A Space for Science - The Development of the Scientific Community in Brazil

Simon Schwartzman

The Pennsylvania State University Press, 1991


Chapter 9
THE GREAT LEAP FORWARD

Science and Tecnology for Economic Development

The 1968 Reform of Higher Education

The New Graduate Programs

High Technology Institutions

Big Science and High Technology

Notes

Science and Tecnology for Economic Development

The involvement of Brazil's main investment bank - the government-owned Banco Nacional de Desenvolvimento Econômico - in the field of science and technology is the most important feature of the new period. For the first time in Brazil's history there was a concerted attempt to put science and technology at the service of economic development, through the investment of substantial resources. In 1964 the bank established a program for technological development known as the Fundo Nacional de Tecnologia (National Fund for Technology), which in its first ten years provided about $100 million for research and graduate training in engineering, the hard sciences, and related fields.

The Fundo Nacional was established with the hope that economic incentives would lead private investors to develop their technologies instead of importing them from abroad, and it soon began to support selected teaching and research programs. With the fund's support, the Universidade de São Paulo acquired its electrostatic accelerator, Pelletron, in 1971; a consortium of institutions began to develop a Brazilian minicomputer; the Centro Tecnológico da Aeronáutica obtained support for its work on airplane engines; the Instituto Militar de Engenharia initiated graduate programs in several branches of engineering and chemistry; and the newly created Universidade de Campinas received substantial grants for a variety of projects. The creation of a complex system of graduate courses in engineering at the Universidade Federal do Rio de Janeiro which came to be known by the acronym COPPE, was an important initiative. The activities of the fund were later transferred to a new, specialized agency, the Financiadora de Estudos e Projetos, which works as an investment bank for technological and feasibility studies and which administers a national fund for science and technology that became part of the federal budget, replacing the development bank's fund. In 1975 the old Conselho Nacional de Pesquisas was transformed into a new and much larger Conselho Nacional de Desenvolvimento Científico e Tecnológico, now under the Ministry of Planning.

The ideological roots of this program can be traced to a combination of two seemingly opposing trends. First, there were the ideas about economic and technological dependency and the need for science planning that were so central to the Brazilian scientists' quest for a new role. José Pelúcio Ferreira - the economist who organized the Fundo de Desenvolvimento, FINEP, and later became vice-president of the national research council - acknowledges the role of physicist José Leite Lopes in shaping his views. Adler reported:
[Pelúcio Ferreira] said that by the middle of the 1960s, although physicists and economists developed their ideas separately they had converged to create an awareness of science and technology dependency. The economists' emphasis on the linkages between technology and economic development was particularly important. Pelúcio has acknowledged that both ídéias cepalinas and the ISEB [Instituto Superior de Estudos Brasileiros, a think tank in io de Janeiro closed by the military in 1964] had a considerable effect on his subsequent work in the science and technology field.(1)
The other trend was the military government's nationalist ambitions, which began to take shape after the 1960s and peaked in the mid-1970s. In the 1960s, South American military regimes were known for their ideological and doctrinaire approximation of the United States, for their economic liberalism, and supposedly for their concern with reducing the role of the state in all spheres of activity - except of course for control of political participation and expression. In that sense, they were opposed to the trend represented by CEPAL or such scientists as Leite Lopes, who supported increasing state planning and intervention to redress the effects of dependency. The most extreme example is probably Chile, which became a test case for Chicago-style economics orthodoxy. Economic liberalism was also central to the first Brazilian military regime headed by General Castelo Branco after 1964. The orthodox economic policies of those years were effective in controlling inflation, increasing the government's tax base, modernizing the government's instruments of economic policymaking, and attracting foreign capital.

Economic liberalism was followed shortly by a parallel, and eventually opposite, tendency toward growth and strengthening of the public sector. In the 1970s a division of labor began to shape up. The liberal economists would continue to run Brazil's economy; political participation would remain under control; and the military, the engineers, and eventually the scientists would speed up their long-term projects through the expansion of the state. The list of projects was impressive: the nuclear program, the large hydroelectric power plants, several ambitious road and railroad construction projects, the expansion of the frontier into the Amazon. The consequences, both positive and negative, are still being evaluated. On the positive side were the modernization of the country's industrial base and the effective growth of national income; on the negative side were the excessive levels of income concentration, the destruction of the environment, the emptying of the countryside, the deterioration of cities, the swelling of the state, the wastefulness of unfinished and overambitious projects, and the economic indebtedness that led to the economic crisis of the 1980s.(2)

What brought intellectuals and scientists on the left together with the military on the right was nationalism and the shared belief in the powers of science and technology. It was not an easy coexistence. Many intellectuals lost their academic positions and were forced into exile. The building of new scientific and research institutions, and the participation of talented scientists who did not submit easily to military authoritarianism, required constant and difficult negotiations with security officers, which were conducted, not always successfully, under the authority of the minister of planning, João Paulo dos Reis Velloso, or by people like Zeferino Vaz. There was also a clear contradiction between the economic policies being pursued by the Ministry of Finance, oriented toward the internationalization of the economy and the introduction of foreign capital and technology, and the projects for technological self-reliance carried on under the Ministry of Planning. Because of this split, investments in technology were seldom based on broader macroeconomic considerations, while economic policies never took the development of national technologies into account.

