1Multilateral development institutions—especially the International Monetary Fund (IMF) and the World Bank—are frequently seen in Latin America as agencies that have imposed their ‘neo-liberal’ or ‘pro-market’ views on countries in the region, working in the interests of Northern countries and multinationals and contributing to persistent underdevelopment, poverty and inequality. On the other hand, some Latin American governments and technocrats have often seen MDIs as useful allies in their pursuit of modernising reforms and sensible policies and in their efforts to combat powerful, rent seeking, domestic lobbies. As with most stereotyped narratives, the true story lies somewhere between these extreme positions. Further, it has considerably evolved over time.
2As we show in Section 2, MDIs’ influence in Latin America has varied significantly over time and across countries, depending on the institutions’ access to international and domestic private capital markets. Their influence has also depended on the economic importance and political sway of individual countries.
3In Section 3 of this chapter, we show how MDIs’ views on macroeconomic and development policy have changed over time, influenced mostly by evolving views in the international academic community, but also on occasion by Latin American policy makers’ own views and innovations. MDIs have thus recognised some of their past mistakes and consequently changed their paradigms and policies. In addition, Section 3 reveals that there has often been significant divergence among MDIs’ views and policies with regard to important development issues; a situation that has permitted individual countries to shop around for financial support, avoiding unwanted conditionality.
4Finally, Section 4 of this chapter illustrates how there has been —with a few exceptions—a significant convergence of views on macro-financial policies in the region, over time and across governments with widely different political inclinations, while major differences remain around microeconomic policy issues. We also discuss cases (in both Section 3 and 4) of Latin American governments from each end of the political spectrum that applied policies that were clearly at odds with views currently held in the MDI community. Some of these policies proved to be successful and were later on embraced by MDIs’ official positions, while others were clearly not successful. Section 5 concludes.
5The degree of influence of MDIs on Latin American countries’ policies (as has been the case in other developing regions) has varied substantially across countries and over time, depending on the scarcity value of foreign exchange resources for individual beneficiary countries at any point in time. Governments of countries with good access to international capital markets, and/or deep domestic financial markets, have not needed to accept MDIs’ advice in order to finance their developmental needs. At the other extreme, governments facing currency and fiscal crises, and/or without access to either international or domestic private capital markets, whether as a consequence of bad policies or bad luck (e.g. sharp drops in the terms of trade or sudden stops of capital flows due to financial contagion), have desperately needed MDIs’ financial resources in order to avoid sharp recessionary adjustments, and thus have had scarce bargaining power vis-à-vis unwanted conditionality.
6Thus, both long-term trends and cyclical variations in international capital markets, and in domestic financial systems, have been important determinants of the degree of influence of MDIs on Latin American countries. Such trends and cyclical variations have also determined both MDIs’ and recipient countries’ attitudes with respect to conditionality.
Figure 10.1 Gross flows from MDIs to LAC7* as percentage of GDP and total capital inflows (1980-2015).
Sources: World Development Indicators, CAF, IDB, IMF Debt Database.
Note: *LAC7 refers to the seven largest Latin American economies, namely, Argentina, Brazil, Chile, Colombia, Mexico, Peru and Venezuela.
- 1 The term MDBs includes the World Bank and Regional Banks.
7Figure 10.1 shows the sharp increase in the relative weight of IMF and multilateral development banks (MDBs)1 gross disbursements for the seven largest Latin American economies during the 1980s, an increase that occurred as a consequence of the interruption and reversal of capital inflows that took place after the US Federal Reserve interest rate hike of 1982 (Végh et al., 1994). This episode led to a series of public debt defaults and many countries in the region remained without access, or with only precarious access, to international financial markets during most of the decade. Regional growth rates were very low during this so-called lost decade. This period coincides with the height of MDI influence in the region (on trade and foreign direct investment (FDI) opening, privatisations, fiscal and monetary adjustments, etc.) through generalised IMF programmes and World Bank structural adjustment programmes.
- 2 Remaining Latin American debt to international banks was exchanged for long-term sovereign bonds ba (...)
8Fiscal adjustment and the restructuring of public debt through Brady bonds2 during the second half of the 1980s, coupled with the rapid growth of international capital markets in the 1990s (Figures 10.2 and 10.3), gave rise to a vibrant market for international placements of Latin American bonds. Consequently, MDI flows’ participation in total gross capital inflows to Latin America decreased sharply, from a peak of 55 per cent during the period 1990‒91 to less than 10 per cent during 2011-13, with temporary increases during regional or global crisis episodes, such as in 1995, 1999/2003 and 2009.
Figure 10.2 Gross Capital flows (global and to LAC7)—1970-2015.
Sources: World Bank, BIS, ECLAC, Broner et al. (2013).
Figure 10.3 Latin America and Caribbean (LAC) Gross domestic product (GDP) and Capital Flows Annual Growth (1970-2015).
Sources: World Development Indicators, CAF, IDB, IMF Debt Database.
9In particular, the number of active IMF programmes in the region peaked during the 1980s at 22, and diminished from the early 1990s on, falling to eight in 2013-14 (Figure 10.4). The IMF disbursements were dominated by the large programmes in Mexico in 1995, following the ‘Tequila’ debt crisis of 1994, and in Brazil and Argentina from 1998 to 2003, when these two countries faced currency crises and sharp capital flow reversals following the 1998 Russian crisis. Similarly, MDB´s gross disbursements to the seven largest Latin American countries’ economies (LAC7), as percentage of total capital inflows, peaked in 1992 (at 45 per cent) and then dropped sharply to around 12 per cent since 1996, with a rebound to 20 per cent during the 2009 crisis, and a temporary rebound during the global crisis of 2009 (Figure 10.1). These disbursements had local peaks around periods of crisis in the largest countries (Mexico and Argentina, 1995; Brazil and Argentina, 1998-2001) and the global crisis of 2009 (Brazil, Mexico, Argentina and Colombia) (Figure 10.5).
