***************************************************************** This document has been made available in electronic format by the United Nations. Reproduction and dissemination of the document - in electronic and/or printed format - is encouraged, provided acknowledgement is made of the role of the United Nations in making it available. ***************************************************************** UNITED NATIONS Distr. GENERAL E/CN.6/1994/2 2 February 1994 ORIGINAL: ENGLISH COMMISSION ON THE STATUS OF WOMEN Thirty-eighth session New York, 7-18 March 1994 Item 5 (a) of the provisional agenda* PRIORITY THEMES: EQUALITY: EQUAL PAY FOR WORK OF EQUAL VALUE, INCLUDING METHODOLOGIES FOR MEASUREMENT OF PAY INEQUITIES AND WORK IN THE INFORMAL SECTOR Report of the Secretary-General SUMMARY The Commission on the Status of Women, at its thirty-fourth session, selected the issue of "Equal pay for work of equal value, including methodologies for measurement of pay inequities and work in the informal sector" as the priority theme in the area of equality to be considered at its thirty-eighth session. On the basis of a study by the International Labour Organization, which is contained in the annex, the present report describes recent developments concerning the subject and makes recommendations for further action. * E/CN.6/1994/1. CONTENTS Paragraphs Page I. INTRODUCTION ............................. 1 - 4 3 II. CONCLUSIONS AND RECOMMENDATIONS .......... 5 - 23 3 Annex. Study by the International Labour Organization..... 8 I. INTRODUCTION 1. On the recommendation of the Commission on the Status of Women, the Economic and Social Council, in its resolution 1990/15, decided that the Commission should consider, as a priority theme in the area of equality, the question "Equal pay for work of equal value, including methodologies for measurement of pay inequities and work in the informal sector". In the annex to that resolution, entitled "Recommendations and conclusions arising from the first review and appraisal of the Nairobi Forward-looking Strategies for the Advancement of Women", the Council stated that Governments and other appropriate parties should renew their efforts to close the gap between women's and men's pay, possibly by 1995, and should take special measures to address the principle of equal pay for work of equal value, and the United Nations system should complete its work on the methodological aspects of measuring pay inequities between women and men (recommendation V). 2. The Convention on the Elimination of All Forms of Discrimination against Women refers to the principle of equal remuneration, with its multiple and complex links to the position and status of women and men in employment and society. In particular, the States Parties to the Convention are to take all appropriate measures to ensure the right of women to equal benefits and equal treatment in respect of work of equal value, as well as the right to equality of treatment in the evaluation of the quality of work (article 11 (d)). 3. The principle of equal remuneration is also included in various conventions and recommendations of the International Labour Organization (ILO). The Equal Remuneration Convention of 1951 (No. 100) and Recommendation (No. 90) focus on discrimination based on sex and call for equal remuneration for men and women workers for work of equal value with initial application to the public sector. The Discrimination (Employment and Occupation) Convention of 1958 and Recommendation (No. 111) deal with all forms of discrimination in various aspects of employment and are not restricted to the public sector. The Workers with Family Responsibilities Convention of 1981 and Recommendation (No. 156) address discrimination and restriction to employment due to family responsibilities. 4. Long-standing work on the issue has placed the ILO at the forefront of United Nations activities in the area of equal pay for women. For that reason, ILO was requested to undertake the study which is contained in the annex to the present report. II. CONCLUSIONS AND RECOMMENDATIONS 5. Historically, legislation on the issue of pay equity has evolved from an initial emphasis on equal pay for equal work to an emphasis on equal pay for work of equal or comparable value. To begin with, equal work legislation attempted to correct gender-based pay inequalities between identical jobs held by women and men. It would be illegal to pay female engineers less than male engineers but quite legal to pay female nurses less than male truck drivers. 6. Subsequently, equal value legislation went a step further to correct pay inequalities between men's and women's jobs that were substantially different but that could be compared in terms of four criteria of skill, effort, responsibility and working conditions. That step became necessary because experience showed that the major factor in wage differentials between women and men was not unequal pay for the same jobs. 7. While the level of inequality varies from place to place, the pattern of unequal remuneration still persists all over the world. Women everywhere, on average, are paid less than men. In addition, men are more likely to have regular full-time work and receive greater seniority and benefits. 8. In the general discourse about development, there is increasing consensus that without achieving equality between men and women it will not be possible to achieve development objectives: closing the wage gap is a development issue as well as an important mechanism for achieving equality. 9. The pattern of unequal remuneration has its roots in a gender-based division of labour according to which women and men are assigned to perform different tasks in the household, the labour market or the community and the value accorded to those activities is different. The value placed on their respective roles affects the status of men and women in society in terms of their differential access to and control over such resources as income and decision-making power. 10. Although women's work patterns have changed during recent years, employers still assume that women are merely temporary workers who will leave the labour force upon marriage and that their income is a secondary supplement to the family economy. It has been assumed that a man's wage is a family wage, which is supposed to be sufficient to support a male worker and his wife and children. The concept of a family wage has had serious and paradoxical consequences for women. As an ideal, it assumed the validity of the middle-class image of the female role as a housewife, devoted to home and children and removed from productive, market-oriented labour. It thus undermined the wage demands of women in the labour force and also strengthened barriers against female access to more highly skilled craft occupations. In addition, the growth in the proportion of households that are female-headed has multiplied the numbers of women seeking jobs; since they bear the main financial responsibility for children and other dependents, inadequate wages have serious consequences for such women. 11. The association between women's paid labour and their unpaid work in the home has served to provide further justification for the very low pay found in professions whose members are predominantly women. The rule in operation in occupations which are segregated by gender often becomes the basis of new forms of constraints in the form of differential vocational or professional training. Where women and men have been educated or trained differentially, discriminatory employment becomes rational from the employer's point of view. Through such mechanisms, the sexual division of labour is transformed into an apparently technical division of labour, resistant to the more obvious anti-discrimination strategies. 12. The overall gender-based organization of labour tends to concentrate economic benefits in one direction and economic losses in another. Although not all women are losers, the overall benefits and opportunities lost are sufficiently large to merit attention. 13. At a basic level, equal pay for work of equal or comparable value proposes a revision of the ways women and the work associated with them have been valued by the society. It challenges both a purely market-based definition of value (i.e., the value of the good or service is whatever the market will pay for it) and deeply rooted cultural assumptions about the value of women's labour within and outside the home. 14. The introduction of policies based on comparable worth promises a potentially fundamental transformation of women's lives by enhancing their capacity to support themselves and their dependants in a time of deepening female poverty. Women's economic autonomy is important, however, not only because it would relieve the burden of low incomes but also because it would allow women to make life choices more freely (education, work, marriage.) It must be recognized that economic independence is the foundation for full citizenship and participation in the broader community. Political rights have little meaning unless women have the resources to act independently and to shoulder their share of communal responsibilities. 15. In that respect, the issue of equal pay for work of equal value - comparable worth - extends beyond a question of employment. The underlying question is the extent to which individuals can control their own destinies and the extent to which those destinies are the result of larger social forces as they play themselves out in individual lives and through individual choices. As an issue, comparable worth encompasses deep divisions of opinion not only about what workers should be paid and thus how they will live but also about how society will determine what workers should be paid and thus how society justifies the payment decisions. It asks how society values men's and women's contributions and recognizes their respective roles. 