The scale and pace of change within the World Bank during Robert McNamara’s presidency begs the question: how successful was the organization in promoting development? The scope of the organization’s operations, not to mention the inherent difficulty in evaluating the effectiveness of development interventions, makes it impossible to answer this question completely.1 In McNamara’s thirteen years as president, the Bank made over 2,500 loans for projects ranging from the construction of dams to the development of tourist industries. The organization also undertook many nonlending activities, from managing international research centers to coordinating consortia of aid donors. The dearth of contemporary analyses of Bank operations, as well as the inaccessibility of much of the organization’s reporting on its activities, makes reaching conclusions about the Bank’s record even more difficult.2
Nevertheless, available evidence indicates that the Bank’s record during the McNamara years was poor. The organization failed to achieve basic objectives in its three main areas: lending for specific projects, advising developing countries on policy matters, and attending to global economic affairs. With respect to the first function, Bank-financed projects were beset by delays and cost overruns, and anticipated rates of return largely failed to materialize. More problematically, during this period the Bank funded a number of coercive interventions, including projects that involved forcibly removing people from their homes. In its role as a policy advisor, the Bank struggled to influence many borrowing governments, particularly those that became less dependent on its capital as the 1970s progressed. The Bank’s efforts to help manage the global economy also left much to be desired, as the organization proved unwilling and unable to redress mounting imbalances in the global financial system as a debt crisis became imminent in the late 1970s.
Of course, not all these problems were entirely, or even primarily, the Bank’s fault. Many projects failed because borrowing governments did not implement them properly. Moreover, it would be strange if failure did not accompany the Bank’s entry into new types of lending. Nor was McNamara directly implicated in all of the Bank’s work. He did not supervise projects and could not singlehandedly prevent governments from borrowing beyond their means. Besides, many of these issues—including the funding of projects that had destructive human and environmental consequences—existed both before and after McNamara’s presidency. Yet it would be wrong to let McNamara entirely off the hook. The strategic and organizational changes he initiated contributed to many of these failures. Specifically, in prioritizing the Bank’s institutional growth, he diverted attention from ensuring that the quality of the organization’s lending remained high and that borrowing governments had the capacity to absorb more loans.
As we have seen, the Bank’s financing operations expanded both quantitatively and qualitatively under McNamara. The organization lent more money to more countries and for more types of projects than ever before.
Rural development was the centerpiece of the Bank’s expanded lending program. McNamara considered smallholder agriculture a key to reducing poverty in the developing world. In his 1973 Nairobi address, he declared that there was “no viable alternative to increasing the productivity of small-scale agriculture if any significant advance is to be made in solving the problems of absolute poverty” and argued that “without rapid progress in smallholder agriculture throughout the developing world there is little hope of achieving long-term stable economic growth.”3 McNamara also feared that failure to share the gains from the Green Revolution would lead to violence.4 To address these problems, he oversaw a tremendous expansion of the Bank’s rural development program. Whereas before his arrival the organization limited its agricultural lending to constructing irrigation works and providing modern equipment to commercial farmers, under McNamara the Bank began to fund the installation of small wells, livestock development, and countrywide rural development programs.
Central to the Bank’s rural development efforts were integrated rural development (IRD) projects, which sought to deliver “packages” of inputs to groups of small farmers in developing countries. The Bank extended credit to purchase modern seed, fertilizer, and equipment, constructed feeder roads to connect participating farms to market, established schools and clinics, and provided agricultural training,5 believing that such comprehensive interventions would greatly increase the productivity and incomes of target groups. The Bank expected one IRD project for cotton farmers in Brazil, for instance, to double yields within six years, while those in Colombia and the Philippines could raise the annual incomes of participating farmers by as much as 400 percent.6
Nevertheless, the Bank’s rural development projects encountered a host of problems. The organization’s PIDER project in Mexico, one of the largest Bank projects in the McNamara era, illustrated many of these issues. In 1972, Mexican president Luis Echeverría established the Programa de Inversiones para el Desarrollo Rural (Program for Rural Development Investments, PIDER) to coordinate the country’s rural development efforts and deliver services to small farmers in some of the poorest parts of the country.7 The following year, the government of Mexico sought Bank funding to assist with these efforts. As a condition for its support, Bank officials had Mexican authorities modify the program to place more emphasis on investments in agricultural production, rather than the provision of social services.8 Having secured this agreement, in 1975 the Bank made the first of three loans to the Mexican government. The organization’s initial $110 million was intended to cover roughly 40 percent of the expenses in 30 of the 45 targeted regions, with most of the funds going toward credit, irrigation, livestock development, fruit production and soil and water conservation. The loan would also provide for the construction of feeder roads, farmer training, and an extension of electricity to participating locales.9 The Bank expected that as a result of its efforts, the annual incomes of 750,000 people would double over eight years, unemployment would decrease, and, though not the main focus, literacy and health would improve.10
Difficulties immediately beset PIDER. Local agencies focused on delivering social services struggled to meet the Bank’s requirement that investments be geared toward increasing agricultural production. In addition, Bank and Mexican officials failed to implement delivery and oversight mechanisms to guarantee that farm inputs reached intended beneficiaries.11 As a result, an analyst who surveyed the project in the late 1970s found that a “large majority” of the targeted population had “received relatively few benefits.”12 Fewer than half the irrigation units functioned properly, and in some places 95 percent of targeted families failed to benefit from the project.13