The entrance of agencies of economic development and planning into the field of science and graduate education intensified the historical tendency to favor applied technology over basic science, a tendency that was dramatized by the change in the name of the Conselho Nacional de Pesquisas, which was far from just nominal. Science-supporting agencies like FINEP and the science council gradually became swollen bureaucracies of hundreds and eventually thousands of functionaries, and scientists had to negotiate with economists and administrators every two or three years, on a project-by-project basis, for renewal of their grants. A two-year national plan for science and technology was promulgated in 1973 and again in 1975, with projected expenditures ranging from $323 million to $824 million a year.(3) These plans were little more than collections of anticipated expenses by sector, most of which - 65 percent for the period 1973-75 - were completely outside the sphere of influence of the planning authorities in such agencies as FINEP, the national research council, or the Banco Nacional de Desenvolvimento. The expectation for the period 1976-77 was that the expenditures under these agencies would be increased slightly. Between 21 percent and 27 percent of the expenditures were to be allocated to graduate training, fellowships, and "scientific development" in general; between 20 and 29 percent to industrial technology; between 11 percent and 15 percent to agricultural research; and between 5 percent and 10 percent to atomic energy projects. There is no known evaluation of how the plan was implemented or how the expenditures were made. The third plan, for the years 1978-79 and already under President João Batista de Figueiredo, was just a broad statement of purposes without any figures attached. At that time João Velloso and José Pelúcio Ferreira had already left their posts, and Delfim Neto, the former finance minister, was running the economy from the Ministry of Planning. For the first time, macroeconomic and technological policies came under the same authority, which gave the latter very low priority.

The 1968 Reform of Higher Education

In 1968 new legislation aimed at a profound reorganization of higher education was introduced. The number of applications was increasing, and it was impossible to keep the system small. This was also a time of intense street demonstrations against the military government, which ushered in several years of student-based urban guerrilla activity and violent government repression, including tight control over political activities at the universities. The 1968-78 decade was also a period of rapid economic growth, with new jobs being created and social mobility intensifying. When combined, these factors led to a complete revamping of the country's higher education, although not necessarily in the directions prescribed by the 1968 legislation.

The 1968 reform adopted the ideas developed in the mid 1960s by the Universidade de Minas Gerais - which were in turn based on the experience of the Universidade de Brasilia in the early 1960s and responded to the desire of highly educated groups to adopt the American research university model.(4) It is possible to establish a direct link of ascendance between these ideas and those tried in the frustrated experience of the Universidade do Distrito Federal: Anísio Teixeira was part of both, and Darcy Ribeiro, who organized the Universidade de Brasilia, was associated with the Centro Brasileiro de Pesquisas Educacionais during the 1950s.(5) There was also a more direct American presence, through the advice of a joint commission established between the U.S. Agency for International Development and the Brazilian Ministry of Education.(6) From an organizational point of view, the 1968 reform introduced many elements taken from the North American research universities: the departments, which led to the elimination of the traditional chair system; the credit system, which did away with serialized, year-by-year course programs; research institutes; graduate programs providing master's and Ph.D. degrees; and a "basic cycle" in the universities, which was designed to provide some kind of general, college-like education in the first two years of school. All higher education institutions were supposed to evolve to this model; no room was allowed for institutional and role differentiation.

TABLE 10. Growth of the Educational System in Brazil, 1965-1980 (1970 = 100)
  1965 1970 1975 1980
Population 87.3 100 115.0 127.8
High-school graduates 49.3 100 163.1 239.6
Vacancies for higher education 39.7 100 240.1 279.3
Applicants for higher education institutions 33.7 100 237.5 548.5
Undergraduate enrollment 34.2 100 212.0 294.9
Enrollment in private institutions 28.3 100 245.3 353.7
Enrollment in public universities   100 164.0 238.2
Source: Education data from Ministério da Educação, Serviço de Estatistica da Educação e Cultura; population data from Brazilian censuses.

Implementation led to unanticipated results, partly because the 1968 reformers did not foresee the explosion in demands for higher education that was gaining speed precisely in those years (Table 10). Applications for higher education institutions increased more than fivefold between 1970 and 1980, partly because of the expansion in secondary education and partly because new social groups (women, older people) were trying to enter the system. Government authorities responded to the pressure by letting private institutions of higher education proliferate without much quality control, moving further and further away from the research university model that was supposed to be followed by all.(7)

The new rules were applied mostly to the public sector, but even there the results were different from what was expected. Previously, power resided mostly with the schools' congregations, and the old professional schools were the only real institutions from a sociological standpoint, within universities or in isolation. The new arrangement sought to transfer power to departments and research institutes and to transform the old professional courses into simply a sum of credits to be obtained by the students in different departments. The stronger and more traditional schools resisted this change and implemented the department - institute organization only within their walls. New and weaker areas were more open to innovation, but their very weakness led to power concentration in the rector's offices. The schools that kept their institutional integrity were the ones that best managed to maintain or improve their quality in the years to follow.

