Report No. 24185-TR Turkey Poverty and Coping After Crises (In Two Volumes) Volume II: Background Papers July 28, 2003 Human Development Unit Europe and Central Asia Region Document of the World Bank Turkey: Poverty and Coping After Crises CurrencyEquivalents ExchangeRate Effective= September 1,2001 CurrencyUnit =Turkish Lira (TL) US$ 1=TL 1,500,000 GovernmentFiscalYear January 1-December31 Weights and Measures Metric System Abbreviationsand Acronyms BK Bag-Kur CPI ConsumerPrice Index DIE StateInstitute of Statistics ES Emekli Sandigi FA0 (UN)Food andAgnculturalOrganization HCIS Household Consumptionand Income Survey HIES Household Income and Expenditure Survey LSA Living StandardsAssessment PPP PurchasingPowerParity SRMP Social Risk Mitigation Project SSK Sosyal SigortalarKurumu SHCEK Social Services and Child Protection Organization SYDTF Social SolidarityFund SYDV(s) Social SolidarityFoundation(s) Vice President: Johannes Linn (ECAVP) CountryDirector: Ajay Chhibber(ECCU6) Sector Director: Annette Dixon (ECSHD) Sector Manager: Michal Rutkowski (ECSHD) Team Members: Jeanine Braithwaite (Team Leader), John Innes, Jennifer Manghinang (ECSHD); Mathew Verghis (ECSPE); Diane Steele (DECRG); Ayge Ayata, Sencer Ayata, Tahire Erman, Kivilcim Ozcan, Y. Ziya Ozcan (Consultants). Turkey: Povertyand CopingAfter Crises Volume Two: Background Papers 1. The Economics o fPoverty. KivilcimMetinOzcan 1 2. Poverty inTurkey. Kivilcim Metin Ozcan 28 3. Poverty inTurkey; The Social Dimension. Tahire Erman 42 4. Measuring Poverty and Inequality inTurkey'. Yusuf Ziya Ozcan 72 5. The Benefit Dependent and the Regular Income Earning Poor: The 104 Analysis o fthe Interview Data. Sencer Ayata and Ayve Guneq Ayata. 6. Interim Report: Assessment o f Social Solidarity FundBeneficiaries 150 Middle Eastern Technical University, Department o f Sociology. .. 11 1. The Economics of Poverty (Kivilcim Metin Ozcan) I.THEORETICALFRAMEWORK Economic growth is an important factor determining poverty reduction. Economic growth improves the living standards o f poor people. The modem economic development made it possible that countries with similar rates o f growth might have very different rates o f poverty reduction. However, this casual relationship between economic growth and poverty reduction i s not valid all the time. Indeed, the relationship among the pattern o f growth, the changes in the distribution o f income and the rates o f poverty reduction is v ery complex due t o differences in growth and the distribution o f assets, state institutions o f varying quality andtheir economic policies, and finally the history, geography and culture o f countries. Therefore, each country should formulate different poverty reduction strategies (World Bank 2000 / 2001, p. 45). Another important causal link can be set up between growth and the income and non-income poverty. Initial inequalities in the distribution o f income also have some consequences for the economic growth o f countries. Highinitial inequalities reduce the poverty impact o f a given rate o f economic growth. There are two theoretical explanations regarding the effects o f inequality on growth. The first one suggests that greater inequality is useful for growth by redistributing income to the rich, who can save from the poor, who cannot. However empirical evidences suggest the second view; that is, lower inequality can increase efficiency and growth through variety o f channels, including political instability, populist redistributive policies, volatilities in several macroeconomic aggregates. This view also suggests that economic policies can lower growth. Evidence on the impact o f inequality in assets, land, education and gender i s clear inthe literature. Some studies have found a link betweeneducation and gender inequality and growth. IEvidence on the effect o f initial income inequality on growth is mixed. While some studies have found negative effects the others have found positive ones. As income poverty declines as average, incomes increase, therefore non- income poverty such as health education, decreases as well. There i s strong evidence that better health and education outcomes contribute to faster economic growth. The greatest poverty reductions would also be achieved through policies that reduce the gap between individuals with varying levels of education and skills. The question is why some individuals are able to accumulate human and physical capital to place themselves in a more favorable relative position compared to other individuals. This i s answered in the literature using alternative household savings theories and one can conclude that several subgroups o f the population behave in totally different ways, e.g., the total 30 percent o f the distribution are not able to acquire income-earning assets because they are subject to survival constraints while the richest tenpercent accumulate to be able to invest and maximize the value o ftheir financial holdings. Effects o f economic liberalization on poverty are also addressed in the literature considerably. In fact, the hypothesis is that the liberalization process provides greater opportunities to the individuals that were initially better endowed with human and physical capital and therefore contributes to increasing inequality and poverty. Szekely (1998) shows that those individuals better 1 endowed with human and physical capital benefited disproportionately more from the economic expansion duringthe reform period." In the following part (section 11) of this chapter, the major findings from the literature on poverty and poverty reduction will be investigated for the across region and c ountries, using the economic framework given above. Then the structure o f poverty in Turkey will be summarized in Section 111. 11. THE MAJORFINDINGSFROMTHE LITERATUREONCASE STUDIESOF POVERTY A- INCOME POVERTYBYREGION The world has deep poverty. Using absolute poverty concept"', o f the world's 6 billion people, 2.8 billion live on less than $2 a day, and 1.2 billion live on less than $1 a day, with 44 percent living in South Asia. Between 1987 and 1998, the share o f population in developing and transition economies living on less than $1 a day fell from 28 percent to 24 percent (see World Bank 2001 p.23 table 1.1). Because o f highpopulation growth, the number o f poor people changes very slowly. However, there are large regional variations regarding the number of poor people. The Middle East and North Africa have reduced their numbers in poverty. East Asia has reduced it dramatically. However, in all other regions the number o f people living on less than $1 a day has risen. In South Asia the number o f poor people rose from 474 million in 1987 to 522 million in 1998, while the share o f poverty fell from 45 percent to 40 percent. In Latin America and the Caribbean the number o f poor people rose by about 2 0 percent. InEurope and C entral A sia the number o f poverty increased from 1.1 million to 24 million. In Sub-Saharan Afiica the number o f poor people increased from 217 million to 291 million over a decade. Blackburn (1994) suggests that poverty comparisons can be very sensitive to how poverty lines are made comparable across countries.'" His findings show that poverty comparisons across industrialized nations can be quite sensitive to whether an absolute or relative concept o f poverty i s considered appropriate. Therefore, estimates o f poverty based on combination o f absolute and relative poverty concepts provide much higher poverty estimates although time trends remain unchanged (see World Bank 2000/2001, p.24, Table 1.2). This fact is also valid for the Middle East andNorth Africa (MENA) and for Europe and Central Asia. Sub-SaharanAfrica 217.2 290.9 46.6 46.3 Total 1,183.2 1,198.9 28.3 24.0 ExcludingChina 879.8 noc 7 Y O J . I I 90 c L0.J I L U . L ? L ? I *Figures depend on the absolute poverty line $1.08 a day at 1993 PPP. Figures takenfrom World Bank 2000/2001 Table 1.1p.23). **Provisional 2 Table 2 Relative Income Poverty by region, selected years, 1987 and 1998" Figures takenfrom WorldBank 2000/2001 Table 1.2p.24.** SeeWorldBank 2000/2001 for the definition of poverw line ***Provisional Detailed studies using national income poverty lines present very large variations inpoverty structure across countries within each region. What are the magnitude and pattems o f poverty in the developed and developing world? How has poverty evolved over the past decade? The rest o f this chapter describes the relationship between income growth, inequality and dimensions o f poverty. Poverty reduction across region and countries is also analyzed. B - DEVELOPED COUNTRTES Inthis section, papers analyzing the antipoverty effects of growth inthe UnitedStates and a couple o f other developed countries are included. They examine the relationship between macroeconomic conditions andpoverty. Discussions about the antipoverty effects o f growth both in the States and other developed countries have largely been predicted on John Kennedy's metaphor that a rising tide lifts all boats. However, in the following papers, it is shown that growth had large antipoverty effect through the early 1970's, but that the more recent experience has been different because growth rates have slowed and inequality has increased. Gottschalk and Danziger (1985) provide a framework for evaluating the effects o f economic growth and transfer incomes on poverty using Annual March Current Population Surveys (CPS) from 1968 to 1983- each containing information on about 50.000 households collected by U.S. Department o f Commerce. Findings show that in the U S economy during the 1967-79 period, changes in transfers were almost as important as increases in market incomes for all persons. Transfers w ere 1ess important for non-aged men, but v ery important for elderly. S ince 1979, the decline in market incomes has increased poverty and the antipoverty effect o f transfer growth has declined. Rising inequality increased poverty during the period o f growth as well as during the recession. 