Similar issues plagued the Bank’s other rural development projects. A few months before the Bank made its initial commitment to PIDER, it lent $50 million to the Mexican government to fund the construction of “small irrigation works, support services for rainfed agriculture, marketing facilities, feeder roads, potable water, sewerage, electricity, primary schools, health facilities and community centers” in the Papaloapan River valley in the eastern part of the country.14 One of the Bank’s first IRD projects, the Papaloapan project was intended “to expand food production, raise incomes and improve the quality of life of about one million people” in the region.15 As with PIDER, problems plagued the project. Bank evaluation reports noted that “insufficient budgetary allocations, limited capacity within the [local executing agency and] . . . lack and poor performance of contractors in the remote project area” prevented the project from producing expected results.16 Half the original components had to be dropped, construction was completed three years behind schedule, and the Bank ended up canceling 48 percent of the loan. The organization concluded in a subsequent evaluation that “the integrated approach proved impossible on such a large project area.”17
In Mexico and elsewhere, the Bank’s difficulty supervising rural development projects led to corruption and the diversion of resources away from intended beneficiaries. In Bangladesh, a large landowner was able to convince local authorities to construct a Bank-financed well on his land, rather than in the village where it was supposed to be located. The company in charge of supplying parts for the project also inflated the prices that it charged the Bank.18 The organization’s Upper Region Agricultural Development Project in Ghana also saw a significant deflection of benefits. The project’s objectives demonstrate the ambition of the Bank’s rural development approach. The Bank intended its $21 million loan, agreed to in 1976 and supplemented with funds from the governments of Ghana, the UK, and the Netherlands, to provide for, among other things, establishment of 90 service centers to provide credit, training, seeds, and fertilizers to small farmers, construction and improvement of 120 small dams and 100 existing dams, construction of 700 village wells, and creation of an adult literacy program.19 Yet rhetoric outpaced reality. The local agency charged with overseeing the project proved unprepared for this task, and two years after ground was broken, few of the support centers were up and running, with much of the material smuggled out of the country.20
Because they had to work through governments, Bank officials felt helpless, meekly noting that the situation was “quite depressing and discouraging.”21 However, an independent researcher who studied the project found that its main flaw was that the intended beneficiaries had no voice in the process.22 Many small farmers did not know how to obtain the benefits, and as a result better-informed large landowners were able to shape the project to suit their interests. One smallholder managed to get this message across to a Ghanaian official who toured targeted areas in the summer of 1979. “A few years back,” he said, “there was word that the World Bank was going to come and help poor farmers in this region.”
We were all happy and thought that at last somebody had heard our cries because as you know we have suffered for too long in this part of the country. . . . However, it has been some time now and we have not heard about them again. It appears, however, that in fact the World Bank people did come to Bolgatanga because we have been seeing many cars and white people and we are told it is World Bank people. This is what is worrying me because we have not received any help from them and in fact these days things are getting very hard for us poor people.23
The Bank also proved unable and unwilling to recognize the importance to its work of ethnic divisions in developing countries. The organization’s Mutara Agricultural Development Project in Rwanda provides one example. Begun in 1974, the Bank intended the project to improve the lives of approximately 10,000 poor farmers in the highland regions in the northern part of the country by resettling them on government-supervised farms, “paysannants,” where they would receive small plots of land, modern agricultural technologies, and training.24 An audit conducted just a few years after the Bank disbursed the loan, however, found that the funds were put to a much different use. Local officials spent most of the money on construction of the implementing agency’s headquarters. They canceled social services and irrigation components. And many of the land titles ended up in the hands of wealthy individuals who lived in other parts of the country.25 To René Lemarchand, a consultant hired by the Bank in 1978 to appraise a follow-up loan, the reason was clear. The Hutu-dominated government had no interest in helping the mostly Tutsi farmers who were supposed to benefit from the project; instead, Bank resources helped the Hutus consolidate power.26 Yet rather than reassess its role, the Bank censored Lemarchand’s report and in 1979 sunk another $8.8 million into the project.27 Not surprisingly, the project continued to fail. Bank evaluation reports later concluded “many pastoralists from the region had left for Uganda out of fear of the project and the intentions behind it.”28
Indeed, the Bank’s rural development projects sometimes forced peasants off their land. An independent investigation of the organization’s work in Nigeria, for instance, concluded that the Bank failed to make adequate preparations for the thousands of people whose lands were flooded as the result of the construction of two dams. Even farmers who were able to maintain their holdings fared poorly. Unable to afford the inputs intended to help them increase production, many had to sell their land at reduced prices.29 Likewise, a project to improve the productivity of small farmers in the São Francisco River valley in Brazil resulted in the relocation of many to smaller plots. In one instance the Bank gave so little thought to helping Filipino villagers displaced by new irrigation works that a staffer involved in the project admitted that “a whole municipality was going under water” as a result of the organization’s negligence.30
Even more troubling were the Bank’s land clearing and population relocation schemes. Between 1968 and 1978, the organization made seven loans totaling over $160 million to the Federal Land Development Authority (FELDA) in Malaysia for projects to move peasants to palm oil and rubber farms in recently cleared tropical rainforest in the central and southern parts of the country. On the surface, these initiatives appeared highly successful. Because most of the land previously lay fallow, agricultural productivity in the region had nowhere to go but up. Thus, the Bank later boasted that “FELDA is undoubtedly one of the most important and efficient settlement agencies in the world.”31 But the economic benefits of these projects came at a high price. According to American environmental writer Bruce Rich, the projects “ignored all provision for pollution control” and ended up eliminating about 6.5 percent of Malaysia’s rainforest.32 The Bank’s Northwest Region Development Program (Polonoroeste) project in Brazil, prepared near the end of McNamara’s tenure, also led to severe environmental damage. Begun in 1981, the project provided for construction of highways and roads in the region to serve its goal of attracting settlers for cultivation of coffee, cocoa, and other export crops. The scheme ended up causing more problems than it solved. Migrants overwhelmed the implementing agency, large parts of the forest burned, and thousands of indigenous people were displaced.33
