Have Poverty Reduction Strategy Papers actually reduced poverty in poor countries?

(1)

Have Poverty Reduction Strategy Papers actually reduced poverty in poor countries?

Master Thesis in Global Studies (30 hec) Spring semester 2019

Author: Markus Bohlers Supervisor: Arne Wackenhut

Word count: 18,709 words University of Gothenburg

School of Global Studies

GS2534

(2)

1

ABSTRACT

Throughout the past 15-20 years, poor countries have been implementing so-called Poverty Reduction Strategy Papers (PRSPs) under the guidance of the International Monetary Fund and the World Bank.

Despite their enormous importance for national development efforts around the globe, the aggregate performance of PRSPs remains understudied. One previous study, however, found that PRSPs were successful in reducing poverty and should therefore act as the ‘mechanism to operationalise’ the Sustainable Development Goals (SDGs) (Elkins et al 2018., 388).

Here, the robustness of these findings is tested in a panel data regression analysis which estimates the effect of PRSP implementation on five primary SDG-indicators.¹ Based on the assumption that the ‘Washington Consensus’ development paradigm has been (partially) abandoned, it is hypothesized that PRSPs successfully reduced multidimensional poverty, and that this success was conditioned on increased economic growth.

While PRSP implementation did not yield a statistically significant effect on most outcome variables, it was associated with small reductions in extreme, chronic undernourishment. It is acknowledged that the results have a relatively weak empirical foundation and that they may suffer from omitted variable bias. The author stresses the need for further research and calls for an increase in resources allocated to the collection of well-developed data on multidimensional poverty.

Keywords: Poverty, PRSPs, poor countries, IFIs, development paradigm, Washington Consensus.

1 Contact the author at gusmarkbo@student.gu.se to access the full Stata-data set.

(3)

2

ACKNOWLEDGEMENTS

Many thanks to my supervisor Arne Wackenhut for his helpful insights and constructive criticism. I would also like to thank Valeriya Mechkova for clearing up some questions about the method, and Meg Elkins for providing me with her dataset.

(4)

3

ABBREVIATIONS:

APR Annual Progress Report

DAC Development Assistance Committee FAO Food and Agricultural Organization FDI Foreign Direct Investment

FE Fixed Effects

GDP Gross Domestic Product

GATT General Agreement on Tariffs and Trade HIPC Heavily Indebted Poor Country

IBRD International Bank for Reconstruction and Development IFI International Financial Institutions

IMF International Monetary Fund IEO Independent Evaluations Office IPL International Poverty Line MDG Millennium Development Goal NGO Non-Governmental Organizations NIEO New International Economic Order NPL National Poverty Line

OECD Organization for Economic Co-operation and Development OED Operations Evaluation Department

ODA Official Development Assistance OLS Ordinary Least Square

OPEC Organization of the Petroleum Exporting Countries PPP Purchasing Power Parity

PRGF Poverty Reduction and Growth Facility PRSP Poverty Reduction Strategy Paper SAP Structural Adjustment Program SDG Sustainable Development Goals WDI World Development Indicator WGI Worldwide Governance Indicator WTO World Trade Organization

(7)

6

1. INTRODUCTION 1.1. Background

The turn of the new millennium marked the beginning of the Poverty Reduction Strategy initiative, launched by the International Monetary Fund (IMF) and the World Bank. Under the guidance of these international financial institutions (IFIs), poor countries subsequently began to implement Poverty Reduction Strategy Papers (PRSPs), and in exchange, these countries received access to debt relief and loans on concessional terms (Elkins et al. 2018, 378). The strategy papers contained macroeconomic, structural and social policies aimed at reducing poverty and stimulating sustained economic growth (Wachira & Ruger 2011, 1957). In the absence of an alternative roadmap, PRSPs also became a main vehicle to achieve the Millennium Development Goals (MDGs)by 2015 (Elkins et al. 2018, 378-379).²

On paper, PRSPs were ‘owned’ by the implementing country. However, strategies were expected to align with the development goals and principles of the IFIs’ Comprehensive Development Framework, and for several years, every PRSP required ‘explicit endorsement’ by the executive boards of the IFIs (Peet 2009, 145-147; Sumner 2006, 1401)³. It is fair to say, therefore, that the IFIs wielded considerable influence over the strategies.

Despite their importance for national development efforts around the globe, the actual performance of PRSPs remains understudied. To my knowledge, the only quantitative appraisal of PRSP implementation that has been published in a peer-reviewed academic journal is by Meg Elkins, Simon Feeny and David Prentice (2018). They estimate the effect of implementation on the main indicators underpinning the Millennium Development Goals (MDGs). Their results show that PRSPs successfully helped poor countries to reduce multidimensional poverty. Based on these findings, the authors suggest that the PRSP approach should remain the “mechanism to operationalise” the Sustainable Development Goals (SDGs), which is set to be achieved by 2030 (Elkins et al. 2018, 388).⁴

2 The 8 MDGs (see appendix A) were announced by the UN Secretariat in 2000. They crystallized the previously adopted Millennium Declaration and provided a plan of action to “meet the needs of the world’s poorest” by 2015 (UN 2019a). In the final MDG report, UN Secretary-General Ban Ki-Moon proclaimed that, although progress had been uneven and many countries had failed to accomplish all the goals, the MDGs still produced “the most successful anti- poverty movement in history” (UN 2015, 3).

3 In at least a few cases, the boards decided not to approve a strategy paper prepared by a government (in collaboration with other stakeholders) (Dijkstra & Komives 2009, 10).

4 The 17 SDGs (see Appendix A) originate from a 2015 UN resolution called “Transforming our World: the 2030 Agenda for Sustainable Development” (UN 2019c). These goals are even more ambitious than the MDGs, covering a total of 169 targets. The SDG commitment also amounts to a broader and more radical economic agenda for the whole world, covering areas such as environmental sustainability, economic growth, industry and production.

(8)

7

1.2. Aim and Research Question

The aim of this thesis is to test the robustness of the results in Elkins et al. (2018). In other words, I aim to uncover if PRSPs actually succeeded to reduce poverty in poor countries. Following the suggestion by Elkins et al., it seemed appropriate to evaluate the performance of PRSP implementation using the relevant SDGs (or rather, the indicators underpinning them) as outcome variables. This will also allow me to capture different dimensions of poverty. My research question reads as follows:

Has implementing Poverty Reduction Strategy Papers reduced poverty in its various dimensions, and if so, to what extent?

1.3. Hypotheses and Theoretical Assumption

As per convention, my primary hypothesis is the null hypothesis (H0): Implementing PRSPs had no effect on poverty in its various dimensions. However, as mentioned, Elkins et al. (2018) provide some empirical evidence for an alternative hypothesis; that (H1) implementing PRSPs reduced poverty in its various dimensions.

