To date, the crisis in Latin America has been analyzed primarily in macro-economic terms and in terms of its impact on the labor market. Very few studies analyze the behavior of men and women in the job market and no comparative studies on Latin American countries exist on this subject.
This article aims to compare the behavior of men and women in the job market in five metropolitan areas: Bogota, Caracas, Panama City, San Jose, and Sao Paulo. The information for the article is based on household surveys2 .
The available information affords only a partial look at how women's production has been affected by the crisis. Also, an important area of women's involvement which remains unclear is the impact of the crisis on reproduction rates. The information which is available, however, shows that government cutbacks in spending for social services, including health, education and housing, and the considerable increase in military spending has greatly affected the population. Women have been most affected because more women than men are employed in the service sector, principally in education and health, and because they are primarily responsible for the household and for child rearing.
Since the 1950's, Latin American countries have had, with fluctuation, sustained growth according to principal economic indicators. But beginning in 1980, that dynamic changed3 . The elements in the chain effect of the crisis were the fall in net capital income, beginning in 1982, and the increase in external interest payments. In that year, due to the reduced loan flow and foreign investment in Latin America and the substantial flight of capital from some countries, more than half of the current deficit was financed by substantially reducing international reserves4 .
For the whole of Latin America there was a strong drop in gross domestic products (GDP). Between 1982 and 1983 the gross domestic product dropped in absolute terms, for the first time in 40 years. Without considering Brazil, which represents one-third of the regional GDP and weighs strongly in the Latin American figure, total Latin American production increased by only 2% from 1980 to 19865 . Between 1980 and 1986 GDP increased at a lower rate than the population. However, an indicator that more appropriately reflects the fall in standard of living is national income per capita. The Latin American total dropped 14% from 1980 and real value in 1986 was equivalent to its worth in 1976, 10 years earlier.
The economic behavior of the countries was not uniform during this period. Except for Brazil, Colombia, Cuba, and Panama, all other countries showed a negative growth rate between 1981 and 1986. The negative impact was much greater on petroleum exporting countries, due to the drop in the price of crude oil.
The year when the crisis hit Latin America the hardest varies by country. This is because the different countries had reached different levels of production and had different types and degrees of ties to the international economy. Both factors came into play when governments in the region applied, to greater or lesser extent, adjustment policies which in some cases complemented public investment programs and other governmental measures. These adjustment policies fit in the framework of the International Monetary Fund which placed fiscal and monetary restrictions on Latin American governments. These policies required the reduction of fiscal deficits by increasing tax revenues, raising tariffs on public sector enterprises and cutting expenditures. In conjunction with this adjustment period the official exchange rate was raised in many countries, real earnings were reduced and high interest rates maintained.
The impact of the crisis and the adjustment policies were felt in the job market in three main ways6 : 1. a lower rate of employment growth; 2. a change in the type of employment generated, with increases in the informal and service sectors; and 3. a reduction in real salaries.
For the workforce as a whole, unemployment rates went up. Between 1980 and 1985, for some countries, the number of unemployed grew by 48%. Regarding the change in type of employment created by the crisis, three trends were noted: increased informal activity; growth in the service sector; and changes in the composition of employment in the public sector versus the private sector.
The clearest impact of the crisis, however, was the fall in real salaries. Between 1980 and 1985, in some countries, the wage index fell between 12% and 18%7 . This fall was caused by rising inflation and unemployment and underemployment. Nevertheless, the impact was not the same for all sectors. The worst affected were public servants and construction workers, while manufacturing employees suffered less.
The drop in real salaries surpassed the drops in per capita income and GDP. Together with the large reduction of the minimum wage between 1980 and 1985, it is clear that the impact of the crisis was substantially absorbed by the workers.
crisis and unemployment by sex Unemployment was one of the most drastic and rapid consequences of the crisis. Urban unemployment grew during the period from 1978 to 1984. Between 1984 and 1986, with the exception of Brazil, where unemployment was reduced by 50% during that period, unemployment fell only slightly in other countries and then stabilized between 1985 and 1986, although at rates higher than before the crisis. Colombia's unemployment rate continued to rise until 1986.
In 1985, unemployment in the five cities reached magnitudes between 5.2% (Sao Paulo) and 13.6% (Bogota). In comparison by sex, male unemployment rates, at 5 to 10%, were substantially lower than, rates for females, which ranged from 5 to 19%. Thus, in the cities studied, with the exception of Caracas, female unemployment rates were much higher than male unemployment rates. This could be due to the difficulty of measuring female unemployment rates.
Historically, female unemployment rates in the region have been higher than those for males. Information collected by the Regional Employment Program for Latin America (PREALC)8 for six countries shows that between 1970 and 1980 female unemployment rates were higher than those of males for all countries. The greatest difference between those rates was in Panama and the least in Brazil.