The Indonesian transmigration project was one of the most harmful interventions in the McNamara era. In 1976, the Bank made the first of a series of loans to assist the Suharto regime’s relocation of half a million families from Java to some of the country’s less populated outer islands. Officials justified the project as a way both to relieve overcrowding on the country’s main island and to increase productivity in less populated areas.34 But the Bank should have been aware of the disastrous consequences this project would entail. Even before the organization made its loan, the Indonesian government had been forcibly removing native populations on the islands to make way for Javanese settlers. The Bank ignored concerns about the coercive nature of the project and made the loan, which like many projects was later found to have suffered from poorly run implementing agencies and lack of environmental monitoring.35 McNamara’s support for the project in the face of its many problems eventually became a source of “bitterness” for some members of the Bank’s staff.36
If the Bank’s population relocation schemes demonstrated the extent of the organization’s hubris under McNamara, more common problems related to its inability to ensure that projects were carried out properly. As much as the Bank grew under McNamara, over 90 percent of staff resided in Washington, meaning that borrowing governments implemented most parts of each project. Since the Bank’s rural development initiatives usually included numerous components, they involved a number of local agencies with little history of working together, and this led to complaints that countries lacked the administrative capacity to implement projects.37 One Bank staffer told a journalist in 1977 that theories about development “broke down” when put into practice. “We’ve over-committed,” another remarked. “It’s more difficult than we thought.”38
Even some of the Bank’s seemingly straightforward interventions ran into obstacles. As noted above, a central component of the rural development program was providing small farmers with instruction in modern agricultural techniques. Many of these efforts took the form of a “training and visit” system (T&V), developed in the early 1970s by Bank staffer Daniel Benor. T&V involved instruction of farmers by experts from Bank-supported national agricultural research institutions. Every two weeks field agents from central agencies would travel to villages to train a handful of “contact farmers,” who would then disseminate this knowledge to others in the area.39
India, which borrowed $200 million from the Bank for T&V projects between 1977 and 1982, was the largest recipient of T&V funds. Despite this substantial commitment, however, observers noted that T&V was a total failure. Few people showed up to the training sessions, and those who did were not aware that they were supposed to pass their knowledge on to others. Indian officials ended up devoting more resources to constructing their headquarters than supporting staff, and the quality of instruction was poor.40 Indian critics came to refer to T&V as “talk and vanish.”41
Like many Bank interventions, T&V projects were self-perpetuating, as officials viewed the “failure” of farmers to participate as a reason to enlarge the program. Thus, while the Bank intended T&V projects to be an inexpensive means of training small farmers, they ended up costing significant money, much of which was paid by developing countries. As one observer at the time noted, “the introduction of T&V has committed the Indian taxpayer to considerably enhanced recurrent expenditure on agricultural extension once the World Bank disappears from the scene.”42 While much of the blame rested with Indian officials who were more interested in securing benefits for themselves than the intended beneficiaries, the Bank was also at fault. In order to maintain an appearance that the projects were a success, the organization suppressed critiques of the program. The Bank “preached T&V like a religion,” one researcher found. “No questioning of the concept was permitted.”43
Problems also plagued the Bank’s agricultural extension efforts in Peru. In 1980, the organization initiated a T&V project in the northeast of the country to help rehabilitate local agriculture, which had suffered from neglect at the hands of the country’s military government. Soon after the program started, however, staff in Peru began to express frustration at their “inability to carry out the training method effectively,” a problem they attributed to “serious financial constraints, lack of relevant training, and poor conditions of service.”44 To one observer, this led to “marginalization” of peasants in the region. Unable to adapt their focus to meet the needs of small farmers, T&V staff provided the bulk of their assistance to large farmers.45 Indeed, the Bank itself later admitted that its agricultural extension efforts in Peru and elsewhere suffered from a “top-downward approach to farmers’ needs, meaning that [small] farmers have been left out of the decision-making process.”46
Although it would take some time for the full extent of the Bank’s rural development failures to come into focus, it was clear by the end of the 1970s that the organization’s antipoverty projects had generally failed to meet their goals. In 1978, the Bank reviewed nine of its initial “new style” rural development projects (defined as those in which over 50 percent of the intended beneficiaries were low-income groups) and found that three were “total failures” that had to be canceled before implementation, while in four others the majority of benefits had accrued to relatively well-off farmers.47 McNamara himself admitted to the Bank’s Board that “agricultural and rural development projects often did increase the skewedness of income distribution and that he did not know what to do about it.”48
The Bank’s rural initiatives were not a total failure. A number of projects led to significant productivity gains. According to one analyst, the organization’s first IRD effort in Paraguay “resulted in production value-added of approximately 150 percent in real terms over the project life.”49 Some regions served by PIDER saw spectacular achievements as well. In the state of Guanajuato, for instance, a researcher concluded that “maize production on newly irrigated plots increased by 245 percent, bean production by 660 percent, and chile production by 1,850 percent.”50 The Bank estimated that a few of its rural development projects in Africa had raised agricultural production sevenfold, while a project for the development of Indian dairy farming was found to have increased the quantity of milk brought to market by almost 250 percent in the span of a few years.51
Yet these cases appear to have been exceptional. In 1989, the organization reviewed eighty-two agricultural projects approved between 1975 and 1982 and rated almost 45 percent as having achieved an “unsatisfactory” economic rate of return.52 In over 40 percent of the irrigation projects that were analyzed, production “declined after the investment phase was completed,” while in a significant number of cases the physical components of the project had proven “less durable than expected.”53 Although the Bank experienced particular difficulty in sub-Saharan Africa, where it estimated that “21 of 36 area development projects failed between 1974 and 1986,” problems were found elsewhere as well.54 In Latin America and the Caribbean, for instance, 89 percent of rural development projects implemented between 1973 and 1977 “had disbursement shortfalls of 10 percent or more.”55 And 30 percent of the Bank’s agricultural and rural development projects implemented between 1980 and 1985, many of which were prepared during the latter stages of McNamara’s tenure, either failed to come to fruition or achieved rates of return below 10 percent, compared with 17–25 percent of projects in other sectors.56