Two other innovations - the basic cycle and the credit system - also had dubious results. Students finishing the equivalent of high school had to take competitive examinations for the career and school of their choice, and those who were admitted were immediately committed to those careers. The basic cycle became sandwiched between the entrance examinations and the vocational courses and was perceived by most as just an annoying waste of time. The credit system ran against the fixed and regulated contents of most careers and collided with the limited resources the institutions had for offering course choices. Both the credit system and the basic courses became, at best, new and more complex ways of doing the same old things and, at worst, administrative and pedagogic nightmares.

The educational authorities also created legal and budgetary conditions that allowed the universities to hire full-time faculty. In the past, university salaries had been low and had not competed with what a successful liberal professional could get from clients. As the universities expanded, new and nontraditional courses were introduced, graduate and research programs were created, and a new stratum of full-time faculty started to emerge. In part, members were called to staff the new graduate programs, but not many had the appropriate qualifications to work at this level, and the new basic cycle, coupled with the expansion of enrollments, required that a large number of new teachers be admitted without delay. The result was that, within a few years, most of the faculty of the Brazilian public universities went from part-time to full-time without necessarily increasing its academic qualifications.

The New Graduate Programs

The reform was much more successful with regard to creation of academic departments, research institutes, and graduate programs. Institutions like the universities of São Paulo, Rio de Janeiro, Minas Gerais, and Rio Grande do Sul - which had benefitted in different degrees from past scientific traditions, the presence of foreign visitors, and opportunities of international exchange-could more easily adapt to and benefit from the new organizational formats. Small, high-quality graduate courses, which earlier hardly existed as an organized endeavor, were easily put together by these same persons and institutions.

This trend was reinforced by new sources of funding and institutional flexibility caused by the entrance of economic planning agencies into the field of science and technology. Suddenly and for several years, the amount of money available for science and technology far exceeded the ability to spend it. The new science and technology agencies created to handle these resources were flexible and modern; free from the bureaucratic and budgetary limitations typical of the Brazilian civil service; and looked with contempt at the complex, conflicted, and bureaucratized university institutions. They tried to stimulate research on Brazilian private and public firms, providing them with low-interest loans and technical assistance services. They soon realized that most of the competent people were in academic institutions and turned a large part of their resources over to them.

The strategy adopted by the science and technology agencies was to identify what they considered good or promising research groups and to provide them with direct support, very often bypassing the established procedures for labor contracts, accounting procedures, and decision making within the universities. Substantive considerations were all that mattered. For the researchers, there was now a market that was sensitive to their qualifications and aspirations. For the universities, new resources became available, but they also flowed completely out of their control. Well-equipped, well-staffed, and well-paid departments and research programs began to exist side by side with poor programs-the first more concerned with research and graduate education, the latter bound to the traditional undergraduate schools and courses. A two-tier system was therefore introduced not only among higher education institutions but also within each of them, leading to tensions and ambiguities that would intensify in the years to come.

The education authorities had their own plans for faculty upgrading and graduate education. The new legislation required that faculty be hired and promoted only if they had the appropriate graduate degrees, and the universities were stimulated to create and expand their graduate programs. Quality was to be regulated through the Conselho Federal de Educação and an agency within the Ministry of Education, CAPES (Coordination for Improvement of Higher-Level Manpower), which predated the reform and was in charge of providing fellowships for faculty and graduate students within and outside the country. For the planning agencies, their strategy worked very well. In 1970 some 57 doctoral programs were available in Brazilian universities; in 1985 there were more than 300, with another 800 providing training at the master's degree level(8). About 90 percent of these courses were in public universities, and both levels combined were graduating about 5,000 students each year. By all accounts, Brazil had begun to build a significant scientific community.

There was only partial overlap between the policies of the educational agencies and the Science and Technology agencies. The premium placed on academic degrees led to a rapid proliferation of graduate programs throughout the country. The universities had the freedom to create them, and the teachers pressed for paid leaves to follow the new courses. Since most of the students were in "soft" fields, graduate education expanded most in those fields (see Table 11). In the process, quality usually suffered. The Conselho Federal de Educação, which was supposed to provide accreditation, was extremely slow and usually not very qualified to do the job. The agencies supported the programs of their liking, mostly in the basic and technological fields, and could not have cared less about the accreditation mechanisms devised by the council. Eventually CAPES established a peer review mechanism that became the de facto accreditation accepted by both sides. According to these evaluations, only about one-fourth of the new graduate programs were of significant quality or had chances for improvement. This system was useful for allocating grants and fellowships but could not force a university to close down a program that did not qualify.(9)

High Technology Institutions

The culmination of this drive was the establishment of brand-new institutions that would be free from the limitations of the past. They were to be as free as possible from institutional and bureaucratic limitations or restrictions; they were to receive large amounts of money from science planning agencies and put them to work in the hands of well-qualified people; and they were to work on the frontier of the modern technologies the country was supposed to need for its economic and industrial growth. Two institutions, more than any others, met these requisites: the Universidade de Campinas and the engineering program of the Universidade Federal do Rio de Janeiro, COPPE.

