strong unequivocal statement about its overall effectiveness.

2.2 The Aid Effectiveness Literature

After fifty years of aid transfers, dozens of studies have assessed the relationship between foreign aid and economic growth through econometric analysis of macro data relating these two variables. These macro studies are bundled in the Aid Effectiveness Literature (AEL), and they consist of three generations quantified in three families of models (Doucouliagos & Paldam, 2009; Roodman, 2007).

The first generation investigated the aid-savings relationship and runs from 1970 to 1972. It was based on a model of economic growth that assumed a linear development from aid to savings to growth. It focused on the importance of capital accumulation, as contemporary thought argued that aid-induced saving will lead directly to investment, and then to growth. This means that each dollar of foreign aid would result in an increase of one dollar in total savings and investment. Consequently, aid was not treated as a component of national income adding to both consumption and investment. Hence, aid fungibility in the form of consumption over savings was not accounted for. According to Hansen & Tarp (2000, p.382) the overwhelming results of these studies argued that aid leads to an increase in total savings, and implicated that aid stimulated growth. There is, however, a note. Since

(9)

9

the model did not account for the fungibility of aid, results of these studies tended to be overly positive, with aid contributing to higher economic growth.

The second generation focused in estimating the link between aid and growth via investment, and spanned the period from the early 1970s to the early 1990s (Hansen & Tarp, 2000, p.382). Contemporary thought assumed that investment was the major direct determinant of growth, and if a positive link can be found between aid and increased investment, one can conclude that aid contributed to economic growth. Miller (2012, p.80) argues that the results of these studies generally concluded a positive link between aid and investment. Thus, aid contributed to higher economic growth (Hansen & Tarp, 2000, p.385).

Finally, the third generation of models relate the aid and growth relationship via good policies, and argue that aid’s impact is conditional on the recipients’ good policy environment. These models were characterized by extended data sets comprising of more countries and years, and see the aid-growth relationship as non-linear. Moreover, since they are heavily influenced by the new growth theory, they also include regressors that cover economic policy and institutional environment. The third generation began with Boone (1994) who found no link between aid stimulating investment and growth. The most prominent study of the third generation of models, however, was the work of Burnside and Dollar (2000) in which they opposed the findings of Boone. The two authors found that aid does positively affect growth, provided that it is distributed to countries that have adopted sound policies i.e. good policy environments. It became an influential study, and its results were defended by researchers from the World Bank Group. However, contradictory results refuting this paper came to the foreground. One of the prominent challenger to the model used were Easterly et al. (2004, p.775) who added four more years and additional countries to the Burnside and Dollar data set, and concluded that the results did not hold. Doucouliagos & Paldam (2009, p.23) also concur that the model is fragile to changes in sample and control variables.

In their survey of the AEL, Doucouliagos & Paldam (2009) boldly argue that after 40 years of development aid, aid has not been effective. Perkins et al. (2013, p.519) also concur that there is no apparent simple relationship between aid and growth. Moreover, Miller (2012) states that the studies over the past fifty years which have assessed the relationship between foreign aid and growth through econometric analysis of macro data have turned up

(10)

10

inconclusive or contradictory. Also, five years earlier Bourguignon and Sundberg (2007) already stated that the empirical literature on aid effectiveness has yielded unclear and ambiguous results. According to the authors, those results were the consequence of conflicting aid motives, limitations of the tools of analysis, and the complex causality chain linking aid to development outcomes.

In response to the overwhelmingly negative view on aid effectiveness, Hansen and Tarp (2000) argue that after reviewing three generations of cross-country studies on aid effectiveness, there is a robust relationship that aid does improve economic performance. That is, aid increases aggregate savings, investment, and there is a positive relationship between aid and growth. Moreover, the authors acknowledge that the results of Burnside and Dollar are delicate (Hansen en Tarp, 2000, p.392). They argue, however, that the results of aid being only effective in a good policy environment is a suggestion how donors and aid recipients can learn from mistakes in the past and improve aid effectiveness in the future in a straightforward manner. In the end, the unresolved issue in assessing aid effectiveness is not whether aid works, but how and whether the different types of aid instruments can work better in varying country circumstances.

2.3 Three Views on Aid and Economic Growth

The discussion of the aid effectiveness literature made clear that the first and the second generation of models have concluded that aid has had a positive impact on growth via, respectively, savings and investment. Whereas there seems to be no consensus in the third generation of models relating the aid to growth relationship. Perkins et al. (2013, p.519) also acknowledge that there is no apparent simple relationship between aid and growth. There is, however, consensus between aid pessimists and optimists to seek for conditions under which aid works or does not work, and on what steps can be taken to make aid more effective. The authors find in the literature that the empirical evidence of aid effectiveness is mixed, as different studies reached different conclusions depending on the chosen time frame, countries, and assumptions underlying the research. They summarize these results of the relationship between aid and growth in three views. In the following paragraph, it will become evident that the second and the third view relating the aid-growth relationship

(11)

11

describe the third generation of models. Whereas the first view more or less captures the results of the first and the second generation of models.