Bank officials attributed these failures to developing countries themselves. They blamed the low success of its initial IRD projects in Nigeria, for instance, on government policies like agricultural price controls that discriminated against the rural sector.57 The Bank offered similar justifications for the failure of its rural development efforts in Tanzania, ignoring its role in supporting the failed collectivized farms, “ujamaa villages,” by faulting the government for paying insufficient attention to the role of private incentives in agricultural production.58
Yet the Bank was also to blame. Leaving aside the organization’s support for coercive initiatives such as the Indonesian transmigration program, under McNamara the Bank regularly overestimated the capacity of local agencies to implement what were often incredibly complex projects.59 In addition, as with T&V in India and the Mutara project in Rwanda, the Bank refused to correct course when presented with evidence of poor results. Finally, the organization failed to use its power to try to overcome the problems that hamstrung its projects. This was particularly true with land reform. The Bank now acknowledges that unequal tenure systems hindered many of its rural development initiatives.60 Yet under McNamara, the Bank repeatedly abandoned its commitments to encouraging its borrowers to adopt land reforms and continued to direct funds to large commercial agriculture.61 For instance, at the same time that Bank management and staff expressed frustration at the slow pace of land reform and other distributional efforts in Brazil, the organization was directing the bulk of its agriculture lending to livestock development projects that benefited large landowners.62
In addition to rural development, McNamara expanded the Bank’s urban development program. Sites and services (SS) projects were the focus of many of these efforts. Like IRD, SS were comprehensive interventions. The projects offered poor city dwellers the opportunity to purchase or rent housing in residential developments that included basic utility and social services. Bank staff also expected the construction of these developments to generate short-term employment.63 The Bank expected its SS to be a cost-effective alternative to the heavily subsidized public housing units that many developing countries had struggled to maintain over the previous years.64 As a Bank policy paper put it, these projects would “limit the burden on public authorities,” thus ensuring that they were “replicable on a wide scale.”65
SS initially appeared to be a success. An early evaluation conducted by two Bank researchers of four of the organization’s initial SS projects found that they led to increased production of low-cost housing, affordability for low-income groups, and the generation of “substantial amounts of employment and income” from rents, taxes, and construction.66 Housing units in one of the Bank’s initial projects in Zambia, for instance, “cost less than one-fifth as much as the least expensive government-subsidized housing,” while those in El Salvador averaged “less than half as much as the cheapest conventional house.”67
A closer look at the Bank’s record during the McNamara era reveals a more complex picture. As with rural development, in a number of cases the organization’s urban initiatives exacerbated inequality in targeted regions. Independent researchers investigating the Bank’s 1977 SS project in Madras, for instance, found that by increasing the value of land in the area, it priced many poor families out of their homes.68 The Bank’s initial SS project in Dar es Salaam suffered a similar fate. Staff estimated that the project would benefit 160,000 low-income residents of the city through the provision of 10,600 newly serviced plots, improvements of existing squatter settlements, the construction of community facilities, and the establishment of a housing bank that would provide credit to help low-income groups purchase or rent the units. Despite assurances to the contrary, the housing bank ended up providing loans to only those individuals who could guarantee that they would repay, meaning that the project ended up excluding most people in the area.69 A Bank evaluation conducted in 1984 found that fewer than half the units the government reported as being completed were actually finished and occupied.70
Problems also beleaguered the Bank’s efforts in Indonesia, the largest recipient of urban development loans during the McNamara era. An independent analyst hired by the Bank to appraise its Kampung Improvement Program in Jakarta in 1978 found that the project’s “top down” nature limited its effectiveness. The decision-making process bypassed intended beneficiaries, and most people were uninterested in maintaining what they viewed as “an imposed package.”71 Moreover, the program also raised the price of land and property taxes in the area, forcing many residents to leave for settlements on the outskirts of the city.72
Similar issues characterized some of the Bank’s urban development efforts in the Philippines. In 1976, the organization made a loan to the government for an SS and slum-upgrading project in Manila’s Tondo district. The project’s proposal provides a sense of the large number of components in SS projects. Among other things, the loan was to fund water supply, sewage, and road repair, establishment of a line of credit for housing materials, the creation of educational facilities, community centers, and a health center, the provision of mixed commercial/residential sites, a new vocational and skills training program, the construction of traffic signals, and equipment for traffic police.73
The reality was much different. According to one report, 4,500 squatter families were removed from their homes to make room for the development, and just 30 percent of the people in the area were able to afford the new units.74 This occurred after the Marcos regime had already evicted 60,000 squatters from another part of Manila to “beautify” the city for the 1976 Bank-IMF annual meetings.75
Even where SS did produce positive results, the projects tended to require large subsidies from local governments. Bank researchers who surveyed the organization’s initial SS projects concluded that their relative success had come at great cost. Without subsidized interest rates on home loans, for instance, many poor city dwellers would have been unable to afford to stay in the new developments.76 Thus, while the projects had provided shelter and services to many, “the goal of large-scale replicability that is so much the object of the sites and services paradigm” went unmet.77 These high costs meant that local governments also struggled to maintain existing developments.78
Although rural and urban development projects were the main parts of the Bank’s poverty-oriented lending program, they constituted a fraction of the organization’s lending during McNamara’s presidency, as indicated below.