10The IMF gross flows have been clearly countercyclical with respect to either private capital inflows or GDP growth3. This is consistent with the Fund’s objective of providing support to countries experiencing balance of payments crises. MDBs’ flows, on the contrary, do not show a consistent pattern in this regard. They should also behave counter-cyclically (Perry, 2011), or at least a-cyclically, given MDBs’ long-term development objectives. But they have often been procyclical because MDBs frequently behave more as private banks than as development institutions (see, for example, Perry, 2009).
Figure 10.4 IMF disbursements and number of countries with active programmes.
Sources: World Development Indicators, IMF.
Figure 10.5 Total MDBs disbursements by country.
Sources: World Development Indicators, CAF, IDB, IMF Debt Database.
11In parallel to these developments, several Latin American countries have been developing deep long-term markets for government and corporate bonds, in their respective domestic currencies. This has reduced the dependence of government finance on external sources, including MDBs (Figure 10.6), with the additional advantage of reducing exposures to currency risk for governments and corporates in non-tradable activities.
Figure 10.6 External public and publicly guaranteed debt stocks (percentage of GDP).
Sources: World Development Indicators, IMF. External debt is defined as all debt owed to non-residents in foreign currency.
12It is thus not surprising that the height of MDI influence and conditionality occurred during the Latin American debt crisis of the 1980s, when most countries in the region requested IMF rescue packages and accepted the highly criticised ‘structural adjustment loans’ of the World Bank (this period is discussed further in Sections 3 and 4 of this chapter). In contrast, influence and conditionality were at an all-time low during the 2003-08 and 2010-14 periods, during which most Latin American countries were blessed with high and increasing terms of trade (due to high and increasing real commodity prices)—especially in South America—and an explosion of private capital inflows, as a consequence of both high international liquidity and good regional growth prospects. Thus, very few countries in the region needed access to IMF funds in the 2003-2014 period and many of them sharply reduced their borrowing from MDBs.
13Also not surprisingly, since the late 1990s, but especially in the early years of the new millennium, the IMF entered a period of ‘soul-searching’, trying to redefine its role and redesign its financial instruments (‘flexibilising’ access to its credit lines, e.g. the Contingent Credit Line) in order to maintain relevance in fast-growing developing countries. During this period, MDBs, and to a lesser extent the IMF, began to talk about ‘partnerships’ and ‘country ownership’ of countries’ own development strategies and adjustment programmes. Consequently, the term ‘conditionality’ faded away significantly from official documents.
14Further, worried about their business prospects, MDBs were busily designing ‘MIC strategies’, intended to lure middle income countries that had achieved easier access to cheaper and longer term private financial resources into maintaining some lending and technical assistance relationships with MDBs (Perry, 2011). For the same reason, most MDBs have ever since been actively increasing their lending to the private sector and to public enterprises or subnational governments without sovereign guarantees (Perry, 2009) (Figure 10.7).
Figure 10.7 MDI Lending to Private Sector.
Source: Perry, 2011.
- 4 There are, of course, exceptions that may eventually require major IMF and MDB financing.
15The 2009 global crisis, with its sharp reduction in private capital flows, constituted a temporary reversal of these trends and was a stark reminder to middle income countries that they may still, on occasion, need to access IMF and MDBs resources, and thus should keep reasonable working relations with these organisations. Something similar, though less marked, has happened following the sharp reduction in commodity prices that has taken place since 2013. However, most Latin American countries have avoided major currency, fiscal and financial crises since 2009 (exceptions include the deep currency and fiscal crisis in Venezuela and, to a lesser degree, the currency crisis in Argentina) as macro-financial vulnerabilities have been significantly reduced in comparison to the past (Perry and Forero, 2014).4 There is a discussion of the post-2013 period in Section 4 of this chapter.
16As a consequence, it is unlikely, at present, that the relative importance of MDI lending and influence return to anywhere near their level of the 1980s.
17Size matters—a complement to this story is that MDIs’ influence has always been greater in countries with comparatively underdeveloped domestic financial markets and scant access to international financial markets, especially if, at the same time, these are small countries with little political sway. It is well known within the international community that large countries such as China, or—in Latin America—Brazil and Mexico, and previously Argentina, have usually received VIP treatment (conditionality has been lower and countries’ views have been more respected) compared to small countries such as Bolivia, Ecuador or Nicaragua. Medium-sized countries with good access to private markets, such as Chile and Colombia, have also been the darlings of MDIs. To be fair, it is also true that countries such as Brazil, Chile, Colombia and Mexico (and Argentina in the 1990s) have had stronger technical teams, more able to design good programmes and persuade MDB technocracies of the relevance of their own views, than other countries in the region.
18In summary, the degree of MDI influence has been a function of both the relative value of MDI resources for individual countries at any point in time (being greater in periods of private capital flow retrenchment and macroeconomic crisis) and the relative lack of technical capacities and political sway of beneficiary countries.
- 5 A view influenced by growth and development theories in vogue in the 1950s and 1960s (Harrod–Domar, (...)
19Easterly (2001) shows eloquently how the World Bank’s views on development have significantly evolved over time. Initially, in the 1950s and 1960s, the World Bank seemed to believe that building adequate public infrastructure was the key to development and, consequently, had a lending portfolio highly concentrated in infrastructure investment projects.5 It should be remembered that the World Bank was created to help finance the reconstruction of European Infrastructure after the Second World War. Increasingly, building human capital was seen as a complementary development approach, thus broadening the investment portfolio to the social sectors. Further, poverty reduction became a complementary objective to growth since Mc Namara´s presidency.
20However, by the early 1980s, the World Bank had adopted the view that what mattered most for development (and for ‘aid effectiveness’) were so-called pro-market policies (the opening up of trade, prudent macro-financial policies, a reduction of the size of the state, and the promotion of private investment), thus broadening again the portfolio with wholesale policy-based lending through Structural and Sectoral Adjustment Programs. This was the period (the 1980s and early 1990s) associated with the so-called ‘Washington Consensus’. It coincided with the height of MDI influence in Latin America due to the region’s currency and financial crisis, as discussed in the previous section.