The first view states that, on average, aid has had a positive impact on economic growth and development. That is, some studies found a positive relationship between aid and growth after controlling for the impact of other factors on growth, although with diminishing returns. This means that small amounts of aid might have a relatively large impact on growth, but each additional dollar of aid might have less effect. These studies found that other variables such as political conflict, policies, geography, and institutions explain much of the variance in growth rates among aid recipients. They conclude that after controlling for those variables and allowing for diminishing returns, a positive relationship emerges between aid and growth. Although with variances around a trend line. Figure 1-a depicts this view.

Contrary to the first view, the second view states that aid has had little or no effect on growth and may even undermine it. Figure 1-b depicts such a situation. Distinct causes include corruption, a country’s limited absorptive capacity, and aid fungibility. First, a prominent problem surrounding corruption is the disappearing of aid funds to government officials’ pocket. Aid may end up in their personal off-shore bank accounts or in the consumption of goods for private use that adds little stimulus to growth. Or in the worst case, aid may support political regimes that further impoverish rather than benefit the country. Second, a recipient’s limited absorptive capacity emerges when it lacks skilled governmental officials who are capable to manage the inflow of larger and larger aid flows. As a result, in the short-term, aid inflows in either money or physical form may stay idle. That is, large donations of medicines might sit unused in warehouses, because they cannot be delivered to clinics. Third, aid’s effectiveness may be undermined when aid finances activities for which it was not originally intended. That is, recipients substitute aid funds from their intended purpose to something else. The prime example of a negative aid fungibility happens when governments and individuals spend aid on consumption rather than development objectives. The authors, however, find in the literature that aid is not perfectly fungible. Much of it goes to its intended purpose, and recipients cannot or do not transform all aid flows into other purposes.

(12)

12

Figure 1. Three Views on Aid and Economic Growth.

(a) View 1: Aid has a positive impact on growth with diminishing returns after controlling for the impact of other variables. (b) View 2: Aid has little or no impact on growth and may have a negative impact. (c) View 3: Aid has a positive impact on growth in some circumstances (circles) but no impact in others (squares). Reprinted from Economics of Development (p.523), by Perkins et al., 2013, New York: W.W. Norton & Company 7th edition.

Finally, following Burnside and Dollar (2000), the third view on aid and economic growth argues that aid has a conditional relationship with growth, stimulating growth only under certain circumstances. This view accepts the idea that aid has had mixed results. That is, studies that have found a generally positive relationship do not claim that aid has worked across all countries all the time. They recognize that aid seems to have stimulated growth in

(13)

13

some countries under certain circumstances, but not in others. Subsequently, they try to decipher the key characteristics that may explain these differences. Figure 1-c captures this view, in which circles represent the positive impact of aid on growth in some circumstances, and whereas squares illustrate that it had no impact in others.

Perkins et al. (2013, p.534) state three “conditional” explanations which suggest this mixed relationship between aid and growth. I.e. the successful impact of aid might dependent on the type of aid being provided, the characteristics of the recipient country, or the way in which donors provide it. First, the type of aid that has been found to have a strong relationship between aid and growth is aid that is directly aimed to stimulate infrastructures, such as roads, electricity generators, or agriculture. Moreover, aid may even had an even stronger impact in countries with stronger institutions. However, for other types of aid, such as humanitarian aid, or aid for education, health, and democracy, the effect on economic growth was less detectable. Second, the impact of aid is found to be highly dependent on the characteristics of the recipient country. As mentioned earlier, Burnside & Dollar (2000) found a significant positive relationship between aid and growth only in countries with good policies and institutions. Despite the debates refuting their results, the findings of Burnside and Dollar (2000) led to a shift among donors to be more selective in allocating foreign aid to countries with relatively stronger institutions and policies. Finally, the third condition affecting the aid-growth relation concerns donor practices. Many analysts argue that differences in donor practices are likely to influence aid effectiveness. Moreover, to make the problem even worse, there seems to be no consensus whether aid must be provided bilaterally or multilaterally, and whether it must be tied or untied. Despite the many studies to date addressing the issue of donor practices, there has been little systematic research or consensus connecting specific donor practices to aid effectiveness.

In short, overall, it seems that aid has been successful in some countries, but not in others, with the overall trend being a subject of debate. Just like Hansen and Tarp (2000), Perkins et al. (2013, p.535) also suggest that further research must focus on revealing what types of aid are the most effective, and under which conditions aid has the largest impact on growth. Even when the findings are mixed, it will become evident that aid’s impact seems to

(14)

14

be successful on a micro-level. Before turning to micro-level studies, the next section will reason why it is meaningless to draw firm conclusions about aid’s impact on the macro-level.

2.4 Macro-level Studies

Like mentioned previously, macro-level studies in the aid effectiveness literature try to answer the question whether aid has been effective on a country or cross-country level. Many authors have concluded that the relation between aid and growth has been either negative, contradictory or inconclusive. Despite contradictory results, economists continue to debate the question of aid’s impact, only to end up with complex disagreements concerning the econometric analysis conducted, merely to establish claims and positions (Gulrajani, 2011). Hence, the debate around macro-level data and the question of aid’s effectiveness in general is empirically irresolvable, since it adds little improvement in the understanding of whether aid works (Miller, 2012). This section will argue that it is exactly the inherent qualities of country and cross-country studies that are guilty of such ambiguous results.

