Report No. 10299-TU Turkey Women in Development Report (In Two Volumes) Volume 1: Main Report September 25, 1992 MICROFICHE COPY CountryOperations Division Report No.:10299-TU Type: (SEC) CountryDepartmenttl Title: WOMEN IN DEVELOPMENT REPORT Europe and Central Asia Regional Office Author: H-ADLER, S. Ext. :32765 Room:H5 103 Dept.:EC1CO 2 VOL. FOR OFFICIAL USE ONLY Dom Of tiw Wo.d Bank This' doctument has a restricted distribution and may be used by recipients only in theeroace of their official duties. its contents may not otherwise be -disclosed without World Bank authorization. TURKEY WOMEN IN DEVELOPMENT REVIEW CURRENCY EOUIVALENTS Currency Unit - Turkish Lira (TL) US$1.00 - TL4100 (1991 Average) ABBREVIATIONS LIST CEDAW - Convention on the Elimination of All Discrimination Against Women DGWSP - Directorate General for Women's Status and Problems DI - Duncan or Dissimilarity Index EDI - Economic Development Institute of the World Bank ESC - European Social Charter FRI - Family Research Institute GAP - Southeastern Anatolian Project GT - Government of Turkey HIPS - Haceteppe University Institute of Population Studies IIBK - Turkish Employment Organization ILO - International Labor Organization IMR - Infant Mortality Rate LFS - Household Labor Force Survey MARA - Ministry of Agriculture and Rural Affairs MOE -. Ministry of National Education MOH - Ministry of Health MOL - Ministry of Labor and Social Security NFTW - National Foundation for the Advancement and Recognition of Turkish Women NGO - Non-Governmental Organizations SII - Social Insurance Institution SIS - State Institute of Statistics SPO - State Planning Organisation SSCPA - Social Security and Child Protection Agency TDF Turkish Development Foundation TFR - Total Fertility Rate TRT - Turkish Radio and Television TSI - Institute of Turkish Standards TUBITAK - Scientific and Technical Research Council of Turkey UNDP - United Nations Develcament Programme UOE - Undersecretariat of Environment WE - Women and Employment Index FOR OMCI4L USE ONLY Table of Content. Volume I. Main Report Bx-e,u,tku lymu 8,"i-xv I. The Situation of Women ln Turkey 1 A. Introduction 1 B. The Social, Polltleal and Feonomlc Posltlon of Women 1 C. Legal Status 2 D. Fertlilty, Nutrition and Health 4 1I. Women in the Labor Foree 10 A. Introductlon 10 B. Labor Force Datat Sources and Issues 1S C. Trends and Patterns ln Labor Force Partielpation 17 D. Gender Segregation in the Labor Force 24 B. Urban-Rural Dlfferences and Dimensions 31 F. Hours of Work 39 G. Bducational Attributoe of the Labor Force 41 H. Male-Female Barnings DLfferences and Soclal insurance Coverage 47 I. Pollcy Impllcations 47 SII. Women in the Eeonomr 54 A. Introduction 54 B Women's Work and Employment Patterns 54 The Agrliultural Sector 55 The IndustrLal Sector 57 The servlees Sector 61 C. Issues and Proposals Concerning Women's Productlvlty 63 Cross-Sectoral ConstraLnts 63 Agrlcultural Sector Constraints 67 Industrial and ServLcees Sector Constraints 69 ThL report is based on the fLndLngs of a World Dank misslon that vLiited Turkey ln July 1991. The miesion consisted of Sandra Hadler, mLsslon leader (BCICO), Yasmeen Mohluddin and Morton Btelener (consultants). Background materlal was provLded by Nlchael Nertaugh (ECiNR), Nedret Duratan (ECITU), Bulma Aksit, YLides Ecerit, amer Karasapan, Nurhan Sural (consultants). Secretarial support was provlded by Barbara Mondestin. of their official duties. Its contents may not otherwiw be disclosed without World Bank authorization. nV. Education and Technical Trainina 71 A. Introduction 71 S. Structure of the EducatLon and Training system 71 C. Statut and Trends in Female Zducation and Training 72 D. Attrlbutes of the Out-of-8chool Population 78 S. Gendor Aspects of Education and Training Programs 80 F. Conclusions and RecommendatLons 83 V. The Role of Government and NGOo in Women's Develoment 85 A. Introduction 85 B. Development Plans and Government Policiea 85 C. Current Role of Government in Women's Development 86 D. Non-Governmental organizations 90 B. Conclusions and Areas for Intervention 92 References & Biblioaranhv 94 Box 1s Employment Equity Programs 31 Box 2: Time Use Surveys 53 Volume II. Annexes Annex Is The Legal Environment Annex 2s Fertility, Nutrltion and Health Issues of Women Annex 3: Labor Force Deflnitions Annex 4s Women in Industry Annex 5s Women Ln Agriculture Annex 6: Credit Survey (ii) 9bantgs ia Table 2.1s Trends in Labor Force PartLolpation by Cender, 1955-90 18 Chart 2.1: Trends in Labor Force Particlpatlon by Gender, 1955-90 18 Table 2023 Female Participation Rates ln Turkey and Comparator OECD Countries, 1950-90 21 Chart 2.2s Female Participation Rate. Ln Turkey and Comparator OCD Countries, 1950-90 21 Table 2.3: Female and Male Partlcipatlon Rates by Age Groups, 1955-90 23 Chart 2.3: Female and Kale Participation Rates by Age Groups, 1955-90 23 Table 2.4: Distribution of Total Labor Force by Worker Status and Gender, 1960-90 26 Table 2.5: Distribution of Total Labor Force by Occupation and Gender, 1965-90 28 Table 2.6: DLitributlin of Total Labor Force by Sconomic Sector and Gender, 1965-90 29 Table 2.7: Participation Rates by Urban-Rural Areas, 1955-90 32 Table 2.8s Urban-Rural Participation by Gender and Age Groups, 1988-90 33 Chart 2 8: Urban-Rural ParticLpatlon by Goender and Age Groups, 1988-90 33 Table 2.9: Dlitrlbution of Urban-Rural Labor Force by Worker Status and Gender, 1990 35 Table 2010s Dlibritutlon of Urban-Rural Labor Force by OccupatLon and Gender, 1990 36 Table 2.11: Dlitribution of Urban-Rural Labor Force by Economic Sector and Gender, 1990 37 Table 2.12: Weekly Hours of Work ln MaLn Job For Those mployed, 1990 40 Chart 2.12: Weekly Hours of Work by Residonce, 1990 40 Table 2.133 DLstrLbutLon of Total labor Force by Education and Gender, 1970-90 42 Table 2*14: Distrlbution of Labor Force by Education, Roeldence and Gender, 1990 44 Table 2.15: Female Participation Ratlo by Educational Attainment, 1990 46 Chart 2.1S: Pemale ParticLpatLon Ratlo by Educational Attainment, 1990 46 Table 3.1: Distribution of Labor Force by Beconomic Sector, 1990 55 Table 3.2: DistrLbution of Employed Women in Cowmunity, Social and Personnal Services by Occupational Groups, 1990 62 cAmeZJIV Table 4.1: Share of Female Xnrollments in Education and TraiLning by Level and Type of Program, 1987-88 73 Table 4.2: Enrollment DefLcits for Males and Females ln Basic Educatlon by Level and Source, 1987 75 Table 4.3: A RegLonal ComparLson of Female Schooling and Llteracy 76 Table 4.4: Promotion, Repetltion and Dropout Rates by grade for Males and Females, 1987-88 77 Table 4.5 Growth of Enrollments ln Formal Education and Training for Males and Females by Level and Type of Program, 1980/81-1987/88 78 Table 4.6. Illiteracy Rates and Gross Enrollment Ratios by Region nd Provinces for Males and Fmales 80 (LiL) 1. This report is intended to serve as the baLss of the World Rank's support to the Government of Turkey for LmplemntLng lts pollcles to enhance the partLiLpatLon of women ln the economlc dwevlopoent process. it provLdes background LnformatLon about the sltuatlon of women ln Turkey, lncludlng legal and health lssues, and examlnes Government strategies to Lneorporate women Lato national polLcLes and program. The maln focus of the report, however, La on lncreaslng and ralolng the quallty of women's partlclpation in the labor force. To thli end, the report glves particular attentlon to enhanclng women's labor sklls and income-earnLng potentLal through educatlon, tochnical tralnlng and credit; and to openLng thelr access to hlgher paying jobs ln Lndustry and agrlculture. Straggaic Considerat ions and Direct ions 2. Turkey can beneflt slgnlfliantly ln terms of lncreased productlvity, enhanced famlly welfare and reduced populatlon growth by focusuLng more on the development potentlal of lt. women. To this end, the report advances three general propositions. 3 investments ln women will promote economlc development. Because of the crltlcal role women play ln their famlly's health and nutrltlon and ln the educatlon of children, the b-neflts from Lnvestment Ln women's development tend to be spread more wldely throughout soclety and across generatLons; o Women contribute far more to the economy and to the family than is reflected Ln offLeLal statletlcs. Recognlzlng the role of women ln the development process (through the adoptLon of better atatLstLcal LndLcators) wl11 help policy- makers in making lnformed decLeLons on resource allocatlon. Simultaneously, lncreaslng the visibLilty and awareness of women's contributions in the public and prlvate sectors, as Government has recently begun to do, and through the use of the medla would facilitate the integratLon of women in the development processa and ° Although women's contrlbution li substantLal thelr productlvlty remalns low because of constralnts of culture and tradltlon, even though the laws and pollcles are, ln general, gender neutral. EasLng these constraints, for instance through "moral suasion", would lead to productlvlty lncreases and increased bousehold income. 3. Turkey's contlnued rapid development and Lategration lnto Burope will be facilitated lf the constraLnts faced by women wlth respect to human and productive resource development are alleviated. Without approprLate - LL - Lnvestment ln women's education, traLning and health and Lncreased female participation Ln productlon, the economy wlll bear the consequences ln terms of foregone production, dimLLnshed famlly welfare and hlgher populatLon growth. The Situ&tion gf Women 4. The posLtLon of women ln Turkey varLoe over an extremely wlde ranges from the hlghly educated, professional women in the major cLtLes to the majorLty of women in the urban and rural areas who are caught between two dLfferent worlds - one, determLned by culture and tradltlon that limlts thelr actLvitLes to the famLly homestead, and the other, shaped ln large part by the contlnulng mlgratLon that brLags them lnto the modern, productlve sectors. Durlng the past decade, however, tradltlonal norms relatlng to women have been breakLng down. Nevertheless, Government hao recognLied that gender-neutral polLciLs are not suffLeLent for fully exploltlng women's potentLal contrLbutLon to the economy, and that women's Lisues thus warrant a specific focus. Notwlthstanding the far-reachlng reforms lntroduced ln the early Republiean era, women's access to servles that can help them to acquLre knowledge, obtaln esentlal technLeal and social services to overcome gender- specifLc constraLnts to thelr labor force particlpatLon remaln limLted. Thus, the majorLty of women face limited economic opportunltles and contlnue to occupy subordinate posLtions in the household and the economy. It is on this majority of women that the report focusses. 5. Although the life patterns of the majorlty of TurkLih women remain condltloned by male-domLnated LnstitutLons relating to the famlly, soclety and the economy, the tradltlonal norm of a woman remaLaing inslde the homstead, and of a male providLng "protectLon", has been contLnuously changLng as a result of economlc and demographlc developmant. Thli evolutlon is resisted, however, by the more conservatlve elements of society, whlch have pushed durLng the elghties for a renewed smphamLs on women's role as wife and mother. Women ln Turkey continue to marry early and, desplte deliLnlng natlonal fertlILty rates, high fertlilty patterns remaln the norm ln ome regLons. overall, the prevaliLng social norms still leave women dependent on men or at rlsk when deserted and continue to produce a strlct dlvlilon of labor and hlghly gender-segregated labor market. 6. Women make a major contribution to the economy and the famlly. Not only do they partLcLpate slgnlflcantly ln agrlcultural and industrLal labor, but also they are charged wlth general household responsLbliLties, chlldcare and caring for the old and lntlfm. Nonetheless, women's contributLon is seriously understated ln the national accounts because lt is unpricied and largely unmarketed. For the majorLty of women, thelr biliLty to be independent and take lnltiatve,, acquLre new ideas, skll. and contacts and to work outslde the homestead remalns restrlcted by traditlon and culture. In thelr labor-force participation, women remain a largely unskllled or semL- skLiled resource. Thus, the majorlty of women ln Turkey have not benefltted to the same extent as men from the far-reachlng reforms of the early Republican era or from economlc development. - LLL - 7. The Isues Turkey faoes ln simultaneously LncreasLng economlc efflclency and the status of women *r virtually the sam as those facod by the OlCD countries followLng World War I and, more recently, by the newly lndustrLalized countries. With the transformatlon from a prlmarily agrlcultural economy to an LncreasLngly lndustrLalLsed one, and wlth the consequent changes ln the soclal fabrlc over the past three decades, Turkey has experienced a doclLne (- -%lo labor force partLcLpatlon, whlch will likely be reversed. if th qwLng ln labor force partlclpatlon is to be accompanLed by growing ecot effLLcency, measures are needed to lncrease the productlvity of the famaiL labor force. Thus, tho focus of the report is on meanures thet will support both an Lncrease ln female partLclpatLon, enhanoe womon's skLils, and open up productlve and non-traditLonal occupatLons to womn. Female labor force partLiLpatLon ln Turkey ia currently around 330, compared to 74% for men. If female partLeLpation were to increase to 50% (somewhat lower than the current OECD norm), ceterie paWLbus, this would amount to an increase ln the current female labor force from 6.5 mllion to. 9.5 millon. At the same tlme, increaslng urbanLzatLon and educational attaLiment wlll result ln slgnlflcant changes ln sectoral and occupatlonal composLtLon of the female labor force. 8. The lmpllcatlons of these changes are complex and will affect all aspects of Turkey's economic and social development. As a result, the coverage and the fLndings of thli report are unusually wLde-rangLng, touchLng on cultural and tradltional issues that are normally beyond the purvlew of Bank reports. The report offers a large number of proposals, but as noted above, priorlty is glven to measures to support an Lncrease Ln female partlilpatlon and employment ln productlve, non-traditlonal occupatlons. To provLde perspectlve, section A below assesses the posltion of women in Turkey today and in a group of middle-Lncome comparator countrles in terms of selected soclo-economli indicators. Sections B-D summarLze the maln fLndLngs and recomendatlons of the report, focusslng ln partliular on the partLclpatian of women ln the labor force and on sectoral iesues that can sLmultaneously strengthen womns *standing and contrlbutlon. sectlon 3 addresses Government's strategLes to incorporate women Lato natlonal policles ad program. A. Procress in Women's Welfare and PartLcLpation 9. AccordLag to almost every socLo-economlc lndlcator, the standlng of women ln Turkey has Improved slgnlfLcantly ln recent decades. Regional varlatlon and comparison wlth a group of mlddle-income countrles (see IndLcatore Table below) indicate, however, some areas where further improvement ls desirable. Pour groups of lndLiators, ln partliular, are useful ln summarlzLng Turkey's progress. The fl st relate to life expectancy and the number of women livLng relatlve to men (the eox ratio). The second relate to health and fertliLty lndLcators - not only as a means of reducing population growth but also of glvLng women greater control over thelr lives. The third relate to human resource development and, ln partliular, educatlon. The fourth group relates to women's participation ln the economy and their contrLbutLon to household incom. - iv - Indicators of the Status of Women in Turkey ond Setected Cowttrfes Indfcator Year Turkey Colombia Greece Mexico Portugal Spain Population (m) Mid-1989 55.0 32.3 10.0 84.6 10.3 38.8 Population growth rate (K) 196s-80 2.4 2.5 0.7 3.1 0.4 1.0 1980-89 2.4 2.0 0.4 2.1 0.6 0.4 1989-2000 2.0 1.6 0.2 1.8 0.4 0.4 Populeaton (m) 2000 68 38 10.0 103 11 41.0 GNP per capita (S) 1989 1370 1200 5350 2010 4250 9330 GNP per capita growth (K) 1965-89 2.6 2.3 2.9 3.0 3.0 2.4 Daily Calorfe tupply par capita 1988 3080 2561 36"9 3135 3382 3543 Life Expectancy at birth Female 1965 55 61 72 61 68 74 1989 69 72 80 * 78 80 Male 1965 52 57 69 58 62 69 1989 64 66 74 66 72 74 Total fertility Rate 1965 5.7 6.5 2.3 6.? 3.1 2.9 1989 3.6 2.9 1.5 3.4 1.6 1.4 2000 2.7 2.2 1.6 2.4 1.7 1.6 glrried Women of Chiledbaring age using contraception (K) 1986 77 63 .. 53 .. 59 Maternal Mortality Rate (per 100,000 live births) 1980 207 130 12 92 1s 10 Itfant Mortality Rate (per 1000 live births) 1965 165 86 34 82 65 38 1989 61 38 11 40 13 8 Low birth wefght babies (a) 1985 7 1s 6 15 8 Births Attended by Health Staff (K) 1985 78 51 .. .. .. 96 Adult Illiteracy Rate Female 1985 38 13 12 12 20 8 Total 1985 26 12 8 10 16 6 % of Age Group Enrolled in Primary School (1988) Female 113 115 102 115 127 110 Total 117 114 102 117 126 111 X of Age Gro*p Enrolled In Secondary School (1988) Female 34 56 93 53 63 111 Totat 46 56 95 53 59 105 X of Age Group Enrolled in Tertfary Educatfon (1988) Total 11 14 28 1S 18 32 X of Central Goverrnent Expenditure on Health 1972 3.2 .. 7.4 4.5 .. 0.9 1989 2.9 .. .. 1.7 8.2 12.5 Education 1972 18.1 .. 9.1 16.4 .. 8.3 1989 15.7 .. 27.5 12.3 10.0 5.1 Total Exp. as X of UP 1972 22.7 13.1 .. 11.4 . 19.6 1989 23.7 14.6 .. 21.2 43.3 34.3 Female Labor Force (X) 1960-65 38.2 12.1 19.0 9.5 16.0 13.0 1988 a/ 30.1 14.4 20.0 18.4 31.4 17.4 Sources: World Development Report 1991, Social Indicators of Development, 1990 A/ Most recent estimte. v Life Zxgctancy and sex R&tUo 10. Women in Turkey constitute 49% of the population, compared to an average sex gato in the industrialized countries of 52% and 48% in the low- income developing countries. This suggests that women are slightly disadvantaged compared to men in Turkey in terms of health and nutrition. The background health study (Annex 2) concludes that this is a result primarily of women's reproductive role. The life exRegtancy differential between men and women in Turkey, however, reveals a very rapid gain in female life expectancy. Recent data indlcate a female life expectancy of 69 years against 64 years for men - compared to 59 and 58 years, respectively, in the early 1950s. The comparator data provided in the Indicators Table show that both the level and improvement in male and female life expectancy in Turkey are commensurate with that of the middle-income countries. Health and Fertility indicators 11. General health indicators of women and their children have improved significantly since 1960. The infant mortality rate, in partLcular, has declined dramatically from 165 in 1965 to 61 in 1989 - although this remains higher than in the comparator middle income countries. Simultaneously, fertility rates have declined significantly. Women give blrth an average of 3.7 times - down from 6.8 in 1960. This also is higher than for the comparator countries. Turkey has made strong progress in family plannLng since the introduction of the first Family Planning Law in 1965 and now has one of the highest contraceptive acceptance rates in the world (77% in 1988, which is on a par with Germany). The population growth rate, however, remains unchanged since the mid-1960. at 2.4% and suggests a margin for improvement. Fertility and health indicators, in general, show a continuing and significant regional variation that warrants greater attention in the design and implementation of health and information delivery services. ZducatL2n 12. Universal education has been an important principle of the country since the founding of the Republic, and the Constitution assures equal educational opportunity for boys and girls. Since 1970, Turkey has made slgnificant progress in reducing the gender gap in education and literacy. The proportion of illiterate women in the female labor force has fallen from 69% in 1970 to 29% in 1990; the corresponding percentages for men have also fallen from 29% in 1970 to 8S in 1990. Gender segregation however exists, particularly in vocational schools, and female enrollments decline with increasing grade level At the secondary level (Indicators Table),, Turkey lags behind the mioule-Lncome comparator countries in female enrollments. PaXt4n44A4ion 13. Female labor force garticioation in Turkey is 33% - significantly higher than in the comparator middle-income countries. Compared however to OZCD countries, the female participation rate is one of the lowest. With a male participation rate of 74%, the gender differential in participation is also one of the highest in the OECD. - Yv. - 14. ignder nartiatatlon rgtes by urlian/rural resdance show marked dlfferences the par tcipation rates for womn and men ln rural ateas are 51% and 78%, respectively, and ln urban areas 16% and 70%, respectlvely. Glven the urban migration since the 1950e - the proportlon of Turkey's population livLng ln urban areas has virtually doubled from 29% in 1955 to 59% in 1990 - the continuing low urban female partlcipation rate La surprLaing. This is attrLbutable, at least partially, to the omiLsLon from offlcial labor statlstics of informal sector workers, as well as more limited opportunities for women's employment ln urban areas and social constraints to their employment (Chapters IZ and Annex IV). 15. aSatXoal emnlove nattrnsa for men and women * -- llffor significantly. Approximately 74% of women ln the labor fort -.e employed ln agriculture, 7% Ln Lndustry and 12% ln servLies, whle mn are more evenly distributed 4cross sectors - 34% in agriculture, 24% in Lndustry and 38% in servLcso. Simllorly, there is a marked differenco ln gender *onient JansąŁt 68% of women in the labor force in 1990 were unpald famlly workers and 25% were wage earners or self-employed, compared to 14% of men who were unpald and 82% who were wage earners or self-employed. Analysis of data by eu2ion "an gegnMic sector, show women concentrated in the lower occupational categories and performing work closely related to their traditLonal household duties. 16. When the lower educational attainment of the female labor force (almost 30% of the female labor force but only 8% of the male labor force were illiterate in 1990) is also taken into consideration, these differential patterns indicate a labor force that is markedly segregated by gender and in whlch women remain a largely unskilled or semi-skilled resource. The high proportion of women who are unpaid family workers also ralseo concerns regarding the recognition of women's contribution to houshold Lncome and their igfluence in household decision-making. . Women in the Labor Force 17. Increasing women's abillty to contribute to a country's development plays a critical role in simultaneously increasing overall oconomic efficLency and improving the relative economlc status of women. While Government recognises the importance of women to Turkey's economic development there has been little consideratLon, to date, by policymakore as to how to further women"s integratLon into the eoonomy. Thli section briefly oumuArizes the major report findaings concerning the labor force partioipation of Turkish women and policy recn.nses that would enhance the ability of women to become effective agents in Turkey's development (chapter 2). The major findings ares There has been a sha= dea lLne in female labor force 2 rtLcLMLtLcn in Turkey - from almost 70% ln 1955 to 33% in 1990. This is consistent wlth the widely-documented U-shaped relationship between economic development and female partLclpation observed in the OZCD countrLes since World War 1U, and where female laWbor force partLcLpatLon now exceed .50%. A comparison of the female participation pattern in Turkey with that of four southern Buropean OECD countries suggests that - viil - Turkey La lkely to experience a secular upturn In female partLeipatlon (par s. 2.13-19). ow qulokly the second stage of the U-shaped relationship materLalLseo will depend largely on two consideratLons the avaLlabillty of employment opportunities for women and the supply of skilled and qualLfied women who could take advantage of these opportunlties. Policy lnitLatlves focussing on woen's partlcipation in non- tradltLonal actlvltLes will be needed to support thLs upturn (paras. 2.55-62). The work force i uevimrklysreaed b edr especlally ln urban areas, where people tend to have hlgher educational attalaments and wider employment opportunltLes. Whlle the gnader segregation is simLlar to that obsrved in OECD countrLes, lt li more lntense and, as the experience of the OECD countries has shown, will be dlfflcult to allevLate wlthout polliy lnterventlons focusaing on, Lnter alla, education and training, the introductLon of employment eulty programs and repeal of dLscriminatory protectlve legislation (paras. 2.22-32). o In addltlon to the potential upturn in female partLeLpatLon, there are2 mAicr mallowcnen eadn h asc JOf Ln the labor for*s. The impact of the lncreasLngly educated female labor force has not been reflected in the labor markets, although this wll be largely a matter of time (pars. 2.44- 52) * For lnstance ln 1990, 50% of the urban female labor force but only 5 of the rural female labor forco had moe than a prlmary school educsatLon. Currently, the educatLonal lovel of the rural women will constraLn their labor force participation, regardless of the availabliLty of employmnt opportunitie's. Based on current enrollment rates, the proportion of rural women wlth more than a prLmary school education will be slgnlflcantly hlgher a decade from now, as will be thelr employment expectatlons. ThLs underlnes the need for pollcy measures deslgned to promote the employmnt of women in hlgher productLvlty jobs ln both urban and rural areas. such measures are Lmportant not only for women' sgeployment but for reglonal development. 16. Polliy lnitLatlves that governmnt may wLsh to consLder Lnclude the following: Aro inortnt - if no the key element - Ls edaatin icluding all forms of schooling: fomal and informal, adult educatlon, vocatlonal and on-the-job. Raising women's educational attainments or earnangs capaclty - especially at the post- prlmary lvelos and in programs, courses, and fLilds of study that would meet the skill requLrments of industrialization - w$il improve both economia effLciency and women'sa conomic status. Increased and Improved education will raLse thelr - vlll - partLiLpation rates, reduce the likellhood of unemployment and lncreass contlnulty ln thelr labor force attachment. Moreover, changLng the patterns of women's educatlon will likely reduce the gender wage gap and segregatLon ln the work place, as well as change attLtudes towards the gender dlvlslon of labor. ChangLng educatLonal polLcies li a necessary but not sufflclent condltlon for LmprovLng tho status of women. If educatlonal polLeLes are to be successful ln removing gender dliparltles Ln the work place, they must be comolemented by labor Market- related inLtlaivau. To thle end, there are several pollcy measures that could be pursued. PoliciLe for integratLna women ln the develooment orocess need s2ociftcally to focus on broadenLna woman's skill ranae. Much of the observed gender dLiparLty ln partLclpation rates and earnLngs, and the persistence -f segregatlon ln the work force can be attrlbuted to gender dlfferences ln educatLon and tralnlng. Reduclng these dlfferences will impact favorably on the status of women and on Turkey's overall development. o Prress towards elLmLnatLna ander segaraation, can be promoted by the Government itself setting an example to ensure symbolically and substantLvely that gender segregation is not perpetuated. The implementation of employment equity LnitLatives, along the lLnes pursued ln Canada (Chapter 2, Box 1), would also represent an important step ln facliltatlng the equal access of women to employment opportunlties. Employment equity programs, not to be confused wlth "affirmatlve action" or quota initLatives, alm at proventlng or redueLng disadvantages experlenced by desLgnated groups by desLgning and implementLng agency-speciflc measures to overcome ldentlfled dLeadvantages. The programs are given "teeth" ln both prlvate and public sector agencLes by being tled to federal funding. ActLon ln thls area, would need to be coordlnated simultaneously wlth other LnLtiatlves. * Inter alia, some protective leaLslatLon desLgned to ensure that woman are not exposed to hazardous physical or moral conditions ln the workplace has the unintended effect of restricting women's job opportunlties. Conslderatlon should be given to repealing measures that provide for dLfferentLal treatment of workers by gender, with the exception of maternity protectlon (Annex 1). The experience of several countrles suggests that the repeal of such legislation will play a slgnlflcant role ln changing the attltudes of men and women towards wsuitablew women's work and will also Lnfluence women's career cholces. * There are also constralnts emanating from the labor market that make lt dlfficult for women to seek formal market work. Labor demand constraints, whlch generally arise from labor standards legislation and collective agreements, are reflected ln minimum working hours and ln rigid work schedules. These constraints - ix - compound the difficulty of combining market work with household responsibilities. Encouraging (or mandating) employers to provide more flexible time arrangements by modifylng labor regulations to permit flexitime and part-time would be a useful policy measure. o Women's employment opportunities are further influenced by th costo of emolovment - the location of the jobs reflected in commute time and expenditures on clothing, childcare and meals. Enforcement of existing regulations regarding enterprise childcare facillties and encouragement to establish community or family childcare facilities, for instance, as in the Bank- supported day care projact in Colombia, would be beneficial (para. 3.41 and Annex 4). o A further and hiahlv sianificant oolicy initiative to increase the awareness of women's contribution concerns the collection and publicatlon of data. A comprehensive and up-to-date data base of socioeconomic indicators on the role and relative situation of women is a sLn2 aug non for analyzing and formulating policy initiatives designed to improve women's relative economic status and for monitoring, inter alia, the impacts of such initiatives. Consideration should be given to both the regular publication of a statistical report on the socio-economic status of women (as 8I is now proposing) and the undertaking of policy-oriented research on a wide range of women's issues (paras. 2.62-64). 19. Additional recommendations to further women's participation, which Government may wish to address, include promotional and publicity campaigns (para. 2.65), and increaeing the proportion of women with social insurance coverage (para. 2.61). 20. Chapter III identifies constraints to women's productivity in the agriculture, industry and service sectors and proposes alleviating measures. In aariculture, where almost 75% of women work, the extension and research delivery systems, with their focus on male activities are a major constraint to women's productivity (3.34-38). Inter alia, there are virtually no female extension agents - as opposed to home economists - outside the provincial directorates and extension agents are required to work with titled land owners. Within MARA, pilot programs are now being established to address these constraints through the re-training of home economists as field agents. The opening up of the agricultural extension training program, currently offered only in boy's vocational schools, and explicit consideratlon of women's role in agriculture in both the central and regional offices in planning, extension, technology and research would be useful first steps. The employment of female home economists throughout the country demonstrates the feaslbility of finding women to also work as extension agents. 21. Although less than 20% of the female labor force li employed ln the industry and service sectors, these sectors are of central importance to the integration of women in generating new (non-traditional) employment -X- opportunltlos (paras. 3.08-22). Since 1970, employment growth of women ln the lndustrLal sector has lagged behind that of men, whle- ln the servLces sector lt has slgaLfLcantly outpaced men's mployment. The LnclusLon of tho Lnformal sector ln offLiLal labor statistics would strengthen understandLng of developmnts Ln theme two sectors. Detalled analysis of employment trends and potentlal la also neded to underpln the formulation of pollcy masures to further women's partLclpatlon ln these sectors. 22. The major constraLnts identified to women's productlvlty ln lndustry and servLoes are the hlgh degree of job segregation - 80% of women ln industry work ln the textlles and food sub-sectors and, of these, 80% work ln the lower ocaupational (production) ranks - and labor market factors (para. 18 above). Proposed pollcy measures would focus on broadening trainLng and employment opportunltles in non-tradltlonal occupatLons; facLlitatLng access to resources, espeoially creditg a revliw of labor logLilatLon and regulatlons and thelr enforcementp re-entry opportunltleso and encouragement of chlldeare faclIlties (paras. 3.39-41). To overcom the relatively low proportlon of women entrepreneurs and self-employed women ln Turkey, training ln basLc business and managment skills should also be encouraged. Programs to promote women's entrepreneurshLp developed with 3D? may provide a useful model. Additional proposals focus on the need to reconaLder Government support for homebased work and the urgent ned for broad-based sectoral research on women's participatLon. 23. The legal environment and access to credit constrain women's productLvlty in all sectors, and therefore warrant speciflc consideration. UMIa znv$ronment 24. The background legal paper prepared for thli report (Annex 1) concludes that the legal status of women is at par with that of men, in that the Constitution and the Clvll Code are, for the most part, gender neutral. However, thes also acknowledge unequal status by reserving the right of making special provisions for women. In fact, under the Civil Code, women function under a Opartlal disability" (Law of Persons, Article 8) and untLi the National Assembly reviews the raco _m ndations of the reform comission for the Civll Code, filed wlth the Ministry of Justice ln 1984, men are designated the legal head of the household and have broader rights than women (Law of Family, Articles 152-158). Moreover, although laws relatlng to marriage, divorce, chlld custody, lnheritance and maintenance are governed by the Clvil Code, traditional practlces continue in many areas. For instance, although the laws of property and Laherltance are gender neutral (with the exception of agricultural land), the continuing practice of having immovable property acquired during the marriag reglstered in the man's name lends to difflculties in the case of divorce or death of the husband and, undoubtedly, also constrains the wlfe's access to credlt (Annex 1). 25. Some of the existing labor leglslatlon raLses serlous concerns rgardLng women's full participation (Annex 1, parae. 35-55). For instance, Article 13 of the Labor Law permlts an employer to dismiss a woman worker on ground of pregnancy, whlle some of the protectlve legislatlon in the industrial sector contributes to conflning women to tradltional and low- - xi - paying occupatLons. Once these concerns are addressed, consideration should also be glven to extending labor legLslatLon to the agrLcultural and Lnformal sectors (whLch together employ more than 80% of the female labor force), where workers are currently without benefit of protectlve legislation or regulatlons regardLng workLag conditions. Simultaneously, more vlgorous enforcement of supportlve labor and civll legislation would help alleviate constraints faced by women in their full partlclpation. Women - and men - need assistance in understanding and purouLng their rights under the law; the establlshment of legal advice bureaux and/or counelling facilities should be considered. 26. Wlth a view therefore to furthering women's integratlon, Government may wleh to consider the following actLonst o accelerate implementation of the Civil Code proposals of the reform comuLsoion; o review existing labor legislation and, in particular, protective legislation to ensure lts gender neutrality; o consLder extending labor legislation to the agrLcultural and informal sectors; and o institute measures to enforce existing beneficial leglslation, both civil and labor; Cered&t 27. To facilitate thelr integration, Turkish women - and small enterprLses, in general - need increased access to credlt. In the absence of gender-specLfic data, a lImited survey of women's access to credlt was undertaken for thli report (Annex 6). The results suggest that the access of women (and small enterprlses) to credit ln Turkey is constrained on the supply side by three factorss traditional collateral requirements of flnancial institutions based on land tltle; standard loan formalitles, Lncluding lengthy appllcation procedures; and an emphasis of financial institutions on providing credlt for traditional entrepreneurial activities and, in agriculture, an emphasis on loans for commercial crops, which tend to be male-inteneive actLvities. On the demand slde, access li also constrained by the tendency of women to invest ln activities wlth inherently low rates of return; exacerbated by their low-skill levels, particularly ln business matters; and the prevaillng view that finance is the prerogatlve of the male kin (paras. 3.28- 31). 