While safety net efforts in adjusting countries have a mixed record, the experience provides some important lessons, both for future safety net policy and for social welfare policy more generally. The effectiveness of targeted social safety net benefits has highlighted the extent to which the allocation of basic social services, such as health and education, is skewed toward wealthier groups in many countries, as well as the need for better targeting of social welfare benefits in general. While social expenditures must be at a realistic level—and in many countries in Latin America these expenditures fell well beyond desirable levels during the debt crisis—the allocation of expenditures is as critical as overall amounts, if not more so. In Brazil, for example, only 18 percent of the poorest groups, who account for over 40 percent of the population, are covered by social security, and they receive only 3 percent of social security benefits. In Venezuela, over 50 percent of the education budget is spent on higher education (Birdsall and James, 1990; Angell and Graham, 1995). Chile, in contrast, provides a good example of how social expenditures can be made far more effective in reducing poverty when they are targeted. With the transition to a democratic regime in 1990, the targeted approach of the military government was maintained, while social expenditure was increased at a rate of almost 10 percent per year. Because the demand of upper- and middle-income groups for social services was now served by the private sector, government increases were able to benefit the poor disproportionately. As noted above, poverty has fallen markedly in Chile: from 45 percent in 1986, to 40 percent in 1990, to 28 percent in 1994—and is projected to fall to 17 percent by the year 2000 if the economy maintains its current trajectory of 6 percent annual growth (Cowlan and de Gregori, 1996). Bolivia, meanwhile, provides an important example of how the incorporation of beneficiary participation can lead to more sustainable poverty reduction efforts, and in particular enhance local institutional and organizational capacity. In Bolivia, the demand-based approach is now being extended to reforms in the education sector, in addition to the Social Investment Fund's cooperation with the ministries.

In many other countries, however, social expenditure remains skewed to wealthier groups. As noted earlier, public focus on the social costs of reform and on short-term safety net measures, particularly in the absence of progress on macroeconomic reform, can divert attention from necessary reforms in the mainstream social sectors. Precisely because many safety net programs are implemented outside the mainstream public institutional framework, they avoid addressing difficult problems within it. This underscores the importance of implementing safety nets in a broader context of macroeconomic reform. When safety net measures are implemented effectively, and introduce key principles such as targeting of the poorest and beneficiary participation, they can provide impetus—as well as some guiding principles—for the broader process of social-sector reform.