This is corroborated by another, unpublished study by Marshall and Bernard Walters (2011), which also found PRSP implementation to be associated with reduced poverty.

In order to establish a theoretical basis for why PRSPs were a success, I also draw on Elkins et al. (2018), and specifically, their results suggesting that successful implementation was, for the most part⁵, not conditioned on closer alignment with the ‘Washington Consensus’—the neoliberal development paradigm pushed in the Structural Adjustment Programs (SAPs) of the 1980s and 90s.

Indeed, the SAPs are generally understood to have failed to reduce poverty and stimulate growth (see for instance Cornia et al. 1987; Johnson & Schaefer 1999, Easterly 2005). The alignment scores utilized in Elkins et al. (2018)—which are built on a scorecard system developed in two separate content analyses by Elkins (2014) and Elkins and Feeny (2014)⁶—suggest that many PRSPs instead embraced more diverse strategies influenced by development paradigms emphasizing ‘good governance’ and

‘pro-poor’ social spending (which will be discussed at length in chapter 2). In fact, Elkins et al. (2018) found some evidence that successful PRSP implementation was conditioned on closer alignment to the ‘New York Consensus’—the development paradigm associated with the MDG commitment.

5 The effect on maternal mortality represents an interesting exception, which I will return to in chapter 4.

6 Since these scores only stretches to 2008, they will not be utilized in this study.

(9)

8 In other words, while many have argued that PRSPs failed to sufficiently distance themselves from the past (see for instance Hermele 2005, Stewart and Wang 2005; Fukuda-Parr 2010), it appears as if the PRSP approach at least represented a step in the right direction—away from a one-size-fits-all formula, unyielding IFI-conditionalities and the neoliberal orthodoxy of market solutions. Here, I assume that this step was significant enough to render the aggregate performance of PRSP implementation a success.

Echoing Marshall and Walters (2011), I also propose a complementary conditional hypothesis (H2); that successful implementation was conditioned on increasing economic growth. Indeed, Marshall and Walters found some evidence that this was the case. It is also in line with my theoretical argument, as abandoning neoliberal orthodoxy does not mean abandoning economic growth as a mechanism to reduce poverty. After all, PRSPs were aimed towards stimulating growth alongside poverty reduction.

1.4. Method and Operationalization

To test my hypotheses, I will estimate the effect of PRSP implementation on poverty-related SDG- indicators in a linear panel data regression analysis using the difference-in-difference method. My regression models will be run with both country-fixed effects and year-fixed effects, as well as with two control variables, aid received and political stability. This will allow me to control for the effect of unobserved differences across countries and changes over time. My full sample is comprised of an unbalanced panel dataset of 103 non-fragile poor countries over 17 years—between 2000 and 2016.

In many ways, my study serves as a replication study of Elkins et al. (2018). The use of outcome variables based on the SDGs will test the robustness of their findings, as will the inclusion of different control variables and the interaction with Gross Domestic Product (GDP) per capita⁷. I will also consider the main effect if PRSP-implementation is measured differently; if outlier countries are removed; if robust standard errors are employed; and if the outcome variables are transformed using the natural logarithm.

1.5. Relevance

The PRSP approach has been at the heart of international development efforts throughout the past two decades. PRSPs have functioned as national development plans for well over 60 countries (see

7 The interaction models will allow me to consider the effect of economic growth without controlling away the effect of implementation. Such potential post-treatment bias is not addressed in Elkins et al. (2018).

(10)

9 IMF 2019). Despite this, we know very little about their aggregate performance. By beginning to fill this gap in the literature, I argue that my study will make a valuable scholarly contribution of great relevance to the field of Global Studies.

Evaluating PRSP performance also places the IFIs under scrutiny. Indeed, the findings of this study may allow us to draw some conclusions about the judgement and expertise of these powerful institutions; whether or not their continued influence over economic policy in poor countries can be justified. To be sure, if we can conclude that PRSPs failed to reduce poverty then that is arguably a serious indictment of the IFIs and their role in leading and shaping international development efforts.

If, on the contrary, we find a substantial reduction in poverty, the suggestion by Elkins et al. (2018) that PRSPs should remain an instrument to achieve the SDGs will be validated.

1.6. Delimitations

Unfortunately, my analysis will not be able to factor in how closely PRSPs were aligned with different development paradigms; whether each strategy was fully and correctly implemented; or to which degree they can be said to have been ‘owned’ by the implementing country. Although it is difficult to imagine a reliable method for quantitively determining national ownership of PRSPs, both paradigm alignment and full/correct implementation could have been determined by systematically evaluating every PRSP and every Annual Progress Report (APR) published for implementing countries. Such an undertaking, however, is way beyond the scope of my thesis project.

Furthermore, it is important to emphasize that even if probabilistic causation can be established using this method—theoretically, panel data regression can show that PRSP implementation probably caused changes in poverty rates—my analysis will need to grapple with potential endogeneity problems and limitations in the internal validity of the data (which I will discuss at length in chapter 5).

1.7. Ethical Considerations

Given that this is an observational study in which only publicly available and widely used secondary data is being handled, no ethical considerations were raised when conducting it. There is, however, a risk that the results of my analysis will be used to justify policies of ethically questionable nature. Once again, therefore, I want to clarify that these results should not be interpreted as definitive proof of the success or failure of the PRSP approach.

(11)

10

1.8. Outline

The rest of this study is divided into six chapters: Theory; Background; Literature Review; Method and Operationalizations; Results and Analysis; and the final Conclusion.

Chapter 2 begins with a brief account of my epistemological point of departure (section 2.1).

This leads us to a discussion about the theoretical frameworks and concepts relevant to the study;

most notably, poverty and development paradigms (section 2.2). The remainder of the chapter outlines and contextualizes the three dominant development paradigms which have been showed to drive policy in PRSPs: The 'Washington Consensus’ (section 2.3), ‘post-Washington Consensus’ (2.4) and ‘New York Consensus’ (2.4) (Elkins and Feeny 2005).

Chapter 3 should be read as an extension of the previous (sub)chapter. It covers the historical background of the IFIs and the conditions and developments leading up to the ‘Third World debt crisis’ (section 3.1), as well as the ensuing era of ‘structural adjustment’ that allowed the ‘Washington Consensus’ to be applied in practice (3.2). The chapter ends with an overview of the PRSPs; how they came to be and what they are (section 3.3).

In Chapter 4, the literature on PRSPs is reviewed at length. The first section (4.1) provides an overview of the literature on the PRSP approach—on the process of implementation and on policy content—while the second section (4.2) offers a more exhaustive account of previous quantitative studies and reports on PRSP performance.

