World Development, Vol. 17, No. 7, pp. 1077-1095,1989. 0305-750X/89 $3.00 + 0.00 Printed in Great Britain. fQ 1989 Pergamon Press plc

Global Feminization through Flexible Labor

GUY STANDING* International Labor Organization, Geneva

Summary. -The emphasis on structural adjustment and labor flexibility in both developing and industrialized economies is rapidly altering the nature of employment. Women are being substituted for men and many forms of work are being converted into the kinds of jobs traditionally geared to women. International data on recent trends in female economic activity reveal that new types of labor data are needed to highlight the mechanisms of control over workers and the actual economic forms of vulnerability to which women are exposed. The paper concludes by posing questions that challenge the traditional concerns of donors and policy makers in many developing countries.

1. INTRODUCTION 2. LABOR IMPLICATIONS OF SUPPLY-SIDE ECONOMICS: The 1980s might be labeled the decade of labor THE CULT OF INSECURITY deregulation. It has also marked a renewed surge of feminization of labor activity. For reasons to For most of the 20th century, and particularly be considered, the types of work, labor relations, after 1945, the dominant development model income, and insecurity associated with “women’s can be described as “social adjustment,” with work” have been spreading, resulting not only in a redistributive welfare state as the long-term a notable rise in female labor force participation, objective. It was to be achieved through a diverse but in a fall in men’s employment, as well as a array of labor rights, protective legislation, and transformation - or feminization - of many other forms of security, and a larger role for the jobs traditionally held by men. It is no coinci- public sector in economic and social policy. In dence that this shifting pattern has been closely the 1970s that model ran into trouble, first losing related to an erosion of labor regulations. There its legitimacy and then being displaced from has been explicit deregulation, whereby formal intellectual hegemony by the early 1980s.’ regulations have been eroded or abandoned In considering the changing economic situation by legislative means, and implicit deregulation, and its effects on women, six developments whereby remaining regulations have been made have been critical. First, whereas previously less effective through inadequate implementation trade took place between countries or regions or systematic bypassing. with similar costs (or labor rights) and was a fair- To elaborate on this thesis, it is necessary to ly small percentage of most countries’ Gross trace the emergence of the supply-side politico- National Product (GNP), in the 1970s the global economic agenda that has dominated policy economy became far more open to internation- making in most of the world in the 1980s. This ally competitive trade, as various low-income agenda, it will be argued, has led to a series of countries became producers of exports and changes in women’s economic roles, increasing potential exports. A second factor was that their use as workers but weakening their income by the late 1970s sustained Keynesian demand and employment security in both low-income management had become associated with rising industrializing and industrialized countries. By inflation, and internationally had contributed focusing on the global spread of flexible labor to excessively rapid lending. The result was practices and the supply-side “structural adjust- indebtedness, as deflationary policies were ment” development strategy, it will be argued adopted in the industrialized countries in the that existing policies - and the data on which mid- and late-1970s. A third related factor was they are based - are inadequate, and that specific alternatives offer far more hope of *Thanks are due to Loretta de Luca, Jo-Ann UnRuh, benefiting working women and men in the com- Caren Grown, and Frances Williams for assistance and ing decade. comments.

1077 1078 WORLD DEVELOPMENT

that the welfare state became perceived by some the leaders have become the led; international as “crowding out” productive investment and by competition from low-income countries where others as ineffectual in redistributing the benefits labor costs and labor rights are least developed of growth. A fourth factor was what some have has been instrumental in weakening the rights called technological stalemate, whereby labor- and benefits of those in the lower end of the labor saving innovations became more predominant market of many industrialized economies.2 In than product innovations, which led to a more effect, within labor markets income security intense search for cheap-labor forms of pro- has been eroded, and economically and socially duction. The subsequent “technological revo- vulnerable groups have been most likely to lution,” associated with microelectronics and suffer. satellites, was a fifth eroding factor, since it gave The supply-side economic model rejects neo- rise to more managerial options and to more corporatist State planning and incomes policy, intensive international competition, partly be- whereas faith in market mechanisms is absolute. cause the new technology was so internationally One consequence is that the strength of “in- mobile. Finally, the growth of open unem- siders” in the labor market has also been eroded, ployment accompanying these developments notably unionized (male) wage workers. That weakened workers’ bargaining power and put in turn has strengthened the pressure for labor welfare states under tremendous pressure. market deregulation, weakening both employ- Although one could quibble with these stylized ment security legislation and customary practices facts, essentially they combined to give intel- preserving job security. In country after country, lectual and political legitimacy to an ideology of including many developing countries, govern- supply-side economics, where market mecha- ments have taken steps to make it easier for nisms and cost competitiveness were given over- employers to dismiss workers or reduce the size whelming emphasis. This crystallized in a global of their labor force, as, for example, in the strategy of “structural adjustment and stabiliza- Philippines, where legislation is planned to re- tion ,” and has been linked to radical changes move most enterprises from coverage by various in labor relations in most parts of the world labor laws. By such means, they have encouraged economy. a more flexible approach to job structures, This argument does some injustice to nuance making it easier to alter job boundaries and the in the interest of brevity. But in essence, the technical division of labor. This has reduced the supply-side model entails a global strategy of job “rights” of existing employees and allowed growth based on open economies, with trade greater resort to so-called external labor mar- liberalization as vital and export-led growth as kets. Because the employment, income, and the only viable development strategy. As such, job security of insiders has weakened, employers cost competitiveness is elevated to utmost signifi- have been able to substitute lower-cost labor. cance, and from that, labor market regulations In many cases, job flexibility also reduces become “rigidities,” which raise costs and thus the premiums that employers usually attach to harm living standards and employment. An irony workers’ employment continuity and on-the-job is that in the 1980s many of the previous experience. objectives of economic growth, notably a whole A further aspect of supply-side economics set of labor and social rights, became perceived concerns income security directly. Governments increasingly as costs and rigidities. have been urged to remove or weaken minimum A few key features of the supply-side agenda wage legislation and institutional safeguards, on are worth noting. The goal of ‘*rolling back the the grounds that such wages reduce employment. State” means focusing on rewarding merit and One might question the logic of that argu- combining fiscal reform with a minimalist rather ment, but among the likely consequences of a than “redistributive” welfare state; poverty alle- weakening of protective machinery is a growth of viation and universal social security are no longer very low-wage employment, consisting of jobs priority issues. A consequence of increasing paying “individual” rather than “family” wages. “selectivity” or “targeting” has been that fewer Research shows that when low-wage jobs spread, people are entitled to state benefits in industrial- it is women whose employment in them in- ized countries. This has given a boost to “addi- creases. Even in many developing countries tional worker” effects (pushing more women where minimum wage legislation was only into the labor market), the informal or “black weakly enforced, it at least set standards and had economy,” and precarious forms of working, demonstration effects. Deregulation sanctions since those without rights to benefits have been and encourages bad practices. obliged to find whatever income-earning work An aspect of the supply-side agenda has been they can. It is scarcely an exaggeration to say that the stabilization and structural adjustment policy FEMINIZATION THROUGH FLEXIBLE LABOR 1079 packages urged on many developing countries has been a shift in these directions and from by the International Monetary Fund (IMF), direct to indirect forms of employment, includ- the World Bank, and other international and ing subcontracting from larger to smaller units national donor agencies, largely in the wake of of production, “networking,” and a revival of the debt crisis. To assess what is happening and homeworking and other forms of outwork. But likely to happen to women in the labor market, these trends have also been occurring in indus- we must appreciate what this orthodox strategy trializing economies, where until recently it had involves. been presumed that the long-term trend of in- First, overwhelming emphasis is put on trade dustrial development would involve a shift from liberalization and the need to orient produc- unregulated, informal labor to secure, regular tion to export-led industrialization. This entails employment. The global pursuit of flexible low- cutting subsidies to domestic “nontradeable” cost labor has encouraged industrial enterprises production, often including staple food items everywhere to reduce their fixed wage labor (with such effects as lengthening women’s work- force, make payment systems more flexible and ing day). It has meant macroeconomic defla- use more contract workers, temporary labor and tion to reduce domestic consumption or living out-sourcing through use of homeworking or standards, so as to shift resources to export subcontracting to small informal enterprises that industries, again often having adverse effects on are not covered by labor or other regulations and low-income women who produce basic consumer that bear the risks and uncertainty of fluctuating goods. The supply-side agenda has meant a focus business. That is the context in which to assess on cost-cutting international competitiveness, in the changing labor market position of both men practice implying a strenuous search for ways and women in many parts of the world. of lowering unit labor costs, which of course At the same time, industrial enterprises means that firms will find ways of employing have been introducing modern technologies that workers prepared or forced to take low-wage have been associated with changing skill and job jobs. Finally, it has also meant a spread of structures. The debate over the “de-skilling” or new production techniques, usually as part of the upgrading effects of modern technology is un- search for least-cost methods. This, no doubt, resolved, but the evidence seems to support two has increased the scope for more refined tech- pertinent trends. The use of craft skills learned nical divisions of labor. Thus, for such conven- via apprenticeships and prolonged on-the-job tional supply-side reasons as improved efficiency learning have declined; such crafts have tradi- and renewed growth, governments have been tionally been dominated by male “labor aristoc- pressed to remove labor market regulations, cut racies.” Second, there is a trend toward skill the public sector, and privatize public enterprises polarization, consisting of an elite of techni- and services, all of which in one way or another cally skilled, high-status specialist workers pos- have eroded employment security and led to a sessing higher level institutional qualifications, reduction of employment. coupled with a larger mass of technically semi- In the context of this global supply-side per- skilled production and subsidiary workers requir- spective, corporate management strategy has ing minor training typically imparted through evolved in clear directions in the past decade. “modules of employable skill,” that is, by short- Stimulated by high unemployment, by new tech- term courses of a few weeks or even by on-the- nology, by more aggressive international com- job learning. This polarization places greater petition (notably from Japan and the newly reliance on external than on internal labor industrialized countries), by deregulation and the markets, since more workers are in “static” erosion of union strength, and by the desire to rather than “progressive” jobs involving little or overcome the uncertainty induced by the inter- no prospect of upward mobility, or firm-specific national economic instability, enterprises every- returns to on-the-job continuity. This, of course, where are devising means of reducing the fixed weakens one reason traditionally given for dis- costs of labor. There is a global trend to reduced crimination against women: that women have a reliance on full-time wage and salary workers higher probability of labor turnover. If there earning fixed wages and various fringe benefits. were less benefit from on-the-job learning and Companies and public sector enterprises in experience, this presumption would not matter, both developed and developing economies are even if true. Indeed, for many monotonous jobs increasingly resorting to casual or temporary high turnover may have a positive benefit for workers, to part-timers, to subcontracting and to employers, since maximum efficiency may be contract workers. In the process, they further reached after only a few months, thereafter erode employment and income security. plateauing or declining. This is one reason Particularly in industrialized countries there for resorting to temporary employees, for 1080 WORLD DEVELOPMENT