How successful, then, were the Bank’s projects as a whole during this period? Available evidence paints a depressing picture. An analysis by Bank researchers of 1,015 projects carried out over the organization’s history found that the gap between staff estimates of economic rates of return versus rates at the time the project was actually completed grew significantly under McNamara. The researchers hypothesized that this was primarily because staff appraisal reports overestimated future gains, but the data also show that even when adjusting for this, rates of return on Bank projects remained below their historical average.79 Although the fact that the Bank expanded its lending to riskier areas may explain some of the decline, circumstantial evidence shows that McNamara’s drive to expand lending entailed reduction in loan quality. Projects were frequently delayed and went over budget.80 And, in the mid- to late 1970s, Bank staff began to complain about management’s emphasis on quantity rather than quality. “The pressures to carry out lending programs, and to bring projects to the Board on schedule, have an undue influence on staff decisions,” a 1977 internal Bank report noted. “In effect, the scales are weighted in favor of doing things on time, even when quality considerations should prevail.”81 As one Bank official later explained, “the main object was to go out and make loans. And many of the projects just folded.”82
Table 4. Distribution of IBRD and IDA Commitments by Sector, 1968–1981
Sector |
Percentage of total |
Agriculture and rural development |
28.3 |
Development finance companies |
8.6 |
Education |
4.7 |
Energy |
16.8 |
Industry |
7.1 |
Nonproject |
5.3 |
Population, health, and nutrition |
0.7 |
Small-scale enterprises |
1.3 |
Telecommunications |
2.5 |
Transportation |
16.3 |
Urbanization |
2.6 |
Water supply and sewerage |
5.2 |
Technical assistance and tourism |
0.5 |
Total |
100 |
Source: IBRD/IDA, Annual Report, 1981, 12–13.
Figure 7. Appraised and reestimated rates of economic return for selected Bank projects, 1961–1980. Source: Gerhard Pohl and Dubravko Mihaljek, “Project Evaluation and Uncertainty in Practice: A Statistical Analysis of Rate-of-Return Divergences of 1,015 World Bank Projects,” World Bank Economic Review 6, 2 (May 1992): 274.
Since its beginning, the Bank supplemented its lending operations by seeking to influence the economic policies of its borrowers, usually under the rationale that a sound policy environment was vital to the success of its projects. The organization helped draft national development plans and brought government officials to its headquarters for training.83 Occasionally, the Bank also included conditions in its loans specifying reforms the borrower was to adopt upon receiving funds. For example, Bank loans for hydroelectric projects typically contained clauses stating the manner in which the local implementing agency was to determine electrical rates, although the organization tended to leave macroeconomic policy advising to the IMF.84
As we have seen, McNamara was determined to increase the Bank’s ability to influence the behavior of its borrowers. During his first years in office, he increased the regularity and comprehensiveness of the Bank’s missions to developing countries, opened Bank offices in some borrowing countries, and conducted his own dialogues with developing country leaders. He soon found that the Bank’s power as a policy adviser was limited. Increasingly able to tap other funding sources, many developing countries ignored the Bank’s advice. Compounding these difficulties was McNamara’s determination to increase Bank lending. Because developing countries knew that the organization was driven to meet lending targets, threats to cut off funding, which historically had been an important source of leverage for the Bank, rang hollow.