- 6 Most so-called development economics, from Harrod–Domar growth models to Rosenstein-Rodan’s ‘big pu (...)
21The changing views held by MDBs were often influenced by academic research on development, as well as by Latin American policy makers’ opinions. Thus, during the 1950s and 1960s World Bank views were heavily influenced, among other sources,6 by Albert Hirschman’s views based on his Latin American experience (Hirschman, 1995). On the other hand, the contents of the Washington Consensus, and its very name, were drawn by John Williamson from the proceedings of a Washington-based conference attended by several high-level Latin American policy makers (Williamson, 2004). Indeed, a new generation of highly influential Latin American technocrats and politicians had become convinced, following the debt crisis of the 1980s, of the need to open up their economies, reducing the scope and intensity of government intervention and adopting prudent macro-financial policies. Thus, they were autonomously pushing for them in their own countries.
22As a consequence, the conventional view that the reforms that took place in many Latin American countries during the 1980s and early 1990s were simply imposed by the IMF and the World Bank is exaggerated in many cases—there was, rather, a concurrence of a significant need of resources and shared views—or plainly incorrect in others.
- 7 The Chilean fiscal rule requires the government to keep a ‘structural’ balance in surplus, thus mak (...)
23Thus, for example, there was very little IMF or World Bank influence in the pioneering Chilean reforms of the mid-1970s (which were associated, rather, with the local so-called ´Chicago boys´ in charge of economic policy making during the authoritarian regime of Pinochet) or in their deepening and development during the centre-left ‘Concertación’ governments of the 1990s. On the contrary, Chile often adopted reforms that only later on were promoted by MDIs in other countries and latitudes, such as the pension reforms of the 1980s (based on individual savings and capitalisation funds, mostly privately administered) or the fiscal rule7 of the late 1990s.
24And though there was MDI influence, there were also significant domestic initiatives in the reforms enacted in the late 1980s or early 1990s by the governments of Salinas in Mexico, Menem in Argentina (the ‘convertibility’ regime was not only home grown, it was not enthusiastically regarded by certain of the MDIs—see below), Fujimori in Peru, Gaviria in Colombia and Cardoso in Brazil (maintained and deepened by the Lula government), among others.
25Colombia, for example, had easy access to private international capital markets and did not have an IMF programme when Washington Consensus-type reforms were enacted. Moreover, in the previous 25 years it had only had a monitoring IMF programme (without IMF financial resources), and only required access to IMF funds during its 1999 crisis, almost a decade after enacting Washington Consensus-type reforms (Edwards and Steiner, 2008; Junguito, 2015).
26Later on, since the mid-1990s, the World Bank (and the Inter-American Development Bank: IDB) evolved towards a more holistic view of development, in which other factors acquired a role at least as important as the core Washington Consensus tenets. In particular, a series of World Bank analytical reports began to emphasise, from the mid-1990s on, the critical importance of institutions for growth and development (see, for example, Burki and Perry, 1998). This led to a new portfolio of loans related to institutional reform in a wide variety of fields and the development of a set of measurements of institutional quality (the Kaufmann-Kraay Worldwide Governance Indicators, see Kaufmann et al., 2010), which have become benchmarks.
- 8 Institutionalist views on development had been popular among many schools (including the Austrian S (...)
27The re-emergence of this key concern, virtually absent during the Washington Consensus period, occurred due both to the growing influence of the ‘new institutional economics’ of North (1990), Coase (1960) and others,8 and to the stark evidence of the initial failure of market-oriented reforms (supported by MDIs after the fall of the Berlin wall) in Russia and other ex-Soviet Union states and areas of influence, where there was a significant lack of market-supportive institutions (Stiglitz, 2002).
28Similarly, MDIs’ enthusiasm for ‘structural adjustment’ through Big Bang-style reforms became more subdued when it became evident that such reforms might as easily be undone in countries in which they were adopted through emergency powers, such as in Argentina during the Menem period, without wide domestic political support. Indeed, most of the 1990 reforms in Argentina were overturned in the early 2000s. In contrast, MDIs began to appreciate formerly despised ‘slow-reformers’, such as Costa Rica and Uruguay, where both the slow pace and the resilience of reforms, and their ultimate success, were intimately connected to the deeper democratic and political processes involved.
29Correspondingly, MDBs undertook a significant re-evaluation of their views on the role of the state. Thus, at the World Bank, the intellectual debate about too large vs. small states, that had characterised the era of the Washington Consensus, shifted towards a discussion about strong and competent vs. fragile and failed states (World Bank, 1995).
30Similarly, public investment in infrastructure, which had fallen into benign neglect during the Washington Consensus years, alongside a naive faith in private participation, was re-evaluated. A more balanced view with regard to the importance of these matters, and to the complexity of public‒private partnerships, emerged thanks to pressure from the academic community and Latin American governments. Latin American countries access to a growing self-managed regional development bank, the Development Bank of Latin America (Corporacion Andina de Formento: CAF), that engaged heavily in substituting World Bank and IDB lending to public infrastructure, played a major role in this revival of interest, as discussed below.
- 9 ‘A World without poverty’ became the Bank’s official vision, embodied in its mission statement. Pov (...)
- 10 As importantly, the concept of equality of opportunity as a basis for higher and more equitable gro (...)
31Similarly, during Wolfhenson’s mandate, poverty reduction became the main explicit objective of World Bank support9 and the share of anti-poverty programmes increased sharply within the Bank’s portfolio. The World Bank has since been associated with most conditional cash transfer programmes (as was the case in Brazil, Colombia and Peru), though it had had nothing to do with the enactment of the most famous and pioneering of these programmes (‘Progresa’, afterwards renamed ‘Oportunidades’, in Mexico).10
32Also, since the 1990s, the World Bank Group has been a pioneer of environmental sustainability and climate change concerns. So-called environmental and social ‘safeguards’ (which the IDB and the CAF have also implemented) have affected the approval, design and operation of public and private projects in infrastructure and other areas, contributing to heightened standards in the region. In addition, there has been a substantial body of research into these areas at the Bank and an important portfolio, financed by the Global Environmental Fund.