job-rotating, or for collapsing job classifications share of nonagricultural employment has grown into more broadly based job clusters such that (see Tables 3 and 4). And it is likely that the Workers can be shifted from one set of tasks female share of production worker employment to another from time to time. But this repre- is also relatively high and rising. The available sents a growth of job insecurity to accompany figures generally bear this out, even though in the income and employment insecurity that some countries, particularly in Latin America, have marked the growth of more flexible labor the relationship may have been weakened by the markets. debt-induced recession (see Table 5). These limited data are still inconclusive, al- though they scarcely support Boserup’s thesis 3. GLOBAL FEMINIZATION? that with industrial development women would be pushed out of production work. In support of In the 1960s, economists commonly argued an alternative thesis that trade liberalization and that the growth of the modern sector in develop- export-led industrialization tend to increase fe- ing countries contributed to the marginalization male employment, it has been observed that the of women as workers.’ But the distinction be- female proportion of productive wage workers tween “modern” and “traditional” or “formal” rose in all countries that had set up large Export and “informal” sectors has become much less Processing Zones - the Dominican Republic, clear, if it ever was clear. In various respects, El Salvador, Honduras, Hong Kong, Republic trends discussed in the preceding section repre- of Korea, Malaysia, Mexico, the Philippines, sent widespread informalization of labor in most Puerto Rico, Singapore, Sri Lanka, and Thai- sectors. This may well explain the absolute and land.’ Productive employment, of course, covers relative growth in the use of female labor around only direct wage earners, and there is grow- the world and a “feminization” of many jobs and ing evidence that. for reasons discussed in the activities traditionally dominated by men. Al- previous section, much of the employment con- though the concepts and measurements of labor nected with export industries, as well as others. force participation are notorious, the interna- is indirect if not concealed altogether. A very tional data strongly suggest that women’s partici- good example comes from a study in Mexico pation has been rising while male equivalent that should be replicated in many more coun- participation has been falling (see Tables 1 and tries. This study showed that production was or- 2). ganized through a complex process of subcon- First, outward-oriented development strate- tracting, with the labor-intensive, lower paid, gies, based on export-led industrialization, have more informal activities being put out to women brought a rapid growth of low-wage female workers, many of whom were not recorded in the employment. Indeed, no country has successfully workforce.h This is a classic instance of “modern” industrialized or pursued this development strat- production relying on what is depicted as “pre- egy without relying on a huge expansion of modern ,” or informal, labor relations. The pres- female labor. And in export processing zones sure to avoid overhead and other indirect labor of many industrializing countries it is not un- costs in the quest for competitiveness has surely common for three-quarters of all workers to be accentuated such tendencies. women. Other aspects of the structural adjustment The reasons are well known. Much of the strategy have also affected women’s employ- assembling and production line work is semi- ment. Consider three: the pursuit of lower wages skilled and low paid; young women, particularly (and greater wage differentials and wage flexi- in the newly industrialized countries (NICs) in bility), labor market deregulation, and the cut- Asia, have been socially and economically op- back of the public sector, either through general pressed for so long that they have low “aspiration public expenditure contractions or privatization. wages” and low “efficiency wages.“J They are Not only do women workers receive lower prepared to work for low wages for long work wages in general, but they are more prepared weeks, normally without agitating to join unions, to work for lower “aspiration wages” for well- and when their productivity declines after a few known reasons. The erosion of minimum wage years of youthful diligence they are replaced by legislation, or of its implementation, and the new cohorts. sanctioning of a general lowering of wages are Typically, in countries that have pursued likely in themselves to lead to a substitution of the export-led industrialization strategy recom- women for men, partly because men are less mended as part of structural adjustment pro- willing to work for sub-family wage rates and grams, the female labor force participation rate is partly because they would be expected to re- high and has risen. In such countries the female spond to lower wages by reducing their “effort Table 1. Variations in adult’ male and female activity rates in the 1980s