28. In recent years, programs have been lnitiated to address these constralnts. Government has introduced a number of small credlt programs, such as the FamLly Credlt Program of Balk Bank, where 80% of credits have been for women involved in handicrafte in the home. The overhead costs of the programs, however, appear to be high, and the rationale for supportlng women's homebased employment warrants reconsideration. NGoo, such as the Turklsh Development Foundation and the FoundatLon for the Advancement and Recognltion of Turkish Women have recently and successfully introduced limlted cash and ln-klnd credlt progrss for women ln non-tradltlonal agrLiultural actlvltles. - xii - 29. To ease the credit constraints facing women (and small enterprises generally), consideration should be given to: o strengthening existing programs by simplifying and facilitating both delivery and recovery procedures; ° expanding and strengthening related non-credit activities, particularly in training and business know-how, possibly through entrepreneurship training courses tied to a revolving credit fund; and o increasing the dissemination of information on existing credit facilities. C. Education and Trainino 30. The significant progress in reducing the gender gap in education and the central importance of education and training to the fuller integration of women have been touched on in sections A and B above: Chapter 4 and Annex 4 address in detail Government's education and training programs and their role in furthering women's opportunities for employment, in addition to the non-market and intergenerational benefits of furthering women's education. The major conclusions are: o Notwithstanding the significant progress made in extending universal education and reducing the gender literacy gap, Turkey still lags behind middle-income comparator countries in the proportion of the school-aged population, and of girls in particular, enrolled in middle and secondary schools. o Regional variations in gender enrollments and literacy rates are significant (Table 4.6). The recent Study of Factors Affecting Girls' School Attendance in Turkey concludes that economic, cultural and logistical (transportation) factors strongly influence girls' attendance and recommends that targeted interventions be designed to encourage their attendance in areas of deficient enrollment including, inter alia, consideration to establishing girls' middle and secondary schools for the current generation. o Currently, one-third of vocational and technical schools are designated for girls. Courses offered focus on traditional women's occupations, with little potential for career growth and without consideration of skills needed for self-employment or business management. To address the bias in education/trainino which constrains women's labor force op2ortunities. consideration should be aiven to eliminatina the designation of separate bovs' and airls' vocational and technical schools and the corresoondina distinction in the administrative structure of the Ministry of Education. Simultaneously, there is a need to set up oroarams to attract airls to non-traditional vocational proarams and to proarams - xiii - that are currently male-dominated, as has been successfully done in Morocco. Simultaneously, the expansion of iob-oriented trainina for women, as has been introduced by the Turkish Employment Office (IIBD), warrants further support. D. Fertility. Nutrition and Health Status 31. The background health study undertaken for this report (Annex 2) concludes that, while considerable progress has been realized in recent decades, women are slightly disadvantaged compared to men in Turkey in terms of health and nutrition, primarily as a result of their reproductive role. The major conclusions (Chapter I and Annex 2 present details) ares ° Women's health and fertility indicators show significant regional variation and suggest a need to strengthen regionally the delivery of health and family planning services. Family planning services, in particular, could be strengthened through a focus on modern methods, the use of languages that would be understood by local inhabitants and increased follow-up and counselling. o 8everal factors suggest a strong need for improved information and education. For instance, a high proportion of contraceptive users rely on traditional methods, one-third of Turkish women are anemic, and one-third are obese. Consideration could be given to the promotion of media campaigns and community education programs addressing, inter alia, family planning, maternal care, nutrition and health. o Simultaneously, there are sevoral measures which could improve the health care of women and pregnant women, in particular. These include strengthening health delivery services, for instance through the distribution of iron tablets and iodine supplements, and strengthening the training of midwives and implementing systems of referral. B. Government Plans and Proarams for Women 32. Since the early years of the Republic, Government's view was that gender-neutral policies were sufficient to ensure what had been provided by law. Coinciding with the resurgence of interest in women's issues in the mid- eighties, however, there has been an acceptance within Government of the need for a special emphasis on women's issues (Chapter V). Thus, the current Sixth Five-Year Plan (1990-94) addresses specifically women's issues. Simultaneously, the Directorate General for Women's Status and Problems (DGWSP) was established in 1990, charged with improving women's status and promoting their full integration into the economy and the Family Research Institute, which also addresses women but as the central element of the family was established in 1989. To date, the former has been handicapped by institutional and physical relocations and budgetary constraints. Nonetheless, its establishment is a very positive step and sends clear signals of Government's intentions to strengthen the standing of women. The new Government in November 1991 created a separate Ministry of State for Women, - xlv - provLding clear indieation of commitment at the hlghest levels. In the near term, the MlnLstry will need conasderable support to realize its advisory and catalytlc role. In lLght of the experlince w$th natLonal machlnerles for women ln the 1980s, and as a fLrat step towards implamentLng polLLegl and *ectoral recLmendations, the followLng should be considered. o Ensure that all line ministries and agenceLs see women's developoent as an integral and prLority element of nutgLnal dAni.elgMnt and therefore as part of their own regular responsibilities - i.e., not a responsibliLty to be passed on to the Ministry of State for Women. o Nstablish an inter-minLasterial committee chaired, at least initially, by the Prime Minisatr with representatives from the llne ministrLes and agencies. Bach representatlve should present a timebound action plan for promoting the welfare ani productivity of women in their minlitry or agency and be required to report progrres in its implementation to the committee on a regular basLi. o Establsh womens units at the hlghest levels wlthin the key line mlnlstrLes (for lnstaneo, Labor, Xducation, Agrlculture) responsLble for imp eme nting and monitorLng the action plans and for ensuring that women's iesues are integrated into all mlnlstry programs and policLes. o Ensure that people of the hlghest callber are appointed to key positLons Lnvolved with the formulatlon, implementatLon and monltoring of policieo and programs to improve the saatus of wmen . o Ensure a sustained effort to raise the vlisbility of woman's issues, ln general, and their contrlbution, in particular. Thls can be realLzed through, lnter alLa, medla campalgns, revisLon of currLcula and textbooks, publication of an annual statistical report on the status of women and undertaklng and publlshLng a broad-based program of research on women's issues. 33. Zn addltion to the Ministry, there are a number of NGOs actlvely working on womens Lesues in Turkey (paras. 5 21-27). Greater support for their innovative activltles and closer collaboration with Government agencLes are indicated. Government also ratified the UN Convention on the Elimination of All Form of Discrimlnatlon Against Women (CBDAW) in 1985 and the Buropean Soclal Charter ln 1989* Further progress in elimLnating Turkey's reservations to these conventions li needed. The now Government has indLcated lts lntentlons to comply wlth CEDAW ln the near future. Progress ln compliance wlth the Charter, and the Charter's Actlon Program that is close to being finalLsed with far-reachLng labor recommendations, will be important for Turkey's acceptance to the BC. - xv - 34. NotwithstandLng the signifiant institutLonal progress that has been made, women's Lssue are not clearly addressed Ln Government'a programs., For instance, the need for women to earn an independent icome La not wldely recognised. men are seen to requlre employment to sustaLn the family, whereas women are seen an needing more ioela -genert Lag aatlvitLes to perform ln thelr sparc tlme as eacenslons of domestlc aatlvitLes to supplement famLly lnome. womense contrlbution to both the economy and the household la not fully approlated. There is an urgent need to assess the impact of structural changes in the economy that are causiLg more women to seek wage labor outside th home and to assme the lmpact on the economy through the longer term of the consequencs of the current gender segregation of the labor force. 35. 1An cm geratlng actlvltles carrLed out by women in connection wlth Government traLning program are generally aharacterised by low productivity actlvltlie that are an extension of women's tradltlonal actLvitLes, low returns end weak demand ln the labor markets. Training programs do not appear to generate subatantLal iprovement ln women's lnomes. 3vLdence from small program lnltlated by non-governental organlsatLons ln non-traditional market activltles (for Lnstance, poultry and be-keping) demonstrate the potential to generate sustaLned and profitable mployment for women and to Lncrease productLvity. However, outside of these limited projects in the agrLcultural sector, few instances of attempts to draw women lnto non-tradltlonal activities were found ln the minLittLes involved ln traLnLng or the internatlonal agency programs. The need to lntroduce tralnlng programs that are in the lnterest both of the long-term development of Turkey's economy, and whlch create productive employment for womna wlth potential for career growth, is one of the central themes of thli report. carna is us 213 az! oF "wM IN iuuxu A. Zniiwkian 1.01 In adopting the Swiss Civil Code ln 1926, Turkey becam the fLrst Islamoc country to eliminate the Sharla, the Islamic legal code that underlies the segregation of sexes and differentlal legal treatment. ThLs step cleared the way for comprehensive changes in the status of women ln Turkey. Nonetheless, more than half a century later, there are a multltude of issues relatLag to the role and status of women. In very general terms, these range from the extent to which legal reform and the separatLon of religion from state has affected the tradltlonal life style of women ln Turklih society, to major economic and soclal factors that are impacting on thelr role and abillty to partlcipate ln the country's development. 1.02 Thle report is intended to serve as the basil of the World Bank's discussions wlth the Government of Turkey to enhance the participation of women ln the economlo development process. In light of these discussions, the Bank will pre par an actlon program, whlih would serve as the basis for lts support to the Government to enhance wamen's partLeLpatLon. Whlle touching on the general sltuation of women, the focus of the report Ls on women's labor force partLeLpation and the constraLats they face therein. In this context, the report addresses Lisues concerning women ln four major areass education and technLeal traLning, agrlculture, industry and credlt, together wlth the impact of the legal framework on women' partLclpatLon. In addltlon, the report addresses women's health, fertLlity and nutrltlonal status. The role and impact of Government on women and lts strategies to lncorporate women lnto national pollcles and programs are also examlned. 3. The Social. PolitLgal and 3eonomlc PositiLn of Women 1.03 As an Islamlc country where relLgLon and state are separate, the status of womn in Turkey is perhaps more than in most countrLes, complex and multlfaceted. The situation of women ln Turkey covers an extremely wide ranget from the hlghly educated, professLonal women ln the major clties to the majorLty of women in the urban and rural areas, who are caught between two dLfferent worlds - one, determlned by culture and tradltLon that largely restr$cts thelr activitLes to the famlly homestead and the other, shaped in large part, by the contlnulng migration that brLags them into the modern, productLve sectors. 1.04 Although the life patterns of the majorLty of TurkLih women remaln conditLoned by male-dominated LnstLtutLons relatLng to the famlly, socLety and the economy, the tradltlonal norm of a woman remaLinig LnsLde the homestead and of a male providing *protectionw has been gradually changLng as a result of economLc and demographLc developments. This evolutLon is resisted, however, by the more conservative elements of soclety, whlch have pushed durLng the eighties for a renewed emphasis on women's role as wlfe and mother. Women in Turkey, contlnue to marry early, partlcularly ln the rural areas, and soon enter lnto a paSetern of hLgh fertlilty. Overall, the prevallng soclal norms continue to leave women dependent on men or at rick when deserted and produce a strLct dlvLiLon of labor and hlghly gender segregated labor market. -2- l OS Women contrLbute a great deal to the economy and the famlly. Not only do they partLeLpate sigalfLcantly in agrliultural and LndustrLal labor but also they are -harged wlth general household responsibliLtLes (LncludLng ln aome areas, collectlng flrewood and water), chlldoare and carLag for the old and Lnfirmo Nonetheless, a woman's contributLon goes unrecognLzed ln the national accounts because lt li unprLced (difficult to value) and Lnvciible (consumed immediately). For the majorLty of women, their abillty to be Lndependent and take LnLtiatlve, acquire ne ideas, skills and contacts and to work outslde the homo li restrcLted by tradltlon and culture. Thus, the majorLty of women have not benefltted to the sam extent as men from the far- reachLng reforms of the early Republican era or from economic development and ln terms of labor-force participatlon, women r maln a largely unsklled or semi-sklled resource. It le on this group of women that the report focu5ses. 1 06 To better understand the posltLon of women ln Turkey, the impact of the early Republican era reforms ned to be taken lnto account. Speclflcally, by elLmlnating segregation and differentlal legal treatment of women the reform cleared the way for comprehensLve changes in the posltion of women. In the 1920's, when even the SC and many Amerlean radio statlons would not conaLder the employmont of a woman announcer, TurkLsh radio employed women as both announcers and tohaLcLans.11 In the 1930s, women began to enter public offSce and the professLons and were glven strong oncouragement and support. In Parliament, the number of female deputles reached a peak of i8 (4.5%) ln 1934. Although flrat admitted to academic positions ln 1932-33, by 1946-47, 44% of all faculty in the Natural Sclinces were women YV Today, amongst the profeusLons there remain a marked absence of gender-typLng, whlch is thought to be the result of the encouragement to women in the early Republican era. Nonetheless, and whle- the educated urban Turklh women are well-represented in the professions and public offLco compared to many middle- lncome developing countries, there has been a declLne ln thelr particlpation since the mld-1940s. The position and statue of the majority of TurkLih women, however, appears to have changed little since the beginning of the Republican era. C. Ldgal Statia 1.07 Notwithstanding the gnder-neutral character of the 1982 ConstitutLon and the 1926 Clvll Code adopted from 8wltzerland, there aro a number of areas Ln which the legal framework constrains women's full particlpatlon. Anex 1 examnes in detail the treatment of women under the country's legal framework, both as the law is wrltten and as it is carrled out ln practlce, in llght of tradltlonal, cultural practLces. An attempt is also made in the Annex to ldentlfy speclflc constraints to the abilLty of women to contrLbute to the development process that are imbeddod ln the legal framework and to lndicate where remedial measures could improve the treatment of women. This sectlon briefly summarLsee the main Lssues. 1J XLrby (1960). V See Acar ln OECD and GOT, 1989. -3- 1.08 Within the ConstitutLon two articles raLse concerns: Article 41 on the Protection of the Family mandates that the State shall take the necessary measures and establish the necessary organlzation to ensure the peace and welfare of the family and the protection of the mother and childrent and Article SO on Working Conditlons and the Rlght to Rest and Leisure states "minors and women" shall enjoy special protection regarding working conditions. While seemingly paternalistic, these articles are less than benign in practice. Although discrimLnation is proscrLbed under the Constitution, instances occur and underline the dlfficulties, in practice, of eliminating discrimination, obtainLng legal redress and to the need for a legal mechanism to provide redress for employment discrimLnatlon. 1.09 The Civil Code also does not provide for the full equalLty of women. Articles 151-158 on the famlly deflne the husband as the head co household, who determlnes domicile and whose views prevail on family matters in case of dispute. The annulment in 1990 of Artlile 159 which required that a woman obtain her husband's p9rmlssion to work, represents a major step forward for Turkish women. While the provisions of the Civil Code regarding women's issues were progressive at the tlme of adoption some have proved to be insufficient for society today. Accordingly, Switzerland passed slgnificant amendments regarding women's rights in 1984. In Turkey, a commlssion was appointed to prepare amendment proposals (includlng Articl- 151-158) and submitted its report to the Ministry of Justice in 1984. The proposed amendments have still to be approved by the National Assembly, but the current Government is proposing their early consideration. 1.10 The laws on Education, Property, Banking and Credit Regulation are essentially gender-neutral. However, the lmplm_ntation of the laws relating to property and inheritance (particularly in the context of immovable property acquired during marriage), in effect, favor men and their access to credit (Annex 1). The Laws of Succession (in partlcular Articles 597-598, relating to agricultural land that give priority to male chlldren) and the Nationality Code (notably regarding acquisitLon or loss of Turkish nationallty other than by birth) do however treat women differently. The former especially, may constrain women's participation, given that women constitute SO% of the agricultural labor force. 1.11 The major concern of this report within the legal framework, however, is the labor legislation. While many of the labor issues relate to both men and women (for instance, job security) others, such as Article 13 of the Labor Law which permits an employer to dismiss a woman worker on grounds of pregnancy, constraLn women's particLpation specifically. Currently a comission within the Cabinet is revlewlng draft legislation pertainLng to job security that will bring Turkish legLilatLon in compliance with IW ConventLon 158; enactment of the legislation is anticipated in the near future. Also, some of the protective regulatlons contribute to constraining women to traditional, and generally, low-paying occupatlons. in addltion to a re- evaluation of discriminatory protectlve legislation, women would benefit from the extension of legislation concerning working conditions to all workers, particularly those in the informal and agricultural sectors, and to the Lntroduction of legislation for part-time and flexltime employment. Enforcement of existing beneficial civil and labor legislation, also warrant further conslderation as lndicated, for instance, by the small number of -4- childcare centers Ln enterprLes and the low proportlon of clvll marrLageo regletered ln rural areas. D. Fertility. Nutrition and Health Status 1.12 The background health study (Annex 2) prepared for thle report concludes that health Lndicatore of women and thelr chlldren have lmproved signlflcantly slnce 1960 but that women are sllghtly dLiadvantaged compared to men ln Turkey in terms of health and nutrltlon, primarily as a result of thelr reproductlve role. Specifically, women ln Turkey constitute approxlmately 49% of the populatlon compared to an average of 52% in the LndustrLalized countries and 48% ln the low-lncome developLng countrles. Nonetheless, the life expectancy dlfferentlal between m*n and women ln Turkey reveals a rapld gain in female life expectancy that suggests the dliadvantage may be short- lived. Recent data lndlcate a female life expectancy of 69 years V agaLnst 64 years for men (the gender dlfferentlal ln the Lndustrialized countrLes li also 5 years) compared to 59 years for females and 58 years for males ln the early 1950S. Of partlcular lgnifLicance ln Turkey le the dlfferentlal between urban and rural sex ratios, reflecting the continued urban mlgratlon sLnce the 1950e. During the 1980's, thLs mlgratLon accelerated wlth the result that populatlon growth ln rural areas became slgnlflcantly negatLve for the flrst tlme. Mortality Rales 1.13 Mortallty rates in Turkey have declined slgnlflcantly ln the past two decades, however, reglonal dlfferences remaln hlghly slgnlflcant. Gender- specLfLc dlfferentlals in Lnfant/chlld mortallty rates are not publlshed ln Turkey. However, recent studLes (see Annex 2) indlcate a sex preference for sons, such that parents glve more attention to nutrltion and health care for boys due, inter alla, to thelr contributLon to family income and thelr importance as old age security. Although dlfflcult to substantlate, Annex 2 argues that this preference affects both gender mortality and morbldlty rates in Turkey. 1.14 Maternal mortality rates in Turkey are not readily available. Eetimates based on hospital births are available but are Lnadequate, elnce approximately 60% of dellverles occur in hospLtals. In 1989, a State Instltute of Statltlces (SIS) survey showed a maternal mortality rate of 132 per 100,000 blrthe for 1981. This compares to a range of maternal mortallty rates of 5-15 per 100,000 blrths in the Lndustrialized countrles. The mortallty rates, however, are subject to major regional varlatlon. For Lnstance, community-based surveys conducted by the Mlnletry of Health (MOR) in 1986 ln parts of eastern Turkey show a maternal mortality rate as hlgh as 284. 1.15 Recent studLes in Turkey lndlcate that the major factors contrLbutlng to high maternal mortallty rates are frequent and young age of pregnancLes. The average age of marrlage for women ln Turkey remains low - 18.2 years in 1988, slightly hlgher than ln 1983 when the average age was 17.6 years. State Planning Organization (SPO) data show that ln 1985 approxlmately V World Bank, WDR (1991). 41% of females aged 1eee than 19 wer, marrLed and 84% by age 24. The age dlifferentlal between husbands and wLves, although also narrowing ln recent yearse, l on average 6 years for women aged 20 years or less. puttlng men at a conelderable advantage ln famLly decisLon-makLng. Other factors contrLbuting to high maternal mortallty rates lnolude malnutritLon (although ln Turkey, wlth an average calorlc lntake of about 3,200 calorles, this is largely a matter of educatLon), the low proportlon of women recelving any pre-natal care (43%) and poorly conducted abortlons. In rural areas, in partlcular, communlties still lack awareness that pregnant women are ln need of speclal care and that they can be at risk. The 1988 Haseteppe University Institute of Populatlon Studies (HIPS, 1989) survey shows that ln rural areas 53% of births occur at home and 35% of births are without health personnel. Also, obstetric technLques In these areas tend to be poor and access to lifesaving health facillties, includlng blood, are limited. FertLiLty 1.16 Fertillty has declined slgnlficantly ln recent years. The Total Fertlilty Rate (TFR), which is the average number of chlldren that would be born alive to women livlng to the end of their chlldbearlng years, has fallen from 6.8 ln 1960 to 4.0 accordlng to the 1988 HIPS Survey. Nonetheless, thLe is quLte high when compared to the comparator middle income countrles ln the Indleators of Women's Development Table (p. lv) .I The TFR for Turkey, however, shows sLgnLfLeant regLonal variation. Speclfically, the urban TFR rose from 3.2 ln 1983 to 3.8 ln 1988 as a result of migration, whlle the rural TFR fell from 5.1 to 4.4 (HIPS, 1989). Across regions, the variation is even greater (Annex 2, Table 10). The causes of Turkey's contlnulng hlgh fertllity in the rural areas lnclude, lnter alia, the percelved economic and soclal utllity of children, the strong preference for sons and the continulng prevalence of cultural and tradltlonal belLefs and values. 1. 17 Government's target in the Sixth FLve-Year Plan is for a further reduction in the TFR to 3.46 for 1990-94, and is likely to be supported by the seemingly declining fertLilty aspiratlons in Turkey. Speclfically, the HIPS surveys indlcate a decllne ln the number of "desired" chlldren from just over 3 in 1978 to just over 2 Ln 1988, wlth the mean number of children desired not varying signlficantly by elther region or urban/rural resLdence. In the 1988 survey, 44% of all ever-marrled women responded that they had more chlldren than thelr ldeal or desired number. The mean number of children deslred was found however to Lncrease wlth the number of lLvLng children and the age of the woman and to fall as the educational status of the woman or husband increases. 1.18 To realize the fertLilty target, Government is endeavoring to lncrease the acceptance of modern contraceptlve methods. Following introduction of the flrst Family Plannlng Law in 1965, the percent of exposed women uslng contraceptives increased rapidly from 22% in 1963 to 77% in 1988. More recent, but unofficlal data show a decllne to 3.7 ln 1989 (Population Reference Bureau, 1989). ,V.Chlle 2.6S Colombia 2.9; MexLco 3.4; and Portugal 1.6; (wDR, 1991). -6- This la one of the highest reported usage ratoe in the World (WDR, 1991). Nonetheless, a population growth rate of 2.4% suggents a margin for improvement. Speclflcally, while the use of modern contraceptive methods doubled between 1978 and 1988 to 38% of exposed women, the use of traditional methods also continued to Licrease and still account for 39% of exposed women (with withdrawal alone Ln 1988 accounting for 31%). Realization of the fertility target could be aided by strengthening family planning service and, ln partloular, through a focus on modern methods, the use of languages that would be understood by local lnhabitants and increased follow-up and counselling. Studles in Turkey confirm, as in other countries, that increases ln ohlld survival rates dLrectly 4ontribute to increased acceptance and continued use of contraceptives. Sven more important is the impact of education ln reducing fertility and increasing contraceptive use. In addition, studles outslde of Turkey indicate that gainful employment opportunities for women show a positive correlation with declines in fertility and that women who work outside the home practice contraception more and have lower fertility than those who work at home. In short, family planning helps women to develop or selze "modern" options - the rationale is not just one of health or demographics. 1.19 While there are few studies on nutrition in Turkey, average caloric intake is sufficLently high (3,200 calories) to suggest that nutrition is a serious health concern only where incomes are low or unstable, or there is lnadequato access to social servlces. Available data, however, provide no basis to determine calorlc intake by age and sex and thus there is no basis on whlch to assume preferential intra-household allocation of food. To the contrary, studies suggest that obesity is a more widespread problem amongst women than men ln Turkey (one recent study by T8nUk in 1987 concluded that 32% of women and 13% of men are obese). 1.20 Limited regional studles (Annex 2) suggest that anemla prevalence in adult women exceeds 40% of the female population in large areas of Turkey and that it is highest in the peak child-bearing (25-34) age group. studies also show goiter prevalence to be high for women throughout Turkey and that vitamin and mineral deficiency are much more frequent in women than men and particularly for pregnant and lactating women. Iron folate anemia can be effectively treated at low cost with iron and folic acld tablets or iron fortification of salt and other processed foods. Iodine fortification of salt or supplementation with iodized oil is also a low cost and effective treatment for goiter. Distribution of these treatments should be considered through the health care system. 1.21 BreastfedndLg of babies in Turkey is almost universal. Both the 1983 and 1988 HIPS surveys show that the percentage of mothers breastfeeding the last child averaged 95% with little varlation by region or rural/urban residence, age of mother or education. While the prevalence of breastfeeding is a positive aspect of nutritlon, mothers in Turkey are not well informed as to an appropriate dlet for infants and often nutrient-poor foods are chosen and prepared under unhygienic conditlons. -7- 1.22 Notwithstanding the high caloric lntake, the importance of women's nutritlonal well-being needs emphasis in Turkey. In addition to the dlitribution of iron and iodine supplements, communlty education progracs can play a crucial role ln promoting women's good nutritlon as well as that of infants and children. Adult Health 1.23 Followlng a decade of adjustment, during which the share of the Government budget for health expenditures fell from just over 3% of GNP to just over 2%, less is known about the prevalence of dLaease and illness in Turkey for the population as a whole than would be expected for a country at thls stage of development. Available sbudLes (Annex 2) suggest a marked difference in gender health problems; for instance, men's health li significantly more likely to be affected by addictlves (tobaoco, alcohol) than women's. For women, however, little is known about the prevalence of gynecological problems and diseases. In light of the fact that women's health problems affect not only their own well-being but that of their families, critical health needs of women in Turkey Lnclude increased availabLilty of health facilities at the local level, dissemination of health information covering hygiene, nutrition (easpecially for Lnfants and pregnant women) and food preparation. Health Cga System 1.24 With the Lncreased priorlty accorded to the health sector in the current, Sixth Five-Year Plan (1990-94), quantltative health sector targets are specLfied for the first tlme. In addition to a reductlon in the TPR to 3.46, the most osgnificant targets are an increase in life expectancy to 68 years (already realized for women) and a reduction in the infant mortalLty rate from 61 per 1,000 births to 50 by 1994. Amongst government agencies involved in the provision of health serviees, the Xinistry of Health accounts for 51% of serviccs; the Social Security Institution (55I), 17%; the universitLes, 13%; and the Ministry of Defense, 12% wlth the balance provided by the SEEs, munlcipalities and the private sector, whose role has been encouraged slnce the Fifth Five-Year Plan (1985-89). There li little coordinatlon of the various institutions involved and the Ministry of Health has only a guiding role. Zach agency has its own standards and systems. From the standpoLnt of designing programs for the sector and addresoLng constraints, the current system has the disadvantage, inter alia, that there is little comparabLilty in the recording of health data. In fact, statistics on morbidity and mortalLty are collected only in hospitals and data collectlon in primary health care facilities io not systomatic. 1.25 Limited data on gender utLilzation of health facLilties suggest that 61% of all in-patients are women, but that just over 30% of all ln- patients were treated for pregnancy complications, deliveries or abortions. Recent studies (for instance, Ozeebe, 1990) suggest that women of reproductive age use primary health services more than men in the semi-urban areas, but that men use rural health facLilties more and account for a much higher proportion of visits to private physicians and dentists. There remains a strong reluctance of women ln many regions to visit a male doctor and also of male doctors to treat females. The studies also suggest a positlve correlatlon between educatlonal level and wlliLgness of women to meek professLonal medlcal care. 2.26 Llttle lnformatLon La available as to the dLstrLbutLon of health personnel by gender. Although women account for 53% of mployment in the public health sector (Cltcl, 1988). women account for only 26% of phyaLcLans and tend to be concentrated ln the paramedLcal ranks. 1027 GLven both the protjcted population growth and the priority currently accorded to the health sector, not only do sector strategy, planning and coordinatLon of resources rquire greater emphasLs in general, but due attentlon to gender issues is crucial ln all aspects of the health sector. As a fLrat step, the establihment of a data base whlch permlts analyaLs of health issues by gender is essentLal for saetoral plannLng and developments simultaneously, the Lntroduction of unlform standards and recording systems for the varLous agencies Lnvolved in the provision of health servLces is needed. Avallable studLes on the utliLmatLon of health facliltles suggest that further consLderatLon to the employment of women at all levels ln the sector is needed and especially ln the plannaig and administrative processes. The health servlces need also to consider more closely the soclal environmant in which servlies are provLded. for instance, consideration should be glven to establishLng gender-balanced health teams (physicians, technLcLans, nurses and support staff) that can provlde servlces to both sexes and to the employment of socLal workers/socLologLits who can help health workers understand the socio/oultural character of their communlty. The communication barrler between health provlder and reclplent is a cause for concern. The development of system to facilitate comiunlcatlon wlth the local population ia central to addressing women's health Lssues ln the rural areas. To thLs end, consideration should be given to the use of languages that would be understood by local Lnhabitants for promotional campaigns, family planning programs and/or the traLLng and aseLgnment of health personnel. 1.28 A woman's health and status are lntricately entwined. ImprovLng the health of women requlres addressing the oclal customs and cultural traditlons of the reglon. Programs to $mprove the health of women need to proceed ln the context of improving the overall status of women in the family and communlty and, ln partlcular, of educating both men and women as to the beneflts of addressing women's health concerns. 1.29 The slgnlfLcant reglonal varlatlon in women's health and fertlilty LndLcators suggests the followLng lnterventLons could further women's particlpatlons o strengthening rA lonal1v the dellvery of health and famLly planning servlces. FamLly plannag servlces, ln partlcular, could be strengthened through a focus on modern methods, the use of languages that would be understood by local Lahabltants and Lacreased follow-up and counselling. * esmultaneously, developLng system to facilitate both the provLsLon of health servlces in dlfferent reglonal envlronments and coumunlcatlon wlth the local populatlon. -9- * promotion regionally of comunlty education program. for aen and women addressing the Importance of famLly planning, maternal care, nutrltlon, health and hyglene Lssuse * strengthen programs to reduce maternal mortallty rates Lncludings strengthenLng the training of mldwlves, partLcularly ln the rural areas, to improve dellvery practLce. and skills to detect conditLons leading to complLcatLonsg lmpleentLng system of referral (including access to transport to reach dlstrlct hospLtals)g development of a health sector data bas (includLng personnel), with gender-specLfic Information, to strengthen the health care system and, in partlcular, to underpln sector planning and researchi * undertaking research, wlth due conaLderatlon to socLoeconomic and cultural factors of, lnter alla, maternal mortallty rates, fertlilty patterns, utllixatlon of exLitLig health care systems and the continued use of traditional practlces with a vlew to strengthenLng the family health and family planning progrmsi and * the dlstrlbutLon of iron tablets and lodlne supplements through the health care system ard the ummunLsatLon of prognant women for tetanurt - 10 - CHAPTIR SI WONCN IN THB LABOR VORCH A. oug!Igtalton 2.01 The vltal role of women ln Turkey's economlc development is well recognLsed by Government. However, to date, there has been little conslderatlon by pollcy makers as to how to reallse won's full LntegratLon lnto the economy. The enhancement of women's capablIltile to contrlbuto to a country's development plays a erLtLcal role ln sLmultaneously LncreasLng overall economLc efficLency nd Improaing the relatlve economlc status of women - there is no "trade-off". 2.02 While both women's market and nonmarket actlvities are indLipensable components in the development process, thLs chapter focuses on women's labor market actlvltles and participation sLace 1955 with a vLew to formulatlng policy lnterventions that will not only increase womens contrLbutLons to economlc development, but enhance their pooltlon wLthin the famlly and soclety, taking into account changes that the economy wlll likely undergo ln the course of the current decade. In the case of Turkey there has been conslderable debate as to what has happened over time to women's labor force participatLon. This chapter, therefore, addresses in detall the dlffering statistical sources and the conceptual losues assesoLng labor market actlvltles wlth a vlew to resolving the controversy. By reason of data avallablilty, however, the focus here ls expository and descrlptlve. 