The limits of the Bank’s influence were evident in Brazil, the site of McNamara’s most intensive effort to encourage poverty-oriented economic reforms. Brazil, one of the Bank’s largest borrowers, was a classic example of the problems of growth-centric development strategies. The economy had grown impressively in the postwar decades, yet significant poverty and inequality remained. As such, McNamara was pleased when, soon after he took over the Bank, members of the military government running the country expressed their desire to improve conditions for Brazil’s poor by, among other things, devoting more resources to the impoverished northeast part of the country.85
The slow progress of these efforts frustrated McNamara, however, and in 1970 he told a Brazilian official that he was “disturbed by the fact that Brazil was not doing very much for better income distribution or for employment.”86 Unhappy with the criticism, the government rejected Bank proposals to establish an office in the country.87 Bank managers encouraged McNamara to cut off lending to Brazil, but he wavered on the grounds that this would do little good. The country is “so big,” he stated. “They are going to tell us to go to hell.”88
Relations between Brazil and the Bank remained strained for the rest of the decade. In his speech to UNCTAD in 1972, McNamara pointed to Brazil’s poverty as an example of the ways growth had not reduced inequality or poverty in developing countries, which prompted a sharp rebuke from Brazilian Finance Minister Antônio Delfim Neto.89 The government’s repression, in turn, disturbed some Bank officials. In a 1973 memorandum to McNamara, Hollis Chenery, the organization’s chief economist, argued that the “brunt of our influence in Brazil at the present time should be directed at social justice.”90 McNamara did not directly address Chenery’s concern. Instead, a few days later he directed the head of the Bank’s Latin America Department to increase funding of poverty-oriented projects in the country.91
The Brazilian government, perhaps aware that the Bank would be unwilling to stop lending, informed Bank staff that it “welcomed” the organization’s antipoverty loans.92 Despite this rhetoric, the government was reluctant to work with the Bank on these projects, and by 1975 McNamara had had enough. In response to an internal Bank report that “the Brazilian government’s commitment to large-scale programs for improving small farm productivity [is] lacking,” he told the Bank’s head of operations that, “if the Brazilian government had no program for increasing the productivity and the level of income of large numbers of its poor,” the Bank should not make new loans to the country.93
But by that point the Bank was so overextended in Brazil that it could not stop lending. Should the Bank cut off Brazil, the country would start paying more money back to the Bank in the form of repayments on old loans than it was receiving from the organization. The possibility that a negative net transfer would jeopardize McNamara’s calls for rich countries to increase their funding of the Bank convinced McNamara to continue lending to Brazil.94 McNamara stopped criticizing the government, and in the early 1980s Bank officials admitted that the organization had “little policy impact” in Brazil.95
In Brazil and elsewhere, governments became less willing to listen to the Bank not just because they knew that lending cutoffs were unlikely but also because they were increasingly able to tap alternative sources of capital. This was particularly true in oil-exporting developing countries. For example, although the Bank’s initial dealings with Iraq were cordial thanks to McNamara’s insistence that the country receive loans over the objections of the Nixon administration, which was upset at lingering disputes from the nationalization of the Iraq Petroleum Company, the relationship turned hostile within a few years. The Iraqi government took issue with certain conditions to the Bank’s assistance, such as a requirement on rate pricing in a telecommunications project, and the country stopped borrowing from the Bank in 1973.96 Similar issues emerged in Iran. While Bank lending to Iran surged in McNamara’s first years, Iranian authorities frequently complained about the advice that came with these funds. The Shah was especially upset that Bank loans for education were allocated to secondary and vocational schools rather than universities, and he bristled at what he perceived as the organization’s critique of the government’s agricultural policies.97 In a meeting with McNamara in 1973, he voiced his objections to what he considered were “evidences of Bank leverage to force policy changes in Iran.” McNamara reassured him that the Bank’s only objective was “to assist him in the dramatic revolution which he [had] underway.”98 Nevertheless, tensions remained high. Although McNamara officially supported the Shah’s plan to establish a concessional lending agency to help oil-importing developing countries cope with the price increase brought about by the first oil crisis, he was never sold on the proposal. For its part, the government of Iran essentially cut off ties with the Bank once the plan fell through. Iran refused to contribute to the third window that McNamara created to augment the Bank’s lending in response to the oil crisis, and in 1976 Iran terminated the Bank’s technical assistance operations in the country.99
The Bank struggled to influence policy in many non-oil exporting developing countries, as well. Efforts to reform the Malaysian transportation sector were fraught with problems, and throughout the 1970s the Bank struggled to convince Colombian authorities to reform the country’s tax system and reduce import restrictions.100 The Bank tried to get Guatemala’s rightwing government to increase social spending, but the government’s aversion to such measures as well as its access to capital from the U.S. government and commercial banks undercut these efforts.101 And, though McNamara was pleased by Indian prime minister Indira Gandhi’s efforts to “effectively address its population problem,” which included forced sterilizations, and was largely uncritical of her suspension of the Indian Constitution in 1977, the organization was frustrated by the country’s reluctance to adopt Bank-prescribed agricultural pricing reforms—a problem that was partially attributable to its desire to lend.102 As one Bank official explained at the time, since “lending levels are sacrosanct targets . . . we may have locked ourselves in a position that impedes leverage.”103
The Bank had more success in Indonesia. As noted earlier, McNamara came to the Bank determined to accelerate the country’s reintegration into the organization. He had the Bank establish an office in Jakarta and met with Suharto soon after assuming the Bank presidency. McNamara was particularly keen on having the Bank advise Indonesian officials on economic policy matters. For instance, in 1968 he told his aides that the Bank’s “technical assistance,” not its funding, was the most important service it could offer the country.104 The Bank’s early efforts focused on helping the government settle its debts with external creditors. Later, the Bank had limited success in encouraging officials to adopt policies aimed at expanding food production for both domestic consumption and export, reducing the power of state-owned enterprises, and tackling corruption.105 Bank staff was also involved in helping to draft Suharto’s third five-year development plan in the late 1970s.106
Although the specific advice the Bank dispensed varied by country, its basic approach remained focused on promoting stability, growth, and international economic integration.107 It thus encouraged governments to remove barriers to foreign investment, relax trade restrictions, adopt anti-inflationary monetary policies, and reduce public expenditures.