The first three sections (5.1; 5.2; 5.3) in Chapter 5 describes which method is used to conduct my analysis and motivates why this method is chosen. The following four sections describe and discuss the sample (section 5.4); the explanatory variable (5.5); the outcome variables (5.6); the control variables and the moderator variable (5.8)⁸; as well as potential endogeneity problems (5.7) and limitations to the internal validity of the data (which is discussed throughout).

Next, in Chapter 6, the results of the regression tests are reported and analyzed. First, we analyze the results of the additive models (section 6.1) and the interaction models (6.2). To check the robustness of the results, a residual analysis is carried out (section 6.3), followed by an analysis of the robust additive models (6.4) and the robust interaction models (6.5). Lastly, the whole analysis is summarized (section 6.6).

As a conclusion, Chapter 7 offers a brief overview of the study, a final discussion of my findings and what they mean for the IFIs, followed by a few suggestions on further research.

8 Descriptive statistics and detailed definitions of all variables can be found in Table 9 and Table 10 in Appendix C.

(12)

11

2. THEORY

2.1. Epistemological Approach

Since the purpose of this study is to try to isolate and observe an empirical link between PRSP implementation and poverty, my epistemological point of departure is heavily influenced by empiricism.

In other words, I take the position that ‘true’ knowledge, by and large, emanates from observations in/of the ‘real’ world.

That being said, I also draw on the critical realist tradition in order to make a distinction between the ‘real’ world, which exists regardless of our interpretations of it, and the ‘actual’ world, which is the world we think we know through our interpretations. These interpretations are tainted not just by our sensory filters but by discursive filters, which, in turn, are always situated in a particular social, cultural and historical context (Haraway 1988). In the following section, therefore, I will not only define the key concepts and frameworks used in this study, but try to contextualize them.

2.2. Theoretical Framework and Key Concepts

Development paradigms. In my study, the concept of paradigm is understood to be roughly synonymous with discourse; that is, a particular way of interpreting and communicating (parts of) the world (Jorgensen & Phillips 2002, 1). A ‘development paradigm’, then, encapsulates certain views on/ideas about/approaches to (economic) development. It could be said that the three development paradigms discussed in the next section are the ones competing for hegemony in the ‘discursive field’

of development economics (Hansen 2006, 7; Jorgensen & Phillips 2002, 66-73). I should add, however, that while there is arguably enough discursive cohesion to justify why distinguishing these three paradigms from each other serves more than merely an analytical purpose, such distinction is in and of itself a discursive construction that may obscure important overlap and therefore cannot be said to perfectly represent the ‘real’ world.

Economics. In a similar vein, the field economics is founded on a number of assumptions about the world, which together, through intersection, create a particular discursive framework. Since my study is primarily situated in development economics, I want to clarify that this framework places inevitable restrictions on our ontological horizons, and some of them, arguably, are quite problematic. Below, I shall discuss these restrictions at some length.

(13)

12 First of all, it has been convincingly argued that the invention of economics in the latter half of the 19^th century effectively served to depoliticize knowledge about the economy, transforming into a less contestable and more mathematically elegant ‘natural’ science (Peet & Hartwick 2014, 56-62). This natural science, consequently, insulated ‘economists’ from other aspects of society and enforced the notion that they acted as objective observers of, rather than participants in, economic life (ibid.). While I have already established my rejection of dogmatic forms of empiricism, the sheer nature of statistical methods and econometric data puts me in a position where I risk reproducing this so-called

‘economistic fallacy’.

Development economics. The advent of development economics in the latter half of the 20^th century added to economics another problematic dimension; the notion that ‘developing countries’ (or, indeed, ‘emerging markets’) are an inferior, incomplete version of ‘developed countries’, most notably Western Europe and North America (The ‘West’). Indeed, today’s ‘developing countries’ will not likely be considered fully developed any time soon, despite the fact that many of them are richer, by conventional measures, than many Western countries were only recently.

The reason for this, according to postcolonial and postmodern theorists, is that the hierarchical binary of developed/underdeveloped represents a repackaging of colonialist thought (see for instance Saïd 2003/1979; Escobar 1995): It is part of an intellectual tradition, rooted in Western modernism, in which history is framed as a linear, universal process wherein the world progresses through different stages (such as in Rostow’s stages of growth) and Western ‘civilization’ acts as the locomotive that pulls the ‘backward’ Others out of a past in which they would otherwise be stuck.

Methodological nationalism. A closely related discursive framing is the conceptualization of social geography as a mosaic of nation-states that each contain their own society and economy (Agnew 1994). This conceptualization has formed an equally important ontological basis for much of modernist thinking about the world: In what is often labelled ‘methodological nationalism’, nation- states serve as the primary unit of analysis and point of reference (ibid.).

Recently, however, the assumption that any state reign sovereign over its bounded territory, and that its national borders simply function as a ‘line in the sand’, has increasingly been put into question (Agnew 1994; Scholte 2005; Johnson et al. 2011; Parker & Vaughan-Williams 2012). Indeed, in the past half century, revolutions in economics, logistics, transport and information technology have facilitated a restructuring of states in favor of transborder markets and movements, eroding

(14)

13 national territorial integrity and bringing about a new configuration of space which calls for new methodological and analytical tools (Scholte 2005). Then again, some parts of the world were probably never a suitable fit for the ‘Westphalian state’ template, especially African countries (the typical PRSP adopters) whose externally imposed national territories have always been a site of contestation, competition and cooperation between both state and non-state actors (Abrahamsen 2017).

Arguably, the combination of methodological nationalism and development thinking has encouraged researchers and policymakers to trace the source of poverty in a ‘developing’ country exclusively to its state, its residents and their purported backwardness and economic illiteracy, rather than to the colonial legacy, neo-colonial arrangements and inter- or transnational structures of power. In fact, the PRSP approach has been criticized for placing the task of development solely in the national realm, thereby undermining efforts to overcome development obstacles embedded in structures beyond the nation-state, such as the rules of global trade, finance and intellectual property (Tan 2011).

While my emphasis on the role and influence of the IFIs clearly illuminates some of the more supranational dimensions of economic policymaking, it will not be possible for me to completely venture beyond the lens of methodological nationalism simply because of the nature of the PRSP approach and the composition of the relevant data. I will, however, avoid using the term ‘developing countries’ and instead opt for the more appropriate ‘poor countries’ or ‘poorer countries’.

Poverty. While being a ‘poor country’ refers to total national (monetary) income, my outcome variables are concerned with poverty on the individual level. Here, such poverty is defined as the lack of access to the basic material inputs necessary to sustain a decent life. Of course, poverty is by no means a normatively and politically neutral and uncontested concept. It is important to note, therefore, that my definition of poverty is rooted in a materialist understanding of the phenomenon which fails to consider how the experience of poverty varies depending on different circumstances, such as the presence of economic inequalities.