16. A major difficulty in achieving equal pay for work of equal value is how to compare the value of jobs performed by men and women. In comparing the value of different jobs, it is important to employ methods and procedures capable of ensuring that the criterion of sex is not directly or indirectly taken into account in the comparison. Particular difficulties for job evaluation are experienced in areas where men and women are in practice segregated into different occupations, industries and specific jobs within enterprises. The ILO study (see annex) describes the steps involved in comparing jobs by means of job evaluations, including the identification of gender-dominated occupations; shows how to conduct a job evaluation; and indicates the relationship between pay and point scores on job evaluation. 17. A second question is how to elaborate methods for implementing the principle of equal remuneration. The difficulties here arise from a lack of knowledge of the true situation concerning inequalities in remuneration, which are poorly researched and not well identified statistically. The ILO study gives a world-wide picture of wage differentials between men and women, explains why they exist and identifies the size, scope and nature of such inequalities, considering historical patterns of industrial development and cultural environment from the point of view of gender. 18. The future situation regarding remuneration and equal pay will be affected by global changes in the context of economic restructuring. The past decade has seen considerable growth in female participation in the labour force. Moreover, the increasing globalization of production and the pursuit of flexible forms of labour to retain or increase competitiveness, as well as changing job structures in industrial enterprises, favour the feminization of employment. As a result, women are gaining an increased share of jobs at many levels. However, in terms of wages, training and occupational segregation, they remain disadvantaged in the new labour markets and the conditions of non-regular employment, which are increasing among both men and women, are still worse for women. 1/ 19. Another factor which contributes to the feminization of the global market is the growing proportion of women entering the largely unregulated informal sector. In developing countries, for example, the majority of jobs created in the export manufacturing sector has gone to women. This is becoming a reality not only for developing countries but for developed countries as well, where the process has taken the form of industrial restructuring and a shift from manufacturing to services. 2/ 20. In contrast to the formal sector, pay and working conditions in the informal sector are unregulated by labour legislation. As a result, the problem of equal pay for equal value in relation to the informal sector becomes even more complicated and raises difficult methodological issues, such as how to compare jobs across establishments, how to define male- and female-dominated occupations, how to avoid gender biases in job evaluation schemes, and most important of all, how to redress inequities in a sector where labour law and regulations are often either not applied, or not enforced, or both. 21. The ILO study offers some thoughts on how it might be possible to transfer the concept of equal pay for work of equal value to occupations and establishments within the informal sector through the use of job content assessments. Although legal action and administrative measures for the most part have no effect on the informal sector, assessments based on job evaluation could provide a useful conceptual frame for analysing and targeting policies aimed at improving women's earnings and status in employment. 22. There is no longer any serious challenge to the notion that women and men should receive equal treatment in employment. Economic considerations, frequently put forward as an excuse for postponing action, are less and less accepted as justifying the perpetuation of injustice. The problem concerns both the developed and the developing countries, although in the developing countries the issue also requires that considerable effort be made to ensure that women benefit from literacy campaigns and basic education and training. 23. However, the problem of how to enforce the principle of equal pay for work of equal value, especially in the informal sector, requires additional consideration and should also take into account emerging trends. In that respect, an expert group meeting to consider the problem may be justified. Notes 1/ See, for example, Guy Standing, "Cumulative disadvantage? Women industrial workers in Malaysia and the Philippines", paper presented at the UNU/WIDER Conference on Trajectories of Patriarchy and Development, Helsinki, 6 and 7 July 1992. 2/ See Bluestone and Harrison, The Deindustrialization of America (New York, Basic Books, 1982); and David Harvey, The Condition of Postmodernity: An Enquiry into the Origins of Cultural Change (London, Basil Blackwell, 1989). Annex STUDY BY THE INTERNATIONAL LABOUR ORGANIZATION I. GENDER INEQUALITIES 1. Widespread gender inequalities in the labour market are receiving increasing attention among researchers and policy makers. The present report briefly considers attempts to explain those outcomes, after first reviewing the evidence on occupational segregation and wage differentials by gender. It observes that patterns of occupational segregation change over time and with industrialization, and are affected by demographic and life-cycle factors, income levels, macroeconomic conditions and institutional factors. There are enough similarities across development levels to indicate that gender roles and the division of labour have some commonalities in the world, while there are enough anomalies to indicate that cultural biases are also important. A. Evidence of gender pay differentials and occupational segregation 2. The ratio of female wages to male wages tends to be about 60 to 70 per cent in the western industrialized countries. According to data for the manufacturing sector from the 1990 ILO Yearbook of Labour Statistics, that ratio ranged from around .50 to .90 in 1990. a/ 3. Occupational segregation varies greatly around the world (see Anker and Hein, 1986; Barbezat, 1993; and Jose, 1987 for reviews). The 1992 Yearbook of Labour Statistics provides information on women's employment distribution by seven major occupation groups: professional and technical; administrative and managerial; clerical; sales; services; agriculture; and production. Such aggregated occupational data reveal considerable segregation by gender but also conceal much segregation. Although there are wide variations globally, women tend to be relatively concentrated in three of the seven groups: clerical, services, and professional and technical workers. In Chile, for example, where women made up 30.5 per cent of the labour force in 1991, they comprised 51.3 per cent of services workers. In Canada, where women represented 45.3 per cent of the labour force in 1991, they comprised 55.7 per cent of professional and technical workers, 80.7 per cent of clerical workers and 56.6 per cent of those in services. By contrast, women tend to be underrepresented in managerial and production- related occupations. 4. A more disaggregated analysis shows even greater occupational concentration of women in what are frequently less attractive positions in terms of pay, status and opportunity for professional and technical advance. For instance, nursing tends to be a female-dominated profession, while the higher-paid and more prestigious occupation of doctors tends to be a male- dominated profession. Furthermore, a recent ILO study (Anker and others, forthcoming) indicates that women are concentrated in a narrower range of occupations than men, a fact that lessens their chances for occupational mobility. For example, of 180 occupations in the Netherlands in 1990, over 50 per cent were male-dominated (defined as at least 80 per cent male). By contrast, less than 20 per cent were female-dominated (defined as at least 80 per cent female), with over 70 per cent of the female labour force employed in that narrow range of occupations. That pattern is in evidence, to a greater or lesser degree, in both developed and developing countries. 5. Other forms of segregation that reinforce women's low earnings are concentration in small businesses, where women employees often earn less than men for the same work, and concentration in casual and part-time forms of employment. There is also limited evidence to suggest that the male-female earnings ratio initially worsens before improving the so-called U-shaped relationship - as industrialization proceeds (Standing, 1982). 6. A study of Lima by MacEwen-Scott (1986) documents how occupational segregation and lower pay for women continue even after women attain higher educational levels than men. MacEwen- Scott finds that almost all obstetricians are women, while there is a slight female overrepresentation for chemists and laboratory technicians. A study conducted by Cohen and House (1993) in Khartoum estimates the proportion of the gender wage gap attributable to occupational segregation at 16 per cent and that attributable to wage discrimination for the same work at 21 per cent, controlling for differences in human capital endowments. Since working women tend to be more educated but also tend to be younger than working men, it appears that experience commands larger wage increments than education. 