2.03 Although labor market actlvltles do not fully capture women's contributLons to development, a notable feature of econoiLc growth Ln almost all countries and ln the posltlon of women ln soclety and the household, li the shlft of women's work actlvltles from household-based, and generally unpald, employment (housework, the family farm, or homebased production) to pald employment ln a market economy. The economlc hlotory of lndustrLalLsed countrles shows a strong associatlon between women's partLclpatLon ln the market economy and economic growth. Moreover, as sucoLnotly put by Nleva (1985, p.164), "the Lndependent flnanclal base provLded by employment provldes women wlth an increased sense of competence, glves women more power wLthin marriage, and lncreases her lnfluence in decLiLon making . . . Couples Ln whlch both husband and wlfe work are more llkely to share decLsLons about major purchases and child rearlng.0 2.04 At the outset, lt li useful to compare Turkey's experlence as regards women in the labor market wlth that of the OECD countrLes where, for the most part, there has been a strong assoclation between the expansion of the female labor force and economlc growth.i/ In the OUCD countrles, the growth ln women's labor force partLclpation is a falrly recent phenomenon, mostly slnce 1960. 1V oldLm (1991, 1989, 1986, 1983) performed a thorough analysls of the relationship between economic growth ln the U.S. and the shift of women's work activities from the home to the labor market. One of the maln conclusLons is that the emergence of a visLble female work force played a slgnlfLcant role in raLiLng lncome per caplta. Between 1890 and 1980, annual national per capita income would have been lower by at least 14% had the female work force not expanded. - 11 - Before World War SI, female participation rates in the OECD countries were relatlvely low (15-30%), much lower than those of men (around 85%). The most notable feature of female labor force patterns in the first half of the century, in fact, was how little they changed. Following World War SS, most OBCD countries experienced dramatic change. in the position of women in the labor market - and unprecedented increases in living standards, in part, as a result of the labor force entry of married women. 2.05 In most OECD countrLes, the participation rate of women now exceeds 50% and continues to rise steadily, while the participation trend of men (which is declining, largely becauce of longer years of schooling and earlier retirement) is close to intersecting that of women. A constellation of well- known aocial and economic forces is strongly associated with the propensity of women to enter the work force. These includes sharply declining fertility, lncreased divorces, later marriages, increased educational attainment reflected in earnings capacity, improvements in the technology of household work, legislative changes, the feminist movement, increased availabllity of part-time work, urbanization, and changes in the structure of the economy reflected in the rapid growth of the industrial and service sectors and the decline in the lmportance of agriculture. In essence, the significant rise in women's participation rates in the OECD countries is essentlally a post-war phenomenon, mainly concentrated during the last two or three decades, and, as will be discussed below, even later in the southern European OECD countries. 2.06 In contrast, the labor force participation of women in Turkey has declined sharply since the mLd-1950s - from almost 70% to 33% in 1990. A particLpation rate that is not only much lower than that of Turklsh men (almost 75% in 1990) but one of the lowest in the OECD. As in most developing countries and at an earlier stage in the industrialized countries, the types of labor force activities that Turkish women pursue differ considerably from that of men: women are les likely than men to work for wages, are employed mostly as agricultural workers, and have lower educational attainments, especially at the post-primary level. Nonetheless, and as will be demonstrated here, the Turkish patterns are similar to the pre-1960s trends in several industrialized OECD countries, and bear a very strong resemblance to the more recent experience of OECD countries in southern Eurqne. 2.07 The economic efficiency implications (societal rates of return) of gender disparities in a structural adjustment context are straightforward. Essentially, structural adjustment is concerned with the efficient mobLilzation and inter-sectoral mobility of a country's non-human and human resources. This suggests that the less than full involvement of women in the process reflects a misallocatLon of human resources. The resulting restrLition of output results in an irretrievable deadweight loss to society due to allocative inefficiency, and io likely to act as a strong impediment to the fulfilment of the overall long-term objectives of the structural adjustment strategy. Thus, a review and assessment of the role of women can make an lmportant contribution to both the development process and the status of women. At a minimum, this requlres an awareness and understanding of the trends in women's labor force behavior in the Turkish economy. The asses_ment of these trends and patterns will provide useful information in designing overall labor market initiatives, especially appropriate female-oriented policies and programs. - 12 - 2.08 Insofar as data permit, this Chapter analyzes for the past three decades: - trends in women's labor force participatLon; - patterns of gender segregation in the work place - by employment status, occupation and economic sector; - hours of work among men and women; - gender differences in the skill or educatlonal content of the labor force; - male-female differentials in earnings and social security coverage. The analysLs makes liberal use of comparisons with men and wlth other countries. The former provides a clearer indicator of "gender neutrality" and indicates areas for improvement, while the latter comparison of Turkey with several Buropean OECD countries, whose development experiences and structural features are fairly similar, provides an important perspective on future developments in Turkey. The countries selected for the comparison are Greece, Italy, Portugal, and SpaLn.V 2.09 The outline of this Chapter is as follows: after a discussion and assessment of the available data in Section B, Section C documents nationwide secular trends in labor force partlcipation. Gender segregation in the labor market is examined in Section D, while Section E considers urban-rural dLfferences in labor force behavior. Section F examines another dimension of labor supply - hours of work. Section G discusses gender differences in the educational attainment of the labor force. Gender differences in earnings and social insurance coverage are addressed in section H. Based on the main findings, the concluding section proposes proactlve policy responses that would further women's integration in the economy. 2.10 The main concluslons of the chapter are as follows: o The sharp decline in female labor force participation in Turkey - from almost 70% in 1955 to 33% in 1990 - is fully consistent with the wldely-documented U-shaped relationshlp between economic development and female participation observed in the OECD countries since World War II, and where female labor force participation now exceeds 504 (paras. 2.13-19). A comparison of the participation pattern in Turkey with that of four southern European OECD countries suggests that Turkey is likely to experience a secular upturn in female participation. Policy initlatives focusslng on women's participation in non-traditional activitles will be needed to support this upturn. 21 Greece joLned the ENC in 1981, Spain and Portugal in 1986, and Italy was a foundLng member. These countries were chosen, rather than other middle lncome countries ouch as Chlle, Colombla, Hungary, Poland, or Venezuela because Turkey le in the same geographical vicinity, a member of the OECD and a candldate for the EC. - 13 - The work force in Turkey Le markedly aeareaated by aender, especially in urban areas, where people tend to have higher educational attainments and wider employment opportunities. To a large extent, Turkish women work in different occupations and industrles than men, and have a different worker status. In 1990, approximately 74% of women employed in the labor force worked in agriculture, 7% in industry and 12% in aervices, while men are more evenly distributed across sectors - 34% in agriculture, 24% in industry and 38% in services. As to employment status, 68% of women in the labor force were unpaid family workers and 25% were wage earners or self-employed, compared to 14% of men who were unpaid and 82% who were wage earners or self-employed. Analysis of data by occupation show women concentrated in the lower occupational categories and performing work closely related to their traditional household duties. Whie the gender segregation is similar to that observed in OECD countries, it is more intense, and as the experience of the OBCD countries has shown, will be difficult to alleviate without policy intervention focussing on, inter alia, education and training, the introduction of employment equity programs and repeal of discriminatory protective legislation (paras. 2.22-32). o And third, in addition to the potential increa8e in participation, there are maior 22ILcv concerns re2g&rdina the wstock" ofwmni the labor force. Analysis of the gender segregation indices indicates that, the impact of the increasingly educated female labor force has not been reflected in the labor markets, although this will be largely a matter of time (paras. 2.44-52). For instance in 1990, 50% of the urban female labor force but only 5% oe the rural female labor force had more than a primary school education. Currently the educational level of the rural women will constrain their labor force participation, regardless of the availability of employment opportunitiesY Based on current t Although hard evidence is not available, it appears that following migration to urban areas, these women resume their customary role as unpaid family workers, but in a different setting - homebased production. Unfortunately, very little is known about the work activities and well-being of Turkish women who migrate from the rural areas. According to Abadan-Unat (1986) the work situation of "gecekondu" women differs from those of their counterparts in Latin American "favelas" or "pueblos jovenes". While Latin American women are visibly engaged in the vibrant informal retail trade sectors (vendLng, street selling, and peddling),. Turkish women are conspicuous by their absence in these activities, as well as domestic work ln private homes. Moreover, fragmentary evidence suggests wgecekondu" women, who are often resettled from small villages and farms, respond to the difficult transitions to urban life by withdrawing from the urban labor market. This phenomenon, no doubt, contributes to the low labor force participation rates that are observed in urban areas, especLally among women with low educational attainments. SLmilarly, little is known about the "returning migrant" women from Europe, who may well serve as a catalyst in the villages and towns. Research into these issues, including, in particular, the services and informal sectors, is essential. - 14 - enrolLment rates, the proportion of rural women wlth more than a primary school education wlll be significantly higher, a decade from now, as will be their employment expectations. This underline the need for pollcy measures designed to promote the mployment of women ln hlgher productivity jobs in both urban and rural areas. 8peclfically, in 1990, rural fmale participation was 51% compared to 15% for urban females. Thus, and while some 60% of Turkey's populatLon lives ln urban areas, 77% of the female labor force llve in rural areas (paras. 2.33-34). Also, as. education and marital atterns differ markedly by urban-rural areas. For instance, ln rural areas, female participation shows a mild N- shape, with participation remaining hlgh until age 60, whereas the urban profile peaks at age 20-24 and declines steadlly thereafter. The U-shaped female profile was found ln most OECD countries during the 1960's and 1970's but has subsequently been replaced by the male bell-shape, which is observed in both rural and urban areas for Turkish males. These differences are extremely significant for the design of policy measures, not only for wcoen's employment but for regional development, in the context of contlnuing urbanization and industrialization. Closely related to the issue of urban-rural dLfferences is that of lattrre aonal disarities. Turkey, the fourth largest country In the ONCD, has an uneven spatial distributlon of its population and economic activltles. These issues have been discussed in detaLl in two recent reports by the OECD (1988) and UNICEF (1991). The main conclusions reached by these studies are that the provinces in the eastern region of Turkey are seriously disadvantaged in term of public services, (e.g. education, health, transportation, water), actual and potential industrial production, as well as agricultural output. These disparities give rise to wide regional dLfferences in educational attainment, fortility, mortality, employment opportunities and per caplta lncome, underscoring the need for policy initiatives designed to reduce inter-regional disparities to also address gender issues. Additional findings are that: ° The number of hours worked by -emloved women, which is in addltion to household chores and childoare responsibilities, is high in Turkey. Employed women on average work 43 hours per week compared to 50 hours for men. The large amount of time devoted to market work partly explains their low participation rates, especially in urban areas. Inevitably, the competing demands of household responsibilitLes and market work may slmply discourage women from entering the labor force. - 15 - o b.Aa&&a are avallable only for the small proportlon (6%) of the female labor force covered by social Lnsurance. A meanlngful analysLi of gender wage differentLals La therefore not posslble. Releare of the LYS wage data would be a useful first step to romedy thli defLeLancy. o Th data on the educational attainment of the labor force suggest that the gWor gan in education has narrowed aLificantlv durLna the goat 20 year. The followlng considerations, however, should be born ln mLnd. First, prLor to the late 1980. the avallable education information is only for labor force participants, and thus excludes a hlgh proportion of women who are nonparticlpants, especLally in urban area.. Second, there are pronounced dlfferences between urban and rural areas as regards women's educational attainments. Zn rural areas, where 5 of the female labor force has more than primary schoollng, education has currently little lmpact on particlpatlon slnce most women are unpald famlly workers. By contrast, ln urban areas where 50% of the labor force has more than primary schoollng, educatLon has a strong Lmpact on partieLpatLon. Thlrd, ln 1990 women accounted for almost two-thlrds of the total labor force that is illiterate. S. Labor Fore Datas S o - Its and Tusu e 2 11 StatistLis on women'e -labor force behavior in developing countries, in general, are plagued by conceptual problem and inadequate data collection methods. It is also LncreasLagly recognlsed that conventional statistics tend to undercount the number of workLag women and their contributions to economic actlvity because of problem lnherent in the definition of weconomlcally active" populatLon, the designatlon of "meaningful economlc activity", and the implementation and tlming of surveys or censuses. These shortcomings should be borne ln mLad when consLderLng the patterns presented in section C below ./ In Turkey lt appears that untll recently, the development of a reliable labor market information system was a low priorlty when compared to data needs concerning populatlon characterlitlcs, LnternatLonal trade, national income and other macro indicators. The hlitorlcal labor market data series are incomplete and unreliable: a careful examLiatLon of labor market and related data published by the ILO, the OECD and the Government of Turkey, revealed serious omissions and W mouldLng (1983), Mueller (1983), Suss (1989) and Standing (1978) provide excellent revLws of the limitations of official data collection systems in approprlately reporting women's contributLons to the economy. A main source of the dLffLiultLes li that the data collectLon methodologLes are often incompatible wlth women's work roles ln the household and ln the market. A particularly serious issue li the unrellable assessomnt of womens actlvitLes in the winformall sector and as unpaLd famLly workers, especially in agriculture. UxLiting Lnastruments of data collection reveal llttle about the time which women devote to economlc activities vli-&-viL their household responsiblltLes. Even leos is known about theLr monetary worth. See also ON (1991) for a summary of estimates wlth housework Lncluded in the Natlonal Accounts for a number of lndustrialLsed and developing countrles. - 16 - inconsiutencies through the late 1980s, when the nationwldo Household Labor Force Surveys were first implemented. In addltion, the hlitorical series provLde only partlal labor supply measures. SpecLflcally they do not indicate hours of work, or even the extent of full- and part-time work among the employed. The amount of tlme devoted to work in an important dimensLon of labor supply behavior and partlcularly of women's work. 2.12 The analysis here Lo based primarlly on two sources of data from the State InstLtuto of StatLitics (SIS): census data for the period 1955-1965S and .for 1988-1990, the published compllations of the Household Labor Foroe Surveys (LFS) ./ The latter provide more comprehensive and consiLtent coverage of various aspects of mon's and women's work activities, including hours of work, than does the census data. It should be noted at the outset that there are signifipant dLicrepancLes between all of the data source*.g These dLeorepancLes are partialliy attributable to dLfferences in statistical definitions, measurement of the labor force and the populatlon count (Annex 3). DefinLng the particLpation rate as the labor force (12 years and over, the age of eligibility to work in Turkey sLnce 1970), the 1985 participation rates summarLzed below, serve to illustrate the problem. DATA FOR 1985 FRON DIFFERENT SOtRCES CENSUS ILO LFS OECD Population ('000) 50,664 50.664 50,959 50,664 Labor Force ('000) 21,580 21.580 18.203 17.894 Participation Rate 42.61 42.6% 35.7X 35.31 1m Population ('000) 25,672 25.672 25.685 25.782 Labor Force ('000) 13.933 13.933 12.712 11.550 Participation Rate 54.3X 54.3X 49.51 4.6X Population ('000) 24.992 24,992 25.274 24.862 Labor Force ('@00) 7,647 7.647 5.491 6.344 Participation Rate 30.6X 30.6X 21.71 25.5S 2/ In addition to the State Institute of Statistics, agences provlding labor force related data in Turkey are the State Planning OrganLzatLon (SPQ), the Turkish Umployment Organization (ZXBK) and the Social Insurance InstLtution (8U). The historical Turkish data are reproduced with appropriate corrections and adjustments ln varlous publications of the ILO and the OECD' For reasons illustrated above, the analysis here relies more heavily on these sources for the period prior to 1988 than on the orlginal TurkLih Census data. Data contalned ln Durand (1975) are also used for 1955 and 1965. Al The Census and the IL data agree, however, for all years, except 1965. The ILO reports a labor force of 12,993,245 (7,697,483 men and 5,295,762 women), while the 1965 census shows a labor force of 13,557,680 (8,420,829 men and 5,137,031 women). The population counts are the saw in the two sources - 15,996,964 men and 15,394,457 womn. If one uses the Census cata, the 1965 female partLcipation rate li 37.8% compared to 39.0% from 1W0 data. Attempts to reconclle the two sets of numbers proved futile. - 17 - C. Patterns In Labor Fote PartLainatUo 2.13 This sectlon reviews the secular patterns of labor force partioipation ln Turkey mince 1955. As noted in Section A, theme patterns appear to follow those observed in both OBCD and tho more LndustrLalized developlng countrLes. Namely, that ln the process of development female partLcipation rates Laitially decrease and then Lncrease wlth industrialization and that education plays a plvotal role in LncreasLng women's participatLon ratesXZ' AWi&tn.part La n Trends 2 14 Table 201 and Chart 2 1 show the labor force magnltudes and labor force partlclpation rates for 1955-199C06 as reported ln the Censuses (1955- 1985) and the Labor Force Surveys (1988-1990). The overall participatLon rates of both men and women in Turkey have decllned durLag the past three decades, but much more so for women - from 69% to 33%. Thli rate is Ln line with developlng country female participatlon rates but well-below the OCD rates of 50% or more. The downward trend in the female rate is to be espected in a country that is undergoing a transformation from a primarily agricultural to an lndustrial urban economy. Somo of the decllne ln participation rates may be attributed to social change, not least of whlch is rising school attendance and educatlonal opportunitLes.V 8ectoral changes in the economy, however, are also a significant factor and, ln particular, the agricultural sector whlch has been the traditional source of employment for Turklsh women, albeit as "unpaid" family workers. While agriculture remains important to the Turkish economy, its importance in CDP and to employment (especially of men) has declined signlflcantly durLng the past 30 years. The steady doellne of agriculture does not necessarily imply that suLtable employment opportunities open up for women in the expandiLng lndustrial and services sectors, located mostly ln urban areas - at least not 4mmediately and not without an appropriate incentive structure. Zt These are the conclusions of Psaeharopoulos and Tzannatoe (1989), using data from 13S countries in the early 1980e. They also provide a thorough discussion of the conceptual and statistical limitations of the labor force particpation rate, especially of women. if Data from the 1950 census are excluded because, according to Durand (1975), the methodology and definitions used are substantially different from those of later censuses. Also, lt appears that there are errors in the 1965 Census, which, resulted in statlstical anomalies. See, for example, Tables 2.5 and 2.6. The table and chart present partLcipation rates normalized in different ways using an eligibility age of 12 years for 1970-1990 and one of 15 years for 1955-1965, following official Turkieh definltions. The table and chart in Annex 2 show labor force trends based on an eligLbility age of 15 years for the entire period. 2Y Illiteracy rates for women fell from 74% in 1955 to 32% in 1985 and to 30% in 1990u the corresponding values for men are 44S, 13%, and 10%. - 18 - TABLE 2.1 TRENDS IN LABOR FORCE PARTICIPATION BY GENDER (Eligibility Age is 12 Years, 1970490; 15 Years, 195565) MEN WOMEN LABOR LABOR FORCE AS % OF MALE POPULATION FORCE AS % OF FEMALE POPULATION YEAR 000 12+ 1504 TOTAL '000 12+ 1544 TOTAL 1955 6944 89.6 99.7 56.8 5262 68.9 77.3 44.5 1960 7697 87.0 98.2 54.3 5296 628 68.3 39.0 1965 8472 85.2 98.1 53.0 5141 54.8 61.5 33.4 1970 9306 79.5 96.3 51.7 5813 50.3 61.2 33.0 1975 11180 80.9 97.9 53.9 6204 47.3 57.7 31.7 1980 12284 79.8 96.3 54.1 6928 45.8 55.6 31.4 198S 13933 78.3 93.3 54.3 7647 43.6 524 30.6 19880 14083 76.5 91.0 52.3 6534 34.9 41.2 24.5 1989A 14175 76.0 90.6 54.1 6941 36.6 43.3 25.6 19890 14253 75.4 90.6 51.7 923 36.0 42.7 25.2 Io199 14165 73.9 88.5 50.7 6486 33.4 39.3 23.3 sz3m: t 151ig. Yubok od Lor 8*, ILO; 105.1i0,. A;u-eheI8 ? EeVsy. 5 . Nola. 0. Ootobsr A- Apo CHART 2.1 PARTICIPATION RATES BY GENDER LABOR FORCE / POPULATION 12 + 00 2a- 20. 19 1860 1966 1970 1975 1960 1S5 1 1m1 1 YEAR (A - April; 0 - October) - 19 - 2.15 The deolInLng trend in female labor force partlclpatlon in Turkey, li entlrely conalstent with the wldely documented U-shaped relationship between economlc development (LndustrLalLation) and female partLcipatLon rates that has been observed ln most of the OECD countrles and a growLng number of the more LndustrLalLsed developing countrles.io The key element ln the U-shaped pattern reflects structural changes ln the economy accompanLed, lnevltably by other soclal transformatlons (for Lnotance, lnternal mlgratlon, urbanLzation, lower fertillty rates, lese stable famlly and kLnshLp systoms and hlgher educatLonal attalnment). As the locus of economio actlvlty shifts from the famLly farm and homebased productlon lnto market-orlented Lndustrial and servLce activltLes, there are slgnlfLoant changes ln labor-demand requlrements, reflected in amployment opportunlties and ln their location - ln urban areas. DurLng the early phases of the process, there li often a "discouraged worker" effect and reported partlilpatlon rates fall, especLally among women, who were active as unpald workers Ln agrlculture and tradltlonal nonagrlcultural actlvltles. Moreover, women may also withdraw from the labor force Lnto areas of employment whlch are not well enumerated by the official statistics, maLnly seasonal work and homebased non-farm productlon. 2.16 At least two addltlonal factors contribute to the labor force wLthdrawal by women. First, lncreased mechanlzatLon ln agriculture reduces the employment of both men and women, but mostly women. Second, ln the growLng induotrial and servlces sectors, wcmen are at a disadvantage ln competition wlth men for jobs because they generally have lower, or different, educational attkiriments and labor force qualiflcatlons, are constrained by famlly, tradition and oulture, and compete wlth men who are more moblle and elther unemployed or underveployed. As the nonagrLiultural sectors expand, the declining trend iln male and female particlpation rate is arrested and then reversed. There ls an increase in demand (higher wages) flrst for male workers, followed by that for female workers, especially in the service sectors. The rlos ln wages increases the propensLty of women to enter the labor force whlch, in turn, is reflected in steadily rising participation rates. The labor force entry of women is further encouraged by declining fertlilty and maller household *Lses, Lncreased educational attainment, urbanizatlon, and changes ln socoetal attitudes towards women who perform paid market work. Also, the character and substance of government pollcies can play a plvotal role in lnfluencing women's labor force behavlor. 1W Durand (1975) was probably the flrst to hypotheslze the relationshLp between development and female particLpatLon ln a cross-natLonal context, and fLnd scme support for it. UsLng more sophisticated emplrlcal technlques and cross-sectlon lnternatlonal data, Pampel and Tanaka (1986) found strong support for the curvillnear relatLonship. Recent studles by KottLi (1990) on Creece and by Pre"ser and Kishon (1991) on Puerto Raco add more convlncLng evldence for the U- shaped form, as does the study by Psacharopolous and Tsannatos (1989) whlch uses data on 136 countries Ln the early 1980. sorrentlno (1983) also provides data supportlng the U-shaped pattern for Italy and Japan. - 20 - 2.17 Flnally, lt Ls noted that lt takes varyLng amounts of tlme for the U-shaped trend to materlalize. In some countries the pattern emerges after three or more decades while in others Lt in telescoped Lnto a comparatlvely short period of 20 years. SorrentLno (1983), ln a nlne-country comparLson for the perlod 1960-81, observed a strong and sustaLned Lncrease ln female partLcLpatLon rates ln AustralLa, Canada, Sweden, the US, and the UK sLnee 1960. PartLeLpatLon rates Ln aGrmany, France, Italy, and Japan decllned until the 1970. and then increased ln the early 19708 Ln France and Italy, and the late 1970. Ln Germany and Japan. Greece, Ireland, the Netherlands, Portugal, and Spaln appear to be "stage two" candldates for the 1980c. 2.18 The hLitorLcal behavlor of female particLpatLon rates Ln Turkey suggests lt li likely to become a "stage-two" candldate begLnnLng ln thle decade - partlcularly lf accompanled by approprLate female-orlented policy LnLtLatLves. To see how thli may develop, secular patterns Ln the southern European OECD comparator countrles - Greece, Italy, Portugal and Spain - were examined for the perlod 1950-1989. Whlile there are many dLfferences among the four countrLes, they also share a number of common characterLiticas in comparlson to other OECD countrLes, they are latecomors to Lndustrlallzation and to export-led development. This shlft in economlc structure - as well as enactments of SC dlrectives that promote gender equalLty - contrLbuted to the comparatively late growth in female partLcLpatLon, reversing a trend whlch was falling, or flat, untll the 1970. and 1980s. As expected, the rLiLng trends ln partLeLpatLon rates were accompanied by increased urbanLzation and educatlonal attaLnment and by declLnes in fertillty. W Table 2.2 and Chart 2.2 suggest that the pattern of TurkLsh female partLeLpation rates Li not markedly dlfferent from the recent hsetorLcal experience of the comparators.. 2.19 The available labor force survey data do not permlt an analysLs of female labor force partLcLpatLon ln relatlon to background famlly characteristics, such as marital status and the number of chlldron, whlle data from other sources provLde essentially no informatlon on the labor force behavior of marrled women. Although the Household Labor Force Surveys collect thls informatLon, the data are not made avallable. Nonetheless, it is useful to note that a hlgh proportion of the female populatlon aged 15 years and older li married. Between 1955 and 1965 just over 70% were marrled whereas between 1980 and 1985 just under 70% were marrled. AccordLng to the 1990 Household Labor Force Survey, 68% of the female populatlon li mA rLed, wlves comprlse 65% of the female labor force 15 years and older and the partLiLpatLon rate of marrled women li about 33%. There is, however, an lmportant dlfference between urban and rural areas, which can best be seen from the summary data on marrled women below. ly Larre and Torres (1991) present an analysis of the experlences of Greece, Portugal and Spaln as regards thelr catch-up and convergence possibliltles. - 21 - TABLE 29 FEMALE PARTICIPATION RATES IN TURKEY AND COMAPARTOR OECD COUNTRIES. 1950-1989. (FEMALE LABOR FORCE / FEMALE POPULATION) YEAR TURKEY GREECE ITALY PORTU SPAIN 190 44.5 21.7 16.8 11.8 1960 39.0 27.8 19.5 13.1 13.5 1970 3X0 20.2 19.6 19.0 13.3 190 31.4 19.3 26.7 29.0 16.5 1965 30.6 27.3 28.2 37.4 21.2 19o 2&0 287 30.2 38.6 25.6 1989 24.9 28.7 30.4 39.1 26.1 SWoueS: 1950-1985. YeaWbook Of LaOr 30aIuIc, ILO; 19861989, HIstforIa StaistIc 1960-969 OECD Notes: The 1950 value for Tukey refr to 1955. CHART 2.2 FEMALE PARTICIPATION RATES BY COUNTRY 4 r . TURKEY E 0 > , ~~~~~~~GREECE z r0rFWEIALY 0 6 / ,,,.,. -PU A R......O UGAL 0 O SPANSPAIN 10- .... 190 1960 15li 19t0 15 1 1 YEAR - 22 - R"udune of FLF 77 23 100 Wtm In FLF 70 50 65 Participation Rate of Narrted Vmi 53 12 33 Whll- the data on marital status of women ln the labor force are imcomplete, they nonetheless lndlcate that the trends examined in thli chapter relate largely to married women, but also that there are slgalficant urban-rural differences ln the particlpailon of marrled women that need to be addressed in terms of policy lnterventlons, and speolfically relating to visible urban employment, flexible formal employment and chlldcare. Partia Laton Patten byheStutr 2.20 A fuller pLiture of trends ln male and female participation rates is gained by disaggregatlng by age groups Table 2 3 and Chart 2 3 dLiplay the cross-seation age profle s of particLpatLon of men and women between 1955-1990. Reflecting a combLnatlon of both life-cycle effects and time trends, both the male and female profiles have shlfted downward over the past three decades, and have been transformed, markedly so for women. For the 1980-90 period, the male pattern assumed a bell or Laverted-U shape, while the female pattern assumed a (mild) twLn-peak or K-shape. The age profiles of women in recent years show that participatlon rates rlse steadlly among teenagers and young women (20-24) as they leave school and enter the labor force, decline during the 25-34 age lnterval because of marriage and childbearLng rLse between 35 and 50 as women are less burdened by young children. After age 50 the rate falls steadily. The male profiles tell a different story. The particlpatlon rate rises steadily through the twenties, as boys complete their education, peaks in the 35-39 age group and as with females, decline sharply after age 50. 2.21 The observed effects of age on labor force participation rates ln Turkey since 1955 are consistent with patterns found in most OECD countries at a similar stage of development. Speclflcally, the profiles observed in the 1980. resmbeb c hose of many of the OCD countrles during the 19605 and 1970s. However, more recently ln the 0CD countrLes, the twin-peaked female profiles have been replaced by patterns resemliLng the bell shape profile of men.W& s See OZCD (1989, 1985, 197S), *nd SorrentLno (1983). TABLE 23 PARlTlP^ATION RATES BY GENDER AND AGE ROP FEMAES MALES AGE 1905 low 1990 19ur 0u urns us AG_E I= urn 197 u9nD 1n5 u990 12.14 Q 0. 4 42.3 3L 21.0 12t14 0.0 QD A9 424 3.1 253 15t19 14.4 65.2 54.4 51A 4" 40. 15.10 63. 73.0 60 S4 6.1 a. 2024 734 S Ul 4. 48 40 am24 6L. 940 65* 9A 9002 02 25S2 72i5 a" 44.7 434 Su 3 07 PA 0. 6.4 65" USI? 3UN 7* 62 M19 44. 42* 352 LS 63* 96* 07.7 6.4 6D 33 71.0 63 5 472 4 37.7 r 9" 64 9N0 W07A 2 07 4044 73.5 I S 57 48.4 4.2 338 44 9" IN4 95.0 " It?7 o 414 hA.4 C45 53. 55.5 40 3" 5U 3* 940 a" 37.1 .l m4 73. 632 510 4 42* am5A7 07 91.9 3a"0 6OL0 790 am19 0.? 65 M0 47. 442 3" am 614 OLD 610 7 722 67 am4 71.1 GSA 47. 4.7 33 M1 Om on W 61 6. a"0 1"0 es + 634 10 31 23 232 0.0 as + U. 3L0 WA 63 41.7 2349 ALL 9 10 2( 50* 45 43 334 ALL a"5 652 73. 73 733 7 CHART 2.3 FEMALE PARTICIPATION RATES CHART 2.3 MALE PARTICIPATION RATES BY AGE GROUJPS BY AGE GROUPS 0~~~~~~~~~~~~~~~~~~~~~~~4 z , 0.~~~~~~~~~~~ '*' 33 535353 43 b. m S.*1,of d me Su' .. .' iem am eo as. AGE GROUP AGEGOUPS - 24 - 0. LedrSircton In theLbrba 2.22 The economic status of women la lnfluenced not only by thelr labor force participation, but also by the types of market work they perform and how much they earn. as this sectlon will demonstrate, the work force ln Turkey li markedly segregated and less than a third of women workers recelve monetary payments. The gender dlfferences Ln worker status, occupational, and sectoral dlstrLbutions are also very strLkLng. What li perhaps more remarkable, li how little they have changed over tLme. In short, since 1955 women workers have remained concentrated ln tho unpaid famlly worker category (two-thLrds) and usually in agrlculture (three-quarters). The gender segregatLon losues are examined by flrst dlicussing how segregation may be measured, and then by consLdorLng male-female dLoparltles by worker status, occupation, and economlc sector, or lidustry (Tables 2.4-2.6). The tables reval several lmportant trends in the gender compooLtLon of w4rk, worker status, occupatLon and industry. They also show a strong and perelstent imbalance ln the types of labor force actlvLtieo pursued by men and women. Meaures of Gede gMoraat Ion 2.23 To obtain an ldea of the trend ln segregation durLng the past 30 years, Tables 2.4-2.6 Laclude two summary lndlces (expressed as percentages): the Duncan or, wdlssLmlarLty", index (DI) and the women and employment index (1). IW The Duncan index shows the percent of women (or men) who would have to change jobs ln order to eliminate gender dLfferences ln the dLitributLons. The DI Lndex has a range of 0, when there is no gender segregatLon to 100%, when there is complete segregatLon ln the sense that each category (workers status, occupatLon or sector) li entlrely filled by men or women. The less famillar WE index varies wLth the dlstrlbutLon of total labor force and the distributlon of women among categories. It shows twLce the mLnimum proportlon of women who would have to hange jobs wlth men to elimLnate segregation, takLng into account female representation in the total labor force. WE has a minimum of 0 when there Li no segregation (DT - 0%). If there le gender segregation (DI > 0), the value of WE lncreasae wlth the male share of the work force# lts maximum value approaches 200% when the DI quals 100% and the labor force consosto of any number of men and only one woman the hlgher the share of women ln the labor force, the smaller the percent of women that would have to change jobs in order to remove gender segregatLon. In general, the DI and Ng measures of dLssimilarity may be lnterproted as follow s the hlgher the value the greater the degree of gender segrgatLon. $ The dLissimlarlty or Duncan index is fairly common ln the literature, whlle the WB index was presented ln an OZCD report on Women and Employment ln 1980. For details see Duncan and Duncan (1955) and ORCD (1985). - 25 - 2.24 The formulae for the Lndiess are as follows Lets Vf the number of females Ln category lt F - the total number of fmales Ln the work forcog Ml * the number of males in category Lo N - the total number of males Ln the work force; Tf- the number of persons Ln category L; T - the total number of people in the work force. DX - 100 x 31 I Vlg/ - Nj/N 1/2i L a l...a. 3 o 100 x sit I j,/ - Tj/T 1; . * 1..... The relation between the two LndLces Lo t W * (Xj/2) x (2 x D1), so that the WB lndex Li the male proportlon in the work force multLplied by twlce the Di lndex. 2.25 The Lndlces, whlch are calculated from data provided by the ILO (1990) and the 1990 Labor Force Ourvey, are based on one-dlglt classifications for occupatlon and lndustry. For the perlod 1970-1990, these claseLficatLons strLetly conform to the International Standard ClassiflcatLon of Occupations (2800) and Industrles (lSCK), whlle for 196S mlnor adjustments were made to the data, so that they would conform to the 2800 and 25C0 format. It should be noted that the lndex values reported here uggntate the extent of gender segregatlon because they are based on broad (one-diglt) defLnltLons of occupatlonal and industrlal categorles. More narrowly-defLned claseLfiLatLons - ln terms of, say, tasks would result ln hlgher magnltudes of gender segregatLon.iW Gender ComouLtLon of Work Cateaories 2.26 As seen ln Table 2.4 below, the share of women ln the total labor force deelLned steadlly frci 41% ln 1960 to 31% ln 1990. It li noteworthy that, over the same period, the femAle proportlon of unpald family workers fell from 75% to just under 70%, whlle that of wage earners rose frozm only 11% to 16%, indLcatLng that unpald famLly workers remain largely women, whlle wages earners remaLn mainly men. As regards self-employment, whlch lncludes both own-account workers and employers, the small female reprosentatLon has changed little over time (7-10%).ff ly Dy way of clarification, the terms "gender composltLono or gender representation" refer to the proportlon of women (or men) ln a glven work category, say in the clerlcal occupatlon. By the tern "segregatlon" we mean a notLe able Imbalance ln the male and female dLstributLons across work categorLes, for example, AL1 occupatLons. 