Moreover, even if we are to agree that poverty should be defined as I define it here, we also need to be cognizant of the fact that there are many different ways to approximate, quantify and measure it (see for instance Kwadzo 2015). No doubt, as I shall return to in chapter 5, the SDG- indicators used in this study are not without their problems. The reason for why I nonetheless choose to draw on the SDGs is that, apart from enhancing the commensurability and relevance of my study,

(15)

14 it allows me to investigate the effect of PRSP implementation on several different dimensions of (material) poverty.

2.3. Washington Consensus

Although any claim on causation in this study remains probabilistic, critical realism contends that careful, rigorous application of scientific methods can help us get closer to observing the ‘real’ world;

to produce ‘true’ knowledge. In order to really explain or understand a phenomenon, however, we typically need to do more than just observe it; we also need to uncover the ‘generative mechanisms’

underneath it. This is where theory becomes useful.

As elaborated in the Introduction, I hypothesize that PRSPs were successful in reducing poverty based on the theoretical assumption that the strategy papers were less aligned with the

‘Washington Consensus’ in favor of more diverse, country-specific strategies influenced by other development paradigms. Below, I will be outlining and contextualizing the theoretical bedrock upon which the ‘Washington Consensus’ emerged—that is, neoliberalism—and the economic policies associated with it. Following this, I will offer a brief account of the criticism levelled against the

‘Washington Consensus’, and how this led to two new development paradigms, the ‘post-Washington Consensus’ and the ‘New York Consensus’, which will also be discussed. Again, it bears repeating that these paradigms are by no means dichotomous. Quite the opposite, there is significant overlap.

While neoliberalism has deep roots—indeed, it is rooted in classical liberalism—it emerged most forcefully as an economic doctrine in the mid-1970s. Up until this point, ever since the global depression, Keynesianism had reigned supreme (Peet & Hartwick 2015, 89-90). This had allowed for considerable state ownership of national economies and, especially, its key industrial sectors; the use of ‘deficit spending’ to pull countries out of economic downturns; a separation of social welfare services from private markets; progressive taxation on income and wealth; and regulation of economic flows through tariffs, quotas, price controls, capital controls, fixed exchange rates, and the lowering of interest rates to stimulate consumption and investment (Peet & Hartwick 2015, 66-76). Throughout its early decades, the IMF also echoed Keynesian ideas, promoting expansionary macroeconomic policies and a regulated market. In fact, the whole Bretton Woods system (see next chapter)

(16)

15 represented a form of institutionalized international Keynesianism⁹ (Stiglitz 2002/2015, 109-110; Peet 2009, 48-59).

However, in the 1970s, the peculiar combination of low growth, high unemployment and high inflation put Keynesian models into question and created an opening for new ideas; specifically, neoliberalism (Peet and Hartwick 2015, 89). Rooted in a deeply individualistic ontological tradition, the founders of this school of thought espoused a view of state interventionism and protectionism as not only inefficient but inhibiting of human freedom (ibid. 90-94). Instead, they advocated for a system in which the state would be relegated to little more than a ‘night-watchman’ role of maintaining the rule of law, as well as performing limited macroeconomic tasks such as setting interest rates: Growth, wealth and welfare were most efficiently achieved through unobstructed, voluntary competition between self-interested, utility-maximizing private buyers and sellers (ibid.).

Contrasted with Keynesianism, these neoliberal ideas had radical policy implications. Some of them had already been reflected in IMF-conditionalities in the 1970s, but it was the during the debt crisis in the 80s that neoliberalism was fully embraced by the IFIs (Peet 2009, 84-85, 136-145). Whereas the World Bank had focused on targeted poverty reduction and basic needs provision throughout the 1970s, the prevailing wisdom now was that if inflation could be kept under control and markets left alone, or at least left unobstructed, then wealth would ‘trickle down’ to the poor (ibid. 134-137; Banks

& Hulme, 2012, 5).

In 1989, economist John Williamson formulated what was, in his view, the ten principal policy recommendations of the neoliberal development paradigm—what he called the ‘Washington Consensus’¹⁰ (Williamson 1990): (1) Commit to fiscal discipline, especially through (2) reduced public spending, in order to avoid large budget deficits and macroeconomic instability; (3) enact tax reform by broadening the tax base and lower marginal tax rates, thereby increasing incentives for investors;

(4) let interest rates be set by the market, but ensure that they are positive when adjusted for inflation to attract and preserve capital; (5) allow for competitive exchange rates and (6) remove trade barriers to channel resources into more productive use and encourage export-oriented growth; (7) liberalize foreign direct investment (FDI) to allow foreign companies to compete on equal terms; (8) privatize state enterprises to maximize economic efficiency; (9) remove unnecessary regulation of the market to increase competition; and finally, (10) secure property rights to ensure an attractive investment environment.

9 It was, in part, a brainchild of John Maynard Keynes himself.

10 A reference to the home of the IFIs and the US Treasury.

(17)

16 Criticism against the ‘Washington Consensus’ has been extensive (and this account is far from exhaustive): For instance, it has been argued that neoliberalism builds on flawed and disingenuous ontological assumptions about human nature, the nature of markets and the science of economics (Peet & Hartwick 2015, 45-62, 112-117). Ha-Joon Chang, most notably, has made the case that the

‘Washington Consensus’ is devoid of historical analysis in that the development path it puts forward runs counter to the one embarked upon by the West and the ‘emerging markets’ in East Asia—a path which included considerable state intervention (Chang 2002, chapter 1 & 2). In fact, continuing protectionism in rich countries have been a major impediment to the export-oriented growth that neoliberalization is meant to facilitate in poor countries (Abrahamsson 2008, 16). To Marxists, such hypocrisy reveals how neoliberal theories merely function as an ideological camouflage for what is really an elitist political project aimed at solidifying the dominance of Western powers and the upper classes, reducing the cost of labor, increasing surplus accumulation and redirecting it to the top (see for instance Harvey 2005; Duménil & Lévy 2004; Varoufakis 2011; Prashad 2014).¹¹

A less radical but more hard-hitting renunciation of the ‘Washington Consensus’ has come from Joseph Stiglitz, who, in 1998, as chief economist at the World Bank, expressed the need to move towards a ‘post-Washington Consensus’ (Stiglitz 1998). A few year later he elaborated on his criticism, arguing that the uniform neoliberal policy packages imposed by the IFIs (which I will return to in the next chapter) often had profoundly negative effects, not least during the 1997 East Asia crisis¹² (Stiglitz 2002/2017, chapter 8). Stiglitz was especially critical of capital market liberalization and argued that it is a fallacy to assume that capital flows are counter-cyclical¹³ (ibid. 161-162, 324-325). In Stiglitz view, the IMF (in particular) put faith in overly simplistic, discredited economic models, leading to reforms (like privatization and trade liberalization) being implemented with no sense of pacing and sequencing and without the added inclusion of necessary regulation (like competition law) or safety nets (like unemployment insurance and worker retraining programs) (ibid., chapter 7).¹⁴

11 This was achieved, it is argued, by transforming and relocating production systems; liberalizing financial capital flows, commodifying public services and privatizing state assets; and undermining unions and the prospects of a New

International Economic Order (Prashad 2014, chapter 1).