2.4.1 Country Level Studies

Despite the improved quality of information and data obtained in country-level studies, Riddell (2007, p.214) argues that the majority of such studies provide contradictory results from which to draw firm conclusions about aid’s impact at the country level. He argues two causes. Sometimes there has been insufficient data on the impact and performance of micro-level studies, which makes it difficult to make well-founded generalizations about overall performance on the country level. Also, despite the significant positive results of micro-level studies, evaluators have been reluctant to draw conclusions about its wider general impact of aid on a macro level. They argue that it is difficult to trace firmly a direct and causal relationship between overall growth and aid flows, or between overall aid flows and changes in poverty (Riddell, 2007, p.220). In particular, it is not possible to separate out the influence of aid from other factors, or to verify what would have happened in the absence of aid. Hence, this makes it difficult for evaluators to link directly the overall contribution of a particular donor’s aid program to the country’s aggregate growth performance or poverty reduction achievements.

(15)

15

Next to country studies, macro-level studies also consists of cross-country studies, which focus on aid across recipient countries, and the relationship between aid and key macroeconomic indicators. According to Rajan & Subramanian (2007, p.3), these cross-country aid-growth studies have followed a continuous cycle of validating or refuting papers. That is, those studies tend to follow a cycle in which one paper finds a result, and is followed by another paper with a slightly different model specification, either refuting or validating the previous result, which is then followed by another, and so on.

2.4.2 Cross-country Level Studies

Riddell (2007) argues three causes why cross-country studies add little improvement in the understanding of whether aid works. First, there is yet not enough knowledge about what triggers growth and reduces poverty to be able to trace the relationship between aid’s impact on key macroeconomic variables. Like mentioned previously, firm conclusions about aid’s general impact at the country level cannot be easily drawn. Hence, drawing conclusions at the cross-country level is even more dubious. Second, the statistical relationships between aggregates do not state the real life causality between variables, nor about whether the relationships observed are due to other influences, or to the interrelationship between aid and other internal influences. These statistical relationships merely capture and summarize regularities in the data used. Those econometric studies are useful in summarizing regularities in the data, but they do not have the richness of institutional and historical data that one can retrieve from a case study (Collier and Dollar, 2004, p.258). Third, the influence and impact of aid on the country level is dependent upon a range of country specific variables, which might change over time. It is difficult to quantify variables such as the recipient’s commitment and ownership to national and sectoral strategies upon which aid is provided. Even when it is possible to draw firm general conclusions about the past relationship between aid, growth, and poverty – which cross-country studies are all able to do – this would not picture a clear-cut blueprint for the future.

As a consequence of this accumulation of negative arguments, a growing number of researchers and aid specialists are beginning to question the value of studies that try to answer whether aid works, as cross-country regressions do not throw much light on aid effectiveness. Among these researchers is Esther Duflo (2003) who argued that the question

(16)

16

of whether development aid contributes to growth cannot really be answered, and trying to look for a cross-country relationship is devious. Moreover, Quibria (2004) an Asian Development Bank economist, reported in his working paper that because of data and methodological problems, cross-country growth regressions should make place for in-depth studies of individual countries that specify a tailored growth model for the country’s economy as the foundation for empirical analysis. In short, the aid industry should put a halt to cross-country studies that are addressed to prove that aid in general works, and to challenge the commonly and deeply held assumption that such studies are necessary to validate or confirm the justification of providing ODA (Riddell, 2007, p.224).

2.5 Micro-level Studies

Regardless of the negative results of aid, success stories on micro-level have been reported. This field of research seek to determine when and why aid is effective by looking at projects and programs on a case-by-case basis. The majority of micro-level studies that focuses on individual projects find that aid is working, whereas macro-level studies report aid to have a negligible, no observable or negative impact (Miller, 2012, p.78). Hence, it seems that aid successes on micro-level are undetectable on the macro-level. A phenomenon called the micro-macro paradox.

Instead of trying to answer the general question whether aid is effective, micro-level studies assume that sometimes it is and sometimes it is not, and try to identify the conditions and policies that lead to improve aid effectiveness. Micro-level studies consist of individual aid studies on a project level. It entails donor interventions in sectors such as health, education, rural development including power and agriculture. The focus is on achieving specific and concrete outputs i.e. to build or supply the ‘hardware’ as schools and textbooks, hospitals and medicine, or by delivering the ‘software’ in the form of skills training and know-how to properly use the hardware. Donors also emphasize the importance of sustainability in the end result of aid. I.e. whether the roads, textbooks, and training contribute to improving the lives and well-being of the recipients. Clearly a project is a waste of aid resources if it is ineffective or fails to achieve its immediate objectives, and if it is irrelevant to the needs of the recipient.

(17)

17

According to Riddell (2007, p.180) the results of project aid are overwhelmingly positive. In particular, where the intended outputs are clearly specified, most projects succeeded in producing or delivering those outputs. The author states that the share of projects that achieved their intended objective varies between donor and recipient countries, with success rates varying between 70% to 85%. Miller (2012, p.78) also concurs that these micro-level studies report high success rates. Cassen & Associates (1994, p.133) state successes in areas such as credit and agriculture projects, education projects in terms of schools built and the number of enrolments, and health projects in terms of declining birth rates. They conclude that the majority of projects have been successful in delivering their intended outputs.