1V HistorLially, the proportlon of employers among the self-employed has been relatlvely small for both men and women. But, the number of women who are employers li negligLble ln comparlson to men. The 1990 Labor Force Survey shows that there were 665,042 men who were employers (13% of the male self-employed), but 12,313 women (30 of the female self-employed), so that the representatlon of women among employers Lo negllglble - less than 2%. in 1980 thwr were 7,218 female employers (2% of the self-employed) versus 169,241 men (4% of the self- employed), whlle in 1960 there were 5,569 female employers (1.5% of the female self-employed) compared to 150,539 male employers (4% of the self-employed) See Yayan (1984). - 26 - TABLE 24 Oe t U0N OF TOTAL LABOR FORCE BY WORKER STUS ANO GENOR 9140. 1910 1915 1970 1975 11#0 1915 1910 MALE FLEW MALE FEMALE MALE FEMALE MALE FELE MALE FEME MALE FEMALE MALE FEMALE WORKESRGATUS mm TTAL W7 5GM On 141 OM 013 111110 9104 1354 _ 139 7547 14163 641 %fmde 41 so 36 36 36 36 31 SELF,EMPLOID 346 371 367 110 3746 37 4A0M 30 4123 1 446 se2 6107 37 S _iun 10 4 10 7 7 7 7 UMPLOYE 2177 9 2765 1S6 3576 4510 77 61 o46 am 1072 69 1211 % em 11 10 14 I$ 11 1S 16 UNPAIDWORKER 166 460 1739 4602 1l64 420 4 aS 2224 an am m 200 4438 % bm 75 73 71 es 70 as 69 UN WNOWN 49 6 5 101 4 0 0 13 4 6M 116 670 165 67 460 -I_0 1 4 0 24 16 1S 4S PERNCNrAGE OIEUMUON OF MALE AND FEA.LE LABO FORCE B'Y WOR KERTATU 16 lS 197 1975 1950 IS 190 MALE PEMAL MALE FEAALE kMLE FEMALE MALE' P5MW MALIE P5MW MALE FEMALE MALEK FEMALE WORKER STATuS TOTAL 100.0 100. 100.0 100.0 10. 10 1 100 100.0 100.0 10 100 100.0 100.0 1010" 100 BELF4.EPLOV 4.1 7.0 45. 29 402 0.6 4.6 22. 4. 323 4.7 33.7 Le EMPLOYS a3 4.9 U5 5 364 10.2 40.3 14.1 4L5 137 434 14.0 462 127 UNPAIDWORKR 3U SLO 0.6 01.1 21.3 69 2L27 0a 16" 79.0 10.1 76.2 142 684 UNKNOVM GA 0.1 1.2 0.1 0.0 00 0.1 0.1 1. 1.7 6a 20 4.0 7.1 a11140C 6tJ7.7 704 61J 572 61.0 1 57.4 WE u 62 7. 76 7 76.0 77.6 76.7 8:urs FQr 1965.1955 Yinoolt dabor I d s11w (19t0). 1L0: br 1910. M.IwId Lor Foro 6uIj,. 912. Not.:Te amo pybwiim4lbsjohsner.0 w 0ufiIpbid. E UeWas Id neswhuiBeeWl o deaot - 27 - 2 27 The rep5resentation of women within occupatlone and eoconomLe sectors (Tables 2. 5 and 2. 6), however, reveal dLverging patternin namely the female share of agricultural workers remained essentially constantt the proportLon of women ln professional occupations rose steadily (from 21% in 3965 to almost 30% in 1990)g the share of women among service workers almot doubled (from 7% in 1965 to 12% in 1990) and that of clerLcal and sales workers also increased dramatically. The share of women in managerial and administrative occupations, however, remained unchangedW - at about 6% between 1970 and 1990 - as did that of women among production workers - about S%-100 over the period. 2 28 As regards female presence in dlfferent econoicL sectors (Table 2. 6), significant increases are found only in womn's share in the manufacturlng and services sectors (both business and personal srvlces) * Women's shae in trade, transport, mining and construction remained low throughout. Gender 8eareaation 2.29 Analysis of trends in the gender compositLon of work activltles is useful, but does not reflect differences in the *Lsi of the male and female labor forces and, hence, their participation rates. The shares of men and woen cannot be expected to be the same within, say, each economLc sector if there are more men than women in the total labor force. Thus, it is useful to compare the work activity distributions of men and women, i.e. the proportLons of all men and of all women in various categories. The lower halves of Tables 2.4-2.6 show these distributions and the correspondlng DI and WE lndLces of segregation and they reveal clearly substantial gender dlfferences. The major changes by gender that have occured since 1965 can be summarized as follows Nale Fema1l % of Male/Female LF in Agriculture 58 95 34 74 % of Male/Female LF that is unpaid 21 91 14 68 % of Kale/Female Workers that receive a wage 33 6 45 19 X of Male/Female Workers who are self-employed 46 3 37 6 DI/WE Index by Worker Status 70.5/87.8 57.4/18.7 DI/WE Index by Occupation 36.2/45.0 41.6/57.3 DI/WE Index by Economic Sector 37.4/46.6 42.6/50.7 I Zn 1965, clerical workers are lncluded in the Oadminlitrative category. - 28 - TABLE 2.5 SO TION OF TOTAL LABOR FORCE BY OCCUPATION AND GENOER 9654O. 96 1970 1175 1960 19*5 1990 MALE FEAL MALE FEMALE MALE FEMALE MALS FEMALE MALE FEMALE MALE FEMALE OOCUPATIONAL GROUP MM TOYAL 8472 5141 OM06 5613 lt11O W4 12264 62 130 7647 14165 6466 %femm 30 36 36 36 35 31 PROFESSIONAL 26 62 415 140 468 173 565 255 714 296 830 328 %fsn.I 21 25 27 30 23 28 AOM RATIVIE 306 46 sO 5 71 4 149 t 15l 10 30 15 % smds 13 6 6 5 6 5 CL UIA 0 0 23 73 401 133 436 211 496 236 676 419 %S 0amel 21 25 32 32 38 SALES 375 a 446 17 536 34 757 34 921 57 1426 113 %tomme 1 4 a 4 6 7 SERVICE 442 33 510 47 S28 56 847 67 1023 s8 1488 252 %htnuml 7 6 10 7 8 14 AGRICUXLTUFE 4903 4636 494 5153 590 5430 5123 5932 5t06 6464 4813 4806 % k W -, 50 51 48 54 54 50 PRODUCTION 1690 15 1106 262 3270 375 3777 305 4140 340 4512 525 % hSruid 6 10 10 7 8 10 UNKNOWN S20 4 152 116 0 0 607 116 875 165 115 28 %Iumnuls 1 7 0 16 15 20 PERCENTAGE OITEIUTrION OF MALEA NO FEMALE LABOR FORCE EY OCCUPATION 1965 1970 1975 1980 1985 1990 MALE FEMALE MALE FEMALE MALE FEMALE MALE FEMALE MALE FEhALE MALE FEMALE OCCUPATIONAL GROUP TOTAL 100 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 PROFESSIONAL 2.8 12 4.5 2.4 4.2 2.8 4.8 37 5.1 3.9 *5.9 5.1 ADMINSTRTIVE 3.6 0.9 0.9 0.1 0.6 0.1 1.2 0.1 1.1 0.1 2.2 0.2 CLERICAL 0.0 0.0 2.0 1.3 3.6 2L1 3.6 3.0 3.6 3.1 4.8 6.5 SALES 4.4 0.1 4.8 0.3 4.8 0.5 6.2 0.5 8 0.7 10.1 1.7 SERVICE 52 0.6 5. 0.5 4.7 0.9 6.9 1.0 7.3 1.2 10.5 3.9 AGRICULTURE 57.9 94.1 53.2 98.6 52. 57.5 41.7 85.6 40.2 84.5 34.0 74.1 PRODUCTION 20.0 30 11.9 4.5 29.3 60 30.7 4.4 29.7 4.4 31.9 6.1 UNKNOWN 1 0.1 16.4 2Lo 0.0 0.0 4.9 -1.7 6.3 2.0 0.8 0.4 t I- 36 35.5 34.7 43.9 44.3 41.8 WE INDEX 45.0 43.7 44.6 56.2 572 57.3 M Mres FOr 06.185 Yowbo4* ot L@Saw lIog (1980. ILO; for In0. Hou;w T*d jbrFore VW" wxy. Sa5 Nobn: 'Pro h Ide. NbhOWl & reAd; Am sv Includi m _geI and In 1965 minrIn : PmdueUon, Incjde tanupa equ'pmFn "kimownltofirdes6t b job iemkem. O f hu OhIwIy Id WE fIs us Wolmean sd Employmem: hiweix. See oWoc for dep - 29 - TABLE 6 DIWRITION OF TOTAL LABOI FORCE BY ECONOMIC SETOR ANO GENDE, 19SO6S0. 1965 19710 1975 1960 1965 1990 MALE FEMALE MALE FEMALE MALE FEMALE MALE FEMALE MALE FEMALE MALE FEMALE ECONOMIC SECTOR 000's TOTAL 472 5141 9G06 5613 ll180 a204 12264 O62S 13033 7647 14165 6486 % f.nme 39 3S 30 30 35 31 AGRICULTURE 4679 46 8031 5200 621o 544 5156 5949 534 6484 4884 4802 %bem o 50. 51 47 54 54 50 mINuNG 64 1 107 4 110 1 131 1 135 2 177 2 % fohma 1 4 1 1 1 1 MANUFACTURUNG 866 75 9es 290 1200 257 1672 304 1653 332 2177 437 % kna1 6 23 16 15 15 17 UTnLITES 26 0 14 1 16 2 31 2 22 1 41 0 % female I 6 12 6 5 0 CONSTRUCTION 352 1 422 9 527 5 760 5 744 7 979 16 % femna 0 2 1 1 1 2 TRADE 364 21 646 27 767 47 1034 50 1301 62 2006 145 % female 5 4 6 5 6 7 TRANSPORTATKON 2 6 3t4 16 465 1S 505 26 5S6 31 774 34 %f.mlso 2 4 4 5 6 4 BUSINESSSERV1CES 697 103 115 31 166 53 218 76 291 99 296 109 %fgmiao 13 21 24 26 25 27 OTHER SERVICES 0 0 1376 161 156 272 2067 359 2406 441 2264 480 %eSmale 0 12 16 15 15 17 UNNOWN 906 46 ,255 54 166 63 710 157 962 18 567 460 %female S. 16 27 1s 15 .45 PERCENTAGE 0DTWBUTON OF MALE ANO FEMALE LABOR FORCE BY ECONOMIC SECTOR 1lo5 1970 1975 1960 1985 1990 MALE FEMALE WALE FEMA MALE FEMALE MALE FEMALE MALE FEMLE MALE FEMALE ECONOMIC SECTOR TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 AGRICULTURE 57.6 95.0 54.1 89.5 5.S 88.4 42.0 U5.9 40.4 84.8 34.3 74.0 MINING 1.0 0.0 1.1 0.1 1.0 0.0 1.1 0.0 1.0 0.0 1.3 0.0 MANUFACTURINO 10.2 1.5 10.7 5.0 10.7 4.1 13.6 4.4 13.3 4.3 15.4 6.7 UTILMTES 0.3 0.0 0.1 0.0 0.2 0.0 0.3 0.0 0.2 0.0 0.3 0.0 CONSTRUCTION 4.2 0O 4.5 0.2 4.7 0.1 8.2 0.1 .3 0.1 6.9 0.2 TRADE 4.5 0.4 6.9 0.5 68 0.8 0.4 0.7 9.3 1.1 14.2 2.2 TRANSPORTATION 3.3 0.1 3.7 0.3 4.2 0.3 4.1 0.4 42 0.4 5.5 0.5 USNESSSERIVICES 6.2 2.0 1.2 5 1.5 0.9 1.8 1.1 t.1 1.3 2.1 1.7 OTHER SERVICES 0.0 0.0 14.6 3.1 13.9 4.4 16.8 5.2 17.3 5.6 16.1 7.4 UNKNOWN 10.7 1.0 2.7' 0.9 1.5 1.0 5.8 2.3 6.9 2.2 4.0 7.1 01 INEX 37.4 X354 32.8 43.9 44.4 42.6 WE INDEX 56- __74 WEINOEX ~~~~ ~ ~~46.6 43.68 42.3 56.1 57.3 5. Smom: For 1968.1565. Youbook of Labor SlAUsile (198W. ILO; for 1990. &sjshlcld Labor Fare. Survey. Sit. Noes Um' Inlude .ie tV. gm & woo r.Trade irulue whdOeWr, trde. res_t & hook;ls e Ifvdude flnanceo In uwrune and eWa st; soIher m's to oamAi. s69 & pwwI sevcs 'UwJW iidm ime job s5o5kWes. 01 is O bt DM*el bIde; WE Is Women and Empoyent bdmL See td for d _auon - 30 - Thus, while ther have been soom changes ln the type of work that women do, the majorlty of women continue to work La agrlculture, s unpaid family workers. 2.30 Whle- the gander distributions n nonL gnicultural occupational and sectoral grouplags (Tables 2.5-2.6) suggest a mall shlft ln the position of women, both the DX, and WE dlm show not only considerable, but LncreaeLng .gender segregatloa over time. 2.31 When compared to the data for other MMOD and developLng countrles, ummrizsed below, the data for wurkey indlote a blhb and perseitent degree of gender segregation ln the workplac ln Surbkys DI .matm _ . . ... .. Australian 49 spain 37 blsdu 42 psrte 2 VSA 4 Or 2r C_ B P Ws 19) 6s mat !!M -3 summt f19") 3OECD, 1901. Sima .u rtebw. 1tw. Given also the observed iras in the lndlm of disolmlarlty ln Turkey over time, lt is unlikely that gander segpgtion will be reduced without pollcy inltiativeo designed to ncourage womnn to enter mno-tradltlonal work actLvitLes. 2.32 RecognLsang that ther are may oomplex factors that generate and perpetuate the gender dlfferenoe Ln the distributlons of work activltLes, there are, nonetheless, several ares whor polliy inItLatives could reduce gender segregation. For instances *ducatLon a * in!s the types of work that pople prform are related to the skills that thoy brLng to the labor mroket. Gender dLfferences ln education and trainng level and LA the content ot courses and programs are ignilfieant determinants in channlllng men lnto one set of work actlvltles and women into another. Reducing thes dlfferences can play an important role ln modlfylng the gonder dlvLsion of the labor market. * Ei& L.s Zasalas s Some of the existing protectlve leglslation, whlch prohLblts women fro workLog in *unsuLtable" jobs or workLng at certaln hours, also contributes to osgreatlon. Repealing the women- spoclflc provisions of such laws will have both a symbolLi and substantlve lmpact on women's work oholces, as well as thelr educational choices. - 31 - _m n Nuyhv Prom: e nlatr .tLon of Oeployment equLty" programs could also be benefioial (ase Box 1 be?@w). 2 I33 In ILh of Tuky inrasn uIazto th pr rtI ilf .\ : w. , * . 3fW :'_>-g2g>.S |~~~~~~~~ * ! fl s tw stun r t g S ffi wwm ! m i 1 . L a in mf patterns L Lmg;pt rtaonts . partLalarly frc the standpoint of designg po.Lay m !a=u . a lleviate ndy T d. U Uran Dinra.l msn an Difer .ns 2.33 Sn light of Turkey's inrerain g urbanriation - the proportion of population living in urban ar as ha t vlrtually doubled fro 29 ln 1955 to 59% in 1990 - a consideration of urban-rural labor force patterns is important, particularly from th. standpoint of designing policy measures to alleviate gender teregation ln the labor force tn the absence of reliable historical urban- rurtl labor force data the analysi here is based on the 1988-1990 Household Labor Force Surveys Trends ln Urban-Rural Particination Rates 2.34 Table 2.7 shows the evolution of participation rates by gender in rural and urban arear. A notable aspect is that the participation rate of rural women remains significantly higher than that of urban women - 51% against 15% but that the rural participation rate has fallen from 90% in 1960 while the urban participation virtually dcublod. - 32 - IAILL 2.7: Participation Rates In Rural and Urban Area, 1955-90 URBAN RURAL YEAR NEN WOWEN NEN WONEN I(X) 1955 78.7 6.3 93.0 92.3 1960 75.2 8.3 91.8 90.4 1965 73.7 8.2 90.8 88.6 1982 68.6 11.2 n.a. n.a. 1983 69.9 12.0 n.a. n.a. 1984 68.9 13.1 n.a. n.a. 1985 69.9 13.1 75.1 47.4 19880 72.9 16.9 80.4 51.8 1989A 71.9 16.5 80.2 55.6 19890 71.0 17.2 80.1 54.3 1990A 70.3 15.2 78.0 51.2 Sources: 1955-1965 Durand (1975. P. 166); 1982-1990 Household Labor Force Surveys. Notes: 1955-65, the eligibility age Is 15 years. otherwise 12 years. A - April; 0 - October. &ae Strugture Patterns of Urban-Rural Particimation -2.35 Table 2.8 and Chart 2.8 show the cross-sectLon age proflles of participation for men and women according to place of residence based on the Labor Force Surveys. The profiles show that the female partiLcipation rate is much higher among the rural population than among the urban population and that the rural profile shows a falrly dietinat K-shape, while the urban pattern has a small peak at age 20-24 and then declines steadily thereafter - reflecting the reluctance of employers to hire women with chlldren. It is also significant that there is little variation in the partieLpation rates of rural women betwoen 15 and 55 years - from 55% to 63%. The urban and rural male profiles have the expected "male" bell shape and show little dlfference except for the relatively sharp decrease in participation rates of urban men after age 45 years, whereas in rural areas it only starts to decline slgnificantly at age 60. 2.36 The urban-rural differences in partlcipatlon rates, especially among women, can b explained in terme of differences in labor market conditions of rural and urban areas namelys differential unemployment rates, the "mix" of economic sectors and occupations, earnings prospects and work opportunitles compatible with educational attainments and skill levels. The low participation rates of urban women (who are generally more educated than their rural counterparts) may also reflect the "discouraged worker" phenomenon documented TABLE t. UlA*AUNWPL PARTCWAION RATES BY GENM AND AGE GROWS FEMALES MALES I URUW' 0RURL WFN RiUAL AGE t911 t9 m 11 im i19 tm im im 99 AGE los 1 19B9 1111 1919w 1 *99 12't4 62 6.6 72 5.3 40.9 39.4 6.7 367 12-14 212 2" 22.0 18.5 9.4 3.5 22 3" 15.19 21.5 18.9 21 16.0 59.9 61.6 Ole 05 15S19 5.0 S6 5.1 SU 732 6W? 702 683 2024 26B 25.9 26.0 24.6 56.8 642 59.7 57.0 244 86*6423.04.1 910 91t 91.7 n9 25S29 21.7 22 21.7 200 552 572 589 55.1 20 962 97 67.3 1 SU. 4 8.0 SA 0&6 3044 22.3 2Z7 232 2D .S57.3 6 5.B 55.6 3034 KG . " 64 1. " 9 0297.9SMA075 3549 1.6 22.7 2.3 22 60.7 66 64. 354 35 935. 9 6.1 0. 9 I A4 0.0 07.0 07 4044 16.9 15.0 172 10.6 5.1 0S 06.6 629 4044 95. 16. 96.0 94 96.6 05.1 97. 963 4549 11.4 11.5 11.0 9.6 50.0 70.9 57 62" 4540 9. WS 852 6.4 04.1 94.4 9.4 04.3 04 9.6 9.3 6I .5 55.4 593 552 504 709 73S 67 06.00 91t * .79.5 00.5 55S9 62 4.3 84 4.1 47.0 51.5 S.9 492 55 9 50.4 0S S.3 EU 64. S5.5 6S.4 6044 3.7 3.5 5.6 2.9 32.8 392 44.6 34.0 044 39.7 38.1 35.4 32.0 74.5 76.0 770 63 65 1.5 1.1 29 1.1 16.9 172 18.0 14. as + 17.4 1.1 1.1 142 4 45.0 47.9 39.9 ALL 16.9 17.2 152 51 59.6 543 512 ALL 17271* 71* 70.3 80.4 802 601 7.0 ooe: Hoshold Lbo Fos &Ws. M. A ApW 0. Oobr. W CHART 2.8 FEMALE PARTICIPATION RATES CHART 2.8 MALE PARTICIPATION RATES URBAN AND RURAL - BY AGE GROUPS URB3AN AND RURAL - BY AGE GROUPS 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~7 0. 0 s. AGE GROUPS AGE GROUPS - 34 - elsewhere, (Lee., urban women may forego participating ln the work force lf they do not flid approprlate employment, at an acceptable wage, ln a favorable work envlronment or ln a reasonable period of time). The low rates may also refleot prevaillng cultural norms that preclude women from associating wlth non-kln males. The low rates, however, also reflect the fallure of offlclal labor statlitics to lnclude workers ln the lnformal sectors. Urban-Rura atensof Activities In the Labor bEra 2.37 Tables 2.9-2.11 conslder the gender compositlon and dlstrlbutlon and segregatlon measures by resldence across worker status, occupatlon and sector for 1990. Essentially there are three relevant measures - women's share of the urban/rural labor forcel the urban/rural share of the female labor force and the urban/rural female partlclpatlon rate - wlth slmilar measures by worker status etc. For simplicity, the key measures contained ln Tables 2.9-2.11 are summarlzed below. in terms of gender composltion, the urban-rural patterns reveal few surprises. it is significant to recall that 77% of the female labor force is in the rural arecs, that the female share of the total rural labor force ia 41%, but only 17% of the urban labor force and that in rural areas, 71% of unpaid workers are womon, ccapared to about 33% ln urban areas. UEhIn ŁuAu 12l Proportion of Femle Labor Force 23 77 100 % of LF that Is Femle 1? 41 31 Female Participation Rate 15 51 33 -roportion of FLF that ts unpaid 3 97 100 Females as % of all unpaid workers 33 71 69 % of FLF that is unpaid 10 6s 66 - 35 - TABLE 2.9 CSTRIMlN OF U 1SAN.RURAL LABOR FORCE BY WORKER STATUS AND GENDER. 90 WORKER STATUS mmLE FEMALE MALE FEMALE TOTAL 7060.0 14HI 7105.2 1020.5 %femele 172 41.4 REGULAR EMPLOYEE SA $6t.0 1312.2 169.4 % femee 17.0 11.4 CASUAlA EMPLOYEE 1"9 97.1 467.4 129.9 %femb 13.6 21.0 EMPLOYER 6.1 11.6 125.2 1.3 %fma 21 1.0 SELF-EMPLOYED 121.6 1019 3t.7 261.0 % fwnW* 7.3 7.5 UNPAIDWORKER 2U2* 144S 1727 4293.2 %ftolw 33. 71.4 UN3WN a09 us 23.0 166.0 % __ __le 472 41.1 PERTSFAE OSTEU?1ON OF UMNARURAL LABOR FORCE BY WORKE STATUS AND GENDER 1990 WPMN RURAL WORKER STATUS MALE FEML MALE FEMALE TOrAL 1OQ.0 1000 100.0 100.0 REGULAR EMPLOYEE 554 Su6 IS5 3.4 CASUAL EMPLOYEE 0.7 6 6.9 2.6 EMPLOYER 7* 0s 1* 0.0 SELFUEMPLOYED 16z 7.0 463 52 UNPAID WORKER 4.0 9A 242 65S UNKNOWN 4.7 2.1 3.3 3.3 oD1INDEX 21.3 61.3 WEINDEX 3S.3 71.8 Soure: Husehold Lbor Foa 4 190N 1 S. Nows: Th.unknown' ouMy bdua *re st jb seeers. DOIl 0is dbsiml Now: WE Is Gms Wom wnd m0theam See $a w efor IFo I - 36 - TABLE 2.10 DRBUTON OF URBANRURAL LABOR FORCE BY OCCUPATION AND GENDER 1990 URBAN RURAL OCCUPATIONAL GROUP MALE FEMALE MALE FEMALE ('oa) (000) TOTAL 7060.0 1465.6 7105.2 5020.5 % emls 17.2 41.4 PROFESSINAL 560.5 242.2 235.3 53.9 %ofeOle 302 18.6 ADMIN18TRATIVE 226.5 132 84.0 0.0 %know 5.5 0.0 CLERICAL 460.3 247.5 124.4 34.6 % mn mdo 35.0 21.7 SALES 1104.7 86.7 308.2 14.8 % fsmeb 7.3 4.5 SERVICE 950.1 126.8 342.6 19.9 % fomd 11.8 5.5 AGRICULTURE 2622 184.8 4587.8 4614.8 %emele 41.3 50.1 PRODUCTION 3130.8 256.9 1168.9 111.8 % fesmb 7.6 .7 UNKNOWN 364.9 307.5 253.9 171.1 %femal. 45.7 403 PERCENTAGE D0TRBTON OF URBAN-RURAL LABOR FORCE BY OCCUPATION AND GENDER. 1960 URBAN RURAL O0CUPATlONAL GROUP MALE FEMALE MALE FEMALE TOTAL 100.0 100.0 100.0 100.0 PROFE SIONAL 7.9 16.5 3.3 1.1 ADMINITATIVE 3.2 0.9 12 0.0 CLERICAL 6. 16.9 1.8 0.7 SALES 15.6 5.9 4.3 0.3 SERVICE 13.5 8.7 4.8 0.4 AORICULTURE 3.7 12.6 64.6 91.9 PRODUCTION 44.3 17.5 16.5 2.2 UNKNOWN 52 21.0 3.6 3.4 DI INDEX 43.7 27.3 WE INDEX 72.3 32.1 Sourc: Hou or Foce uy (1990, Mt. N:719femionar Inkluds Wghrdca & mIsled;AdmInstatve'includes mn.negi 'Production' in takhpoet equIment lmmim' in GM rn job seekesm. I is te DssImonty IWndex: WE i tw Wonen wnd Emplyment In See te for descrption. - 37 - TABLE 2.11 ODISTRIBTION OF URBAN-RURAL LABOR FORCE BY ECONOMIC SECTOR AND GENDER. 10 URBAN RURAL ECONOMIC SECTOR MALE FEMALE MALE FEMALE (000) (000) TOTAL 7060.0 1465.6 71052 5020.5 % t f.m 172 41.4 AGRICULTURE 247.5 185.0 4616.3 4617.4 %femule 42.8 50.0 MININO 772 0.0 100.2 1.5 %fe_al 0.0 1.5 MANUFACTURING 1811A 328.0 365.1 111.2 % f.nml 15.3 23.3 UTILITIES 20.9 0.0 20.5 0.0 %hfnab 0.0 0.0 CONSTRUCTION 624.5 136 354.7 1.9 % fh_ 2.1 0.5 TRADE 1568.0 124.3 418.2 21.0 % f.mab 7.3 4.8 TRANSPORTATION 517.9 28.1 255.8 5.9 %fenab 5.1 2.2 BUSINESS SERVICES 252.? 95.8 43.7 13.5 %1 mmna 27.5 23.7 OTHERSERVICES 158.9 397.4 694.6 83.0 %fsmnuI 2Q.O 10.7 UNKNOWN 330.9 295.3 236.0 165.0 %male _ 472 41.1 PERCENTAGE DISBUTION OF URBAN4RURAL LAOR FORCE BY ECONOMIC SECTOR AND GENDER. 1990 URBAN RURAL ECONoMIC SECTOR MALE FEMALE MALE FEMALE TOTAL 100.0 100.0 100.0 100.0 AGRICULTURE 3 12.6 65.0 92.0 IINING 1.1 0.0 1.4 0.0 MANUFACTURING 25.7 22.2 5.1 22 UTILTES 0.3 0.0 0.3 0.0 CONSTRUCTION 8.6 0.9 5.0 0.0 TRADE . 22.5 8.5 5.9 0.4 TRANSPORTATION 7.3 1.9 3.6 0.1 BUSINESS SERVICES 3.6 6.5 0.6 0.3 OTHER SERVICES 22.5 27.1 9.8 1.7 UNKNOWN 4.7 20.1 3.3 3.3 01 INDEX 32.1 27.0 WE INDEX 53.2 31.6 80ur : Houmeluold Fame Survey (1990). SIS. Nuam: UEe' Inludex gm & water; de includes whol6es_leretell uda. reStRurts and hotel: uL'n sevices Ineude bL wand eae 'o.p v 'refem to community. sooW and p- earvice. The 'unknwn' aegoy hcludes fdgt Mn pb ekers. 01 s Oh Dlsbnmlet Index; WE 1s th Women and Eloment index. See text for desrpIon. - 38 - 2 38 The urban and rural patterns of gender sagregatLon by worker status, and the measures of disLsmLlarity reveal some LterestLng dLfferences. Theme can best be -umarsied as follows: Rural Un ble Femle MaI. Female % of LF classified as Unpaid worker 24 86 4 10 Employee 25 6 65 62 Self-Employed 45 5 16 7 Employer 2 0 8 1 in short, there is a marked difference in employment status between urban-rural females and rural males and females, but a greater siLmlarlty betwen urban males and females. GLven the continuLng urbanLzatLon of Turkey's population, thli pattern has important implicatLons both for the rgularLsation of female employment condltlons wlth urbanization and for agrlcultural productlon. The urban-rural dLfferences in the gender distributions of worker status are refleoted ln the Lndices of segregatlon and, not surpriLsngly, show a much hlgher degree of segregatLon in rural areas - the DI Lndex ln urban areas is 21% compared to 61% ln rural areas). 2 39 As regards gccuations and sectors, the tables show a hlgher degree of gender segregatLon in urban areas than in rural areas because farmlng li the maln actlvity of both workLng men and women ln rural areas - 92% of women and 65% of men work ln agrlculture. As a result, the segregatlon measures are relatively low (le than 30%) in rural areas. Gender segregation ln urban areas for occupatLon and industry is higher, ln part, becaus women comprise only 17% of the urban labor force. Tn ter of occupatLons, the share of women ln the more "modern- activitLes (proa..dsions, clerlial, sales) is slgnlficantly higher ln the urban than ln the rural areas. 2 40 Thus, the data lndicate that, notwLthstanding the dLffeorences Ln the urban and rural patterns, the work actlvitles of men and women are markedly segregated ln both urban and rural areas. The urban-rural dLitrLbutLon of the female labor force, ln conjunctLon wlth the dlfferences in gander egregatLon, and the declining rural, but riasng urban partlclpatlon rates, is of special pollcy importance - it lndicates the existence of two largely self-contalned female labor market segments. Hence, the policy inLtLatlves that would enhance the employment standLag of the two groups of women are lkely to be dlfferent. - 39 - P. Hour- sf Work 2.41 The accurate measurement of labor supply requires lformatlon on varlous dimensions of work behavlor. In addltlon to labor force partLclpatLon, data on hours of market work are especLally Lmportant.Wv It La posuibl- for partLiLpatLon rates to vary over time and between groups because of dlffering amounts of time worked - hours per week, weeks per year, and over the lit cyole, as well as because of unemployment spells. For Lnstance, the contLnuLng growth of female labor force partLcipation in the OC3 countries ha ben accompanied by a general reductLon ln average annual hours worked and by a pronounced increase ln part-time employment. Women account for most of th growth ln part- tlme work, whlch ia concentrated ln typlcally female actlvltles - clerlcal, sales and services .i 2042 Untll the implementatlon of the rvlsed Household Labor Force Surveys ln the late 1980s, the measure of labor supply in Turkey was limted to one dimension - labor force partLcipatlon. The surveys now provlde lnformatlon on weekly hours worked in the maln job by thos who are employed.Wv Thli Lnformatlon, dLiplayed ln Table 2012 and Chart 2.12 reveals sveral important poLnts, especially as regards workLng women. Flrst, as can be seen there are no major gendet differences ln the average weekly hours of work among the employed. Women on average work 43 hours per week compared to 50 hours for men, wlth slLght varLatLon across urban and rural areas and age groups. The percentage dlitrlbutlon of working hours by gender La also noteworthys most men work at least 40 hours per weeks natlonwlde, 90%g urban, 95%g and rural 83%. The pattern for women i somewhat dlfferents nationwLde, 32% of women work 35 hours or les and 36% work at least 50 hours, whether wage erners or unpaid family workers. It is also important to note that 50% of unpald famLly workers work more than 50 hours a week. WV Ideally, a measure of labor supply would include lnformatLon not only on the proportlon of people seekiLng or engaged ln market work (labor force particLpants) and hours worked, but also on the level of skills and competence that workers brLng to theLr jobs, the contlnuLty of labor force attachmnt, and the effort devoted at the job. Skill level are measured imperfectly by educatlonal attalnment, whlle LnformatLon on work contlnulty or lnter-labor force turnover is generally not avaLlable ln developLng countrLes, and measures of effort are basically nonexistent. F Vor example, ln Canada, prime-age womn represent about 75 percent of part- time employment. The other important groups of part-tlm workers are youth and older men. The Canadlan pattern li typlcal of other O0CD countrles. See OCD (1985). 12 The LnclusLon of multiple job holders does not change the magnltudes because less than 3 percent of men and almost no women hold more than one job. However, there are no data on lnter-labor force turnover, l.e. the number of week per year worked. - 40 - TASUIE 2L12 WEEKLY HOUR F O WORK IN MAIN JOB FoCR EMPLOYED. 190 TOTAL URAN RURAL AGE GROUP MALES FEXALES MALES FEMALES MALES FEMALES f CO RS X100 OS 0 HOURS X00 HOURS 1W0 HOURS VW0 HOURS. 12.14 S601 4tJ 302L7 42.1 124.0 53.2 31.6 44. 236.1 46.5 271.1 41.8; 15.13 1417.7 49.9 100 43.1 m.6 52.3 125.5 46* 7892 47A 663.3 42L7 0.24 126*U M4 763.1 442 62A* 60.1 177.2 45.1 6725 50.7 565.9 44.0 21.2 1795.6 512 6423 439 1003.5 51.7 163.5 43.6 7SC* 0.5 4768. 439i 944 16l962 50.1 614* 434 976*8 49.9 170.9 41.3 721.4 50.3 443.7 42.8 1 Alil 1640U3 50U 6872 43U5 92M.3 506 164* 40.6 711.0 505 423.0 44.7 40Wt4 12:E1 49A4 45@.4 4&I 71&.9 46A 108.9 39.6 &D*.a 50 5 =5x 44.1 44 .I 5 0.4 30 qLO4 471.1 OO;D "4. 42.0 4118.0 50.7 354.4 42.0 064 On?7 49.6 372,9 41.6 36.7 512 27.5 41.4 46.0 48.4 345.4 41.6! S59 674.8 492 321 432 214.0 51.1 16.0 42 . 469.9 482 313.7 43.0 664 464.8 43.1 166.7 40.5 101.* 22 9.1 402 36.0 482 179.6 40.0: U5+ 3767 44.0 106.5 37.0 73.5 49.5 6.4 50.3 3032 42.7 101.1 36.3 LLMIS 1StO. 49.9 5774.1 429 6200.4 50.6 1041.7 4S' 6s09.7 492 4733.4 42.9! S- I*ii" i Xoi F uiwy (1990\. 816. > NeW: 9~nb muv n d v. W rWd CHART 2.12A HOURS OF WORK- TURKEY BY GENDER AND AGE GROUP AME GOP ,HART 2.12B HOURS OF WORK URBAN CHART 2.12C HOURS OF WORK - RURAL BY GNDER AND AGE GROUP BY GENDER AND AGE GROUP lo"AGE .AGE ROUP - 41 - 2.43 Thus, the hours worked dimension of labor supply suggests that female labor force partLeipants devote a conslderable amount of time to market work, whlch is ln addltLon to household chores and chlldcare responsibilitieo. The hours worked dimension may partially explain the relatlvely low participation rates of women in Turkey. This underscores the usefulness of time-use data to assesu how women allocate their time between nonmarket and market work, and to determine the extent to which the household responsLbilities act as deterrents to employment opportunltles in the labor market. Moreover, time-use lnformation would be valuable in designing polliy initiatives that would reduce women's workload ln the home and in the market. These polLcies, in turn, could result in significant productivlty lncreases in both the household and the workplace, as well as creating employment opportunities. 0. Nducational Attributes of the Labor Porce 2.44 The economic slgnificance of educatlonal attainment, as a proxy for labor force qualiflcations and skillo, is widely recognized. It is perhaps the most important determinant of individuals' labor market outcomes - part˘icpation, type of job, unemployment and earnings. Thus, thls section considers gender differences in educational attainment in the labor force for the period 1970- 199O.W. During this period, signlficant progress has been realized in reducing the gender gap in education. Table 2.13 presents the statistical informatLon for the country as a whole, while Table 2.14 displays urban-rural differences for 1990. NatLonwide Patterns 2.45 Several interesting trends merge from the Table 2.13. Although Ilteracy rates of women have lncreased sharply during the past few decades, women comprise a signlficant share of the labor force that is illiterate. The female portlon of the illiterate work force rose slightly from 60% ln 1970, to 63% in 1990. The female share of the functionally" illiterate (literate but no diploma) work force also rose from 23% in 1970 to 30% ln 1990. Women's representation ln the labor force with primary schooling rose from 21% ln 1970 to 27% ln 1990. The patterns for junior hLgh school and hLgh school show a relatively constant female repreosentatlon - junior high school, 13-15%, hlgh school, 24-27%. The proportion of women ln the unLversLty-educated labor force fluctuated between 1970 and 1990 but in 1990 was about 26%. In sum, the results show that the illiterate labor force is domLnated by women, while the labor force wlth schoolLng consists mainly of men. 2.46 Examlnlng gender dlfferences ln the distrLbutions among different levels of education shows more positlvely the progress ln women's education. The proportion of illiterate women in the female labor force has fallen dramatically from 69% in 1970 to 29% ln 19901 the correspondLng values for men have also fallen from 29% in 1970 to 8% in 1990. At the same tim, the proportlon of women wlth primary schoollng rose from 20% in 1970 to 48% in 1990, whlle that of men Lncreased from 46% to 60%. The proportion of women with post-prLmary schooling rose from 4% in 1970 to 15% in 1990, compared to 9n and 24% for men. These statistics on male-female trends in educatLonal attainment suggest declining gender dlfferences, whlch is conflrmed by the marked decline in the DI and WE indices between 1970 and 1990. Informatlon was not avaLlable for the years prLor to 1970. - 42 - TALE 213 OIgST JItON OF TOTAL LABOR PORCE BY EDUCATION AND GENR. 197040 Iwo 15 1SuD iw 1900 EUCTN MALE FEMALE MALE FEMALE MAL FEMALE MALE FEMALE MALE FEMALE TOTALWIAOR fOC on" 631 11179 3 204.3 1234.3 6937.0 19 13 7 764.3 14165.3 66.2 %%_ Ja84 35. 361 35.4 31.4 NOT REPORD 0, 0 °3 6. 6.1 1102 67 15.7 0.0 0.0 % _- Wma0.0 20.9 26.3 24.6 0.C TOTALREPRTD 680".3 5as 111 617.7 117062 6611.7 13062.0 7491. 14165.3 64662 %_1 kw&%.4 36.7 36 36.4 31.4 TOTAL U TA 68162 1777.1 004 SK05.7 9661 806.1 11654 404.9 1304.? 41.9 %fos . 21.3 232 242 20.4 26.0 LrRTE 3740.2 45 26361 35120 =m1 37237 1456 286 1116. 11002 %Sumds 59.0 Ws 64 . 69 UTIEATE NO DPLOMA 14404 42.5 12630 351.5 637.6 802 911.0 692 1192.6 4.8 %fsw'ds 2.0 21. 242 40.1 29.5 PRIMARYSCHOOL 469 11338 004 19t9.1 5.7 20. 71.5 3294.8 519.4 3123.8 %I_Mi 21.1 24.2 25.3 29.5 26.6 JUNIOR HON SCHOOL Ws 51.5 56.6 964 73 106.1 1014.3 156. 1276.7 2123 %S ba t12.0 14.7 I6 1t3.4 14.3 HIGHCHOOL 3447 10. 54 167.6 6412 2e0.4 1201 374.6 1341.5 499.7 _ mm L260 26. 25.0 23 27.1 UNIVERSilY 163.1 50 22".0 51.1 505 16a. 617.4 101.5 716.3 250.4 ~fsnids 23.2 17.6 234 21.5 25.9 PUCEffAG T RISnON OF MALE AND FBEMLE LASOR FORCE BY EDUCATION 197 1975 1960 135 1tWO EDUCAIONYY L L LE ALE MALE FEMALE MALE FEMLE MALE ' FEMLE MALE FEMA TOTAL LASO FORCE 100.0 100.0 100.0 10.O 10O 100.0 10. 100.0 100.0 100.0 NOTREPORTED 0.0 0.0 0.3 0.1 4.7 1.7 6.2 2.0 0.0 0.0 TOTAL REPORE 100.0 100.0 100.O 1OO. 100.0 100.0 100.0 100.0 100.0 100.0 TOrAL UIfrTE 70. 80.0 77.3 42.0 6a5 463 66a 61.5 921 70.7 LLErTE .4 69.4 2.7 81.0 17.5 64.7 112 36.5 7.9 29.3 LITERATENO PLOMA 1L5 7.4 11.3 6. 6.0 4.4 7.0 - .1 8.4 7.7 PRIAIRY SCHOOL 45I 9iO6 53.9 31.0 56.3 32.7 e02 44.0 60.1 48.2 JUNIP4HCHOOL 0. .0 1.6 6.3 1.6 7.6 1 9.0 3.3 HIGH SCHOOL 3.7 19 4* 3.0 72 4.1 92 6.O 9.6 7.7 UNINE I 1.J 0r 2. 0.6 4.7 2.5 4.7 2.3 5.1 3.9 Di - -D-- 40i0 352 372 28.5 214 wEHD( 46.2 45.3 47.0 36.3 9 1h fpg hgo* wdt 91inmuw.D 9* Olnmoli kszWEs eWmt wnd Em1vmmnt ndea. Se tadfrdniUn - 43 - Urban-_fural Diffeeos-19 2 47 An expected, there are signiflcant urban and rural differences in educatLonal attainment of men and women (Table 2 14). xn rural areas, the female shar of the illiterate labor force is 68%, but only 35% in urban areas. The corresponding rural and urban portions for the functLonally illiterate labor forco li 36% and 13%. 2.46 There are also notable urban-rural differences Ln the educational distributions of men and women, whlch can be summarized as follows Rural - pbn Kale Femle Nale Female X of Labor Force that In: illiterate 12 35 4 11 Fuctionally Illiterate 11 9 6 4 Compltd Primry School 63 52 57 34 Total of Above 66 96 67 49 It is surprlsing to note that a signlficantly hlgher proportion of the rural rather than the urban femAle labor forco completed only primary school and that in the urban areas, the female labor force boyond junior high school has a significantly blgher educational level than thelr ma-le counterparts. However, it needs to be remmbered that women account for only 17% of the urban labor forco. 2.49 This apparent declne in gender differences in schooling, however, conceals som important differences that may impact on women's labor market outcomes. irst, the information on educational attainment concerns only people in the labor forcol it thus excludes the large numbers of women who are not in the work force, espcially in urban areas. Unfortunately, except for 1988-1990, thero is no hietorical informatlon on schooling attainment of the economlcally inact$ve population.32 Second, the data on educational attainment are used as eorsats* LidLcatore of labor market qualifications and sklls. A serlous drawback of the schoolLng measure is that lt does not reveal elther the quality of schoollng, the types of courses that men and women take, and their fields of specialization or programs, especlally at the post-primary levels. A related lmportant information gap concerns the nature of formal or informal vocational and technlcal training acqulred by men and women. Here too, there is no detalled information on aspects where there are likely to be substantial gender differences. Wi The Hou"ehold Labor Force surveys do not provLde lnformatLon on. partLcipation rate cross-classlfied by age, marLtal status and educatlonal attainment. - 44 - TABLE 214 DISTRIBUTION OF TOTAL LABOR FORCE, 1990 BY RESIDENCE EDUCATION AND GENDER URBIAN *RA EDUCATION LEVEL MALE FEMALE MALE FEMALE TOTAL LABOR FORCE 7W0.0 1465.6 71052 50.5 % FEMALE 172 41.4 TOTAL LITERATE 6761.0 1304.6 687 3281.4 % FEMALE 162 34.3 LITERATE 299.0 161.1 819.5 1739.2 % FEMALE 35.0 68.0 LUIERATE NO DIPLOMA 391.4 57.1 801.4 442.8 % FEMALE 127 35.6 PRMARY SCHOOL 40492 502.0 44702 2621.8 % FEMALE 11.0 37.0 JUNIOR HIGH SCHOOL 772.3 131.5 425.1 59.3 % FEMALE 14. 12.2 VOCATIONAL JUNIOR HIGH 452 15.3 33.1 62 % FEMALE 25.3 15.7 HIGH SCHOOL Z23 2m5 250.3 81.9 % FEMALE 312 24.7 VOCATIONAL HIGH SCHOOL 342.2 111.7 146.7 33.5 FEMALE 24.6 16s UNNERY 558.4 214.4 157.9 35.9 IS FEMALE 27.7 165 8 :. WM LborF uy (190) s818. PERCENTAGE DITRBUTION OF LABOR FORCE. 190 BY RESIDENCE, EDUCATION AND GENDER UR1AN FRURAL EDUCATION LEVEL MALE FEMALE MALE FEMALE TOTAL LABR FORCE 100.0 100.0 100.0 100.0 TOTAL LITERATE 956 89.0 88.5 65.4 ILITERATE 4.2 11.0 11.5 34.6 LITERATE NO DIPLOMA 5.5 3.9 11.3 8.8 PRIMARY SCHOOL 57.4 34.3 62.9 S22 JUNIOR HIGH SCHOOL 10.9 9.0 6.0 1 2 VOCATIONAL JUNIOR HIGH 0.6 1.0 0.5 0.1 HIH SCHOOL 8.5 18.6 3.5 1.6 VOCATIONAL HIGH SCHOOL 4.8 7.6 2.1 0.7 UNVERSITY 7.9 14.6 2.2 0.7 I1 OIINDE 26.7 23.1 WEINOX WM 44.2 27.1 8oLiowPlnhol GorFoU,e (1990). 1Si. Note: aI b Is DWOls tft ridex: WE b te Wmen end Emplynt Iex. - 45 - Educatlonal Attalnment and Female Labor Force Particetion 2.50 Table 2.15 and Chart 2.15 shows the relatlonship between schoolLng and female labor force particlpation rates. One would expect labor force partlclpation rates to vary more by educatLonal attalnment among women than among men. It has been found ln numerous econometrlc studles of female labor supply, in both developing and industrialized countrles, that schooling Le a consiLtent and among the most effectlve determLiants of female particlpation, but much less so of men. The effects of schooling are reflected ln two general ways. First, since schoollng is an important determinant of market wages and work opportunitles, thls makes lt more attractive for women to perform market rather than nonmarket work. Second, education can affect female labor supply, Lndependent of lnfluences through wages, by breaklng down the noneconomlc conistralnts related to the woman's (as well as her family's and soclety's) attltude. (or so-called ftaste.") toward market work as opposed to housework. Therefore, one would expect that hlgher educatlonal attainment would lnduce greater participation among women. 