In 1965 Zeferino Vaz left the Universidade de Brasilia to organize a new state university in the city of Campinas, São Paulo. At first the new university was to be mostly an institutional alternative to the Universidade de São Paulo, which was already too crowded with students and personnel. With Zeferino Vaz and heavy federal support, however, it became a project of a new and modern research university.

Table 11 - Students Enrolled in Graduate Programs, by Field (1975-1983) (thousands)
Year Hard Sciences Biological Engineering Health Agriculture Social, applied Social, humanities Total
1975 2,898 2,196 2,421 2,111 1,811 10,808   22,245
1976 3,751 2,172 3,491 3,028 1,942 11,871   26,255
1977 4,362 2,405 3,969 3,370 2,374 15,052   31,532
1978 4,829 2,761 5,442 3,612 2,857 14,130   33,631
1979 4,755 2,951 5,459 3,771 3,018 16,654   36,608
1980 4,936 3,054 5,644 4,216 3,145 17,611   38,606
1981 5,170 3,137 5,715 4,677 2,709 18,776   40,184
1982 4,385 2,852 5,391 4,658 2,728 6,479 12,737 39,230
1983 4,264 2,913 4,990 4,561 2,709 6,452 9,961 35,850
SOURCE: Paulinyi et al. 1986.Social sciences and humanities were counted together until 1981; from then on, applied social sciences (social work, administration, communications, and so forth) are counted separately

His conception for the new university was romantic. A new campus was to be built, and Vaz asked the architect to make "a large central plaza 300 meters across," and said:
I will make it a wonderful garden, with the natural beauty of flowers, trees, stones, and water. This will be the Greek agora, and all units will converge on it... The agora is attractive, and students and professors meet there to discuss and exchange ideas and concepts. You can find there the geneticist, the physicist, the physician, the botanist, the chemist, and the Faculdade de Engenharia de Alimentos. Multidisciplinary programs emerge everywhere, stimulated by the layout - because the circle provides a concept of unity: there are no privileged positions or sides. All converge on this plaza, which symbolizes the well-being of mankind. I wanted to create a university like an organism in which the different organs - physicists, mathematicians, naturalists, philosophers, artists - all worked together for the preservation of the community's physical, mental, and spiritual well-being.(10)
Less romantic was his practice. Based on his reputation, and with financial support from the state and federal government, Vaz initiated an effort to bring back Brazilian scientists who had left the country in the previous years. In his interview Vaz told about his personal contacts and the support received from Finance Secretary Dilson Funaro at the state level and from powerful names in the federal economic and planning agencies, such as Minister of Finance Delfim Neto, Marcos Viana at the Banco Nacional de Desenvolvimento Econômico, José Pelúcio Ferreira at the Financiadora de Estudos de Projetos, and João Bautista Vidal at the Secretaria de Tecmologia Industrial.

Physics was to be a central activity, and Marcelo Damy de Souza Santos was invited to organize the institute, which was christened with the name of Gleb Wataghin. Cesare Lattes, who had left the Centro Brasileiro de Pesquisas Físicas, was also offered a position. Next came Sérgio Porto, followed by Rogério Cerqueira Leite, José Ripper, and several others. This group had in common old links with the Instituto Tecnológico da Aeronáutica and their years of work at the Bell Laboratories in the United States.

Sérgio Porto was to play a central role in the new project. He was born in 1926, studied chemistry at the Faculdade de Filosofia in Rio de Janeiro, and received a Ph.D. in physics from Johns Hopkins University in 1954. From 1954 to 1960 he worked in the physics department of the Instituto Tecnológico da Aeronáutica, and then he joined the technical staff of Bell Telephone Laboratories until 1965. From there he moved to the University of Southern California. A series of studies on the Raman effect based on the utilization of lasers led to many publications and an international reputation.

In spite of his achievements in the United States, in the mid-1960s Porto began to consider the conditions under which he would return to Brazil. In 1970 he came to Brasilia to testify before a parliamentary investigative committee about the brain drain of Brazilian scientists. "I told them: The Brazilian people do not want us. We did not run away, you threw us out." For him, inadequate working conditions, not politics, were the reason so many Brazilian scientists lived abroad (his list included Sérgio Mascarenhas, Luis Valente Boff, Rogério Cerqueira Leite, José Ripper, Roberto Salmeron, and Fernando de Souza Barros). Several contacts with Planning Minister Reis Velloso followed, and in 1972 Porto agreed to return based on a pledge of $2 million for his projects.(11)

Sérgio Porto was supposed to become head of the physics institute, but when he arrived the position was already occupied by Rogério Cerqueira Leite.(12) Disputes over leadership and orientation led to Damy's dismissal, and Leite and Porto divided command of the new institution with Vaz. Eventually, the physics institute became what Leite described as "the largest physics institute in a university in the world" - and a good one in terms of quality - working in all areas related to semiconductors, from crystals growth to practical applications. Porto continued with his research on lasers and worked on ways to utilize them in a great variety of tasks, from eye surgery to atomic fusion. A company for technological joint ventures with the industrial sector, CODETEC, was also established under the leadership of Aldo Vieira da Rosa, an air force officer and scientist originally also from ITA.

