Government's Role in Development: The Case of Brazil under the Workers Party


Sociologist Peter B. Evans modifies his "embedded autonomy" theory to take account of the pressures of international finance on Third World states and alternatives to an industrialization strategy.

In 1995 University of California, Berkeley,  sociologist Peter B. Evans published a book with the somewhat obscure title "Embedded Autonomy: States and Industrial Transformation." In it he elaborated a theory that well-ordered state bureaucracies were an important factor in their own right in speeding industrial development, independent of their particular policies, provided they forged a strong alliance with a domestic productive class. In elaborating this theory Evans had looked at the economies of South Korea, Brazil, and India. On May 25, 2004, he revisited his topic with a talk at UCLA's Sociology Department on Brazil  under the Lula administration. Evans defended his theory of a decade before, but made some modifications to indicate how a Third World government might have to do some dodging to contribute to economic development while accommodating pressures from international bankers. Evans'  talk was cosponsored by the Sociology Department and the UCLA Latin American Center.

"'Embedded autonomy,'" he began, "is a delicately balanced combination of (1) capable, coherent bureaucracies characterized by meritocratic recruitment, long-term career rewards, and high espirit de corps, with (2) dense ties to industrializing elites, which provides access to information, agents for implementation, and catalyses a more coherent, forward looking entrepreneurial class."

The state's role, he said, is to be autonomous of society, the autonomy part of the theory's title, "capable of constructing long-term projects of social change that transcend short-term interests of specific groups." But the state could not be heavily insulated from its society. It had to be embedded in a deep alliance with people active in the economy. "It is their dense ties to potential partners in shared projects that is crucial in terms of making states effective. Even the most capable state apparatus lacks the capacity to implement general social transformation. No bureaucratic apparatus can accumulate the information necessary to carry out such a project."

How Big Is the "Embedded Autonomy" Effect?

Peter Evans summarized the best data he and his collaborators had found in 1995, which showed there really is an embedded autonomy. They had had developed a "Weberianness scale" of governmental bureaucratic efficiency and run regressions on the economic performance of 35 countries. At the high end of efficiency were Singapore, Taiwan, Korea, and Hong Kong. At the low end, counting from the bottom, were the Dominican Republic, Kenya, Guatemala, Nigeria, Argentina, Haiti, Zaire, and Ecuador.

The results? "An increase of one half of a standard deviation in the Weberian score is worth a 26% increase in GDP from 1970 to 1990 controlling for human capital and initial GDP per capita," Evans said. "An increase of one standard deviation in the Weberian score is effectively the same as a shift in average education of from 3 to 6 years. So bureaucratic capacity has a substantial effect on a country's chances for economic development."

Brazil under the Workers Party

Leftist trade union leader Luiz InĂ¡cio Lula da Silva was elected president of Brazil in October 2002. How did this sit with an embedded autonomous state bureaucracy? Peter Evans said that he needed to modify his theory to reflect the power of international financial institutions to put pressure on developing country governments. "Looking at the Workers Party period only, the first thing that becomes clear is the extent to which that experience gives us a new appreciation of the economically coercive constraints placed on transformative strategies by the global neoliberal regime," Peter Evans said. He traced the sharp fall of the Brazilian real against the dollar leading up to and immediately following Lula's election. From 3.43 reals to the dollar in April-May 2002, it fell to 3.26 in October 2002. The currency restabilized several months into the Lula administration in May 2003 at 3.34 to the dollar and remained at about that rate for the next year.

Financier and author George Soros, Evans recounted, "said the international financial markets would punish Brazil if they elected Lula, so badly that Lula would not be elected." International banks "advised their investors not to invest in Brazil in the later spring of 2002, responding to a perception of the Workers Party, of being a left party with a large contingent of Marxists. The real was sinking like a stone. This suggests that the coercive impact of the global neoliberal machine is not something that falls on countries randomly or uniformly but is driven by the political perceptions of the international financial leadership. If the political leader is perceived as leftist or unfriendly to capital, they will bear the brunt of pressure from global neoliberal institutions."