2.51 As seen in Chart 2.15 the expected relationship between educational attainment and partLeipatLon rates is clearest for women ln urban areas, but the pattern differs ln rural areas. The data for urban areas shows that schoolLng has a powerful effect on the partlcipatlon status of women. The partlcipatlon rate of urban women wlth less than primary schooling is about 8%, but it rLses steadlly and dramatlcally wlth lncreased education - to 11% for those wlth primary education; 16% wlth junLor high school (22% with vocational trainlng)t 39% wlth high school (44% wlth vocatLonal hlgh school), and 78% wlth unlverslty educatLon. It is clear that urban women who have more schoollng are much more Ilkely to be ln the labor market than their less-educated counterparts. The proportion of women in the working age populatlon (12 years and over) in urban areas wlth leso than secondary schoolLng, however, is large - almost 80% (23% are illiterate, 7% are functionally illiterate, and 48% have completed primary school). 2.52 In rural areas, the relationship between education and partLiLpation is dLfferent. PartLcipatLon rates are relatively high among illiterate women (47%), literate women wlth incomplete schooling (48%), and among primary school graduates (57%). The proportlon of women in the labor force in rural areas with post-primary educatlon iL very amall - les than 5% - but their partilcipation rates are higher than those of their urban counterparts. TAE 2.15 FEMALE PARMCIPAT10N RATES BY EDUCATIONAL ATFTAhENT. 1990 ______ _TOTAL WN I RUAML EDUCATIONAL ATTADIMEN7 NOTU -- NIW N -W- POP U. tAB FM IA fOA l- U. L IAIO fO P 12 .-+ *# LM UIWIf 000_ %V0 % VW0 s PRS%VW % mm0 %oo % PRS%00 VW % S I0L % PR% 0 ILLITERATE 5919 305 1900 29. 4019 31.0 321 217 227 161 tl.0 248 7.4 3 3 .1 t739 34.6 I9 41.7 4_6 1 IJTERATE NO DILOMA 636 8.4 50 7.7 11U GA 30" 703 7.3 57 39 646 7* &1 03 9.5 443 6A 490 10.3 475 2 P ARYf SCH40L 97 47)3 124 462 6103 47.1 3 49 476 92 34 4100 50.2 IDS*43 2 24 2 032003 419 56.7 3 JUOS10HI6HCHIOOL 1090 5.6 191 2.9 0S 6.9 17. 842 6. 132 0 710 6. 1& 5 . 246 S5 59 12 16B 3* 239 4 VOCATIONAI JUIORN IGH 103 05 21 0.3 81 0.6 2 71 0.7 15 1. 55 0.7 215 32 03 6 0.5 25 0. t9.5 S l10SCHOOL 653 4.4 354 55 49 3.9 415 M 7.3 273 I" 430 5.3 368 150 15 E2 16 6B 1.4 5s 6 VOCATIO FSHI#4SCH0OOL 294 15 145 2.2 140 12 4A 252 2. 12 7? 140 1.7 44.4 43 0.4 34 07 9 Q2 6 7 UINER1TY 313 I. 20 3.9 62 05 60. 274 2 214 14*5 0.? 7.a 3S 0.4 33 07 3 Ql 24 ALL LEVELS 19435 100.0 6486 1 12906 1000 33.4 93 100.0 146 108.0 6168 100.0, 152 9t01 1004 51 10Q0. 4781 100 512 Smw.: thwwI tLar wt19". Nose: 1m _m nl add * bcreurm; PR- Puco An CHART 2.15 FEMALE PARTICIPATION RATES I BY EDUCATIONAL ATAINMENT, 990 aD) 40- DA 2 A e Tbe EDUCATIONAL ATTAINMENT (See Table 2.15) - 47 - H. Male-Fema Ragnings Diffen irane _Coverage 2.53 The gender gap in earnLngs li an lmportant statistic relevant to the economlc status of womon. However, analysis of the gender gap in Turkey is not feasible with avallable data for a number of reasons. First, the proportlon of wage-earning women in Turkey is smalls natlonwide women comprlse 16% of total wage earners and 19% of workLag women are wage earners. Since only 4% of rural working women but 57% of urban worklng women are employees, the gender wage gap issue is essentially one of relevance in urban areas. Second, avilable wage data from the social Insurance InstLtution (SIX) relate only to workers who are covered by social insurance schemes; in 1990, 22% of the male labor force and 6% of the female work force was covered by the SIX. Thus, available data not only fail to support vlgorous analyels of the gender wage differential but highlight the fact that the vast majorLty of working women do not enjoy the benefits of social Lnsurance and their representation ln the system is small. Namely, of wage earners, almost one-half of men were covered compared to just over 30% of women. The gender compositLon of social lnsurance reclpients is similarly skewed: in 1990 women comprised only 10 percent of all workers covered by social insurance. 2.54 The available wage data from SIX suggest, furthermore, that the gender gap in earnings is a non-issue - women's average (daily) earnings are 85% of those of men. Thli magnitude exceeds that found in all OECD countries (where women earn about 20-40% lses than men). Statistlcally, the comparison is also spurious since lt compares the average earnings of a select group of 3.07 million men and 377,000 women, nor are the data callbrated for similarity of tasks or years of experLiece. Although not rLgorous, two data sets are lndicative of the need for further analysis. FLrst, data on agrlcultural day-laborers wages from the Provincial Agricultural Dlrectorates for various provinces in 1990 (see Annex 5, Table 13), suggest a wage differential ln the range of 40%. Second, using SII data, the ratio of female/male wages for various manufacturing subsectors for 1983 and 1988 show that women on average earn 20% to 30% less than men (Annex 4, Table 14). Simultaneously, the need for further research on the gender wage issue and the impact of adjustment is underscored by the marked decline in the ratio of female/male wages (averaging 25-30%) for all manufacturing subsectors (except leather goods and fabricated metals) revealed by the SIX data for 1983 and 1988. EarnLngs data for wage and self-employed workers are collected in the Household Labor Force Surveys; whlle these data would underpin the necessary analysis, to date, they have not been released. More attention in terms of data collection, analysis, and policy inltiatives must be given to this question because a higher proportion of the female work force will engage in paid market work activities during the process of urbanization and industrialization. Finally, it should be noted that the lssue of pay equlty cannot be separated from that of gender segregation of the labor market. - 48 - I. Policy lmolications 2.55 Based on the patterns and findings concerning the labor force activlties of Turkish women, thle section proposes proactive pollcy responses that would enhance the ability of women to become effective agents in Turkey's development. The economic transformation of Turkey will likely contlnue to be reflected in rapidly falling rural and slowly rLaing urban participation rates of women during the 1990. Whether or not, and how quickly, a U-ahaped relationship materialLes depends largely on two considerations: the availability of employment opportunlties for women and the supply of skilled and qualifled women who could take advantage of these opport-nities. 2.56 An important - if not the key element - in education, which is lnterpreted here to include all forms of schoollng: formal and informal, adult education, vocational and on-the-job training. Raising women's educational attainments or earnings capacity - especially at the post-prlmary levels and ln programs, courses, and fields of study that would meet the skill requirements of industrialization - will improve both economic efficiency and women"s economlc status. Increased education will raise their participation rates, reduce the likelihood of unemployment and increase continuity in their labor force attachment. Moreover, changing the patterns of women's education will likely reduce the gender wage gap and segregation in the work place, as well as provide young girls and boys - and adults - wlth different attitudes towards the gender division of labor. The oollcv response for integratina women ln the develonment ocean need- therefore to focus on broadenina women's skll range. 2.57 Much of the observed gender disparity in participation rates and earnings, and the persietence of segregation in the work force can be attributed to gender differences in education and training. Futher narrowing these differences wlll impact favorably on the status of women and thereby, Turkey's overall development. If educational policies are to be successful in removing gender disparities in the work place, they must be comolemonted by labor market- related initlatives. In other words, changing educational policies is a necessary but not sufficlent condition for improving the status of women. To this end, there are several policy measures that could be pursued. 2.58 For lnstance, oroarees towards eliminatina gender segreaation can be made by the Government itself setting an example to ensure symbollcally and substantively, that gender segregatlon is not perpetuated. The imolementation of emnlovment eguWt& lnltiatives, along the lines pursued in Canada (see Box 1 below), would represent an important step in facilitating the equal accees of women to employment opportunitles that have traditlonally been dominated by men, as well as the newly-emerging work prospects that will become available during the continuing transformation of the economy. Action in this area, would need to be coordinated with other initiatives. - 49 - 2.59 Some of the existlng urotective leaLolatLon whlch was designed to ensure that women are not exposed to hazardous physical or moral conditions in the workplace, precludes their employment of women ln jobs that are deemed OunsuLtable" for them. Consideration needs to be glven to repeallng these measures as they have the unintended effect of restrLting women's job opportunltles. As evldenced by the negligible female representatLon ln industrLes including heavy industry, mlning, constructlon, utilitles and transportatlon (Table 2.6), these measures contrLbute to gender segregatlon. Thie logislatlon also lnfluences educatlonal choicos of women and dLscourages them from even entering particular occupatLons. The experience of several countrles has shown that the repeal of thls legislatLon is unllkely to result in a massive and immedLate influx of female truck drlvers, construction workers, heavy equipment operators and miners. Rescinding the laws will, however, play a sLgnLficant role as a slgnal in changing the attitudes of men and women towards suitable" women's work; Lt will also influence young women's educatlonal cholies for their future careers. The negative sociologLcal and psychological impacts of protectlve legislatLon should not be underestimated. 2.60 There are also constralnts emanatina from the labor market that make it dlfflcult for women to seek market work. Women are hampered in their market work opportunlties by labor demand constralnts and by employment costs. Labor demand constraints, whlch generally arlse from labor standards leglslation and collectlve agreements, are reflected ln minimum working hours and by rigid work schedules. These constraints compound the difficulty of combinlng market work with household responslbilities. Encouraging (or mandatlng) employers (especLally in the industrial and servlces sectors) to provlde more flexlble time arrangements by modlfying labor laws governing working hours would be a useful policy measure. Women's employment opportunlties are further lnfluenced by the costs of employment - the location of the jobs reflected in commute time and expenditures on items such as clothing, childcare and meals. Whlle the necessary data on Turkey are not available, the informatlon from other countrles clearly show that Oflxed costs" of employment are in practice very significant for women (especially among wives and female-headed households) and play a prime role ln influenclng the entry of women lnto the labor force and the types of jobs they will seek. Enforcement of existing regulatlons regarding enterprise childcare facilities and encouragement to establish community or family childcare facilitles would be beneflcial (see also para. 3.41). 2.61 The vast majority of women ln the labor force - 94% have no social insurance coverage. As a first step, a target could be set to equalLze the proportlon of worklng women covered by social insurance wlth that of worklng men (22%). The inferior access of women, especially unpaid family workers, to social securlty rlghts suggests a vlolation of Article 60 of the Constltutlon of Turkey ("Everybody has the rlght to Social Security"). As a first step, and in light of the need to re-vamp the exliting insurance system, a review of the present system and its future adequacy should be undertaken, glvlng speciflc conesderation to means of increaslng women's coverage wlth the objectlve of equalizLng their coverage with that of men within a specifled period. - 50 - 2.62 A further. and highiv sianificant area of nolicv initiatives to increase the visibility of woen concerns the collection and nublication of data. A comprehensive and up-to-date data base containing socioeconomic lndicators on the role and situation of women vie-&-vie men is a sine aua non for analyzing, evaluating, and improving women"s relative economic status and for monitoring the impacts of female-oriented and structural adjustment initiatives. Consideration should therefore be given to introducing the following measures: Publication of an Annual Statistical Report on the Socioeconomic Status of Women X This would be a useful and inexpensive means, inter alia, of reducing female invisibiiity. The report, which could follow TLo, OECD, and/or UN standards, would collate and summarize existing information on socioeconomic indicators that are already avallable. The Canadian approach is illustrative. Each year, since the late 1950e, the Canadian Women's Bureau has published a "Women ln the Labour Force" report, whlch contains contemporary and hLitorical gender-disaggregated information on labor force activities, earnlngs and income, educatlon, benefits and unionization. In additlon, StatLstics Canada (the Federal government statistical agency) publishes a document called "Women in Canada", which, in addition to labor market aspects, also provides information on health, education, the family, housing, the legal system and minority groups. 8I has recently announced its plans to publish a document tltled "Women in Statistics" in October 1992; the frequency of publication will be decic-d by demand. Undertakina Policy-Relevant research on women's labor force behavior either within SIS or by outside researchers. There are numerous such studies for various developing countries and at least one econometric study of female labor supply for each OECD country, excepting Portugal and Turkey).VW Simultaneously, analysis of gender dlfferences in wages, schooling attainment, poverty, unemployment and the impact of migration and legislation could usefully be undertaken. 2.63 The implementation of the revised nationwide Household Labor Force surveys beginning in 1988 represents a significant improvement over the Census in providing labor market information, especlally in reducing the statistical invisibility of women's work. The continued implementation of the surveys should be strongly encouraged. However, Lnsufficient published data are currently made available to permit a thorough assessment of the role and contributions of women in the Turkish economy. In particular, data on gender-disaggregated average earnings, family characteristics and broken down by the major metropolitan areas should be made available. W For collections and summaries of recent studLes on female labor force behavior in OECD and other industrLalized countries (the Soviet Union, Israel, Sweden) see: Layard and Mlncer (1985), Killingsworth and Heckman (1986), and Moffitt (1990). As regards developing countrles, recent studies include those by Arriagada (1990) on Peru; Behrman and Wolfe (1984) on Nlcaragua; House (1985) on Cyprus; Khandker and King (1990) on Peru; Mackinnon-Scott (1991a, 1991b) on Bolivia and JamaLca; Mohan (1986) on Colombia; Ng (1990) on Argentina, Terrell (1989) on Guatemala, and Steloner, Smlth, Breslaw# and Monette (1991) on Brazil. - 51 - 2.64 Modifications of, and addltions to, the well-designed labor force survey or the planned Income and Expenditure Survey could provlde valuable pollcy-relevant informatlon at low incremental cost. Speclflcally, the addltion of lnformatlon on time use by household members (see Box 2 below); retrospective informatlon on lifetime and, recent (5 years) market work history; data on inter- labor force turnover (weeks worked per year); data on spells and duration of unemployment and wlthdrawal from the labor force and reasons (e.g. pregnancy, return to school, husband does not want me to work); housing characteriatics (e.g. water supply, electrlcity, fuel and major household appliances); education (e.g. fleld of study at the post-prlmary level, reasons for leavlng school); fertility (e.g. how many chlldren have you given blrth to, do you plan to have any more children, how many, use of birth control); mlgratlon status and reasons; and soclal security coverage and unLonLzatLon by gender. The benefits of a more detailed annual survey along these lines would readlly exceed those of the current blannual surveys, while the costs would likely be of the same magnitude. These changes would support analysis and research into the consequences of economic change, and the character and effectLveness of pollcy interventlons, ln additlon to issues relatlng to women's fuller LntegratLon lnto the economy. 2.65 In addition to family responsibilitLes, women ln the labor force work long hours - frequently, unpald. In rural areas, household responsibilltLes, may be shared by other female famlly members. However, thls is less llkely to be the case ln urban areas where familles are smaller and female school enrollment rates are hlgher. Women, regardless of educational attalnment, face the lssue of time allocatlon between market work and household responslbillties. If they perform market work, they carry a double burden. As ln the OECD countries, and notwithstanding improvements in the technology of housework, their education, or even household income, women's responslbililtles for malntalnlng and managlng the household are unlikely to change. The social conventlon that women are responsible for household work hinders their labor market activity. changing socLal conventions and ingrained attitudes among men and women is not an easy task. However, pol$cy resoonses ln the form of "moral suaslon" are 2o2sLble. For example, through promotlonal and publiclty campalgn. (e.g. TV programs, documentarles and medla advertlsing by government) showing men helplng in household chores, women working ln "male-dominated" occupations, and publicizing the dllemmas that worklng women face ln juggling thelr time. While the lmpact of thls publicity will probably be greatest on younger people, a message wl11 also be conveyed to employers of large numbers of women. 2.66 Another set of nonmarket pollcies that would reduce the double load of working women concerns imorovements in llv$na conditions, especially among the urban poor and ln the South-Eastern provlnces. Increased provislon of amenities such as water, sewerage, heating and cooking fuels would ease the burden of housework. Also, improvements in health services, especially for chidren, and the availability of childcare facilities would make market and nonmarket work more compatible. - 52 - membears of thef Eu oropa Couneil spento ctiviie thteneK)adoted Thet earmnin tyn Chenearter of~~eore tho Fnaetal S~oia~ Raightis of Workers (the wSoe,ias Cohkactivter" (bTheI tharo mandatesol the ComsLno the Eurpa iemmeratied ~Eto phrodc 'iank atonk progrhi*am and egularle~y reporthon thiplnemenitiatio (ohf thre arndioles ork adithe ChorknthefrmA Lmportan.t feture oŁ thf ChmrterpVeeqnt i8 pd so˘tio 6onttlono ooen Euae Therteatment for Msen an oe wpiehf sta4~tnes .d~eof thats.lg~.5~ aesn. ae tiona. 1~, h.t4 shoul tba wnesfe omensure@ impifecantacnrbtion tf the prle$plbeof eulty5~fsis b tween men and somiety asc fregardsn.cn pateulart aeceit emlomn.te iets, remunert ionan,e wourceng eond$to lounf s, foeLal proter doltion, edetLn voaioa triigadeareril dsvelpet NoW1roeasurens shoduld$t alsoaflctonpot~.o4 be doveloped eynabling mnadwenthe feolncil thpei occqupationalo h and amiyst hobl anyon "OFSdyuuulydvt o nm ae okn ils hl ae - 53 - The EC is now finalizing a Social Charter Action Program which contains proactive policy recommendations that are particularly relevant to the status of women. In addition to fostering the enforcement of the principle of pay equity between men and women, several other women-related measures of the Program deserve mentlon. First, initiatives must be developed to improve the training, employment and career development needs of women by providing full access to vocational training and retraining on a strictly non-discriminatory basis. Second, gender equality in matters of soclal security must be ensured in all work activities, regardless of the size of the enterprise, worker status, or economic sector of activity. Third, to reconcile market work and family obligations and recognizing the growing importance of working women with children, the Action Program advocates improving the quantity and quality of childcare services for parents who are employed or in education or training programs. It is also recommended that the member states take measures to enable parents to have access to the childcare services by instituting flexible work schedules and leave arrangements for employed parents, and to encourage the equal sharing of household responsibilities between .aen and women. 2.68 Turkey has ratified the Social Charter (with some reservations) as well as the UN Convention for the Elimination of All Discrimination Against Women (CEDAW) and numerous international labor conventions. Ratification, as seen from the statistics cited in this report (Chapter V below), does not necessarily and readily translate into implementation and enforcement. Nonetheless, Government's ratification of the social Charter and the anticipated adoption by the SC in the near future of the Action Program, will contribute positively to an environment supportive to women's further integration. - 54 - CHAPTER III: WoME In THU ECONOMY A. IZntrdiction 3.01 While Chapter II above focussed on the issues relating to women's particLpation arising from the official labor force statistics over the past four decades, this chapter goes beyond the data to address the issues and constraints currently faced by women in the major sectors, including women working in the informal sector - an estimated 0.5-1 million women - who do not appear in the official labor force statistics. This chapter first describes women's contribution to the agricultural, industrial and service sectors and then addresses, inter alia, the structural and instLtutional constraints to women's productivity, their access to resources and inputs, includlng health, education and training. Alleviation of these constraints would simultaneously increase economic efficiency, promote development and improve the relative status of women. Furthermore, enhancing the participation of women in economic development has intergenerational implications since the greater economic independence of women is shown to be positlvely related to lmprovements in chlldren's health, nutrition and education. S. Women's Work and Emoloyment Patterns 3.02 According to the most recently published Labor Force Survey (April 1990) there are approximately 6.5 million women in the labor force and 14.2 million men - a femAle labor force participation rate (FLFPR) of 33% compared to 74% for males (Table 2.1). As noted in Chapter I above, there are marked differences in particlpation by residence: the participation rates for women and men in rural areas a-e 51% and 78% respectively, and in urban areas, 15% and 70% respectively (Table 2.7). aiven the urban migration since the 1950's and the economlc standlng of migrant families, the contlnuing low urban female participation rate is surprising. This is attributed, at least partially, to the omlision from official labor statlstice of informal sector workers, as well as more limited opportunitLes for women's employment in urban areas and social constraints to their employment. As seen in the educational attainment levels (Chapter U), women's limlted skills and training relative to men may also make it more difficult for them to obtain employment in the urban formal sector. These dlfferentlal urban-rural partlcipatlon patterns underline the need to address separately the issues facing women in the urban and rural areas. In thli chapter, the Lndustry and services sector and agriculture are used as proxies reflecting the participation patterns in these areas. 3 03 Tab.e 3 1 below, which summarizes the sectoral employment patterns for men and women, shows that a majority of women ln the labor force are concentrated in the agrlcultural sector, while men are more evonly dlstrlbuted across economlc sectors. Also, as noted ln Chapter II, 68% of the women in the labor force in 1990 were unpald family workers and 25% were wage earners or self-employed, compared to 14% of men who were unpald and 82% who were wage earners or self-employed (Table 2.4). - 55 - IhJILLA: Distribution of Labor Force by Economic Sector, 1990 Total Vomen Hen Fales Hales Foulas Hales ( of ( of ( of ( of Fumlo Male Total Total ('OOOs) Labor Labor Labor Labor Force) Force) Force) Foare) Agriculture 9.666 4,802 4,864 74 34 50 50 Industry 3,829 455 3374 7 24 12 es Services 6.128 768 5360 12 38 13 87 Unknown 1.027 460 567 7 4 45 55 Total 20,650 6.485 14,165 100 100 31 69 Source: Table 2.6. Note: Industry includes manufacturing, mining, construction and utilities. Services Includes trade, transportation, businss services and other services. Unknown refers to first time job-seekers. (S) The AarcLultural sector 3.04 The agricultural sector employs almost three quarters of the female labor force and, of the total agrlcultural labor force, one half La female (Table 3.1). It is also significant that of all women worklng ln agrlculture, almost two-thLrds work 40 hours or more and 20% work 60 hours or m6ore1, but that only an estimated 3% of women in agricultur, receove a wage, 92% are unpald family helpers and St are self-amployed.I/ Notwithstanding the involvement of women ln this setor, their role is not addressed in available data, studies or programs relating to the agricultural sector. Thus, Annex 5 represents probably the firat attempt ln 3nglish to analyze women's involvement in agrLculture. In the absence of concrete lnformation, the analyLsi of women in agrlculture in this report is based on mocro-level data and case studles, prlmarlly sociological and anthropologlial. 3.05 Although women are engaged in almost all actLvltLes ln agriculture, the extent of thelr partlcipatLon depends to a large extent on the lncome and size of the hometead, the topography of the region and the extent of mechanlsatLon (Annex 5). In general, womens partLclpatLon is hlgher in lower income and small farm households, ln hllly or mountainous. areas than in the plains,&V and more in crops/activities that are leso see 8Oouehold Labor Force Survey. S18, Aprll, 1990. Table 12. ZV IbLd, Table 11. ] Thli li because hand harvestLag and hoeLng, primarily female tasks, are required in these areas slnce the topography li not sulted to mechanization. Also, livestock productlon, whlch is female-Labor intensive, Ls important ln thes reglons. - 56 - mechanized.5 Women's partlclpatlon li limited ln commercial crops (such am cereals) whore most operatlons are mechanized and greatest in self-consumption crops. (partlcularly, vegetables, frult, pulses and small llvestock). Actlvitles that are female-lntenslve lnclude hoeing, weedlng and hand harvestLngVt male-intensive activlties lnclude prunlng, graftlng, and mechanLzed operatlonsM (such as soil preparation, seedlng, fertliLzatLon, chemical spraying, LrrlgatLng and harvestLng) and marketlng. With respect to livestock, men focus on field actlvltles and marketing, whleo women are responslble for the remalnlng tasks (mlkLng, feodlng and cleanLng the stalls and animals). For homestead poultry production, women carry out most tasks but for market-orlented, larger-scale operations men assume overall responaLbLilty. Women are also Lnvolved to a large extent in serlculture, and increaeLngly ln aplculture. 3.06 Far those women employed ln the agrlcultural sector who receive a wage, there is llttle information on wages and even less on gender dLfferentials. Untll 1989, the mLntmum wage for agrLcultural workers was set separately from other workers and was lower. The only regular source of wage data is the Social Insurance InstLtutLon (SIX) . Socal securlty was extended in 1983 to self-employed agricultural workers (Law No. 2926) and unpald family workers (Law No. 2925). The coverage of women, however, is limlteds ln 1990, whlle women constituted S0% of the agrlcultural labor force, they constituted 9% of insured workers in the agrlculture, animal husbandry and forestry sub- sectors and lese than 1% ln fLihery. Whlle the sample populatlon li limited, wage data for insured workers show a narrow wage dlfferentlal: female wages were 80% of male wages ln agrLculture and animal husbandry, 89% ln forestry and 73% in fisheries (Annex 5, Table 14). Recent data on dally wages of male and female agrLiultural workers in 18 provLnces collected by the Provlncial AgrLcultural Directorates suggest a wage dLfferentLal of 40% (Annex 5. Table 13). 3.07 Notwithstanding the importance of women to the agrlcultural sector, women do not enjoy ready access to agrLiultural resources and support servLces such as extension and trainLng, lnformatlon, credlt or approprlate technology. Wlth respect to the extenslon and tralnlng system, women are constrained since the vlllage extensLon agents who work with the farmers dlrectly are invariably male whlle the home economists, who also work in the vlllages, are all female. This occupational segregation is, ln large part, the result of the trainLng system whereby village extenslon tralnlng programs are glven in the Soys VocatLonal Schools (dormitory programs) and the Home TI The regressLon results in a study of 290 women in 32 village. of Erzurum also support thli, Brturk (1988), Women's Partlcipation ln Agrlculture ln the Villages of Brzurum, Turkey", PAO. p. 31. 2t Threshing, wlnnowlng, crop processLng, cleanlng, drying, storage. h There li some evLdence of a change in this practice. Women ln many regLons of Turkey are actlvely enrolllng ln tralnlng courses on mechanizatlon and tractor drlving organLied by the MARA (Annex 5). - 57 - Economics programs Ln the Girls Vocatlonal Schools.Z Whlle the male exteneion agents are undoubtedly willing to work with women, they are required to contact a partLcular number of farmer., who are defLned as the tltled owner. Thus advLiing women ln thelr farm actlvltles would be in additlon to the agent's routine work load. Simultaneously, while there are women who can talk freely wlth the extension agents in the absence of a male relative, thLi is the exception rather than the rule. The curricula of extensLon programs and research undertaken also reveals a blas towards comercLal crops, which tend to be male-intensLve actlvltles. Village extenslon meetLngs are also attended primarily by men, thus women rely on theLr menfolks for dissemination of information. Xn fact, the case studies are replete wlth Lnstances of extensLon agents provLding traLning to men for women's actlvltles. In terms of access to credit, the results of the credlt questLonnaire submitted to Zlraat Bank regional branches (Section C below and Annex 6) suggests that women recelve little agrLcultural credlt. SLmilarly, technology improvements ln the agricultural sector have tended to beneflt commercLal crops with only limited development of technology for female-Lntenslve activitLes. In light of the Lncreaeing educational attalnment of rural women over the past decade, whlch may eerlously affect female participation in agriculture in the coming decade, consideratLon need to be given not only to the likely availability of female labor in the sector over the decade, but to the consequent need for technology, extension servLces, credit and research. (LL) The Industrial 8ector 3.08 According to the 1990 Labor Force Survey, 7% of employed women worked iA the LndustrLal sector and they constituted 12% of the LndustrLal labor force (Table 3.1). The differentLal employment growth by gender Ln the industrLal sector between 1970 and 1990 is sLgnLficants the female industrial labor force grew by 50% (from 304,000 to 455,000) while the male Lndustrial labor force more than doubled (from 1.5 million to 3.4 million). It L also important to note that three-quarters of the women employed in the lndustrial sector live Ln urban areas and that over three-quarters are wage-earners (compared to 3% ln agriculture), 9g are self-employed, less than 10 are employers (compared to 4% of men) and 12% are unpaLd family workers. For the industrial sector - and the service sector - it li also important to know whether the person works ln a formal or informal enterprise and the locatLon of employment - homebased, workshop or formal enterprlse.IY Based on LFS defLnitLons, approximately 53% of all women working ln rural industrLes worked I At a higher level in the ProvincLal DLrectorates, extenaLon workers (includLng home economLets) are unLversity graduates and gender segregatLon of trainLng is not a factor; nonetheless, at thle level, women are rarely found worklng ln the extension services other than as home economLits. IV The 1990 LWS defines an informal enterprise as one employLng less than flve people. However, in practice, an informal enterprLse is dLstLnguLshed rather by its observatLon of labor regulatLons (for instance, minimum wage, payment of overtime, job securLty, working condltLons). - 58 - ln liforoml enterples and 24% ln urban lndustries.2V Of tho women workLng ln laformal entrprlse ln both urban and rural areas, more-than three- quartrs amr hombasd. The data on lnformal workers in both the lndustry and servlce sectors, however, ar consldered to be a srlous under-estlmate of the number of people workLng, and partLoularly of womn. 3.09 The lndustrlal sector conslits of four subuectors: manufacturlng, minLzg and quarryLng, construction, and gas and electricity. In term of female employment, the industrial sector io vlrtually synonymous with the manufaoturLng lndustry since the 1950 some 97% of all women ln the lndusttrLl sector have been employed in manufacturLng (Annex 4) .V WLthin the manufacturLag sector, there is an extremely hlgh concentration of women ln the textiles (66%) and food (14%) lndustrLes, followed by machinery and equlpment (7%) (Annex 4, Table 4). The first censuses (1913 and 1915) show that this concentration is a longstandLng one, reflecting, to a large degree, the employment of women in activltLes that are an extension of thelr traditional household dutles. 3.10 As regards to occupational status, the majorLty (80%) of employees in manufacturlng - both male and femAle - are in the 'productLon and related workersw category. Bmployment of both men and women in thLs category has fallen consistently sLnce 1950, when approximately 95% of employees were in 'production". The greatest growth of female employment ln the manufacturlng sector has taken place ln the "OscLntlflc, techniLal and profeseLonal", clerical and servlce workers categorles,W where"a for men the employment growth has been concentrated in 'administrative and managerial" and servLce workers. The gender dlfferentlal ls greatest ln the 'admlinitrative and managerial' category - and the dlfferentlal appears to be lncreasing (Annex 4, Table 5). WLthin the "productlon and related workers' group lt is important to note that women re concentrated ln four subgroups: 45% work as splnners, weavers, knltters, dyers and related workerol 25% work as taLlors, dressmakrs, seamstresses, upholsterer r 6% work as food and beverage processors and 5% in tobacco proceosang. 3.11 Data sumearLsang amployment status by gender in the manufacturLng sector show a significant increase ln the proportLon of fmale wage earners Ln the manufacturLng sector between 1970 and 1965W - from 40% to 79% and a steady decllne in the proportlon of unpald famLly workers from 18% ln 1970 to 2V Labor Force Survey, 8I8, April 1990, Table 47. W The data on the manufacturLng sector here are based primarLly on the 1985 Populatlon Census rather than the Labor Porce Surveys since for dLiaggregation at the sub-sectoral level the sample size is not sufficient (Annex 3). ly Untll 1965, clerical workers wer lncluded ln the admlniLtratlve category. W There are slgnlflcant dLsocontlnuLtLes wlthln the Census data for the manufacturLng sector (notably for 19SS, 1960 and 1970) and between the Census and the Labor Force Survey data. - 59 - 12% in 1985 (Annex 4, Table 7). These trends are much stronger than for the industry sector as a whole (see Chapter U1). The trends in female employment in the manufacturing sector suggests that while their employment growth was not rapid, that the sector has created significant job opportunities since 1970. The trends also suggest that to the extent that employment opportunities develop for women, that women take the opportunities regardless of factors such as perceived inferior status (for instance, of factory work) or social pressures. The data also supports trends in other countries - that initially, the majority of women work as unpaid workers, asslsting their male relatives, mainly due to the lack of job opportunities in modern organized sectors. When opportunities do emerge, they prefer working in the factories rather than in homebased industries. Working outside of the home corresponds with an improvement in their status within the family. 3.12 Tbe decline in the proportion of self-employed women from just over 40% in 1970 to 9% in 1985 deserves elaboration. While the fluctuations undoubtedly reflect changing statistical definitions, their decline is, at least, partially real. For instance, looking at sub-sectors, the proportion of tailors (seamstresses) as a subgroup of self-employed women was 82% in 1970 but fell to 32% in 1985. As ready-made clothing dominates the market, fewer people continue to wear tailored clothes and this inevitably reduced tailors' lLvelihoods. A continuing decline in this category is expected as ready-made clothLng increases its share of the clothing market. The share of women who are classified aa "employers" is very small and has remained stable in the Census data whereas the male share has grown steadily. This reflects in part the prevailing view that it is inappropriate for a woman to run a workplace by herself, given the competitive nature of the manufacturing industry, the type of work involved and the conditions of workplaces. 