7. In the informal sector, findings are qualitatively similar, with a concentration of women occurring in many cases, especially in sub-Saharan Africa. Evidence, however, is still subject to much debate, partly because of measurement issues (Anker and others, forthcoming). 8. In Narsapur, India, home-based workers, who are generally women, receive the lowest wages of all categories for work that requires a great deal of skill, such as dressmaking and lace- making (Bhatt, 1985). In West Bengal, India, Bardhan (1989) reports that among casual workers, women receive about 50 per cent of men's daily wage, while among regular workers the ratio is about 30 per cent. For housekeeping, the female to male earnings ratio is 50 per cent for casual workers and 63 per cent for regular workers. Earnings differentials increase with educational level and family income. In fact, daily wages for women with middle to secondary education are about the same as that of illiterate casual male workers. B. Determinants 9. Explaining wage differentials by gender has received much attention in economics. Some economists stress non-labour-market factors, while others stress factors originating in the labour market, on either the supply side or the demand side. In practice, however, those factors are difficult to separate because of the feedback effects that operate between them. 1. Labour-market determinants 10. On the supply side, economists often emphasize differences in human capital endowments. However, male-female differences in educational attainment, timing and duration of work experience, and labour force attachment only partially account for gender wage differentials and gender segregation (Barbezat, 1993). Furthermore, human capital factors themselves depend on expected wages and discrimination in a vicious cycle. For instance, human capital theories start with the assumption that women plan for flexible work arrangements and work interruptions over the life- cycle and therefore choose educational attainment and occupations accordingly (self-selection), that is, they choose occupations that require less education, provide relatively high starting salaries, place a low premium on experience and do not seriously penalize temporary withdrawals. However, that does not explain the lower wages or occupational segregation of women who are well educated and who do not plan any work interruptions. 11. On the demand side, several models compete for an explanation: compensating differentials, efficiency wages, segmented labour markets, occupational crowding and models of discrimination per se. Compensating differential theories state that workers are compensated for unattractive or dangerous working conditions. Efficiency wage theories justify very high rewards for work experience as a productivity-raising device: people work harder to obtain a promotion and the rewards it brings. In addition, since the penalty is high, costly labour turnover and work interruptions are reduced. 12. Segmented labour-market theories distinguish between internal labour markets consisting of progressive jobs - in which efficiency wages prevail and vacancies are mainly filled through internal promotions - and external labour markets characterized by static jobs - mainly filled through external hiring, with high turnovers, little reward for work experience, short promotion ladders and short learning curves. Occupational crowding theories argue that the outcome of women being segregated into a limited number of occupations depresses wages in those occupations because of the plentiful supply of labour that is available. 13. Again such theories still fail to explain why women are crowded into relatively few jobs, why women who do not plan any work interruptions are not hired into the progressive jobs of internal labour markets and do not receive either efficiency wages or promotions, or why men are well compensated for unpleasant working conditions they must tolerate while women are not. 2. Theories of discrimination 14. Theories of discrimination distinguish different sources of discrimination: employers, customers and co-workers. The final result is the same, since the disutility associated with hiring a non-traditional worker translates into lower wages for identical levels of productivity in order to compensate for that "cost" (models of personal prejudice are based on that presented in Becker, 1971). That disutility, of a subjective nature, can result in real costs to the extent that it affects the behaviour of others. For instance, customers may choose to buy their products from a competitor. 15. Another theory of discrimination states that employers engage in statistical discrimination. Unlike the above-mentioned case, in which hiring a non-traditional worker may result in perceived (and real) disutility, in that case employers discriminate because they lack information and hold preconceived concepts concerning the characteristics of individual workers from particular population subgroups. Hiring decisions depend on screening applicants for average characteristics, such as expected (presumably higher) labour turnover of women on average. That has the effect of discriminating against all women, even though many women are above average and perform better than many men. Furthermore, such expectations have a tendency to become self-fulfilling prophecies, since women who are denied promotions and salaries commensurate with their performance and experience may become discouraged and withdraw from the labour force. They also ignore the possibility that many of the supposedly higher costs of women workers to employers may simply not exist; for example the supposedly higher labour turnover and greater number of absences of women workers were not found to be significant in a series of developing country studies, because the higher turnover and greater number of absences of women workers due to maternity, marriage ad child care was roughly offset by male workers' great propensity to switch jobs and get drunk (Anker and Hein, 1986). 3. Non-labour-market determinants 16. Non-labour-market determinants include demographic factors, such as marriage and migration; fertility and related work interruptions; the burden of household responsibilities; education; health; and non-labour-market policies and regulations. But many of those factors themselves depend on expected wages, such as the preference given to boys in the provision of education: there is a problem of circularity in that line of reasoning. 17. It is necessary to acknowledge the role that stereotyping and social norms play in valuing women's work. Psychological factors, perceptions of the feminine and masculine identities, and the association of women and men with such abstractions (misinterpretations of archetypes) fail to recognize the individuality of people. They also rob both men and women of the freedom to choose their occupation based on their actual capacities and personal identity. Research in psychology challenges stereotypes concerning the alleged differences between the sexes in behavioural patterns and innate aptitudes, and questions to the extent to which such differences are genuinely built in or are the result of socialization. There are greater differences among the sexes than between them; stated differently, there is much overlap. The challenge, then, is to devise a more objective set of criteria for evaluating people's work. Such an assessment could assist in planning for reducing male-female pay differentials and hopefully in breaking the circularity of self-fulfilling presumptions. II. WHAT IS EQUAL PAY FOR WORK OF EQUAL VALUE? 18. Equal pay for work of equal value is a widely accepted concept but it is a broader concept than equal pay for the same work or equal pay for broadly similar work. The equal pay for work of equal value concept allows for the comparison of what may be quite different types of work, for example, that of secretaries and that of truck drivers. A. Equal value 19. The gender wage gap can be divided into four components: (a) pay differential within job-cells, for identical work in the same enterprise; (b) pay differentials within establishments, for jobs that are comparable in terms of combinations of skill, responsibility, physical effort and working conditions but, that constitute different occupations; (c) differentials across establishments for comparable work; and (d) different mixes of occupations for men and women (women usually being segregated into low-skill and low-wage occupations, sectors and small firms). 20. Policies for ensuring equal pay for the same work within job-cells have a rather limited potential for redressing the gender wage gap in industrialized countries. One problem in industrialized countries is that employers can get around legislation by relabelling jobs. For instance, a woman might be a secretary and a man a management assistant even though their job content might be similar. In that case, within job-cell pay differentials would be artificially reduced while differences across occupations would be increased. In the less industrialized countries, gender pay differentials for the same work, both on average and within occupations, are sometimes sizeable and therefore the potential for legislative redress seems greater. 