Page 1 of 37
Choosing to Target
What Types of Countries Get Different
Types of World Bank Projects
By Matthew S. Winters*
bEginning with the Pearson Commission Report in 1969,
through the seminal Assessing Aid report from 1998 and its involve- ment with the 2002 Monterrey Consensus on Financing for Devel- opment, the World Bank has advocated selectivity in the provision of
foreign aid.1 The criteria for selection have been based, first, on the
quality of economic policy and, more recently, on the quality of govern- ing institutions. Given research claiming causal priority for rule-of-law
institutions in bringing about economic growth,2
governance-based se- lectivity seems an important mechanism for making foreign aid effec- tive as an engine of development.
Evidence of donor selectivity exists. Recent research shows that
some—but not all—donors take account of governance, rule of law,
and democracy when making lending decisions.3 However, as some of
*Previous versions of this paper were presented at the 2008 annual meeting of the Midwest Politi- cal Science Association, the 2009 annual meeting of the International Studies Association, the Prince- ton University ir Faculty Colloquium, and the Brooks World Poverty Institute’s Advanced Graduate
Workshop on Poverty, Development and Globalization at the University of Manchester. Thanks to
Kate Baldwin, David Epstein, Macartan Humphreys, Helen Milner, Kevin Morrison, Pablo Pinto,
Shanker Satyanath, Miguel Urquiola, Milan Vaishnav, Joseph Wright, and the World Politics review- ers and editors for comments. Thanks also to Jake Bowers, Devesh Kapur, Alex Scacco, Tracy Sulkin,
Michael Tierney, and Rebecca Weitz-Shapiro for useful conversations. The author wishes to acknowl- edge the financial support of the Institute for Social and Economic Policy and Research at Columbia
University and the Niehaus Center for Globalization and Governance at Princeton University. 1 The Pearson Commission says, “[I]ncreased allocation of aid should be primarily linked to per- formance.” Cited in Easterly 2007, 637. Assessing Aid argues that “financial assistance must be targeted
more effectively to low-income countries with sound economic management”; World Bank 1998, 4.
The Monterrey Consensus proclaims, “[S]ound policies and good governance at all levels are necessary
to ensure oda effectiveness”; United Nations 2002. 2 E.g., Acemoglu, Johnson, and Robinson 2001; Rodrik, Subramanian, and Trebbi 2004. 3 E.g., Neumayer 2003b; Berthélemy and Tichit 2004; Dollar and Levin 2006; Easterly 2007; Ber- meo 2008; Freytag and Pehnelt 2009. One could also read the empirical results this way in Bueno de
Mesquita and Smith 2007; and Bueno de Mesquita and Smith 2009. They find that larger winning
coalitions in recipient countries result in more aid. Although the authors argue that this supports
their theoretical model of foreign aid being used to purchase policy concessions, it could indicate that
donors reward countries for being more democratic.
World Politics 62, no. 3 ( July 2010), 422–58
Copyright © 2010 Trustees of Princeton University
doi: 10.1017/S0043887110000092
Page 2 of 37
choosing to target 423
4 A vast literature has established the dominance of political motives in determining bilateral aid
flows (e.g., McKinlay and Little 1977; McKinlay and Little 1978; Maizels and Nissanke 1984; Schrae- der, Hook, and Taylor 1998; Alesina and Dollar 2000; Bueno de Mesquita and Smith 2007; Bueno de
Mesquita and Smith 2009). 5
Maizels and Nissanke 1984 and Girord 2008, for example, argue that multilateral aid is more likely
to be given for development purposes. However, for evidence that the United States and other large
donors exercise extensive political influence inside multilateral institutions, see Frey and Schneider 1986;
Thacker 1999; Neumayer 2003a; Stone 2004; Barro and Lee 2005; Andersen, Hansen, and Markussen
2006; Fleck and Kilby 2006; Dreher and Jensen 2007; Dreher, Sturm, and Vreeland 2009. 6 The World Bank group comprises five organizations. The International Bank for Reconstruction
and Development (ibrd) and the International Development Association (ida) lend to member gov- ernments, the latter at concessional rates and using a significant amount of grants. The International
Finance Corporation (ifc) lends to the private sector in the developing world, whereas the Multilat- eral Investment Guarantee Agency (miga) provides insurance for investors largely from the developed
world. The International Court for the Settlement of Investment Disputes (icsid) is an arbitration
body that does not engage in the direct transfer of resources.