Riddell (2007, p.181) focuses on project aid performance of World Bank projects, and argues that World Bank project aid performance increased from 60% in the mid-eighties to roughly 76% in the beginning of the millennium. However, with varying performances across different areas and countries. In terms of sectoral performance, the World Bank studies report high rates of success of over 85% in rural development, transport, and financial sector projects. Whereas comparatively lower rates of 70% have been achieved in environmental projects. The author also note that by comparing project performance between two time periods – early-millennium and end-nineties – an encouraging image emerge. That is, improvement in education projects rose from 74% to 81%, and health and nutrition projects rose from 76% to 83%. Relatively the biggest improvement was achieved in sanitation and water supply projects, where the success rate rose from 52% to 74%.

In sum, it seems that micro-level studies have been successful in delivering their intended outputs in specific areas. Moreover, as better conditions and policies are being implemented, higher success rates are likely to increase in the future.

3. Aid Related Problems

One of the world’s most renowned development economist Riddell (2007, p.357) argues in his book that institutional or systemic factors cause problems in the aid system, which impede the potential impact of aid. This chapter will discuss these problems that originate from three sources. The first section will argue that foreign aid in itself might undermine

(18)

18

principles that the aid industry is trying to promote. This means that the accountability, participation, and ownership of development programs by and between donors and recipients might be hindered. The second section will discuss several fundamental problems at the donor end. Finally, at the recipient end, aid effectiveness might be impeded by political and institutional constraints. Which, consequently, pose dilemmas for donors. That is, the central challenge surrounds between the willingness of donors to put aid to good use, and recipients’ weak institutions and governance to use aid effectively.

3.1 Aid Undermines Exactly the Principles it Intends to Promote

According to Moss et al. (2006, p.14) foreign aid may undermine the very principles the aid industry intends to promote between donors and recipients, i.e. accountability, participation, and ownership. The argument goes that large sustained external financial flows change the incentives of the receiving government and its citizens, regardless of the precise nature of donor practices. That is, aid flows an sich, apart from inefficiencies in the aid system itself, can influence the evolution of state-society relations. Hence, if donors provide the majority of public finance, and recipient governments are primarily accountable to those donors, the development of a credible social contract between the state and its citizens to foster those aid principles will be undermined.

First, governmental accountability to its citizens can be drastically reduced when large aid flows replace the need to raise revenues from the local economy. As a consequence, the government will no longer need to ensure the support of their citizens and its approval of their legislatures. This reliance on aid as a substitute for taxation means that the flow of resources to the state is uninfluenced by government efficiency, which results in a tendency for governments to underinvest in their developmental capacity. This is the moral hazard effect that arises as a consequence of this aid dependence. In other words, aid dependence reduces the incentive to reform inefficient institutions and implement good policies, which, consequently, weakens the government’s developmental performance and encourages rent-seeking (Moss et al, 2006, p.15). This decline in governmental accountability can also negatively influence the degree of democracy. Countries that rely on a greater proportion of ‘unearned’ income tend to be less democratic, and have less effective institutional mechanisms and accountability. Hence, building strong democratic

(19)

19

states depend on the positive correlation between the progressive growth of a democratic and accountable government, and the emergence of a state apparatus that has the extractive capacity to collect tax. Section 3.3 will further elaborate on this accountability problem at the recipient end.

Second, in providing aid, the participation and accountability of donors, recipient governments and its citizens are since 1999 explicitly promoted via a poverty reduction strategy paper (PRSP). This contract subjects the national budget to multiple rounds of consultation with civil society groups. However, participation by citizens can be undermined as a result of large flows of aid over a sustained period of time. That is, citizens’ approval of government policies will be less important to governments that receive large amounts of external support. As a consequence, governments will underfund public institutions that encourage citizens’ participation in development projects, and devote less time and resources to explaining and defending unpopular policy decisions to their citizens.

Third, large amounts of aid flows might also cause ownership of development goals by citizens to decline, since decision-making is triggered externally. That is, local officials responding to pressures from donors will result in departicipation of citizens with the local government and its development projects. As a consequence, citizens may view the local legislature as the place to demand for favours and patronage for their own causes, rather than pressing for developmental policy outcomes, thus reinforcing the patrimonial elements in the local political economy.

Another patrimonial example concerns the increase in social services provision by donors in notably health and education. Moss et al. (2006, p.18) state that it is not uncommon for donors to finance over half of the country’s public investment budget. This means that donors help recipients to implement basic social services that their local legislatures fail to undertake. In other words, donors are actually helping those legislatures to escape their accountability and ownership for their own developmental progress. As section 3.3 will argue, these accountability and ownership problems at the recipient end can strongly impede aid’s impact. Historically, increases in social services provision only belongs to democratic governments facing substantial participatory pressures from their citizens. That is, pressures from below encouraged governments who wanted to remain in power to provide better education and health services to their citizens. However, in most aid receiving

(20)

20

countries those pressures to build strong social services have been external. Consequently, malicious governments can escape their accountability, and spend their resources on non-developmental expenditures that help them to remain in office.

3.2 Fundamental Problems at the Donor End

This section focuses on three fundamental problems at the donor end constraining aid’s impact. This section looks among others at the information and incentives problem in the aid delivery process, the variability of aid volumes, and the multiplicity of donors and their activities.