12 The obsession with cutting deficits and increasing interest rates (despite high levels of indebtedness) had, in Stiglitz view, exacerbated the downturn and helped spread the crisis to other countries.

13 Money enters when an economy is growing and leaves when it is shrinking, not vice versa.

14 Stiglitz put the bulk of the blame on the IMF which, he argued, was ideologically blinded by a naïve ‘market fundamentalism’ and manifested a paternalistic mentality towards poorer countries reminiscent of colonialism (2002/2017, chapter 6 & 12). At the same time, Stiglitz found the approach to be intellectually incoherent as the IMF would sometimes push for certain government interventions (such as bailouts or exchange rates manipulations), seemingly to serve the interests of the ‘financial community’ (ibid.).

(18)

17

2.4. Post-Washington Consensus

Throughout the 1990s, the World Bank had started putting more emphasis on ‘market-friendly state intervention’, ‘good governance’ and ‘inclusive’ and ‘pro-poor’ growth (Peet 2009, 145-149).

Following Stiglitz’ criticism, there was a deepening division within the Bank over the merits of the

‘Washington Consensus’ (Mosley 2001), and by 2004, its president James Wolfensohn proclaimed that the consensus “has been dead for years” (as quoted in Sumner 2006, 1401). This message was brought home forcefully in a major 2005 World Bank report which rejected the ‘one-size-fits-all’ formula of previous programs and instead advocated for a more flexible, selective, moderate and experimental approach to reform (World Bank 2005).

Meanwhile, at the IMF, its managing director Anne O. Krueger took a different position: The problem with neoliberal programs was that they had been poorly implemented (Krueger 2004). She doubled down on the importance of fiscal discipline but added that, if development is to be successful, macroeconomic stability should coexist with “sound governance—at the national and corporate level;

effective and respected institutions; a well-established legal system; recognition of, and protection for, property rights; a well-functioning financial sector” (ibid).

It is this emphasis on governance and institutions that is said to embody the ‘post-Washington Consensus’ that Stiglitz called for (Rodrik 2006). While this development paradigm amounts to little more than an augmented version of—or at least a complementary addition to—the original consensus, it offers an ambitious policy agenda: According to the ‘post-Washington Consensus’, governments need to step in to control corruption; make the judicial system more effective; increase transparency and accountability within public institutions; give independence to the central bank; adhere to World Trade Organization (WTO) principles; improve regulatory framework around firms—specifically regarding capital requirements, bankruptcy and competition—while continuing to weaken labor market regulations (Elkins & Feeny 2014, 233). In addition, under the ‘post-Washington Consensus’, liberalization of capital accounts needs to be more ‘prudent’ (cautious and gradual), inflation rates need to be targeted at 3-5%, while exchange rates fluctuate within a limited margin. Lastly, the state needs to provide some social safety nets, alongside certain targeted poverty reduction efforts (ibid.).

(19)

18

2.5. New York Consensus

A no less ambitious but more social-welfarist and Keynesian approach to development has been offered under what economist Andrew Sumner labeled the ‘New York Consensus’¹⁵ (Sumner 2006, 1405-1406). This development paradigm is closely related to the MDGs—which were formulated by the UN secretariat but in collaboration with the IFIs (Kwon and Kim, 2014, p. 354)—and it draws heavily on the work of Jeffrey Sachs (2005) and the UN Millennium Project. Its point of departure is that the poorest countries are stuck in poverty ‘traps’ with low levels of innovation, FDI, domestic savings and tax revenue—among other ills—and to escape these traps, a ‘big push’ fueled by increased foreign aid is required (Sumner 2006, 1406; Rodrik 2006, 980-981).

Although the ‘New York Consensus’ aligns with the other consensuses in that good governance, open trade and economic growth are viewed as necessary preconditions for development, it also allows for greater state interventionism, including considerable increases in public spending on physical and social infrastructure; a commitment to develop rural areas more efficiently; launch programs to combat and prevent diseases; ensure representation and participation of women and girls; provide access to free, universal primary enrollment and to more decent work programs (Elkins and Feeny 2014, 235).

In accordance with the ‘New York Consensus’, the IMF has emphasized the need to increase social spending in the poorest countries, but, they add, such spending may also undermine “debt sustainability and private sector activity” (IEO 2007, 37). Researchers have found an increased presence of ‘social spending floors’ in IMF programs, but they also argue that the commitment to pro- poor policies remains limited, and on a whole, changes to these programs are little more than cosmetic (Kentikelenis et al. 2016). Others, however, have concluded that PRSPs have been increasingly aligned to the ‘New York Consensus’ and a ‘social protection agenda’ (Elkins & Feeny 2014; Elkins 2015). I will return to this in the Literature Review.

3. BACKGROUND

Before reviewing previous literature, I will chronicle the historical background of the IFIs; the conditions and developments leading up to the ‘Third World debt crisis’; the ensuing era of neoliberal

‘structural adjustment’; and finally, the birth of the Poverty Reduction Strategy initiative. It goes without saying that any exhaustive account on the subject is way beyond the scope of this study, and

15 A reference to the home of the UN headquarters.

(20)

19 necessarily, important nuances will be smoothened over. Moreover, when explaining how the debt crisis came to be, emphasis is put on exogenous factors rather than those found within affected countries¹⁶. Of course, there is no single culprit here, but as I will try to demonstrate, it is my conviction that these factors were the most important ones.

3.1. The IFIs and the Debt Crisis

The World Bank and the IMF were formalized in 1944 at the UN Monetary and Financial Conference in Bretton Woods, New Hampshire. Although both institutions are international in name, they have, by and large, been kept under the sway of their North American and Western European members—

especially the United States¹⁷ (Peet 2009, 69, 177; Financial Times 2018; Weisbrot & Johnston, 2019).

The World Bank (originally the International Bank for Reconstruction and Development) was set up to help finance the reconstruction of war-torn Europe, but it was quickly refashioned into a development bank that would issue concessional loans for infrastructure projects in the South (Peet 2009, 129-131).

The IMF, meanwhile, was tasked with regulating the exchange rates among its members and to offer short-term loans to those facing balance-of-payments problems (ibid., 66).

