What, finally, of the most important test of all—the indices of social progress under Cardoso’s rule? Eagerly associating himself with Clinton and Blair as a companion spirit in the South, Cardoso—who in the eighties had spoken of himself as a social-democrat, and headed a party (the PSDB) calling itself such—preferred the vaguer recipes of the Third Way once in power. Few have many illusions in the Anglo-American versions of this nostrum, now widely discredited in its homelands. In the very different conditions of Brazil, has it acquired more substance?
Table 3 compares the rates of growth and industrialization under Kubitschek and the ‘miracle’ years of military rule, with those under Cardoso. The contrast is arresting. In the former, GDP per capita grew at an average rate of 6.1 per cent between 1957–61 and 8.4 per cent between 1968–73, whereas between 1995 and 2001 Cardoso’s neoliberal experiment delivered an average growth of GDP of 2.4 per cent, and per capita of just 1.0 per cent.
Comparing not periods but countries, the results are also striking. Of the four main neoliberal experiments in Latin America, Brazil certainly did better than Argentina, where the model eventually collapsed, yielding overall figures of 0.9 and –0.3 per cent. But it did markedly less well than the other two stars of neoliberalism: Chile registered GDP growth of 4.8 and per capita of 3.4 per cent, Mexico 3.1 and 1.4 per cent. More starkly still, under the successive impact of the Russian crisis, the collapse of the Plano Real in 1999, and initial contagion from the Argentine crisis of 2001, per capita income in Brazil fell in dollar terms from $4,940 in 1997 to $2,920 in 2001—a regression to the levels of the 1980s; while GDP itself dropped from $807 billion to $503 billion, demoting the country from its rank as the eighth largest economy in the world, and placing Mexico as the first in Latin America in total economic output, with a GDP in 2001 of $618 billion. 
The inevitable consequence of such sluggish long-term growth, as elsewhere in Latin America, has been massive urban unemployment. According to official figures, which systematically underestimate the real rate, open unemployment rose from 4.6 per cent in 1995 to 7.6 per cent in 1998 and 1999, and 9 per cent by March 2000.  Other estimates reckon that open unemployment in São Paulo, the industrial heartland of Brazil, rose from 13.2 per cent in 1995 to 19.3 in 1999, and reached 20.4 per cent in May 2002, higher than the official 17.4 per cent reported for Argentina in 2001 just before its collapse.  This social failure has been a direct result of the massacre of small and medium enterprises, under the twin pressures of very high interest rates and sweeping trade liberalization. Such unemployment figures reveal a depth of social exclusion that does much to explain the rise of urban violence in Brazil during the Cardoso presidency. The latest data show that Brazil has climbed the ranks of Latin American countries for homicides to near the very top, with a coefficient of 19.12 per 100,000 inhabitants in 1992 rising to no less than 26.18 today.
If we turn to education, Cardoso has made some progress in reducing illiteracy in Brazil, according to official definitions, from 18.3 per cent of the population in 1990 to 14.7 per cent in 2000; although some 18 million Brazilians still cannot read or write and the average number of years of schooling of the economically active population is only 5.5, putting Brazil alongside Guatemala and Nicaragua. Argentina, Chile, and even Mexico come off much better, with illiteracy rates of 3.1, 4.3, and 9.0 respectively, and much higher average number of years of schooling: 10.1, 9.0 and 7.4. Infant mortality has also fallen, from 38.4 per cent in 1994 to 29.6 per cent in 2000, to the credit of the regime. But the Northeast, the poorest region of the country, continues to suffer from the appallingly high rate of 44 per cent. If infant mortality in Argentina and Chile is lower, 24.3 and 14 per cent, in Mexico it is higher than in Brazil, at 34.0 per cent. Set against these modest gains are the reverses of Cardoso’s second term. Between 1999 and 2001 average real wages fell 10 per cent, and government researchers themselves estimate that by 1999 absolute poverty had increased to 34.1 per cent of the population, some 53 million people, with another 22.6 million, or 14.5 per cent of the population, condemned to indigence. If Brazil does not compare well in these respects with Argentina or Chile (19.7 and 4.8, and 20.6 and 5.7 per cent), absolute poverty is more widespread in Mexico, 46.9 per cent in 1998, and indigence stands at 18.5 per cent. 
Lastly, what of income distribution? Despite claims by Cardoso and his economic team that currency stabilization was the best social policy, inequality remains one of the most scandalous in the world—the countries with the four highest Gini coefficients are, in rank order, Nicaragua, Brazil, South Africa and Malawi. As Table 4 illustrates, in 1999 the richest 10 per cent of the population enjoyed 47.4 per cent of the national income and the poorest 40 per cent a mere 8.1 per cent.
In fact, the neoliberal model in Latin America has increased inequality everywhere, the Gini coefficients in Argentina, Chile and Mexico all rising through the nineties. But according to ECLAC, in 1999 Brazil was the only country in Latin America in which more than half of the population had less than 50 per cent of the mean income.  Modest progress in literacy, health and reduction of poverty have not been able to redeem mediocre growth rates, high unemployment, vertiginous inequality and spreading urban violence and criminality.