3.13 The female labor force in the industrial sector is relatively young. The highest activity ages of women are between 15-24 years (almost 40% of the sectoral labor force) and if the 12-14 age group is also included, women under the age of 25 comprise 50% of the female manufacturing labor force (Annex 4, Table 10). The male participation rates for the 15-24 years group is 28% and for the 12-24 age group, 31%. The difference in gender participation rates at these ages reflects very largely parents' differential schooling preferences for sons and daughters. After 25, female partielpation rates decline consistently with age, reflecting both their roles as wives and mothers but also, a marked preference of employers for younger women, who are considered to be both more productive and obedient. Frequently, an age limit of 25 is set for applicants. 3.14 The educational attainments of the female labor force in the manufacturing sector has increased considerably in recent years - the percentage of "illiterate" and "literate without a diploma" women fell from 45% ln 1970 to 16% in 19851 while the correspondLng decline for men was from 20% to 7% (Annex 4, Table 11). Although considerable progress has been realized in narrowing the educatlonal gender gap, the female labor force remains at a dLsadvantage. Data on the source of tralnina of the skilled labor force also underscores the importance of schooling in the emplovment of women - 45% of sklled women received their trainina at school comp.red to 18% of men (Annex 4, Table 12). It is also significant that 59% of men compared to 34% of women received their training on the job. - 60 - 3.15 Hlnimum wages are set at a mLnimum every two years by the Minimum Wage Board ln the MlniLtry of Labor. In July 1991, the mLnimum monthly wage for workers aged 16 and over was set at TL801,000 (US$195), or ln net terms about TL502,911 (US$123). Data on wages whlch would permit analyaLs of the gender differential are not published. The only regular source of wage lnformatLon ln Turkey is the Social Insurance Institution, (SIX) but these data need to be treated cautlously slnce only 6% of the female labor force li covered by 8II and the sample size in the sub-sectors can be extremely small. Nonetheless, the ratio of men's and women's wages usLng SIX data for different sub-branches of the manufacturlng industry in 1983 and 1988 show that women on average earn 20 to 30% lesa than men (Annex 4, Table 14). The 1988 data suggest that women do relatLvely less well in tobacco, textlles, clothlng - all of whlch are relatively female-dominated sectors - and in, petroleum, and transportation and communication. They recelve the highest relatlve pay ln leather manufacturLng, furniture and fabrlcated metals - which are relatLvely male-dominated sectors. Perhaps of greater Lnterest, li the delilne ln the female/male wage ratio for almost all sectors - except leather manufacturing, furniture and fabricated metals between 1983 and 1988. 3.16 According to the Turkish Labor Code, the offLcLal working week ln factorLes is forty-five hours, although there are exceptlons. Durlng the peak seasons, factorles may resort to overtime work and employees are not always in a positLon to refuse working. Small flrms which are dlfflcult to regulate, when extending the working week, may or may not pay their workers for overtime work. In contrast, larger firms usually comply wlth regulations and pay overtime. The practlce of shlft work also depends on firm size and the nature of the business. During the 1980s, wlth lncreased productlon stemming to a large extent from increased capaclty utliLzation, shift work increased. Both the legal environment and tradltLon work to minimize female shift work, thus closLng off many sectors to thelr employment. 3.17 Turning to the lnformal sector, it should be noted that the workshop is most usually found in the textile and leather garment industries, where sub-contracting by large enterprises is prevalent, and in carpet- weaving. Workshops permit larjer flrms to reduce costs, circumvent regulations and to recrult women who may have dlfflculty working outside of their community. Workshops may be family workshops, in which case the women may not be paid slnce their work is part of the family effort, or commercial, where wages would be paid. 3.18 Although technically homebased 2ro˘tciwLon lncludes self- employment, lt le used ln Turkey primarlly to refer to the putting-out system, whereby agents (middlemen - who are often relatives) supply raw materials to the producers, mostly women, who are paid on a piece-rate basie. The putting- out systems covers a wide range of activities in Turkey: carpet and rug weaving, handloom weaving of local textiles, lace making, embroidery, crocheting, knitting, tailoring, souvenirs and food production. The system is advantageous to the agents who have mlnimal overhead costs and no capital expenditures. From the point of vlow of wom3n, homebased work provldes flexibliLty, does not interfere wlth their household responeiblilties, and provides an opportunlty to earn - although wages are low. For Lnstance in Ankara province in mld-1991, women were found to recelve TL5,000 ($1.25) for - 61 - knltting an Lntrlcate sweater - a mLnimum of 20 hours work. Simllarly, ln Konya ln 1991 women weavLng carpets for a cooperative (without mLddlemen) recelved approximately TL270,000 ($65) for a carpet whlch takeo about one month worklng 50-60 hours a week. The reasons for the low earnLngs of women ln the Lnformal sector reflect both the nature of thelr work and status: specifically, their productivity li low (Ayata, 1987)1 there are no alternatlve sources of employment regulatlons as to minimum wages and workLng conditions are dlffleult to enforce; and, reflectlng the patrlarchal structure of the household, the male members of the household generally set or negotLate the rates wlth the mlddlemen or employers and recelve the earnlngs. Whether they work at home or in a workshop, the wage earned by women tends to be seen as an addltLonal income supplementlng that of thelr husbands or fathers; for example, Berlk (1987) found that the male relatives received the earnlngs dlrectly in 70% of cases. 3.19 To summarime, whlle only 7% of women ln the labor force are employed ln the LndustrLal sector, the analysls shows a high degree of job segregation - 80% of women work in the textles and food lndustrles and of these, 80% are in the lower occupatlonal (production) ranks. The data also show that the rate of growth of women ln the LndustrLal sector sLnce 1i70 has lagged behlnd that of men, although offlilal statlitlce largely exclude Lnformal employment ln workshops and at home. Zmployment ln the lnformal sector is difficult to gauge, but is wldely belleved to have grown rapldly durlng the past decade as real lncomes fell in the course of adjustment; unoffclial estimates add 0.5-1 milLon women to the offlical estimates. 3.20 In short, and ae ln many mLddle-Lncome developLng countrLes, a signLfLcant part of the female industrial labor force is statistically invlsible. This underlLnes the need not only to conduct time-use studles and to strengthen the labor force surveys, but more importantly, to revlslt and enforce labor legLilation, to reconsLder Government support for homebased work and to improve and broaden women's employment prospects, LacludLng thelr access to resources, and ln partlcular to credlt. (LLL) The Servicoe Sector 3.21 Approximately 12% of employed women work ln the services sector, and they constitute about 13% of the documented labor force ln the sector (Table 3.1). The services sector is deflned to include four subsectors: wholesale and retail trade; transportation, coummunication, and storage; flnancial and buslness services; and community, socLal, and personal servLces. About 63% of all women employod in the sector are in communlty, social and personal services, followed by trade (18%), business servlces (14%), and transport and communlcation (5%) in that order. Compared to lndustry, the services sector has been one of relatlvely rapid employment growth for women. Between 1970-1990 the female labor force in the servcoes sector grew three- fold (from 255,000 to 768,000) compared to a 50% Lncrease ln industry ( from 304;000 to 455,000). In the same perlod, the employment of men Ln the servlces sector doubled (from 2.5 million to 5.4 million). In the services sector, as ln the Lndustrial sector, there is a serlous under-estimatlon of workers, and especLally of women. 62 - 3.22 The omounity, personal nd oaoal lorvieo subseoctor ncompasoes a wide rang. of ocaupatlons from gavarmesnt emloyqee, profesilonals and clerlcal staff to domstics. The dltributlon of women wlthln thlo category (Table 3.2)o, oharactrlses so of the Lsvueso tfaig Turklsh womn today. The proportLon of protesslonal, o scientii and tochaLnal personnel La hlgh - 51%, and yet the proportion La ad4nletrative or managerlal pooLtLons Ls low - 1%. NotwithstandLng the hlgb proportlon -f professionals, oicro-level data suggesto that women are seg*rated in the loww ranks of *eoh profesiLon. For Lnstance, ln acade" women account for 32% of all faculty - but 20% of full professors. 30% of istructors, 35% of r earoh assistants and 54% of *other (mostly language teachers) n t, women account for just over 25% of professLonal staff, but thr r o fmale undersecretarles, 3-5 director-generals, and a mall percentage of dlrectors and chiefs ln the lLne mlniLtries (OXOD data fortbcamng). SdIilarly ln the health sector, there is a greater concentration of women ln the paramdical ranks (Annex 2). ME JS. 2 Distribution of Epleyyd WI_n In Conity. Social am Pend l Service by Oooutioeul Grou. 1990 0cc stional Groups Scientific, technical, professiona, and related wwkers 51.0 Acinistrative, cxecutive, and mmnWril workers 1.0 Clerical and related workers 27.0 Sales 0.3 Services 20.0 Agriculture. Forestry and Fihories 0.0 Production and related workers, transport equipment 0.7 operators and laborers TOTAL 100 Source: Household Labor Force Survey. SIS, 1990, Tahle 9. W Acar, ln Abadan-gnat (1981) and -GO and MMW (1989). YW Yayan (1984), The Role of Women in the Turkish 3conomy", SocLal Planning Department.o Ankara. - 63 - C. Zssues and Promosals Concernina Womepk' Productivity 3.23 DrawLag on the issues ralsed in Section B, thie section attempts to identify both the constralnts facLng women ln all sectors of the economy and allevlating measures. Many of the constraints are cross-sectoral, others sector specific. (L) Cross-Sectoral Constraints 3.24 The major cross-sectoral constraine LdentifLed are: awareness of women's contribution; access to resources and, ln particular, credit; the legal environment; and the general lnfrastructure of the economy. a) Awareness of Women's Contribution 3.25 Notwithstandlng the major contrlbution of women both in the marketplace and the homestead their output ie not fully recognized. There are many reasons for thies their role ln the labor market; thelr statistical invLsibLlity; and the lack of knowledge as to womens role. For instance, in the rural labor force, the majority of employed women work in the agricultural sector and are unpaid. Similarly, for women working in the rural industries, their husbands or fathers tend to recelve their wages directly. In the urban areas, the female particLpatLon rate le low (15%), although this understates women's partlcipation since the offLeLal labor statietics do not fully reflect women in the informal sector. The statistical LnvLisbility reflects a number of factores under-reportLng of women'e work by male proxy respondents, as well as by women themselves; the woverlap" between women's market work and housework; and the dlfflculty of both deslgnlng surveys and training enumerators to fully reflect women"s work. The lack of research into women's role has also contributed to the failure to fully recognize women's contributlon. As noted above (SectLon 3), there has been little analysis of women's work in the agrlcultural sector or research into improving productivLty or technology for female-Lnteneive actlvities. Similarly, the design of extension programs has not taken into account the need to provide informatLon or traLning to women (other than ln traditional wifely activities through the home-economice field staff), although MARA has recently introduced pilot programs to re-train the home economlets as field agents to assist women ln thelr agricultural activIties. In addition, in the past few years, MARA has begun to consider women's role in theLr rural development projects. The international agencies whlch have been most actlve in the developing world in promoting women's activities (PAO, IFAD), do not have a women's focus in Turkey although the UNDP has in the past two years lnitiated several innovative projects dlrected at women (for instance, geese-breeding, aplculture). similarly, the NCO0 in Turkey have only recently become active in addreselng the sltuation of the majority of Turkish women (Chapter 5) For lnstance, the Foundation for the Advancement of RecognitLon of Turkish Women (NFTW), founded ln 1986 has been active in lnitiating at the grassroots level non-traditional, as well as traditional activities in both urban and rural areas and established (with support from PAO) the first women's agrLcultural project ln Turkey (Giresun), whlch provides credit and establishes cooperatives for women. similarly in the urban areas, there is little knowledge of the role and contributLon of women in the Lndustrial, service or - 64 - informal sectors. In particular, other than soeiological or anthropological case studies, there has been no analyLss of women ln the service or Lnformal sectors. 3.26 Ironlcally, the seeming fallure to address women's contribution stems, in part, from the far-reaching nature of the women's reforms in the early Republican era. Specifically, slnce women were granted full rights, there was no need for women-oriented programs or policies. Rather, gender- neutrality has been the norm in Turkey since the early 1930s. The status of women in Turkey had become synonymous with that of the visible and articulate educated, urban women who enjoyed a high status in terms of labor force participation, remuneration, and their representation in public office and the professions. However, the position of the majority of women remaLns essentially unchanged by their legal emancipatlon. For these women, gender- neutral policles were not sufficient to ensure what had been provided by law. In the mid-eighties, with the resurgence in intereat in women's issues, there has been an acceptance within Government of the need for special emphasis on women's isues and a renewed level of activlty of non-governmental organizations - nonetheless, to date, the impact on the majority of Turkish women is minimal. 3.27 Thus, until recently, women's issues have recelved scant consideration be it in government, academic or private circles. Increasing awareness of women's issues involves, as a first step, not only integrating these Lesues into the consciousness of planners and decision-makers, but overcoming ingrained attitudes about w on's role in government and society, in general. In the interest of increasing the awareness of women's contribution and reducing their statistlcal invisibility (Chapter II), the collation and publication of existing data on gender would constitute a useful step, followed by strengthening the household labor force surveys, conducting time-use studies and training enumerators in gender sensitlvity. There is a need, in addition, for basic research - economic and social - lnto many aspects of women's contributLons quantltative research lnto labor-force participation, working conditlons and wage issues, and the impact of technology. In the rural areas, research is needed lnto women's agrlcultural actlvlties (includLng undertaklng gender-actlvity analyses)WV and means of strengthenLng the extenelon servLces and training programs to better serve women. In the urban areas, research is needed on the lnformal sector, especially the contribution of homebased work and the putting-out system; the service sector; the lmpact and contributlon of returnlng migrantsF and of the status of female-headed households. b) Access to Credlt 3.28 The revlew of Banklng and Credlt Laws ln Annex I lndicates that there are no legal constraints to women"s access to credit, 2e jo Nonetheless, ln light of the small number of loans granted to women by the banks catering to small and medlum enterprises and in the absence of systematic collectlon of credit data by gender, a small survey of six banks - DJ See for instance Feldsteln and Poats (1991). - 65 - Balk, 8Lrast, Vaklflar, lmbank, SKID and TSXB - was undertaken ln the summer of 1991 for thls report wlth the objectlve of obtaining some understandLng of the credit constraints women face. The survey (Annex 6) covered some 340 branches in all reglons of Turkey.J A separate survey was deslgned for Elraat Bank (the Agrlcultural Dank) ln vlew of the dLfferent loan types offered. The survey shows that ln Zlraat Bank women account nationally for 26% of deposits and accounts ln the branches surveyed but 3% of all loans and borrowers. Women's share of deposits and loans is slgnlflcantly lower in the FLrst and Second Prlority RegLons. In Halk, VakLflar and SYKB Banks, women's share of deposits and accounts varled between 12-15% whlle they accounted for 6-7% of loans and 3-4% of borrowers. The results further suggest that while the financial institutLons do not dLicriminate on the basls of gender, that these formal instltutlons and their programs do not meet the credlt needs of small entrepreneurs, includlng women, and that both supply and demand factors are at play. Supply factors include the structure of loans - average loan size is too large, terms are not flexLble to the borrower's circumstancesF transactions, processoing and monltorLng costs are too hlgh for small loans rigid collateral procedures and perceLved hLgher rliks of default on small and inexperienced borrowers. On the demand slde, women are constrained by custom from seeking credlt since tradltionally, male family members deal wlth flnanclal matters. Other factors constrainLng womens demand for credit in Turkey Lnclude strict collateral requLrements - partLcularly in agrlcultural areas whorm land is frequently'tltled in the male's name (Annex 1)t lengthy application procedures; dlocomfort ln dealing wlth banks, and in some cases, the absence of female staff to serve femAle borrowers. Two additlonal factors, conslitently identLfLed ln the survey responses as helpful to improvLng women's access (and small enterprlsoe in general) are advertising and the increased dlssemlnatlon of information about loan facliltles through exsLeting channlos, and the need to strengthen the baslc buaLness sklll of borrowers. 3.29 StudLes of women's access to credit and of programs to facliltate theLr accoss, from countrLes such as IndoneoLa, Malaysla, the Phlilppines, Peru and Mexico, demonstrate that whenever loan lnstruments and lendlng crLterLa *ulted to small-scale entrepreneurial activltLes in the lnformal or rural sectors are developed, that women demand such credlts and that the rate of return and repayment records are as good as or better than those of men. Such programs have demonstrated that streamlinLng of paperwork, modlfliatlon of conventlonal loan repayment schodulseW, collateral requlrements, through greater reliance on the repayment capacity of the borrower or of broadenLng the concept of collateral to encompass loan seocurlty through group lendLng, livestock, farm euipment or jewelry are helpful. 3.30 In vliw of the unsultablilty of the formal flnanclal lnstltutlons' programs for the need for capltal, Turkish women and small entrepreneurs have relled on limited Liformal credLt sources about whlch little lo known. One such source ls the rotatLng savings and loan groups, called "Para Li Gunw. 262 responses were receiovd ln total; lobank and TSKB did not respond. F Vor instance, allowLng a cholce if repayment in frequent small amounts or a few large payments dependLng on the expected incom stream. - 66 - These are essentially informal groups wlthln communLties, whose members make regular, specified contrLbutions to a fund, whlch li given in rotation to each participant. Contributions are frequently made ln foreign exchange or gold, as a hedge against inflation. The funds may be used for consumer purchases (cars, household equlpment), house purchaso or investment ln businesses. 3.31 In an e'fort to enhance the productivity of familLes, in recent years, several small programs to facilAtate access to credit have been introduced. For instance, MAR has lnltlated ln conjunction with PAO, a small program to provide in-kind credits to farmers - male or female - for small- ecale poultry,. livestock and milk-processLng activities. In addition, both the Turkish Development Foundation (TDF) and the MYTW have also begun programs to extend ln-kind credlt to women ngaged in rural and agricultural activities. The TDF program, targeted for 11,000 famliles, is based on a revolving fund for interest-free, in-kind credit repayable in equal annual installments over five years with a one-year grace period. The Family Research Institute in conjunctLon with Balk Bank lnitLated in 1989 a Family Credit Program, whereby families can apply for loans for the production of handicrafts within the homeJ as of mld-1991, 2,700 loans had been made for an average loan amount of just over $100 - wlth 85% of all loan to women. y Experience to date ln Turkey, however, suggests that lower transactions costs will be essential if credit programs for small entrepreneurs are to be developed within the formal financial Institutional seator. c) Access to COMnlMentar Inuts aNd IngrastutueCensftraints 3.32 Besides credlt, women also face constraints in thelr access to complementary inputs - marketing outlets, improved technology, technical support services - especially agrliultural extenslon and skill traiaLng. These constraints are dealt wlth more specflcally, in sectlons (LL) and (LL1) below. More generally, improve mnts i houalng and the natlonal infrastructure (water and sanitation, fuel and electricLty) will simultaneously help promote the productlvity and integration of women. l The Famlly Credit Progaram was lnitiated by the Family Research InstLtute in 1989 to finance home productlon of handicrafts. The Government contrLbuted TL5 billion to a revolving fund, with addltLonal funds provLded as neeed by Balk Bank, whleh is responsible for implementing the program. As of mid-1991, 2,700 loans had been approved for a total amount of TL212 bllion, of which TL9 7 billion had been dlibursd. Of the total loans ap2roved, 85% were for women. The mlnimum loan amount under thli program ie TL500,000 (approximately $125) and the maximum SL5 million (around $12,500). The maximum maturity is two years with a six month grace perlod. Government funds are restriLted to credits Ln the Prlority Provinces wlth an interet rate of approximately Sit; Halk Bank funds are for all other provLaces and the lnterest rate is approximately 68%. As of mid-July, the repaymnt experience was good - less than 2% of the total disbursed was in arrears, although repayment experience was poorer in the Prlorlty Provinces. Operatlng costs ln terms of processLng applications, monitoring loans and marketing are considered to be too hlgh to make the program commercLally feasible. - 67 - d) Ihe Leosi Framework 3 33 Chapter I summarises the legal environment as it affects women and Annex I reviews in detail the legal framework and the constraints it imposes on women's particLpation. While the Civil Code constralne women in so far as the husband remains the head of the houshold, and thereby, the decision- maker, the annulwgsnt ln 1990 of Article 159 by the ConstLtutLonal Court by which women were requlred to obtain their husband's permission to seek employment and to obtain credlt for their work, removed a major barrier to women's partLcipation. Nonetheless, an seen above, the exliting Labor Legislation constraLns women's participation, over and above the differential treatment accorded to women under the protective legislation. Specifically, Artlcle 13 of the Labor Law permits an employer to dismiss a %oman worker on ground: of pregnancy and wLthout severance pay. Similarly, some of the protective logislatLon regarding nlght work and heavy and dangerous work contributes to confining women to the tradltional femal i_Lntensive ind:astrial activities and should be reconsidered. The absence of legislation concerning working conditions in the agricultural and Lnformal sectors, whlch together account for more than 80% of the female labor force, and the difficulty of enforcing existing regulatlons, ln general (for instanco, mLnimum wage and the establlohment of childcare facilitles) constrain women's full participatLon. Simultaneously, the introductlon of leglelation/regulations permlttLng part- tLme and flexitims employment and to preclude dlicriinatlon agalnet older workers would also facilitate women's integration. In the agrlcultural sector, the Law of Succession (Articles 597-598, Annex 1, p. 6), which in the case of the partLtion of agricultural land gives priority to male heirs, disadvantages women, especially in the context of collateral for credLt. Li) ricultural a) Agriculunral Extension and Research 3.34 As noted in Section 3 above, women are constrained by the current delivery systems desplte their slgnlficant partLclpatLon ln the agricultural sector. Factors contributing to the continuation of these systme include: the focus of extenslon servLces and research on commercial crops (in which men tend to be involved) rather than the subsiltence crops, such as small livestock, vegetables, frults and flowers - all of whlch have conelderable commercial potential for Turkey; the vlrtual absnce of female extension agents or subject-matter specialist (S8'o) - as opposed to female home economlsts - outside of the provLn*Lal dlrectorateug the absence of women as contact farmrer, in large part sance they are not the titled owner the prevalence of tradltlonal attltudes that restriLt contact between the women and non-kln males; and, the male identity of the bureaucracy responslble for extenaLon and training - from the centrol plannLng offices to the village level. 3.35 Whlle women are largely excluded from agricultural extension servLces, they are not excluded from bm ssaKx na extension ervLces. ReflectLng the perception of women as homeworkers, the government programs and those supported by -nternational agencies, promote the trainLng and recruLtment of home economlits for dellverang servlos focussLng on foodmaking - 68 - and preservation, handlirafts (sewing, knitting), childcare and homemaking. In part, this system results from the tralning system. The agricultural extension training program for village extension agents is a 3-year dormitory program offered only ln Boys' Vocational Schools. The home-economics program is a 3-year program offered only in Girls' Vocational Schooles the main purpose of which is to increase the livlng standards of the rural families by increasing the homemaking skill level of the womer. Currently, the Ministry of Agriculture offers thre home economics vocational high school programs (graduating about 75 home economists a year) and employs 1,336 home economists in 73 provincees. Home-economics is one of eight Subject-Matter Specialist (SMS) groups within the extension program responsible for providing information and training at the provincial level. The home-economice SMS group, however, is the only group authorized to contact women and the only one wLth a significant number of women agents. Thus, while extension ls provided to women, it is not productivity-related. In recent years, MARA has lntroduced on a pilot basie programs to re-train home economists as field agents so that they can assist women farmers in their agricultural activities. It is also significant to note that the widespread employment of the home economice specialists also counters the prevailing view that women cannot be found to work in rural areas. 3.36 Although women continue to account for more than 25% of students in agriculture and forestry at the universities and most of the students (over 90%) are employed by the Minietry of Agriculture and Rural Affairs (MARA), iesues relating to women's agricultural activities receive little attention in either Government or academia. In both Government and the universities, agricultural research has tended to focus on commercial (male-inteneive) crops rather than on self-consumption (female-Lnteneive) crops. Case studies, however, also cite numerous inetances of productivity losses resulting from this focus. For inetance, hlgh yieldlng wheat varieties were not introduced in the eastern part of the country because of their unsuitability for home- consumption (the selection criteria of research departments focussed on criteria such as incremental yield, dloease resistance but not on criteria relevant to women - taste, grlndability, storage life) with a consequent loss ln ylield. Similarly, women were reluctant to use fertilizer on potatoes, despite the yield Lncreasing effect because of the shorter storage life with fertilizer. 3.37 Thus, consideration of women's lssues in both the central and regional offices in planning, extension and research is needed. The introductlon of OadvLseres on women's iesues or a women's unit in MARA to focus attentlon on these issues could contribute elgnificantly to raiolng awareness. Also, consLderation should be given to reviewing regulations concerning gender enrollment ln MARA coursees the curricula of the home economics programes as well as re-training both home-economiets and agrieultural S18's to address women's agricultural activLties and training in gender-senitivity; and to the establlihment of a target-number of female farmers for field extenslon agents to contact. A greater focus on female- lntensive agrLcultural and research activities in extension services would also be benefiLialp namely, crop storage poultry and livestock management and food-processLng. In research, and in technology development, a greater focus on women's actLvlties and crops is desLrable. In particular, consideration - 69 - should be given to undertaking time-use or gender-activity analyses (see Feldstein and Poats, 1991). b) Technoloay 3.38 Women are constrained in their access to new technology in several ways. For technology improvements relating to tasku performed by women (or women and men) which are addressed by extension agents, women rely on receiving information from their male-kin. In general, for tasks that are predominantly performed by women, (hand-hoeing, harves'ing by sickle., nut- shelling, silk-spinning) there has been relatively littls development of appropriate technology, in large part because research and development has focussed on commercial agriculture. However, where technology changes for primarily female tasks occur, frequently the traditional dlvision of labor is restructured to permit men to take on the newly productive task (for instance, electrical milking equipment; mechanized hoeing for sunflower production in the Thrace region). (iii) The Industrial and Services Sector Constraints 3.39 The major constraints facing these two sectors, besldes the general constraints touched upon ln Section C (i), are the need for trainlng, particularly ln the more modern, "higher tech" fields; the need to understand the impact of Turkey's economlc development and of technology on women ln these sectors; and supportive services, such as childcare. a) TraLning 3.40 For women in the industrial and service sectors, the need for appropriate training is perhaps the greatest constraint to their fuller participation. As noted ln Chapter II, the gender differential ln educational attainment has narrowed continuously over the past thlrty years, but gender segregation in the labor force has not yet begun to narrow. Approximately 80% of women in lndustry continue to work in food and textiles - sectors in which employment growth will likely be limited. Thus, there is an urgent need to broaden the training opportunities offered to women and, thereby, their skil. range. The small proportion of female entrepreneurs and self-employed women in Turkey also suggests that training in basic business and management skills. should be encouraged. Programs to promote women's entrepreneurship developed, inter alia, by the Economic Development Institute of the World Bank may provide a useful model. Annex 4 and recent studies (for inetance SudL, 1989) demonstrate a great demand by women at all educational levels for skill training. The da_a (Annex 4, Table 12) a2'so indicate that men receive significantly more on-the-job tralning than women and that more women ln skilled positions received their training in school. Simultaneously, however, data on applications by gender to vocational and technical schools show that enrollments rates for glrls schools are significantly lower than for boys schools, which outnumber girls schools by approximately 3:1. The discrepancy between desire for training and enrollment of girls may reflect the percelved utility of programs offered. Thus, a reassessment of the training policies and courses offered for women by all agencies, both in the context of Turkey's lndustrialization and service-orientation over the coming decade and in terms - 70 - of their remuneration and skill/career growth la needed. SimLlarly, a re- evaluation of the segregation of Boys/Glrls VocatLonal and TechnLial Schools is neded, together wLth the ratLonale for separate Dlrectorate Generals for Boys/GLrls VocatLonal Xducatlon and the oontinuLng gender segregation of courses offered by the Dlrectorate General for ApprentlceshLp and Non-Formal IducatLon. ThLs reassessment needs also to address specifically women's lLfe-cycle- needs, Lncludlng re-entry potentLal and the dLfforLng urban/rural partLclpatLon patterns of womn. In addltlon, the assessent needs also to address the issue of 'educating employers as to the advantages of hlrlng women. b) ChLldcar 3.41 According to Turkish Labor regulatLons, work places employlng between 100-150 women workers are requlred to establish nursLng rooms and those employLng more than 150, must have nursing rooms and creches for the care of children leos than 6 year of age. However, the regulations do not appear to be enforced. AccordLng to avallable data from the Directorate General for socLal Services and Ch$id Protectlon Agency (SSCPA), ln 1990 there were 262 creches and day-car centers run by workplaces (including all industrial and non-industrial establishments, both in the private and public sectors) and 17,321 chlldren enrolled. In private factorLes there were 12 chlldcare centers wlth a capaclty of 814 chlldren. Simultaneously, there were 486 prlvately run chlldcare centers ln Turkey wlth a capacLty of 20,591. However, slnce these are costly, they are only available to middle class and/or profesilonal women. Unlike ln many other parts of the world, lt does not appear ln Turkey that wLthin neLghborhoods, famlies operate lnformal day- care facliLties. grandmothers or other female relatlves appear, however, to play an important role ln provLdLng day-care. This suggests a serLous shortage of appropriate chlldcare facilities ln Turkey, whlch affects the employment of women ln two ways: first, it may prevent women who want to work, from doLng sop second, if women declde to work, they tend to look for work with reasonable access to day-care facilitLes. Unforcem_nt of exLstLng enterprLse regulatlon. will contrLbute to the establishment of a day-care mentality but, in addltLon, the developmnt of neighborhood and munlcipal day- care is needed to facilitate women'o productlve work outside of the home. For thli the use of the media, extension of credits (such as the Famlly Credlt Program) to establish community or hom day-care facliLtLes, provisLon of Lnformation on e*tablishing day-car and also, consideration to entablishing short training courses for community mothers. The Bank-supported day-care project in Colombia (No. 8029-00 May 1990) provides a useful model for this. - 71 - CRAPTER IVs EDUCATION AND TECHNICAL TRaINING A. Introdugtin 4.01 In Turkey, as in most countries, education and training are cruclal instruments for integrating women into development -- partLcularly Lnto paLd employment. The significant progress realized in Turkey since 1970 ln reducing the gender gap in educatlon has been dLscussed ln Chapter 1I (Section 0). Nonetheless, there are aspects of Turkey's educatLon and traLning programs where minor changes could strengthen consLderably thelr role ln improvLng opportunitLes for women. Thle chapter describes briefly the structure and coverage of education and training programs and examlnes theLr implicatLons for women's work. It then examines particular features of the system which are llkely to Lnfluence girls, enrollment in school and training programs and choLce of specialization. it ends with recommendations for improving opportunltLes for females in edu 2ation and training and, through them, in employment. B. 8tructure of the Educatlon and Trainina System 4.02 Universal educatlon has been an important prLnciple of the country since the founding of the Republic. The 1982 ConstitutLon states that *Primary education ie compulsory for all citizens of both sexes .... y law, education in public schools is free, and school attendance through the eighth gradey li now mandatory. But the coverage of education at the lower grades falls well short of this statutory requLrement, particularly for glrls (para. 4.7 and 4.8). Depending on the type of education, faclilty and requlrements, some schools, however, are allocated only to girls or to boys." In fact, gender stratLfication is qulte pronounced in vocatlonal education (paras. 4.17 and 4.18). 4.03 Formal education progrms at the prlmary and secondary level are the responsibllty of the Nlnistry of National Education (MOE). They comprlse a five-year primary cycle, a three-year middle-school cycle, and a three-year or four-year secondary cycle. Children normally enter primary schooling at age six in urban areas, and at age seven ln rural areas. There is also a limited program of pre-school education. Formal education throughout the country follows a common curriculum through prlmary and mlddle schoollngZl, wlth the exceptlon of Anatollan schools. Anatolian schools are established to meet the needs of the more successful students and equip them wlth a forelgn language whlch will be helpful for thelr higher educatlon ln Turkey or abroad. AnatolLan schools generally are better equipped than conventional schools and attract better J/The Government now refers to basic education as the elght-year perlod of compulsory education, comprlsed of the flve-year primary educatlon cycle plus the three-year mLddle-school cycle. VA once-extensive program of vocational educatlon at the middl_-school level has largely been replaced with a uniform program of general studles. A number of mlddle schools etill carry the vocatlonal-school desLgnation, but the vocational instructlon programs which they offered are being phased out. - 72 - qualified teachers and students. Zntrance to the Anatollan schools is hlghly selective, based upon a national entrance examlnatlon whlch students may take at the end of primary schoolLng. 