Men, rose Men, fell Men. no change Developing developed Developing Developed Developing Developed

Women Barbados Canada Algeria (-)t Australia (0) Honduras Austria rose Chile iouth Africa Costa Rica (-) Finland (-) Indonesia Denmark Egypt Ecuador (-) France (0) Mauritius New Zealand Guam Israel (-) Fed. Rep. of Pakistan Sweden Jamaica Korea, Rep. (0) Germany (0) Puerto Rico Peru Kuwait (+) Greece (+) Seychelles Senegal Netherlands Italy (+) Sri Lanka Thailand Antilles (0) Japan (0) Venezuela Singapore (+) Netherlands (+) Trinidad and Norway (+) Tobago (0) Portugal (-) Spain (+) United States (+) I

125% (60% ( 22% 1 20%

Women Cameroon (-) 0 Haiti 0 0 0 fell Argentina (+)

in Women Bolivia Iceland Bahrain Hong Kong 0 no change Panama Guatemala Syrian Arab Rep. F Philippines Zambia E

*Age coverage is 15-64 except as follows: 15-49: Cameroon (1985), Syrian Arab Republic (1984); 15-59: Costa Rica, Panama, Seychelles (1985), Sri Lanka (1981), Thailand, Zambia; 1659: Guam; 16-64: Puerto Rico, Norway, Spain, Sweden; lM4: Israel (1980); 20-59: Algeria; 20-64: Finland (1980). Italy, Jamaica, South Africa. tSymbols in parentheses indicate net direction ofchange, male and female combined: (+) Net increase; (-) Net decrease; (0) Zero net change. *Percentage of countries in the category. Source: ILO, Year Book of Labor Starktics (various years). 1082 WORLD DEVELOPMENT

Table 2. Variations in activity rates* (persons 15354). more labor market flexibility implies implicit 1980s by percentage of countries with each type of deregulation, and could be expected to lead to change, total and by sex a widening of sexual earnings inequality. The available data, for what they are worth, suggest a Developing Developed more complex story. While the sex differential is Population Type of change countries countries relatively great in economies pursuing export-led industrialization, the differential shows signs of Women Increased 69 90 narrowing in Korea, unlike other NICs (see Decreased 8 0 Table 6). Not too much should be read into those No change 22 10 figures, since money earnings are only part of Total 99 100 total compensation; a loss of entitlement to fringe benefits is most closely associated with Men Increased 36 15 labor flexibilization. Decreased 33 65 It is often claimed that regulations designed to No change 31 20 protect women workers generously contribute Total 100 1Oi.l to discrimination against them by employers. Total Increased 61 65 This has been reported from countries where, Decreased 22 15 for example, maternity leave benefits are paid Compensatedt 8 20 mainly or wholly by the employer;’ the problem No change 8 0 is apparently less acute in countries in which Total 99 100 the benefits are funded through worker and employer contributions to the social security system.X Nevertheless, informalization of em- *For national definitions of activity rates and labor force participation, refer to the IL0 Yearbook of ployment reduces labor costs in that respect and Labor Statistics. For a critique of the relevance of this means that fewer women workers receive such concept in developing countries, see Standing (1981). benefits. Legislation targeted at relatively large Figures have been rounded. industrial enterprises stipulating that they must tActivity rates of men and women changed in the provide critches has also been cited as deterring opposite directions. involving a fall in male and a rise in the recruitment of women wage workers.’ Once female activity rates. so that they approximately offset again, explicit and implicit deregulation may each other. mean more employment of women, but on less favorable terms. Mothers of young children are left to find alternative - and probably costly or bargain.” So, employers are inclined to hire inadequate - child care arrangements. Or they women more readily. While the promotion of may take informal. low-income work that can be female employment may be desirable, this is combined with childrearing, or drop out of the surely not the way to achieve it. labor force altogether. In effect, labor deregula- As for labor market deregulation, it has af- tion of this kind means transferring labor costs fected the economic position of women in from the firm, or even the State, to the individual various ways. Consider the principle of “equal workers. most of whom are in the poorer strata pay for equal work.” As of late 198X. 108 of society. countries had ratified the International Labor In various countries regulations have long Organization’s (ILO’s) Equal Remuneration existed to limit the working time of women Convention No. 100. Among those that have not workers or to prohibit night work, as stipulated ratified it are a disproportionate number of by IL0 Convention No. X9 of 1948. These countries that are pursuing an export-led in- regulations have been criticized by supply- dustrialization strategy. especially those with side and structural adjustment theorists on the large export processing zones: Hong Kong, the grounds that they reduce employment of the Republic of Korea, Malaysia, Mauritius, Singa- “protected” groups. It is interesting that 62 pore, Sri Lanka, and Thailand. Other countries, countries have ratified the Night Work (Women) like India, have different legal minimum wages Convention, but whereas 26 of those did so in the for men and women in certain industries, on the 1950s and 22 did so in the 196Os, only 10 ratified official presumption that women do the less in the 1970s and only 4 in the 1980s; five have arduous work, which is far from the case. Besides denounced it in the past eight years. With the such loopholes in official regulations, any in- spread of shift work, often done by women, formalization of labor relations can be expected governments have not only moved away from to undermine whatever protective effect regula- such regulations, but have failed to implement tions might have on equal wages. Accordingly, those that exist or have granted exemptions FEMINIZATION THROUGH FLEXIBLE LABOR 1083

Table 3. Percentage share of women in nonagricultural employment,* 1975-87 country Sourcet 1975 1980 1985 1986 1987

Africa Botswanat 19 24 30 31 na. Egypt5 10 11 16 n.a. n.a. Gambia11 10 12 15 15 n.a. Kenya n.a. 17 20 21 21 Malawi 7 9 16 14 n.a. Mauritius 20 26 35 36 36 NigerS 4 4 7 7 n.a. Swaziland 22 26 31 31 n.a. TanzaniaO 12 17 17 n.a. n.a. Zimbabwe 13 13 16 n.a. n.a. Latin America and the Caribbean Barbados$,fl 42 43 44 45 45 Bermuda na. 43 46 47 47 Brazil$,n 33’ 35 38 39 n.a. Chile na. 34 36 36 35 Colombia 37 39 38 39 40 Costa Rica n.a. 30 34 34 35 Cuba n.a. 36 41 41 n.a. Haiti 66 71 n.a. n.a. n.a. Jamaica 46 48 48 48 n.a. Netherlands Antilles** 35 n.a. 37 37 n.a. Panama11 38 39 40 40 40 Paraguay 39 35 n.a. n.a. n.a. Peru n.a. n.a. n.a. n.a. 40 Puerto Rico 35 38 39 40 40 Trinidad and Tobago 28 31 34 34 34 Venezuela 32 32 32 32 32 Asia and the Pacific Bahrainll,tt n.a. 10 11 n.a. n.a. cypruss 30 33 35 35 36 Hong Kong 40 39 40 40 41 India 10 11 12 12 n.a. Indonesia$ 37 34 37 39 na. Israel 33 37 39 39 40 Jordan8 14 17 23 n.a. n.a. Korea, Rep. 33 35 38 38 39 Malaysia n.a. 30 33 34 n.a. Philippines 47 46 48 48 47 Singapore 30 35 36 38 38 Sri Lanka 18 18 25 28 n.a. Syrian Arab Rep.P,J( 8 9 9 n.a. n.a. Thailand 42 42 44 44 n.a.