3.2.1 Information and Incentives Problems

Martens et al (2001, p.4) argue that the performance of aid programmes is determined by the incentives embedded in the institutional environment of the aid agency and its delivery process. That is, individual agents’ incentives can differ substantially from those of the organisation that they work for, which thus lead to very different behavioural outcomes. However, this informational problem in the aid delivery process is not always easily noticeable. This may induce biases in the behaviour of agents, which may negatively affect the performance of aid programmes (Martens et al., p.10, 2001). Riddell (2007, p.362) argues that there is a dichotomy between the commitment towards more aid effectiveness and the incentive systems that drive and reward agency staff. That is, agents are not rewarded and assessed for their contribution towards achieving the goals of the agency in enhancing aid’s impact and reaching sustainable aid programmes and projects. Rather, agency staff are stimulated to disburse funds quickly, while few resources are allocated to designing and monitoring those funds. In addition, Gunning (2008, p.3) also claims that this behaviour is the result of the embedded incentives structures in donor agencies. There seems to be the reluctance to withhold aid from countries with weak institutions, a poor policy environment, and strong incentives to use aid in ways which are unlikely to reduce poverty. These moral hazard and adverse selection issues are consequences of the principal-agent theory that result from the delegation of tasks in organisations. There are two causes that induce information and incentives problems that affect agency behaviour. Namely,

(21)

21

having multiple principles and objectives, and the existence of a broken feedback loop between taxpayers-donors and the intended beneficiaries.

First, because large-scale aid organisations consist of multiple principles and objectives, incentive problems may arise at different levels of delegations. That is, the joint delegation of aid programmes may result in confusion over objectives, or agencies pursuing their own interests. These execution of tasks may create perverse incentives and divided loyalties to different principals. Moreover, a multilateral aid agency might exaggerate the problems of a beneficiary to justify and boost the importance of its existence. These problems require for the right institutional design to consist of mitigating them by installing incentive structures that both motivate agents to reveal relevant information to their principal as well as mechanisms which reduce biases in their behaviour. According to Martens et al. (p.14, 2001), this principle-agent relationship can be aggravated, since foreign aid organisations differ in two prominent characteristics from private enterprises.

a) Foreign aid organisations have usually multiple objectives, such as building schools and roads, and the finance of small-scale enterprises. Private enterprises, on the other hand, focus primarily on making profit.

b) Private enterprises have multiple principals (shareholders) who share the same objective to make profit, whereas foreign aid organisations have multiple principals (e.g. politicians and legislators) who rarely share the same objectives. One politician may for example allocate more resources to road construction, whereas another focuses more on medical research. Both being heavily influenced by their constituency.

These two characteristics complicate the internal organisation of foreign aid agencies, as they affect organisational performance and incentives of their individual agents. Consequently, aid officials will have to find a middle ground between their multiple political principals and different objectives.

Second, agency behaviour can also be affected since aid beneficiaries are not the same as the people who finance it, as the latter live in different countries and different political constituencies. E.g. this may cause agents to focus on tasks that are easily observable and will advance his career, even when this do not lead to the best outcome. As a result of this geographical and political separation between beneficiaries and

(22)

taxpayers-22

donors, the normal performance feedback process is hindered. That is, dissatisfied taxpayers-donors and beneficiaries cannot rely on their direct political representatives to exercise political pressure on the public administration or agency to improve the performance of particular transfer programmes. This broken feedback loop increases the costs for taxpayers-donors to obtain reliable information on the outcomes of the programmes that they finance (Martens et al., 2001, p.18). The intended beneficiaries, on the other hand, have no political leverage over their domestic politicians who approve these aid programmes, as they did not pay for the aid. Consequently, those politicians are almost free to use the aid money for different political purposes other than development. That is, the broken feedback loop induces strong incentive biases in foreign aid, which divert its original purpose.

These incentive biases are also aggravated since the interests of domestic suppliers of aid goods and services, the contractors, such as consultancy firms and experts who has a contractual relationship with the donor, dominate decision making. This moral hazard may be caused because donors have imperfect knowledge of contractor’s activities and are not entirely able to monitor them. This makes them the direct beneficiaries of aid. In addition, since they are part of the constituency of the decision-makers in the donor country and possessing first-hand information on the outcomes of the aid programmes in beneficiary countries, they also have direct leverage on domestic political decision-makers. This informational advantage may indirectly make them the ultimate beneficiaries of foreign aid. The intended beneficiaries’ interest, on the other hand, suffers from the geographical and political gap, and are too remote to overcome the direct beneficiaries’ leverage.

3.2.2 Variable Aid Volumes

According to Riddell (2007, p.359), the variability of aid volumes can undermine the potential impact of aid. This variability is caused by insufficient aid volumes, high volatility of aid volumes, and the voluntarism in aid-giving. First, like mentioned previously, the allocation of aid is primarily influenced by humanitarian and development considerations. However, commercial, strategic, and political interests of donors also determine and shape to whom aid is given. Those different interests of donors entail aid tying to the purchase of goods and services like technical and consultancy services originating from the donor

(23)

23

country. The author claims that almost 60 percent of all ODA remains tied (Riddell, 2007, p.358). Consequently, causing recipients to accept resources which are not a high priority for development, which erodes the impact of development aid. According to Riddell, however, this impact would be greater if the total amounts of aid provided match the humanitarian and development needs of poor countries (2007, p.359). This means that the international community are providing insufficient essential amounts of aid. Therefore, he states that the total amount of aid volumes has found to be twice or possibly more than three times as high as currently provided. Also, since the broken feedback loop already stated that the incentive biases of contractors can negatively affect aid’s impact, there must be less interference from commercial, strategic, and political influences on aid decisions.