4.04 In addltLon to formal programs of educatLon, there are also numerous non-formal programs of educatLon. Most of these are managed by the M.OB, but specialLzed, sectoral traiaLng is also provlded by the MinListrLes of Industry, Defenos, Agrliulture, Culture, Labor and TourLim, zion-formal education programs are of three types* a) literacy training for adu't and out-of-school children, b) employment-oriented training designed to d&1el.op skills for home production, to improve employabilty and to promote entrepreneurship, and c) llfe sklills in safety, hygiene, etc. The non-formal training provided by the Ministry of Labor's amployment Office (11BK) is particularly promising because it assures job placement upon completion of the tralning. The X1BK tralned and placed some 11,000 iob applicants between 1988-90, 60% of whom were women. Much of thls trainLng has been for the service sector (tourlim and commercial traLning). Many of the women's courses, however, focus on tradltlonal women's occupations. C. _tatus and Trends in Female Education and Trainino 4 05 Current 8tatu Table 4 1 summarizes the status of female enrollments ln formal educatlon and training by level and type of program, based on enrollment data from the 1987/88 school year. While there has been a signLficant reductlon in the enrollment and literacy gender gap in the past two decades (see para. 4.12) lt le apparent from the flgures ln Table 4.1 that the proportion of female enrollments continues to decllne as the level of education and traLning lncreasee, although there is lese inequality between male and female enrollmnts in general secondary education than in mlddle schooling or secondary vocatlonal education. Female enrollments constitute almost half of recorded enrollments ln pro-school and prLmary education, but represent only about a thlrd of enrollmnts ln middle schoollng, secondary schoolLng, and hlghor education. Female students are partlcularly under-represented ln vocational and technlcal education. Moreover, the flgures ln Table 4.1 on the overall situation understate the presence of females in many occupatlonal areas, due to the differentlatlon of programs ln glrls' and boys, vocational schools (para. 4.17). There is no clear pattern of glrls' enrollments in Anatollan schools. The proportion of glrlo' enrollmnts in Anatollan middle schools li hLgher than for conventlonal middle schools, but the reverse li true of Anatolian secondary general schools. - 73 - IULlE A.t Share of Finote Enrottl.nts In Formt Education and Training by Lwet and Type of Program, 1987-1988 Proaram TMre ,iuoaL IAL2 Femla fmle me X of Total Pro-School 103,204 54,257 48,947 47.4 Kindorgarten 15,153 8,138 7,015 46.3 Nursery School 88,051 46.119 41.932 47.6 Primary (grades 1-5) 6,860,304 3,646,0S4 3,24,250 47.2 Pubtle 6.843,335 3,615,708 3,2f?7627 4T.2 Private 36,969 20,346 iS,623 45.0 Middte (grades 6-8) 2,091.178 1,354.42 736.7 35.2 General 1.870,244 1,207,332 662,912 35.4 Pubifc 1,815,110 1.175.701 639,409 35.2 Anatolian 33,424 21.398 12.026 36.0 Private 55.134 31,631 23.503 42.6 Vocational 220,934 147.110 73.824 33.4 Pubilc 220,410 146,839 73.571 33.4 Anatoltan 1,069 680 389 36.4 Private 524 271 253 48.3 Secondery (grade 9-12) 1,168,357 743,089 425,268 36.4 General 69r,227 396,885 300.342 43.1 Pubtlic 673.130 384.246 288.884 42.9 Anatotlan 14,713 8,984 5,729 38.9 Private 24,097 12,639 11,458 47.5 Vocational 471,130 346,204 124,920 26.5 Pubilc 469,923 345,548 124,375 26.5 Anatotlan 4,229 2.620 1.609 38.0 Private 1,207 656 551 45.7 University Undergraduate 3107m3 199.968 111,095 35.8 Technieal/Vocationlt 59,421 40,578 18,843 31.7 Open University 133,139 95,057 38.082 2S.6 GradUate/Professionat 30,836 20,584 10,252 33.2 sources: statistics of Nationat Education, Youth, and Sports 1967-1988 (Ministry of Ntional Education). Higher Educ tfon Statistics, Student and Placement Center - OlIN. - 74 - 4.06 What do these flgures lmply about room for improvement in glrls' enrollments ln formal programs of education and trainlng? At the prlmary level, the hlgh (47.2w) percentage of female enrollmentu suggests satlifactory coverage, but estLmates of enrollment ratLos show that there remalne considerable room for improvamnnt. The normal indlcator of educatlon coverage ln Turkey is the gross enrollment ratlo. The gross enrollment ratio expresses the number enrolled at a given level of schoolLng as a percentage of population ln the age group which li concldered normal for that level of schooling. A flgure frequently cited ln Turkey li the primary-school gross enrollment ratio of 102 percent, derlved by divldLng recorded 1987/88 primary-school enrollments, public und prlvate, by the SPO's estLmate of 1987 populatLon for the 7-11 year age group. If one performs the same exercLse separately for male and female enrollments, it leads to a gross enrollment ratlo of 105 percent for males and 99 percent for females. 4.07 These flgures and flgures like them have been Lnterpreted in Turkey as evidence that primary-school coverage is complete, notwlthstanding the contrary evidence ln literacy ratloo (paras. 4.9 and 4.14). But ln the presence of hlgh grade repetition (para. 4.11) and numerous children who start school late, gross enrollment ratloas exaggerate actual school coverage because they Lnclude overage chlldren in the numerator but not in the denominator. In thli situation, a more accurate indLiator of school cua-rage li the net enrollment ratLo. Llk the gross enrollmnt ratlc, the net. erollment ratlo is a ratLo of enrollments to populationg but unlike ts'% gr"a enrollment ratio, lt excludes overage students from the numerator so that t 'tii the numerator and denominator are limlted to the sam age group. The primary-school net enrollment ratio for 1987 ia 79.3 percent, based on SPO population estimates for the 7-11 year a group. The separately estimated net enrollment ratios for males and femalesa in 1987 were 81.7 percent and 76.7 percent, respectively. 4.08 The enrollment defLclts whlch these flgures imply are summarLsed ln Table 4.2, below. In primary schooling, the deficit ln female enrollments ln relation to male enrollmentsW ln 1987 was 164,000. Another 1.2 million enrollments, about half of them female, would be necessary to attain full enrollment at the primary level. At the mlddle-school level, the female enrollment gap is even larger. There, an addltlonal 335,00 female enrollments would be requlred to attain parlty wlth male enrollments, and an additional 2.1 mllion enrollments -- agaln,, half of them female -- would be necessary to reach full enrollment at the mlddle-school level. Thus, based on the latest avallable data, lt would require an addltlonal half mllion female enrollments merely for female enrollments to catch up wlth male enrollments Ln basLc education. Full compliance wlth the Government's pollcy of universal baslc educatlon would require an addltlonal 3.8 mllion enrollments ln primary and mlddle schoolLng, includLng 2.1 million addltlonal female enrollments. There is, then, very important unfLnashed business ln provldLng basic educatlon to all chlldren in Turkey. ThL li particularly true for girls, sance these estLmates indicate that 41 percent of glrls of primary .and mlddle-school age currently do not attend school. ygecau" published HOg enrollment data by students' age are not gender speciflc, these estimates are based on the assumption that the proportlon of enrolled males and females aged 7 to 11 years is the same as the reported proportlon of total primary enrollments aged 7 to 11 years. Given the tehdency to wLthdraw glrls as they reach puberty, this assumption overetimates female enrolLnmnts. VoDefined as the number of addltlonal female enrollments necessary to attaln the net enrollment ratlo for males Ln primary schooling. - 75 - ZADLLh4 2s Enrollment Deficits for Helt end F_les in lastc Education by Level and 8ource 1987 PRINAR1 HALE FEMAlE TOTAL (1) Total Enrollments 3.636,054 3,244,ZO 6,880304 (2) Enrollments Aged 7-11 2,821,064 2,517,079 5336.143 (3) Estimated Population Agsd 7-11. 3,451,000 3.200000 6m731,000 (4) Estioated Net Enroltment Ratfo (NER). (5) Deficit in Femalo Enrollents with 0 164.000 164.000 Respect to Hale Enrollments. (6) Remaining Oeflcit In Enrollents with 629.936 59J 921 1.228.657 Respect to Full Enrollment (7) Total Enrollment Defict 629.936 762,921 1.392.857 MIDDLE SCHOL (1) Totat Enrollments. 1.354,693 737,097 2W091790 (2) Enrollments Aged 12-14. 875179 476,191 1.351,370 (3) Estioated Populeatfn Aged 12-14. 1.995.000 1 .JOO000 3,805.000 (4) Estmated Net Enrollment Ratfo (NER). 0.443 0.260 0.355 (5) Deficit In Femalo Enrollments with O 334.890 3346890 Respect to Hale Enrollments. (6) Remaining Defieit ln Enrollments with 1.099,J21 1.0t1.*i. 2,118,740 Respect to FUll Enrollment. (7) Total Enrollment Deficit. 1.099,821 1.353.809 2,453,630 RAsIC EDUCATION (t7 (1) Total Enrollments. 4.990.747 3,981.347 8,972,094 (2) Enroltlents Aged 7-14. 3,696,243 2,993,270 6,689,513 (3) Estimted Populstion Aed 7-14. 5,426,000 5,110.000 10,536,000 (4) Estimated Net Enrollmnt Rotfo (EKR). 0.661 0.585 0.635 (5) Deficit fn Femlo Enrollmnts with 0 498.890 498,890 Rospect to Hleoi Enrollments. (6) Remaining Doffict In Enrollments with 1,729,757 1,617.840 3,347,59r Respect to Full Enrollmnts. (7) Total Enrollment Deficit. 1,729,757 2,116,730 3,846,487 (1) Statistics of NatIonal Eduostfon, Youth, ad Sports: 1987/1986, 01OE/u81). (2) The figures for moles ad fmemles assua th proportion of enrotled males and fmles in the relevant ag grozp fs the a as the reported propwrton of total nrollents fn the releant ae group. (3) SF0, uTurkfeh Population Projections for the Sixth Ffve-Year Develtpmnt Plan (1990-1994). February 1989. (4) Row 2 divided by Row 3. (5) (Hale HER - Feualeo ER) x female populatlon In the relevant ag growu. (6) Rom 3 -Row 2 -Row 5. (7) Row S Row 6. - 76 - 4.09 How does Turkey's performance in providing education for girls compare wlth other countries? Table 4.3 presents a summary compariaon in terms of female enrollments and female literacy -- a cumulat've record of past enrollments -- for seven countries in the region. TABLEg 4.3 A Reionat Coeparison of Femate Schooling and Lfteracy. Per Capita Adult Literacy Prlmary Net Gross Enrollment Ratio Co GNP (12989 Rote i 1985. Enrottment Ratfo Primarv Secondary CS) ale Fonte , Total at Fet Male Femael Portugal 4250 88X 80X lOOX 125X 127X 55X 63X Spafn 9330 96X 92X 100X 112X 110X 99X 111X Yugoslavia 2920 96X 86X N.A. 94X 94X 81X 79X Greece 5350 96X 88X 97X 102X 102X 97X 93X Jordan 1640 87X 63X N.A. N.A. N.A. N.A. N.A. Turkey 1370 86a 62X 79X 105X 99X 52X 34X Iraq 91X 87X 84X 105X 87X 57X 37X Source: World Develoanmnt port, 1991. except for the primary enrollment ratios for Turkey, ihich are from the Ninistry of National Education. it is apparent from these figures that the differential between male and female literacy is greatest where literacy rates are lowest. Adult literacy for males and females Ln Turkey is the lowest of any of the countries in the reference group. Per capita income is also the lowest, but the estimates nonetheless provide some indication of the progress that is required. The primary-school net enrollment ratio in Turkey is also the lowest of any country in the group. Primary-school gross enrollment ratios are misleadingly high in Turkey because of the prevalence of overage students due to grade repetition (para. 4.11). The secondary-school gross enrollment ratios for males and females in Turkey are lowest of any country in the group, although these figures are overestimates -- again, due to the prevalence of overage students. 4.10 The differences between male and female enrollments at every level result almost entirely from differences in the proportion of 7-year-old boys and girls who start primary school, and, thereafter, in the proportion of primary- school graduates who proceed to middle schoolingY The number of girls who began primary schooling in 1988 was 13 percent smaller than the number of boys who began primary schooling ln that year. Vsingle-year population estimates which might be used to estimate these proportions are subject to large variance because of errors in the exact age of the census population in the base year from which the population is projected. - 77 - 4.11 Table 4.4 summarisee the promotlon, repetition, and dropout rates for primary and secondary schoollng. Repetltion rates decline *qually for boys and girls throughout prlmary schooling, and rise sharply thereafter -- mout markedly, for boys. Dropout rates also rise sharply -- agaLn, particularly for boys -- after primary schooling. Taken together, these repetltlon and dropout rates conatitute unusually high internal inefflciency by LnternatLonal standards. This is most apparent in the first years of mlddle schoollng and secondary schoollng, where only 52 percent of male students progress to the next level of schooling on schedule. In every grade beyond primary schooling, girls perform better than boys ln terms of promotion rates. Yet the transition rates from primary schoollng to middle schooling are much lower for girls than for boys. The proportion of primary-school graduates who progressed to middle schoollng ln 1988 was much lower for females (45.7 percent) than for males (65.8 percent). In contrast to this, the proportion of middle-school graduates entering secondary schooling in 1988 was virtually the same for males (27.9 percent) as for females (27 3 percent). The absolute number of secondary-school entrants, however, is much smaller for girls than for boys because of the cumulative effect of lower rates of girls, entry into primary and middle schooling, with a consequent smaller number of female middle-school completers. The principal challenge for reduclng the male/female enrollment gap in primary and secondary schooling is to raise the proportion of 7-year-old girls who start primary school, and, particularly, to raise the proportion of girls completing prLmary school who progress to middle schooling. TABLE 4.4: Promotion, Repetition ad Dropout Rates by Grade for loes and Females (1987*1988) Pr-omotro-n Recetitin -gronoms Grade flays Girls Boy Girls Boy Girls Primary 1 0.83 0.84 0.15 0.14 0.02 0.01 2 0.90 0.90 0.08 0.08 0.02 0.02 3 0.89 0.89 0J9 0.09 0.02 0.02 4 0.94 0.93 0.04 0.04 0.02 0.02 5 0.97 0.96 0.01 0.02 0.02 0.02 Niddle 6 0.52 0.65 0.39 0.29 0.09 0.06 7 0.63 0.70 0.33 0.26 0.05 0.04 a 0.76 0.82 0.23 0.17 0.01 0.01 Secondary 9 0.52 0.58 0.39 0.36 0.10 0.07 10 0.76 0.81 0.20 0.17 0.04 0.03 11 0.78 0.82 0.16 0.16 0.08 0.02 Source: Ministry of Educatfon Data. - 78 - 4.12 Trnda. Although sLseable male/female aaroilment dlfferentLals are apparent at every level, there has been progress ln reduclng the gap between male and female enrollments. Table 4.5 sumarises the growth of male and female enrollments and graduates by level and type of program durlng the 1980s. Female enrollments and graduates grew faster than male enrollments and graduates in every schooling category except secondary vocatlonal education. The pace at which female enrollments caught up wlth male enrollments was fastest ln secondary general education, where female enrollments and graduates grew more than twice as fast as male enrollments and graduates. TABLL 4.Ss Growth of Enrollmnts in Formal EducatLon and Tralning for Males and Females by Level and Type of Program, 1980/81 - 1987/881/ NaLE FEMALE _ N~~~~ME C,) Primary Enrollments 2.2 3.2 Primary Graduates 2.9 4.0 Middle-School Enrollments 6.7 7.4 Middle-School Graduates 4.1 5.0 secondary General Enrollments 2.3 6.1 Secondary General Graduates 2.5 5.0 secondary Vocational Enrollments 8.8 5.2 Secondary Vocational Graduates 7.8 3.7 (1) Rates of growth of graduates are for the period 1980/81 - 1986/87. Sources Statistical Yearbook of Turkey, 1989, Tables 89 and 90. D. AttrLbutes of the Out-of-School PpInUlaio 4.13 In view of the Government' cooducation polley, we would expect to see roughly equal numbers of boys and girls in school if school availability alone determined school attendance. The figures on female enrollment deficits shown in Table 4.2 Lllustrate the Lipact of factors other than school availability which influence school attendance. The deficit of 500,000 girls in basic education Lndicates that efforts are needed to encourage girls' school attendance. What evidence do we have as to who the out-of school children are and why they are not in sciool? For girls in partLcular, what interventions appear most appropriate for improving school attendance? Unfortunately, there is little Lnformation avallable to answer these questLons. A potentlally valuable source of information, the semi-annual Household Labor-Force Surveys conducted by the State Institute of StatLitLcs (I8B), collects information on school attendance and household characterLstLce, but the school enrollment - 79 - information in not currently published. Xnrollment information on houcuhold members is presented in the semLannual survey reports oniy ln the form of highest level of schoolLng completed. The Bank li worklng wlth the SIS to try to change the format of presentation such that reglonal varlations Ln school attendance can be dlitlngulshed and characterlstlce of households wlth dlfferentlal school attendance can be analysed in order to help ldentlfy approprlate interventions to ralse school attendance. Also the Bank, together wlth NO0 carried out in 1991 a selectlve fLold study ln areas of the country wlth low glrl' eschool attendance to ldentlfy the factors responsible. it ie planned to use the results of thls study an an early lnput to preparatlon of a project ln support of baelc schooling expansLon ln areas of greatest need. Study fLndLngs will be used to develop targeted lnterventlons to improve gLrls' school attendance ln areas of the country wlth deflclent enrollments. 4.14 Publihed data on reported illiteracy rates from the 1985 Census of Populatlon and on mlddle-school and secondary enrollmentes for the 1987/88 school year provide a useful vlew of regLonal dlfferentlals -- including male/female dlfferentlals. Table 4.6 presents these data by province and region. Generally, male/female dlfferentials ln illiteracy and school attendance are greatest ln provLaces and reglons wlth the hlghest levels of illiteracy and lowest levels of school attendance. Thus, the Bastern regions -- the Black sea and Bast and Southeast Anatolia -- had the lowest enrollment ratios and the highest reported illiteracy rates for males and female. of all the regions, and the largest male/fmale differntlals in enrollment ratios for both middle schoollng and secondary schooling. The Marmara Region (Lncluding Istanbul) and Central Anatolia (includLng Ankara) had the hlghest enrollment ratios and the lowest reported illiteracy rates for males and females, and the smallest male/female enrollment differentLals of all the regions. In every region, the difference between male enrollments and female enrollments was greater at the middle-school level than at the secondary level. This pattern illustrates a general tendency for males to benefit first from an initial expansion in educational capacity, with girls' enrollments catching up at higher overall levels of educational coverage. it also illustrates the particularly rapld growth of female enrollments relatlve to male enrollments in secondary general education (para. 4.12). Yuquivalent data for primary schooling are not available. - 80 - TABIE 4.6: Illteracy Rates and Grou Enrollmnt Ratios by Region and Gener tEGIOIIS Ilttteracy Rate C1) Gross Enrollment Ratfos. 195185 (2) (198S) .idi S Fmhgg ecb Eastem Anatolia 21.9 49.4 51.4 21.2 33.1 12.6 Black Sea 14.8 33.8 62.0 30.4 41.7 18.5 Nediterransen 13.1 31.2 61.5 39.8 38.1 24.3 Narmera 8.1 20.6 82.9 60.5 45.1 38.4 Aeean 10.9 26.5 70.4 46.2 38.2 27.5 Central Anstolls 10.4 26.2 65.2 43.1 46.0 26.9 Southeast Anatoltf 28.0 60.3 45.2 18.1 25.4 9.8 Source: (1) DefIned as the proportion ef population over 6 years reported as unable to read and wrIte. 1985 Cns of Population, Socfa. and Economie CharacterIstics, Table 27. (2) TUSAID Educatfon Report, Turkiye de Egitimm* 1990. Table 25. B. Gender Agggots of Education and Training Proarams I.15 Within the Turkish National Education system, secondary school students make their own cholces whether to attend general schools or vocatLonal/teehnical schools according to thelr skills and Lnterest areas. The students are oriented to the most appropriate program according to their skills and interest. Students who successfully complete mlddle schooling may proceed directly to a three-year program of general secondary education or may take an examlnatlon to enter a speclfic vocational secondary program. The purpose of the exam is to select appropriate students for the program. In Turkey, as in most countries, general secondary educatlon tends to be the preferred form of secondary education, reflecting the fact that it is the normal route to higher educatlon and higher-level jobs. But although general education attracts the best students completing mlddle schooling, many students opt for vocational education in the belief that lt will improve their chances of getting a job at the completlon of secondary schooling. 4.16 About one-thlrd of students entering secondary schooling enroll in vocational programs. At the end of a common flrst year (grade 9), vocatlonal students with good course performance in science and math subjects may enter a technical secondary program consisting of three addLtional years of classroom and workshop tralning ln one of sixteen specializatLons. Other vocational students continue wlth two more years of vocational instruction, in whlch training is provided alternately in the vocational school and ln an enterprise outside the school. Both the three-year vocatlonal program and the four-year tchnLcal program are meant to be terminal, but students from both may proceed to hlgher education if they pass the natlonal university entrance examLnations. A slgnlfiLant number of technical secondary graduates progress to higher education, few vocational secondary graduates do so. - 81 - 4.17 About one-quarter of vocatlonal and technLcal secondary schools are gender neutral. ror the most part, these are commercial and tourlsm schools, ln whlch enrollments are about equally dlvided between boys and girls. most vocational and technlcal schools are deslgnated as boys' schools or glrls' schools, dependlng upon the speclali2atlons whlch they offer. Orlglnally, the separatlon of girlr' schools in Turkiy was undertaken to attract girls to the schools. About a thlrd of vocational and technlcal schools are desLgnated as girls' sohools, and are admlnistered by the MOE's General Dlrectorate of Technical IducatLon for Glrla. These schools offer traLning primarily ln occupations (such as textiles, weavlng, food procegsLng, haLrdressLng and ceramics) whlch tradltionally have attracted women (Chapter IU and Annex 3)* In addltlon to the traditional occupatlons, however, more recently programs for the lndustrLal and servlce sectors such as travel agency, hotel management and tourism servLce; textle-design; electronLscp offlce management, secretarlal traLning and computers; food technology, food control and analysLs have been Lncluded ln the currlculum as a result of needs stemmlng from the social, economia and cultural development of the society. Boys can attend theme schools, but few -- amountLag to less than 1 percent of enrollments -- do. Analogously, boys' vocatlonal and technLcal schools, admlnlstered by the MOE's General Directorate of Boys' Vocatlonal Educatlon, offer tra_nlng in occupatlons which are typlcally male-domlnated. Very few of these are hazardous or strenuous occupatlons (such as mlning and foundry work) whlch may be sald to be naturally limLted to male workers; most are occupatLons (LncludLng most manufacturing and construction occupations) ln which women can perform productlvely, and do so ln most OECD countries. About 6 percent of enrollments ln boys' vocatlonal and technLcal schools are gLrls -- a larger proportlon than ln the converse case, but still a very small proportlon of total enrollments. Wlthin the non-formal education system, gender segregatLon also appears wldespread, wlth boys enrolled exclusively ln ApprentLceshLp Training Centers and glrls prlmarlly lnvolved ln Girls Applied Craft Schools (Annex 4, para. 32). 4.18 The contlnulng gender segregatLon of vocatlonal educatlon ln Turkey, in name and ln fact, reflects a still wldespread vlew that many occupatlons are not suitable for women. The continuing separation of boys' and gLrls' vocatlonal schools exaccerbates thle view. By eliminating the gender designation, the Government would help reduce perceptlons that women belong in certain occupations, but not in others. Government intervention could also help to eliminate the existLng gender blas in vocational educatlon and training programs A possible model for such efforts is a USAID-supported project ln Morocco / whlch set out to attract females to tralnlng programs ln male-domlnated occupations Ln manufacturlng, constructlon, and the commrclal sector. With relatively modest efforts to explicltly lnvlte female applicants and to provide inltLal screenlng and counselLng to them, the project achleved LmpressLve results in attractLng females to tralnlng programs for these occupatLons, and in placLng them ln approprLate jobs after completing the courses. ZDescribed ln Margaret Lycette, 'The Industrlal and Commercial Job TraLnLng for Women Project ln Morocco', USAID worklng paper, June, 19861 and in Laurel Ulmer, aGuLdelLnes for IncreasLng Female PartLcLpatLon ln AID TraLnLng Programs for Asla and the Near East", USAID Working Paper, September, 1990. - 82 - 4.19 One of the early fLndLngs to _mrge from the Jolnt Bank/MOE study "Factors Affectlng Glrls' School Attendanco Ln Turkey iL that some regLons ln particular are willing to sentd thelr adolescent daughters to separate glrls' schools, but not to coeducatlonal schoolly. For these groups, the Government's policy of coeducatLon may frustrate the attaLnment of lts goal of unLversal basLc educatlon. Moreover, the fact that all general secondary echools are coeducational constraLns the career optlons of girls whose parents are opposed to coeducation, since these schools are the normal route of entry to most higher- level jobs. Parents who consider coeducatlon to be culturally unacceptable for their adolescent daughters can elther send their daughters to a girls. vocational school (if one is available), withdraw them from school altogether, or send them to a religlous school (Imam Hatip) - an optlon whleh has grown ln Lmportance recently. Although coeducatLonal, these schools are often seen as acceptable by parents who would not oend thelr daughters to a secular coeducational school. For the current generation, conaLderatLoh could be glven to establihLng separate girls' boarding schools -- both mlddle schools and secondary general schools -- ln remoter parts of the country wlth slgnlfLcant senaLtlvLty to coeducatlon. This could be done wlthout calling lnto questlon the basLc polley of coeducation, since the establishment of slngle-sex schools would be temporary. It is expected that female educatlon itself, together wlth increased attention to gender issues in the curricula would, over time, diminish the cultural sensltlvlty to coeducatlon. 4.20 In low-density rural areas, the dlsperslon of population favors boardlng schools, partlcularly at the mLddle-school and secondary levels. The Ninlstry of Education has recently started a program to provlde regional boardlng schools for children from remote vlllages.Y Although thli :d a promising approach to the problem of hlghly dlopersed students, som aspects of the program in practice limit its effectiveness. For lts regLonal boarding schools, as for its conventional schools, the Ministry's now pollcy ls to establish Lntegral eight-year "basic education schoolow ln place of exlsting five-year prlmary schools and mlddle schools (which are typlcally comblned with secondary schools). The motlvation, for this change derlves from the Government's wLsh to provlde all the classes of compulsory schooling ln the same school. While the neatness of thLs plan may be appealing, it is not necessary on pedagogical grounds to ensure quallty educatlon or integration of the basle schooling curriculum. The integrated basic education curriculum whlch the Mlnistry plans to develop (with support from the Natlonal Education Development Project) could as well be implemented under the existing formula of dlitlnct prLmary and mlddle schools. The Ministry also intends to phase out the Lncomplete primary schools whlih exlit I/in several cases, the centralized cooducatLonal boardLng schools whlch the Mlnlstry of Educatlon has bullt to provlde schoollng to children ln romote villages have attracted only boys, and have consequently become de factg boys' schools. 2/Currently, there are 102 of these schools, enrolling 19,000 males and 7,000 females in grades one through eight. There are no fees for attendance at these schools. Boardlng students are also provLded meals and unLforms at no chargel they also receive modest stlpends to help defray the cost of transportatLon and educatlonal materlals. - 83 - ln many rural areas. These schools have fewer than fLve classrooms -- one for each grade -- and thue resort to multl-grade teachlng.V The Ministry (and many others ln Turkey) consLder that these Lneomplete primary schools cannot offer prlmary education of acceptable quality. FLndLngs from other settLngs, however, show that small schools with multigrade teachLng can provLde educatlon of equal or hlgher quallty than conventional schools if teachers are selected and tralned properly, and if the necessary educational materials are avaLlable.11 There is, then, a compelling cost-effectiveness argument ln favor of not changlng the format of basLc educatLon as the MLnLetry proposes. most parts of the country could be reached through conventional, fLve-year coeducational primary schools and incomplete primary schools wlthout recourse to boardlng, partlcularly if the Mlnistry is willing to make a serlous effort to make multi-grade teaching work. Boarding schools are least 4pproprlate for chlIdren ln the first five grades, where parental and sLbling support are vital to healthy chlild development. Moreovez glrls" boarding schools are not necessary to ensure female school attendance at the prLmary level because there appears to be llttle reslstance to coeducatlon at thli level. sensitlvlty to glrlsa attendance at coeducatLonal schools gnnerally begLns at puberty, which often coLncLdes wlth the start of middle schoolLng. The creatlon of boardlng schools to serve a dLipersed cllentele makes most sense at the ml4dle-school and secondary levels. To ensure that these schools attract girls, separate boardlng schools could be establlshed for the current generatlon of adolescent glrls ln areas of partlcular sensLtivity to coeducation. 4.21 Other notable features of formal primary and secondary education whlch may affect glrls' attendance ares a) the'very limited role of prlvate schools, accounting for less than I kae ent of enrollments at the prlmary level, and less than 4 percent at the middle-school and secondary levels, and b) the prevalence of out-of-school, pald tutorlal instruction to impr6ve secondary students' chances of success in the national university entrance examinations. Private schooling ln Turkey, while limited, benefits boys disproportionately. Greater attention to the provision of private, schooling for glrls should be encouraged, especially it the role of prlvate education increases. There is no dlrect evidence avallable on male-female dlfferences in tutorial instruction, but it is likely that here, as ln private schooling, boys benefit from household education expendltures more often than girls. P. Conclusions and Recommendatioms 4.22 The paragraphs above have summarized some of the principal issues of female education ln Turkey. ThLs seation provides a brief recapitulation of findlngs and recommendatLons ln the two main issue areas. These relate, first, to the -coverage of the educatlon system ln terms of girls' enrollments, and, WVIn whlch several classes are taught simultaneously in the same class by the same teacher. 11The evLdence on thli Lisue is revliwed in Marlaine 3. Lockheed and Adriaan M. Verspoor, Improving Primary 3ducation in Doveloplna CountrLes: A Review of Policy Octions, Background Report for the World Conference on Education for All, Bangkok, 1990, page 109. - 84 - ascond, to the biases of the rxistLng system in terms of orientation to future careers. 4.23 educatl on. gmso. The current coverage of the education system is incomplete, especially for glrls. This is a particular concern at the basLc education level, wher the Government appropriately aspires to provlde schoollng for all TurkLih ohildren. There are now about 2.1 million girls of primary and middle-school age who should be attending school but are not (para 4.8 and Table 4.2). Most of tse -- about 1.6 million girls -- are not attending school because no school 's available. But it appears that a large number -- about 500,000 -- are not ln school because of other factors. Work is now in progress to determine what these constraints are. Evidence from other settings suggests that the reasons are likely to be cultural and economic, and that Government policy could help reduce the constraints to girls' school attendance through actions, inter alia, such as raising the proportion of female teachers and providing better transportation.M. To improve girls, school attendance, the following specific actions are recommended: o Expand the schooling infrastructure, starting at the primary level, then at the mlddle-school level. * In doing so, retain the existing structure of school delivery, with generally separate primary schools and middle schools. There is no convLncing reason to establish integral eight-year basic education schools an the Government proposes, elther for reasons of cost- effectiveness, pedagogical effectiveness, or access (para. 4.20). o In the program to expand basic education infrastructure, use the res6ltz of the study on factors influencing girls, school attendance (para. 4.13) to develop targeted interventions specifically designed to reduce the barriers to girls, school attendance in areas of the country with defLoient enrollments. o on an exceptional basis to improve girls' continuation in schooling beyond the primary level in the remoter areas, develop girls, middle schools and girls' general secondary schools (para. 4.20). 4.24 BiAs in Educati2n PrZKgrm. To address the bias in education programs whleh llmits girls' educatlon and career prospects (paras. 4.17 and 4.18), the followlng actlons are recommendeds o Blimlnate the deslgnation of separate boys' and girls' vocational schools, and the corresponding dlstinction ln the admlnlstratlve structure of the Mlnistry of Education. o Conduct active programs to attract girls to male-dominated secondary vocational program. o Support the expansion of job-orLented training for women, such as that offered by IIBR (para. 4.4). lsug, e.g., Rosemary Bellew and ElLsabeth M. King, Promotina Girls' and Wbmens. Educadion: Leaos- from the Past, World Bank PRE Working Paper No. 715 (July, 1991). - 85 - CHAPTER V: THE ROLE OF GOVERNNENT AND NGOS IN WOMEN'S DEVELOPMENT A. Introduction 5.01 In 1926, Turkey became the first Islamic country to elLmLnate the Sharia, the legal code that underlies the segregatLon of the sexes and dlfferentLal legal treatment. ThLi was followed ln the 1930e by the rlght to vote in natlonal electlons and stand for office (1934) and the opening up of public offices and the professlons (law, medicine, flnance) and the universLties to the appointment of female faculty ln 1932-33. Wlth these far- reaching reforms and the tremendous progress realized by women ln entering public office and the professions through the mld-1940s, the prevailing vlew was that there was no need to address women's issues lndependently of the general development process. For the majorLty of women, however, thelr positlon was unchanged - gender-neutral pollcles were not sufflclent to ensure what had been provided by law. Coincldlng wlth the resurgence in interest ln women's issues in the mld-eighties, there has been an acceptance withLn Government of the need for a speclal emphasis on women's lssues. B. Develovment Plans and Government Policles 5.02 In the first five Five-Year Plans (1960-1984), women's Lssues were not explLcitly addressed. Nonetheless, Government plans and pollcies aimlng at soclal change affected women ln four main areast populatLon pollcy, education, employment and working conditions and social security. 5.03 In terms of oovulation wollcv, the maln iesues relating to women since 1960 have been the measures to reduce population growth. Wlth the establishment.of SPO in 1960, the existing pro-natalist polLcLes were brought into question. The first plan introduced a pollcy advocatlng low population growth and the second, followlng the passing of the 1965 Population Planning Law, introduced national family plannlng programs and set a plan target of provldlng family planning servlces to 25% of women of chlld-bearing age. These targets have been raLsed in subsequent plans. The current Sixth Plan (1990-94) contains no targets for the provision of family planning services but with a contraceptive acceptance rate of 77%, Turkey has one of the highost acceptance rates in the world (WDR, 1991). 