- 11 See, for example the reports published by IDB (2015) and the World Bank (De Ferranti et al., 2003).
33A more recent development relates to innovation and industrial policies. From the 1950s to the 1970s, the World Bank supported many public industrial projects, as well as private manufacturing, through lines of credit to central banks and local development banks. Such operations were, however, suspended in the 1980s and 1990s as part of the Washington Consensus era, which favoured free trade and the reduction of government intervention in industry. However, critiques from academia (Hausmann and Rodrik, 2003) and the growing evidence and belief in Latin America and other regions that opening to trade is, by itself, not enough to promote export-led growth, as suggested by early proponents of such a strategy, led to an intellectual revival of ‘industrial policies’ (more precisely, sectoral policies), oriented towards providing specific sectoral public goods and overcoming coordination failures. These views are beginning to make inroads within the development banks,11 though they are not yet an important part of these banks’ ‘official’ doctrines or lending portfolios.
34Simultaneously, there has been some reconsideration of the role of national development banks, which, ironically, were seen as superfluous or inconvenient by the multilateral development banks during the Washington Consensus era. Successful initiatives, including the Chilean Economic Development Agency (Corporación de Fomento de la Producción de Chile, CORFO), the Nacional Financiera (NAFIN) in Mexico, the Brazilian Development Bank (Banco Nacional de Desenvolvimento Econômico e Social: BNDES), the Bank of Foreign Trade (Banco de Comercio Exterior: BANCOLDEX) in Colombia, and the Development Finance Corporation S.A. (Corporación Financiera de Desarrollo: COFIDES) in Peru, have exerted a significant influence over this revision of MDBs’ doctrine. Further, the International Finance Corporation (IFC) and CAF have even become minority shareholders in COFIDES, in Peru, and in the newly created National Finance Development institution (Financiera de Desarrollo Nacional, FDN) in Colombia.
35As a final example, there is an emerging discussion within MDBs about the need for, or convenience of, approaching basic service delivery (education, health...) from the perspective of citizens’ rights, and not just from the standpoint of traditional ‘social policy’. This initiative is still in its infancy within the MDB community, but it is probable that, as is the case for other aspects just mentioned, it may eventually become part of the official doctrine.
36IMF ‘orthodoxy’ has also undergone an important evolution over time, on diverse topics central to the Fund’s core business. These changes have usually been influenced by academic research, but also by policy debates with MDBs and with Latin American (and other developing countries’) authorities, as discussed below.
37The IMF was created to support the Bretton Woods system of fixed exchange rates and capital controls by providing financial support during periods of balance of payments crisis. However, Latin American and other developing countries’ attempts to maintain fixed parities resulted in frequent currency crises, followed by massive devaluations towards a new fixed parity, and, as a consequence, in very high output and inflation volatility. The effects of adverse external (or domestic) shocks were exacerbated by procyclical monetary policies that attempted to defend fixed parities and contain the erosion of reserves, until it proved necessary to enact or permit massive devaluations when it became obvious that the former parity could not be maintained. IMF advice and programmes were important factors behind these highly inconvenient ‘stop-and-go’ policy cycles, which exacerbated volatility in countries subject to frequent terms-of-trade shocks and capital flows reversals. There were major confrontations between the IMF and some countries with regard to these policies, the most notable being when Colombia adopted a crawling exchange rate system and generalised capital controls in 1966 (Espinosa, 1970; Junguito, 2015).
- 12 Due to space constraints, we cannot refer in detail here to the abundant literature on this topic. (...)
38After the demise of the Bretton Woods system, IMF advice evolved towards a position that increasingly supported exchange rate flexibility. In particular, during the following decades, the IMF supported the managed crawling pegs instituted by Colombia, Brazil and Chile. However, due to rising concerns that such regimes contributed to price and wage indexation and persistent moderate inflation, these countries’ policies eventually shifted (at the end of the 1980s or beginning of the 1990s) towards ‘crawling exchange rate bands’, within which the currency was allowed to float. Later on, as high volatility of capital inflows made it difficult, or very costly, to defend such bands, IMF advice and these countries’ policies increasingly turned to favour full exchange rate flexibility. There was a parallel evolution of views regarding monetary regimes, progressively leaning towards support of inflation-targeting regimes, which are consistent with exchange rate flexibility and permit the deployment of countercyclical monetary policies.12
- 13 See, for example, the Board Statement about the completion of Article 4 consultations in Colombia i (...)
39However, it should be noted that even until recently IMF programmes supported—on several occasions—unsustainable, overvalued fixed-exchange rate systems (such as the Argentinean currency board from 1997 until 2001) or bands (as in Brazil in 1998/99). Evaluations delivered by the IMF's Independent Evaluation Office have been especially critical of the fact that the IMF failed to point out the high degree of vulnerability of the Argentinean currency board, did not promote an early, ordered exit, and continued to support the regime in 2000-01 when it was clearly unsustainable (Independent Evaluation Office, 2004; Perry and Servén, 2003). These cases highlight the fact, noted in Section 2 of this chapter, that large countries with significant political sway were often treated in a more deferent manner, even when IMF staff had serious misgivings about their policies. This evolution of IMF advice with respect to exchange rate management proceeded in tandem with the Fund’s views regarding capital account opening. Article 4 consultations criticised the capital controls instituted by Brazil, Chile and Colombia, and other countries, though—on occasion—the IMF Board recognised that use of these controls could be a legitimate tool, when they acted as complements to sound macro-financial policies, in order to avoid destabilising surges of short-term capital inflows.13
- 14 The main author of the current chapter was personally involved in this dispute.
- 15 See, for example, Forteza et al. (2006).