While Zeferino Vaz was in charge, the Universidade de Campinas was considered to be in a provisional stage, which in practice meant that Vaz and his closest associates had full power to hire faculty and designate the university's authorities. This situation, combined with the extremely high proportion of "soft money" that became part of its budget, made the Universidade de Campinas an unstable, ambitious, and innovative institution, the closest Brazil ever came to a true research university.

While Campinas was planned as a classic university to which high technology was attached, COPPE was precisely the opposite - that is, an attempt to graft a high-technology teaching and research program to a traditional Brazilian university, the Universidade Federal do Rio de Janeiro.(13)

COPPE's history is inseparable from that of its founder and first director, Alberto Luis Coimbra.(14) Thanks to Coimbra's efforts, in the early 1960s Frank Tiller, his former professor at Vanderbilt (by then working at the University of Houston), was invited to teach at the Escola Nacional de Química. A series of visits by Brazilian chemical engineers to the United States followed, the goal being to establish a graduate program in chemical engineering in Rio de Janeiro This interchange was supported in part by the Escola de Química, in part by American institutions, and in part by the Organization of American States. In 1962 Coimbra went to the United States to become more acquainted with the American model of graduate education. In the next few years he sent some of his best students Giulio Massarani, Afonso da Silva Teles, Carlos Augusto Perlingero-to study in Houston.(15)

The project for a new institution was already being designed. Early in 1964 the project began to receive support from the recently created Fundo Nacional de Tecnologia and from the Banco Nacional de Desenvolvimento Econômico This influenced its direction.
A friendly relationship between professors and the officers at the BNDE was established. It was very gratifying, very pleasant. They shared the problems of the university and helped out when payment was delayed for some reason. The first bylaws for the fund were written at COPPE. At first they had only engineering, but we included physics, mathematics, and chemistry because one cannot do graduate education in engineering without the basic sciences... We also helped establish the graduate program in mathematics at the Instituto de Matemática Pura e Aplicada.(16)
Money began to flow in 1967 and peaked in 1973. COPPE established a bewildering variety of cooperation agreements with persons and institutions all over the world. Its catalog for 1971 mentions the Organization of American States, the Fullbright Commission, the Rockefeller Foundation, the U.S. Agency for International Development, and the governments of France, England, the Netherlands, the Soviet Union, and Germany.

Through these agreements, foreign professors were invited, students went to get their doctoral degrees, and joint research projects began.(17) From chemistry, COPPE expanded into biomedical engineering, mechanical engineering, metallurgy, civil engineering, production engineering, nuclear engineering, naval engineering, urban planning, production engineering, and business administration. Several hundred professors were hired with better salaries and working conditions than those of the Escola de Engenharia, to which COPPE was formally attached. As the program's general coordinator, Coimbra moved to concentrate most decisions in his own hands. "We had to create almost an island to protect ourselves against the mold that surrounded us. We had to grow and to go up very quickly - like a balloon rising so quickly that no stone could reach us - to be strong and not tainted by the university's inefficiencies... We had to make use of unorthodox means to do all that. This of course was not in accordance with the university's bureaucracy."(18)

Bureaucracy fought back. Coimbra was charged with mishandling public funds and left his post with bitterness in 1973. In 1977 the Universidade do Rio de Janeiro approved a new statute for COPPE that placed it under the direct authority of the university's rector. After the first years of strong entrepreneurial leadership, it was the time for management, and COPPE became a permanent and important component of the university.

How well did COPPE meet its initial goals? According to Coimbra, "COPPE was created to form a kind of professional Brazil did not have at the time, one at the master's and doctoral levels. We believed that people with these qualifications were necessary to the country's technological development. We had middle-level personnel, engineers, but we lacked graduate people who could create new technology."(19) This was precisely what the Fundo de Tecnologia defined as its own goals. According to the bank's economists, Brazil showed a clear lag between investments in basic sectors of the country's economy and investments in education. Industrial development, however, depended on the qualifications of the personnel, the strengthening of the country's scientific and technological competence, and the reduction of payments for imported technology, know-how, technical assistance, and patents. The fund was supposed to both invest in graduate education and provide incentives and technical assistance to stimulate Brazilian companies in their use of modern technology to meet the competition of the local branches of multinational corporations.(20)

The balance was not very positive. Coimbra, for one, showed his frustration:
We created the graduate programs for a Brazil that did not exist and still does not exist, which did not correspond to what we expected to happen. We were throwing into the market a sophisticated product meant for the country's technological development. We imagined that if we did our part in forming creative people in engineering, they would be absorbed by a country that really wanted to create its own technology. But it never happened... Brazil does not need M.A.'s and doctors, not even five-year engineers. Operational engineers are enough, since we will keep handling imported factories forever.(21)
Other opinions were more balanced but pointed in the same direction:
I am almost sure that COPPE is full of defects, in the sense that it goes too far ahead of Brazil's reality and is too sophisticated with regard to the practical side of the productive activities in many fields... The industries were completely resistant to any participation of this kind, now or in the past. Our industrial development was based exclusively on the importation of foreign technology, on multinational corporations, or on the acquisition of foreign patents by Brazilian companies.(22)
A more precise picture can be gathered from an analysis of data on student graduation and their future work (see Table 12). Between 1964 and 1978, only twenty-five students received doctoral degrees. Only about 20 percent of the master's students ever got their degrees, and 50 percent abandoned the courses without getting their credits. These low rates of completion and graduation are similar to what is found in most graduate courses throughout Brazil. In that sense, COPPE is not exceptional, but it is obviously no less troublesome.