these papers acknowledge, political incentives, particularly for bilateral
donors, might shift the focus from the governance characteristics (or
other measures of deservingness) of aid-receiving states.4 Even though
some argue that multilateral aid—as compared with bilateral aid—is
less likely to be driven by strategic motives, substantial evidence sug- gests that it ultimately is.5
As in existing research, I, too, examine the impact of borrower
governance characteristics on overall World Bank aid flows and find
that, controlling for other country characteristics, the Bank gives more
money to well-governed countries. However, I also move beyond this
aggregate focus to look at specific aid project decisions, using an orig- inal data set of World Bank projects that allows me to see whether
the World Bank alters its lending strategies in terms of the modalities
through which it gives aid.
If the World Bank has a preference for aid being used for poverty re- duction and development, then, according to standard principal-agent
models, it should use more constrained forms of aid in situations where
it anticipates government malfeasance. Specifically, for countries with
worse governance characteristics, it should demonstrate a preference
for project aid (that is, specific investment loans) over programmatic
aid (that is, budgetary support or structural adjustment lending). In
addition, when the Bank uses project aid in countries with more severe
governance problems, it should target this aid at specific constituencies,
as more focused projects can be better monitored, thereby reducing
government discretion.
The data provide evidence that well-governed countries receive a
larger proportion of their project lending as nationwide projects. With
regard to programmatic lending, I find that the pattern varies across
the two main branches of the World Bank.6
For the World Bank’s
Page 3 of 37
424 world politics
concessional-lending wing, the International Development Associa- tion (ida), good governance surprisingly predicts less programmatic
lending. In contrast, for the World Bank’s market-rate-lending wing,
the International Bank for Reconstruction and Development (ibrd),
good governance predicts a higher proportion of programmatic lend- ing. I explore the extent to which these discrepant patterns might be
the result of political influence. Given that the ida relies on regular
budgetary replenishments from donor countries, it may be more sus- ceptible to such influence.
In the next section I discuss how the World Bank has conceptualized
the importance of governance in development and describe specifically
how governance characteristics are supposed to operate in World Bank
lending. I also offer an intuitive explanation for why a poverty-respon- sive institution should prefer more discretionary (that is, programmatic
and national) lending in well-governed states. After reviewing relevant
research on aid allocation, I describe the data that I use in this study. I
then examine trends in the total amount of aid that countries receive
and in the proportion of this aid that they receive as programmatic and
national projects. I check the robustness of the results by adding addi- tional potential confounding covariates, paying particular attention to
variables that other scholars have used to proxy for political influence
among aid donors.
I. The Role of Governance in World Bank Lending
For much of its history the World Bank’s discussion of governance in
borrower countries was limited.7 The Bank’s charter explicitly states:
“The Bank and its officers shall not interfere in the political affairs of
any member; nor shall they be influenced in their decisions by the po- litical character of the member or members concerned. Only economic
considerations shall be relevant to their decisions.”8
Bank staff inter- preted this “apolitical clause” in the extreme to mean that even the dis- cussion of corruption in recipient country governments was forbidden.
Since the charter also says that it can operate only through national
governments, the Bank generally has been constrained to work with
borrower governments as they currently exist.
7
For an excellent review of these issues and a description of how the World Bank came to develop
its current governance and anticorruption agenda, see Weaver 2008. 8 Article IV, Section 10.