3 Intheir follow up study, Danziger and Gottschalk (1986 ) also analyze the impactof secular and cyclical changes on poverty in the U.S economy using Census of Population Reports, for the years 1950-1980 for households headed by men aged from 25 to 64. It has been emphasized that there are limits to the antipoverty effects of improved economic conditions. Risinginequality during 1970's and 1980's was a major factor contributing to increases inpoverty. Growth has a diminishing impact as povertyrates fall. Inthe absence o f an unexpected increase inthe rate o f economic growth or an unforeseen decline in inequality, it seems unlikely that growth will substantially reduce poverty inthe near future. Sawhill (1988) examines poverty in the U.S and answers the question on why it i s so persistent? Inthis study Sawhill uses Current Population Survey, U.S. Bureau o f the Census (1987), Panel Study o f Income Dynamics, Survey o f Income and Education and Survey o f Income and Program Participation. Findings show that income poverty declined only modestly after the mid- 1960s and actually increased inrelative terms. A dverse demographic trends, high unemployment rates o f the early 1 9 8 0 ' ~failure o f average real incomes to grow very much over the past 20 years ~ are some o f the facts to explain the rather modest progress in poverty reduction. There i s considerable uncertainty about how effective the anti-poverty programs were. There was a decline in poverty among the elderly, induced by the growth o f social insurance programs. Within the non- elderly population, the distributiono f earnings became more unequal and pre-transfer poverty rates increased. Balke and Slottje (1993) study the relationship between poverty and change in the macroeconomic aggregates by estimating a dynamic macroeconometric model. They use U.S Bureau o f the Census, Current Population Reports(C.P.R) for 1959-1989 and poverty data for 1947- 1958 were obtained from an unpublished study by Gordon Fisher of the Census Bureau. Balke and Slottje (1993) found that in the U S economy the unemployment rate, transfers payments, money supply growth and inflation have significant direct contemporaneous effects on the growth rate o f poverty. Only inflation has a significant lagged impact on poverty growth. The white poverty subgroup yields results consistent with the overall poverty model. For blacks, poverty appears to be less sensitive to shocks in the economy and transfers appear to have stronger impact on black poverty over time. Pendakur (2001) examines consumption poverty in Canada, for the period o f 1969-1998. In this study, poverty rate i s estimated as t o the proportion o f individualsw h o have consumption- rather than income- lower than an absolute poverty line. As for income poverty measures, the consumption poverty rate declined over the 1970's and 1 9 8 0 ' ~in~other words, all boats rose in the risingtide. However, the 1990s show a different picture. The consumption poverty rate increased by more than half between 1992 and 1998. The situation was even worse for children. The rate o f consumption poverty among childrenmore than doubled between 1992 and 1998. Trends in China in 1996-1998 period are very sensitive to the choice o f welfare and poverty measures which based on the $1 a day line. However consumption based poverty measure shows a poverty reductionbetween 1996 and 1998, suggestingthat poor householdsandpoor people inrural areas had beensaving a large share o f their incomes (World bank 2000/2001, p.26). Chen and Wang (2001) also investigate China's economic growth and poverty reduction by 1ooking at the trends between 1990 and 1999. Poverty declined significantly in the 1990's and there was a slight increase in the poverty headcount between 1997 and 1999. Economic growth contributed significantly to poverty reduction, rising inequality worsened rural and urban income distributions. The poor benefited far less thanthe rich from the economic growth. 4 C- DEVELOPING COUNTMES I)EuropeandCentralAsia Income, inequality, and povertyduring the transition from planned t o market economy is examined by Milanovic(1998). The study covers eighteen post-communist countries, seven in Eastern Europe and eleven former Soviet states, from the period o f 1987-88 until 1996, using the best empirical data to be found, largely drawn from various World Bank studies. In the course o f transition, real per capita income increased in one single country, Slovenia, which was not part o f the Soviet Bloc. The greatest declines were recorded inMoldova and Kyrgyzstan, where per capita incomes fell by two-thirds. Personal incomes fell less than GDP, because both social transfers in cash and private incomes rose sharply. Contrary to the expectation, the share ofhealth and education inthe GDP increased by about 2 and 3 percent inbothEastern Europe and Slavic Republics ofthe former Soviet Union. Inequality increased in all countries in the region with the exception o f Slovakia. Inequality did not increase significantly on average. It was only approaching the OECD average, but the increase in inequality was very sharp, and the dispersal of income differentiation increased sharply (Aslund, 1999). Certain former Soviet countries, notably Kyrgyzstan, Russia, and Ukraine, attained Latin American levels o f inequality (Gini coefficients at 0.56 in 1996) while Slovakia remain as one o f the most equal countries in the World. In short, Eastern Europe maintains a comparatively low Europeandegree o f inequality, whereas income differentiation is greater inthe former Soviet Union. UsingUkraniandata Kakwani(1995) analyses the trends inincome inequality, welfare andpoverty. The standard o f living increased significantly inthe late 1 9 8 0 ' ~and then fell inthe 1990's. Income ~ inequality declined inthe 1980's, to rise again in 1991-92. Poverty declined over the period o f 1980- 91, from 38 percent o f the population to 9 percent. But in 1992, 30 percent o f the population was poor again. World Bank (2000/2001, p. 25) also suggested that in Europe and Central Asia the proportion o f the population living on less than $2 a day ranges from less than 5 percent inBelarus, Bulgaria, Estonia, Hungary, Lithonia, Poland and Ukraine to 19 percent inRussia, 49 percent inthe Kyrgyz Republic and 68 percent inTajikistan. It shows incredible variation from one country to the others. The main cause o f the higher inequality is wages as they are unequal everywhere, especially inthe former Soviet Union. Social transfers enhance equality inEastern Europe. However, inRussia andUkraine, social transfers are focused on wealth. Indeed, o f all non-pension cash social transfers, only 12 percent goes to the poorest quintile inRussia and 8 percent in Ukraine. Pensioners are the main winner o f the transition. They have gained more compared to workers and farmers. The unemployed and the farmers are most likely to be poor everywhere. Milanovic (1998) calculates that the total number o f the poor in eighteen countries rose twelve-fold from 14 million before the transition or about 4 percent o f the population, to 168 millionbetween 1993 and 1995. Many arejust below the poverty line and they are very well education. An analysis ofpovertyin Slovenia usingthe 1993 HouseholdBudget Survey is implemented by Pentaraki andMergos (1999). Datawas gathered throughthe 1993 surveyo fthe Statistical Office of the Republic o f Slovenia. The analysis i s related to the socio-economic characteristics o f Slovenian households. The major finding o f the study i s that 17.9 percent o f the Slovenia sample i s classified as poor. 5 The social safety net in Hungary and other transition economies has undergone significant changes. The traditional measure for poverty in Hungary was the subsistence minimum which has lost much o f its relevance because o f the transition to a market economy. Using 1992-1995 Household Budget and Panel Surveys, Rudolf and Zsolt (1996) and Grootaert (1997) analyze poverty and social transfers inHungary. Grootaert proposes two benchmarks: the minimum pension set by an absolute poverty line and a relative poverty line set at two-thirds o f mean household spending. Although unemployment benefits and social assistance are well-targeted to the poor, Grootaert still distinguishes six components o f social safety net as pensions, unemployment benefits, family allowances, child care allowance, social assistance and child care fee. Approximately 91 percent o f households receive one or more transfers. Hungary's social safety net represents 54 percent o f spending in an average household, and provides 38 percent o f spending in an average income. The average transfer i s eight times smaller than that would be needed under perfect targeting. T herefore, there is s ignificant room for reallocating funds for improved w elfare o f the poor. Since January 1990, Poland's social safety net has changed. Unemployment benefits have been introduced and the cost o f social safety net has risen sharply both absolutely and as a fraction o f GDP. To analyze the effects o f the new system on income distribution, Grootaert (1995) uses Household Budget Survey Data for January-June 1993. H e examines the relationship between poverty and social transfers in Poland. In 1993, 26.3 percent o f the population had an expenditure level below the minimumwage, and 14.4 percent spent at a level below the minimumpension. This study discusses proposals which are budget-neutral or which imply only modest increases inthe total amount o f transfers, for improving the ability o f social transfers to reduce poverty. Falkingham (1999) examines welfare in transition, trends in poverty and well-being in Central Asia's five Republics o f former Soviet Central Asia- Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan and Uzbekistan, and the Republic o f Azerbaijan. The regional population faces severe economic, physical and psychological stress. Over half of the population lives inpoverty, real wages have fallen, unemployment has increased, school enrolment has dropped and general health has deteriorated. 