Leslie Elliott Armijo; email@example.com
Draft of July 24, 2013
Prepared for inclusion in Peter Kingstone and Timothy Power, eds.
Democratic Brazil Ascendant,
What is Brazil’s development model for the twenty-first century? This chapter argues that the ubiquitous term “new developmentalism” is best used to identify the large degree of consensus on economic policies that has characterized all of Brazil’s governments since President Fernando Collor. To understand Brazil’s development model, I examine the recent trajectory of the country’s industrial development bank, the BNDES (National Bank for Economic and Social Development) through the lens of the “public bank trilemma.” The trilemma encapsulates the larger conflict inherent in state-led capitalist development: that of balancing autonomous, technocratic decisionmaking (“Expertise”), and the immediate demands of ordinary citizens for a better-life, as expressed through their political representatives (“Democracy”), with the delicate challenge of using state monies and regulatory authority to stimulate voluntary, decentralized, competitive, and efficient private investment (“Markets”). The fights over BNDES operations conveniently illustrate many of Brazil’s on-going debates over the pace, biases, and goals of mixed-capitalist economic growth in a middle-income, but still highly unequal, country.
Sections one and two introduce the new developmentalism, reasons for the focus on the BNDES, and the public bank trilemma. The next three sections summarize the Bank’s history, then review the evidence for the trilemma in recent debates over Brazil’s external economic insertion and the BNDES’ relationship with private banks and investors. The conclusion comments on Brazil’s development model.
I. Brazil’s “new developmentalisn” and the BNDES
The old developmentalism in Brazil and Latin America was the import-substituting industrialization (ISI) of the 1950s through 1970s, which promised, but did not deliver, rapid catch-up with the wealthy countries through protected industrial production for the home market. As contrasted to ISI, the new developmentalism (ND) of the 1990s through the present has emphasized macroeconomic stability (low inflation), greater participation in global markets, and reducing domestic inequality.1 Some prominent proponents of ND have championed, in addition, the need to rely on domestic savings rather than foreign loans or investment and, especially, to maintain a competitive (non-overvalued) exchange rate (Bresser Pereira 2009). The ND is not a sharp break with the past, sharing with the ISI-era a belief in the essential economic role of the state in identifying strategic goals and providing public goods. In Brazil, the state has played the central role in financing infrastructure since the early 1950s. Astonishingly, Brazil’s federal government, through the BNDES, also has been virtually the only source of medium-to-long term bank credit for private industry during the entire period from the late 1960s to the present. The BNDES has been central to each phase of Brazilian developmentalism.
Why adopt the BNDES as the symbol of the ND? First, the BNDES, founded in 1952, spans both the old and the new developmentalism.2 Second, it is a dominant institution that has recently become even larger. Its recent growth in annual loan disbursements, shown in nominal reais in Figure 1, has been truly extraordinary. In 2010 the annual flow of gross new BNDES loans briefly reached 4.6 percent of GDP, before falling back to “only” 3 percent in 2011.3 As of late 2009, the BNDES accounted for 19 percent of the stock of all outstanding loans from the entire national financial system to households and firms in Brazil. The Bank also supports Brazilian businesses through consulting and long-term financing through equity participation. By end 2011, the bank held assets worth 15 percent of Brazil’s R$4.14 trillion (about US$1.44 trillion) GDP, including loans (10 percent of GDP), corporate shares and debt securities (another 3 percent), and other assets and investments, including Brazilian public debt (2 percent).
< Figure 1 about here >
Brazil’s BNDES also looms large in comparative international perspective. As shown in Table 1, the BNDES, although legally limited to support only Brazilian companies and their customers, is significantly larger than the Inter-American Development Bank, and roughly equivalent in size to the World Bank. Although its loan portfolio per se is only half the size of that of the China Development Bank, BNDES President Luciano Coutinho confidently notes that this understates BNDES’ influence and versatility: “ ‘The Chinese and South Korean development banks lend more than us, but they only do monolines. …We do credit, we support capital markets, and we operate through the private banking sector itself.’ ” (Wheatley 2009).
< Table 1 about here >
Finally, there are clear sub-periods within Brazil’s new developmentalism. Thus, some observers term the governments of Presidents Fernando Collor, Itamar Franco, and Fernando Henrique Cardoso in the 1990s through 2002 “neoliberal,” because of their emphasis on fiscal responsibility, privatization, and freer capital movements, reserving the “new developmentalist” label for the post 2002 period of PT Presidents Luiz Inácio Lula da Silva and Dilma Rousseff (von Mettenheim 2010). In my view, the Collor to Cardoso period was clearly pro-market (Armijo and Faucher 2002), yet Brazil’s policies were less ideological and extreme than contemporaneous reforms elsewhere in Latin America. There has been substantial continuity in Brazil’s ND economic policymaking, as practiced if not always as articulated, from the early 1990s through the present. Partisan differences within the post-1990 consensus, although they exist, look more important from within the country than from without. Nonetheless,
topical controversies over the future trajectory of the BNDES go to the heart of important nuances currently being debated within Brazil, within the larger ND consensus. The politics of the BNDES encapsulate arguments about the country’s future.