The Bank’s relationship with the government of the Philippines during the 1970s illustrates the basic thrust of the organization’s policy advice.108 Like many developing countries, the Philippines maintained high levels of protectionism to foster domestic industry. Ferdinand Marcos, elected president of the country in 1966, was eager to attract foreign capital to the Philippines, however, and in his first years proposed legislation that offered multinational companies various incentives to invest in the country.109 The Bank supported these moves and was thus disappointed when local industrialists and leftists were able to defeat some of the measures.110 When Marcos declared martial law in 1972, the Bank threw its support behind the president, viewing the suspension of democracy as vital to the Philippines’ economic prospects and increasing its lending to the country.111 The Bank also encouraged Marcos to pursue a development strategy centered on increasing the country’s export of labor-intensive manufactured goods, such as textiles.112 Marcos adopted some of these measures, and exports began to rise. Nevertheless, Bank officials expressed frustration at the fact that some protectionist measures remained in place. Thus, in 1977 an internal Bank report argued that “tariff reform is becoming increasingly urgent since high effective protection rates appear to be sheltering various inefficient industries.”113 The absence of such reform led organization officials two years later to argue for a total “restructuring of the economy itself.”114 In other words, despite the fact that the government had adopted many recommended measures, Bank officials remained frustrated by the pace of reform. As we will see, the belief that organization needed to do more to promote comprehensive policy change in the Philippines and elsewhere contributed to McNamara’s decision to begin making loans that were conditioned on borrowing governments adopting Bank-prescribed reforms.
The Bank’s project lending and policy advising efforts took place amidst the backdrop of a dramatic increase in the indebtedness of developing countries. At the beginning of the 1970s, the total external debt of the world’s developing countries was about $69 billion, which constituted approximately 10 percent of their combined incomes. By 1980, these figures stood at $494 billion and 20 percent. Populous middle-income countries like Brazil, whose external debt skyrocketed from $6 billion in 1970 to $72 billion in 1980, and Mexico, which despite being an oil producer saw its obligations grow from $7 billion to $57 billion during this time, were the largest debtors. Much of this debt came in the form of variable interest rate loans from private creditors. Low-income developing countries, whose debt was held mainly by official lenders like the Bank, joined the party as well. Kenya, for example, saw its external debt jump from $478 million in 1970 to over $3 billion by the end of the decade.115 Although some were worried about these developments, most did not fully understand the consequences. Developing countries needed to finance their oil imports somehow, and high levels of inflation and low interest rates made it relatively inexpensive to borrow. Meanwhile, private creditors were largely unconcerned about lending to sovereigns. Although developing countries’ debts had risen significantly in the 1960s, sovereign defaults were rare. As Citigroup chair Walter Wriston famously quipped, “countries don’t go out of business.”116
Figure 8. External debt of developing countries, 1970–1982. Source: World Bank, International Debt Statistics.
McNamara demonstrated concern about sovereign debt issues at the beginning of his presidency. In discussions with aides in 1969, he noted that projections indicated that there would be “serious debt servicing and credit-worthiness problems for the future” in many developing countries. He viewed this as indicative of an “increased need for funds of the developing countries on soft terms.”117 As McNamara explained to the finance minister of Malawi in 1970, “the emergence of major international indebtedness problems in recent years has shown that, without the full awareness of the governments concerned, private debt has built up in a number of cases to magnitudes which have presented governments with serious problems for the management of their balance of payments.”118
McNamara viewed debt not as an obstacle to be overcome but, rather, as something that needed to be managed. Specifically, he felt that the lack of information about the debt servicing capacities of developing countries—at the time no international organization systematically collected this information—distorted the market. With lenders unable to assess sovereign creditworthiness, usually measured as the ratio of a country’s interest and principle repayments to its export earnings, they might make unwise investment decisions. Poor information would also lead borrowing governments to over or under extend themselves.
As such, in McNamara’s first years he attempted to improve the way the Bank collected and disseminated information about the creditworthiness of its borrowers. Early in his tenure he complained that “the debt service problems of individual debtor countries had still not been adequately examined within the Bank” and shortly thereafter directed staff to create an “early warning system on external indebtedness.”119 This resulted in the creation in 1970 of the World Debt Tables, an annual report that published detailed information on developing countries’ external debts.120 McNamara also sought to increase the Bank’s own lending to help countries manage their debt repayments. To this effect, McNamara was more interested in maintaining the Bank’s ability to help countries roll over their debts than in reducing those debts themselves. In the early 1970s, he repeatedly rebuffed calls from developing countries to reschedule their obligations to the Bank by claiming that the organization’s “difficult position as a capital market borrower” precluded it from doing so. The Bank “did not have the flexibility of a commercial bank nor did it have the taxing capacity of a government,” he explained, and rescheduling Bank loans would have “potentially destructive effects” on the organization’s creditworthiness.121
The Bank’s ambivalence toward sovereign debt issues remained even as the external debts of developing countries began to explode in the wake of the 1973–74 oil crisis. McNamara was aware that commercial loans often carried terms less favorable than those provided by public creditors like the Bank. Yet he thought that high levels of inflation made servicing this debt manageable. He and other Bank officials also worried about the effects that would ensue if the organization were to raise red flags. For instance, while the Bank began warning the Mexican government about the country’s growing debt-service ratio in the mid-1970s—in 1975 McNamara told Mexican officials the country “might encounter serious debt management problems” in the near future—the Bank greatly expanded lending to the country during the period in part because it wanted to avoid informing other lenders about the precariousness of Mexico’s position.122 “The Bank profile in Mexico should not be lowered precipitately,” an internal report explained. “Probably the most important [reason is] . . . the impact on Mexico’s creditworthiness of a sharp withdrawal by the Bank; Mexico had stretched its credit in the world’s capital markets to a point where any explicit loss of confidence on the part of an agency like the World Bank would have disastrous consequences on the country’s creditworthiness.”123