11)Sub-Saharan Africa During 1990s, among seven African countries, four (Burkina Faso, Nigeria, Zambia, and Zimbabwe) experienced an increase in poverty, while three (Ghana, Mauritania, and Uganda) had decline. Ghana is the country that has been extensively investigated with regard to poverty and its reduction among the other African countries. Canagarajah and Coulombe (1997) examine the relationship among employment, labor markets and poverty in Ghana This study particularly analyses the changes during e conomic decline and recovery periods. The formal s ector is only a small part o f Ghana's labor market. This should be taken into account in forming a n appropriate employment policy. An active labor policy and employment creation is necessary for sustainable poverty reduction. Inthe other study, Canagarajah, Mazumdar and Ye (1998) examine the structure o f poverty reduction in Ghana, for the period o f 1988-92 and the determinants o f inequality. Three rounds o f the Ghana Living Standard Survey between 1988 and 1992 are used. Findings suggest that the adjustment policies, aiming at decreasing public sector employment and stimulating activities in the 6 private sector, succeeded in raising the living standards in the rural areas and cities but not in the capital city. Thus, an integrated regional study, taking into account the local socioeconomic structure, i s necessary. Finally, Blunch and Verner (2000) investigate the link between poverty and child labor in Ghana. The findings prove that the positive relationship between poverty and child labor i s reinstated. There i s evidence o f a gender gap in child labor linked to poverty. Girls are consistently found to be more likely to engage in harmfil child labor than boys. The incidence o f child labor increases with age, especially for girls. Grootaert, Kanbur and Oh (1995) investigate the dynamics o f poverty in Cote d'Ivoire. In urban areas o f Cote d'Ivoire, human capital is the most important endowment explaining welfare changes over time. Inrural areas, physical capital matter the most. Households whose heads worked in the public sector maintained welfare better than other households. Migrant non-Ivorian households tend to be better at preventingwelfare losses compared to Ivorian households. Mackinnon and Reinikka (2000) summarize the lessons from Uganda's Strategies to Fight Poverty. Uganda was one o f the first low-income countries to prepare comprehensive national strategy for poverty reduction usinga participatory approach. Ugandabenefited from the new World Bank-International monetary Fund approach to strengthen the impact on poverty o f occasional assistance in low-income countries. The new approach features a more inclusive participatory process for helping recipient countries to develop poverty reduction strategies which would bring the Bank-Fund assistance to the country. 111) LatinAmerica andthe Caribbean Poverty and poverty reduction policies in Latin America and the Caribbean were extremely poor in 1996. Inequality also remained high (weighted average of Gini at 0.56) in 1996. The elasticity o f poverty to growth was unitary. The high level o f poverty was inpart due to the adverse impact o f macroeconomic shocks. The poor were socially excluded and the state institutions were not sufficiently pro-poor. Available national poverty estimates for Latin America showed that between 1989 and 1996 poverty rate fell in Brazil, Chile, the Dominican Republic, and Honduras and rose inMexico and Venezuela. Inanother group o f countries for which only urban surveys were available, poverty rate rose inEcuador, s tayed st able inUruguay, and fell inArgentina, Bolivia, Colombia and Paraguay (see World bank 2000/2001, p. 25). Wodon et.al (2001) examine trends (1986-1998) and determinants of poverty in Latin America. In 1998, about a third o f the population was poor, and one sixth extremely poor. Such household characteristics as demographics, education, employment, geographic factors, location, migration and ethnicity have large impacts on the probability o f beingpoor or extremely poor. Hentschel, Waters and Webb(l996) study rural poverty in Ecuador by using a rural Qualitative Assessment o f living conditions in rural communities in three o f Ecuador's diverse regions. The findings indicate that rural communities with the same characteristics (such as area, soil quality and ethnic background) are very heterogeneous in their command o f land resources, definition o f well being, range o f economic activities and recommendations for what i s needed to overcome poverty. Intimes o f hardship, families complement income from traditional sources such 7 as subsistence agriculture and small animal husbandry with earnings from new activities. Spending less on food is alarming as malnutrition rate in Rural Ecuador i s already very high. Poor rural families demand very practical solutions to reduce poverty. They suggest measures that will make available land and human resources more productive inorder to overcome poverty. Gragnolati (1999) examine the relationship between children's growth and poverty in rural Guatemala. Gragnolati shows that poor child growth outcomes in Guatemala are the result o f widespread poverty. The better the parent's level o f education and household income are, the less the children suffer from malnutrition. Children are cared better where there are better community infrastructure and health care facilities. The effects o f economic liberalization on poverty are addressed in the literature to a great extent. It is hypothesized that the liberalization process provides greater opportunities to the individuals that were initially better endowed with human and physical capital and therefore contributes to increasing inequality and poverty. Szekely (1998) shows that those individuals better endowed with human and physical capital benefited disproportionately more from the economic expansion during the reform period, thus confirming the working hypothesis, using ten household surveys spanning 42 years (from 1950-1992) and concentrating particularly on the last three surveys from 1984 to 1992 that cover the reform era inMexico. The effects o f poverty alleviation programs were largely ineffective inMexico during the reform era (Szekely (1998). Thus, the liberalization in Mexico contributed to greater poverty and inequality. Cord and Wodon (2001) look for an answer to the following question: D o agricultural programs inMexico alleviate Poverty? The evidence from the Ejido S ector which is based ona panel survey for the sector inMexico for the years 1994 and 1997 indicates that [l] participation in PROCAMPO-"a cash transfer per hectare o f cultivated land" program-significantly reduces the likelihood that a ejido household will be poor and the program may have a positive multiplier effect on household income and that [2] ALIANZA-a matching grant program for investments in agriculture- has no significant impact on poverty. IV) East Asia and The Pacific East Asia poverty trends in the 1990s were influenced by the impact o f economic crisis. Poverty increased inIndonesia, Republic of Korea and Thailand, while inVietnam poverty based on national poverty line decreased from 58 percent in 1993 to 37 percent in 1998. Suryahadi et. a1 (2000) evaluate poverty during the crisis in Indonesia for the period between o f 1996-99 by using Susenas surveys. This study shows that poverty increased from 11.34 percent in 1996 to 18.9 percent in 1999. Datt and Hoogeveen (2000) investigate poverty and income distribution in the Philippines using 1998 Household Survey Data. The recent economic crisis led to a 5 per cent reduction in average living standards and a 9 percent increase in the incidence o f poverty. There is evidence o f consumption smoothing by the households affected by the crisis, but the poor were less able to protect their consumption. 8 V) SouthAsia Poverty reduction also varied in South Asia in the 1990s. Bangladesh showed good performance owing to the good rice crop in 1988-1989 despite the terrible floods o f the same year. Wodon (1999) examines growth, poverty and inequality using a regional panel data for Bangladesh. Wodon shows that growth reduces poverty inboth urban and rural areas. Growth i s associated with rising inequality only inurban areas. Promoting growth inrural areas rather than urban areas would reduce poverty more. Pakistan and Sri Lanka made almost no progress in poverty reduction in 1990s. Some aspects o f poverty in Sri Lanka are investigated by Datt and Gunewardena (1997) for the period o f 1985-90 using 1985-86 and 1990-91 household surveys. Poverty in Sri Lanka i s largely a rural phenomenon. Between 1986 and 1991, national poverty rates declined modestly, due to a decline in rural poverty. But, urban poverty increased. Poorer households tend to have higher dependency ratios, fewer years o f schooling, lower participation in the labor force, and significantly higher unemployment. Datt and Ravallion (1996) try to find out why some Indian States did better than others at reducingrural poverty? Datt and Ravallion use pooled state-level data for 1957-91. Findings suggest that reducing rural poverty requires both economic growth (farm and non-farm) and humanresource development. Jayaraman and Lanjouw (1998) calculate inequality in Indian villages and analyze the evolution o f poverty by using longitudinal village survey data. Results reveal that if India i s to continue reducing rural poverty, it is essential to have continued agricultural growth, diversification into non-agricultural activities and awareness o fpersisting inequality within villages. VI) MiddleEast andNorthAfrica The rural economy o f developing countries used to be regarded as synonymous with agriculture but in recent years this view has been changing. Such diverse activities like commerce and services can be seen as providing some income in rural households. Applying decomposition analysis and using household level data from Egypt and Jordan, Adams( 2001) examine the effects o f non-farm income on inequality and poverty in rural Egypt and Jordan. Non-farm income decreases inequality in Egypt and increases it in Jordan. The poor in both countries depend heavily on government employment t o decrease inequality. InEgypt, c ultivated 1and b ase is v ery highly productive, thus Egypt's large rural population seeks access to land. Unlike the rich in Egypt, those inrural Jordan earn less than 10percent of their total per capita income from agriculture and more than 55 percent o f it from non-farm sources. 