TABLE 12. COPPE: Graduation, Enrollment, Desertion, and Destination of Students, 1965-1978
Field and year of creation Ph.D.'s Master's Degrees Master's Enrollment Desertion Destinations*
Chemical engineering (1963) 4 124 437 52.6% Teaching, public companies
Mechanical engineering (1966) 3 66 268 48.5% Teaching, private companies
Electrical engineering (1966) 1 86 699 60.9% Public companies, teaching
Metallurgy (1966) 2 59 368 30.2% Teaching, public companies
Civil engineering (1967) 6 139 942 65.5% Teaching, doctoral programs
Production (1967) 3 140 852 68.9% Public companies, teaching
Naval (1967)   25 97 43.3% Teaching
Nuclear (1968)   87 329 44.3% Doctoral programs, teaching
Systems engineering (1971) 6 109 660 49.1% Teaching, doctoral programs
Biomedical (1971)   21 120 33.3% Teaching
Business administration (1975)   12 160 20.6% Public companies
SOURCE: Nunes, Souza & Schwartzman, 1982:241-42. *Main Occupation of those who obtained their master's degrees. Two occupations are given when the figures are close.

If we leave aside the more recent courses, it is obvious that the highest rates of desertion were in fields where enrollment was also highest. These figures suggest that, for most students, graduate education was just a way of extending somewhat their student life, most of the time with a fellowship, while they waited for a place in the job market. These figures can mean also that COPPE students were in such high demand that they had no time to finish their degrees.

Either way, it is obvious that the level of education offered by COPPE was much higher than what was demanded by the students who do not get their degrees or by the companies that employed them. The same or better results could be obtained if COPPE were geared to provide more modest levels of training for the bulk of its students and concentrated its effort in graduate education on the 20 percent or so who finish their degrees. For those who do get their master's degrees, the destination is the university itself (37 percent) or public employment (21 percent). Only 13 percent enter the private sector, while 19 percent continue their student life at the doctoral level. Those who became teachers after getting a master's degree probably worked with undergraduate students and were not able to continue a professional life geared toward research and technological development. Desertion rates for those who work for their doctoral degrees was still higher than at the master's level.

In short, COPPE seemed to become a good program of specialization in the most traditional and operational fields of engineering: civil, electric, systems, and production engineering. Its original goals-to develop a truly national and internationally competitive engineering capability were mostly frustrated. The ideals of research, academic work leading to dissertations, a constant flow of professors to provide the highest possible levels of technical and scientific competence, the permanent interchange with Europe and the United States-these notions, what many call "the COPPE spirit," contrast sharply with the reality: the desertion of 80 percent of its students and the employment patterns of the others.

Big Science and High Technology

The whole rationale for investment in technology is its practical utilization, and a survey of experiences of this kind would have to include the whole field of agricultural research; recent achievements in biotechnology, the airplane and weapons industry, and the technology for steel production;(23) and the role of such institutions as the Instituto de Pesquisas Tecnológicas in São Paulo. Such a survey would be beyond the scope of this study, but two extreme cases of high technology-atomic energy and computers-stand out from the others, in part because of their proximity with the basic sciences, and require a closer look.

We have seen how research in advanced physics, which provided the basis for some applied work during the war, could not perform the same feat afterward in the much more complex, expensive, and politically charged field of nuclear technology. Confronted by the alternatives of trying to develop its own technology with the help of the existing scientific community or acquiring foreign technology, the Brazilian government chose the second. In 1975 an ambitious agreement for nuclear cooperation was signed with West Germany, which implied the construction of several nuclear energy plants and the transfer of enriched uranium technology. The agreement drew strong opposition from Brazilian scientists, because it consisted mainly of the transfer of engineering technology and did not incorporate the acquired or presumed competence of Brazilian scientists. In time the agreement proved to be overambitious, and it is now limited at most to the construction of two power plants, neither of which is near completion at this writing (the Westinghouse plant, meanwhile, is plagued by successive technical difficulties and may have to be scrapped before going into full operation). A so-called parallel program of atomic research was also undertaken by the Brazilian military, outside the restrictions built into the German-Brazilian agreement. Rumors that Brazil is developing its own atomic bomb have never been confirmed; still, the government has acknowledged development of nuclear engines for ships and submarines, and in September 1987 it was formally announced that Brazil had developed all the needed technology for the production of nuclear fuels for peaceful purposes. The method of ultracentrifugation was said to be similar to the one used by the URENCO consortium in Europe, and the grade of enrichment, which was announced to be of 1.2 percent, was supposed to increase to 20 percent in one or two years, when an industrial plant was supposed to begin working.(24) The work had been going on for eight years at the Instituto de Pesquisas Nucleares at Universidade de São Paulo with support provided by the navy and at a stated cost of $37 million. The announcement was received with widespread skepticism by the Brazilian press and Brazilian scientists. The level of enrichment was considered too low for any practical purposes and the expenses too high, given the country's deep economic crisis and the other needs of scientific institutions. The concentration of resources in military research at the expense of civilian institutions was seen as a worrisome trend.(25)