II. The public bank trilemma
As one of the world’s largest development banks, the BNDES illustrates what I term the “public bank trilemma,” which is particularly acute for economic policymakers in developing countries seeking rapid industrial catch-up. The public bank trilemma is similar to the better known “central bank [regulatory] dilemma.”4 On the one hand, it is accepted wisdom that central banks should be independent of the political authorities in carrying out their day to day policy implementation of monetary policy (Goodhart 1990; Bernhard, Broz, and Clark, eds. 2003). In this way they can resist pressures to sacrifice price stability for electoral expediency, as by easing liquidity in the months prior to elections, generating the infamous “political-business cycle.” Yet on the other hand, monetary policy is not purely a technical matter. Only the legitimate political authorities can decide whether central bank governors should target only inflation, or also should consider how money and credit aggregates influence other outcomes, such as unemployment or the nation’s exchange rate. Moreover, it cannot be assumed that central bankers, like other experts, have no interests and preferences of their own—which may or may not coincide with those of the larger body politik. For example, central bankers typically come from, and share in an epistemic community with, private bankers and financiers, a circumstance that should skew their preferences toward low inflation at all costs, even when they are asked to consider growth as well (Epstein 1992; Henning 1994). The central bank regulatory dilemma is thus that of balancing the imperatives of independence of the technical experts (“Expertise”) with oversight and goal-setting by the legitimate, democratically-chosen, authorities (“Democracy”).
The challenge of regulating a public bank in any modern mixed-capitalist economy is more complex still (Caprio et al. 2005; Bonney, ed. 1995; Allen and Gale 2000; Aghion 1999; Bacha 2002; Schleifer 1998). The core rationale for creating a public bank (or having a national industrial policy more generally) is to empower competent technocrats, operating above the political fray, to aggregate financial resources, typically subsidized by citizens, and make investments (usually long-term) resulting in public goods that otherwise would not be provided (“Expertise”) (cf. Shapiro 2007). Like central banks, effective public banks should be independent of partisan and short-term political pressures. Nonetheless, decisions about the outcomes public bank activities ought to promote involve choices about values, and thus belong with the country’s legitimately-elected political authorities, charged to execute the will of the citizenry (“Democracy”). Public sector banks have the additional job of coexisting with, and promoting, a vibrant, innovative, profitable, capitalist economy, including competent private financial sector. To fulfill their core expert task of promoting economic and industrial growth, successful state banks must support, rather than undercut or substitute for, private credit and capital markets (“Markets”). A fair evaluation of the performance of any public bank must consider each leg of this complicated regulatory trilemma.
The public bank trilemma frames our analysis of the BNDES. The initial rationale for public banks, and thus the apex of the triangle, is Expertise. A public bank that cannot plausibly claim to be providing public goods in a competent fashion has no reason to exist. Nonetheless, Expertise alone is insufficient. Following a short history, two sections analyze contemporary tensions in the BNDES between Expertise and Democracy, and then Expertise and Markets.
III. BNDES’ shifting roles over time
Founded in 1952 to furnish project finance for big infrastructure projects, later in the decade the BNDES also began financing capital goods production through its wholly-owned subsidiary, FINAME.5 Following the 1964 military coup, which the new government justified by reference to economic disarray under its civilian predecessors, a barrage of liberalizing and institution-building reforms attempted to construct the regulatory framework for macroeconomic stability and capitalist economic growth (Fishlow 1973; Skidmore 1973). However, inflation was not controlled, and long-term, voluntary private financing did not appear. Instead, in the late 1960s large business-owners, especially from São Paulo, successfully pressed the government to reorient BNDES activity away from infrastructure and toward long-term lending for industrial and commercial firms across all sectors. Ever since, long-term credit for business borrowers sourced from within Brazil has been a virtual monopoly of the BNDES.
The BNDES maintains an enviable reputation for competence--and flexibility. In the 1970s, the bank sought to ease fiscal pressures on Brazil’s central government, hitherto its primary source of funds, by borrowing cheaply in international markets and on-lending these resources within Brazil. Accumulated foreign loans in the 1970s lead to the debt crisis in the early 1980s, when rates shot up and foreign banks suddenly refused to renew working capital loans and trade credits. Throughout that decade, the BNDES helped the country cope with stagnation and firms starved for credit by inaugurating another wholly-owned subsidiary, BNDES Participações (BNDESPar), which would support companies by investing in their shares, rescuing illiquid firms from bankruptcy through debt-to-equity conversions. When markets recovered, BNDESPar managers traded these shares, in most years turning a profit.
In the post-debt-crisis 1990s, Brazil, like most of Latin America, entered a phase of pro-market economic reforms. Called upon to run Brazil’s large-scale privatization process, the BNDES ultimately oversaw sales of US$105 billion (BNDES 2002b:2). As part of the process, economic policymakers usefully cleared out US$18 billion in public debt securities of questionable value (moedas podres, or rotten monies) that had been issued by various state firms and entities over the previous twenty years: these securities could be spent in the privatization auctions. From the late 1990s through the present, the BNDES has promoted exports and Brazil’s global economic insertion, while also inaugurating new programs for small businesses. The Bank has seldom been free of controversy: earlier arguments over its proper role were as fierce as those that now rage. Throughout, the BNDES has retained its reputation, both within Brazil and internationally, for skill and integrity. Recently, Brazil has made two significant transitions, to institutionalized democracy in the mid-1980s and a stable macroeconomy a decade later. Together they render Brazil’s public bank management trilemma more acute.
IV. Expertise and the claims of Democracy: The BNDES and the internationalization of Brazilian business
The conflict between autonomous, technocratic policymaking (Expertise) and political oversight by the elected authorities, responding to public preferences (Democracy), is illustrated by the way the BNDES has promoted greater engagement with the world economy.