9 References Adams R.H. (2001) Nonfarm Income, Inequality, and Poverty in Rural Egypt and Jordan" Policy ResearchWorking Paper 2572, World Bank, Washington, D.C. Aslund, A. (1999) "Review Of: Income, Inequality, and Poverty duringthe Transition from planned to Market Economy" Journal of Economic Literature, vol. 37 3, 1186-1187. Balke, N. S. And Slottje, D. J. (1993), " Poverty and Change in the Macroeconomy: Macroeconometric Model", The Review o f Economics and Statistics, Vol. 75, Issuel, pp. 117-122. Bhorat, H. (1999) "Distinguishing between individual- and household level poverty" Development Southern Africa, vol. 16, 157-162. BlunchN. H. and Verner D., 2000, "Revisiting the Link between Poverty and ChildLabor: The GhanaianExperience", PolicyResearchWorking Paper 2488, World Bank, Washington D.C. Canagarajah S.and Coulombe H. , 1997, " 1845.Employment, Labor Markets and Poverty in Ghana: A Study o f Changes during Economic Decline and Recovery" Policy Research Working Paper, World Bank, Washington D.C. Canagarajah S., Mazumdar D., Ye X., 1998, "The Structure and Determinants of Inequality and Poverty Reduction in Ghana, 1988-92, Policy Research Working Paper, World Bank, Washington D.C. Chen S. and Wang Y., 2001, "2651. China's Growth and Poverty Reduction: Trends between 1990 and 1999", PolicyResearchWorking Paper 2651, World Bank, Washington D.C. Clark, S., Hemming, R. And Ulph, D. (1981), " On Indices for the Measurement of Poverty" The Economic Journal, Vol. 91, issue 362, p. 515-526. Cord L. and Wodon Q., 2001, "Do Agricultural Programs in Mexico Alleviate Poverty? Evidence from the Ejido Sector", Cuadernosde Economia, August 2001, v. 38, iss. 114, pp. 239-56 Datt G. and Ravallion H., 1996, "1594. Why Have Some Indian States Done Better Than Others at ReducingRuralPoverty?", Policy ResearchWorking Paper, World Bank, Washington D.C. Datt G. and Gunewardena H., 1997, " 1738. Some Aspects of Poverty in Sri Lanka: 1985-90", PolicyResearch Working Paper, World Bank, Washington D.C. Datt G. and Hoogeveen H., 2000, "El Nino or El Peso? Crisis, Poverty and Income Distribution in the Philippines", Policy Research Working Paper 2466, World Bank, Washington D.C. Danziger, S. and Gottschalk, P. (1986)" Do Rising Tides Lift all Boats? The Impact of secular and Cyclical Changes on Poverty" The American Economic Review, Vol. 76, Issue 2, Papers and Proceedings of the Ninety-EighthAnnual meeting o f the American Economic Association, pp. 405- 410. 10 Dollar, D and Gatt, R. (1999) " Gender Inequality, Income and Growth: Are Good Times Good for Women?" Policy Research Report on Gender and Development Working Paper Series, no.1, World Bank,Washington D.C. FalkinghamJ., 1999, "Welfare in Transition: Trends in Poverty and Well- Being in Central Asia", Centre for Social Exclusion Discussion Paper: CASE/20London School o f Economics and Political Science, London, England Foster, J. (1984), "On Economic Poverty: a survey of Aggregate Measures", Advances in Econometrics, vol. 3, pp. 215-25 1. Gottschalk, P.And Danziger, S . (1985) " A Framework for Evaluating the Effects of Economic Growth and Transfers on Poverty" The American Economic Review, Vol. 75, Issue 1, pp. 153-161. Gragnolati M., 1999, "Children's Growth and Poverty in Rural Guatemala", Policy Research Working Paper 2193, World Bank, Washington D.C. Grootaert C., 1995, "Poverty and Social Transfers inPoland", PolicyResearch Working Paper 1440, World Bank, Washington D.C. Grootaert C., 1997, "Poverty and Social Transfers in Hungary", Policy Research Working Paper, WorldBank, Washington D.C. Grootaert C., Kanbur R, Oh G., 1995, "1499.The Dynamics o f Poverty: Why Some People Escape from Poverty and Others Don't? An African Case Study", Policy Research Working Paper ,World Bank, Washington D.C. Hentschel J., Waters W. F. and Webb E., 1996, " 1576. Rural Poverty in Ecuador- A Qualitative Assessment", Policy Research Working Paper, World Bank, Washington D.C. Jayaraman R. and Lanjouw P., 1998, "The Evolution o f Poverty and Inequality in Indian Villages", Policy Research Working Paper, World Bank, Washington D.C. Kakwani N. ,1995, "Income Inequality, Welfare and Poverty: An Illustration Using Ukranian Data.", Policy Research Working Paper 1411, World Bank, Washington D.C. Klasen, S. (1999) "Does Gender InequalityReduce GrowthandDevelopment? Evidence from Cross Country Regressions" Policy Research Report on Gender and Development Working Paper Series no. 7 World Bank, Washington D.C. Mackinnon J.and Reinikka R., 2000, "Lessons from Uganda on Strategies to Fight Poverty" Policy ResearchWorking Paper 2440 ,World Bank, WashingtonD.C. Blackbum, M.L.(1994) " International Comparisons o f poverty" The American Economic Review, Vol. 84, issue 2, Papers and Proceedings of the Hundredand Sixth Annual meeting o f the American Economic Association, 371-374. 11 Milanovic, B. (1998) Income, Inequality, and Poverty during the Transition from Plannedto market Economy, Regional and Sectoral Studies. Washington, D.C.: World Bank. Pendakur K.,2001, " Consumption Poverty in Canada,1969-1998", Canadian Public Policy, June 2001, v.27, iss 2, pp.125-49 Pentaraki K.N. and Mergos G.J., 1999, "An Analysis o f Poverty in Slovenia Using the 1993 Household Budget Survey", East-West Series in Economics, Business, and Environment 1999, v.2, iss. 1-2, pp. 96-104 Rudolf A. andZsolt, S. (1996), "Poverty in Hungary in 1992-1995", Review of Sociology o f the Hungarian Sociological Association, specialissue, pp.3-29. Sawhill, I.V. (1988), "Poverty in the U.S.: Why I s I t so Persistent? " Journal of Economic Literature, vol. 26, issue 3, pp. 1073-1119. Sen, A. K.( 1976), "Poverty: an 0rdinalApproach t o Measurement ",E conometrica, v 01.44, pp. 219-231. Suryahadi A., Sumarto S., Suharso Y. and Pritchett L, 2000, " 2435. The Evolution of Poverty during the Crisis in Indonesia, 1996-99 ", Policy Research Working Paper 2193, World Bank, Washington D.C. Szekely,M. (1998) The Economic Poverty, Inequality and Wealth accumulation in Mexico, St Antony's Series. New York: St. Martin's Press; London: Macmillan Press; in association with St. Antony's College. Thomas ,V. and Wang Y. (1998) " Missing Lessons of East Asia: Openness, Education and the Environment" Paper presented at Annual Bank Conference on Development in Latin America and the Caribbean, World Bank, June, Montevideo. World Bank 2000 / 2001, World Development Report 2000 / 2001 Attacking Poverty, New York: OxfordUniversity Press Wodon Q., 1999, "Growth, Poverty and Inequality: A Regional Panel", Policy Research Working Paper, World Bank, Washington D.C. Wodon Q. et.al, 2001, "Poverty in Latin America: Trends (1986-1998) and Determinants", Cuademos de Economia, August 2001, v. 38, iss. 114, pp. 127-53 Endnotes: Thomas and Wang 1998; Klasen 1999; Dollar and Gatti1999 12 Szekely(1998) argues that the financial liberalization ( key element o f the new strategy inMexico) failed to ease the liquidity constraints for most households - benefiting only the richest quintile of the population. ... 111The literature on poverty measurement has grown apace since Sen's seminal paper, (Sen 1976). A useful survey i s provided by Foster (1984) iv They made international comparisons of poverty using Luxemburg Income Study (LIS) data 11 industrialized nations inthe 1984- 1987 period. " The level of aggregation is an important component in the analysis of poverty measurement. Utilizing the Headcount Index and cumulative distribution functions, Bhorat (1999) examined earnings among individuals and households o f South Africa. It is shown that property measures at the individual level compared to the estimates at the household level will transmit differential rankings of indigence. Both are important predictors o f poverty and hence should be coupled when tryingto understandlow earnings indifferent societies. vi The characteristics associated with poverty through the construction of poverty profiles are found to be educational and occupational disparitiesas well as geographic and demographic characteristics ( Szekely 1998) 13 APPENDIX I:POVERTY AND INCOME INEQUALITY MEASURES Internationally ComparableAbsolute Poverty Line (the "One -Dollar-A-Day Per Capita Line)': For absolute poverty line a minimum food basket composed o f number o f commodities which can be chosen regarding the consumption habits o f an adult can be used. Then, generally, monthly cost o f food basket i s calculated int e m s o fboth domestic currency and also interms o f its dollar equivalent. Per capita poverty line can be adjusted for purchasing power parity (in certain base year prices) amounted to a value in terms o f domestic currency per period o f interest. The cost o f this minimum food basket allowances for non-food items, as suggestedby actual consumption patterns o f the less affluent can be added to arrive at the economic vulnerability line. Household whose monthly consumption i s below the absolute line or the economic vulnerability line are considered to be absolutely poor or economically vulnerable