Second, the volatility of aid volumes forms another factor that can impede aid´s impact. That is, the amounts of official aid flow vary from year to year depending among others on the economic state of donors. Like illustrated in figure 2, aid flows have been steadily declining in the nineties, whereas the first decade of the century has showed an increasing overall trend. Riddell (2007, p,359) claims that the aid literature seems to conclude that the volatility in the allocation of aid is particularly present at the receiving end of countries emerging from conflict, with aid levels often falling sharply at precisely the time when aid could be used more effectively. Therefore, the more unpredictable and volatile the amounts of aid that recipients receive, the harder it becomes for them to plan their development budget. As a result, aid recipients anticipate this by setting up spending plans on the assumption that they will receive lower quantities of aid, which reduces their ability to absorb higher amounts of aid if it does become available. More on this topic will be discussed in the next section covering the aid-institutions paradox at the recipient end.

Lastly, a key factor underlying the insufficient and high volatility of aid volumes are the result of the voluntary nature of aid-giving. That is, official aid given by donors for development and humanitarian purposes is on a completely voluntary basis. There is no system or mechanism to either enhance each individual donation, or to punish donors who decide to reduce their donations. Even international aid-pledging conferences provide only a loose framework where some attempt in favour of aggregate donations are made. Hence, whether donations are given individually or in the aggregate, they are not based on a system which effectively matches what each country needs, and what is provided.

(24)

24

Figure 2. Net Official Development Assistance Received in Current US$

Source: http://databank.worldbank.org “World Development Indicators”

3.2.3 The Multiplicity of Donors and Their Activities

With more than 200 official donors and agencies, and new aid donors, funds, and mechanism being formed annually, Riddell (2007, p.360) states that each year over 35,000 independent official aid transactions take place, with each aid recipient dealing with more than 25 different official donors. Donors individually, independently, and voluntarily decide how much aid they will provide in the form of projects or programmes. Like mentioned previously, those activities lead to the variability in aid volumes. Moreover, they also have to decide how much of this amount will be divided among those different aid projects and aid programmes. All in all, marking the complexity of the aid system. Also, donors often

(25)

25

compete with one another to fund projects and programmes. Alongside with this competition comes their entourage of administrators and consultants that share the same skills and specialities. This creates parallel systems and processes, which oversee similar sorts of projects and programmes, hence undermining and reducing the overall potential impact of aid.

Additional to the complexity of the aid system is the presence of NGOs engaging in development work. Riddell (2007, p.360) claims that their activities add another facet in the multiple aid transactions, as numerous NGOs are operational within the same country, and providing the same skills and services. This increases the risk of duplicating the activities of other NGO’s and official aid projects, which reduces the potential impact of aid. Notably larger NGOs have tended to dominate and overshadow the activities of smaller grassroots organisations with their skills and professionalism. Also, the quality of NGOs development projects and programmes have been mixed, as NGO activity is subject to no or very little regulation or monitoring. That is, some are run and managed by individuals with insufficient skills, resulting in few tangible gains to the beneficiaries, whereas others make a major difference to poor people and are cost-effective.

3.3 Aid Problems at the Recipient End

One of aid’s problems is its failure in helping to build the capacity and strength of institutions in recipient countries. A failure that covers the most fundamental problems and dilemmas of the aid relationship between donor and recipient. This section will discuss the negative effects of foreign aid impeding the development of domestic institutions, and several other incentives problems at the recipient end that together impede the impact of aid.

3.3.1 The Aid-Institutions Paradox

It is a widely accepted argument that good institutions and governance structures are critical for economic development and growth (Johansson, 2010, p.1205). Countries with better institutions not only invest more in human and physical capital, but they also use these factors more efficiently to achieve higher growth. Moreover, from a donor agency’s and recipient’s point of view, having an appropriate institutional set-up is essential to deliver aid in a credible manner (Martens et al., 2001, p.7). Gibson et al. (2005, p.14) argue that

(26)

26

development will occur only if political and economic institutions generate incentives that facilitate individuals’ achievement of development goals. These authors, however, believe that such institutions are absent in most countries which receive large amounts of development aid. They are either missing, weak, or bad.

Like mentioned previously, the provision of aid has distorting effects on outcomes that donors are trying to achieve via aid. Moss et al. (2006) focused their research on Africa, and argue that the disappointing progress of aid on capacity building is the result of political instability. While over a quarter of the World Bank’s credit to Africa was explicitly targeted to stimulate capacity building in that region (OED, 2005, p.9). It seems that the slow progress of state capacity building is due to the nature of African bureaucracies for being patrimonial and corrupt. As a consequence, they show little interest in the provision of public goods that are essential to development.