There are a variety of ways to implement an outward economic orientation, ranging from the least to most state-interventionist. One may unilaterally remove tariff and non-tariff trade barriers, on the theory that all trade (even when it involves a surge of imports) is good trade, because post-liberalization the country’s consumers and firms enjoy better or cheaper goods. A somewhat more activist state will seek reciprocal removal of trade barriers. More statist still is the belief that “good” trade means the export of manufactures, not commodities, and that it is the government’s job to promote this “higher value-added” trade. Both the right-leaning (Collor through Cardoso) and left-leaning (Lula and Dilma) strands of Brazil’s new developmentalists have been are in agreement: it is better to export airplanes than coffee or soya. Even a recent World Bank report worriedly notes that Brazil’s high technology exports grew 35 percent (by value) in the decade following 2000—as compared to over 800 percent for China and almost 400 percent for India (Canuto, Cavallari, and Reis 2013). Meanwhile, the share of Brazil’s total trade that could be classed as high-tech fell from 10 to 5 percent during the decade, while simple commodities increased their share from 46 to 63 percent.
The BNDES has sought to promote exports, particularly manufactured ones. Since foreign customers will not pay the extraordinarily high nominal and real interest rates that have reigned for decades in the free segments of Brazilian markets (on which see below), funds to provide credit to foreign purchasers of Brazilian exports either must be borrowed abroad or the Brazilian government (and ultimately taxpayers) must subsidize the additional interest cost. Brazil’s private banks provide short-term trade credit, funding themselves from abroad (Rossi and Prata 2009). However, longer term credit for big-ticket manufactured exports, as well as export financing for firms too small to access global financing, increasingly comes from the BNDES. The BNDES website actively encourages borrowers to combine resources from multiple different credit lines, so the total of financial resources dedicated to assisting exports likely is many times larger than official totals for “export” financing. BNDES Exim loans (earlier called FINAMEX), nonetheless have increased dramatically, rising from 4 percent of total BNDES loan disbursements in 1996 to 11.5 percent in 2009.6 Real resources going to support exports have increased even more, given the large expansion in BNDES disbursements during this period (as in Figure 1). The U.S. Eximbank expresses concern that the trio of China, Brazil, and India all subsidize exports significantly (U.S. Eximbank 2005: Appendix F).
More intriguing--and controversial--are the further steps the BNDES has taken since the mid-2000s, termed promoting the “internationalization” of Brazilian business. The BNDES, at the behest of Brazil’s long-serving finance minister, Guido Mantega, has promoted “national champions,” that is, specific firms judged able to compete with the largest transnationals in global markets. De facto, this means that the taxpayer-supported BNDES acts as a full-service merchant bank to Brazil’s largest transnational firms, financing almost any activity that forms part of one of these firms’ strategic corporate vision. Most innovative--or merely startling—have been BNDES loans to giant companies such as the meatpacker, JBS Friboi, now one of the largest firms in its industry worldwide, to gobble up foreign rivals through mergers and acquisitions (Latin Finance 2009). The BNDES assisted the wood products company, Aracruz, to recoup foreign exchange bets gone wrong: in 2009 the bank arranged a merger with Brazilian rival Votorantim Papel e Cellulose, forming the new company Fibria (Valle 2012). The BNDES also has opened large credit lines for foreign governments to acquire Brazilian goods and services, often from one or a handful of companies. From 2006 to 2012, the BNDES extended US$ 3.2 billion in loans to the government of Angola, of which almost half purchased consulting and construction services from Brazil’s mega-engineering firm, Odebrecht (Fellet 2012).
A few firms have received a lot of money. From 2008-12 the BNDES lent a cumulative R$40.8 billion (approximately US$22.7 billion) to only six firms, all privately-owned: JBS Friboi, Marfrig (both frozen foods), Oi (telecommunications), BRF Brasil Foods (food and beverages), Fibria (pulp and paper), and Ambev (food and beverages) (O Globo 2011). Of the BNDES’ total loan portfolio of outstanding loans as of mid-2012, 39 percent had gone to the top 5 borrowers, and a further 28 percent to the 50 next largest borrowers, revealing a high degree of concentration of funding (BNDES 2011:53). 80 percent by value of BNDES’ equity holdings are of shares of state-owned or de facto controlled firms (such as Petrobras and Vale, in petroleum and mining, respectively), each of which would fit any list of plausible national champions (BNDES 2012 Acctng Rpt: 4). Yet the Bank also owns large blocs of shares in firms generally understood as private, including 30.4 percent of Fibria, 20.25 percent of Klabin (pulp and paper), 17.3 percent of JBS Friboi, 13.9 percent of Marfrig, 12.2 percent of America Latina Logística (transportation infrastructure), and 3.5 percent of Gerdau (construction) (Bugiato 2012:31).
A related goal of Brazil’s ND policies from the mid-1990s onwards has been South American economic integration, by both reducing intra-continental trade barriers and promoting the construction of new transportation and power infrastructure. Presidents of both center-right and center-left have supported the complex of 31 projects, all involving two or more sovereign states, organized under the aegis of the IIRSA (Integrated Regional Infrastructure for South America) program, an ambitious multilateral collaboration process inaugurated in 2000.7 The BNDES has provided whatever funds it can, within its legal mandate to support only Brazilian firms. (Thus, the BNDES will finance, for example, up to 80 percent of a loan to Ecuador’s government to hire a Brazilian construction company to build a dam.) For Brazilian ND governments, the focus on South American integration makes economic as well as political sense: While Brazil’s exports to Asia are overwhelmingly commodities, and its sales to the U.S. and EU contain a mix of high and low-value-added products, Brazilian exports to Latin America and the Caribbean are overwhelmingly manufactures, as shown in Table 2. The outward foreign direct investment (“internationalization”) of Brazilian firms also has been focused on South America, with Argentina and Venezuela being particularly lucrative host markets, a circumstance that explains the otherwise puzzling support of Brazil’s conservative business community for the entry of left-leaning Venezuela into MERCOSUR (Lissardy 2011; ESP 2012).