respectively. Inthis measure, poverty lines o f other countries are also set in order to maintain a constant purchasingpower o f the poverty lines across countries. For converting currencies, a set o f international price indexes are required. This i s not an indicator o f aggregate poverty but it can help to revealboth the risk and the patterns o fpoverty. Relative Poverty Line2:The relative line i s set at one-half o f the monthly median expenditure per equivalent adult. Household with monthly income less than the corresponding relative line are called relatively poor. This measure suggests choosing a poverty line which varies according to the average income o f each country. Relative poverty lines are established as a fraction o f medianincome o f the society as a whole. Relative poverty comparisons are mainly comparisons o f the dispersion o f income at the low end o f the distribution. This allows overall poverty and poverty among groups to be studied. Income Gap3;The other measure o f poverty i s the average income gap. This i s defined as the difference between the poverty lines and the disposable income o f the poor, as a percentage o f the poverty line. Gini Index4:This is the index o f the distribution o f income among the poor. Gini Index is a function o f the rank order weighted sum o f individual income shares. Gini index is drived from the Lorenz curve, which plots cumulative shares o f the population, from the poorest upwards, against the cumulative share o fincomes that they earn. Sen Index'; The other measure o fpoverty i s Sen Index. This index is defined as the Head- Count Ratio (i s the percentage o f p ersons with a n equivalent household disposable income below a poverty line) multiplied by the sum o f the income gap ratio and Gini Coefficient o f the poor weighted by 1 minus the income gap ratio. The Sen index o f relative poverty reflects the implication o f some poor individuals having incomes which deviate from the poverty lines by more than i s the case with other poor individuals. 1-2 -3 Blackbum, M.L, International Comparisons o f Poverty, The American Economic Review, V.84, Issue 2, Papers and Proceedings of the Hundredand Sixth Annual Meeting o f the American Economic Association (May, 1994), 371-374 ECO/WKP(98)2 4-5 Clark S., HemmingR., Ulph D., On Indices o f the Measurement o f Poverty, The Economic Journal, V. 91, Issue 362 (June 198l), 515-526 ECO/WKP(98)2 14 n n n 3 w m v, 3 UY UY 0 cd .s .9 8 n n N M Lr. a;, . % $ n n m * nIA F Gj 3 n nn d c\Im 1 3 . f 'U b a E n + nN n n n 4 " P 0 a .$t n n n n h + + c \ l m * * 0 n n n n n n 3 c\1 m " N m 3 n n n n 3 N " N N 3 N ha, 9 9 d 00 2. m 2 h zl 8 2 cd .i Y W 3 n2 v) e, 3 ? s b v, uj e, .i 21 cd s Y v) 3 e, ti 3 d, c.' cd Y % v) 3 0 pc 0 a 5 -Elf w s0 P .I4 d E51 &.a Ycd 43 3a 4 4cd .,-ba 8 2 2D m n 4 hh 3 + N m 0 c, agcd w m I I I cz. n N 2. Poverty in Turkey (Kivilcim Metin Ozcan) The development level o f a country is mainly determined by the income inequality between the rich and the poor living inthat country. Inequality matters because it determines how the benefits o f growth are to be distributed and who receives the benefits o f that growth. If poverty was a concern, economic growth would be the main tool to alleviate it. Growth mostly affects the top o f the distribution and may improve poverty very little. An increase in inequality may reduce the positive effect o f growth. However, when there are significant improvements in the distribution o f income, even very slow growth may have a big impact on growth. Although there i s no single inequality measure, one potential measure can be defined as the relationship between the incomes o f the rich (persons at the 90thpercentile o f the distribution) and those o f the poor (those at thelothpercentile) or decile ratio. It i s common to break decile ratio into its bottom and top portions: the distance between the poor and the middle class (or median); and the distance between the middle class (median) and the rich. The second measure i s a single coefficient that gives an integral view o f the overall distribution o f income or consumption between households in a survey. The most well known i s the Gini index, which i s mostly sensitive to the middle of the income distribution. The higher the value o f Gini coefficient is, the more unequal the distribution is. Finally, inequality can be interpreted...as the inequality in shares o f total incomes. The population is dividedinto equal groups"' with different levels o f income or consumption and the share o f total income or consumption accruing each group i s calculated as an inequality measure. Turkey is a country with large inequalities. Income differentials across regions, provinces and social groups are wide and persistent. The distribution o f income inequality between 1987 and 1994 remained the same using Household Income and Consumption Expenditure Survey (HICES)" o f 1987 and 1994. When we look at provinces rather than households, we see that Turkish provinces are diverging in an absolute sense inthat rich areas are gettingricher, while the poor ones are gettingpoorer. Inthe following subsection, the trends in inequality and the factors affecting inequality will bediscussed indetail. I)Dimensions,TrendsandtheSourcesofInequality Inequality in Turkey i s above the median for upper and upper middle income economies. The social distance between the poor and the rich inTurkey i s higher than in other countries." Inequality is coming from the bottom and the top o f the distribution. The income o f rich people i s almost 7 times higher than that o f the poor. Indeed, inequality in Turkey is high and close to the levels observed in economies such as Peru or Russia. Inequality in Turkey is not as high as Latin American countries but it is higher than the levels o f inequality observed in Central European and Mediterranean countries (see World Bank 2000, p. 19). 28 A decomposition o f inequality by sources o f income shows that labor income (wages and self employment income) i s the main driving force o f inequality. Labor account for 77 percent o f the total income inequality between households in Turkey. Income from self-employment i s especially unevenly distributed (World Bank 2000, p.19 Table 2). Capital and property income i s also unequally distributedwhile it represents a small share o f total income. Wage differentials in Turkey largely reflect the differences in education ,gender, experience and occupational status. Much o f the observed inequality in wages also reflects the differences in endowments, productivity or working conditions. One o f the most striking facts o f the last two decades in high and middle income countries alike, i s the rise in the wage premium for education. This can be seen as a sign o f rising demand for skilled labor. Given the relatively modest increase inthe wage premium by education and slow changes in the structure o f employment, calculated Gini coefficient for wages increased very little between 1987 and 1994: from 0.395 to 0.418. This can be defined as highbut stable inequality. Ifone excludes university level education ,wage differentials between education groups appear to have increased. Therefore rising educational differentials could be one factor contributing to higher wage inequality in Turkey (see World Bank 2000, p.29). There is very little redistribution o f income through either transfers or safety nets. None o f the sources o f household income in Turkey contributes to reducing inequality. Although State transfers which regressively distributed the sources o f income, play a n important role in preventing some groups such as the elderly, from falling into poverty, the rich still receive a larger share than the poor. Rural and urban differences are also important determinant o f inequality in Turkey. The differences between urban and rural areas account for more than 10 percent o f the total inequality in Turkey. These differences are due to the existence o f large and growing productivity differentials between agriculture and the rest o f the economy. Therefore, the income o f the rural households is much lower than that o f the urban households. Filiztekin(l999) summarizes a number o f factors that can be attributed to inequality between regions as: inter-regional differences inaverage GDP p er c apita t o differences in sectoral structure and differences inproductivity across sectors.Vi Filiztekin also finds large cross regional differences inproductivity within the same sector: not only poor regions have 1arger share o f their resources engaged inagriculture, but they also exhibitmuchlower productivity within agriculture thanricher regions. The comparison o f the 1987 and 1994 HICES reveals that income inequality for household monetary income increased significantly duringthis period. This can be due to the impact o f macroeconomic imbalances and structural changes in the economy. Whether the existing income differences increase or fall inthe future depends mainly on the ability o f Turkish Government eliminating macroeconomic imbalances and 29 improving its redistributive policies and the most importantly targeting o f its social transfers. Inflation has a significant distributional impact on inequality by affecting both within wage inequality and relative factor incomes.v11Since consumption data from 1987 and 1994 surveys are not directly comparable, it i s not meaningful to compare consumption inequality. The Gini coefficient for household money income increased from 0.411 to 0.453, which corresponds approximately to 10 percent increase in 7 yearsviii However, inequality in total income increased much less. The Gini coefficient for total income did not change between 1987 and 1994 and quintiles shares remained stable despite highlyvolatile economic conditions. Both He(1999) and World Bank(2000 p.26) suggest that the impact o f inflation on the poor is transitory. Unanticipatedjumps