21. Comparable worth policies address pay differentials within establishments, which in industrialized western countries are of greater magnitude than pay differentials within job-cells. They were first implemented in the State of Washington, United States of America, although they have since found greater application in Canada. In both countries, they are generally restricted to comparisons between gender-dominated occupations within establishments. Hence, all-female establishments or occupations for which no male comparators of equal value can be found may often be excluded from coverage (Gunderson, 1993). 22. Comparable worth policies, as applied in Canada and the United States of America consist of comparing jobs in different occupations based on job demands. Four steps are involved: (a) choosing the occupations to be covered; (b) evaluating job demands on employees by means of gender-neutral job evaluations and assigning point scores (the usual criteria are skills, physical effort, level of responsibility or accountability, and working conditions); (c) imputing a pay value to the point scores obtained from the job evaluations (estimating shadow wages); and (d) adjusting pay across occupations so that jobs with the same point scores have the same pay. 23. The job evaluation methods that are currently practised in most businesses are often not suitable for comparable worth cases because they are deliberately gender-biased. The evaluation criteria and numerical ratings often used contain a built-in gender bias, whereas the essence of comparable worth is to evaluate men and women's jobs on the basis of a gender-neutral system (Bergman, 1989). When a gender-neutral job evaluation is used, large differences in the pay of male-dominated and female- dominated jobs appear for similar point scores. For instance, in one study a secretary was found to have twice as high a point score as a delivery truck driver, yet she earned 20 per cent less. The procedure raises several technical and methodological difficulties, some of which are discussed in section IV below. B. Equal employment opportunity and facilitating policies 24. Opponents of wage-fixing policies argue that by interfering with the market, inefficiencies result. They favour a second group of policies that target different sources of discrimination and seek to increase the demand for discriminated groups, hence reducing occupational crowding (which lowers wages) in order to bring about a more even distribution across sectors and occupations. 25. Equal employment opportunity policies, including affirmative action, seek to reduce occupational segregation by prohibiting discrimination in the various phases of employment (recruitment, hiring, training, promotion and dismissal). They also often involve setting targets or quotas and changing hiring procedures. Targets and quotas can be set either in general or with reference to the relevant regional labour market (Gunderson, 1993), in which case they do not correct overall segregation but only relative segregation in one firm compared to others in the area. An alternative approach is to give preference to women in hiring and promotion, perhaps whenever they are about equal in qualifications. 26. Facilitating policies are a third group of policies, which can address a number of sources of discrimination, including non- labour-market factors that result in pay inequalities, such as access to education; credit; land ownership; birth control; and other community services. Such policies are most relevant in the context of the informal sector, although other policies are also important. 27. The above-mentioned policies sometimes have limited coverage; for example, they might be limited to comparable work for public service employment in a state or provincial government. Still, they can have a demonstration effect reaching much further: comparable worth, once restricted to public companies and rejected by private corporations in the United States of America as unworkable and undesirable, is now quietly being applied by such major private companies as AT&T, Bank of America and Motorola. They are compelled not by law or by a philosophical change of heart but by pragmatic realities (Berstein, 1986). III. EMPHASIS ON THE INFORMAL SECTOR AND DEFINITION 28. The remainder of the present paper focuses on the informal sector, not because gender wage differentials are necessarily greatest there (evidence to support that is sketchy), but because many poor women are employed there, although it should be noted that the earnings distribution of the formal and informal sectors overlap each other. In fact, there is evidence that people move between the two sectors, sometimes leaving the formal sector to become successful entrepreneurs in the informal sector or moving from the informal to the formal sector. Hence, the informal sector can play a key role in the formation of human capital in developing countries by providing a forum for both the acquisition of skills and experience, and the growth and development of entrepreneurship; it can also act as a residual sector for poor individuals with no other opportunities. Furthermore, large proportions of women and children work in the informal sector and, depending on the definition used, in some countries a majority of its workers are women. The informal sector, particularly in the developing countries, should therefore be the prime target of concern to relieve female poverty, gain a more complete picture of male-female earnings differentials and improve the status of women in the labour market. A. What is the informal sector? 29. There is considerable debate over the definition and characterization of the informal sector. Everybody knows what it is, but a precise definition that is applicable in all settings, is difficult to make. One review study found 50 different definitions (see Haan, 1989). Sethuraman (1981) has suggested 10 criteria. Key characteristics are flexibility and family ownership. The sector is characterized by microenterprises that have limited capital stocks, with much of the employment consisting of self-employment. Informality characterizes several but not all of their features, such as ownership, operation, labour contracts and sources of capital, raw materials, technology and skills). Yet there are numerous exceptions: some enterprises are licensed and regulated; some use imported raw materials; some are owned by educated persons; and some obtain a loan from a formal source of credit. 30. The difficulty with defining and analysing the informal sector arises mainly from its heterogeneity: various distinctions must be drawn in order to give credit to its complexity. The recent International Conference of Labour Statisticians has made an important contribution to the process of defining the informal sector. B. Definition of the informal sector adopted by the Fifteenth International Conference of Labour Statisticians 31. The informal sector was broadly defined by the Fifteenth International Conference of Labour Statisticians as consisting of units engaged in the production of goods or services with the primary objective of generating employment and incomes for the persons concerned. Such units typically operate at a low level of organization, with little or no division between labour and capital as factors of production, and on a small scale. Labour relations, if they exist, are mostly based on casual employment, kinship or personal and social relations rather than on contractual arrangements with formal guarantees. 32. The production unit - not the individual - is therefore taken as a reference point. Production units are defined as units engaged in economic activities as defined by the United Nations System of National Accounts (SNA), that is, irrespective of the number of individuals working in that unit, the type of premises, whether fixed assets are owned, or the seasonality of its activities. The definition places production units in the subsector of the household economy. 33. The household sector consists of unincorporated, family- owned production units. Assets and working capital are owned by the household, not by the enterprises themselves (as are corporations); liability is unlimited. From an accounting point of view, they cannot be clearly separated from other activities of the household to which they belong (no separate and complete sets of accounts are available); otherwise they would be considered as quasi-corporations and could no longer be classified as belonging to the household sector. That fungibility between the production unit and the household extends to other areas as well, such as expenditures, the use of capital goods and equipment (vehicles, tools, premises), financial transactions and the taking out of loans. The Conference definition replaces size standards (used for defining microenterprises) with the criterion of fungibility. Thus, not all microenterprises belong to the informal sector. 34. In the Conference definition, the informal sector excludes domestic employees working on a regular basis in one or a few households but includes self-employed domestic workers rendering their services on an ad hoc basis to a large number of households; it also excludes outworkers who are paid a wage for hours of work (not a piece-rate). b/ As stated above, the Conference takes as its reference not individuals but the production unit and therefore excludes informal sector employees as a production unit. Further excluded are units that are exclusively engaged in subsistence and home production for their own final use. However, if any part of that production is sold on the market, such units belong in the informal sector, irrespective of the regularity of such transactions or the amounts involved. Agricultural activities are also excluded. 