Page 4 of 37
choosing to target 425
Nonetheless, it has been easy to see how poor governance has im- peded aid from meeting development objectives. Consequently, the
discourse on governance in the World Bank’s official lending policies
has grown over time. Since 1977 the Bank has divided ida funds us- ing the Performance-Based Allocation System (pba), which “tak[es]
into account country performance,” as defined in terms of “a borrower’s
implementation of sound policies and institutional arrangements con- ducive to sustainable economic growth and poverty reduction within a
framework of good governance.”9
In the early 1990s governance—defined as “accountability, openness
and predictability of government actions”—began to play a larger role
in the country policy and institutional assessment (cpia) measure that
is part of the pba: that index started to reward community participation
in projects and to punish the diversion of resources for nondevelop- ment purposes.10 These elements disappeared from the cpia in a 1995
revision, only to be reintroduced in 1997. In 1998 the Bank introduced
a “governance discount” to the cpia formula based on six governance
criteria that “reflect[ed] the importance attached to governance by ida
management and ida deputies.”11
The signal moment at the World Bank for the emergence of “good
governance” as the post–Washington Consensus development paradigm
was World Bank president James Wolfensohn’s October 1996 “cancer
of corruption” speech, in which he described corruption as inimical to
development. Under Wolfensohn’s leadership, members of the Bank’s
research department started producing the leading research on the re- lationship between governance and development,12 including the late
1990s papers suggesting that foreign aid works best in well-governed
countries.13 Although the original statement of the Burnside and Dol- lar hypothesis—that aid results in economic growth only conditional
9
World Bank 2001, iv. The ida lends about $10 billion per year to a subset of around eighty
eligible countries. The pba is used to allocate about two-thirds of ida resources. Lending to blend
countries—those that are also eligible for ibrd borrowing—and postconflict states falls outside the
purview of the system. World Bank 2007, 2, fn. 7, 24. Since its inception in 1960 ida credits and
grants have totaled over $180 billion. See World Bank, “What Is IDA? ” at http://go.worldbank.org/
ZRAOR8IWW0. 10 World Bank 2001, 5–6. It is worth noting that this change predates James Wolfensohn’s arrival
as president of the Bank. 11 World Bank 2001, 20.
12 E.g., Knack and Keefer 1995; World Bank 1997; Kaufmann, Kraay, and Zoido-Lobatón 1999.
Another early and influential paper in this area came from the imf; Mauro 1996. 13 World Bank 1998; Burnside and Dollar 2000; Burnside and Dollar 2004; Collier and Dollar
2002. For a general review of the role of politics in the foreign aid and growth literature, see Wright
and Winters 2010.
Page 5 of 37
426 world politics
on good policies—defined “good policies” as an index of macroeco- nomic outcomes (that is, fiscal deficit, inflation, and trade openness),
in their later research the authors shifted the focus from policies to
institutions, arguing that aid results in economic growth conditional on
good governance.
This focus on governance is related to the idea that where there are
clear rules of the game— expectations about government actions and
legal enforcement—then investors, firms, and households make more
efficient economic decisions that, in turn, bring about faster economic
development.14 In addition, good governance is taken to include ac- countability mechanisms through which a government responds di- rectly to the needs of its citizens, leading to better antipoverty policies
and less rent seeking and corruption. (These accountability mechanisms
may include democratic institutions, but this does not necessarily have
to be the case.) Today, the received wisdom asserts that economic de- velopment is hindered when states lack stable property rights and are
not accountable to their citizens.15
Another strain of research provides evidence that national gover- nance quality affects the implementation quality of development proj- ects. World Bank projects in countries with high levels of civil liber- ties or superior institutions perform better.16 The quality of governance
institutions appears to affect even the quality of World Bank project
preparation.17 One paper has concluded that reallocating aid to well- governed countries with the most extreme poverty could approximately
double the effectiveness of global aid.18
Therefore, insofar as the World Bank is a development institution
with the stated goal of eliminating poverty in the world, we would ex- pect World Bank staff to be motivated to target aid and to design aid
projects based on national governance characteristics. In addition to
the explicit incorporation of governance concerns into the pba system,
the Bank has also described its market-rate lending through the Inter- national Bank for Reconstruction and Development as “strongly influ- enced by the Bank’s assessment of countries’ policies and institutions,”
14 See the useful discussion in Birdsall 2007. Other studies show that foreign aid leads to better hu- man development outcomes in more democratic countries; Kosack 2003; Rajkumar and Swaroop 2008. 15 Some authors question whether good governance really leads to development (given the dif- ficulty of knowing whether to define fast authoritarian developers like China and Korea as having had
good governance) and therefore question whether the emphasis on governance leads to an efficient
allocation of donor resources. See, for example, Chang 2002; and Khan 2002. 16 Isham, Kaufmann, and Pritchett 1997; Dollar and Levin 2005. For a review of these issues, see
Winters 2010. 17 Wane 2004.