ininflation or periods o f accelerating inflation hurtthe poor andworsen the distribution. But,nominalincome adjusts to new stable level o f inflation quickly due to the wage indexation mechanisms which might help to explain why, Turkey's income distribution has not deteriorated despite years ofhigh inflation. However, theoretically, high inflation, even if it i s stable, i s bound to have a negative impact on poverty and inequality inthe longrunthrough its dampening effect on growth. State transfers by households are also other important determinants o f inequality and they are classified basically under four headings, namely, state pensions, tax return, old age income and scholarships and in-kindtransfers from the state in the 1994 HICES (see Table 1below). Some important transfer programs such as agriculture subsidies, are aimed at producers rather than households and therefore are not represented in the 1994 HICES. The most important state transfer i s pensions, which are rather unequally distributed with the richest 20 percent receiving almost 50 percent o f all pensions. The pension system in Turkey is designed as an insurance scheme; it has almost no redistributive aim. To an extent, pensions are determined by past contributions and therefore past wages, a larger fraction, i s bound to accrue to the top o f the distribution .Pensions are important since they prevent certain groups o f population, particularly the elderly, from falling into poverty. However, not many pensioners are that old as a result o f the absence o f a minimum retirement age and a short period o f minimum contribution.ix Similar to pension, old-age income transfers and scholarships appear to be unequally distributed. Incontrast, in-kindtransfers from the State favor the poor, but they are small interms of their impact on household income (see World Bank 2000, p.32). Table 1 Incidence analysis: Distributionof State Transfers by HouseholdIncome - Note: Table is obtainedfrom World Bank 2000, p. 31 and annual income datafor 1994,from SISAnnual Income Survey. 30 Income inequality andpoverty are two highlyrelated concepts and they are isomorphic to each other. Ifgroup of countries and societies suffer from severe income inequality, one can expect highpoverty problemas well. Therefore, poverty andpoverty reduction gain importance and should be separately analyzed. Although, many studies were conducted on inequalityo f income distribution or determinants o f income regarding Turkisheconomy using 1983, 1987 and 1994income and consumption surveys data, studies on poverty and poverty reduction inTurkey are quite rare and recent. Inthe following section, studies ofpoverty and poverty reductioninTurkey will be summarized. 11) Dimensionsand CharacteristicsofpovertyinTurkey Earlier studies o f poverty in Turkey by Dansuk (1997), Dumanli (1996), Erdogan (1997), Erdogan (1998), Uygur and Kasnakoglu (1998) use HICES data to show relative and absolute poverty deprivation in Turkey. However, with the exception o f Erdogan (1997, 1998), none o f these studies use data for the whole country from both recent surveys. With a different approach, Akder (1999) uses aggregated geographical data at the level o f administrative districts to arrive at a poverty map o f Turkey based on the Human Development Index. Studies on poverty have been carried out by using the household income and c onsumption surveys, and major studies on finding the poverty line and poverty rate inTurkey are summarizedbelow indetail. Income distribution and structural change The first study t o investigate t he problem o f income distribution and structural change inthe Turkish Economy is Dagdemir (1992). Inhis study, Dagdemir investigates the problem o f poverty and the socioeconomic structure o f poverty. H e uses absolute poverty criteria and the criteria o f earning less than the average level o f income in this study. H e stresses the fact that, in order to attain a fair level of income distribution, it i s necessary to increase the living standards o f the poor. Table 2 illustrates that during the 1968-1987 period, poverty declined and the income share of the poorest 40 percent and 60 percent in the total level o f income increased by 5.3 percent and 8.3 percent respectively (see Table 2). Table 2 also clearly shows that the percentage o f households who earn less than the average level o f income decreased from 49.1 percent to 34.2 percent during the two decades. It i s reported that poverty i s mostly common among the households who earn wage income ( see Table 3). Dagdemir identifies strong trade-off between growth and poverty reduction and points out that, the planned period showed poverty reduction, which corresponds to the period when a sustainable economic growth was achieved. Therefore, he concludes that in order to alleviate poverty, income distribution should be improved, and production capacity and growth rate o f income shouldbeincreased. 31 Table 2: Householdsat the povertyline. 1968 1973 1987 Poverty Line (TL) 6.940 12.000 1.357.000 Income share of the poorest 40% (%) 9.5 11.5 14.8 Income share o f the poorest 60% %) 20.6 24.0 28.9 Households who earn less than the average level of Income (%) 49.1 41.6 34.2 Absolute poverty (%) 51.5 38.4 22.5 Table 3: The Socio-economicstructureof povertyaccordingto thejob status of the householdhead(1987) Acc. to the criteria of earning less Acc. to the absolute poverty criteria Job statue of the than average level of income household head Within the Within all the Withinthe Withinall the Group (%) Poor (%) Group (%) Poor (%) Wage earner 48.5 52.46 36.23 51.71 Self-employed 35.12 46.32 28.45 47.15 Employer 6.67 1.22 3.54 1.15 Source: Dagdemir (1992) The impactof regionalfactors on inequality Inhis study, Dumanli(1996) reports the dimensions ofpoverty over years across several regions using 1987 HICES data. The poverty lines are computed by calculating the minimum amounts of calorie needed, which i s a very frequently used method o f finding the poverty lines'. It was found out that as of year 1987, Eastem and South Eastem regions had to fight poverty -both at household level and individual level- much more severely than the other regions (see Table 4). It is also observed from the poverty lines in terms o f U S dollars that poverty in Turkey reveals an upward trend during the 1987-1994 period; the annual figure being 332.2 $ inyear 1987, and 541.4 $ inyear 1994 (see Table 5). 32 II Table 4: Poverty in Rural and Urban Regions in Turkey (YO),1987. Urban Regions I1Rural+Urban Household I individual iI Rural household I Individual 1 household I individual Aegean- 3.74 3.01 5.43 4.36 0.00 0.00 Marmara Medit. 19.80 18.55 22.27 21.03 14.36 13.00 Central 32.01 26.30 29.14 23.33 38.33 32.9 Anatolia Black Sea 10.49 8.34 11.28 8.76 8.75 7.41 East & 33.97 43.80 31.88 42.53 38.57 46.64 , Southeast , Anatolia Source: Dumanli (1996) Annual Years T.L U.S. $ 1987 284.700 332.2 1988 473.405 332.2 1989 777.085 365.0 1990 1.388.433 532.4 1991 2.135.740 513.4 1992 3A30.809 556.2 1993 6.433.490 585.6 1994* 14.696.360 541.4 Inthis study, by using 1987 HICES data for consumption expenditures, poverty rates are computed by accepting the minimum level o f consumption expenditure as the poverty line. Also, the differences in price levels and calories across regions are taken into account and "regional absolute poverty lines and rates" are calculated (see Table 6). Results show that regarding the regional characteristics, absolute poverty rates change from one region to another and from rural to urban substantially. 33 Poverty lines (T.L/ month) Poverty rate (%) Regions Urban Rural Urban Rural Total Aegean 27.202 23.348 16.26 5.43 9.56 Mediterranean. 24.160 21.457 11.79 26.2 1 20.72 Central Anatolia 24.595 21.744 29.84 22.16 25.09 Black Sea 22.723 21.349 6.71 10.91 9.3 1 East & Southeast 22.340 20.246 35.4 35.28 35.32 Anatolia L I I I I I I Source: Dansuk (1997) Erdogan (1996) also computes several poverty lines and the number o f households below these poverty lines for: a) Turkey as a whole b) Ruralregions o fTurkey c) Urbanizedregions o f Turkey d) Seven geographical regions o fTurkey separately using 1994HICES data. The cost o f food baskets - that give the minimum amount o f calorie needed to survive- is computed separately for each regional classification using several different methods. These costs are compared to the household expenditure on food and total household expenditure. I n thefirst method, namely minimum food expenditure approach, households who cannot even cover the cost o f the food baskets are determined and they are classified as "extremely poor".x' I n the second method, namely basic needs approach, the minimum level o f income that the households need are determined by taking into account the necessary amount o f expenditure on housing, clothing, transportation and hrniture besides the expenditure on food. The households with income below these levels are classified as "low level poor". I n the third method, poverty lines are obtained by"food rate" approach andthe poor households are classified as "high level The results obtained through 1994 household income and consumption survey data show that 11percent o f the households in Turkey are extremely poor, the following 20 percent are "low level poor" and 12 percent are "high level poor". People who live in rural areas face the problem o f poverty more severely than those who live in urbanized areas. Finally, looking at the results o f the analysis on geographical base, confirming the results o f the previously held studies, she concludes that East and Southeast Regions have to fight poverty much more than the remaining regions (see Table 3). 