*Coverage refers to total employed except as follows: Employees - Botswana, Gambia, Kenya, Mauritius, Niger, Swaziland, Tanzania, Zimbabwe, Cuba, India, Jordan, Sri Lanka; All persons engaged - Malawi, Bermuda, Hong Kong. Figures have been rounded. t(l) Labor force survey; (2) Social insurance statistics; (3) Establishment surveys; (4) Official estimates. $-WFigures were not available for the years specified, and those of the closest years were given as follows: $1975 = 1976; 01985 = 1984; 111980= 1979; 71980 = 1981; **1975 = 1977; tt1985 = 1982. Source: ILO, Year Book of Labor Statistics (various years). 1084 WORLD DEVELOPMENT

Table 4. Percentage* of women among manufacruring employees, developing countries, 1975-87

Country Sourcet 1975 1980 1985 1986 1987

Africa Botswana n.a. 17 27 24 n.a. Kenya n.a. 9 10 10 10 Mauritius 56 62 59 57 Swaziland 26 27 31 n.a. Tanzania n.a. n.a. n.a. Zimbabwe 7 n.a. n.a.

Latin America Costa Rica n.a. 27 30 30 31$ Cuba n.a. 26 31 316 n.a. Mexico n.a. 21 25 26 n.a. Panama 25 n.a. 24 26 n.a. Puerto Rico 48 49 48 48 Venezuela 24 n.a. n.a. ma.

Asia and the Pacific China n.a. 40 40 41 41 Hong Kong 52 50 50 50 50 India (3) 9 10 10 9 9 Jordan 10 n.a. n.a n.a. Korea, Rep. ii; n.a.l2 45 42 42 n.a. Singapore _ 47 46 47 48 Sri Lanka I:; 4132 31 39 45 n.a. Thailand (1) 41 42 45 45 n.a.

*Figures have been rounded. t(1) Labor force survey; (2) Social insurance statistics; (3) Establishment surveys; (4) Official estimates. $Prior to 1987: including mining. BPrior to 1986: including water. Source: ILO, Year Book of Labor Stafistics (various years). to various industries, often those operating in have accelerated the erosion of notions of voca- export zones, as in Malaysia, Mauritius, and tional skill and of job security (traditionally Pakistan. “’ defended by demarcation rules). Within many The shift away from large centralized work- industries, skill polarization favors the femi- forces toward more decentralized, flexible sys- nization of employment. Traditionally, sexual tems implies less emphasis on those behavioral inequality within the labor market has been characteristics that traditionally have been cited perpetuated through sexual segregation in entry as justification for discrimination against the to specific jobs, covering both the level of recruitment of women, such as women’s alleged recruitment and subsequent promotion. How- higher absenteeism and labor turnover.” In- ever, if a growing proportion of all “jobs” have deed, in the desire to avoid overhead and other no promotion potential - that is, are “static” nonwage labor costs, the decentralization - rather than “progressive” - then one mecha- including putting out, contract labor, and sub- nism intensifying sexual inequality is reduced. l2 contracting to subsidiary enterprises - typically Moreover, the well-known “overcrowding” ex- puts more pressure on workers to cut their labor planation of low female earnings - due to supply “price.” We should, therefore, give more women being crowded into a smaller number of attention to the mechanisms of control that force sex-typed jobs - may well lose force if the jobs workers to labor with high intensity for miserably being whittled away are predominantly “male low incomes, an issue discussed more fully in the jobs.” Of course, this might be offset if male next section. workers proceeded to bump out women from What about the implications for skill for- other slots or took the major share of the new mation? Both deregulation and “flexibilization” jobs. However, the latter seems unlikely to FEMINIZATION THROUGH FLEXIBLE LABOR 1085

Table 5. Proportion of women among production workers (all statuses*), (percentage, 1970s and most recent)

country sourcet country sourcet

Africa Trinidad and Tobago LFSS 1978 1986 Botswana C 1981 198&X5 LFSS 7 23 Uruguay LFSSc 12% 197513 1;:5 CaIllerocln C 1976 1982 C 20 18 OE 12 8 Venezuela C 25% 1971 1987 Egypt LFSS 1975 1984 HS 10 10 LFSS 2 6 Virgin Islands cs 1970 1980 Ghana C 1970 1984 (United Kingdom) C 2 5 C 35 45 MOrOW.? C 10% 1971 1982 Asia and the Pacific C 5% 16 23 Bahrain C 1971 1981 Mauritius C 1972 1983 C 0 1 C 6 21 Bangladesh C 1974 1984 Seychelles C 1971 1981 C 5 17 OE 10 15 Brunei C 1971 1981 South Africa cs 1970 1985 C 7 13 Hong Kong CCS 19763 19486 Tunisia C 1975 1980 C 37 33 LFSS 24 22 India C 1971 1981 C 12 13 Latin America and Indonesia C 1971 1985 the Caribbean HS 27 26 Bahamas HS 1970 1980 Israel cs 1972 1987 C 11 12 LFSS 12 13 Barbados HS 1977 1987 Jordan OE 1976 1979 LFSS 22 26 C 3 1 Belize C 1970 1980 Korea, Dem. C 10% 1970 1976 C 10 13 People’s Rep. LFSS 24 30 Costa Rica C 1973 1987 Korea, Rep. C 1975 1987 HS 12 20 LFSS 28 31 Chile C 1970 1986 Malaysia C 1970 1980 LFSS 12 15 C 17 22 Dominican Republic C 1970 1981 Nepal C 1971 1976 HS 9 21 Ecuador Cc 10% 197422 1;:2 Philippines C 1970 1987 C 15 12 HS 33 22 El Salvador C 1971 1980 Singapore C 1970 1987 HS 19 24 LFSS 19 29 Guatemala C 1973 1981 Sri Lanka C 1971 1981 C 14 12 C 15 13 Guyana LFSS 1977 1980 Syrian Arab Republic LFSS 1970 1984 C 15 9 LFSS 5 4 Haiti C 1971 1982 Thailand LFSS 1970 1980 cs 43 32 LFSS 29 30 Jamaica LFSS 1976 1986 LFSS 26 23 OCellflill Mexico C 1970 1980 Cook Islands C 1976 1981 C 24 17 Panama C 1970 1986 Fiji C 197622 $6 LFSS 11 12 C 4 10 Paraguay C 10% 1972 1982 French Polynesia C 1977 1983 C 28 18 C 8 17 PerU C 1972 1981 Samoa C 1976 1981 C 14 11 C 6 9 Puerto Rico LFSS 1975 1988 Tonga C 1976 1986 LFSS 20 24 C 5 13 St.-Pierre and Miquelon C 1974 1982 C 7 5

*Includes conventional categories: own-account workers, employees, employers, and unpaid family workers. Figures have been rounded. tC = Census; C . . . % = Census: sample tabulation, size specified; Cs = Census: sample tabulation, size not specified; HS = Household survey; LFSS = Labor force sample survey; OE = Official estimates. Source: ILO, Year Book of Labor Statistics (various years). 1086 WORLD DEVELOPMENT

Table 6. Female earnings as a percentage of male earnings in manufacturing (selected developing countries)* country 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987