18 Collier and Dollar 2002.
Page 6 of 37
choosing to target 427
saying that “[n]onconcessional lending (ibrd) also goes primarily to
countries with good ratings for policies and institutions.”19
In a policy paper suggesting ways in which donors can improve aid
effectiveness, Steven Radelet argues, “Aid delivery mechanisms should
differ significantly between well-governed and poorly governed coun- tries.”20 He calls for well-governed countries to receive programmatic
aid—budgetary support—since they have well-functioning govern- ments that can allocate this money to appropriate development-ori- ented destinations, whereas he argues for poorly governed countries
to receive short-term project aid that can be implemented through
nongovernmental organizations (ngos), thereby avoiding the grabbing
hand of the state.21 Radelet writes: “Providing recipients with more
flexibility, greater latitude, and more ownership is precisely the right
way for donors to move in well-governed countries.... [I]n weak, fail- ing, and poorly governed countries, donors should retain a strong role
in setting priorities and designing programs.”22
This idea corresponds to the classic principal-agent issue of how
much discretion to allow an agent in implementing policy. When prin- cipals and agents have similar preferences, the principal should allow
more discretion because this reduces transaction costs and increases ef- ficiency. However, when preferences are different, the principal should
reduce discretion so that the end result will be more proximate to her
preferences.23 In well-governed countries, a donor can expect aid funds
to be used to build and maintain schools, roads, and health clinics. In
such cases, programmatic aid to the government is likely to allocate
resources more efficiently than would a series of specific projects de- signed in consultation with a foreign donor. By contrast, in a poorly
governed country, donors should expect a significantly lower likelihood
of the “right” type of projects being implemented. There is instead an
increased likelihood that a government will take advantage of its discre- tionary power to use aid funds to purchase military goods, to improve
the presidential palace, or to line the pockets of political allies.24 Since
19 However, “[i]t does not typically go to countries with the highest ratings [because] these coun- tries tend to have good, if volatile, access to capital markets”; World Bank 2002, 98, 101. 20 Radelet 2004, 3. 21 For a formal model that suggests that budgetary support is a superior instrument in countries
that are committed to development, see Cordella and Dell’Ariccia 2003. 22 Radelet 2004, 13. 23 The classic statement of this idea is Holmström 1984. For applications to foreign aid, see Svens- son 2000; Azam and Laffont 2003; and Hefeker and Michaelowa 2005. 24 In thinking about bilateral donors trading foreign aid for policy concessions by recipient govern- ments, Bueno de Mesquita and Smith propose that the bargain is more effective when aid-receiving
states have small winning coalitions—that is, when there are lower levels of democratic accountability
Page 11 of 37
Table 1
World Bank Lending Instruments and Project Typea
(1996–2002)
Lending Instrument Project Lending Programmatic Lending
Investment Project Instruments
Adaptable Program Loans 136 2
Financial Intermediary Loans 21 0
Emergency Recovery Loans 55 13
Sector Investment and Maintenance Loans 51 0
Specific Investment Loans 969 7
Learning and Innovation Loans 110 0
Technical Assistance Loans 145 0
Programmatic Instruments
Debt and Debt Service Loans 1 3
Rehabilitation Loans 0 4
Sector Adjustment Loans 3 87
Structural Adjustment Loans 11 183
Programmatic Structural Adjustment Loans 4 17
Poverty Reduction Support Credits 0 7
Total 1,506 323
Sources: Author’s coding of World Bank projects approved between 1996 and 2002 based on
relevant Project Appraisal Documents, Project Information Documents, Staff Appraisal Reports,
Technical Annexes, or Implementation Completion Reports. a
The programmatic column includes general budget support at the national level and debt relief
projects; the project column includes subnational structural adjustment loans for reasons described
in the text. The table does not include twenty projects from this period where the project involved
multiple national governments or an intergovernmental organization as the borrower.