< Table 2 about here >
Is there a problem? The Expertise displayed continues high: for example, BNDES officials take great care to demonstrate that the BNDES is profitable and a good steward of the funds entrusted to it—although this judgment takes as a given its low cost of funds. But some critics maintain that the Bank’s heavy support for oligopoly players undercuts Democracy.. Critics from the left, both within and outside Brazil, conceptualize Brazil’s continental policies as expansionist and neoimperialist (Luce 2007; Zibechi 2012; Costas 2012). Others question BNDES financing Brazilian firms to move their production—and thus potential Brazilian jobs—abroad. Building on an index developed by UNCTAD, researchers at the Fundação Dom Cabral calculated a firm’s “transnationalization,” as the mean of three qualities: assets abroad as a share of total assets, revenues abroad as a share of total revenues, and employment abroad as a share of total employment. By this metric, the list of the most transnationalized Brazilian firms--including such names as JBS Friboi (54 percent transnationalized, with 62 percent of employees abroad), Gerdau (52 percent), Stefanini IT Solutions (46 percent), Metalfrio (45 percent), Marfrig (45 percent), and Odebrecht (42 percent)--overlaps quite closely with the list of the largest borrowers from the BNDES (Cretoiu et al. 2012:17). The question is not whether these firms ought to be expanding abroad, but whether their corporate strategies ought to receive public subsidies.
In 2011 one incident sparked particular public ire. Brazil’s large supermarket chain, Pão de Açucar, proposed to merge with a major competitor, the Brazilian subsidiary of the French supermarket chain, Carrefour—in a €2 billion plus deal arranged by the BNDES. The Bank’s press statement highlighted the prospect of promoting sales of Brazilian products through the French parent firm’s worldwide retail network (Rosas 2011). Five term federal deputy Darcísio Perondi (PMDB-RS), despite being nominally a member of the government’s coalition, was among those who quickly attacked the project in Congress, calling it crazy to spend monies derived from payroll and social security taxes on a supermarket, particularly given unmet health and education needs, and noting that the merged company could use its oligopoly share of 32 percent of the Brazilian retail market to squeeze both small farmer suppliers and customers (Perondi 2011). Public opposition was sufficiently intense that the BNDES backed off, offering the bland excuse that not all of the anticipated private co-funding had materialized (Leahy 2011).
However, and despite understandable differences of opinion over BNDES support for the internationalization of firms, there is not a problem from the viewpoint of the public bank regulatory trilemma—because the Bank’s policies have been those of the legitimately-elected executive branch political authorities. BNDES policies support Brazil’s ND economic strategy of increasing manufactured exports and also the country’s de facto foreign policy strategy of being perceived of as a regional leader, thus enhancing its global political status. This emphasis has been consistent since the 1990s. In the words of Brasília business consultant Thiago de Aragão, “Brazil’s number one interest is to make itself influential within the region, and to have its neighbors recognize it as an instrument of regional development” (Lissardy 2011, my translation). Were the center-right PSDB-led coalition to recapture the central government, incremental downshifts in the BNDES’ currently assertive international role could be likely—but a dramatic change of direction is improbable. Overall, BNDES activity in support of the internationalization of firms, although controversial, has been consistent with the need to balance the demands of Expertise and Democracy.
V. Expertise and the claims of Markets: BNDES and the private financial system
The balance within the Expertise versus Markets vector of the public bank regulatory trilemma also has become contentious, revealing clear differences between the center-right and center-left within the larger ND model. For decades, from at least the 1930s through the mid-1990s, Brazil was in the peculiar position of having an economy that suffered from chronic high to very high inflation, yet always managed to avoid the disastrous breakdown associated with true hyperinflation (Armijo 1997). One of the consequences of high inflation was the disappearance, or failure to develop, of long-term credit in the economy, as bankers would not loan long-term in a currency that they expected to be devalued when repaid. When the military took power in the mid-1960s, the administration of President Castello Branco attempted to end inflation and stimulate voluntary private long-term finance (Syvrud 1975; Trubeck 1971; Armijo 1993). As stabilization failed, subsequent military policymakers pushed the BNDES to move into this role. This sets up today’s policy conundrum. Since the Plano Real, the successful stabilization program of the mid-1990s, Brazil’s annual inflation has averaged in the high single digits—in other words, the era of chronic crazy inflation is over. Continuing the wholesale substitution of Expertise for Market provision of long-term financing unbalances the public bank triad of goals. However, like other organizations, particularly ones whose personnel perceive their institutions to be at the top of their games, the BNDES resists shrinking.