Africa Egypt 68 75 63 92 Kenya 66 77 56 54 70 63 59 76 80 77 76 73 63 Tanzania 71 79 88 82 82 79 78 Swaziland 66 78 71 83 81 82 81 61 55 72 73 Latin America El Salvador 90 86 81 82 79 81 86 89 77 84 82 Netherlands Antilles 51 66 65 67 68 64 Asia Burma 89 82 103 87 89 86 89 91 92 94 99 86 Cyprus 47 49 50 48 50 50 54 56 55 56 56 56 58 Hong Kong 78 79 81 79 78 76 Jordan 54 60 58 64 Korea, Rep. 47 49 45 44 44 45 45 45 46 47 47 49 50 Singapore 62 62 63 64 65 63 56 58 Sri Lanka 81 87 82 71 69 72 78 71

*Blank spaces indicate no available data. Figures have been rounded Source: ILO, Year Book of Labor Statistics (various years).

happen. Although much less so in Latin America there, 24 of the 30 jobs formerly barred to than in other regions, not only do the data women have been recently opened to them. suggest that women are being substituted widely This decrowding process extends to both for men in various occupational categories, manual and nonmanual employment. Tradi- including manufacturing and production work tionally, women have comprised a comparatively in countries as diverse as Ghana, Swaziland, large proportion of all professional and technical Bangladesh, and Costa Rica (see Tables 4 and workers in developing countries, largely because 5), but there are good reasons to suppose that of their prominence in professions such as substitution will continue as, a long-term trend. teaching and nursing. Traditionally, with de- This is further supported by a remarkable velopment the growing professions were in in- change between the 1970s and 1980s. In the dustry, where in most parts of the world women earlier decade, women’s unemployment rates have been underrepresented. However, it seems rose relative to men’s in many more countries to be a global trend that enterprises have been than where the reverse occurred. In the 1980s in restructuring to erode middle management, fol- the vast majority of both industrialized and lowing an era when this category had mush- developing countries, female unemployment de- roomed almost everywhere. Now, many middle- clined relative to male, so that in a substantial management functions are being delegated either number their open unemployment rate became to clerical workers, most of whom are women lower than the male equivalent. This marked a (except in some South Asian countries), or to tremendous shift. ” production workers. Whether one interprets the Some observers attribute substitution to grow- trend as one of upgrading or reskilling or simply ing labor force attachment of women, some intensifying clerical work, the long-term trend is to improvements in schooling or access to train- likely to be a further substitution of women for ing, some even to the beneficent effects of men. It would be a mistake to think this trend has antidiscrimination legislation. But it has almost relevance only for industrialized countries such certainly more to do with the feminization of as the United States. It is further reason for labor, a desire to have a more disposable (or believing that the importance of sexual segrega- flexible) labor force, with lower fixed costs, and tion per se in perpetuating sexual inequality in so on. In other words, for women there will be the labor market will decline. Policy attention less problem of job entry. As so often, among should be focused elsewhere. developing countries the Republic of Korea has Before turning to that issue, it is worth been at the forefront of change in this respect; reiterating that for manual production jobs, FEMINIZATION THROUGH FLEXIBLE LABOR 1087 technological and organizational changes have Careful study of this phenomenon is still re- tended to create a jobs polarization, with an elite quired, since we do not know which groups suffer of technicians (or “crafticians”) in specialist jobs relatively from the direct effect.” However, it is coexisting with a growing mass of semiskilled interesting to note that the scanty data available flexi-workers and a dwindling number of jobs for do indicate that in many developing and in- technically unskilled workers. This may increase dustrialized countries, women not only have female industrial employment - because semi- accounted for a greater share of public sector skilled, static jobs are reserved largely for women than private sector employment - with notable - but the disappearance of low-wage jobs at the exceptions, such as India - but their share of bottom may hit the most vulnerable groups of public employment has, if anything, grown in all; impoverished, uneducated women are left recent years (see Table 7). to “crowd” into those jobs, pushing down their The indirect effects are even harder to unravel. wages even further. As a remedial policy, it may In countries where men have predominated in be too late or impractical to stress basic educa- the public sector, as in most of Africa, they tion for such women. Basic education may be will make up the great majority of redundant part of the long-term policy answer, but for workers. Will those men drive women out of currently uneducated women forced to compete small businesses or those few private-sector for the low-paying, unskilled jobs in industry, wage jobs that women have obtained? Ex-pub- minimum wage and other forms of protective lit-sector workers often have received large regulations and/or strong trade unions repre- redundancy payments with which they could senting their interests are surely essential. Those acquire petty capital and drive out existing who wish to see such regulations dismantled need businesses, many of which are dominated by to offer some more viable alternative. women. I6 This tendency may be compounded by In the more flexible labor markets envisaged the contacts men gained from their long presence for the 1990s the attendant insecurities will be in the public sector. intensified by a web of dependency relations. In short, women have good reason to fear the Decentralized or individualized relationships be- marginalization effects of privatization, at least in tween workers and employers, or between Africa. But this is not all. Falling public employ- workers and impersonal enterprises or middle- ment may, as in Peru, lead to more women men, give much greater scope to exploitative entering the low-income labor market as “addi- and oppressive forms of control, including debt tional workers” because of higher male and bondage, outright coercion, and casual beck-and- female unemployment.” If they enter an already call labor relations. This is where policy will have crowded sector, one can predict that average to focus. incomes will decline, along with the security of Along with deregulated and decentralized participating in such activities. labor in industry, a key theme of the orthodox Finally, stabilization and adjustment strategies structural adjustment strategy is “privatization.” have involved deflation of aggregate demand, This has severe implications for women. Not only leading to higher unemployment and more wide- are women’s wages and employment conditions spread resort to informal survival responses better on average in the public sector than in the among the poor, particularly in urban areas. private, but wage differentials between men and Women have long been concentrated in such women are smaller in the public sector.14 There activities, both as petty traders or “pre-entre- is prima facie reason to suppose, therefore, that preneurs” and as dependent workers, whether in women’s wages will fall absolutely and relatively familial enterprises or as wage workers. In eras by virtue of cutbacks in the public sector or as a of stagnation and recession, these activities will result of privatization. But the extent of decline normally become more precarious, and the vul- will vary. The tendency will be relatively less nerable groups will be hit hardest. However. in likely where there is a so-called policy of cutting many countries women are well entrenched in public expenditure with a human face, so that self-employment and in small-scale production smaller cuts are made in social programs like and trade. Indeed, although there is little time- health and education, sectors where women series information available, the data suggest that public employees are concentrated. in most countries for which such statistics are A complicating aspect in assessing the impact available, women have comprised a growing of privatization and public expenditure cuts is proportion of the self-employed (see Table 8). that the incidence of redundancy may hit women Whatever else the data might suggest, they do disproportionately hard, especially where they not indicate that women have been squeezed out make up a relatively large proportion of less of own-account activities by recent economic secure, nontenured posts or marginal positions. developments. What such activities include, and Table 7. Female share of public service employment, * selected developing countries (percent)f

Country 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 _ Africa Benin 13 15 Botswana 19 20 35 37 36 Burkina Faso 16 20 Burundi 41 38 Ethiopia 20 22 23 Kenya 18 18 19 19 18 18 19 Malawi 12 12 11 12 12 Morocco 29 28 28 29 Nigeria 11 11 11 13 Rwanda 32 33 Swaziland 27 27 26 25 31 30 32 33 34

Latin America and the Caribbean Barbados 43 45 42 Bolivia 24 24 24 24 Brazil 23 24 Cuba 30 31 30 30 31 33 33 37 38 39 39 Jamaica 48 50 48 48 Panama 46 47 41 43 43 43 45 Trinidad and Tobago 32 34 Venezuela 41 41 42 43 43 44 45 46

Asia and the Pacific Bahrain 32 31 Cyprus 28 29 30 31 32 32 32 32 33 33 Hong Kong 20 21 21 22 24 23 25 28 28 28 28 29 29 India 10 11 Indonesia 23 23 23 24 27 27 29 29 30 Kuwait 31 34 35 Qatar 11 12 19 21 20 Syrian Arab Rep. 20 24

*Public service employment in the total public sector except: Central government - Burundi, Ethiopia, Mali, Rwanda, Kuwait, Bahrain; Government - Botswana, Morocco, Mexico, Trinidad 8~ Tobago; Federal government - Nigeria; Public administration - Brazil. tBlank spaces indicate no available data. Figures have been rounded.