Page 12 of 37
choosing to target 433
programmatic loans that I classify as project lending because they are
targeted at a subnational entity.38
In terms of classifying World Bank investment projects as being
national or subnational, the Bank itself does not describe projects ac- cording to how they are targeted. I therefore code investment projects
according to nine classifications—(1) a single city, (2) multiple cities,
(3) a single region, (4) multiple regions, (5) the rural sector, (6) the
urban sector, (7) a specific social group, (8) business or industry, and
(9) national.
The Bank sometimes—although rarely—lends to a subnational en- tity.39 In all of these cases the national government must also sign off
on the loan, even if responsibility for its implementation rests with a
subnational entity. I have coded these projects based on where the ben- efits were expected to go.
Figure 1 shows the distribution of projects in the data set across these
targeting categories, both in terms of project numbers and in terms of
dollar amounts. In general, there is a close correspondence between the
number of projects of a given type and the total amount of aid spent
by the World Bank in that category. With national technical assistance
projects, there are many very small projects, and with structural adjust- ment projects, the loan size tends to be quite large.
In the analysis that follows, I make use of a strict geographic defini- tion of subnational targeting, considering those projects targeted at one
or more cities or one or more regions as subnational projects. In a later
section, I examine the robustness of the results to an expansive defini- tion that also includes those projects targeted at the urban or rural sec- tors, specific social groups, or business.
The two histograms in Figure 2 show the variation that exists across
countries for the time period 1996–2002 with regard to World Bank aid
flows. In the upper histogram we can see that a large number of countries
(38 of 127) received no programmatic aid at all from the World Bank
during this period and that only two countries—the Slovak Republic
38 The other four projects are a commercial debt reduction project that benefits businesses in Côte
d’Ivoire (P053328), a technical assistance project that is a component of a structural adjustment project
in Malawi (P045739), and two large loans to Turkey that include either a social fund or other direct
grant components for the agricultural sector (P070286 and P074408). 39 As noted earlier, the Bank, by its charter, must lend to a national government, but it sometimes
enters into an arrangement where an entity other than the national government has responsibility for
a project. For example, a 1998 power system project involved money going directly to the Macedonian
national electric company, and the Bank made a series of structural adjustment loans to Argentine
provinces in 1997. In the late 1990s, according to former World Bank chief economist Joseph Stiglitz,
this type of arrangement was considered an option for countries with poor national governance or poor
national bureaucratic capacity (personal communication; June 2008).
Page 18 of 37
Figure 3
Explaining World Bank Commitments to Individual Countriesa
(1996–2002)
a
Point estimates and 95 percent confidence intervals. Outcome variable is ln(World Bank commit- ments + 1). All models are random effects ols models with HC3 robust standard errors. All models
include a quadratic time trend and regional fixed effects.
Overall World Bank, ida, and ibrd Allocations
(1) (2) (3) (4) (5) (6)
Governance
imf Program
Log (Population)
Log (Land Area)
Log (gdp
per Capita)
Log (Investment
Ratio)
Log (External
Debt Ratio)
WB WB IDA IDA IBRD IBRD
7.5
5
2.5
0
–2.5
3
0
–3
4
0
–4
3
2
1
0
4
2
0
4
2
0
1
0
–1
–2
N=934 N=857 N=552 N=486 N=382 N=371
j=120 j=118 j=72 j=71 j=48 j=47
R-Sq=0.17 R-Sq=0.16 R-Sq=0.20 R-Sq=0.20 R-Sq=0.15 R-Sq=0.15
Page 25 of 37
Figure 5
Explaining World Bank Commitments to Individual Countries, 1996–2002: Accounting for Political Variablesa
aPoint estimates and 95 percent confidence intervals. Outcome variable is ln(World Bank commitments + 1). The coefficient estimate reported in the row for Donor
Strategic Interest Measure refers to the variable listed on the horizontal axis. All models are random effects ols models with HC3 robust standard errors. All models
include the covariates found in Figure 3, a quadratic time trend, and regional fixed effects.