35. The Conference definition also covers secondary activities, which it refers to not as enterprises but as establishments; establishments are defined as being engaged in one principal activity. Therefore, secondary activities carried out in the same location (most detailed geographical area) as an enterprise, unless they are similar to the principal activity, are considered as forming a separate establishment within the enterprise. Hence, a household may run several production units or establishments. IV. APPLICATION TO THE INFORMAL SECTOR 36. One way to objectively apply the concept of equal pay for work of equal value is to perform job content analysis. The present section covers the practical aspects of applying the concept, with an emphasis on applying it in the informal sector. As the area of application is new, some of the discussion is speculative, although the consensus of the authors is that the concept is transferable to the informal sector even if legal implementation is not. A. Identification of gender-dominated jobs and selection of job classification system 37. The first step in the process of job content analysis is to select the occupations that are to be covered. In the United States of America and Canada, only gender-dominated jobs are considered under comparable worth, the most common cut-off point being a share of 70 per cent for one of the genders (Gunderson, 1993). In less industrialized countries, where the overall participation of women in the labour force is quite low in some cases, such a cut-off point would need to be replaced with a relative scale. For example, an occupation with 20 per cent more women than the overall labour force could be considered female- dominated, while any occupation with 20 per cent less women could be considered male-dominated (Barbezat, 1993). Or a scale of 50 per cent more and less than the average percentage of female representation in the labour force could be used (Anker and others, 1993). 38. To reduce the incidence of exclusion from coverage under such arbitrary cut-off points, a scale may be further adjusted to accommodate wide disparities in women's share of employment across regions and industrial sectors in a country; it could differ according to location (urban/rural), industrial sectors (agriculture, manufacturing, etc.) or perhaps also by size of enterprises, given that women tend to be concentrated in small enterprises. 39. For application to the informal sector, which consists of small enterprises, job evaluation comparisons must be made across establishments c/ because in small enterprises there are by definition only a few workers, which makes it impossible to find comparator occupations and raises some difficult methodological issues. 40. First, different enterprise characteristics may be associated with differences in job content due to differentials in productivity or in the quality of the goods and services produced. Wage differentials associated with such differences in job content are legitimate, a major reason why comparable worth policies in the United States of America and Canada are usually done for one company or government level. It is therefore necessary to adjust existing occupational classifications for comparison across establishments, as well as to differentiate the type of markets served, such as neighbourhood/city and export markets or poor/wealthier persons. 41. Second, in small enterprises jobs are often multi-purpose in nature because too great a specialization is not possible, which implies that it may be useful to designate multi-purpose occupations. B. Measuring income and work hours 42. A major difficulty, when surveying household enterprises, is how to measure pay per unit of work-time in order to get an equivalent measure of wage rate. For this, it is necessary to measure both the net annual or monthly income and the number of work hours. Unfortunately, conventional survey methods and statistical concepts, such as "work" or "labour force", tend to underestimate women's work. For instance, in a methods test survey of rural households in the state of Uttar Pradesh in India, the reported labour-force participation rate of women was only 32 per cent when the concept of "market labour force" was used, whereas the reported activity rate was 89 per cent when the internationally accepted labour force definition was used. Furthermore, the reporting of female labour-force activity was found to be very sensitive to the types of questions asked, with activity rates ranging from 16 per cent, when respondents were asked about "main activity", to 48 per cent, when asked about "work" and 89 per cent, when asked whether or not various specific activities were performed (Anker, Khan and Gupta, 1988). 43. To measure activity rates and work hours, detailed time-use surveys and activity listings are required. Otherwise, simple questions are likely to lead to underreporting of female work activity. Although time-use surveys based on recall are generally less accurate than direct observation methods, they are much easier and less costly to implement and appear to provide time-use data with an acceptable degree of accuracy (Anker, Khan and Gupta, 1988). 44. The length of the survey recall period and the number of revisits required to control for seasonal variations in work also need careful consideration when collecting time-use data; a distinction should be made between changes in activities and variations in work hours for the same activity. In practice, there is also difficulty with the reporting of activities that are performed at the same time, which is quite common for informal sector work, especially in family businesses and home production. For instance, women often watch their children while producing goods and services for sale. The usual criteria for assigning time when multiple activities are performed is to distinguish between primary and secondary activities or to use a priority criterion, such as giving priority to income-earning activities. In the example of a woman shopkeeper who also watches her children, both the primary activity and the priority activity would be shopkeeping. 45. Next, some measure of net income must be obtained. Several issues arise: (a) estimating the value of payments in kind (a good or service in exchange); (b) adjusting for output that is self-consumed yet none the less an income generated by the business; (c) imputing costs to a production unit and separating them from household expenditures; (d) assigning income from household enterprises to each of the family members working in the business; and (e) accounting for seasonal variations income. 46. For payments in kind, exchange work and unpaid family work, the income equivalent of the work performed is usually assessed by means of valuation methods that fall into two broad categories: evaluating inputs (labour hours) or evaluating outputs (goods and services produced), in both cases minus production costs. Some of those production costs, including the use of assets, must themselves be assigned a monetary value (e.g., self-made tools and equipment). Sales, input costs and self-consumption are usually measured to arrive at a net enterprise income. However, those methods consider only the net income of the enterprise. They do not disaggregate to the level of the individual, as would be necessary in order to examine the pay differentials between men and women. Hence, household enterprise income needs to be divided among individual household members. Hours of work would be a convenient criterion except that it fails to account for differences in productivity and quality of output. Other differences - in levels of responsibility, tasks performed and working conditions - would also be ignored if time inputs alone were used to allocate household enterprise net income among working household members. An alternative procedure would be to assign household income to family members on the basis of point scores obtained from a job evaluation, a considerable improvement over the cruder method of counting only work hours. For example, if there are two persons who work the same number of hours in a family business and one person's point score is 30 per cent higher, she or he would be assigned 30 per cent more of the net family income. Unfortunately, however, that approach is generally not practical, especially on a wider scale. 47. To obtain an accurate measure of value-added (net income) generated by a business, expenditures need to be assigned to either the production unit or the household; otherwise, measurement will be inaccurate. For instance, if a vehicle is used both for the business and for household consumption, its price and cost of maintenance should be partitioned between the two, perhaps based on the amount of time it is being used for each purpose. 48. Again, income, like hours of work, is subject to seasonality and therefore requires appropriate standardization and recording over an extended period of time, although that period need not be as long as a year, as is required for agricultural activities; a one-month reference period is sufficient for most businesses in the informal sector (Khan and others, 1992). C. Job evaluation 49. The usual criteria for job evaluation are: skills, physical effort, level of responsibility and working conditions. In addition, it is important to consider labour relations, especially when comparing jobs in very different settings and enterprises in the informal sector; and gender biases in typical job evaluations warrant special attention. 