Former senior economic policymakers associated with the President Cardoso--including former BNDES and Brazilian Central Bank (BCB) President Pérsio Arida, now a partner in one of Brazil’s two largest private investment banks, former BNDES President Luiz Carlos Mendonça de Barros, and former BCB President Gustavo Loyola—have been prominent critics of the BNDES’ role in Brazilian financial markets (Arida 2005; Leonel 2010). The critics make these main points. First, the majority of BNDES funding derives from Brazilian taxpayers. As recently as 2007, about 10 percent of funds were shareholder’s (the central government’s) equity, and another 75 percent transfers from the federal government, the bulk in the form of obligatory “investments” by the FAT (Workers’ Support Fund, whose resources come from payroll taxes) in the BNDES (BNDES 2011:43). The BNDES has a statutory obligation to employ FAT funds for projects to increase Brazilian employment and production. Although these investments (loans to the BNDES) are safe, they are remunerated at a special “long-term interest rate,” the TJLP, which is far below the free market cost of funds in Brazil. There is thus a question of the opportunity cost of directing these monies to the BNDES. Moreover, much of the increase in BNDES disbursements in recent years (see Figure 1) has been funded by direct transfers from the National Treasury, through special programs known as PAG (Accelerated Growth Program) and PSI (Sustainable Investment Program), which by 2011 accounted for half of all BNDES funding. Unlike the FAT, whose connection to the BNDES the 1988 Constitution mandates, PAG and PSI funding reflects high priority, multiyear, and massive--yet intrinsically temporary--funds created by Executive Branch decisions.
Second, it is the BNDES that is pushing up real interest rates in the free segment of the economy (Arida 2005; Romero 2011; Mello and Garcia 2011). There is only a set quantum of savings in Brazil: therefore, the larger the share that the BNDES accesses at below-market rates, the smaller the share available for all other borrowers—and therefore the higher the price at which credit in the free market will be offered. A rough approximation of the extent of the subsidy enjoyed by the BNDES may be seen by the difference between the SELIC, Brazil’s policy interest rate, which is the rate at which the central bank, BCB, makes short-term loans to commercial banks needing liquidity, and the TJLP, the rate at which the BNDES borrows. Normally, a central bank’s policy rate (in the U.S., the federal funds rate) is below the best rate available to the most-credit worthy business borrowers (in the U.S., the prime rate), as banks need to borrow more cheaply than they lend. In Brazil, however, the TJLP is notably less than the SELIC.8 True, the gap between the two has shrunk as the economy has become used to low inflation. In 1995, the year the Plano Real took effect, the SELIC was 87 percent annually and the TJLP 25 percent, representing an approximate subsidy of a stunning 62 percentage points, while in mid-2011, the SELIC had been brought down to 12 percent, while the TJLP stood at 6 percent, for “only” a 6 percentage point subsidy (Mello and Garcia 2011:10).9 Meanwhile, the free market interest rate for 30-day working capital for excellent corporate borrowers had fallen from 42 percent in January 2009, at the height of the global financial crisis, to “only” 32 percent in mid-2012 (EIU 2012:41). Naturally, those firms able to do so would prefer to borrow either from the BNDES or abroad. In traditionally very high interest rate Brazil, when Brazilian consumers see real (not nominal) interest rates fall to about 70 percent a year, as they did in 2012, they rush to purchase cars, refrigerators, and other big ticket items.
Third, over the past twenty odd years the BNDES has demonstrated that it can capture resources through bond issues in private capital markets, both within Brazil and internationally. Nonetheless, market-rate funding currently makes up only a small segment of the BNDES’ total funding sources—less than 10 percent in 2011 (BNDES 2011:23). The critics propose that the BNDES be obliged to move gradually to market sources of funds, at least for many of its programs. If the consequence is a shrinking of the total size of the institution’s assets and liabilities, then so be it. That is, Expertise, which almost no one questions that the bank displays, should supplement private financial Markets, not replace them.
Fourth and finally, critics of the BNDES’ funding strategy worry that the recent expansion of BNDES lending imperils Brazilian public finances. While monies booked as investments by various social insurance funds paid into by workers (FGTS, PIS/PASEP, and FAT, all of which have provided resources to the BNDES over the years) have no direct fiscal implications, transfers from the National Treasury (only 7 percent of BNDES resources in 2007, but fully 49 percent in 2011, as per BNDES 2011:43) represent money borrowed from the public—and at significantly higher rates than the TJLP. Mello and Garcia (2011:10-11) conclude, “If BNDES keeps expanding as it did since the 2008 crisis, it will surely constitute a major threat to the solvency of the Brazilian government.”
The response from defenders of the BNDES’ current profile, including Finance Minister (and former BNDES President) Guido Mantega and BNDES President Luciano Coutinho, as well as former BNDES President Carlos Lessa and many others, coheres around five main points (Lessa 2005; Torres Filho 2005, 2009; Torres Filho and Costa 2012). As Coutinho patiently explains, the reason for large grants of public money to the BNDES is that Brazil has urgent needs for centrally-prioritized investments, which provide an invaluable public good (Wheatley 2009). President Lula da Silva also frequently reiterated his support for the BNDES as essential to the country’s commitment to public investment in exports, in industry, in basic research, asserting that, “the world will continue to need more food, and Brazil has all the proper conditions to produce part of that food. .. [W]e have just discovered a lot of oil, and we do not want to use oil as traditionally the oil countries have used oil, … We want to be exporters of oil derivatives, not exporters of oil, because we want to develop a strong oil industry and a strong shipbuilding industry together. We want to build our own drilling rigs, our own offshore platforms, and our own ships. And we want to develop a strong petrochemicals industry..” (FT November 8, 2009). The argument is that these projects are simply too big, too risky, and too long-term for the private sector to handle: hence the need for public financing.
