32872 TheWorld Bank D e c e m b e r PREMnotes 2 0 0 4 n u m b e r 9 4 Gender Labor market regulations for women: are they beneficial? Labor market regulations specific to female workers are common in developing countries. What are the labor market effects of maternity benefits and working-hour restrictions for women? And does such legislation offer female workers the protection it promises? This note considers the impacts on female Maternity benefits workers of two forms of labor market regula- Maternity benefits allow working mothers to tions common in developing countries: man- recover from childbirth and to care for their dated maternity benefits and working-hour newborn infants. But when these benefits are The full impact of restrictions for women. Anecdotal evidence financed by employers, they raise the cost of suggests that maternity benefits are unpopu- female labor. Such benefits generally include maternity benefits lar with employers. For example, in Mexico, employment protection during a specified there are numerous accounts of women being period of absence and some level of cash ben- and working-hour required to undergo pregnancy testing as a efits (to compensate women for lost earn- condition of employment in maquiladoras ings), and sometimes medical benefits as well. restrictions for (Human Rights Watch 1996, 1998). In China, Virtually all OECD countries and most devel- there are many reports of rural migrant women oping countries have some form of mater- female workers being fired from jobs in export processing and nity leave provision, with varying levels of other industries when they marry or become cash and medical benefits, lengths of post- need to be clarified pregnant (CASS 1995; Huang 1999). In both natal leave, and financing mechanisms. countries, employers point to the extra cost What is the expected impact of maternity associated with maternity benefits as the rea- benefits on female wages and employment? son for these practices. But it is unclear how Employers are expected to respond to man- such benefits affect women's labor market out- dated employer-financed maternity benefits comes, and whether there are conditions under by reducing the wages they offer to eligible which maternity leave does not result in such women by the amount of the expected cost of discriminatory behavior. the mandate (Summers 1989). If female work- Many developing countries also have work- ers value the benefits, they will accept a lower ing-hour restrictions specific to female work- wage in return for the benefits. Thus wages ers. Such restrictions are usually motivated by are unambiguously lower in the short run. concerns for women's health and safety (see, Over time, however, maternity benefits may for example, reports on the harmful effects lead to stronger labor force attachment and of excessive overtime work in China's export increased investment in the firm-specific processing industries, such as Pan 2002). But human capital of female workers. The result- this type of legislation may negatively affect ing increase in productivity can lead to higher female labor market outcomes because it con- wages, offsetting the initial reduction in wages strains women's working options. As with caused by maternity benefits (Waldfogel 1998). maternity leave, the full impacts of such reg- Maternity benefits can therefore result in ulations need to be clarified. increased, decreased, or unchanged levels of from the development economics vice presidency and povert y reduction and economic management network female employment, depending on employ- rate as a result of the introduction of mater- ers' and female workers' relative perceptions nity leave provisions in Taiwan (China). Sim- of the costs and value of maternity benefits. ilarly, Ruhm (1998) found that both short Over time, job protection during maternity and extended maternity leave increased leave may help women maintain favorable women's employment rates in nine European job matches and avoid the cost of search- countries by 3­4 percent. Winegarden and ing for alternative employment--increasing Bracy (1995) also found that maternity leave the likelihood that female employment rates provisions in industrial countries are associ- will rise (Waldfogel 1998). The expected ated with higher labor force participation effect of maternity benefits on female work- rates for young women. On the other hand, ing hours depends on two opposing effects. Gruber (1994) found that the introduction Cost sharing To qualify for (higher) benefits, women may of mandated medical benefits covering preg- increase their hours prior to taking mater- nancy and childbirth did not have a signifi- between employers, nity leave. On the other hand, women's work- cant impact on female employment rates in ing hours upon returning to work from the United States. employees, and maternity leave could fall if women work In situations where employers are unable fewer hours to have more time with their chil- to adjust wages in response to maternity leave government can dren (Zveglich and Rodgers 2003). mandates (for example, due to equal pay leg- Empirical studies from OECD and devel- islation), they may substitute male workers reduce the potential oping countries have generally found that for female workers instead (Gruber 1994). maternity leave has negative or insignificant An abundant supply of labor might