Following the 1971 decision to de-link the American dollar from the value of gold, the so- called ‘Bretton Woods system’ collapsed and the fixed exchange rate regime that IMF had presided over was replaced with a flexible one, sending shock waves through the world economy (Buckley 2008, 9-10). Shortly after that, in 1973, came a second shock: The oil-cartel OPEC (Organization of the Petroleum Exporting Countries) imposed an embargo against the US and other Western countries, thereby creating a spike in oil prices. The spike allowed the OPEC-members to reclaim some of the revenue lost through the depreciation of the US dollar, yet much of the oil rents ended up in Western commercial banks. Prompted by their sudden enrichment, these banks went on a lending spree in the South, especially in larger, middle-income countries in Latin America (Peet 2009, 82-87; Sachs 1989, 10-11). Meanwhile, many governments in lower-income countries turned elsewhere for cheap credit, most notably to the IFIs, which began to increase their lending after the (first) oil shock (Birdsall &

Williamson 2002, 15-17).

16 Of course, economies in the South suffered from a range of domestic ills, including militarization; corruption and (neo)patrimonialism; maldistribution of land and other productive resources; and overreliance on protectionism, dirigisme and planning—just to name a few (South Commission 1990, 37-55; Sachs 1989, 13-17)!

17 Despite some recent changes in favor of China, Russia, India and Brazil, the United States has ensured that it maintains a de facto veto power of key decisions (Financial Times 2018; Weisbrot & Johnston, 2019). In addition, the US still appoints the president of the World Bank and the deputy managing director of the IMF, while the European members appoints its managing director (Peet 2009, 69, 177). Both institutions are based in Washington.

(21)

20 Emerging out of colonial rule, few countries in the South had managed to escape their dependency on specialized primary sector exports: It was easier to reinvest foreign exchange earnings in a sector that was already developed, even as the market price for most primary commodities was low, increasingly unstable and in decline (Prashad 2008, 181-182). Western countries had offered little guidance or assistance out of this dependency. On the contrary, there had been a strong conviction in the economics departments that ‘developing’ countries should continue to specialize in the primary sector, at least before ‘taking off’ to the next development stage (Peet & Hartwick 2015, 141-146).

As the economic shocks and changes in the 1970s threw global commodity markets into violent fluctuations—with the spikes in oil prices drastically increasing import bills for non-oil producers—many countries in the South struggled with mounting deficits (Walton & Seddon, 13-15).

Calls for a New International Economic Order (NIEO) fell on deaf ears in the richer countries, and instead, Western governments went to great length disrupting international efforts aimed at improving the terms of trade of poorer countries, such as cartelization, while also raising protectionist walls around sectors of special importance to the South, such as agriculture and textile¹⁸ (Stiglitz & Charlton 2007, 41-45; Prashad 2014, 24-34; South Commission 1990, 60-61). On top of that, Western governments had begun pursuing contractionary macroeconomic policies to combat inflation, which had the dual effect of decreasing demand for goods produced in poor countries and increasing interest rates on their loans (South Commission 1990, 56).

There were of course many other reasons behind the mounting debt burdens in the South:

Even in times of relative stability, Southern governments had proven unwilling or unable to properly balance their budgets and keep inflation down, and the massive borrowing that began in the 1970s was in many cases always unstable as much of the funds did not go to productive use (South Commission 1990, 37-56). When commercial banks realized that their loans were not being repaid—

except with new loans—they simply stopped lending more (Peet 2009, 83). In addition to this lack of credit, some Latin American countries also experienced massive capital flight leading up to, and exacerbating, the crisis (Sachs 1989, 9-13).

Nevertheless, in their 1979 annual report, the IMF acknowledged that the deficits in non-oil exporting low-income countries were driven by falling primary commodity prices, on the one hand, and rising interest payments, on the other (IMF 1979, 23). Then, later that same year, the US spiked

18 In fact, these sectors had been conveniently left out of General Agreement on Tariffs and Trade (GATT)—a multilateral free trade agreement which grew out of the Bretton Woods conference and would serve as the precursor to the WTO (Stiglitz & Charlton 2007, 41-45).

(22)

21 the interest rates on the American dollar, plunging several countries into near-bankruptcy (Prashad 2014, 52-56).

3.2 Structural Adjustment

This is when the World Bank and the IMF really steps into the picture: In coordination with the US Treasury Department and major commercial banks, the IFIs began to offer concessional loans, debt restructuring and, eventually, partial debt cancellation to the crisis-ridden countries (Peet 2009, 86-91, 136-146). In exchange, these countries had to undergo (short-term) ‘stabilization’ and (long-term)

‘structural adjustment’, that is to say, governments had to implement a program of structural and macroeconomic reforms in accordance with the ‘Washington Consensus’ (ibid.). Implementation of these reform programs—labeled Structural Adjustment Programs (SAPs)—would also become a precondition to receive bilateral grants and loans (ibid., 115-116).

Throughout the 1980s and 90s, the roles played by each of the IFIs became increasingly intertwined, largely because the IMF increasingly moved beyond its traditional concern with macroeconomics and into areas which were more micro and structural in nature—where the World Bank focused its attention (Mosley 2001, 310; Stiglitz 110-112, 2002/2017). Gradually, the number of policy conditions that the IFIs imposed through their different programs also increased (Hermele 2005, 2). While these so called ‘conditionalities’ embodied the core principles of the ‘Washington Consensus’ laid out in the previous chapter, it is worth noting that the IFIs often went further, employing a ‘shock doctrine’ approach of rapid structural adjustment which included (but was not limited to) de-unionization, layoffs, wage cuts, social spending cuts, and the elimination of subsidies on basic goods (Peet and Hartwick 2015, 99-105; Sumner 2006, 16).

The consensus view is that SAPs did not deliver on their promises. It is of course difficult to determine how countries would have fared in the absence of the programs, especially considering that the external economic environment continued to deteriorate, at least throughout the 1980s (which is often labeled the ‘lost decade’). We also have to consider the diversity of the places under ‘adjustment’. For instance, it has been argued that SAPs made more sense in Latin America—for which the programs were primarily designed—than they did elsewhere (Stiglitz 2002/2017, 149). Others have argued that countries in Africa may have been worse off without the programs, or at least that they could have been better off if the policy prescriptions had been fully and correctly implemented (Sahn et al. 1997;

Scott 2010).

(23)

22 That being said, there is ample evidence that SAPs failed to spur sustained growth and reduce poverty (see for instance Johnson & Schaefer 1999, Easterly 2005; UNCTAD 2000; 2002). In fact, studies show that the programs often did enormous damage to the already poor and vulnerable; to local industry and agriculture; and to labor rights and human rights (Cornia et al. 1987; SAPRIN 2002;

Lloyd & Weissman 2010; Rodwan Abouharb et al. 2010). Even studies from within the IFIs concluded that many of their policies and projects were unsuccessful (World Bank 1992; OED 1992; Khan 1990).