40In 1997, IMF management proposed changing the IMF charter to require full capital account opening from its members. This initiative was strongly opposed by an active coalition of Latin American and Asian countries14 and, as a consequence, failed to gain approval at the IMF Assembly of that year in Singapore. Shortly after, the Asian crisis erupted, demonstrating, once more,15 the dangers of fast and inadequate capital account opening. As a consequence, the IMF position on capital controls evolved into a more nuanced standpoint. Recently, the Fund has come to explicitly recognise the advantages of dirty floating and central bank foreign exchange interventions in ‘dollarised’ economies, such as Peru (see, for example, IMF, 2015), as a means of avoiding excessive appreciations and depreciations that may lead to balance sheet problems and financial instability.
41As this account suggests, some Latin American innovations and positions had an influential role in this evolution of IMF policies on exchange rate management and the occasional use of capital controls.
- 16 Haussman et al. (1996) and Talvi and Végh (2005). The IMF has recently acknowledged these contribut (...)
42Similar changes have taken place with regard to IMF positions on fiscal issues. IMF recommendations and programmes were strongly criticised for placing excessive emphasis on short-terms goals, without due consideration of business cycle or long-term effects. In particular, the focus on short-term fiscal goals in Fund programmes and advice often led, in Latin America, to inconvenient procyclical fiscal policies, excessive cuts in public infrastructure and a lack of concern given to the term and currency composition of public debt (Perry et al., 2008). A significant body of research has demonstrated the highly procyclical character of fiscal policies in Latin America16, drawing attention to the political economy and financial sector determinants of these phenomena, but also to the role played by IMF short-term fiscal goals that did not adjust for cyclical factors.
43Chile was the first country in the region to break out of this vicious circle, which it did in 2001 by enacting a fiscal rule responsive to the domestic business cycle, as well as to deviations of the price of copper (its main source of foreign exchange and fiscal revenues) from its long-run trend (Perry, 2003; Frankel, 2010; and Engel et al., 2011; among others). Colombia in 2011 and Peru in 2015 enacted similar rules. All three of these fiscal rules are at least cycle-neutral and they allow for some degree of counter-cyclicality in special circumstances.
- 17 See Independent Evaluation Office (2003). This evaluation indicated that most programme fiscal adju (...)
44The IMF gradually, though hesitantly, became a supporter of these policy innovations. Since the 2009 global crisis, it has become a strong advocate of countercyclical fiscal policies in industrialised and developing countries with enough fiscal space, and of more gradual, procyclical fiscal adjustments when such adjustments are unavoidable for long-term sustainability (Blanchard and Leigh, 2013). Still, a 2013 evaluation (Independent Evaluation Office, 2013), concludes that although the IMF has made progress in this area since the previous, highly critical assessment delivered in 2003,17 it continues to fail to take sufficiently into account the critical need for countercyclical fiscal policies in commodity-dependent countries, such as most South American nations.
45Similarly, pioneering studies from the World Bank on the ‘public infrastructure lags’ in many Latin American countries have shown that they often originated in the sharp fiscal adjustments of the 1980s and early 1990s, often under IMF programmes, after most countries in the region had to default on their public debt (Calderón and Servén, 2010). Analysts in MDBs drew attention to the high-long term growth costs of such adjustments (Calderón and Servén, 2010) and to their inefficiency and eventual failure from an intertemporal perspective (Easterly and Servén, 2003). Recent advice from the IMF to European countries, recommending the implementation of a major infrastructure investment programme as a means of stimulating aggregate demand in the short run and increasing productivity in the long run, suggest a significant reconsideration of these issues.
- 18 This is, for example, the case with the ambitious 4G highway programme currently being implemented (...)
46More generally, IMF advice and programmes have been increasingly sensitive to the quality of fiscal adjustments, trying to limit their effect on anti-poverty programmes and badly needed infrastructure investment (Independent Evaluation Office, 2013). Consequently, the Fund has begun to accept that infrastructure concessions’ investment costs may be spread out during the economic life of projects if construction risks are born by private firms, thus not showing up as sharp increases in fiscal deficits and public debt in the short run.18
47In a similar vein, in the late 1990s the World Bank and the IDB began to call attention to the fact that several recent fiscal crises, in Latin America and elsewhere, did not have their origins in excessive fiscal deficits and public debt levels (at least as commonly measured at prevailing exchange rate levels) but in currency mismatches that led to unsustainable debt levels when adverse shocks made sharp currency devaluations necessary (IDB, 2007; Fiess, 2004; Perry et al., 2008). The IMF has gradually incorporated these issues into its surveillance work and programmes.
48Equally important, responding to World Bank and Latin American critiques of the adverse effects of IMF-sponsored fiscal adjustment programmes on poverty levels, IMF programmes have begun to protect anti-poverty programmes from overall required fiscal expenditure cuts, from the mid-1990s on. Also, in 1996 the Fund joined MDBs in launching the Heavily Indebted Poor Countries programme, which since then has engineered 36 debt alleviation programmes, designed to reduce the burden of debt and of required fiscal adjustment on growth and poverty levels. It also launched, in 1999, the Poverty Reduction Strategy Papers, as a joint effort with the World Bank and low-income countries’ authorities, with the aim of guaranteeing not only that required macroeconomic adjustment programmes had only a minimum potential effect on low-income groups, but that coordinated MDI support would ensure faster poverty reduction over the medium term. These programmes were prompted by debates over the effects of IMF programmes in the poorest countries of Africa and Asia (Stiglitz, 2006). In Latin America, a region dominated by middle income countries’ concerns, they have benefited countries such as Haiti, Bolivia and Nicaragua.
49Overall, all this shows that the IMF has gradually accepted certain critiques from the academic and policy maker communities with regard to its previously flawed or excessively simplistic analytical frameworks and programmes. It also indicates the need for additional revisions with respect to several policy issues.
- 19 The IMF’s initial actions during the Indonesian crisis of 1998 (closing banks in the midst of the c (...)