(continued) Table 7. (continued) Sources: Bahrain Statistical Abstract, 1985 Barbados “Labour Force Report,” 1975-83 Belize “1980-81 Population Census of the Commonwealth Caribbean” Benin “Revue de Statistique et de Legislation du Travail,” July 1984 Bolivia “Anuario de Estadisticas de1 Trabajo,” 1982 Botswana “Statistical Bulletin,” regular publication; 1977: Employment Survey 1982 Brazil “Anuario Estatistico do Brasil,” regular publication. 1986: Government reply Burkina Faso “Annuaire Statistique du Burkina Faso,” Oct. 1984 Burundi Government reply. ILO; General Report, JCPS, 3rd Session, 1983 Cuba “Anuario Estadistico de Cuba,” 1986 Cyprus 1977-82: “Statistical Abstract,” 1985 and 1986; 1983-86: “Labour Statistics Bulletin,” Dec. 1986 Ethiooia Government reply Hong’ Kong “Hong Kong Monthly Digest of Statistics,” regular publication India “Pocket book of labour statistics,” regular publication. Data supplied to ILO; Government reply Indonesia Statistical Yearbook of Indonesia, regular publication Jamaica “The Labour Force,” regular publication Kenya “Statistical Abstract,” regular publication Kuwait Government reply Malawi “Reported Employment and Earnings Annual Report,” regular publication Mali IL0 research data Mexico “Encuesta Continua sobre Ocupacion, 2nd Semester 1978” Montserrat “9th Statistical Digest,” 1984 Morocco “Annuaire Statistique du Maroc,” regular publication Nigeria Digest of Statistics, regular publication Pacifii Islands “Quarterly Bulletin of Statistics,” 1980 Panama “Situation Social; Estadisticas de1 Trabajo,” regular publication Paraguay “Encuesta de Hogares par Muestra: Mano de Obra,” 1977 Peru “Censos Nacionales de Poblacion y de Vivienda,” July 1984 Qatar “Annual Statistical Abstract,” regular publication Reunion “Economic de la Reunion, Panorama,” 1987 Rwanda Government reply Swaziland “Employment and Wages,” regular publication Syrian Arab Rep. “Statistical Abstract,” regular publication Trinidad and Tobago “Quarterly Economic Report,” regular publication Venezuela “Encuesta de Hogares por Muestreo,” regular publication 1090 WORLD DEVELOPMENT conceal, is quite another matter, and may repre- acute in the context of what appears to be a sent a deterioration in the labor market position growing diversity of forms of labor relations. of women in general. Unfortunately, the data are The basic starting point is that the conven- rather unhelpful in this regard, which brings us to tional “work status” classification (used tenta- a crucial issue for the further understanding of tively earlier) is grossly deficient. Too many what is happening to women in developing types of labor relations are compressed into the countries, and thus, what policies donors and four statuses: own-account, employer, unpaid planners should consider for the near future. family worker, and wage worker. To understand the mechanisms of labor force participation, data are needed on different forms of control. 4. DATA PRIORITIES While we cannot go into details here, seven critical aspects of control are insufficiently Much has been made of the “invisibility” covered by conventional labor statistics. The first of women workers in conventional labor force is control over self, over one’s own labor power; statistics. It is time that we concentrated on bonded laborers or serfs have no control over the distortions provided by the available data on their choice of activity, and as we know many women’s labor activity. This issue is particularly women are extremely vulnerable in this respect

Table 8. Share of women in self-employment (nonagricultural sectors) in selected developing countries (percenfage, 1970s and most recent)*

Country Source? Country Sourcet

Africa Asia and the Pacific Ghana C 1970 1984 Bangladesh C 1974 1984 C 73 77 c 3 8 Seychelles C 1971 1981 Hong Kong CS 1976 1986 OE 23 19 C 16 20 India C 1971 1981 C 9 8 Latin America Indonesia% C 1971 1985 and the Caribbean HS 24 41 1973 1987 Costa Rica C Korea, Dem. C 1970 1976 HS 13 27 People’s Rep. LFSS 25 39 C 1970 1986 Chile Korea, Rep. C 1975 1987 LFSS 28 28 LFSS 29 35 C 1970 1981 Dominican Republic Kuwait C 1975 1985 C 23 27 c 1 1 C 1974 1982 Ecuador Nepal C 1971 1976 C 25 22 HS 13 31 C 1971 1980 El Salvador Singapore C 1970 1987 HS 48 65 LFSS 13 19 C 1973 1981 Guatemala Sri Lanka C 1971 1981 C 29 25 C 12 9 C 1970 1980 Mexico Thailand LFSS 1970 1980 C 28 33 LFSS 40 44 C 1972 1981 Peru United Arab Emirates C 1975 1980 C 31 29 c 1 1 Puerto Rico LFSS 1975 1988 LFSS 16 15 Oceania Fiji C 1976 1986 Venezuela C 1971 1987 HS 17 23 C 1.5 23 French Polynesia C 1977 1983 C 31 33 Samoa C 1976 1981 C 30 27