1. Skills 50. The definition of skills is frequently subject to gender bias, resulting in lower pay for women because their skills are not always recognized as such. Skills are usually defined as acquired aptitudes that should be rewarded with higher pay, whereas innate aptitudes are not usually thought to warrant higher pay. But that distinction is generally difficult to make in practice: most skills build on innate talents which can themselves command higher salaries. The example often cited is men's physical strength; though largely natural (i.e., innate), it is considered and rewarded as a skill. In comparison, Heyzer (1981) explains how many of women's so-called innate skills, which are not well rewarded, are actually acquired through long training at home beginning in early childhood. She mentions dexterity in making laces and embroidery, a physical skill, but one could also cite human relations skills, communication skills, patience and the ability to perform simultaneous tasks and sustain frequent interruptions. d/ 51. A related issue concerns education, training and skills accreditation. The labour market rewards skills that are formally accredited and acquired. Yet women are much more likely to acquire skills informally or to receive less formal education than men are and in the informal sector such skills and education are relatively important. In considering the skills required in the informal sector, information on the tasks performed may be sufficient, supplemented with questions on work history and how long it takes workers to learn the job. However, some information on productivity and the quality of the goods and services produced is also important because certain jobs may seem identical yet require quite different skills because of higher quality standards and more sophisticated production techniques, which would presumably be reflected in higher productivity. 2. Physical effort 52. The criterion of physical effort tends to be limited to physical strength expended and is an area where there is generally significant gender bias in job evaluations and reward/pay structures in favour of men. However, a more complete measure of physical demands would include the strain of performing repetitive tasks; holding the body in constraining positions, such as bending over all day to pick fruits or make carpets; overexerting the eyes because of poor lighting or exposure to glaring light, such as in checking eggs for cracks; sitting all day behind computer screens; and doing minute work, such as silk embroidery. Related dimensions of physical effort include mental exertion, such as concentration and the need for precision, and emotional exertion, such as responding to the need of people requiring care or confronting frustrating work situations (New Zealand, 1991). Such additional forms of physical effort should be recognized and rewarded at least as much as physical strength expended. 3. Responsibility 53. The level of responsibility is frequently measured in fiscal terms, i.e., accountability and sums involved. However, in the informal sector, the sums involved underrate the impact of error on business survival. Other aspects of responsibility that could be important include the ability to work independently without supervision; responsibility for another person's well-being; and, where the job holder has a supervisory responsibility, direct responsibility for quality control, information and material resources (New Zealand, 1991). 4. Working conditions 54. Working conditions are usually measured in terms of dirt, extreme temperatures, noise levels and vibrations. Again gender biases creep in: the kind of poor working conditions that women are often asked to tolerate, such as in handling faeces, blood and vomit in the case of nurses or in cleaning toilets and soiled clothes in the case of maids, are often ignored (OECD, 1991); moreover, such conditions also involve contact with infectious diseases. 55. Many other criteria for gauging hazards and work environment could be used for a more complete and relevant measure of working conditions, particularly in the context of the informal sector, including health hazards, sexual harassment, moral hazards, and irregularity of unemployment and earnings; such conditions can often be inferred by the status of labour relations and the presence or absence of labour contracts. 56. It is clear that working conditions in the informal sector vary greatly among different occupations and enterprises and can be very detrimental to workers' health, future earnings capacity and mental well-being. 5. Labour relations 57. Standing (1983) offers an interesting set of criteria to distinguish various labour relations. e/ Such relations affect the vulnerability of workers and hence the risks they must assume and tolerate; they can also result in distorted wage levels. In extreme cases, involving forced labour and bonded labour, the relationship is akin to slavery. 58. The preceding discussion of suggested criteria for job evaluation is by no means exhaustive. Job content varies so greatly that other criteria are also relevant, while in certain cases some of those indicated here could be redundant. Some prior assessment could help devise a suitably detailed job classification scheme and list of criteria. D. Relationship between pay and score points 59. A monetary value can be assigned to point scores from a job evaluation in two ways: either by comparing jobs with the same total score points (the so-called point-by-point comparison), or by fitting wage baselines. In both cases, a wage must be selected as the norm in order to provide a standard, to which the wage of the undervalued/underpaid female occupation can be adjusted. For example, if nurses (female-dominated occupation) and architects (male-dominated occupation) had the same point score and architects earned 30 per cent more, then it would be necessary to raise nurses' pay by 30 per cent (assuming that one did not want to lower architects' pay and that such a measure was enforceable). 60. A second, more accurate procedure consists in estimating shadow wages by regressing point scores on observed (actual) wages. When this is done separately for male and female- dominated jobs, two wage baselines are obtained, which can then be used for proportional comparisons. For instance, as mentioned above, if no male comparator with the same total point scores was available in a particular job evaluation (which would be required for a point-by-point comparison) and a woman's job score was 60 per cent of a man's, then the woman's job could be assigned that proportion of the man's pay. The appropriate value could be read off the estimated wage baseline for male-dominated occupations. 61. Such proportional evaluations cannot be made using the first method, point-by-point comparison; such comparisons must be restricted to jobs with the same total point scores. Moreover, if the male comparator job is itself underpaid relative to other male-dominated occupations (below the wage baseline), then the corresponding woman's job will also be undervalued. V. STRATEGIES FOR REDUCING PAY DIFFERENTIALS BY GENDER IN THE INFORMAL SECTOR 62. Within the framework of equal pay for work of equal value policies, job evaluations set standards for remodelling wage systems based on greater equity. Administrative measures and legal action provide the tools for implementation. In the context of the informal sector, however, there is limited scope for such measures, if any at all, because it largely escapes regulations and the poor have only limited access to legal action. Nevertheless, job evaluations in the informal sector could be useful tools for identifying which occupations are the most undervalued and they could provide valuable information on the sources of wage differentials. They might also help challenge stereotypes by providing a more objective assessment of people's work. A. Supporting microenterprises and self-employed women 63. Many women are self-employed in the informal sector and many others are interested in starting their own business. Such women can be helped by providing an enabling environment in terms of credit, regulations etc. as well as by providing direct assistance, such as training, technical advice and loan guarantees. f/ Policies and programmes to address the needs of women operating in the informal sector could focus, inter alia, on: (a) Improving infrastructure, by, for example, providing piped water, electricity, roads and transportation; (b) Providing more supportive institutions and regulations, by, for example, reducing police harassment and allowing women to own property and take credit; (c) Improving macroeconomic conditions, by, for example, eliminating restrictions on access to import and export markets, and reducing subsidies for the formal sector that put the informal sector at a competitive disadvantage. 64. Such policies or programmes can be general, for the benefit of all informal sector businesses, or they can be targeted on informal sector businesses run by women. Direct assistance should focus especially on training, including basic literacy for women without education. Training should also cover key aspects of running businesses, such as marketing, accounting and purchasing. The ILO, for example, has run very successful training programmes in the Czech Republic and in Slovakia for women entrepreneurs; evaluation reports show that of the 10,000 women who attended a two-day course, 30 per cent were still in business after six months (Paukert, 1993). B. Organizing women 65. In many cases, women's wages are unduly low because women lack bargaining power. Women generally work in sectors characterized by a highly elastic supply of labour and highly competitive firms themselves facing very intense competition, conditions which do not favour their organization. 