¹ Sáinz, J. P. Pérez and Larín, R. Menjívar (eds.), Informalidad urbana en Centroamérica: Entre la acumulacion y la subsistenica (Caracas, 1991)Google Scholar.

² Berger, M., ‘La mujer en el sector informal’, in Berger, M. and Buvinic, M. (eds), La Mujer en el Sector Informal. Trabajo Femenino y Microempresa en América Latina (Quito, 1988)Google Scholar.

³ PREALC, Cambio y polarización ocupacional en Centroamérica (San José, 1986)Google Scholar.

⁴ Sáinz, Pérez and Larín, Menjívar, Informalidad urbana en Centroamérica, p. 24. García-Huidobro, G. (Empleo femenino en Centroamérica, Documento, Panama, 1989)Google Scholar, in an article where delimiting criteria for the informal sector are not explained, provides higher percentages for 1986 and 1987. The highest participation by women in the informal sector is given as 57.2% for Honduras, and the lowest, 41.7% for Guatemala. García, A. I. and Gomáriz, E. (Mujeres Centroamericanas, vol. I, San José, 1989)Google Scholar in their statistical compendium point out that women's participation in total informal employment was 22% in Costa Rica for 1983, 53.3% in El Salvador for 1986, and 36% in Nicaragua for 1985. Again, delimiting criteria for informal employment are not clearly stated. In his regional analysis of female employment, Dierckxens, W. (‘Mujer y fuerza de trabajo en Centroamérica’, Cuadernos de Ciencias Sociales, No. 28, San José, 1990)Google Scholar ignores the formal-informal distinction, so that it is not possible to find estimates.

⁵ In a review of existing work, only three papers published towards the end of the eighties were found. See Menjívar, R. and Saínz, J. P. Pérez (eds), Informalidad urbana en Centroamérica: evidencias e interrogantes (Guatemala, 1989)Google Scholar. There are two Master's theses on the Honduran case, one dealing with food-vending in a Tegucigalpa market (Véliz, A., ‘La venta de comida en el mercado como forma de inserción laboral y reproducción de la fuerza de trabajo femenina’, Master's thesis in Social Work, UNAH, Tegucigalpa, 1988)Google Scholar and another on street hawkers, also in Tegucigalpa (Veras, I., ‘Inserción de la mujer en el sector informal de la economía. El caso de las vendedoras ambulantes de Tegucigalpa (1984–87)’, Master's thesis in Social Work, UNAH, Tegucigalpa, 1988)Google Scholar Female market vendors have also been researched in the Nicaraguan case, with an emphasis on their feminist consciousness (Redondo, A. and Juárez, M., ‘Las vendedoras de los mercados y su conciencia feminista’, Cuadernos de Investigación, Managua, 1987Google Scholar).

⁶ Larín, R. Menjívar and Saínz, J. P. Pérez (eds), Ni héroes ni villanas. Género e informalidad en Centroamérica (San José, 1993)Google Scholar.

⁷ Tokman, V. E., ‘El sector informal: quince años después’, El Trimestre Económico, No. 215 (1987)Google Scholar.

⁸ Tokman, V. E., ‘Dinámica del mercado del trabajo urbano: el sector informal urbano en América Latina’, in Katzman, R. y Reyna, J. L. (eds), Fuerza de trabajoj movimientos laborales en América Latina (Mexico, 1979)Google Scholar; Mezzera, J, ‘Notas sobre la segmentación de los mercados laborales urbanos’, Documentos de Trabajo, no. 289 (Santiago, 1987)Google Scholar.

⁹ Castells, M. and Portes, A., ‘World Underneath: The Origins, Dynamics and Effects of the Informal Economy’, in Portes, A., Castells, M. and Benton, L. (eds), The Informal Economy: Studies in Advanced and Less Developed Countries (Baltimore, 1989)Google Scholar.

¹⁰ Portes, A. and Schauffler, R., ‘Competing Perspectives on the Latin American Informal Sector’, Population and Development Review, vol. 19, no. 1 (1993)CrossRefGoogle Scholar.

¹¹ de Soto, H., El otro sendero (Lima, 1966)Google Scholar.

¹² Portes and Shauffler, ‘Competing Perspectives’.

¹³ Tokman, V. E., ‘Sector informal en América Latina: de subterráneo a legal’, in PREALC, Más allá de la regulación: El sector informal en América Latina (Santiago, 1990)Google Scholar.

¹⁴ Sáinz, J. P. Pérez, Informalidad urbana en América Latina: Enfoques, problemáticas e interrogantes (Caracas, 1991)Google Scholar.

¹⁵ Mezzera, J., ‘Informal sector as PREALC’, Mimeo (1990)Google Scholar. Also published as PREALC Documento de Trabajo.

¹⁶ Of the three countries, the most extreme case is that of Costa Rica, where GDP in 1982 fell-by 7.3%. See Gallardo, M. E. and López, J. R., Centroamérica: La crisis en cifras (San José, 1986), p. 47Google Scholar.