Accordingly, Moss et al. (2006, p.4) argue the existence of an aid-institutions paradox, in which the provision of large amount of aid has a negative effect on the institutional development of public institutions and state capacity. The authors identify three examples of this phenomenon. First, macro-economic effects due to large volumes of aid might cause institutional implications on policy making. One central problem has been the possibility of large ODA inflows affecting the real exchange rate and undermining the competitiveness of the export sector, i.e. Dutch disease. This potential loss of competitiveness means lower economic growth and exports, increased dependence on external assistance, and fewer jobs. Another economic problem focuses on the volatility of aid flows, which undermines the ability of recipient governments to budget appropriately. That is, aid inflow is found to be more volatile than domestic fiscal revenues. Since aid commitments often overestimate actual disbursements, aid volatility leads to poor budgeting and the underestimation of revenues. This complicates public policy making in important areas such as budgeting and planning. Hence, this fiscal uncertainty of dependence on external assistance makes long-term planning very difficult.

In addition to this difficulty, Moss et al. (2006, p.8) state that additional public spending tends toward consumption rather than investment. According to these authors, high levels of aid flows function as a ‘soft budget constraint’ where it encourages fiscal indiscipline. That is, when aid is seen as a subsidy, the aid fungibility problem emerges. As a

(27)

27

consequence, the building of institutional development for revenue collection, or the investment in human and physical capital might be discouraged. Instead, aid money is spent on consumption rather than development, potentially undercutting those efforts. Also, aid flows might potentially stimulate excessive and unsustainable levels of government consumption, which can lead to macro-imbalances.

Second, large volumes of aid might impede institutional change through its effect on the local policy-making dynamics. That is, certain aid practices can in fact reinforce the patrimonial behaviour of recipient governments at the expense of investing aid in core developmental areas. When donor projects are poorly integrated into national budgetary processes, and not subject to much transparency or effective control by donors, aid money can help sustain anti-developmental practices within the state apparatus (Moss et al., 2006, p.7). Local officials may see aid just as a set of scarce private goods to be allocated. This is especially the case when repeated fiscal crises are cured by short-term adjustment and debt management. These anti-development practices are disastrous for state institutions, as they worsen the patrimonial attraction of aid resources. Hence, in countries where political power means access to state privileges and rents, and political systems are sustained by complex clientelist relationships, aid becomes the fuel for political mismanagement. In other words, political elites have little incentive to change a situation where large volumes of aid provide substantial resources for patronage and nepotism. As a result, aid will be turned into a mechanism where it increases government consumption rather than public investment.

Finally, large volumes of aid might have negative effects on a country’s state capacity through its tax revenue collection. That is, in the long-term aid is found to hinder the process of institutional maturation essential to development, as it decreases a country’s taxation effort. In order to successfully move from aid dependency to economic self-sufficiency, it is crucial for the state to learn how to tax and improve its tax administration, as taxation is an useful indicator of state capacity. Hence, the collection of public revenues from citizens, in contrast to unearned income via natural resources or external assistance, is considered as an essential factor to establishing accountability between the state and society (Moss et al., 2006, p.10). Aid can also have a negative impact in the short-term when it is used as a substitute for domestic revenue, while keeping the same level of expenditure. This impact is worse when aid concerns grants instead of loans, since grants are considered to be a free

(28)

28

substitute for tax revenue, whereas loans must be repaid. Consequently, countries with weaker institutions are more likely to suffer from a larger negative impact of grants on tax revenues than countries with better institutions. In sum, one can argue that the high levels of aid flows correspond to low levels of tax revenue in the short-term, whereas countries suffer from a weak tax administration and institutional capacity in the long-term.

3.3.2 Politics

According to Riddell (2007, p.370) both donors as well as aid researchers share the same view that the crucial condition for aid effectiveness is the recipient government’ strong ownership and commitment to build and strengthened their institutions. They argue that ownership and commitment form the basis for the aid relationship between donors and recipients. Moreover, the impact of aid is not only dependent on the trio of ownership, commitment, and capacity, but also on good governance. The practice of good governance, and the commitment to it, are now regarded as the reference to which the performance and effectiveness of aid is assessed.

Governance encompasses institutions, activities, and processes that are involved in managing and running a country’s economic and political affairs (Riddell, 2007, p.372). Donors agree that good governance plays an essential part to aid effectiveness, and that there is a relationship between the ineffectiveness of aid and the quality of governance. Since the importance of good governance is increasing, donors have gradually begun to focus more closely at the way that politics influence the impact of aid (Riddell, p,374, 2007). Over the past decade researchers came to the conclusion that the root cause of governing problems lie in the nature of its political systems. Donors often provided aid without sufficiently understanding the political context in which their aid was meant to work. I.e. donors knowledge of country contexts and history were too shallow to fully understand the political sphere of the recipient country. That is, countries that are characterized by a neo-patrimonial or a clientelistic government suffer the most from aid ineffectiveness. These are countries where a small group of powerful individuals use state resources to secure the loyalty of their clients. Hence, aid that is provided with the aim of promoting good governance and strengthening institutions, not only risks failing to achieve its objectives, but

(29)

29

can have the unintentional effect of aggravating several of the very problems that it was supposed to solve.