5.04 Although all of the plans provided for some improvement in the general educatlgn system, until the Sixth Plan, there had been no special consideratLon of women's educational needs. The Second Plan, however, was critical of Girls Technical and Vocational Schools on the grounds that these schools channeled girls into "female* occupations. Although not part of the Fourth Plan, eR se, the adult literacy campalgn initiated by the Government in 1981 is slgnificant for helping 2.9 million women learn to read and write between 1981-87. 5.05 With the greater majority of women working in the agricultural sector as unpaid family labor, for the most part women were not specifically addressed in the emoloyment oolicies of the early plans. The Fifth Five-Year Plan, however, provided manpower projections by gender. Specifically, the - 86 - male labor supply was projected to increase by over 20% between 1980-89, whlle the female labor supply remalned unchanged. The current plan provides no breakdown by gender. Both the Fifth and Sixth plans, however, note that increased female *mployment will increase the demand for day-care and that growth in such facliltles is to be encouraged. 5.06 in accordance wlth ;onstitutlon, whlch states that everybody has the rlght to social insura.. ne Sixth Plan has as an objective, increasLng the proportion of the . 2or force covered by so-ial insurance, and of women explicitly. As noted in Chapter II, approximately 6% of the female labor force is covered by the Soclal Insurance Institute, compared to 22% of the male labor force. The agricultural labor force, however, where three- quarters of employed women work, is largely outside the social insurance system, nor ls it covered by labor legislation. The majority of working women are, therefore, without such protection. Notwithstanding the Plan objectives, little progress has been realized, to date. 5.07 The current Plan's broad objectives, include the "protection and education of women..." (para. 29). More speclfically, the Plan states that "the appropriate environment will be created for promoting the employment of women in non-agricultural sectors by raising their education levels and providing women wider opportunities for vocational training" (para. 766). Unlike the Flfth Plan, however, no projections of the labor supply by gender are provided. "Measures will be developed to eliminate the difficulties that workLng women encounter in such areas as social insurance, health, vocational training and re-employment and increased facilities for childcaro services for them will be encouraged" (para. 767). To date, however, little has been done to further these objectives. The 1991 Annual Program contained three specific women's objectives, namelys to design measures to improve women's education and to facliLtate their employment in sectors other than agriculture; to initiate a system for collecting gender-diuaggregated data; and to determine the problems women encounter in their working life and the use made of their legal rights. To date, little progress has been realized in implementing these measures. C. Curren Role gf Governmen in Women's D_velogMent 5.08 Following several decades of relative quiescence, women's role in Turkey became once again a central issue in the mid-1980s, in large part as a result of two diametrically opposed groups - the religious, conservative groups and the secularist women. For the former, the issue concerned women's role withLn the famLly, as mother and wife, although publicly the debate was over the right of women to wear headscarves in educational institutions and government offlces. Simultaneously, proposals by some conservative groups to segregate men and women in public life, Lncluding hospitals and schools, heightened the debate. The women's group addressed the issue through its focu on women's rlghts and articles of the Constitution and the Civil and Criminal Codes that dlscriminate against women. While the women's movement remained largely informal and apolitical, the conservatives through their lnfluence in the then ruling party, acquired considerable influence through early 1991. - 87 - 5 09 In response to these two movements, Government in the second half of the decade took a number of steps to address women's issues. These included ratification of two LnternatLonal legal conventlons; a speciflo focus on women's Lssues Ln the Sixth Plan and the establishment of a Government Dlrectorate to address the role of women. 5.10 Spolflcally, at the end of the Unlted Nations Decade for Women ln 1985, the Government ratlfled the UN Conventlon on the Elimination of All Forms of DLscriminatLon AgaLnst Women (CEDAW), whlch commltted OT to remove any logislatlon dlscrimlnatory to women, to promote qualLty between men and women and to Lntrodu(. measures to elimLnitte customary practlces that relnforce the subordLnation of women to men. In 1989, GT also signed the European SocLal Charter of the EC. Both measures were ratLfLed wlth a number of reservatlons (Annex 1, Sectlon 0). In December 1989, ln !6sponse to the conservatlve movement, the Government established the Famlly Research Institute (FRI) under the Prlme Minlitry. The FRI was charged wlth preservLng and strengthenLng the Turklih famLly, whlch was seen as threatened by social and economlc development. In Aprll 1990 the Dlrectorate General for Women's Status and Problms (DGWSP) was established under the MinLatry of Labor (para. 5.16 below) wlth the sole task of Lmproving women's status and promoting thelr full integratlon lnto the economy. 5.11 Taken together, these developments suggest an uncertain osmmLtment of the prevLous Government to the role of women. In view of their significance to the role of women over the coming decade, these legal and institutional developments are revLewed briefly below. (i) EDAW 5.12 The ConventLon was adopted by the UN General Assembly ln 1979 and went into effect ln 1981t Turkey was the 83rd country to ratify ln 1985. Together wlth Turkey's human rlghts record, women's status has always been perceLved as an important determLaant of RC membership and the ratification of C_DAW is seen, ln part, as an attempt to strengthen Turkey's applicatlon. As with other UN conventlons, governments are allowed to place reservatLons on artLeles lf they conflict wlth existlng natlonal legLilatLon, wlth the expectatLon that governments will qulckly comply. The Unlted Nations, however, has no means of ensurLng compliance other than applying public presoure lf a country's periodic lmplementatlon reports reveal Lnadequate progress. 5.13 Turkey along wlth almost half of the CEDAW slgnatorLes, submLtted several reservations to specLfLi articles which conflict wlth the TurkLih clvll Code.iV One of the major issues concern the definition ln the Civil Code of the husband as the head of the famLly. As such, the husband chooses the place of domicile, li responasble for the economlc support of the family and ln the case of dlsagxexement, the husband's view prevails Ln all matters V/ Until November 1990, when the ConstLtutional Court ruled lt unconstitutional, marrled women could engage ln pald work only wlth thelr husband's permiosson (ArtLcle 159). - 88 - over issues relatLng to the household and famlly. In addltion, there are numerous contradletlons (Annex 1 Sectlon C) wlth Turkey's labor leglslatlon. 5.14 The lnltlal Government progress report to CEDAW (September 1989) states that work on amendments to the Clvll Code that would remove the reservatLonb is underway. The legal commlsslon appointed to revlew the Civil Code, however, submitted lts recommendatlons to the MinLitry of JustLce ln 1984 but the recommendations have not "een brought before the Natlonal Assembly, notwLthatandLng pressure from womens organLzatLons. The newly appolnted Government, however, li plannlng an early consideratlon of the commLssion's proposals. (LL) lurogean Sagial Charter 5.15 In 1989, the Government also became a slgnatory to the European SocLal Charter, placLng a reservatLon on numerous articles. Article 8, on the rlght of employed women to protectlon presents the major lesue for women (see Annex I for details). Progress in compliance wLth the Charter, and the Charter's Actlon Program that li close to beLng fLnalized wlth far-reachLng labor recommendatlons, will be laportant for Turkey'" acceptance to the BC. (LLL) Tit Directorat genergal fog the Status and Problems of Women 5.16 The DGWSP was authorised by decree under the Prime MinLitry ln AprlU 1990 and subsequently by Law 3670 under the Mlnlitry of Labor on October 25, 1990. In June 1991, the newly appolnted Cabinet re-instated the Directorate wLthln the Prlme Mlnletry and ln November 1991 the Dlrectorate was placed under the newly created Mlnlstry of State for Women, Famlly and Social ServLees. The Dlrectorate General was established, ln large part, as a result of efforts of the then Miniater of Labor, the only woman ln the cabinet, and pressure from women's groups. The goals of the Dlrectorate General lnelude improvLng the educatlonal level of TurkLsh women; lncreaslng their employment in productive sectors and thereby strengthenLng their economlc Lndependencea ensurlng the legal, health and soclal rlghts (LncludLng soclal lnsurance) of women; and promoting the equallty of women ln the soclal, economle, cultural and polltical arenas. 5.17 To realLse these goals, the DGWSP li to develop and lmplement programs and polLoLes ln line wlth Ataturk's legacy to enhance women's status; act as an advocacy center for women; coordlnate, direct and support the actlvltles of both government and non-governmental organLzatLons and regLonal authorlties responslble for addressLng lesues and problems concernlng women; undertake research and develop a databaee to underpln efforts to improve the status of women, includLng thelr legal rlghts; undertake studles of women's Lseue ln the Buropean communlty and other countrLes; and to "contrlbute and partLcLpate in studies implemnted by lnternatlonal organLzatLons. .on women's issues'. The Directorate is also charged wlth ensurlng compliance wlth LnternatLonal conventlons on women's Lesues. 5.18 Wlth the establishment of the Mlnistry of State for Women, the DoWSP ls now actlvely purouLng lts objectlves of developing programs and polliles and consLderLag means of successfully lntegratlng women's lssues Lnto - 89 - the llne mLinstrLes. Specifically, the DOWWP will ensure progress through polliles and programs to lmprove the health, soclal, eoonomlc and educatlonal posltlon of women. The DGWSP is belng asctLvely supported ln lts actlvltles by a recently agreed fLve-year cooperatLve program wlth UNDP. (lv) Family Research InstLitute 5.19 The fLrst step towards lmplementlng the recomendatLons of the SPO report was the establishment by decree of the Family Research InstLtute (FRI) ln December 1989 under the Prlme Mlnlitry. The Decree states ;'t the objective of the Institute li to channel lncreased public aw-- as on family problems, undertake research on the effeots of economlc, sccLU.- and cultural change on the famLly and to "carry out research on the effects of famlly planning and to develop a natlonal pollcy on family planaLng". These actlvltles are to be implemented "Ln accordance wlth exlstlng laws, natlonal securlty pollcy .... 5.20 In contrast to the DLrectorate General, the InstLtute was both well funded and staffed. By March 1991, the Institute was fully functionLng wlth 75 higher level staff on board, an actlve research program and a full schedule of conferences and seminars and media events programmed. In addition, the Institute had established ln conjunctLon wlth Halk Bank the Family Credlt Program to provLde credit to famliLes to engage ln lncome- generating activltles withln the home (Chapter III). By mld-1991, 2700 credits had been extended, 85% to women and the majorlty for carpet weaving. Homebased productlon was advocated by the FRI on the grounds that lt mLnlmLzes conflicts wlth women's domestLc responsLbliLtLes and, ln particular, wlth chlldeare. D. Non-Governmental OroanLatLOns (NaOsm 5.21 At the turn of the century ln Turkey, there were several actlve and articulate groups, such as the Red-Whlte and the Ottoman Assoclatlon of Women's Rights, calling for the elimination of discrimination ln the famlly, laws, career development and social norms. With the grantLng of rights, the openlng up of the professLons and the gender-neutrallty of government polLeLes and plans, women's groups Ln Turkey became less actlve until the mLd-1980s. The 1982 ConstitutLon which forbade the assocLatLon of youths and women also reduced thelr effectLveness to respond to issues in the 1980s. In part as a result of the austerLty measurer -f the early 1980s, but also as a result of the resurgence of conservatism %_thin Turkey and the focus on women's Lesues Lnternationally, women's groups are again becoming Lncreasingly actlve. 5.22 In 1991, there were approximately 25 regLstered NOOs with an orLentation towards women's issues. As of July 1992, there are now over 50. For the most part, these are organizations of profeseLonal groups (jurLsts, academics, nurses or village teachers), cultural groups (Turkish/Japanese, TurkLah/American women), or charitable organizations (protection of chlldren, assistance to the sick and disabled, womens association against separatLon). The NGOs most involved ln addreosLng Lesues relatlng to the LntegratLon and status of women ares - 90 - (i> Th Natlonal aounaal of Turkiph woman 5.23 The Councll was founded ln 1959 and lts amAs Lnclude addresLing lesues of women as mother, wLfe, cLtLien and lndLvldual. The Councll li Lnvolved ln partiLular ln addresiLng the problems of working women and ln promotLng womn's educatLon and ln advocatlng legal reform. To thls end, the Councll provldes some fLnanclal support to women of limlted means, initiates programs to promote the eduoatLon of women and provldes some scholarshlps. The Councll brLngs attentlon to women's issue through the organlzatLon of conferenacs, seminars and through lts publicatLons. its work is carrlid out through oom 20 sub-ccomLttees focualng, Lnter alla, on Lisues such an the law, status of women, women and employment, mLgratLon and the medla. The Coun.i. Lnoludes amongst lts membership a number of women who were actlve ln the 1930. with Ataturk and later ln the 1950s wlth the women's movement. Many of these women have made a slgniflcant contrlbutlon to the standLng of women and, most notably, through thelr efforts at legal reform. Nonetheless, the Moo li very much an Ankara/Istanbul organliatLon of professional and Lntellectual women. Although lt maLntaLns reglonal offloes, including some ln the south-eastern areas, tho organlzatLon does not have a grass roots focus, nor does Lt appear to have a slgnifLcant followLng amongst the current generation of women. The National Councll le affliLated to the InternatLonal Councll of Women and the Nuropean Center for the XnternatLonal Councll of Vomeng through these organliatLons the Counoll malntalns close contact wlth both the ON and the 3r. (LL) The Turkish Federation of Women's !nutituteg 5.24 Founded ln 1976, thli le an umbrella organizatLon for some 10 to 12 women's groups. The membershlp of the governlng board of the Federatlon appears to overlap wlth that of the Natlonal Council, as do many of lts activities. (ill) The Poundaton for Sunnort for Contemporary Llvina 5.25 Nstablshed in the mid-1980s, the Foundatlon was eotabllshed to address and allevlate the problems faced by women today and to advocate legal reform and change. The Foundation Lncludes among its members prominent academics and professionals in the major clties. (iLL) The Foundation for the Advancement and RecoanLtion of Turklih Women 5.26 The Foundatlon was established ln April 1986 wlth the objective of supportlng women in all walks of life. specifLcally, the Foundatlon iB working to strengthen the economlc posLtion of women through generating additLonal means of obtaining income, increasing thelr skill levels and establishLng marketing organLzatLons for thelr production. The Foundatlon's work is accompliabed through 12 activltLes groups covering topics such as LnternatLonal relatLons, press and promotion, agrlcultural work, tourlsm, handLerafts and health. The Foundation has realized a considerable measure of success ln its actlvltles ln large part because of Lto fund raisLng *kills and medla ablilty. The" actlvltles have been further facilitated by the establlhment of som 70 regional offlces throughout most of the country. - 91 - Major activities include: - The Foundation has also been active in the area of woun's health through the establishment of mobLie health buses, each of whlch are staffed with a gynecologist, nurse and support personnel to provlde gynecological services and instruction ln blrth control in the remoter areas to families of limited incomes. More recently, the Foundation has opened mother/child health centers in prLorLty rural areas; - Recognizing the importance of educatlon, the Foundation has been active in providing scholarships for talented and needy girls, setting up vocational training courses and establishing and operating boarding houses for female students who need a place to stayJ 5.27 Although aot a women's organization per so, the Turkish Development Foundation (TDF) administers income-generating projects directed at both men and women in a number of areas in the priority development reglons. The Foundation has been particularly innovative and active in involving villagers in the design and implementation of their projects. The TDF is developing a project concept to strengthen women's position ln rural communlties p:imarily through the establishment of an animator system, by which women with strong leadership qualities within the villages are recruited and trained to act as a catalyst in overcoming traditional attitudes. B. Conclusions and Areas for Intervention 5.28 The institutional measures adopted by the previous Government reflected the prevailing incongruent views on women's issues. Beginning wlth the delays in ratification of CEDAW and the amendments to the Turkish Civil Code that would have brought Turkey in compliance with CEDAW, and the establishment of the Directorate General for Women simultaneously with the Family Research Institute, the measures suggested that women's rights and status were pursued, at least in part, as a means of attalning other objectives of government policy. The establishment of the Ministry of State for Women is a very positive step that sends clear signals as to Government's intentions to st.rxngthen the standing of women. However, the DGWSP remains understaffed and underfinanced, handicapped by the relocations and needs still to develop a clear agenda. Support to this end, however, is being provided by a recently agreed UNDP project and is expected to assist the DGWSP in becoming fully operational. 5.29 Effective institutions and procedures are needed to monitor and Improve the situation of women in Turkey; to identlfy the causes (both traditional and new) of discrimination; to assist in the formulation of policies and implementation of measures to end discrimination. These policles and procedures need to be an integral part of Turkey'e development strategy. As noted in earlier chapters, the alleviation of dLscriminatLon and constraints to women's full participation will require both enforcement and evaluation of legislation, the education of the entire population as to the contribution of women through both formal and non-formal channels, Lncluding government agencies, the media, non-governmental organizations, political - 92 - party platforms and executlve action. To be effective, the DOWSP need to be assured of adequate resources, commltment and authority. Wlth this, the DGWSP can play a. vltal role in improving the posltlon of women through formulating a clear agenda; collaboratlve action with the llne mlnlstries and other agencies, both governmental and non-governmental; in undertaking research both to increase awareness of women's contribution and to increase their productlvlty; and dissemlnatlng lnformation on women's rights and entitlements. 5.30 In light of experlence wlth natlonal machineries elsewhere in the 1980., the following measures should be considered. o Ensure that all line mlnlstrles and agencles see women's development as an integral and prlority element of national develooment and therefore as part of their regular responsibliltles - not a responsibility that can be passed on to the Ministry of State for Women. o Establish an inter-ministerial committee chaired at least inltially by the Prime Mlnlster, wlth representatives from the line mlnlstries and agencies. Each representative should present a timebound actlon plan for promoting the welfare and productivity of women in their minlstry or agency and be required to report progress in its implementation to the committee on a regular basis. o Establioh women's units at the highest levels wlthin the key line ministries (for instance, Labor, Education, Agriculture), responsible for implementing and monitoring the action plans and for ensuring that women's iesues are integrated into all ministry programs and policies. o Ensure that people of the highest caliber are appointed to key pooitions involved with the formulation, implementation and monltoring of policies and programs to improve the status of women. o Ensure a sustained effort to raise the visibility of women's issues, in general, and their contribution, in particular. This can be realized through, inter ala, moral suasion (media campaigns, revision of curricula and textbooks, TV, etc.), publication of an annual statistical report on the status of women and undertaking and publishing a broad-based program of research on women's issues. . - 93 - References and Selected BiblioaraDhv of Regent Studies Abadan-Unat, N. (ed.), (1981). Women in Turkish Society, Leiden, E.J. Brill Press. Abadan-Unat, N., (1986). Women in the Develoging Worlds Evidence from Turkey, Denver, University of Denver. Afetinan, A.,. (1962). The Emancipation of the Turkish Woman, UNESCO. Akadli, E.8. (1990). "Unmet need for Family Planning in Turkey". The Journal of Ponulation Studies, Vol. 11, Hacettepe University, Institute of Population Studies, Ankara. Akqit, B.T. and B. Akqit, (1989). "Sociocultural Determinants of Infant and Child Mortality in Turkey". Social Science and Medicine. Vol.28 No.6, pp.571-576, Ankara. Aksit, B.T. and H. 6zcebe, (1991). "Women's Health in Turkeyt Review of Literature and Discussion of Research Gaps in the Status Context". Paper to be published in a book, Women in Turkey. (Ede. C. Balim and A.G. Ayata), Cambridge University Press. (forthcoming in 1992) Amsden, A.A., (1980). The Economic. of Women and Work, New York, St. Martins Press. Anker, R., (1980). Research on Women's roles and Demooraghic Changet Sgry Ouestionnaires for Households. Women, Men and Communities with Back Zxglanatoi2n, ILO. Arriagada A-M., (1990). "Labor Market Outcomes of Non-formal Training for Male and Female Workers in Peru," Economics of Education Review, 9, 1, 331-342. Behrman, J.R. and B. Wolfe, (1984). "Labor Force Participation and Earnings Determinants for Women In the Special Conditions of Developing Countries," Journal of Develooment Economics, 46, 259-288. Bergmann, B.R., (1986). The Economic Emeroence of Women, New York, Basic Books, Inc. Bhasin, K., (no data). "Agriculture Without Woment A Scenario in the Making." FAO Report. Blau, F.D. and M. A. Ferber, (1990). Women's Work. Women's Lives: A Comoarative Economic Persoective, National Bureau of Economic Research, Working Paper No. 3447, Cambridge, Mass. Blumberg, R. L., (1988). Income Under emale vs. Male Control, Mimeo, University of California. Boratar, K., & Galip Y, (1989). "A Study on the Political Economy of Structural Adjustment, Workers & Peasants During a Major Reorientation of - 94 - Economic PolLcies-Turkey, 1980-87", Himeo, ieuearch paper prepared by lDRC, Ottawa, Canada.Bellew, Rosemary and Elizabeth M. King, (1991). 2Promtina Girls' and Women's Education: Lessons From the Past. World Bank PRE Working paper No. 715. Boserup, X., (1970). Women's Role in Economic DeveloRment, New York, St. Martin's Press. Boulding, B., (1983). "Measures of Women's Work in the Third World: Problems and Suggestions" in M. BUIbUl, A. S., (1990). "Educational & Training Level of Turkish Adult Women Affecting their Access to and kind of Employment", Mimeo, prepared for ILO, Ankara. Buvinic, M.A. Lycette and W. P. McGreevey (eds.), Women and Poverty in the Third World, Baltimore, Johns Hopkins University Press. Buvinic, M., M.A. Lycette and W.P. McGreevey (eds.), Women and Poverty in the Thid Wold, Baltimore, Johns Hopkins University Press. cagatay, N. and 0. Berlk, (1990). "From Import-Substituting Industrialization to Export-Led Growth in Turkey: Feminization of Employment?", mimeo, New York, Department of Economics, New School for Social Research. Charlton, S.M., (1984). Women in Third World Develonment, Westview Press, Boulder, Colorado. Vitgi, O., (1988). "The Quantitative View of Public Workers", Amm Cochrane, S.H., (1988). The Effects of EducAtion. Health and Social Security on Fertility in Develonina Countries: Their implications for Policy, World Bank. Commonwealth Secretariat, (1989). Engendering Adiustment for the 1990s: Renort of a 'Commonwealth Expert GrOuo on Women and Structural Adiustment, London. lonway, P.J., (1987). Economic Shocks and Structural Adiustments: Turkey after,127. . North Holland, Amsterdam. Davidson, N.J. and C.L. Cooper (eds.), (1984). Workino Women: An MItegnational Survey, New York, John Wiley. Dixon, R.B., (1978). Rural Women at Work: Strategies for Develooment in South Asia. John Hopkins University. Press for Resources for the Future, Baltimore and London. Dogramaci,-3., (1989). Status of Women in Turkey, (3rd edition), Ankara, Meteksan Co. Inc. Duncan O.D. and B. Duncan, (1955). "A Methodological Analysis of Segregation Indexes," American Socioloaical Review, 20, 2, 210-17. - 95 - Durand, J.D., (1975). The Labor Force in.Economic Develooments A Comoarison of International Censua Data. 1946-1966, Princeton, Princeton University Press. Ertlirk, Y., (1988). "Women's Participation in Agrlculture ln the Villages of Erzerum, Turkey", prepared for FAO. ErtUrk, Y., (1989). "Rural Social Structure and the Promotlon of Community Particlpatlon ln Development". Paper prepared for IFAD, Rome, Italy. European Communlties - Commlssion, (1990). Community Charter of the Fundamental Social Riahts of Workers, Luxembourg, Office for Official Publications of the European Communitles. Feldstein, H.8. and Poats, S., (1990). Workina Toaethers Gender Analysis in Aqriculture, Kumarian Press, Connecticut. Fuchs, V.R., (1988). Women's ouest for Economic Eaualitv, Cambrldge, Mass, Harvard University Press. Goldin, C.D., (1991). "The Role of World War II ln the Rise of Women's Employment", Amerlcan Economic Review, 81, 4, 741-7 Goldin, C.D., (1989). "Life-Cycle Labor-Force Participatlon of Marrie Women: HLstorical Evidence and Implications", Journal of Labor Economics, 7, 11, 20- 47. Goldin, C.D., (1986). "The Female Labor Force and American Economic Growth, 1890-1980" and comments by S.B. Carter, in S.L. Engerman and R.E. Gallman (eds), Lona-Term Factors in American Economic Growth, Chicago, University of Chicago Press, 557-604. GoldLn, C.D., (1983). "The Changing Economic Role of Women: A Quantitative Approach:, Journal of Interdisciolinarv HLstory, 13, 41, 707-33. G8nen, E. and Hablemltoglu, S., (1989). "The Socio-Economic Status of Women in Turkey", The Turkish Review, Summer, pp. 143-150. Government of Turkey (GT), DG for Status & Problems of Women, MLn. of Labor, (May 1990). -Women's Participation in Education & Training System in Turkey", Gul Aykor, Ankara. GT, D.G. for Status & Problems of Women, Min. of Labor,. (May 1990). "The Status & Problems of Women Workers in Turkey", Gul Aykor, Ankara. GT, "Initial Report of Turkey on CEDAW", September 1, 1989 and "Replies and Explanations", January 31, 1990. GT, (Dec. 1990). Toward the Year 2000: 1991 National Education Report to Paliament, Avnl Akyol, MlnLter of Education. - 96 - OT, SIBK (Turkish Employment Organization) and OECD, (1990). kmorovin0 Bmuloyment Prosgects for Women in a Changing Society, Conference Proceedings, Istanbul, November 7-8, 1991. Ankara, IIBK. GT, UNICEF (1991). The Situation Analysis of Mothers and Children in Turkey, Ankara. OT, State Institute of Statistics (SIS), (n.d.), 1980 Census of Aqricultures Result of Household Survey, Ankara. GT, 8IS (1989). Statistical Yearbook of Turkey - 1989, Ankara. GT, SIS, (1991). Statistical Pocketbook of Turkey - 1990, Ankara. OT, SI8 (1991). Household Labor Torce Survey Results - ADril 1990, Ankara. GT, SIS (1991). Household Labor Force Survey Results - October 1989, Ankara. OT, 8I8 (1991). Household Labor Force Survey Results - April 1989, Ankara. GT, 8I8 (1990). Household Labor Force Survey Results - October 1988, Ankara. 0T, SIS (1989). Census of Population. 1985: Social and Economic Characteristics , Ankara. GT, SIS, 1990. "Maternal Deaths" unpublished research, Ankara. CT, Social Insurance Institution (81I), (1990). Statistical Yearbook - 1990, Ankara. 0'T, SPO, (1990). Social Indicators: Turkey j990), Ankara. 0GT, SPO, (1989). Tugkish Population Projections for the Sixth Five-Year Develonment Plan 1199Q-1994). GT, SPO (1985). Women in Turkey: Country Revort, Ankara. Hacettepe University Institute of Population Studies (HIPS), (1989). 1 Turkish Po2ulation and Health Survey, Ankara. Hacettepe University Institute of Population Studies (HIPS), (1987). 1 Turkish Population and Health Survey, Ankara. Hansen, B., (1989). "Unemployment, Migration and Wages in Turkey, 1962-85." World Bank Staff Working Papers, no. 230. Washington, D.C., World Bank. Herz, S. and Measham, A.R., (1987). "The Safe Motherhood Initiative: Proposals for Action." World Bank Discussion Paper. Washington, D.C., World Bank. House, W.J., (1985). Cvvriot Women In the Labor Market, Geneva, ILO. - 97 - International Labor Office, (ILO), (1991). Status of the Rural Women in Turkey, Ankara. ILO, (1990). Yearbook of Labor Statistics: Retrospective Edition on Pooulation Censusess 1945-89, Geneva. Jacobson, J.L., (1990). "The Status of Family Planning in Developing Countries". In (Ed. H. M. Wallace and K. Glri). Health Care of Women and Children in Develoyina Countries. Third Party Publishing Company. Oakland, California. Joekes, S.P., (1987). Women in the World Economy, Oxford, Oxford University Press. Juster, F.T. and F.P. Stafford, (1990). The Allocation of Times Emoirical Findinas. Behavioral Models. and Problems of Measurement, Institute for Social Research, Working Paper, Ann Arbor, Michigan, University of Michigan. Kagitgibasi, C., (1981). "Value of Children:. Some Social-Psychological Determinants of Fertility in Turkey". In (Ed. N. Abadan Unat) Women in Turkish Society. pp.74-95. Lieden. Kagitcibasi, C., (Nov. 1989). "Women's Intra-family Status, Education & Employment Turkey", paper prepared for OECD & Turkish Employment Organization Continuance on Improving Employment Prospects for Women in a Changing Societys The Years Ahead, Istanbul. xhandker, 8.R., (1990). Labor Market Particigation. Returns to Education. and Male-emale Wage Differences in Peru, PRE Working Paper Series, No. 461, Population and Human Resources Department, Women in Development Division, Washington, D.C., The World Bank. Killingsworth, M.R. and J.J. Heckman, (1986). "Female Labor Supply: A Survey", in 0. Ashenfelter and R. Layard (eds.), Handbook of Labor Economics, New York: Elsevier Science Publishers, pp.103-204. King, E.M., (1990). Does Education Pay in the Labor Market? The Labor Force Partigipation, Occupation. and Earnings of Peruvian Women, Living Standards Measurement Study, Working Paper, No. 67, Population and Human Resources Department, Welfare and Human Resources Division, Washington, D.C., The World Bank. Kottis, A.P., (1990). "Shifts Over Time and Regional Variation in Women's Labor Force Participation Rates in a Developing Country: The Case of Greece," Journal of Development Economics, 33, 117-132. Larre, B. and R. Torres, (1991). "Is Convergence a Spontaneous Process? The Experience of Spain, Portugal and Greece," OECD Economic Studies, 16, 169-198. - 98 - Layard, R. and J. Mincer, (ode). (1985). Tgends in Women' Work. Education. and Familv Buildina, Conference Proceedings, Chelwood Gate, Sussex, England, May 31-June 3, 1983, published in Journal of Labor Economics, Vol.3, No. 1, Part 2. Chicago, University of Chicago Press. Leslie, J. et al., (1986). Weatherino Economic Crisis: The Crucial Role of Women in Health. Inter..atlonal Center for Research on Women. Washlngton, D.C. Lockheed, Marlaine E. Adriaan M. Verspoor, (1990). Prlmary Education in Develocino Countries: A review of Policy ODtLons. Background Report for the World Conference on EducatLon for All Bangkok, p. 109. MacKinnon-Scott, K., (1991a). "Women Ln the Labor Force in Bolivia: Participation and Earnlngs," Mimeo, Latin America and Caribbean Region Technlcal Department, Washington, D.C., The World Bank. MacKinnon-Scott, E., (1991b). "Female Labor Force Participation and Earninges The Case of Jamaica," Mimeo, Latin America and Caribbean Region Technical Department, Washington, D.C., The World Bank. M1artLn, P.L., (1990). International Migration for Employment: an Assesoment of Turkish Labor Miaration to Western Euro2e. Mazumdar, D., (August 1976). "The Urban Informal Sector," World Develo2ment Report. Washington, D.C., World Bank. Mlncer, Jacob, (1974). Schooling. ExPerience & Earninaos, Columbia University Press, I.Y.. Moffltt, R., (ad.), (1990). Taxation and Labor Sup2ly in Industrial CountrLes, Conference ProceedLngs, Racine, Wisconsin, published in Journal of Human Resources, Vol. 25, No. 3, Part 2. Madison, Unlverslty of Wlsconsin Press. Mohan, R., (1986). Work, Waces. and Welfare in a Develoolna MetropolLes Conscuencgs of Growth in Boaota. Colombia, New York, Oxford University Press for the World Bank. Mueller, E., (1983). "Measuring Women's Poverty in Developing Countries" in M. Buvinlc, M.A. Lycette and W. P. McGreevey (ede.), Women and PovertX in the ThLrd Wrld, Baltimore, Johns Hopkino University Press. Mg, Ying Chu, (1990). "Female Labor Force Participation and Gender Earnings Differentials in ArgentLna," Nimeo, Latln America and Caribbean Region Technical Department, Washington, D.C., The World Bank. Nieva, V.F., (l985). "Work and Family Linkages," Women and Work, 1, 1962- 1990. - 99 - Muse, S., with E. Denti and D. Viry, (1989). Women in the World Of Work'" Statistical Analysis and Proiections to the Year 2000. Geneva, ILO. Oaxaca, R., (1973). "Male-Female Wage Differentials in Urban Labor Markets", International Economic Review, 1673, pp. 693-709. OECD, (1991). Sconomic Surveys. 1990/1991 - Turkey, Paris. OECD, (1990). Economic g tlook - Historical Statistics: 1960-1989, Paris. OECD, (1989). Emolovment Outlook. 198, Paris. OECD, (1985). The Integration of Women into the Economy, Paris. OECD, (1975). The Role of Women in the Economy, Paris. dzcebe, H., (1990). "A Study on Factors Effecting the Usage of Curative Health Care in Etimesgut Health Center Area". Unpublished Specialization thesis. Hacettepe University, Department of Public Health, Ankara. Pampel, P.C. and K. Tanaka, (1986). "Economic Development and Female Labor Force Participation: A Reconsideration", Social Forces, 64, 3, 599-619. Presser, H.8. and S. Kishon, (1991). "Economic Development and Occupational Sex Segregation in Puerto Rico: 1950-80", pooulation and Develooment ReyLew, 17, 1, 53-85. Psacharopoulos, G. (ed.), (1987). Economics of Educationt Research and Studies, New York, Pergamon Press. Psacharopoulos, G. and Z. Tzannatos, (1989). "Female Labor Force Participation: An International Perspective," The World Bank Research Obsleaer, 4, 2, 187-201. Saglam, F., (1989). "The Prevalence of Obesity in Women and the Factors Affecting" Jburnal of Nutrition and Dietetics. Vol. 18, No.2. Schultz, T.P., (1989). "Returns to Women's Education." Women in Development Division Background Paper, no. PHRWD/89/001. Washington, D.C., World Bank. Schultz, T.P., (1989). Women and Develooment: ObiectLves. Frameworks, and Policy Interventions, Working Paper, No. 200, Population and Human Resources Department, Women in Development Division, Washington, D.C., The World Bank. Sen, A.K., (1989). Women's Survival as a DovoloQment Problem. Harvard University, Mimeo. Shabon, A.M. and Zeytinoglu, I.U., (1985). The Political. Economic and Labor Climate in Turkey, Industrial Research Unit, Wharton School, University of Pennsylvania. - 100 - Sirman, N., (1991). "Rural Women ln the Process of Developments Approaches and Issues", ILO -TDF. Ankara. Sorrentino, C., (1983). "International Comparisons of Labor Force Participation, 1960-1981," Monthly Labor Review, 106, 2, 23-36. Standing, G. and a. Sheenhan (ads.), (1978). Labor Force ParticiyatLon in Low-Tncome Countries, Geneva, International Labor Office. statistics Canada, (1990). Women in Canadas A Statistical Revort, Ottawas Statistics Canada, Houslng, Family and Social statistics Division. Stelcner, M., J.B. Smith, J.A. Breslaw, and G. Monette (1991). "Labor Force Behavior and Earnings of Brazilian Women and Men, Mlimeo, Latin America and Caribbean Region Technical Department, Washington, D.C., The World Bank. Stichter, S. and, J.L. Parpart (eds.),(1990). Women4 Emolovment and the Family in the International Division of Labor, Philadelphia: Temple University Press. Terrell, T., (1989). "An Analysis of the Wage Structure in Guatemala City," Journal of Develo&ina Areas, 23, 405-424. Toros, A. (ed.), (1990). The Study of Female Particioation in the Formal Sectors Turkey, Report Submitted to the World Bank, mimeo, Ankara: Bigtas. UN, (1984). Comoiling social Indicators on the Situation of Women, Series F, No. 32, UN, NY. UN, (1989). 1989 World Survey on the Role of Women in Development, UN, NY. UN, (1990a). Miethods of Measurina Women's Particgiation and Production in the Informal Sector, UN, NY. UN, (1990b). Handbook for National Statistical Data Bases on Women and ngLrev1MMnt, UN, NY. UN, (1991). The World's Women: Trends and Statistics. 1970-1990. UN, NY. Ward, K.B. and F.C. Pampel, (1985). "Structural Determinants of Female Labor Force Participation in Developed Nations, 1955-1975", Social Science 2uarterly, 66, 3, 654-67. Women's Bureau, (1990). Women in the Labor Force. 1990-91 Edition. Ottawa: Women's Bureau, Labor Canada. World Bank, World Develogment Re2ort, Washington, D.C. (various annual issues). Yayan, V., (1984). The Role of Women in the Turkish Economy, Ankara, SPO.