¹⁷ GDP growth rates for 1980 were 10.0% in Nicaragua and 15.1% for Panama, while for 1988 they were — 8.0% and — 16.4% respectively. See. ibid., p. 47; FLACSO/IICA, Centroamérica en cifras (San José, 1991).

¹⁸ Poverty levels for Guatemala, El Salvador and Costa Rica have been estimated through the indirect method of constructing a poverty line, while for Honduras and Nicaragua, they were estimated through the method of unsatisfied basic needs. See Menjívar, R. and Trejos, J. D., La pobreza en América Central (San José, 1992)Google Scholar.

¹⁹ Data for Panama must be treated with caution due to methodological problems. See ibid., p. 75.

²⁰ Ibid.

²¹ PREALC, Cambio y polarización. This would explain the paradox in the Panamanian case, where in 1980 the Gini income distribution coefficient was higher (0.58) than in Guatemala (0.46), Honduras (0.51) and Nicaragua (0.51) in spite of having lower poverty levels. See Gallardo and Lopez, Centroamérica, p. 157.

²² This information has been obtained through regional research carried out in Guatemala City, San Salvador, Tegucigalpa, Managua, San José and Panama City. An area or group of areas, thought to represent the informal phenomenon in each of these metropolitan cities, was identified. A census of the economic units was made and, on the basis of this delimitation of universes, surveys were carried out. Representative samples were designed on principles of proportional allocation according to two kinds of strata. These correspond to the two key analytical dimensions: gender (men and women) and types of economic unit (dynamic, intermediate and subsistence). During a third stage, in-depth case studies were carried out. In this article analysis is limited to data gathered from these surveys in relation to the links existing between informality and gender.

²³ Necessity of income would mirror situations where both individual and family factors would have similar incidences. The residual category, obviously, gathers all instances where the interpretation of causes of mobility in terms of this scale was not possible.

²⁴ A Chi-square test was applied to this hypothesis. Although the different types of causes can be ranked, the existence of the residual category (‘other’) prevented the use of an ordinal scale. 0.5 was the significance level chosen for all hypotheses.

²⁵ Classification criteria for economic units, and therefore for types of informality, have been based on investment (machinery and/or equipment and means of transportation) and management (type of accounting). Thus, the economic units designated as dynamic are characterised by investment in any of the two criteria and by the use of some type of formal accounting. The intermediate situation meets one of these two criteria while subsistence activities are characterised by lack of investment and formal accounting. A third criterion of exclusion was also applied. No establishments with less than two years of existence were considered, since their inclusion would have introduced a bias favouring dynamic informality.

²⁶ The type of economic unit can be regarded as an ordinal variable, since the categories used express degrees of accumulation. The test applied to the hypothesis is therefore the Mann-Whitney U.

²⁷ A domestic labour index has been elaborated on the basis of four basic domestic activities and on the frequency with which they occurred. Cooking and the care of children have each been assigned 30 points, while washing clothes and food buying have been assigned 20 points each. Thus, it is an index whose highest value is 100. Also, when activity frequency is not a daily one, the score for each activity has been halved.

²⁸ The test applied to this hypothesis, since it is a non-categorical variable, is a t-test.

²⁹ This number expresses the quotient between the domestic labour index for women and the same index for men.

³⁰ The estimated Eta coefficients range from 0.744 for Panama City to 0.501 for Managua. Only the Honduran case shows a weaker association (0.304).

³¹ In a case study of working women from popular sectors carried ou t in three Mexican cities, B. García and O. de Oliveira (‘El significado del trabajo femenino en los sectores populares urbanos’, Paper presented at the workshop on ‘Mercados de trabajo: una perspectiva comparativa, tendencias generales y cambios recientes’, Mexico, 23–26 Oct. 1991) conclude that in the majority of situations female insertion in the labour market is due to a need for complementary income. However, these authors have found cases where such participation is also due to personal aspirations.

³² In the Honduran case, this percentage is a mere 7.6 %, but in San José it reaches 40.9%.

³³ In regard to this, the Salvadorean case is the one that shows the longest average working day for women, while it is shortest in the Costa Rican case.

³⁴ In this type of situation, a deterioration of men's self-esteem may occur. The effects may resemble those formulated by de Barbieri, T. and de Oliveira, O. (La presencia de las mujeres en América Latina en una década de crisis, Santo Domingo, 1987)Google Scholar in relation to the greater impact of open unemployment upon men during the crisis, which may result in an increase in domestic violence.

³⁵ This typology was designed by Olga Goldenberg for the Costa Rican case during the regional study.

³⁶ We put ‘entrepreneurial’ in quotation marks because the use of this term, which is found in a certain kind of literature that seeks to underline the inventiveness of the informal workers in facing their daily difficulties, exhibits a certain formal ethnocentrism. That is, an attempt has been made to explain informal rationalities in terms of formal ones.

³⁷ The analysis of variance of the domestic labour index according to types of economic unit shows significant differences in all cases in the sense that when the index is lower (that is, when there is less participation in domestic labour), dynamism will be higher and vice versa.