make sub- discrimination and effects on female wages, while the effects on stitution for female workers relatively easy, female employment have been found to be which may partly explain the anecdotal prac- negative wage mostly positive. Many studies have examined tices in Mexico and China cited earlier. the impact of maternity benefits on female The evidence on the impacts of maternity effects of maternity wages. Gruber (1994) found that U.S. state benefits highlights the importance of cost legislation mandating insurance providers sharing between employers, employees, and benefits to cover pregnancy and childbirth (with pre- government. Cost sharing is likely to reduce miums paid by employers) led women's discrimination and dampen negative wage wages to fall by the full cost of providing ben- effects by redistributing some of the costs efits. In contrast, Zveglich and Rodgers of maternity leave from women and employ- (2003) found that the introduction of mater- ers to families and society as a whole. nity leave benefits had no significant effect Cost sharing can be undertaken in a vari- on female wages in Taiwan (China). Ruhm ety of ways. First, the costs of maternity ben- (1998), using data from nine European coun- efits can be shared equally between employers tries, found that the duration of benefits mat- and the social security system. The adminis- ters: paid leave of up to three months had trative costs of this approach can be signifi- little effect on women's wages, but paid leave cant in developing countries. Nonetheless, of nine months reduced hourly wages by 3 some developing countries--such as Costa percent. If the same relationship holds in Rica and Thailand--have been successful in developing countries, maternity leave would instituting this kind of cost sharing for mater- likely result in lower wages. This is because nity benefits. Second, private medical insur- many developing countries--particularly ance coverage can be expanded to pay for transition economies--allow maternity leave the medical benefits that may accompany of up to six months (ILO 1998). maternity leave provisions. Although this The effects of maternity benefits on female approach increases the medical insurance employment have been found to be mostly premiums paid by families, it reduces the positive. For example, Zveglich and Rodgers costs incurred solely by women and employ- (2003) found a 4.5 percent increase in ers (World Bank 2001). Third, maternity women's working hours and a 2.5 percent- leave provisions may have broader social ben- age point increase in women's employment efits (such as reduced early childhood mor- PREMnote 94 December 2004 tality rates; Ruhm 2000) that could justify which prohibited night work by women contributions from general revenues or pay- (from 10 p.m. to 6 a.m.) and limited women's roll taxes. overtime to 2 hours a day and 24 hours a month. (Men's overtime was limited to 3 Sex-specific working-hour hours a day and 46 hours a month.) Zveglich restrictions and Rodgers used differences in coverage of Many countries have introduced working- the labor law between industrial sectors and hour restrictions for female workers to pro- demographic groups (men, women) to iden- tect their health and safety and to allow time tify the impact of working-hour restrictions. for their responsibilities as wives and moth- They found that once the labor law was ers. But such restrictions may negatively enforced, the working-hour restrictions re- affect women's labor market outcomes. In duced women's working hours by 6.1 per- Policymakers the 19th century such regulations, together cent, reduced women's employment rate with occupational bans on female workers, by 0.9 percentage points, and had no imme- should consider were among the first labor market laws to be diate significant effect on female wages. introduced in Europe. But during the 20th Restrictions on night work by female work- eliminating century restrictions on female workers were ers are usually motivated by a desire to avoid increasingly deemed discriminatory, and their exploitation in export-oriented indus- sex-specific most industrial countries abandoned such tries (Politakis 2001). But many of the coun- legislation. Working-hour restrictions for tries prohibiting night work by women also working-hour women can take two forms: prohibitions on have statutory work hours and overtime lim- night work by women and lower overtime its that apply to both women and men. If restrictions limits for women than for men. Both types enforced, these regulations can protect male of restrictions continue to be prevalent in and female workers from unlimited overtime developing countries (Nataraj and others and exploitation. Additional restrictions on 1998; Hijab and Fawzi El-Solh 2003). night work by women can then be eliminated, How are working-hour restrictions (night as they discriminate against women. By lim- or overtime) for women likely to affect the iting employers' flexibility, any overtime lim- labor market? Such restrictions will reduce its are expected to have labor market effects. female workers' working hours. Because total But if overtime limits are of the same mag- labor demand will remain the same--and as nitude and are equally binding for women long as male labor supply is not perfectly and men, they will not affect male and female elastic--the reduction in women's