The slippery slope : explaining the increase in extreme poverty in urban Brazil, 1976-96

Despite tremendous macroeconomic instability in Brazil, the country's distributions of urban income in 1976 and 1996 appear, at first glance, deceptively similar. Mean household income per capita was stagnant, with minute accumulated growth (4.3 percent) over the two decades. The Gini coefficient hovered just above 0.59 in both years, and the incidence of poverty (relative to a poverty line of R$60 a month in 1996 prices) remained effectively unchanged over the period, at 22 percent. Behind this apparent stability, however, a powerful combination of labor market, demographic, and educational dynamics was at work, one effect of which was to generate a substantial increase in extreme urban poverty. Using a decomposition methodology based on micro-simulation, which endogenizes labor incomes, individual occupational choices, and decisions about education, the authors show that the distribution of income was being affected by: 1) Three factors that tended to increase poverty-a decline in average returns to education and experience, a negative"growth"effect, and unfortunate changes in the structure of occupations and participation in the labor force. 2) Two factors that tended to reduce poverty-improved educational endowments across the board, and a progressive reduction in dependency ratios. The net effect was small and negative for measured inequality overall, and negligible for the incidence of poverty (relative to"high"poverty lines). But the net effect was to substantially increase extreme poverty-suggesting the creation of a group of urban households excluded from any labor market and trapped in indigence. Above the 15th percentile, urban Brazilians have"stayed put"only by climbing hard up a slippery slope. Counteracting failing returns in both self-employment and the labor market required substantially reduced fertility rates and an average of two extra years of schooling (which still left them undereducated for that income level).