3.3. Poverty Reduction Strategy Papers

Structural adjustment—especially the removal of subsidies on basic goods—ignited massive protests across the Global South and, beginning in the late 1980s, even in parts of the North (Peet 2009, 99- 104; Walton & Seddon 1994). There was a growing understanding that neoliberal prescriptions were failing—perhaps by design—and that the debt burdens in poor countries were not just unfair, but unpayable (ibid., 104-107). In response, the IFIs launched the Heavily Indebted Poor Countries (HIPC) initiative in 1996. The purpose of the initiative was to start writing off some of the debt owed by the poorest countries, but predictably, there were strings attached (ibid. 107-108):

After the HIPC initiative had been enhanced in 1999 to cover a larger number of countries, those that could and wanted to access debt relief were expected to prepare and implement a so-called Poverty Reduction Strategy Paper (PRSP) (Elkins et al. 2018, 378). A preliminary interim-PRSP was to be followed by a full-PRSP, and throughout the implementation period—which typically lasted three to four years—the recipient country had to follow up with Annual Progress Reports (APR) (Elkins

& Feeny 2014, 230).

Apart from an analysis of the poverty situation in the recipient country, PRSPs needed to include the macroeconomic, structural and social policies required to reduce poverty and stimulate sustained economic growth (Wachira & Ruger 2011, 1957). The strategies were supposed to focus on results and outcomes that benefitted the poor, including in the long-term, and recognize that poverty is a multidimensional phenomenon (Guimarães & Avendaño 2010, 320).

Since lack of local commitment was identified as a major obstacle to successful implementation of IFI-supported programs, the PRSPs were to be formulated by the recipient country (Booth 2005, 1). The national government should take the lead, but the PRSP was also expected to be developed in collaboration with a range of domestic stakeholders—including opposition parties, trade unions, businesses, NGOs and religious institutions—as well as international donors (Fraser 2005, 317; Dijkstra & Komives 2009, 3). As articulated in the IFIs’ Comprehensive Development Framework,

(24)

23 country ownership, broad-based participation and partnerships were touted as core principles of the PRSP approach (Peet 2009, 145-147).

Shortly after the PRSP approach had been launched, even countries that were not eligible for HIPC debt relief were urged to prepare PRSPs in order to access certain concessional lending arrangement from the World Bank’s International Development Association and the IMF’s Enhanced Structural Adjustment Facility (Peet 2009, 147)¹⁹. Following the 2002 International Conference on Financing for Development, the PRSP approach also became an important framework through which other donors organized their aid to poor countries (Elkins & Feeny 2014, 230).

Apart from access to debt relief and concessional loans, then, PRSP implementation brought with it the prospect of increased bilateral aid, including as budget support (Elkins et al. 2018, 382). In addition, PRSPs provided a “unique opportunity for poor countries to pursue an integrated, national development strategy, something that had been missing since the late 1970s and early 1980s when these countries embarked on market-oriented reforms” (Gottschalk 2005, 440). The strategies also became an important vehicle to achieve the MDGs (Elkins et al. 2018, 378-379).

While the IFIs would continue to wield influence over the PRSPs, it was declared at the 2005 G8 summit that ‘explicit endorsement’ from their executive boards were no longer required to receive loans for a strategy paper (Sumner 2006, 1401). In many ways, the Poverty Reduction Strategy initiative allowed the IFIs to correct some of their past mistakes and to signal a shift in priorities, as epitomized by the rediscovery of poverty. The World Bank even adopted the slogan “Working for a World Free of Poverty”. Today, however, the fate of the PRSPs seems somewhat uncertain: The IFIs are still committed to poverty reduction, and there has been no formal end to the initiative, yet no new PRSPs appear to have been issued since 2014 (IMF 2019).

4. LITERATURE REVIEW

Did the promises of the initiative materialize? The literature on PRSPs has dealt with its different aspects—policy content, performance and process of implementation—in different ways, most frequently in the form of country case studies. For instance, in 2003, Development Policy Review ran a special issue on PRSPs consisting of qualitative case studies from Benin, Kenya, Malawi, Mali, Mozambique, Rwanda and Tanzania. Contributors found that the PRSP process were leading to some

19 The latter has been renamed into Poverty Reduction and Growth Facility (ibid. 110). Recipients of PRSP loans primarily included countries in the South, but also a few former Soviet states.

(25)

24 gains in areas such as budgetary efficiency, aid coordination, government accountability and civil society participation (Booth 2003a). However, political turbulence, bad governance, poor state capacity, and limited and partial national ownership of the strategies remained serious constraints to successful implementation (ibid.).

Studies such as these offer an invaluable contribution: Not only do they allow for more detailed examinations of PRSP policies, but they also contextualize the PRSP experience, revealing the complexities and specificities that marked each implementation. To be sure, a successful implementation of even the most well-crafted PRSP hinges on different factors that may be more or less prevalent in each country and region. The problem, of course, is when the examination of a PRSP or its implementation—and the theoretical arguments explored through and developed from that examination—is generalized to all other cases.

As such, the following section will begin with a discussion on the PRSP approach based primarily on country case studies and other qualitative studies, and then proceed to a review of the few studies which have analyzed PRSP content using a larger quantity of cases. This will allow for a smooth transition to the next section, in which previous quantitative studies on PRSP performance will be outlined.

4.1. Process and Content

The questions of ownership and participation have attracted particular scholarly interest: Can PRSPs really be said to have been homegrown and ‘owned’ by the implementing countries, and if so, did this ownership go beyond the government’s ministries of finance or planning to also include key stakeholders in civil society? While there is certainly no consensus among scholars, many agree that the PRSP approach represented a step in the right the direction, but that it did not go far enough (see for instance Booth 2003b; Driscoll & Evans 2005;Gottschalk 2005; Cheru 2006; Mouelhi & Rückert 2007): IFIs continued to exert ‘backstage’ influence over the PRSPs, with ‘participation’ often looking more like consultation, and key stakeholders such as trade unions were sometimes excluded altogether from the process (Oxfam 2004, 1; Stewart and Wang 2005, 456-457).

Moreover, while the formulation of the strategies was in some cases characterized by a high

‘degree of national ownership’, there could be quite a bit of distance between what was strategized on paper and the actual implementation (Booth 2003b, 157; Dijkstra 2005; Dijkstra & Komives 2009, 18). In those cases where ownership was indeed strong, the fact that a range of different actors were expected to participate in the formulation may have had the effect of weakening the legitimacy of

(26)

25 sitting governments (Stewart & Wang 2003, 27). This was especially problematic when the government, unlike civil society actors, had a legitimate democratic mandate to act upon.