The explanation for higher female unemployment is due to the fact that the Latin American economies are not sufficiently dynamic to absorb the growing labor pool. Furthermore, there is an excess supply of female labor for the limited number of jobs women can fill, given the great extent to which the labor market is segregated by sex. Failure to employ females is also justified by the argument that since the majority of women are not heads of households they can "opt" to be unemployed because they are not the principal source of family support. This is the classic argument used by employers to contract less women or to pay them less9 . But this argument is starkly contradicted by the fact that in recent years women have come to represent between 1/4 and, in the Caribbean, 1/3 of heads of households.
The economic situation in Brazil differs from the general regional trend. Until 1986, Brazil was able to resolve trade imbalances without paying the high price of recession as in the rest of Latin America. Before the crisis, Brazil made great progress in terms of per capita GDP and in the creation of new jobs for men and women. However, this was accompanied by notoriously inequitable income distribution and labor market access for different sectors10 . Recent data shows that the impact of the crisis on the labor market, both informal and formal, did not affect the rate of women's participation. To the contrary, women's participation continued to rise during the period of the worst crisis in the Brazilian economy, which affected the formal sector more than the informal sector11 .
Between 1982 and 1985, female unemployment practically quintupled in Bogota (from 4% to 19%), and in Caracas it almost doubled (from 5% to 9%). In Sao Paulo it rose slightly (5% to 6%), while it dropped in Panama City (from 14% to 12%) and in San Jose (9% to 8%). Unemployment rates were higher for young women between 15 and 19 years of age, reaching 30% in Bogota and Panama City.
As female unemployment rates were rising in most of the cities, actual employment rates were increasing12 . This trend is observed in three of the five cities: Bogota, Caracas, and Sao Paulo, and shows, in keeping with an historical trend, that women's participation in the workforce, both as employees and as unemployed workers looking for work, increases when family income decreases. In other words, women's entry into the job market is anti-cyclical, rising in both formal and informal markets, particularly for young women, when there is a drop in household income13 .
The above data show that unemployment rose sharply in Latin America during the crisis, particularly among women and youth, while the capacity of the economy to absorb the labor pool diminished. Differences between the countries is due to the different ways their economic systems have generated new jobs, the adjustment policies the governments have applied, and the amount of public and social investment in each country.
The trend evident since the 1950's of a broadening labor market for women continues14 . But this growth was not enough to absorb the female workforce, which had a high growth rate in the period from 1950-1980. This particularly affected young women entering the market for the first time.
job occupation by sex The crisis has affected the job market by segmenting it according to sex. Over time, many jobs came to be defined as "women's jobs"; the crisis has influenced this process. Between 1970 and 1980, in six countries, the number of jobs increased in the formal sector both for men and women, but particularly for the latter. However, the case was different in the informal sector. In four countries the number of jobs in the informal sector increased for men and remained the same in two others, while women's share of jobs in this sector diminished in all cases and dropped drastically in Paraguay, Ecuador, and Panama. Also, the proportion of women in the entire domestic sector, formal and informal, decreased in all cases15 .
The division of the labor market by sex between 1982 and 1985 shows no spectacular changes. The formal sector shrank slightly in Bogota and San Jose, remained stable in Caracas and expanded in Panama City. Coincidentally, the size of the informal sector increased in Bogota and San Jose and decreased in Caracas and Panama City. The impact of the crisis becomes evident when domestic service activity is analyzed. This sector increased in most of the cities up to 1985.
The distribution of the female labor force in different occupations between 1982 and 1985 was also modified. In the short term, the percentage of working women who were professionals showed a marked increase, while the percentage of working women who were clerical workers and salespersons grew in two cities and dropped in two. The percentage of women working as salespersons, including street vendors, decreased in all cities. Rather than reflecting a drop in the number of women in the informal sector, however, these figures may reveal the "statistical invisibility" of the sector, particularly where street vending is punishable by law. Also, domestic and service workers constituted the largest group of women workers in 1982 as well as in 1985. The only exception is San Jose, where in both years, the largest occupational group of women was that of clerical workers.
In general, the nature of "female" occupations remained the same in Latin America in spite of the crisis and it is not likely that, in the short term, there will be significant changes in the division of work by sex. The types of activities women undertake in the informal sector should be studied in more detail because, as much qualitative research has already shown, in times of crisis, the poorest women who join the workforce are not appropriately included in population censuses or household surveys16 .
income distribution by sex Income distribution in Latin America has always been unequal varying according to the level of a country's development and the social and political model in effect. On the regional level no studies have been carried out which examine the difference in incomes of men and women. Some research has been done at the national level, however, and in all cases show that women's incomes are less than men's, with variations from country to country. In the case of Chile, for example, between 1960 and 1985 women's average income was between 68% and 38% that of men at the same educational level. This ratio has not changed in the last quarter of a century17 .