50The previous passages illustrated the significant differences between MDBs’ technical positions and IMF official positions on several topics that have affected Latin American and other developing countries. Some of these debates originated around experiences in other regions, as was the case of that regarding IMF programmes’ effects on poverty (mostly in Africa) and the flawed treatment of the 1997 Asian crisis (Stiglitz, 2006). Although these debates were not centred on Latin American issues, there were also instances of such disagreements about Latin American programmes. In particular, strong criticism from the World Bank with regard to the way in which IMF programmes dealt with financial crises eventually led to a significant strengthening of IMF technical capacity in this area and to the creation of the joint Financial Sector Assessment Programs.19
51In summary, Latin American and other developing countries have often not faced a unified ‘orthodox’ view from the Washington based institutions. Such divergence of views between the IMF, the World Bank and the IDB have on occasion led to the IMF changing or moderating initial policy positions that had had negative effects on Latin American economies.
- 20 Environmental and social safeguards are also responsible for underinvestment in infrastructure proj (...)
52Diversity of views and competition among MDBs played other useful roles. As already mentioned, at the height of ideological excesses of the Washington Consensus, World Bank and IDB support for public infrastructure projects faded away, in the naive belief that, with adequate regulation and concession contracts, private investment in infrastructure could largely replace public funds.20 Although this became true in the areas of telecommunications and energy, it was never the case in transport or water supplies and sewerage infrastructure. Fortunately, the CAF—an MDB created, funded and managed exclusively by Latin American countries—did not buy into these extreme views and decided to fill the vacuum left by the World Bank and the IDB. By 2010 it was lending more to public infrastructure projects in Latin America than those two institutions together. Later on, the Washington-based MDBs recognised their mistake, and the business opportunities lost to the CAF, and began to engage again more forcefully in this area.
53This is only one case, albeit the most notable one, in which diversity of views and competition among MDBs have facilitated Latin American countries’ ability to ‘shop around’ and find the necessary cheap, long-term financing without having to accept unwanted MDB conditionality.
54The existence of the Latin American Reserve Fund (Fondo Latinoamericano de Reservas, FLAR) has played a similar role, though significantly more limited, for small countries in Latin America and the Caribbean (LAC) in need of balance of payments support. However, more often than not, FLAR support has played the role of a complementary source to IMF resources, acting as a ‘lender of first resort’ while an IMF programme is put together, taking advantage of the fact that FLAR can proceed much faster than the IMF (FLAR, 2015).
55The fact that the IDB and, especially, the CAF have often been more flexible than the World Bank in supporting home-grown regional government programmes and policies, has probably had something to do with the fact that a larger proportion of their staff come from the region, and hence are more knowledgeable with regard to specific countries’ circumstances and more sympathetic to them. However, in the last analysis, it reflects the stronger weight of Latin American countries in their boards. This is particularly true of the CAF (and FLAR), which are fully owned and managed by Latin American governments and, as a consequence, are not influenced by extra-regional agendas. Further, each has built governance structures that give more flexibility to their administrative bodies in order that they may accommodate themselves to specific and changing country circumstances (Perry, 2016).
56Though ‘orthodox’ views in the MDI world have changed over time, as illustrated in Section 3 of this chapter, the main tenants of their recommended macroeconomic policies (fiscal and monetary discipline and trade openness) have remained fairly constant since the 1980s. Indeed, though changes in MDIs’ official doctrine over time include the adoption of new views on the desired degree of exchange rate flexibility and financial integration (capital account opening), countercyclical fiscal policies, and the quality of fiscal adjustments, the most significant changes in the official doctrine of MDIs have been related to the microeconomic determinants of growth (especially with regard to the role of institutions) and to the increasing weight placed on equity and long-term sustainability issues.
- 21 A similar development took place in Europe after the Second World War, with the ascent to power of (...)
57During recent decades most Latin American governments, independent of their political orientation, have increasingly embraced those basic elements of ‘orthodox’ macroeconomic policies: fiscal and monetary discipline and trade openness.21 This has been the case, for example, with the ‘left-wing’ administrations headed by Lula in Brazil, Lagos and Bachelet in Chile, Tavaré Vasquez and Mujica in Uruguay, Ortega in Nicaragua, Evo Morales in Bolivia, and Toledo and Humala in Perú, and to a lesser extent with Correa in Ecuador. All these governments clearly followed fiscal and monetary policies as strict or more strict than those of their predecessors, and similar to those of contemporary centre-right governments in other Latin American countries, such as the Pastrana, Uribe, and Santos administrations in Colombia and those of Fox, Calderón, and Peña Nieto in Mexico.
58Indeed, although fiscal deficits and inflation do not depend only on government policies, the figures are quite suggestive22. They show, for example, that the process of fiscal consolidation and reduction of inflation rates initiated by the Cardoso administrations in Brazil was deepened under Lula, though the fiscal situation deteriorated under Dilma Rousseff.
- 23 See an interview with Andrés Velasco by Julian Nahuelhual (2010) in La Tercera concerning the Chile (...)
59They also show the tight fiscal stance maintained under the centre-left ‘Concertación’ governments in Chile, and the successful reduction of inflation rates during the terms of Alwyn and of Frei. It is particularly noticeable that the Lagos administration established and strictly followed the now famous Chilean ‘fiscal rule’ and that the first Bachelet administration strengthened this regulation, making it legally binding, and kept high surpluses during boom years (an average of 3 per cent of GDP; a maximum of around 6 per cent) in spite of high political pressure to weaken or abandon the rule. In contrast, the right-wing Piñera administration that followed engineered some weakening of the fiscal stance by changing the structural deficit goal and other parameters.23
60These data also show that presidents elected on populist platforms in Peru—including Toledo and Humala—maintained fiscal balances under control and the lowest inflation rates in the region. Just as surprising, Alan García, who had, during the period 1985‒90, presided over the most populist government in Peru, leading to hyperinflation, a currency crisis and recession, maintained—in his more recent term—fiscal surpluses and a low inflation rate.