The IMF has acknowledged that the emphasis on ownership and participation in the PRSP process would “make it easier to generate domestic political support for the program, since it is likely to be seen, at least in part, as an indigenous product, rather than a foreign imposition” (IMF, 2001, 14). Several scholars, however, have taken issue with the participatory approach (see Craig and Porter 2003; Brown 2004; Fraser 2005; Rückert 2009; Dutta & Rastogi 2016): They argue that, in practice, this approach merely serves as an instrument to disseminate market-based knowledge about poverty reduction and development—presented as universal, ‘technical’ knowledge—thereby rationalizing the imposition of policies and performance goals defined elsewhere (Rückert 2009; Dutta & Rastogi 2016).

Accordingly, IFI-led participation is understood as a technology of social and political control that seeks to deepen the hegemony of neoliberalism: “Through participatory exercises, the World Bank reaches deep into the cultural fabric of developing country societies, with the ultimate goal of producing complicit neoliberal subjects” (Rückert 2009, 68).

No doubt, the questions of ownership and participation are closely linked to policy content. To some, the PRSP approach has been little more than a public relations effort to allow the IFIs to escape criticism and re-legitimize themselves and their controversial reform agenda in the court of public opinion (Guimarães & Avendaño 2010; Saeed et al. 2015). But while the apparent survival of

‘Washington Consensus’ has been interpreted as evidence of the lack of national ownership (Stewart and Wang 2003), others have argued that certain governments in the South have come to internalize the worldview of donors, turning them into ‘post-conditionality regimes’ (Harrison 2001). In addition, it is wrong to assume that people in the South cannot independently find an appeal in neoliberal economics or associated policies²⁰.

It has also been pointed out that, even if commitment to a radically different development agenda existed, the PRSP approach was by its very nature unable to address (and may have even undermined efforts against) extra-national obstacles to poverty reduction (Tan, 2011). For instance, neoliberalization under SAPs made countries more vulnerable to the whims of global markets and the power of transnational corporations. This have, in turn, almost certainly affected the content of the PRSPs.

20 Indeed, for many governments, IFI-programs have served as a convenient excuse to implement uncomfortable but desired reforms.

(27)

26 A few studies have sought to analyze the content of a larger group of PRSPs. Early on, Frances Stewart and Michael Wang (2003; 2005) examined the content of 30 first generation PRSPs and found that not only were there a remarkable uniformity among the strategy papers, but no meaningful break with the ‘Washington Consensus’ had occurred: Although many strategies emphasized good governance—

including measures to combat corruption and increase public transparency and accountability—most PRSPs also endorsed continued privatization of state-owned enterprises and liberalization of trade, capital accounts, the banking sector, exchange rates and interest rates; as well as contractionary monetary and fiscal policies to keep inflation and deficits under control (ibid., 462-467).

Moreover, argued Stewart and Wang, the macroeconomic section in the PRSPs often seemed detached from or even contradicting the poverty analysis section (2005, 467). This is in line with a 2005 report by the UN Millennium Project which criticize the PRSPs for letting pre-established macroeconomic frameworks and budgetary ceilings dictate—and thereby limit—MDG progress, rather than letting such policies be dictated by the requirements necessary to meet the MDGs (UN Millennium Project 2005, 58-61).²¹

In another cross-country content analysis of 15 PRSPs focusing on the macroeconomic section of the strategies, Ricardo Gottschalk (2005) noted a “strong agreement among the PRSPs that broad-based growth should be at the centre of a development strategy” (emphasis added, 440). However, he also found that the strategies in question did not “really support economic growth and poverty reduction in a direct, clear way” (ibid., 440): The emphasis was still on achieving macroeconomic stability, with fiscal and monetary policies lacking the flexibility required to deal with exogenous shocks (such as a sudden drop in demand for a country’s exports’) and leaving little room for governments to spend their way out of downturns.

The understanding that PRSPs generally championed strict fiscal and monetary policies was confirmed in a content analysis of 50 PRSPs conducted by Andrew Sumner (2006). Privatization was another policy pushed for in the vast majority of PRSPs under study, while liberalization of trade, agriculture, capital accounts and FDI were somewhat less common features in the strategies, especially in African HIPCs (Sumner 2006, 1406-1411). Sumner concluded that PRSPs had made some strides away from the ‘Washington Consensus’, yet very few of the PRSP-policies could be labelled

‘unorthodox’ (ibid. 1408).

21 The World Bank and other donors, it was argued, should facilitate such re-prioritization by increasing development assistance.

(28)

27 In another study, Sakiko Fukuda-Parr (2010) examined 22 PRSPs based on alignment to the MDGs.

She found that commitment to meeting the goals was selective and that the majority of PRSPs lacked strategies for broad-based growth and pro-poor investment in social infrastructure. Instead, the PRSPs reflected “an assumption that ‘trickledown’ would achieve the poverty reduction objectives of the MDG agenda” (Fukuda-Parr 2010, 33).

The conclusions above have been somewhat disputed by Meg Elkins and Simon Feeny (2014).

In their analysis—the most comprehensive to date—81 PRSPs were scored based on its overall alignment to the ‘Washington Consensus’, the ‘post-Washington Consensus’ and the ‘New York Consensus’. Their results suggest that PRSP policies were in fact marked by some diversity; that PRSPs in poorer countries were generally less aligned with the ‘Washington Consensus’; that a shift away from neoliberal orthodoxy had occurred, especially after 2004; and that, instead, the ‘New York Consensus’ was driving content. Elkins and Feeny interpret their results as clear evidence that the MDG commitment “has been successful at shifting policies in the PRSP process towards the achievement of the MDGs” (2014, 243). In a separate but similar analysis of 87 PRSPs, Elkins also found evidence that a ‘social protection agenda’ including policies and programmers aimed directly at combatting poverty and vulnerability was outspoken in many strategy papers (Elkins 2014). This was particularly true in richer and more ethnically homogenous countries (ibid.).

In conclusion, then, it appears as if PRSPs did indeed move beyond the neoliberal one-size- fits-all formula of the SAPs, especially in the second and third generation strategies. However, we need to keep in mind that many PRSPs may have been less aligned with the ‘Washington Consensus’ simply because the neoliberal reforms that would otherwise indicate such alignment had already been carried out in the country. In addition, it appears as if many strategies remained dedicated to strict macroeconomic stability, which may of course have undermined commitments to increased social spending.

4.2. Performance

When it comes cross-country studies of PRSP performance, the literature is surprisingly sparse. This is particularly so if we consider the prominent role that the PRSP approach has had in shaping the international development discourse and, more importantly, shaping national development within poor countries. Perhaps it is the understanding that PRSPs were less uniform than previous programs that have led scholars to refrain from conducting quantitative appraisals of their impact: If all PRSPs

Have Poverty Reduction Strategy Papers actually reduced poverty in poor countries?