Second, BNDES officials insist that the bank has nothing to do with Brazil’s problem of high interest rates or the related problem of high bank spreads between their deposit and lending interest rates. The origins of Brazil’s high interest rates are historically-generated expectations being extrapolated forward, combined with oligopoly in the commercial banking sector, the latter compounded by significant barriers to entry into banking. In BNDES President Coutinho’s words, “The distortion is that short-term borrowing rates in Brazil are much higher than in other countries. Our long-term rates are a bit higher than in the US, the UK or South Korea. There’s no distortion there. The problem is in the short term. But thanks to the macro-economic advances under President Lula, the spread between short and long term rates has fallen. This is our last big challenge: to have civilised interest rates” (Wheatley 2009). That is, since the BNDES loans only long-term, and at low rates linked to its low cost of funds (the TJLP), it hardly can be taxed with responsibility for Brazil’s abusive short-term rates.
Third, the activities of the BNDES do not conflict with the private financial sector—on the contrary, the BNDES “crowds in” private financing and investment by working closely with a wide range of private financial actors. For example, around 50 percent of BNDES lending is “indirect,” consisting of on-lending through a nationwide network of private banks. The BNDES helps keep these smaller banks, who are close to their small business customers throughout Brazil, in business. For big projects, including major new infrastructure undertakings such as hydroelectric dams, the BNDES does the project design and provides partial financing, but also offers lucrative co-financing opportunities for Brazil’s largest financial institutions banks, whether the big universal banks or smaller investment/merchant banks. Moreover, the BNDES has been a great help in the expansion and technical improvement of Brazil’s capital markets—which are finally beginning to provide an alternative domestic source of long-term financing. Thus, for example, in December 2010 Brazil’s securities’ exchange, the BM&FBovespa, and the BNDES, together launched the Carbon Efficient Index, a tradable index asset consisting only of the shares of “green” Brazilian firms, just in time to be announced at the meeting of the Cancún, Mexico United Nations Conference of the Parties (COP-16) of the Framework Convention on Climate Change (BNDES News, December 8, 2010). Moreover, BNDESPar is one of Brazil’s major institutional investors, and its trading activity assists in keeping the market liquid.
Fourth, the defenders of the BNDES’ current activities agree that the bank’s current overwhelming dominance of long-term financing in Brazil must change, but suggest that such changes already are in progress (Torres and Costa 2012). BNDES President Coutinho notes that BNDES currently is responsible for about 20 to 25 percent of all investment in Brazil, but that he’d like to see this share eventually reduced to only 10 to 12 percent (Reuters February 15, 2011). Moreover, at least one form of long-term financing is beginning to appear: the capital markets. Between 1990 and 2009, total stock market capitalization in Brazil expanded from just under 7 percent to 113 percent of GDP, suggesting a major and salutary recent expansion of this alternate source of long-term finance in Brazil (Beck, Demirgüç-Kunt, and Levine 2000 ).
Fifth, Brazil’s government needs a strong, competent, flexible instrument to respond to changing national and international circumstances, and the BNDES has proven its worth on this score time and again (Lessa 2005; Torres Filho 2006, 2007). To take a recent but stunning example, Brazil performed well in the recent global financial crisis of 2008-9, which the finance minister responded to by two main actions: the government instituted a series of new, temporary capital controls, while also expanding liquidity both directly and by transferring a large sum of money from the National Treasury to the BNDES, leading to the expansion shown in Figure 1 above. Because the BNDES was competent, honest, and had (with its nationwide network of private bank partners) a long list of plausible investment projects more or less ready to go, the stimulus quickly got out to the real economy, with most of it going for genuinely valuable investments, although consumer credit markets in automobiles and residential housing did begin to overheat. Even the IMF praised Brazil’s crisis management, observing, “Due to deft policy responses and built-in financial system buffers, the financial system weathered the global crisis remarkably well,” and highlighting the “quasi fiscal stimulus through the national development bank” and “measures to channel liquidity to small and medium-sized banks facing stress” (Madrid 2012:6). In December 2012, President Dilma Rousseff again used the BNDES to implement emergency economic stimulus measures, including lowering the TJLP from 5.5 to 5 percent and extending for another year an emergency credit line of R$ 100 billion from the National Treasury to finance the purchase of capital goods (Biller 2012).
To complete this discussion, I report the four-point rebuttal from critics of the current management of the BNDES. First, if the root cause of Brazil’s high interest rates is the dominance of credit markets by oligopoly players, then we should recall that the BNDES itself is a monopolist, with the unfair advantage of access to a large—and growing—pool of below-market resources. Second, the close, cooperative relations of the BNDES with many important private banks and financial institutions may have solved the problem of the BNDES’ monopoly of long-term credit markets for these financial actors—but it has done so by pushing the costs of this cozy arrangement off onto ordinary Brazilian citizens, who continue to pay extraordinarily high real interest rates. For the largest firms with access to either BNDES or lower-cost international financing, Brazil’s high free market rates are not necessarily a problem, but for other borrowers they are. Overall--and notwithstanding the BNDES’ recent efforts to expand lending to micro, small, and medium firms--Brazil clearly has a socially regressive national financial policy framework. Third, the very large presence of state-owned investors, including the BNDES, in Brazil’s capital markets suggests that these markets also remain dependent. In 2011, public sector entities controlled 48 percent of all pension fund assets and 30 percent of mutual fund assets (EIU 2012:21-2). Finally, the BNDES’ critics emphasize their advocacy of a smaller BNDES refocused on its core competencies in infrastructure, heavy industry, innovation, and special purpose funds. The institution still will be available to be employed by the state for crisis management or to initiate new, strategic planning and projects. No one, they reiterate, is suggesting that the BNDES be abolished.