‘Figures have been rounded. tC = Census; Cs = Census: sample tabulation, size not specified; HS = Household survey; LFSS = Labor force sample survey; OE = Official estimates. #Includes agriculture. Source: ILO, Year Book of Labor Statistics (various years). FEMINIZATION THROUGH FLEXIBLE LABOR 1091 in many parts of the world. Second, there is Seventh, there is control over labor repro- control over labor time. Many women have no duction. This too is critical, and refers to the choice over whether they work the number of ability to develop and maintain the woman’s own hours that would suit their particular needs and, “skills” and work capacity. As such, it is not the as in typical export processing zones, have to same as control over one’s labor power. In the accept 60 hour workweeks, drop out of the case of labor reproduction, the concern is pri- workforce, or lose a disproportionate part of marily with workers’ capacity and with education their income for failing to fulfill the required and training. Many impoverished women have so quota.” The number of workers who are in such little control over their working lives that their onerous situations, where, for example, a flexible working capacity is debilitated by what they are payment system means that they would forfeit forced to do. They may seem to earn a reason- their bonus share of income for failing to work able income today, but it is earned only by loss the full, long workweek, is unknown. of future earning capacity. Although ample Third, there is control over means of produc- anecdotal evidence on this process exists, far tion, such as land, tools, and spaces in which more analytical and policy attention should be to work. Women usually have little or no control focused on this aspect of work.” That aside, in that respect, and may have to rent machines or control over labor reproduction concerns the skill have their labor status determined by others such development potential of specific activities. Most as middlemen who own equipment. Although workers - and the vast majority of women, in some attention has been paid to women’s lack of particular - are trapped in work activities in control over means of production and its con- which they have no possibility of developing sequences for incomes and work activity, the full skills. The controls and constraints may consist of range of control mechanisms should receive the type of work or external mechanisms. In far more attention, for they have adverse con- some places, for example, social norms dictate sequences for equity, efficiency, and economic that to remain in the social structure women do growth. the gathering and weeding, sewing, dairying, and Fourth, there is control over raw materials, animal husbandry; they have no access to other inputs purposely transformed into output. Many skills. How are such behavioral patterns re- women, in particular, are exploited by monopo- produced? Although the social pressures are list merchants or manufacturers who charge them crucial, efforts should be made to identify the excessive amounts for their material so that their direct mechanisms. More statistical information net income is much less than it may appear from is needed on such forms of control. piece payments. In sum, if more relevant policies are to be Fifth, there is control over output. If a woman developed and if international donors are to carpet maker were allowed or enabled to sell her increase their effectiveness, then analyses and wares herself, she could often do far better than data gathering should concentrate not just on the when she is forced to accept payment for it from relative “invisibility” of working women, but also relatives, a merchant, or a manufacturer. No on the distortions of conventional data in other, decent study of working women should leave out even more crucial respects. Access to and control this dimension of the production process. over means of production are often cited as vital Sixth, there is control over proceeds of output. for raising the autonomy and real living standards This usually means control over the income of women. This is surely correct, but should not derived from work. Crude earnings data can be lead to neglect of other aspects of the production far more misleading indices of net disposable and distribution process, where access and con- income for some groups than for others. In trol are equally important. many situations, women workers receive very little net disposable incomes for themselves or for their immediate needs because relatives or inter- 5. LABOR VULNERABILITY mediaries deduct large parts of it. Thus, if a woman receives only 20 percent of the income If there is an international feminization of from her work because some intermediary takes labor relations, is there a corresponding growth the remainder, her welfare would scarcely be of vulnerability, precariousness, and insecurity? improved by policy assistance focused on raising The notion of vulnerability is complex, and it her gross income (through training, for instance), may be useful to proceed by recognizing that since most of her poverty would be “structural.” specific groups are vulnerable by sector, by social It is for such reasons that policy makers should stratification, and by labor status.“’ focus far more on the multiple mechanisms of Women are, first, vulnerable to income and control. employment insecurity by virtue of the sector of 1092 WORLD DEVELOPMENT their involvement, and their relative disadvan- women workers fare relatively better in public tage varies by sector. In most low-income sector enterprises. If privatization schemes are to countries women in agriculture and rural areas proceed, international agencies and others must generally are the most desperately affected, be persuaded to introduce measures that pro- particularly on plantations. Outside agriculture, tect the incomes and employment conditions of as the data reviewed earlier showed, women women workers affected by the change. Too few comprise a disproportionately large number of privatization initiatives have paid attention to the employees of small-scale enterprises, precisely vulnerability of the workers who lose the relative where vulnerability to bankruptcy and chronic security of public sector employment. impoverishment are usually greatest. Often wo- Finally, there are those sectors in which men in those units are unpaid or severely women are vulnerable to “invisibility.” In most underpaid, a problem made worse by implicit parts of the world women represent large propor- and explicit labor deregulation of small-scale tions of the workforce in agricultural small- businesses; sex-related income differentials tend holdings, in family farms or businesses. and in to be much greater in small units, perhaps petty trade, all of which tend to be inadequately because larger enterprises are affected by formal recorded or recognized when policies are being regulations and social exposure that lessen wage devised or reformed. Perhaps in international discrimination. This suggests that efforts to pro- and national policy debates this invisibility mote small-scale enterprises - often in pref- problem is less than used to be the case. But erence to large concerns - are likely to worsen everyone should be wary about the issue be- women’s economic vulnerability unless counter- coming passe, for adequate solutions have not vailing measures are implemented. yet been found to the conceptual and statistical Women have been extensively employed by problems. multinational export-oriented enterprises, often Women are also peculiarly vulnerable by vir- in export processing zones. Critics have been tue of their social status, besides their system- vociferous in highlighting the adverse conse- atic oppression in most cultures of the world. quences, although other observers point out Social stratification takes many forms, some of that conditions are often much worse for women which relate to “life-cycle” events, some to other not employed by such multinationals. Anybody stratifying social influences. One common strati- who has been into, say, modern electronics fying influence is migrant status. Women make factories in Southeast Asia, has to recognize the up a majority of migrants in many developing force of the latter argument. Nevertheless, one countries, and many of them travel alone in must also recognize the forms of vulnerability search of work. Too often this aspect has been to which women workers are exposed in such neglected in women’s studies or in donor pro- enterprises. First, they are employed largely in grams. Women migrants and, most of all. women semiskilled, static jobs with little or no chance to labor circulants are acutely vulnerable. They develop skills or aptitudes that they could use tend to gain entry to labor markets only by taking subsequently. They are then vulnerable to a loss the most precarious jobs and have little prospect of vital social skills and working capacity, be- of upward mobility. The plight of such women is cause they must work excessively long work- desperate. In India, for example, it has been weeks and often face exposure to health-sapping widely reported that hard-pressed rural families working conditions. This problem has been com- send wives and daughters to cities to support pounded by the overwhelming preference of them through prostitution.** And in Thailand employers for young single women who are thousands of young village girls flock to Bangkok expected and encouraged to leave their jobs after to work as prostitutes until their “charms fade,” a few years or when they marry. Compared to often in notorious conditions.2’ Other young other workers, their employability declines fairly women enter urban labor markets through quickly, making them vulnerable to aging earlier domestic employment, for little pay, often un- than most people. Have policy makers really known to any labor authority or conveniently addressed the question: what then? overlooked. Far more attention should be given Such industries are also vulnerable to interna- to the mechanisms of control used in the migra- tional trade fluctuations and to productive in- tion process and to the kinds of policies that vestment and disinvestment decisions taken in could ameliorate the vulnerability of what are, another part of the globe. Enterprises may have typically, ill-educated teenagers (see Heyzer, this to lay off production workers with little warnin volume). - one more reason that they prefer women. B Two other dimensions of social vulnerability The other major sectoral dimension is the deserve even more emphasis than they have public-private sector one, since as noted earlier, received in recent years. Many women, probably FEMINIZATION THROUGH FLEXIBLE LABOR 1093 the vast majority, are vulnerable to poverty and projects are in progress on “putting out” in- labor market marginalization by virtue of their dustries and other forms of outwork. But so far, dependency status in families. Others, however, few effective strategies to enhance the status and are primarily or wholly responsible for their living standards of such workers have been household income and subsistence. They tend to implemented. be among the poorest of the poor. In Costa Rica, Of course, many “self-employed” women are for example, although only 16 percent of urban far from the image that the term conventionally households were recorded as being headed implies. At the very least, statistically, own- by women, more than 37 percent of indigent account workers should be separated from (the poorest) urban households were female contract or piece-rate outworkers, whose net headed.24 In the Sudan, a quarter of all low- earnings tend to be minimal, and even negative income households were headed by women, half in slack periods or in periods of illness or family of whom were over 50.25 In Calcutta, the poorer mishaps. These contract workers are the most the household, the more likely it was to be “flexible” labor force of all, bearing a large share dependent on female earners. Across the world of the risk of business fluctuations. In many the story is essentially the same: single mothers, countries, the international recession in the 1980s widows, and migrant women dependent on their and the adoption of stabilization and structural own resources are the most vulnerable of all, not adjustment policies led enterprises to shift from only to poverty, but to exploitation and labor direct wage employment to such forms of labor.27 marginalization. As noted earlier, the renewed growth of such Finally, workers are vulnerable by virtue of indirect, flexible labor will’necessitate a consider- their labor status. Earlier sections of this paper able policy reformulation in the near future if the have depicted the international feminization of previous international trend toward social pro- labor as closely associated with growing labor tection is to be revived. flexibility, by which secure, protected well-paid wage employment is once again being displaced by other labor relations. One can predict that 6. CONCLUSION when enough data have been generated on these issues, the two trends will be shown to be linked It would be pretentious to pose a full policy closely. As it is, women are most vulnerable to and research agenda on the basis of this brief insecure labor status employment. For instance, overview. Nevertheless, some leading questions they are more likely to be hired as marginal, do suggest themselves. First, if the era of labor casual wage workers. In some countries, the flexibility persists, will women and men be introduction, strengthening, or enforcement of pushed far more into labor statuses in which they minimum wage regulations has led to the substi- face multiple forms of control? To answer that tution of women casual workers for male per- question, we need to review critically the type of manent workers. In Zimbabwe. for example, statistics that are collected, and consider alterna- underpayment persisted through a conversion tive concepts that could identify potential points of permanent to casual, contract or piece-rate of policy intervention. Second, to what extent workers, since they were not covered by the are earnings or income data valid measures of regulations; this was most likely to happen in women’s net disposable income, that is, the cash small-scale enterprises.26 or other income that they could use for their own In India, women are also concentrated in needs? Taking account of deductions by inter- casual wage labor, and relatively high female mediaries, including relatives, it is likely that a unemployment has been attributed in part to large and growing proportion of women workers women being able to obtain only short-term receive much lower net incomes than the scanty casual jobs. Elsewhere, women wage workers are data on earnings suggest. Third, whither labor more often than men classified in that peculiar regulations in labor markets characterized by category, “permanent casuals,” and thereby un- informality and flexible work statuses? If regula- able to acquire legal entitlement to benefits or tions on, for instance, wages, safety, and mater- statutory forms of protection. nity leave are being eroded or bypassed, how Beyond recognized wage labor, evidence is could workers organize to obtain socially desir- mounting that women are increasingly utilized as able benefits, and how could governments assist? “outworkers.” Particularly with respect to this Fourth, the same questions arise with respect heterogeneous category one needs to focus on to social security, especially bearing in mind the diverse mechanisms of control and exploita- that even in many industrialized economies a tion (bearing in mind that “control” and “exploi- dwindling proportion of women actually are tation” are not synonymous). Valuable research entitled to existing social security. 1094 WORLD DEVELOPMENT