While a key feature of the nuclear programs was the exclusion of university-based scientists and the creation of large, state-controlled bureaucracies, the computer industry began with people coming out of the universities and gave rise to a large number of privately owned companies. The origins of this industry can be traced back to the physicists and engineers trained by the Instituto Tecnológico da Aeronáutica. Later it drew from those educated or engaged in research in such places as the Universidade de Campinas, the Escola Politécnica of the Universidade de São Paulo, the Universidade Católica in Rio de Janeiro, and COPPE. At the end of the 1960s the Banco Nacional de Desenvolvimento Econômico began to support research and development in computer science and microelectronics. In 1971 the navy contracted with the Universidade de São Paulo and the Universidade Católica in Rio de Janeiro "for the planning, development, and manufacture of a computer prototype suitable for naval operations, preferably in association with Ferranti," an English company.(26) In 1972 the government created an agency to Coordinate the whole area of computer acquisition and data-processing for the public sector, but with powers to control imports for the whole industry. This commission led to the creation of the Secretaria Especial de Informática, established in 1979 under the National Security Council with full powers to decide everything related to computers and microelectronics, from imports to the establishment of manufacturing firms or the acquisition of data-processing equipment by universities or state corporations. In 1973 a state-controlled holding, Digibrás, was established with capital from state-owned corporations to promote a Brazilian computer industry, and in 1975 it gave rise to Cobra S.A. With these instruments, a policy of market reserve for the production of microcomputers for Brazilian firms was established. For minicomputers, the strategy was to stimulate joint ventures under the control of Brazilian firms and a full transfer of technology; the market for mainframes remained open to IBM and other multinational firms, but all their actions came under scrutiny.

This policy coincided with the worldwide explosion of the microcomputer industry, and in Brazil it started with imported components and was patterned on well-known products manufactured by Sinclair, Tandy Corporation, Apple, IBM, and their clones. Research was mostly on reverse engineering, software adaptation, and the development of less complex circuitry and parts. In 1983 Brazilian computer firms, which had barely existed five years before, were employing about 16,000 people and selling about $690 million in equipment, while multinational firms in the country were employing about 10,000 people and had sales of $800 million.(27)

By the end of 1984, a bill defining a "national policy for informatics" was approved by the Brazilian Congress by large majorities from the government and opposition parties. It was preceded by a lively debate reminiscent of the one leading to Brazil's state oil monopoly in the early 1950s Once more the issue was presented as a dilemma between national autonomy and self-determination, on the one hand, and control of the country's economic resources by international Companies and their local associates, on the other hand. Again the proposed policy gathered intense support from those ranging from the nationalist military groups to intellectuals, students, scientists, trade unions, and politicians.(28)

The many remarkable elements in this policy generated several studies, international press coverage, and threats of retaliation from the Reagan administration. The main novelty was that this was the first attempt in Brazil's history to develop an industrial policy based on local technology and purely Brazilian firms. Market protection had existed for many years, for instance, for a small group of multinational corporations in the automobile industry; and state monopoly for oil had also existed for many years, but based on internationally available technologies. Here, however, some level of technological competence had been previously built, and it was to provide the foundations for the new policy.

The Brazilian policy for computers is an extreme case of "backward integration", in which production starts with assembly of the final product with imported parts, with the expectation that the product will incorporate an increasing amount of locally produced components.(29) Such a policy requires, among other things, a corresponding investment in research and development, which in the Brazilian case does not seem to have existed.

A picture of the current stage of university research in computer sciences (which does not include such related fields as electronics or automation) can be gathered by examining a three-year research plan put forward by the Sociedade Brasileira de Computação, a scientific association, and the Centro Tecnológico de Informática, a research outfit created by the Secretaria Especial de Informática near the city of Campinas. There are five institutions providing doctoral degrees for about ten people a year, and the total number of researchers with doctoral degrees is 108. There are also fifteen institutions providing master's degrees and a small number of people being trained abroad and returning to the country. At most, fifteen new Ph.D. level researchers are expected to enter the field each year. The total number of researchers outside industries is estimated at 750, five hundred of whom work in software. The research plan projected an increase in the number of Ph.D.'s to 750 by the year 1997, against 300 if the current growth rate stands. For this, proportional increases in technical personnel, equipment, library facilities, and so forth will be needed. The total cost for all research projects, infrastructure, interchange, and follow-up was estimated at about $40 million in three years, a very small amount if compared with what is being invested in the developed countries, and close to the value of the equipment IBM was willing to provide Brazilian universities in the same period.

A list of high-priority research projects was also drawn up. It included the design of digital systems, time-sharing systems, software engineering, data bases, CAD/CAM, artificial intelligence, sign processing, mathematics applied to computing, and computer theory. In all, sixty-five projects were identified. The research plan was an aggregation of existing research projects, to which a weak order of priorities was attached. The plan was never funded as it stood.