This section has described a sometimes acrimonious policy debate, whose major proponents are past and present government economic officials, over the way that the BNDES has pursued its mandates to employ Expertise to push the economy forward, while also ostensibly promoting private financial Markets. Views on the desirable role of the BNDES—and the larger new developmentalist model of which it is the most obvious manifestation—fall out along a partisan (but not terribly wide) divide. The intellectual debate remains unsettled—but meanwhile the center-left remains firmly in power in Brasília.
This paper began by observing that the principal differences between Brazil’s old ISI developmentalism of the 1950s through 1970s and the new developmentalism of the 1990s and after lay, first, in the ND’s greater enthusiasm toward participation in global markets and, second, in its basis in a radically different underlying Brazilian macroeconomy: one with credibly stable prices and a renewed attention to the evils of intense inequality. There exist two recognizable policy camps within the larger, centrist Brazilian economic policy consensus on capitalist developmentalism. The public bank trilemma is a useful framework for thinking about not only the BNDES, but the internal strains within Brazil’s ND more generally.
The rationale for a national development bank turns on the autonomous exercise of Expertise to identify public goods that the private financial sector will not provide voluntarily, from promoting a unified South American market to enabling implementation of a rapid counter-cyclical macroeconomic policy response to a global financial crisis. There is a potential conflict with the imperative of Democracy, most especially when that bank draws heavily on taxpayer funds which it lends out at subsidized rates to Brazil’s largest firms, operating in oligopoly markets. However, in the case of the BNDES, which has a deserved reputation for competence and integrity, Bank presidents have worked closely with government finance ministers to implement executive branch priorities. Disgruntled members of the public have worked through Brazil’s legislature and the press—and occasionally through the courts, as when non-governmental environmental organizations have sued the BNDES for its support of large hydroelectric dams. The focus on supporting “national champions” with taxpayer funds tends to reinforce income inequality. However, such public policy disagreements are a normal part of the democratic process—and are to be resolved through the political system.
Accusing the BNDES of being anti-democratic thus seems to miss the mark. For example, the plans for extensive infrastructure investments necessary to link the interior of Brazil with the industrialized Southeast and South, as well as with the rest of South America, are correctly identified as policy priorities of both the Cardoso and the Lula da Silva and Rousseff governments—although these investments have many critics within Brazil and internationally, particularly among environmentalists and supporters of the rights of indigenous peoples.10 In late 2012 the BNDES announced a multiyear loan of almost US$11 billion, its largest ever, to Norte Energia, a construction consortium created to build the controversial hydroelectric dam known as Belo Monte (to be the world’s third largest) on the Xingu River (Amazon Watch 2012). Another state bank, the Caixa Econômica Federal, will co-finance, as will the private investment bank, BTG Pactual--which has, in this case, been “crowded in” by BNDES leadership (Place 2013). Although the Belo Monte project’s opponents are not satisfied, the dam has passed through multiple environmental impact processes within Brazil and had its design modified as a result (see Hochstatler in this volume). There is thus not a problem of an inappropriately autonomous Bank escaping from legitimate political oversight. On the whole, Brazil has managed the Expertise-Democracy vector of the public bank regulatory trilemma well.
The Expertise-Market vector is more problematic, particularly since the mid-1990s end of astoundingly high inflation, long the plausible rationale for the BNDES’ dominance of financing for long-term investment in Brazil. Nearly two decades later, Brazil’s private financial sector, which benefits from on-lending, co-financing, and securities markets support coming from the national development bank, is comfortable with its relationship with the BNDES. But whether the BNDES’ dominance is good for financial market functioning (as opposed to the profits of financial institutions) is less clear. The BNDES’ role in public finance is also ambiguous: for example, journalists recently decried a large loan to the state electricity firm, Eletrobras, whose timing suggested that it would be used to pay stock dividends—including to both the BNDES and central government—thus serving as “creative accounting” to improve the look of public finances (Campos 2013). There is a clear bifurcation between those, mostly on the center-right, who would oblige the BNDES to rely on increasing shares of market financing, consequently shrinking its activities, and those of the center-left, including incumbent President Dilma Rousseff, who remain extremely proud of the BNDES’ accomplishments—and of its size per se.
Nonetheless, there are extensive areas of new developmentalist agreement across most partisan divides in contemporary Brazil, and weak economic and industrial growth rates in Brazil under the Rousseff government have sharpened the debate (see also Peter Kingstone’s chapter in this volume). Still, the differences that loom large from within the country appear modest from outside. No politician with a realistic opportunity to capture national policy influence wants to attack or attenuate private property rights: all agree that capitalism promotes growth. Yet neither are there influential voices raised that would radically dismantle the extensive apparatus of state credit and investment. Overall, the underlying national consensus on an activist state as the best option for achieving rapid growth is unlikely to falter, except in the presence of a major economic crisis—which is just the sort of situation that proponents of Brazil’s ND and a strong BNDES believe themselves uniquely well-equipped to handle. The implicit model of state-led capitalism looks likely to continue to guide Brazilian public policy for the foreseeable future.
Aghion, B. 1999. “Development Banking.” Journal of Development Economics, 58, pp. 83–100.
Allen, F. and D. Gale. 2000. Comparing Financial Systems. Cambridge, MA: MIT Press.
Amazon Watch. 2012. “Amazon: Financing Ignores Violations of Human Rights and Environmental Safeguards,” November 28. At: www.amazonwatch.org
Arida, Pérsio. 2005. Mecanismos Compulsórios e Mercados de Capitais: Propostas de Política Econômica, Texto para Discussão 8, Instituto de Estudos de Política Econômica, Casa das Garças, May. Preliminary version available at:
Armijo, Leslie Elliott. 1993. "Brazilian Politics and Patterns of Financial Regulation, 1950-1991." In S. Haggard, C. Lee, and S. Maxfield, eds., The Politics of Finance in Developing Countries. Ithaca, New York: Cornell University Press.