61Even more impressive, the Evo Morales administrations in Bolivia kept the largest fiscal surpluses in the region, in sharp contrast not only to what their main ally and mentor (Chávez) did in Venezuela, but also to the policy outcomes of their own centre-right predecessors. The same has been the case with the Ortega administrations in Nicaragua. Fiscal and inflation outcomes also suggest that Correa in Ecuador did not shift towards excessively expansionary fiscal and monetary policies, in comparison to his predecessors.
- 24 Argentina maintained a tight fiscal stance during the Kirchner government and until 2009, but allow (...)
62These facts indicate that these ‘orthodox’ views on macroeconomic policies have become an entrenched conviction in Latin America, irrespective of the political inclinations of governing coalitions. Of course, there have been exceptions to this rule, such as the Chávez and Maduro regimes in Venezuela and, to some extent, the Cristina Fernandez government in Argentina24 (see below).
63Most recent left-wing governments have also maintained relatively open trade policies, at least similar to those of their predecessors, as well as prudent financial regulation and supervision, which avoided bankruptcies or credit crunches during the recent global financial crisis (Perry and Forero, 2014). In particular, the ‘Concertación’ administrations in Chile maintained lower average effective tariffs than had the Pinochet regime and engaged in negotiating a plethora of free trade agreements with developed and developing countries in most regions. Consequently, trade openness is the highest in Chile among LAC7 countries and has continued to rise. Though other LAC7 countries (with the exception of Mexico) maintain more restrictive trade policies than Chile (especially Brazil and Argentina), and exhibit lower scores on indexes of trade openness, figures do not indicate that left-wing regimes have been significantly more prone to protectionism than were their predecessors.25
- 26 Following the major capital outflows that took place at the end of the electoral campaign, the Worl (...)
64Reality checks may have played a significant role in these trends. Thus, for example, Lula may have been forced at the beginning of his first administration to nominate a very ‘orthodox’ economic team (in Finance and at the central bank), in order to restore market confidence in the face of the considerable volatility and capital flight that had been taking place since the latter stages of the election campaign.26 Later on, he appointed a more ‘heterodox’ Minister of Finance, but was keen to maintain the significant autonomy that the central bank had enjoyed during Cardoso’s term and kept fiscal policy within reasonable bounds.
65Lula also kept the pro-market microeconomic reforms of Cardoso (for example, with respect to openness to trade and FDI and public–private partnerships in infrastructure), though he did not continued Cardoso’s reformist drive with regard to other microeconomic issues, except on social policies (see below). His successor, Dilma Rousseff, was more prone to interventionist policies and fiscal expansion in her first term and, partly because of this, ran into serious market confidence problems from 2013 on (Figure 10.8), which together with a negative shock to the terms of trade have brought Brazil into a deep recession (Figure 10.11). These outcomes prompted Rousseff to nominate, as Finance Minister, a well-known ‘orthodox’ candidate—Joaquim Levy—who immediately announced a sharp fiscal adjustment package. Political turmoil impeded its approval in congress and Minister Levy had to resign.
Figure 10.8 LAC7 EMBIs* 2000-16.
*Emerging Market Bond Index.
Source: JP Morgan.
66In Nicaragua, Ortega, in spite of his close relations with Venezuela (which has provided Nicaragua with significant financial support), has followed strict macroeconomic policies and strongly pro-market microeconomic policies. Indeed, he has become the darling of private investors in Central America. Such a dramatic shift from his earlier ideology and practices may either reflect new convictions, having witnessed the economic failure of interventionist and expansionist policies during the Sandinista regime, or a pragmatic stance, given the commitments and opportunities related to the Central American Free Trade Agreement and the obvious importance of large FDI and private investment flows for Nicaragua’s growth.
67As already mentioned, the Morales administrations macroeconomic orthodoxy is noteworthy. In spite of a very close relationship with Venezuela, and initial left-wing and populist microeconomic policies (his ‘nationalisation’ of gas production and his attempts to impose agrarian reform in Santa Cruz and to take away oil royalties and autonomy from provinces), the Morales regime has applied the most austere fiscal and monetary policies in the region (highest fiscal surpluses and accumulation of international reserves; see Figure 9), certainly much more orthodox that any of his predecessors.
68In contrast to a relative convergence with regard to macro-financial policies, there have been more differences between Latin American governments’ respective microeconomic and social policies. At one extreme, following the return of democracy in Chile the victorious centre-left coalition decided to maintain and deepen most of the pro-market macroeconomic, trade, financial and microeconomic policies that had succeeded in delivering high growth and stability during the second half of the Pinochet dictatorship, but to complement them with aggressive and highly innovative social policies. Thus, the ‘Concertacion’ governments, in addition to the world-famous ‘fiscal-rule’, developed the best financial sector regulation and supervision in the region, promoted more free-trade agreements than any other country in the region, implemented highly ambitious and well-designed public–private partnerships in infrastructure, engaged in a second (and now a third) generation reform of its pioneering pension system and put in place ambitious policies to promote innovation. At the same time, they implemented successive ambitious educational and health reforms and the first programme in the region that attempted to eradicate extreme poverty (Chile Solidario). Though MDIs had little to do with most of these initiatives (or with those implemented under the Pinochet regime), they were later on adopted as part of MDIs’ recommended ‘package’ of policies.
- 27 Building on Bolsa Escolha, a more limited programme instituted in some states during the Cardoso ad (...)
69Uruguay’s Frente Amplio administrations, and the Lula administrations in Brazil, considerably strengthened social policies and introduced innovations into them (including Bolsa Familia and Primer Emprego in Brasil).27 However, although they maintained the pro-market microeconomic reforms of previous administrations, neither deepened them considerably. As previously mentioned, there were some setbacks in several of these areas during the Rousseff administration.
Figure 10.9: International reserves/percentage GDP, 2000-15.
Source: World Development Indicators.
70All of these governments had very close and productive relations with MDIs. This was in particular the case for the close collaboration between the World Bank and the first Lula government. For example, the World Bank suggested to the Lula government that it institute an aggressive Conditional Cash Transfer programme (Bolsa Familia, based on an expansion of the previous Bolsa Escolha programme) in lieu of the set of heavily interventionist social programmes initially envisaged by the administration, and helped in designing and financing it.