wb Governance
wb - Log/U.S. Military Assistance wb - Log (Trade with U.S.) wb - Log (Trade with oecd) wb - U.N. Voting with U.S. wb - U.N. Voting with G7 wb - Rotating UNSC Seat wb Log (oecd Migrant Stock) wb Former dac Colony wb Log (Oil Production) ida Governance ida - Log/U.S. Military Assistance ida - Log (Trade with U.S.) ida - Log (Trade with oecd) ida - U.N. Voting with U.S. ida - U.N. Voting with G7 ida - Rotating UNSC Seat ida (oecd Migrant Stock) ida Former dac Colony ida Log (Oil Production) ibrd Governance ibrd - Log/U.S. Military Assistance ibrd - Log (Trade with U.S.) ibrd - Log (Trade with oecd) ibrd - U.N. Voting with U.S. ibrd - U.N. Voting with G7 ibrd - Rotating UNSC Seat ibrd (oecd Migrant Stock) ibrd Former dac Colony ibrd (Oil Production)
7.5
5
2.5
0
12
9
6
3
0
–3
–6
Donor Strategic Interests and Overall Allocation Amounts
Governance
Donor Interest
Variable
Page 26 of 37
Figure 6
Explaining the Proportions of World Bank Programmatic and National Commitments to Individual Countries,
1996–2002: Accounting for Political Variablesa
aPoint estimates and 95 percent confidence intervals. Outcome variables are the proportion of programmatic lending (versus project lending) and the proportion
of national lending (versus subnational lending) in terms of the total value of lending over the period from 1996 to 2002. The coefficient estimate reported in the row
for Donor Strategic Interest Measure refers to the variable listed on the horizontal axis. All models are linear regressions including the covariates found in Figure 4
and regional fixed effects.
wb Governance
wb - Log (Trade with U.S.) wb - Log (Trade with oecd) wb - U.N. Voting with U.S. wb - U.N. Voting with G7 wb - Rotating UNSC Seat wb Log (oecd Migrant Stock) wb Former dac Colony wb Log (Oil Production) ida Governance ida - Log/U.S. Military Assistance ida - Log (Trade with U.S.) ida - Log (Trade with oecd) ida - U.N. Voting with U.S. ida - U.N. Voting with G7 ida - Rotating UNSC Seat ida (oecd Migrant Stock) ida Former dac Colony ida Log (Oil Production) ibrd Governance ibrd - Log/U.S. Military Assistance ibrd - Log (Trade with U.S.) ibrd - Log (Trade with oecd) ibrd - U.N. Voting with U.S. ibrd - U.N. Voting with G7 ibrd - Rotating UNSC Seat ibrd (oecd Migrant Stock) ibrd Former dac Colony ibrd (Oil Production) Prog Lending Pct Natnl Lending Pct wb - Log/U.S. Military Assistance
0.3
0.2
0.1
0
–0.1
–0.2
0.9
0.6
0.3
0
–0.3
–0.6
Governance
Donor Strategic
Interest Measure
Donor Strategic Interests and Proportions of National and Programmatic Lending
Page 36 of 37
choosing to target 457
Knack, Steve, and Philip Keefer. 1995. “Institutions and Economic Performance:
Cross-Country Tests Using Alternative Institutional Measures.” Economics and
Politics 7, no. 3: 207–27.
Kosack, Stephen. 2003. “Effective Aid: How Democracy Allows Development
Aid to Improve the Quality of Life.” World Development 31, no. 1: 1–22.
Lahiri, Sajal, and Pascalis Raimondos-Moller. 2000. “Lobbying by Ethnic Groups
and Aid Allocation.” Economic Journal 110, no. 462: C62–C79.
Maizels, Alfred, and Machiko K. Nissanke. 1984. “Motivations for Aid to Devel- oping Countries.” World Development 12, no. 9: 879–900.