Armijo, Leslie Elliott. 1997. "Brazil: Business-Government Financial Relations in the Land of 'Super-Inflation.'" In R. Bingham and E. W. Hill, eds., Government and Business Finance: Global Perspectives on Economic Development. Newark: CUPR Press of Rutgers University.
Armijo, L.E. and P. Faucher. 2002. "‘We Have a Consensus’: Explaining Political Support for Market Reforms in Latin America." Latin American Politics and Society, 44:2, pp. 1-40.
Bacha, E. 2002. ‘Do Consenso de Washington ao Dissenso de Cambridge’ In A. C. Castro (ed.), Desenvolvimento em debate: novos rumos do desenvolivmento no mundo. Rio de Janeiro: BNDES.
Ban, Cornel. Forthcoming 2013. “Brazil’s Liberal Neo-Developmentalism: New Paradigm or Edited Orthodoxy?” Review of International Political Economy.
Banco Central do Brasil (BCB). 2009. Boletim Mensal. Brasília: BCB, December. At www.bcb.gov.br, accessed February 2013.
Banco Nacional de Desenvolvimento Econômico e Social (BNDES). 2002a. BNDES: Cinqüenta anos de desenvolvimento, June. At www.bndes.gov.br, accessed February 2013.
Banco Nacional de Desenvolvimento Econômico e Social (BNDES). 2002b. “Privatization in Brazil,” Powerpoint presentation, September. At www.bndes.gov.br, accessed March 2013.
Banco Nacional de Desenvolvimento Econômico e Social (BNDES). 2011. “Financial and Institutional Aspects,” Powerpoint presentation, June. At www.bndes.gov.br, accessed February 2013.
Banco Nacional de Desenvolvimento Econômico e Social (BNDES). 2012. Annual Report 2011. Rio de Janeiro: BNDES. At www.bndes.gov.br, accessed February 2013.
Banco Nacional de Desenvolvimento Econômico e Social (BNDES). 2013. “Relações com Investidores.” At www.bndes.gov.br, accessed February 2013.
2010. “BNDES and the Stock Exchange launch Efficient Carbon Index at COP 16,” Press Release, December 8. At www.bndes.gov.br.
Beck, Thorsten, Asli Demirgüç-Kunt and Ross Levine. 2000 . "A New Database on Financial Development and Structure." World Bank Economic Review 14, 597-605. Dataset update of May 2009, available from the World Bank.
Bernhard, William, J. Lawrence Broz, and William Roberts Clark, eds. 2003. The Political Economy of Monetary Institutions, Cambridge, MA: MIT Press.
Biller, David. 2012. “Rousseff’s Stimulus Measures to Spur Brazil Economy: Timeline.” Bloomberg, December 6.
Bonney, Richard, ed. 1995. Economic Systems and State Finance. Oxford: Oxford University Press.
Bresser-Pereira, Luiz Carlos. 2009. “From Old to New Developmentalism in Latin America.” In José Antonio Ocampo, ed., Handbook of Latin America Economics. New York: Oxford University Press.
Burges, Sean W. 2009. Brazilian Foreign Policy After the Cold War. Gainesville: University Press of Florida.
Campos, Eduardo. 2013. “BNDES empresta e Eletrobras paga dividendos.” Valor Econômico, July 10.
Canuto, Octaviano, Matheus Cavallari, and José Guilherme Reis. 2013. “The Brazilian Competitiveness Cliff.” Economic Premise, No. 105, Washington, D.C.: The World Bank.
Caprio, Gerard, Jonathan L. Riechter, Robert E. Litan, and Michael Pomerleano, eds. 2005. The Future of State-Owned Financial Institutions. WDC: Brookings.
Castelo, Rodrigo. 2009. “O novo desenvolvimentísmo e a decadência ideológica do pensamento econômico brasileiro.” Serviço Social & Sociedade 112, December, pp. 613-36.
Costas, Ruth. 2012. “Aos 60 anos, ‘BNDES do século 21’ atrai polêmicas e atençao internacional.” BBC Brasil, August 20.
Cretoiu, S.L.,L.L. Barakat, V.S. Nogueira, and L.M. Diniz. 2012. “Ranking FDC das Transnacionais Brasileiras 2012.” Caderno de Ideas CI 1215. Nova Lima, MG: Fundação Dom Cabral.
Doctor, Mahrukh. 2013. “Assessing the Changing Roles of the Brazilian Development Bank.” Revision of paper presented at Annual Meeting of the Latin American Studies Association, Washington, D.C., May 29-June 1.
Epstein, Gerald. 1992. “Political Economy and Comparative Central Banking.” Review of Radical Political Economics, 24:1, pp. 1-30.
Estado de São Paulo [ESP]. 2012. “Empresários brasileiros esperam ampliar negócios com a reeleição de Chávez.” October 8.
Fellet, João. 2012. “Com BNDES e negócios com politicos, Odebrecht ergue ‘império’ en Angola.” BBC Brasil, September 18.
Financial Times (FT).
2009. “Interview transcript: President Luiz Inácio Lula da Silva.” November 8.
Fishlow, Albert. 1973. “Some Reflections on Post-1964 Brazilian Economic Policy.” In Alfred Stepan, ed. Authoritarian Brazil: Origins, Policies, and Future. New Haven: Yale University Press.