An analysis of the differential in average income by sex and age reveals that among the youngest workers, the difference is less, and increases with age. Thus, women progress less than men in increasing their income as they gain work experience with the result that the gap widens with age18 .
Looking at the 5 metropolitan areas, in 1985 the female population received between 53% and 84% of the average income earned by males, in the following order: Sao Paulo 52.8%; Caracas 60.8%; Bogota 66.3%; San Jose 79.9%; Panama 83.5%.
Women's lack of skills could account for the difference in the incomes of men and women. However, for the five cities studied there were also large differences between average income for men and women with the same level of education. Interestingly, in all cities there was a great difference in income between men and women at higher educational levels. For example, at the university level (more than 13 years of instruction) the income differences were marked between men and women. The extreme case is Sao Paulo where average male income is double average female income at that level of education. Among the illiterate population, however, the differentiation by sex does not exist and in two cases average female income is slightly higher than male income in that category (Caracas and San Jose).
Whenever the indices for average income are analyzed by occupations for men and women, incomes for males are higher than for females in the same occupation. Again, the greatest differences are observed in average male and female incomes in high-paying professional and management positions.
Heads of households, both male and female, have income levels which are much higher than the average income levels for the general population. However, the difference in income levels between men and women heads of households is much greater than the differences by sex in the general population. This pattern is repeated in every country to a similar degree and reflects a need to analyze the group of women heads of households whose standard of living is among the lowest in the region.
The greatest disparity in female and male income both for heads of households and the population as a whole is in Sao Paulo. While there was a spectacular increase in female employment over the last few years, this came at the cost of much lower salaries for women than men. Thus, women in this instance played the role of a "reserve army" of laborers.
Between 1982 and 1985, the relative average income of women dropped in Bogota and Caracas, remained the same in Panama City and increased only in San Jose where there was also a sharp reduction in employment of women, the opposite of what occurred in Sao Paulo. The women whose incomes dropped the most were those with intermediate educational levels, between 7 and 12 years of instruction. During this period, the difference in average incomes for those with the highest and lowest educational levels tended to become smaller so that the differential in incomes due to level of education is much less in 1985 than it was previously. However, this balance was attained by an overall lowering of average female incomes in the higher educational levels.
With respect to occupations, women who experienced the greatest decline in income were office workers, who had an intermediate level of education, and the self-employed vendors. Average incomes of professionals dropped only in Caracas while in the other cities they remained the same or even increased. A comparison of the average incomes for those in occupations paying the lowest and highest incomes shows that in Bogota the difference between the two categories became greater. In Caracas and San Jose, the difference in salaries between these two groups diminished while it remained stable in Panama City.
A first impression of what occurred as a result of the crisis, therefore, is that incomes of working men and women became more unequal. Among women the crisis has produced a decline in salaries resulting in an overall decrease in income available to them.
conclusions The impact of the crisis has been felt differently by working men and women. It produced higher unemployment among women than men in almost all the cities studied, and increased or maintained the high rates of female unemployment, particularly among the young. In this situation, the activity of women has been "anti-cyclical". As drastic reductions in household income occurred, women entered the formal and informal job markets in increasing numbers. However, as we have seen, while their rate of participation increased, unemployment rose significantly, particularly for those aged between 15 and 19.
The sexual division of the job market did not change greatly as a result of the crisis. The informal sector has grown, even though we do not have an accurate picture of it from household surveys. Domestic work which declined between 1970 and 1980 increased in the period 1982-1985. Therefore, the structure of "female" occupations remains basically unchanged in spite of the crisis and it is unlikely that the sexual division of the labor market will change in the short term.
Last, it is in the income distribution category, that the disparity between workers by sex is most apparent. This inequality has been accentuated by the crisis so much so that average female incomes in some cases are half of the average male earnings. This gap is greater among heads of households. According to level of education and type of occupation, average incomes declined most among women in low level non-manual jobs: self-employed vendors, office workers, etc. which correspond to the intermediate educational levels of 7th to 12th grade education.
In sum, while the crisis affected all workers, it had a particularly negative effect on women. This has specific socio-political implications for women, and particularly for those women who have been hurt the most, namely heads of households and young women.
Many issues remain to be researched on the impact of the crisis on women in Latin America. Among them is the issue of how national budget cuts affected women, the consequences of these for families, and the accompanying development of new forms of social and family organizations. Also, the situation of rural women has been left out of the analysis. Information about their situation by country does not exist, making it impossible to do an adequate assessment. However, it is evident that the multiple problems which have affected rural areas in the last decade have also provoked radical differences in this sector's standard of living.