34 cost o f 1 minimum food 11 7 14 mending: I I Basic needs 1 1I I 31 28 32 Foodrate Source: Erdogan (1996) In her (1998) study, Erdogan employs two methods for calculating the poverty line and establishing the poverty profile o f the whole country: "cost o f minimum food spending" method and ''cost o f basic needs" method. According to the first method, 5.66 percent o f the households and 8.37 percent o f the individuals are poor. According to the second method, 19.31 percent o f the households and 24.30 percent o f the individuals are poor. When regional poverty rates are considered, it i s observed that people livinginrural areas have a living standard much lower than those living in urbanized areas. As the education level increases, individuals face much less poverty. Also, among those who are not in the labor force; unemployed, old and ill people have a higher poverty rate compared to others in the same group. Finally, those who have agricultural income have lower living standardsthan people working inother sectors. Senses (1999) investigates the role o f The Social Assistance and Solidarity Encouragement Fund (SSF) in poverty reduction. SSF was founded in 1986 as an institution and a measure to alleviate poverty. Senses analyses the reasons and the timing o f t he emergence o f the Fund, its monetary s ources, i t s areas o f effectiveness and the problems faced while implementing the aims o f the Fund. It i s pointed out that the Fund has a formal appearance, an unstable source structure, and that it does not rely on information related to the level and profile o f poverty and has often been subject to political interferences. Therefore, Senses concludes that, although the Fund has some positive aspects, its contributions to poverty fight will only be limited and therefore it i s necessary to have rapid technological and industrial growth. InDagdemir (1999) study, the effects of the macroeconomic contraction on the problem o f poverty is analyzed for the period o f 1987-1994. A numerical change in poverty over years, poverty gap index and an index o f income inequality across the poor are computed in order to measure the change in the intensity o f poverty. The problem o f poverty ins everal regions o f T urkey is examined by using different measuresby also taking the structural change o f poverty into account. In this study, Minimum Food Cost (MFC) and Basic Needs Cost (BNC) methods are usedto calculate the poverty lines. 35 It is observed that, in Turkey, poverty in urbanized regions increased over the 1987-1994 period. In rural regions, besides increasing poverty, it i s also observed that poverty problem becomes deeper and the income inequality across the poor people increases. The results are presented inTable 7 below. Source: Dagdemir (1999) RuralPoverty Akder (1999) shows that productivity differences across rural areas are in tum a reflection o f the differences in factors o fproduction such as labor, capital and land usage. &der classifies provinces as a) high mechanizatiodlow labor intensity provinces, engaged in basic intensive agriculture; b) low mechanizatiodhigh labor intensity provinces, characterized by a scarcity o f arable land and the lowest agricultural productivity; c) high mechanizatiodlow labor intensity provinces, which tend to specialize in oil seeds and some tuber crops; d) highmechanizationhigh labor intensity provinces, which are the most productive and specialize in the vegetables, fruits, tubers and industrial crops. Akder states that poor provinces are disproportionately concentrated inthe low mechanizatiodlow labor intensitygroup, (with some poor districts found also inthe low mechanizatiodlow labor intensity category) (see World Bank 2000, p.22 Table 4) - Inher study, Pamuk (2000) uses the rural areas data of 1994 Household Income Distribution Survey of the State Statistical Institute. Relative poverty definition o f OECD i s adopted to construct the poverty lines. The relative poverty line i s set at one-half (50 percent) o f the annual median income per equivalent household member (according to the OECD equivalence scale). People with disposable income lower or equal to the poverty threshold are called relatively poor. Using the OECD methodology, it i s computed that 14.8 percent o f the rural population and 14.15 percent o f the rural households are called relatively poor; that is, their total annual incomes are below the relative poverty line. It is also detected, confirming the findings o f the previous studies, that poverty differs a lot across regions. Poverty risk i s at the highest level in Southeast Anatolia with 30.46 percent. The population o f this region constitutes the 9.94 percent o f the rural population and the contribution o f this region t o poverty is 2 0.44 percent. The 1owest poverty risk i s in Marmara with 6.50 percent and in Aegean with 7.67 percent. The risk 36 o f the individuals living inthese regions constitutes approximately the halfo f the average poverty risk (see Table 8). Measuresof Poverty (?A) Structure of Poverty (%) ~ Regions Head-Count Poverty GapIndex Total Population PoorPopulation Index Aegean 7.67 23.36 15.10 7.83 Mediterranean 17.42 27.77 12.77 15.03 Central Anatolia 15.79 26.73 16.28 17.36 Black Sea 16.44 29.44 20.01 22.22 East Anatolia 12.97 26.17 13.15 11.51 Southeast 30.46 26.93 9.94 20.44 Anatolia Marmara 6.5 21.76 12.75 5.60 Total 14.80 26.92 100.00 100.00 It is also detected that educationhas the strongest correlation to poverty risk; that is, the education level o f the household head plays a key role in determining whether the household i s poor or not (see Table 9). Table 9: Povertv related to the education level of the household head Education level of the Measuresof Poverty ("A) Structure of Poverty ("!) householdhead Head-Count Poverty Gap Total Population Poor Population Index Index Illiterate 26.25 30.23 18.65 33.08 Literate without a 18.14 28.16 11.86 14.53 diploma Primary school 12.92 24.60 57.29 50.01 Junior school 4.40 19.85 5.44 1.62 High school 2.3 1 27.37 4.91 0.77 University 0.00 0.00 1.85 0.00 Total 14.80 26.92 100.00 100.00 Another finding o f t he st udy is that poverty risk differs among the households depending on the sector in which they are employed. The findings show that those who work inagriculture and construction are subject to more risk o fpoverty (see Table 10). 37 Branches Transtlortation I 7.79 I 20.41 I 3.90 I 2.05 I Financial institutions 0.90 9.38 0.62 0.04 Society services 4.3 1 25.48 8.82 2.57 Out of work 20.77 30.24 7.33 10.28 TOTAL 14.80 26.92 100.00 100.00 Source: Pamuk(2000) Economic Reforms, Living Standards and Social Welfare Study o f World Bank (2000) Report aims at providing the Turkish policymakers, the Bank audience and other interested parties with the information necessary to find the solutions for social dilemmas and policy questions such as: Can Turkey continue to progress in the fight against poverty? I s it possible to accelerate the improvement? What needs to be done in order for the GDP growth to show itself in terms o f increasing employment and wages? I s it possible to obtain a future growth path if 40 percent o f the employment i s in the agricultural sector? Can the huge disparities between regions, in terms o f poverty, be eliminated? Because o f the fact that the main concern o f the Report i s on the impact o f economic reforms on living standards, poverty and welfare, it also touches major labor market issues, income distribution, inequality aspects and the impact o f government expenditures. Interms of labor market issues, it i s concluded that Turkey faces a serious challenge in generating employment. It i s pointed out that, although Turkey has been successful in sustaining positive GDP growth rates recently, it has been less successful at generating employment, and that meetingthe employment challenge requires faster GDP growth. Another detection o f the Report i s that income inequality between regions in Turkey is high, the Gini coefficient for income being 0.45 and for consumption expenditures 0.41. A significant share o f the total inequality in Turkey i s claimed to be explainedby the differences inthe endowments, geography and the opportunities faced in the labor market. On the other hand, it is observed that when looked at the households rather than the regions, income inequality has remained roughly unchanged between 1987 and 1994. The Report also provides a wide elaboration on the poverty issue in Turkey. It i s stated that absolute poverty inTurkey, based on a country-specific minimumfood basket, 38 i s low. However, economic vulnerability is found to be widespread (see Table 11). Inthe absolute sense, a household i s considered to be poor if it consumes less than the cost o f absolute minimum required to provide nutrition and shelter for its members. But, some households may not be absolutely poor, yet may be vulnerable to poverty and an economic shock like loss of employment or unexpected disability may push them into absolute deprivation. Methodology Poverty Line Poverty Incidence Absolute poverty Local cost o f minimum food 7.3 basket Economic Vulnerability Local cost o f basic needs basket 36.3 (including non-food) Relative Income Poverty One-half o f national median 15.7 income The Report also claims that poverty in Turkey i s linked mainly to the education level o f the household head. It i s computed that, one half o f all households headed by an illiterate person are economically vulnerable and nearly 15 percent are poor in an absolute sense (see Table 12). Education Poverty Indicators Structure and decomposition ("A) o f Incidence o f Incidence o f Average Population Vulnerable Poor household economic poverty shortfall o f population population head vulnerabilit the poor Y Illiterate 0.526 0.149 0.329 13.6 19.7 27.8 Literate 0.453 0.105 0.305 7.4 9.3 10.7 without a -diploma Primary 0.382 0.072 0.301 55.8 58.8 55.0 Secondary 0.23 1 0.025 0.261 17.9 11.4 6.2 Higher 0.056 0.003 0.159 5.3 0.8 0.2 Total 0.363 0.073 0.301 100 100 100 Note: average shortfaliis the gap between the average consumption ojthe poor and thepoverty line (0.3 means that an averagepoor person has consumption 30 percent below thepoverty line) Labor market status o f the household head i s claimed to be another major determinant o fpoverty. Ifthe head i s employed inseasonal or casualjobs, the households are subject t o a higher risk o f poverty. Moreover, t here are big differences inpoverty incidence between the regions o f the country; East and South-East Anatolian have a vulnerability risk which i s 50% above the national average and the differences in the absolute poverty rates across regions are even wider. Contrary to the findings o f previous studies, the Report finds s mall differences inv ulnerability and poverty betweenurban 39 and rural areas because o f the fact that different poverty lines are calculated for urbanand rural areas, prices beingmuch lower inrural areas. Overall, it i s concluded that inorder to improve the living standards and to reduce the poverty problem, government expenditures should be reallocated. In doing this, economically vulnerable people should be targeted, increasing growth and price stability should be reached, biases against employment creation outside o f agriculture should be removed and investments ineducation should be increased. 