It is clear that strong commitment and ownership, along with good governance, are crucial for aid effectiveness. Consequently, failures of governance initiatives to achieve their objectives can be attributed to weak commitment and ownership. On one hand, weak commitment is caused by the inability of political leaders to deliver on their promises, as they do not have the means to comply. On the other hand, developing and achieving ownership of a development strategy is complex, as it is a long-term process which involves negotiations with different and competing interests groups. Hence, the test of commitment and ownership lies not only on what is formally agreed, but also in the outcome of the interaction between agreement and the actual implementation to pursue those actions. Which means that tangible and sustainable gains can only be detected when donors extend their time-frame in assessing aid’s impact.

All in all, aid’s impact will be greater if donors grasp an understanding of the political context in which their aid was meant to work. Focusing especially on isolating and understanding the nature and effects of the constraints which limit the ability of aid to work more effectively. The following paragraph, however, will clarify that politicians have incentives problems that can heavily impede aid’s impact at the recipient end.

3.3.3 Incentives Problems

Erbeznik (2011, p.875) argues that the rule of law is a fundamental element that developing countries must foster in order to escape poverty. It is a system of rules that is transparent, public, and enforced regularly. Also equally against all persons, both for citizens and political elites, and with government accountability to the rule of law as the most essential factor. Hence, successful rule of law reforms to escape poverty depends on political elites’ willingness to enact reforms.

People respond to incentives, and it takes great willpower to resist the rewards in case of malicious actions. Hence, before promoting the will to reform on the part of political elites, their incentive structures must be changed in advanced. It is, however, difficult to remove a deep-rooted culture of corruption at the highest level, since the payoff of corruption to an economic agent is greater than the potential costs to that same individual.

(30)

30

Consequently, the will to reform suffers from a collective action and commitment problem, as those who currently benefit from the status quo resist to see institutional changes. Even if it is better for the society as a whole if systemic corruption is restricted or government accountability is strengthened. This means that political elites who benefit from the old institutions will face new costs and lose old benefits in times of reforms. They will, consequently, work against reform efforts which change the status quo, and try to undermine any incentives for change.

Keeping the aforementioned incentive problems in mind, Erbeznik (2011, p.882) argues that foreign aid offers a challenge to rule of law reform, since it can maintain the negative incentive structures of government officials and political elites. That is, foreign aid can undermine rule of law reform by reducing the costs of maintaining the status quo. In theory, foreign aid can promote good governance in assisting rule of law reform by, respectively, increasing salaries and reducing corruption, and by training the police and the judiciary. However, the unintended consequences of aid may overshadow any positive returns in those areas. Instead of promoting good governance, foreign aid can alter the incentives of governments and political elites. This may result in the ruling elite staying in power, since whoever controls the government also controls the money. As a result, a moral hazard problem arises in which people who are in power tend not to be the same people who have sincere reform efforts in mind. That is, according to the prisoner’s dilemma problem, it is always in the individual’s self-interest to promote personal gain and not to make sacrifices, regardless of whatever anyone else does. Thus, in the equilibrium position few political elites and government officials are prepared to make the sacrifices necessary for reform.

3.4 Summary

This chapter has discussed several aid related problems that find their origin in three sources. It argued that aid flows have undermined the establishment of accountability, participation, and ownership of aid programmes between donors and recipients. Moreover, fundamental problems at both the donor and recipient end have showed that incentives problems play a crucial role in impeding aid effectiveness. Those problems, subsequently, have altered an optimal aid relationship between donors and recipients.

(31)

31

The next chapter will put forward policy suggestions that attempt to improve aid’s impact by altering the way aid is provided, along with the matching criteria to do so. The chapter proposes the reorganisation of aid relationships as well as the emphasize to improve selectivity and accountability procedures. Moreover, by providing insurance, aid effectiveness can be improved.

4. Necessary Improvements

Enhancing the impact of aid is a multi-million-dollar business, which extends to the annual publication of thousands of reports written by aid officials and consultants focusing on how to make aid work better (Riddell, 2007, p.381). These reports comprise of numerous recommendations that are meant to enhance the impact of a particular aid project, sector, programme, or country. Donors, however, now realise that the impact of aid depends on the context in which it is provided, and on the capacity and commitment of recipients to use it wisely. That is, on one side, its impact is highly dependent on the competence, quality, and incentive structures present among their own agencies. On the other, aid’s impact depends on the quality of governance in the recipient country. Hence, how can foreign aid be strengthened, so it can reach it intended goals of catalysing economic growth and poverty alleviation?

According to the previous chapter, notably politics and incentives problems strongly impede aid effectiveness. Subsequently, this chapter looks at several policy suggestions comprising the provision and use of aid which attempt to solve such problems. The first topic emphasizes the urge to reorganise the aid relationship between donors and their recipients. It tries to solve the political and incentives structure in the aid delivery system by favouring one unified agency which will facilitate this. This first topic forms the basis for the following two sections. The second section underscores the importance of implementing a stricter country selectivity procedure, whereas the third section argues that donors need to enhance their accountability relationship between themselves and the end users. Finally, by proposing a complementary mechanism of aid provision, aid effectiveness can be improved by means of implementing insurance in developing countries.

Foreign aid - an analysis on its effectiveness and future application.

Foreign Aid - An Analysis on Its Effectiveness and Future Application

Contents

1.

Introduction

2.

How Effective has Aid been so Far?

3.

Aid Related Problems

4.

Necessary Improvements