66. Nevertheless, there is some scope for organizing women. For instance, Strassmann (1987) reports on a home-based enterprise in Colombo, Sri Lanka. Another example is SEWA in India, which has been successful in organizing home-based piece-rate workers, setting up trade unions and obtaining legal protection through the courts. Similar projects have been implemented by the ILO in rural India (ILO, 1989). 67. Organizations for women workers can be a key player in articulating women's needs in the informal sector and in many cases they have acted as catalysts and focal points for self-help initiatives. They can thus become useful bridges for channelling assistance and pooling local resources. C. Facilitating policies for improving access to employment 68. More generally, the earning capacity of women and their access to employment can be enhanced in numerous ways by facilitating policies that influence non-labour-market determinants of pay in order to provide, inter alia, (a) Access to education; (b) Family planning, resulting in the reduction and spacing of births; (c) Improvements in health and nutrition; (d) Transfer to the market of domestic chores that would significantly relax time constraints on women, by such means as the provision of day-care centres; (e) Changes in laws, regulations and customs that constrain women's physical movements. 69. Last but not least, much can be done to change perceptions about gender roles and the stereotyping of jobs. Everyone is affected - workers, employers, trade unions, Governments and the public at large. Most people are not even aware that they use double standards and gender stereotyping. All too often, jobs and occupations are restricted to men either on the mistaken assumption that women cannot take responsibility or are too weak, or on the basis of some other erroneous stereotype. In that respect, job evaluation can play a role in more objectively assessing people's work and challenging such stereotypes. It can also provide useful information about the sources of pay differentials. Changing occupational stereotypes also raises interesting questions about alternative ways of doing things, alternative styles and their effectiveness; but those avenues can be opened only by more information, sensitization and greater tolerance. VI. CONCLUSION 70. The present paper has been concerned with equal pay for work of equal value, a widely accepted concept of equity included in national laws and international standards, such as ILO Convention 100. That concept is broader in scope than the concept of equal pay for the same or very similar work, in that it allows for the comparison of pay across dissimilar occupations. That broader scope is important because male-female pay differences for the same or very similar work are now believed to be fairly insignificant in the developed countries, although they apparently remain significant in some developing countries. 71. Data have been presented demonstrating that there are large gender differentials in the labour markets of all countries. Women receive considerably lower pay than men on average. Also, labour markets are highly segmented by gender in all regions of the world, with many occupations more or less reserved for one gender and unavailable to the other; at the same time, the choice of occupations available to men is far larger and more attractive in terms of remuneration and opportunity for promotion. 72. Those labour-market inequalities can be explained by various economic and non-economic theories. One such economic theory is the human capital model, which stresses the lower levels of human capital that women both bring to the labour market (e.g., less education) and acquire in the labour market (e.g., fewer years of experience). That reasoning is used to explain and all too often justify the lower pay that women receive and their segregation into less skilled, lower paid jobs with few opportunities for promotion. On the other hand, labour-market theories of efficiency wages, bargaining power and compensating differentials argue that the jobs that women hold pay less because women tend to work in non-unionized and smaller-size enterprises. A third line of reasoning is known as the occupation crowding theory: it stresses that the restricted job opportunities available to women crowds them into relatively few occupations, causing earnings in those female-dominated occupations to be depressed. However, notwithstanding those economic theories, there is little doubt that non-economic theories related to societal and cultural perceptions and standards are also very important. They help to explain why so many occupations are considered to be "correct" for one gender and not the other; why the labour market apparently undervalues the skills that women acquire in the home; why families and society under-invest in women's education and training; and why family and household responsibilities, which greatly constrain and burden working women, are not shared equally by men. 73. Comparable worth policies, as applied in the United States of America and Canada, have been described. Those policies have resulted in sizeable settlements, typically involving wage increases on the order of 20 per cent for workers in female- dominated occupations; they have, however, been limited by practical considerations, generally being restricted to large enterprises and the public sector. 74. The paper then considered possible approaches to applying the concept of equal pay for work of equal value to the informal sector. Discussion began by describing how the recent Fifteenth International Conference of Labour Statistics defined the informal sector. It went on to discuss how four aspects of comparable worth cases might be applied in the informal sector: identifying gender-dominated jobs; measuring net annual income and hours of work for the self-employed; performing job evaluations to eliminate gender biases; and examining the relationship between pay and job evaluation point scores. While it is clearly impossible to apply job evaluations and comparable worth policies in the informal sector in the way that they are being implemented in the industrialized countries, since the informal sector is not subject to regulatory mechanisms, those procedures never the less offer useful measuring rods for identifying especially undervalued informal sector occupations and thus women in need of assistance. 75. Finally, the paper described some labour-market and non- labour-market policies and programmes that can help improve the position of women in the informal sector by improving women's skills through training and education; assisting women entrepreneurs through direct assistance, such as guaranteeing loans, providing training and helping with regulatory obstacles; organizing women workers and businesswomen in order to increase their bargaining position; and providing facilitating services, such as water, birth control, electricity and fuel, in order to reduce women's additional burden of employment and family responsibility. 76. Last but not least, there is a need to educate and sensitize everyone - workers, Governments, employers, trade unions, and the general public - in order to change traditional stereotypes about what work women should and should not do and thereby help open up the world of work to being more equal for men and women. Notes a/ A study by Horton (1993) for seven Asian countries decomposes gender wage differentials into those which can and cannot be ascribed to differences in human capital endowments. There is no clear trend over time and the unexplained wage differences are large, ranging from 127 per cent for urban workers in the Philippines (1988) to -4.5 per cent for self- employed workers in India (1989/90). b/ The Conference recommended that domestic workers be included in a separate subcategory to facilitate international comparisons. Also, the decision whether or not to include self- employed domestic workers in informal sector statistics was left to the discretion of individual countries. c/ Extending job evaluations to cover different establishments also avoids the exclusion from coverage of all- female and all-male establishments. d/ See New Zealand (1991) for a job evaluation scheme specifically aimed at providing gender neutral assessment of job content. e/ For labour contracts, Standing (1983) suggests a number of detailed criteria, namely, duration: casual, daily, piece work, fixed term, "permanent"; basis of payment: time rate (hourly, etc.), piece rate, product share; frequency of payment: at regular intervals, upon completion of job, irregular, including timing of bonuses; medium of payment: cash, kind, meals etc. (provision of a service, work, or favour); relative autonomy: timing and amount of work, and intensity of effort/productivity per-specified or not; freedom of employment relations: ability to change employers, ability to combine with other (work) activities; linkages with other contracts: credit, land, employment of family members. f/ The informal sector does not necessarily operate outside existing regulations. In some countries, a sizeable proportion of informal sector enterprises are registered or pay taxes (Haan, 1989). In theory, they are also required to observe safety and health regulations, statutory minimum wages, working hours, restrictions