NBER Reporter: Spring 2001
Thirteenth Annual Inter-American Seminar on Economics
The NBER's Thirteenth Annual Inter-American Seminar on Economics took place in Santo Domingo, Costa Rica on November 16-17. Sebastian Edwards, NBER and University of California, Los Angeles, and James J. Heckman, NBER and University of Chicago, organized this year's program around the topic "Micro-data Research in Latin America":
Heckman and Pages document the high level of job security protection in Latin American labor markets and analyze its impact on employment. They show that job security policies have a substantial impact on the level and the distribution of employment in Latin America, reducing employment and promoting inequality. The institutional organization of the labor market thus affects both employment and inequality.
Saavedra and Torero analyze the effect of several aspects of labor legislation that were modified in Peru after 1991. Through the progressive elimination of job stability regulations, the introduction of temporary contracts, and changes in the severance payment structure, firing costs diminished sharply. Simultaneously, nonwage labor costs increased. Using data on 10 sectors observed bimonthly between 1987 and 1997 and data on establishments for 3 subperiods, the authors find that labor costs have a negative and significant effect on labor demand. The coefficient of their measure of firing costs, the expected severance payments, is negative and significant, and it decreases in the post-reform period. After the reforms, the price and output elasticities are larger and there is evidence of a speedier adjustment in labor demand. Using household surveys for Lima, the authors also find that mean tenure has fallen since 1992. The fall is larger--and statistically significant--for formal salaried workers than for informal workers.
Cardenas and Kugler examine the reduction in firing costs on worker turnover in Colombia. Its labor market reform of 1990, which reduced severance payments substantially, affected worker flows into and out of unemployment and reduced severance payments for all workers hired after 1990 and for those covered by the legislation (formal sector workers). The Colombian Household Surveys provide information about formal and informal sector activity and allow for estimating hazard rates for formal and informal workers, before and after the reform. The results indicate that hazard rates into and out of unemployment increased after the reform for formal sector workers (covered by the legislation) relative to informal workers (uncovered). Moreover, the increase in worker turnover was greater among younger, more educated workers employed in larger firms who were likely to have been affected most by the changes in the legislation.
The U.S. temporary help services (THS) industry grew at 11 percent annually between 1979 and 1995, five times more rapidly than nonfarm employment. Contemporaneously, courts in 46 states adopted exceptions to the common law doctrine of employment at will, limiting employers' discretion to terminate workers and opening them to litigation. Autor assesses whether the decline of employment at will and the growth of THS are causally related. To aid the analysis, he considers a simple model of employment outsourcing that shows that firms will respond to externally imposed firing costs by outsourcing positions requiring the least firm-specific skills rather than those with the highest expected termination costs. Author's analysis indicates that one class of exception, the implied contractual right to ongoing employment, led to 14 to 22 percent growth in excess temporary help in adopting states. Unjust dismissal doctrines did not significantly contribute to employment growth in other business service industries. The decline of employment at will explains as much as 20 percent of the growth of THS between 1973 and 1995 and accounts for 336,000 to 494,000 additional workers employed in THS on a daily basis as of 1999.
Income inequality in Latin America has increased over the past several years. Attanasio and Szekely ask whether there is a supply-side explanation for this experience. Specifically, they explore whether the dynamics of inequality relate to life cycle or to cohort effects, and they ask what underlying household decisions are behind these cohort and life-cycle patterns. They also assess the role played by labor force participation, fertility, household arrangements, household formation, investment in education, and other related family decisions, and compare these with the role played by changes in the returns to schooling.
Over the last 40 years, Peru has achieved significantly lower fertility and mortality rates, bringing population growth rates down to less than 1.2 percent per year. These improvements have led to a demographic transition with lower dependency ratios. Saavedra and Valdivia find that household size is now smaller for the younger cohorts, in all but the households with less-educated heads. To explain these differences, they argue that reductions in fertility have not yet reached the less-educated. However, family living arrangements appear to change throughout the life cycle; for example, extended families are more common for households with very young (under 25) or older (over 60) heads. The authors also show that intergenerational family arrangements over time limit the ability of the life cycle hypothesis to explain household saving behavior. Peruvian households, especially the less-educated, smooth consumption over the life cycle, not only through the typical saving-dissaving mechanism, but also by smoothing income. Net cash transfers, or living arrangements between parents and their offspring, play an important role in this income smoothing.
Neal presents an economic model that captures Wilson's general hypothesis -- marriage markets matter -- while simultaneously fleshing out details of the interactions between sex ratios, male income, welfare generosity, and other factors that may influence nonmarital fertility. Neal's model highlights how marriage markets and government aid programs interact to determine choices about marriage and fertility. It shows how the existence of government aid shapes the effects of changes in marriage market conditions. Neal's model also illustrates how commonly used regression models may yield inferences about the links between marriage market conditions and observed family structures.
Behrman, Sengupta, and Todd study the effects of a new antipoverty program in Mexico called PROGRESA that provides families with monetary transfers contingent upon their children's regular attendance at school. The benefit levels are intended to offset the opportunity costs of not sending children to school and vary with the grade level and gender of the child. The authors show that the program effectively reduces drop-out rates and facilitates progression through the grades, particularly during the transition from primary school to secondary school. Their results also indicate that if children were to participate in the program between ages 6 and 14, the program would increase average educational attainment levels by 0.6 years and increase the share of children attending secondary school by about 19 percent.
Bravo, Contreras, and Crespi establish a method of evaluation tailored to training courses for small-scale entrepreneurs. They also develop measures of training success and design a questionnaire. They propose a pilot evaluation; document the selection of the samples (trained by FUNDES Chile and a control group); elaborate on the questionnaire and its application; and finally report their results.
Bravo and Contreras then offer empirical evidence on the impact of the introduction of a monetary incentive in the Chile Joven (youth labor training) Program. This program, initiated in 1991, focused on unemployed low-income youths between the ages of 15 and 24. Its goal was to raise the probability of employment of youths through job training and work experience. Starting in 1995, a financial incentive was added: an additional quantity of money went to training organizations for each beneficiary of the program who carried out his work experience under a labor contract in a private firm. The authors estimate that the probability that the beneficiaries will finish the work experience phase of their program is 11 percent. However, after controlling for differences in age, gender, occupational status, educational level, and geographical location of the beneficiaries, and for characteristics of the training received, and differences in the macroeconomic environment in a particular year, the impact of the monetary incentive appears closer to 8 percent. On the other hand, when the authors estimate the impact using Matching Techniques, which allows the construction of a better control group, the impact of the incentive increases to 13 percent -- again, measured as the probability of the change in the labor status.
These papers will be printed in a special issue of the Journal of Development Economics.