Fifth, whither antidiscriminatory legislation? Finally, when will donor agencies and national Discrimination is a particularly insidious process; policy makers turn to that most neglected of even if one form of discrimination is partially groups, older women? Already a majority of the tackled, other forms are likely to grow, which is elderly (aged 55 or older) are in developing why it needs to be attacked on many fronts at countries. Their marginalization in almost all once. It is possible that the sexual segregation of respects is a somber specter, too often result- jobs is becoming relatively less important, and ing in premature death, thereby concealing the that career paths and differential access to fringe seriousness of the process. With urbanization and benefits and bonus payments are becoming rela- industrialization, kinship support networks are tively more critical.*s being eroded, yet very few women workers have Sixth, has too much attention been devoted to pension rights; nor do women have employment schooling and training as the avenue to higher security or access to retraining or labor market incomes, labor market security, and mobility? assistance in times of recession or structural Besides the likelihood that most forms of labor adjustment. If this paper made just one plea, it market inequalities are structural, it is possible would be that international agencies should that in recruiting for a wide range of jobs, devote more resources to assisting older women screening devices other than formal schooling are workers. the growing number who will soon be becoming more important, while prior vocational in that category, and those pushed out of the training is scarcely needed for access to most labor force prematurely - the easily-ignored jobs. This is because a more developed technical “potential” workers in their 50s and 60s. Given division of labor and growing emphasis on today’s flexible, insecure labor processes, and semiskilled labor must surely diminish the rele- weakened social support systems, the needs of vance of vocational training. Of course we should older women have never been greater. not dismiss automatically the need for training These eight sets of issues are by no means policies, but “training” should be put in its exhaustive, but they do suggest some reordering proper context - a minor component of any of priorities. Although women may be gaining strategy to improve women’s economic and labor economically in some crude senses of that term, market status. the crucial point is that feminization in the sense Seventh, how could unions protect women in used here represents pervasive insecurity. Tradi- the more flexible labor markets of the 199Os? tionally, women have been relegated predomi- Unless communal unionism develops in place of nantly to more precarious and low-income forms craft or industrial unions, protection of vulner- of economic activity. The fear now is that their able groups will be partial at best and easily increased economic role reflects a spread of those circumvented. Male-dominated trade unions forms to many more spheres. That is scarcely must fully incorporate women and struggle for what should be meant by progress. “women’s issues” at least as strongly as for others; otherwise their collective strength will continue to dissipate.

NOTES

1. Standing (1988). 7. For example, di Domenico (1983); Date-Bah (1986). 2. That is not an argument for cutting exports from industrializing economies, as some might claim. 8. See, for example, Scott (1986).

3. Boserup (1983). 9. International Center for Research on Women (1980). 4. The “aspiration wage” is the level at which a person would be prepared to accept employment; the 10. See, for example. Dror (1984). p. 709. “efficiency wage” is the wage level at which a worker would work with optimum efficiency once in employ- 11. Anker and Hein (1985). ment. Many people fail to make this distinction. 12. Standing (1981). 5. Anker and Hein (1986). p. 95. 13. An important caveat here is that the data do not 6. Beneria and Roldan (1987). include observations for many African countries. For FEMINIZATION THROUGH FLEXIBLE LABOR 1095 those which have observations, the trend is in the same 21. Lim (1985), p. 30. This was a joint UNCTC/ILO direction as that in other regions. study.

14. See, for example, Papola (1986). During the past 22. Pandhe (1976), p. 52. decade of adjustment, private-public sector wage dif- ferentials have apparently narrowed or disappeared in 23. Phongpaichit (1982). many countries, particularly for “educated labor.” 24. Pollack (forthcoming). 15. An official US study found that women were disproportionately affected by job cuts in the privatiza- 25. IL0 (1976). p. 70. tion of federal agency programs. Rein (1985), p. 132. 26. Shopo and Moyo (1986). 16. Collier (1988). 27. See, for example, Penouil and Lachaud (1986). 17. Scott (1986), especially p. 357. PP. 37-38.

18. See, for example, Fuentes and Ehrenreich (1983) 28. Relatively more attention should be paid to the p. 23. factors that cause women to shaft horn male-dommated to female-dominated jobs rather than to those that 19. See Standing (1981), Ch. 4 for one review of cause initial sex segregation. See. for example, Jacobs evidence on this issue. (1983).

20. A more detailed discussion of these dimensions is presented in Standing (1987).