There is no equivalent information for research in industries, but its scale could be inferred from the existence of about 4,000 employees with university diplomas in all Brazilian companies, most of them working on sales, maintenance, quality control, and administration. The Centro Tecnológico de Informática was supposed to become the starting point for a large research and development establishment, to be supported with a special tax and placed directly under the authority of the Secretaria de Informática The tax was vetoed by President Joao Figueiredo from the 1985 law, and the Centro remained as a small outfit, with about 300 people and a budget of around $1 million a year. It is not a purely research institution, since it is supposed to sell services to the private sector, to provide technical assistance, and to develop joint projects with the universities. It is also responsible for overseeing the fulfillment of nationalization targets of IBM's computer assembly plant nearby (geared only to the foreign market) and for developing standards and providing certification for the national computer industry. The weakness of the research effort is one reason the Brazilian policy for the computer industry came under severe criticism, not only from foreign competitors who would like to sell in the Brazilian markets but from end-users and manufacturers in Brazil who see this policy as a growing roadblock to their access to state-of-the-art technologies. In computers, as in atomic energy and in other applied fields, the great leap forward was far shorter than originally expected.

Notes

1. Adler 1987:210; Lopes 1978.

2. Lessa 1978; F.M. de O. Castro 1985; Schwartzman 1980.

3. Schwartzman 1978:574.

4. Pimenta 1984:24.

5. Mariani 1982a.

6. Carneiro et al. 1969.

7. Schwartzman 1988a.

8. Paulinyi et al. 1986.

9. Castro and Soares 1986.

10. Vaz interview.

11. "This was his commitment right there near the swimming pool [at Porto's Los Angeles house]. He [Velloso] would tell Zeferino Vaz to provide the building and gave him assurance that he would hire thirty Ph.D.'s. These were my conditions in return: thirty Ph.D.'s, a building, and $2 million. And I got them. Unfortunately, I was naive, and the $2 million became just one. FAPESP had promised $300,000 but gave only 400,000 cruzeiros (about U.S.$67,000), which means that unfortunately I could not finish my laboratory" (Porto interview).

12. Cerqueira Leite had been a student of Porto's at ITA and went to work at the Bell Laboratories in 1962 after earning a degree in physics from the University of Paris. His return to Brazil was part of the same initiative reported by Porto and included also the promise of financial support and equipment.

13. For a full account, see Nunes, Souza & Schwartzman 1982.

14. Alberto Luis Coimbra graduated from the Escola Nacional de Química in Rio de Janeiro in 1946. Re studied for a master's degree in chemical engineering at Vanderbilt University and from 1949 to 1953 worked at the Escola de Engenharia Industrial in São Paulo, a private institution. In 1953 he returned to Rio de Janeiro to compete for a chair at the Escola de Química. For the next several years he "taught at the Escola de Química, at Petrobrás [the Brazilian state-owned oil company], worked as a consultant for two American firms, taught at the Universidade Católica, and dealt with chemical engineering and the mechanics of fluids" (Coimbra interview).

15. "It was like a direct order from Coimbra. He was a very good teacher at the Escola de Química. People liked his courses, and it was a privilege to be treated well by him. I remember that Coimbra came into the library and gestured to me: 'Come here. You are going to Houston.' We believed in him so much that we did not think twice... We had no idea what we were doing or what graduate education really meant" (Massarani interview). Massarani and Teles went to the United States "to get their M.A.'s and return to Cod knows what. If the project for a new graduate program went through, they would have a job, a career in graduate education at the university. If not, they would work in industry, where, at the time, they could do very little with the knowledge they obtained with their M.A.'s" (Perlingero interview).

16. Coimbra interview.

17. "Coimbra was really concerned with getting all tendencies. Americans were somewhat prepotent in the organizations they supported-the OAS and others-and he did not like that because he felt restricted in his freedom... Today COPPE has more European influences than American. But in some areas the Americans are stronger-for instance, in Systems engineering. Europe and the United States are similar in terms of knowledge, and the Europeans are stronger in chemistry. But it is difficult to say; it is not clear" (Massarani interview).

18. Coimbra interview

19. Coimbra interview.

20. BNDE 1974.

21. Coimbra interview.

22. Pinguelli Rosa interview.

23. Dahlman and Fonseca 1987.

24. The entire nuclear program was put in disarray with the economic crisis of the late 1980s, and the industrial plant never became operational. In 1990 the Fernando Collor government sealed a deep well drilled by the military for underground nuclear testing and signed an agreement of mutual inspection with Argentina to ensure that the military side of the nuclear testing would stop

25. Guilherme 1957; Sales 1958; H. G. Carvalho 1973; Gall 1976; Leite 1977; Morel 1979; Adler 1987; Jornal do Brasil 1987.

26. Adler 1987:245.

27. Secretaria Especial de Informática 1984.

28. What follows is based on Schwartzman 1988b; see also Tigre 1983; Piragibe 1985; Frischtak 1986; Evans 1986; Adler 1987.

29. Nau 1986.