working workers differently. hours will increase wage rates for both men and women. Employers will substitute men's Policy implications hours for women's in production, increas- The evidence above suggests that maternity ing men's average working hours. As male benefits and working-hour restrictions for earnings (hours times wage) rise, the level female workers have opposite effects on of male employment will increase as well. female employment. Working-hour restric- The effect on the level of female employ- tions specific to women can lower their work- ment will depend on the combined effect ing hours and employment rates. To prevent on earnings of reduced working hours and these adverse effects, policymakers should increased wages (Landes 1980; Zveglich and consider eliminating sex-specific working- Rodgers 2003). hour restrictions and protect male and There are few studies of the labor mar- female workers through overtime limits that ket effects of working-hour restrictions in apply equally to both sexes. developing countries. Zveglich and Rodgers In contrast, mandated maternity benefits (2003) used data from Taiwan (China) to can have a positive effect on female employ- assess the impact of working-hour restrictions ment rates and may increase women's attach- on the female labor market. In 1984, Taiwan ment to the labor force. The potential (China) introduced the Labor Standard Law, negative wage effects and discrimination due PREMnote 94 December 2004 to employer-financed maternity leave provi- Workers in Industrializing Countries." sions can be mitigated through cost sharing International Review of Comparative Public between employers, employees, and govern- Policy 10. ment, and by limiting maternity leave to 12­16 Pan, P. P. 2002. "Few Protections for China's weeks (in line with recommendations by the New Laborers." The Washington Post, 13 May. International Labour Organization). Politakis, G. P. 2001. "Night Work of Women in Industry: Standards and Sensibility." Further reading International Labour Review 140 (4). CASS (Chinese Academy of Social Sciences Ruhm, C. 1998. "The Economic Conse- Research Group). 1995. "Studies of Fe- quences of Parental Leave Mandates: male Workers: Migration for Work and Lessons from Europe." Quarterly Journal of Maternity benefits Development of the Countryside and Peas- Economics 112 (1). ants." Sociological Studies 4 (in Chinese). ------. 2000. "Parental Leave and Child can increase Gruber, J. 1994. "The Incidence of Mandated Health." Journal of Health Economics 19 (6). Maternity Benefits." American Economic Summers, L. 1989. "Some Simple Econom- women's Review 84 (3). ics of Mandated Benefits." American Eco- Hijab, N., and C. Fawzi El-Solh. 2003. "Laws, nomic Review 79 (2). employment rates Regulations and Practices Impeding Waldfogel, J. 1998. "Understanding the `Fam- Women's Economic Participation in the ily Gap' in Pay for Women with Children." and attachment to MENA Region." World Bank, Gender and Journal of Economic Perspectives 12 (1). Development Group, Washington, D.C. Winegarden, C. R., and P. Bracy. 1995. "Demo- the labor force Huang, X. 1999. "Divided Gender, Divided graphic Consequences of Maternal-leave Women: State Policy and the Labour Mar- Programs in Industrial Countries: Evidence ket." In J. West and others, eds., Women of from Fixed-effects Models." Southern Eco- China--Economic and Social Transformation. nomic Journal 61 (4). New York: St. Martin's Press. World Bank. 2001. Engendering Development. Human Rights Watch. 1996. "No Guaran- A Policy Research Report. New York: tees--Sex Discrimination in Mexico's Oxford University Press. Maquiladora Sector." New York. [http:// Zveglich, J. E., and Y. Rodgers. 2003. "The hrw.org/press/08/mexicomag96.htm]. Impact of Protective Measures for Female ------. 1998. "A Job or Your Rights--Con- Workers." Journal of Labor Economics 21 (3). tinued Sex Discrimination in Mexico's Maquiladora Sector." New York. [http:// This note was written by Alexandra van Selm hrw.org/reports98/women2]. (Consultant, PREM Network Gender Unit and ILO (International Labour Organization). World Bank Institute Human Development Divi- 1998. "Maternity Protection at Work." sion) under the supervision of Nayantara Muk- World of Work 24 (April). erji and Susan Razzaz. The author is grateful for Landes, E. 1980. "The Effect of State Maxi- useful comments from Amy Luinstra, Karen mum-Hours Laws on the Employment of Mason, Yana Rodgers, and Cecilia Valdivieso. Women in 1920." Journal of Political Econ- If you are interested in similar topics, con- omy 88 (3). sider joining the Gender and Development The- Nataraj, S., Y. van der Meulen Rodgers, and matic Group. Contact Wendy Wakeman, x33994, J. E. Zveglich. 1998. "Protecting Female or click on Thematic Groups on PREMnet. This note series is intended to summarize good practices and key policy findings on PREM-related topics. The views expressed in the notes are those of the authors and do not necessarily reflect those of the World Bank. PREMnotes are widely dis- tributed to Bank staff and are also available on the PREM Website (http://prem). If you are interested in writing a PREMnote, email your idea to Madjiguene Seck. For additional copies of this PREMnote please contact the PREM Advisory Service at x87736. PREMnotes are edited by Paul Holtz and laid out by Suzanne Luft. Prepared for World Bank staff