Brazil's currency began to go back up shortly after Lula's electoral triumph. "Why? Because Lula and the Workers Party understood that the average Brazilians had lost 40% of their purchasing power, so they knew they must do something to change the perception of the international community. And it must be something credible, concrete. So they put a genuine, bona fide international banker in charge of the central bank."

The central bank raised interest rates, "a sign of responsibility. If currency is in trouble raise interest rates to attract capital." Lula also committed to a 4% of GDP fiscal surplus in his budget, "even a little more than the IMF wanted. The world image of Lula changed, from a revolutionary to a global statesman. The result was a stable exchange rate for the real and the pressure of the global neoliberal regime was neutralized."

These conservative fiscal policies may have pleased the bankers in London and New York, but they did not make local entrepreneurs happy. "The people screaming about high interests rates are above all industrialists. The Workers Party was able to resist elite demands for special interest economic policies. This shows that the party in power with a strong nonelite base may have a better chance to rebalance where embeddedness is too strong, i.e., where elites are too powerful."

Alternatives to a Strategy of Promoting Industry

Resisting the demands of Brazil's capitalists was one thing, but there were also broader costs to Lula's cautious economic policy that affected prospects for industrial expansion. High interest rates and a large government surplus "makes it near impossible to generate economic growth," Evans said. In 2003 Brazil's growth rate was -0.2%, and there was 19.1% unemployment in Sao Paulo.

Here Peter Evans proposed that boosting GNP was not identical with building the industrial sector. The problem in Brazil, he said, was not so much industrial growth, at which Brazil has been quite successful, but reducing unemployment. More than 50% of Brazil's exports are manufactured goods." Brazil has been a very successful industrializer." With the shift to service jobs in industrialized countries of the northern hemisphere and outsourcing of production, "The North is less industrial than the South. There has been a big growth in industrialization in the Third World, but little effect on GNP."

A Question of Domestic Allies

According to Evans' theory the well-oiled political machine needs to be embedded in society through strong ties to a practical class doing things in the economy. This is presumptively the industrial bourgeoisie, but Evans suggested that for Brazil this might not be so. "The industrial bourgeoisie may be an undependable ally in a project of local transformation. It may want to ally with transnational capital and no longer focus on local transformation, which will make it less interested in cooperating with projects of the local state. Second, focus on the industrial bourgeoisie produces isolation form other social groups that make them see you as an enemy. This is a self-limiting political strategy."

A solution for Brazil, caught between requirements of international finance and an industrialization that is having little impact on living standards, could be a turn toward transformation projects using human capital he suggested. This, Evans said, would better suit the real social base of the Workers Party.

"Think about what you want as a development strategy. You need productivity that delivers what people need. You also need collective goods like health care and education, plus productivity enhancing investments. In the case of a broad investment in human capital there is a large overlap between productivity enhancing investment and the things people need. Health care and education promote production, but they are also high on the list of things people want from life."

Evans gave an example of a program in Ceara state in northeast Brazil, where young women were sent out as health agents accompanied by a state orchestrated propaganda campaign over the radio. "They managed to reach 850,000 families, tripled the coverage of vaccinations, and reduced infant mortality by 36%. This had a huge impact on potential productivity at a cost of $2 per head. A huge rate of return. In contrast they built a Ford plant in Bahia. It cost $400 million in government subsidies and created a few dozen jobs. That was not as good a rate of return."

Evans said he was adjusting his theory in light of this experience. "I was a little too mesmerized by the industrial transformation aspect of embedded autonomy and I should have thought a little more broadly about the implications of building different kinds of social ties. In the case of the Workers Party we have a political party with a broad-based mobilization that has an advantage in organizing embedded autonomy that needs more independence from the industrial bourgeoisie. They would have an extreme difficulty of pursuing a traditional industrializing strategy." He suggested that this alternative human capital strategy might work well for highly organized mobilizational parties in other countries.

"Global neoliberalism is not only a threat but an opportunity. By making the futility of certain kinds of strategies inescapable it pushes Third World countries to seek other options. The health agent version of development strategy is far more applicable to Haiti than a semiconductor strategy. In Brazil the Workers Party is in a better position to organize a health agent program than an industrialization one."

Published: Wednesday, June 02, 2004