71Something very different happened in the case of Venezuela, and—to a lesser extent—in the cases of Argentina and Ecuador. The Chávez and Maduro governments had the largest deviations from ‘orthodox’ macroeconomic policies and pro-market microeconomic policies. In spite of benefiting from the highest foreign exchange earnings and fiscal boom in Venezuelan history (Figure 10.12)—due to the large increase in oil prices from 2003 to 2014—Venezuela incurred significant fiscal deficits during part of this period, lost market access (since 2009) and ran out of international reserves (Figure 10.9), even while imposing strict capital controls and devaluing, on several occasions, the official currency rate to the dollar (Figure 10.10), in spite of which the difference between the black market and the official exchange rate skyrocketed. In addition, Venezuela has had the highest inflation in the region, as a consequence of a huge monetary expansion to finance the growing fiscal deficits, and the sharp devaluations of the currency.
72Microeconomic policies in Venezuela have been characterised by a strong anti-private sector and populist tone: expropriations (including of supermarkets and cement factories), generalised price controls, severe restrictions on trade flows with neighbouring Colombia, and so on. Though there has been significant expenditure on social programmes (the so-called Misiones), some of these programmes have been questioned regarding corruption and ineffectiveness, and have used less public resources than the huge subsidies applied to gasoline prices and electricity, subsidies that favour the upper middle classes and the rich. This combination of populist macro- and microeconomic policies led the country into a severe contraction, which began even before the fall in oil prices (Figure 10.11).
Figure 10.10 Real exchange rates—2000-15.
Source: Bank of International Settlements. Argentinean data was calculated using the official exchange rate.
Figure 10.11 Growth rates by country.
Sources: World Bank, WEO October 2014 and WEO Update 2016.
73Argentinean microeconomic policies since 2000, and especially since 2012, show some similarities to those of Venezuela (expropriations, generalised tariff and price controls, export taxes and prohibitions, and import restrictions), but have never been as extreme. Macro-financial management, however, remained very prudent until 2009. Indeed, low fiscal deficits, limited monetary expansion and a competitive exchange rate (Figure 10.10) facilitated a very fast recovery from the 1999-2002 crisis (Figure 10.11), in spite of the adverse effect of some microeconomic policies.
74However, since 2009 fiscal deficits and inflation rates increased, the country lost access to international capital markets (Figure 10.8) and the central bank lost reserves rapidly (Figure 10.9), even before the fall in commodity prices and in spite of the imposition of capital controls. Consequently, there has been a more significant slowdown than that experienced by other countries in the region that were affected by more severe terms of trade shocks than those that Argentina suffered (Figure 10.12).
Figure 10.12 Net Barter Terms of Trade Index.
Source: World Development Indicators.
- 28 Indexes of institutional quality can be questioned though; because they are often based on entrepre (...)
75The significantly larger distortions caused by microeconomic policies introduced by Argentinean governments since 2000, and especially by Venezuela during the Chávez and Maduro eras, are reflected in the weakening of the Kauffmann–Kraay indicators of institutional quality28 in these countries. These figures also show the significant Chilean advantage over other Latin American countries in these indexes (see Figure 10.13 for the example of regulatory quality); an advantage gained thanks to the continuous process of pro-market microeconomic reforms that accompanied the Concertación administrations.
76This brief account suggests that the few governments that have engaged in openly populist macro- and microeconomic policies during the early 2000´s, have eventually led their economies into output contractions that other countries have so far avoided even while suffering similar adverse external shocks. It also illustrates the close and productive collaboration of MDIs with most recent governments in the region, governments of widely different political inclinations.
Figure 10.13 Governance indicators.
Source: World Development Indicators.
77This chapter shows that there have been significant changes over time in MDIs’ paradigms and policies and in their relationship to, and degree of influence over, Latin American governments’ development policies.
78MDIs, influenced by academic research findings and, on occasion, by policy makers’ views—in Latin America and other developing regions—have often recognised their previous mistakes and excessively simplistic views, and, consequently, have adapted their official positions. The analysis presented here also illustrates how a divergence of views on particular issues has also been common among MDIs themselves, and how regional MDIs (the IBD, the CAF, and the FLAR) have been more responsive to regional concerns and views and have played an increasingly important role and exerted growing influence within the international community.
79The chapter also shows how the degree of influence of MDIs has naturally varied with the development and cycles of international private financial markets and individual countries’ access to them, among other factors. Thus, that influence was at its height during the Latin American debt crisis of the 1980s (the period of ‘structural adjustment’ and the so-called Washington Consensus, when most countries in the region lost access to international private capital markets) but has significantly declined since the early 1990s, with short-lived resurgences after the 1998 Russian crisis’s impact on capital flow reversals in the region and during the recent global financial crisis in 2009. MDI influence is likely to increase somewhat again in the current environment of lower terms of trade and capital inflows.
80Further, the chapter illustrates how even strongly pro-market governments often had serious confrontations with what they considered the simplistic views of MDIs with regard to important issues of policy. Some of these governments’ policy innovations, in either macro- or microeconomic and in social policies, were later adopted and promoted elsewhere by MDIs. At the same time, we have seen how many pro-market and social policy reforms resulted from shared views between Latin American policy makers and MDIs, or were the sole initiative of Latin American governments.
81Finally, the chapter shows how governments in the region—governments of a wide variety of political inclinations—have, with some notable exceptions, increasingly applied orthodox macroeconomic policies and how they have enjoyed close collaboration with MDIs and have found that relationship useful in their pursuit of economic and social objectives, but that there still remains considerable divergence in views on microeconomic and social development policies across the region.
82The landscape that emerges is, thus, more complex than a division of the world between ‘orthodox’ and ‘unorthodox’ views or strategies, either among ‘left-wing’ or ‘right-wing’ governments, or among MDIs themselves. It is better described as a changing landscape with mutual influences exerted between MDIs and regional views.