McNamara spelled out his thinking on debt issues in his 1976 address to the Bank-IMF annual meetings. He acknowledged that developing country debt had risen significantly over the previous years but insisted that such “increases are not large when adjusted for inflation, and they appear still smaller when account is taken of the growth in exports.” As a result, McNamara predicted that “even in 1980, the debt service burden of these nations, in relation to their exports, is likely to be about the same as in 1973.”124 Rather than debt relief, more lending was needed to encourage “productive investment.”125
But the Bank’s analyses of developing country debt, which some in the financial community looked to when making investment decisions, were flawed, relying on overly optimistic growth projections to argue that external debts of developing countries were manageable.126 As one Bank official noted in 1977, Bank reports sought to paint a “reassuring picture of the debt situation” that would serve as “a useful antidote to the loose talk about a debt crisis.”127
The Bank thus downplayed concerns about the possibility of a debt crisis. While some Bank officials acknowledged that a growing number of developing countries were having difficulty servicing their debts, they dismissed such problems as “the concern of the financial community.” McNamara agreed. “What was required,” he told Bank management in 1977, was not a reduction of lending but “continued rollover and expansion.”128 McNamara reiterated these points at that year’s Bank-IMF annual meetings. After declaring that “the private capital markets [had] responded well to the emergency needs of the developing countries for credit in the wake of the oil crisis,” he argued that the belief that “this dramatic growth in external borrowing—particularly the borrowing from commercial banks—is unsustainable, and that if it is allowed to continue there will eventually be a generalized debt crisis,” were misplaced. “Such a crisis was not inevitable,” he insisted. “The debt problem is indeed manageable.”129
Nevertheless, concerns about the stability of the international financial system began to mount. By 1978, Bank officials were warning about significant “riskiness in [the Bank’s] loan portfolio,” and many countries began asking the Bank to reschedule their obligations to the organization.130 Members of the Bank’s Board worried about “the heavy dependence of middle income countries on commercial credit . . . since a sharp reduction in the availability of financing could trigger the kind of debt crisis that all wanted to avoid.”131 And Ernest Stern, the Bank’s Vice President of Operations, noted that the organization’s country analyses had regularly overestimated developing country creditworthiness.132
World events drove home the precariousness of the situation. The destabilization caused by the Shah of Iran’s downfall in January 1979 touched off a second round of oil price increases. That fall, U.S. Federal Reserve chair Paul Volcker began to raise U.S. interest rates in order to rein in inflation in the country. These developments had a devastating impact on the economies of developing countries. Oil-importing countries were suddenly forced to devote more of their resources to energy, and the real value of the dollar-denominated, variable-interest rate loans that many of these countries had taken out over the previous years skyrocketed.133
By the end of 1979, concerns emerged in the Bank about a wave of sovereign defaults. An October memorandum drew attention to the likelihood that, if private banks failed to increase their lending to developing countries, those nations would face the possibility of default.134 In November, McNamara confided to his aides that he was “uneasy about delinquent repayments by Bank borrowers.”135 And the following month, staff reported concerns about the ability of developing countries to continue servicing their external debts.136
McNamara made every effort to keep these projections from becoming public. In an interview with the Times of London in 1980, he argued that the situation remained manageable. “Never underestimate the complexity and sophistication of the international financial system, its ability to cope. The system as a whole hasn’t yet reached its limit.”137 McNamara also scrubbed the 1980 World Development Report clean of findings that might trigger concerns about global financial stability. Whereas the draft version predicted that real interest rates would stabilize at around 3 percent in the near future, enough to provoke significant debt service problems in many countries, McNamara insisted that the final report list this figure as 1 percent.138
Nevertheless, concerns within the Bank continued to grow. In January 1981, Bank Treasurer Eugene Rotberg reported that, unlike in 1974, commercial banks who were absorbing OPEC funds did not appear willing to expand their lending to developing countries. McNamara responded that “the Bank ought to raise red flags now.”139 However—as he had done ten years earlier when presented with evidence of impending disaster in Vietnam—McNamara refused to make his concerns known. The 1981 World Development Report insisted that the organization “expects the borrowing needs of the middle income countries to be met largely by the commercial banking system.”140
McNamara was wrong, as private banks curtailed their sovereign lending in the early 1980s. Unable to rollover its debts, in 1982 the government of Mexico declared that it was defaulting on its external debt. A number of other developing countries soon followed suit. With private banks overextended abroad, the inability of developing nations to service their loans precipitated a global debt crisis that, at the time, posed the greatest threat to the stability of the world economy since the Great Depression.141
Accounts of the origins of international debt crisis of the 1980s tend to downplay the Bank’s role. The tale is usually explained as a classic case of boom and bust, with private bankers and developing country officials sharing blame.142 When observers have assessed the Bank’s record, criticism is usually muted. As one observer wrote, “it would be unfair to fault the Bank for failing to foresee the debt crisis. Few predicted the gravity of the global recession of the early 1980s or foresaw soaring interest rates and the sudden retreat of the commercial banks.”143
But it was more than a coincidence that the conditions that led to the debt crisis coincided with McNamara’s presidency of the Bank. Although the organization was not alone in failing to predict the crisis, it shares some of the blame. This is not just because of the poor performance of its own portfolio but also because it regularly overstated the creditworthiness of developing countries and, in so doing, sent incorrect signals to a market that often relied on its forecasts. Throughout his tenure, McNamara remained focused on increasing the Bank’s lending and, when he realized that many countries would have trouble servicing their debts, decided to keep this knowledge secret lest it precipitate a crisis of confidence. In other words, under McNamara the Bank was unwilling and unable to recognize a bubble and take action. Although such shortsightedness is common in financial history, it bears striking similarities to his refusal to speak out or change course when, as U.S. defense secretary a decade earlier, he realized that the war in Vietnam was unwinnable.