Mauro, Paolo. 1996. “The Effects of Corruption on Growth, Investment and
Government Expenditure.” International Monetary Fund Working Paper
96/98.
McKinlay, R. D., and R. Little. 1977. “A Foreign Policy Model of U.S. Bilateral
Aid Allocation.” World Politics 30, no. 1 (October): 58–86.
———. 1978. “A Foreign Policy Model of the Distribution of British Bilateral
Aid, 1960–70.” British Journal of Political Science 8, no. 3: 313–31.
Neumayer, Eric. 2003a. “The Determinants of Aid Allocation by Regional Mul- tilateral Development Banks and United Nations Agencies.” International
Studies Quarterly 47, no. 1: 101–22.
———. 2003b. The Pattern of Aid Giving: The Impact of Good Governance on De- velopment Assistance. New York: Routledge.
Olson, Mancur. 1971. The Logic of Collective Action: Public Goods and the Theory of
Groups, 2nd ed. Cambridge: Harvard University Press.
Radelet, Steven. 2004. “Aid Effectiveness and the Millenium Development
Goals.” Center for Global Development Working Paper 39.
Rajkumar, Andrew Sunil, and Vinaya Swaroop. 2008. “Public Spending and Out- comes: Does Governance Matter?” Journal of Development Economics 86, no. 2:
96–111.
Reinikka, Ritva, and Jakob Svensson. 2005. “Fighting Corruption to Improve
Schooling: Evidence from a Newspaper Campaign in Uganda.” Journal of the
European Economic Association 3, no. 2–3: 259–67.
Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi. 2004. “Institutions
Rule: The Primacy of Institutions over Geography and Integration in Eco- nomic Development.” Journal of Economic Growth 9, no. 2: 131–65.
Schraeder, Peter J., Steven W. Hook, and Bruce Taylor. 1998. “Clarifying the For- eign Aid Puzzle: A Comparison of American, Japanese, French, and Swedish
Aid Flows.” World Politics 50, no. 2 ( January): 294–323.
Stone, Randall W. 2004. “The Political Economy of imf Lending in Africa.”
American Political Science Review 98, no. 4: 577–91.
Svensson, Jakob. 1999. “Aid, Growth and Democracy.” Economics and Politics 11,
no. 3: 275–97.
———. 2000. “When Is Foreign Aid Policy Credible? Aid Dependence and
Conditionality.” Journal of Development Economics 61, no. 1: 61–84.
Thacker, Strom C. 1999. “The High Politics of imf Lending.” World Politics 52,
no. 1 (October): 38–75.
United Nations. 2002. Final Outcome of the International Conference on Fi- nancing for Development: Note by the Secretariat. Vol. A/CONF.198/3
Monterrey, Mexico.
Page 37 of 37
458 world politics
van de Walle, Nicolas. 2001. African Economies and the Politics of Permanent Crisis,
1979–1999. New York: Cambridge University Press.
Wane, Waly. 2004. “The Quality of Foreign Aid: Country Selectivity or Donors
Incentives? ” World Bank Policy Research Working Paper 3325.
Weaver, Catherine. 2008. Hypocrisy Trap: The World Bank and the Poverty of Re- form. Princeton: Princeton University Press.
Winters, Matthew S. 2010. “Accountability, Participation and Foreign Aid Ef- fectiveness.” International Studies Review 12, no. 2.
World Bank. 1997. World Development Report 1997: The State in a Changing World.
Washington, D.C.: World Bank Group.
———. 1998. Assessing Aid: What Works, What Doesn’t, and Why. New York: Ox- ford University Press.
———. 2001. “Review of the Performance-Based Allocation System, ida10-12.”
Washington, D.C.: World Bank Group.
———. 2002. A Case for Aid: Building a Consensus for Development Assistance.
Washington, D.C.: World Bank.
———. 2007. “ida’s Performance-Based Allocation System: Options for Simpli- fying the Formula and Reducing Volatility.” Washington, D.C.: World Bank
Group.
Wright, Joseph, and Matthew Winters. 2010. “The Politics of Effective Foreign
Aid.” Annual Review of Political Science 13: 61–80.