on child labour, and other labour-related regulations (Tesfachew, 1992). References Anker, Richard (1994). Measuring women's participation in the African labour force. In Gender, Work and Population in Sub- Saharan Africa, A. Adepoyou and C. Oppon, eds. London: Currey Press. , and others (forthcoming). Occupational segregation around the world. Geneva: International Labour Organization. , and Catherine Hein (1985). Why third world employers prefer men. International Labour Review, vol. 124, No. 1. , and Catherine Hein (1986). Sex inequalities in third world employment: statistical evidence. In Sex Inequalities in Urban Employment in the Third World, Richard Anker and Catherine Hein, eds. London: Macmillan Press. , M. E. Khan and R. B. Gupta (1988). Women's participation in the labour force: a methods test in India for improving its measurement. Women, Work and Development, No. 16. Geneva: International Labour Organization. Barbezat, Debra (1993). Occupational segmentation by sex in the world. Working paper prepared for the Interdepartmental Project on Equality for Women in Employment. Geneva: International Labour Organization. Bardhan, Pranab (1989). Poverty and employment characteristics of urban households in West Bengal, India: an analysis of the results of the National Sample Survey, 1977-1978. In Urban Poverty and the Labour Market, Gerry Rogers, ed. Geneva: International Labour Organization. Becker, Gary S. (1971). The Economics of Discrimination, 2nd ed. Chicago: University of Chicago Press. Bergman, Barbara R. (1989). Does the market for women's labour need fixing? Journal of Economic Perspectives, No. 3 (Winter). Bernstein, Aaron (1986). Comparable worth: it's already happening. Business Week (28 April), pp. 52-56. Bhatt, Ela (1985). "Home-based women workers". Asian Labour, No. 33 (212). Blau, Francine D., and Marianne A. Ferber (1992). The Economics of Women, Men, and Work, 2nd ed. Prentice-Hall. , and Lawrence M. Khan. The gender earnings gap: learning from international comparisons. American Economic Review (May). Burton, Clare, Ravan Hag, and Gay Thompson (1987). Women's Worth: Pay Equity and Job Evaluation in Australia. Canberra: Australian Government Publishing Services. Camazon, Daniel, Guillermo Garcia-Huidobro and Hugo Morgado (1989). Labour market performance and urban poverty in Panama. In Urban Poverty and the Labour Market ... Cohen, Barney, and William J. House (1993). Women's urban labour market status in developing countries: How well do they fare in Khartoum, Sudan? The Journal of Development Studies, vol. 29, No. 3. De Los Angeles Crummet, Maria (1988). Rural women and industrial home work in Latin America: research review and agenda. Working Paper of the Rural Employment Policy Research Programme (WEP 10/WP.46). Geneva: International Labour Organization. Dessing, Maryke (1990). The urban informal sector in developing countries: labour supply and the family. Ph.D. dissertation (No. 468), Graduate Institute of International Studies, Geneva: University of Geneva. , (1990). Support for microenterprises: lessons for sub-Saharan Africa. World Bank Technical Paper, No. 122 (Africa Technical Department Series). Washington, D.C.: World Bank. Economist (1992). Women in Management. 28 March-2 April. Eyraud, Fran?ois (1993). Equal pay and the value of work in industrialized countries. International Labour Review, vol. 132, (1). Fischer, Stanley, and others (1988). Economics, 2nd. ed. New York: McGraw Hill. Flanagan, Robert J., and others (1985). Economics of the Employment Relationship. Glenview, Illinois: Scott, Foresman and Company. Goldin, Claudia (1990). Understanding the Gender Gap: An Economic History of American Women. New York: Oxford University Press. Goldschmidt-Clermont, Luisella (1982). Unpaid work in the household. Women, Work and Development, No. 1. Geneva: International Labour Organization. , (1990). Economic measurement of non-market household production: relating purposes and valuation methodologies. Working paper of the Population and Labour Policies Programme (WEP 2-21/WP.174). Geneva: International Labour Organization. Gunderson, Morley (1993). Comparable Worth and Gender Discrimination: International Aspects. Geneva: International Labour Organization. Guyer, Jane I. (1981). Household and community in African studies. African Studies Review, No. 24 (June-September). Haan, Hans (1989). Urban Informal Sector Information: Needs and Methods. Geneva: International Labour Organization. Heyzer, Noleen (1981). Women, subsistence and the informal sector: towards a framework of analysis. Discussion paper of the Institute of Development (No. 163). Brighton: University of Sussex. Horton, Susan (1993). Women and industrialization in Asia: overview of 7 country studies. Institute for Policy Analysis. Toronto: University of Toronto. Hundley, Greg (1993). The effects of comparable worth in the public sector on public/private occupational relative wages. The Journal of Human Resources, vol. 28 (2). International Labour Organization (1989). Employment Opportunities for Rural Women Through Organisation. Programme on Rural Women, Rural Employment Policies Branch. Geneva: International Labour Organization. , (1993a). Active labour market policies in a wider policy context. Committee on Employment and Social Policy, 258th session. Geneva: International Labour Organization. , (1993b). Report III: Statistics of employment in the informal sector. Fifteenth International Conference of Labour Statisticians (19-28 January). Geneva: International Labour Organization. , (1986). Equal remuneration: general survey by the Committee of Experts on the application of conventions and recommendations. International Labour Conference, 72nd session. Geneva: International Labour Organization. , (1988). Equality in employment and occupation: general survey by the Committee of Experts on the application of conventions and recommendations. International Labour Conference, 75th session. Geneva: International Labour Organization. , (1991). The dilemma of the informal sector: report of the Director-General. International Labour Conference, 78th session. Geneva: International Labour Organization. , (1993c). Workers with family responsibilities. International Labour Conference, 80th session. Geneva: International Labour Organization. Jose, A. V. (1987). Employment and wages of women workers in Asian countries: an assessment. ILO/ARTEP, No. 643. New Delhi: ILO/ARTEP. Khan, M. E., and others (1992). Methodological issues in collecting time-use data for female labour force. Indian Journal of Labour Economics, vol. 5, No. 1. Macewen-Scott, Alison (1986). Economic development and urban women's work: the case of Lima, Peru. In Sex Inequalities in Urban Employment in the Third World, Richard Anker and Catherine Hein, eds. London: Macmillan Press. New Zealand (1991). Equity at Work: An Approach to Gender Neutral Job Evaluation. Department of Labour and State Services Commission. Oppong, Christine (1993). ILO standard setting, policy studies and technical cooperation relating to population issues and women. Working paper on labour and population. (POP/WP.185). Geneva: International Labour Organization. Organisation for Economic Cooperation and Development (1991). Equal pay for work of comparable worth: the experience of industrialized countries. Labour Market and Social Policy Occasional Papers, No. 6. Paris: Organisation for Economic Cooperation and Development. Paukert, Liba (1993). Women in the labour market in the CSFR: plan of action for women's employment during the transition to a market economy system. Geneva: International Labour Organization. Rubery, Jill, and Colette Fagan (1993). Occupational segregation of women and men in the European Community. Manchester: Manchester School of Management. Schultz, T. Paul (1990). Women's changing participation in the labour force: a world perspective. Economic Development and Cultural Change, 38 (3). Schwartz, Felice N (1989). Management, women and the facts of life. Harvard Business Review, (January-February). Seager, Joni, and Ann Olson (1986). Women in the World: An International Atlas. London: Pan Books. Sethuraman, S. V. (1981). Urban Informal Sector in Developing Countries: Employment, Poverty and Environment. Geneva: International Labour Organization. Standing, Guy (1982). Labour Force Participation and Development, 2nd ed. Geneva: International Labour Organization. , 1983. A labour status approach to labour statistics. Working paper of the Population and Labour Policies Programme (WEP 2-21/WP.139). Geneva: International Labour Organization. , 1989. Global feminisation through flexible labour. Working paper on labour market analysis and employment planning. (WEP 2-43/WP.31). Geneva: International Labour Organization. Stone, Nan (1989). Mother's work. Harvard Business Review (September-October). Strassmann, Paul W. (1987). Home-based enterprises in cities of developing countries. Economic Development and Cultural Change, vol. 36, No. 1. Tesfachew, Taffere (1992). Government policies and the urban informal sector in Africa. Working paper of the Urban and Informal Sector Programme (WEP 2-19/WP.59). Geneva: International Labour Organization. Taylor, Debbie (1985). Women: an analysis. In Women: A World Report. New York: Oxford University Press. Tilly, Louise A., and Joan W. Scott (1987). Women, Work and Family. New York and London: Methuen. Terell, Katherine (1992). Female-male earnings differentials and occupational structure. International Labour Review, 131 (4-5). Treiman, Donald, and Heidi Hartmann, eds. (1981). Women, Work, and Wages. Washington: National Academy Press. -----