40 References Akder, H.(1999) Dimensions of RuralPoverty inTurkey, World Bank. Dagdemir, 0.(1992), Turkiye Ekonomisinde Yapisal Degigimve Gelir Dagilimi, UnpublishedPh. DThesis, Anadolu Universitesi, Eskisehir. Dagdemir, 0.(1999) Turkiye Ekonomisinde Yoksulluk Sorunuve Yoksulluk Analizi 1987-1994, H.U.IIBFDergisi, Cilt 7, Sayi 1. Dansuk,E. 1997, Turkiye'de Yoksullugun Olqulmesi ve Sosyo-ekonomik Yapilarla Iligkisi, Uzmanlik Tezi, DPT, Ankara. Devlet Planlama Tegkilati. (2000) UzunVadeli Strateji ve Sekizinci Beg Yillik Kalkinma Plani 2001-2005, Sekizinci Bolum, Ankara. Dumanli, R. (1996) Yoksulluk ve Turkiye'deki Boyutlari, DPT UzmanlikTezi, Ankara. Erdogan,G.(1996) Turkiye'de Bolge Ayriminda Yoksulluk SiniriUzerine Bir Caligma, DIEUzmanlik Tezi, Ankara. Erdogan, G. (1997), Poverty Definitions andPoverty in:Turkey, Ankara, SIS. Erdogan, G.(1998), Poverty inTurkey: Level and Profile (inTurkish), Ankara, SIS. Erdogan,G.(2000) "Tiirkiye'de ve Dunyada Yoksulluk Olqumleri Uzerine Degerlendirmeler" inDIE, IggucuPiyasalari Analizleri 1999(l), Devlet Ankara: Istatistik EnstitusuMatbaasi. Filiztekin, A. (1999)," Convergence Across Turkish Provinces and Sectoral Dynamics", World Bank He, D.(1999), "Turkey- Inflation and the Distributionof Income, World Bank Pamuk, M.(2000) "Kirsal Yerlerde Yoksulluk" inDIE, IggucuPiyasalari Analizleri 1999 (l), DevletIstatistikEnstitusuMatbaasi. Ankara: Senses, F. (1999), "Yoksullukla Mucadele ve Sosyal Yardimlasma ve Dayanigmayi Tegvik Fonu", ODTU GeligmeDergisi, 26(3-4), 427-51. Uygur, S. and Kasnakoglu, Z. (1998), Estimation of Poverty Line: Turkey 1994, Ankara, SIS* World Bank. (2000) Turkey: Economic Reforms, Living Standards and Social Welfare Study, Poverty Reduction andEconomic Management Unit,Europeand Central Asia Region. 41 3. PovertyinTurkey; The Social Dimension (TahireErman) Research on poverty has been accelerating since the 1980s, and more so in the 2000s when economic restructuring at the global scale and 'post-fordist production system' (Jessop, 1994) led to changes in the socio-economic conditions of people, challenging the older social order. In Western literature, the term 'underclass,' which refers to those who are marginal and excluded from the system since they are not economically needed, has been with us for many years (Myrdal, 1963). But recently, research pointing to the increasing economic gap between the rich and the poor in Western metropolises (Towsend, 1993), has triggered research on the 'underclass' (Mingione, 1996), which i s now called the 'new urban underclass' (Musterd, 1994). Today's underclass, under the shrinking welfare state, has become an important problem to address. In the Western context, the increasing poverty in cities, and particularly the increasing gap between the rich and the poor in 'global cities,' such as New York, London and Tokyo (Sassen, 1991), hasbeenobserved. As manufacturing moves out o ftowns inthe 'First World' to 'Third World' societies in the search of the capitalist class for cheap and unorganized labor, as new type of service industries, namely, producer services (such as, international legal and accounting services, management consulting, financial services) come to dominate cities, and as the economic basis of highly industrialized countries changes from goods-producing to services- and information- processing ('deindustrialization') (O'Louglin and Friedrichs, 1996), available jobs do not match the qualifications o fpotential labor, and the middle classes erode (Mingione, 1991). The global city has come to contain two major groups, namely, the well-educated andwell-paid people who occupy importantpositions intransnational multinational companies, and the unskilled andpoorly-paidpeople who cater for the needsandwishes o fthe former group. Inbrief, theeconomic restructuring processatthegloballevelthat hasbeenaffecting manysocieties andpeople, has created arecent attention to increasing poverty, mostly inthe urbancontext. A new concept, namely, 'the new urbanpoor' has come to beusedto refer to the 'disadvantaged' (the UnitedNations), the 'non-labor' (Parkin, 2000), andthe 'underclass' (Myrdal, 1963) who lack any opportunity for social mobility and hence who are imprisoned ina 'caste system.' Thus, thenewpovertyis the advanced stageofpovertythat is severe and persistent. The urban poor are mainly women, Blacks and newcomers (Parkin, 2000). Thus, race, ethnicity and gender are the three distinguishingfeatures o fthe newurbanpoor. Turkey, where economic restructuringhas beenthe aim of successive governments since the 1980s which have favored liberal policies to this end, is no exception in experiencing increased poverty and income inequality. While income inequality decreased between 1973 and 1987, it increased in the period o f 1987-1994 (Dansuk, 1997). Poverty intensified between 1987 and 1994. Average annual poverty rate increased by 4.00 % in this period(Dumanli, 1996). Particularly following the November2000 andFebruary2001 crises, poverty in Turkey has become acute. The data support this claim. The amount of money needed by a household o f four persons to survive was 567,095,000 liras in 42 September, and it reached 870,862,000 liras in September 2001 (Kamu-Sen, 2001). Those who face the risk o f povertyinterms o f food and basic needs have reached 24 % (25 % in rural areas and 21 % in urban areas) (SPO, 2001b). And those who are poor despite their employment have reached about 50 % (SPO, 2001b). Inthe last four years, 25 % o f the population applied for and...received financial assistance from the State's hnd for the poor, namely, the Fak-Fuk-Fon""' (Sallan-Giil, 2001). Not only poverty intensified, but also the inequality in the distribution o f income increased. The gini coefficient was 0.58 in 1994, up from 0.43 in 1984 and from 0.38 in 1978 inthe case o f Istanbul(SIS, 1994). In 1994, the wealthiest 17,800 households forming 1% of Istanbul'spopulation received 29 % ofthe total income, and their income was 1437 times as much as o fthe poorest households (Sonmez, 1998). The wealthiest andthe poorest are spatially differentiated in Istanbul. The results o f an analysis done by using the 1990 National Population Census (Guvenq and Igik, 1996) suggest that well-to-do sections o f each category are found to cluster inthree specific enclaves at the sea side, one ineach side o f the metropolitan area, while the poorest are scattered in and around transition zones or aroundvertically disintegrated production complexes (p.59). Furthermore, the results indicate that the spatial variation inhome ownership rates is directly related to the geography o f metropolitan employment, that is, on Istanbul's east coast (Anatolian side) where fordist production (regular payment, steady employment, insurance) predominates, home ownership rates are high, whereas on its west coast (European s ide) where flexible production (responsive t o fluctuations inthe market, and hence entails temporary employment without social security benefits) i s widespread, there i s highconcentration oftenants. The same economic polarization is observed inother metropolitan cities in Turkey. For example, in Adana the wealthiest 20 % of the population received 64.4 % of the total income, whereas the poorest 20 % o fthe populationreceived only 4 YO.This meant that the wealthiest group received 1391 times as much as the poorest group (Sonmez, 1998). Erzurum, Antalya and Ankara follow Istanbul and Adana intheir extreme inequality inthe distribution o f income. Izmir, with its wealthiest group receiving 99 times o f the poorest group, emerges as a city with less inequality between the wealthy and the poor (See tables on pages 26, 27, and 30-48 in Sonmez, 1998 for the unequal income distribution in metropolitancities). Other data (Tiiziin, 2001) further demonstrate that the poorest 5 YOof the urban population receives 246 dollars annually, whereas the wealthiest 5 YOo f the urban populationreceives 10,376 dollars annually. There are various reasons behind this economic polarization in the Turkish urban context. Under the umbrella term 'economic restructuring' and 'neo-liberal policies,' the state's role in economy and politics i s changing. Before the 1980s, the State acted as the referee inthe socio-economic system, controlling redistributionmechanisms, and keeping a strong position bothdue to the strong state tradition the Turkish Republic inherited from the Ottoman Empire and to the characteristics specific to import-substitute industrialization (the 43 state protecting the national private sector through tariffs and quotas) (Pinarcioglu and Ipk, 2001). However, the State has started to take on less active roles since the 1980s. The economic crisis inthe mid-l970s, which affected 'the West and the Rest' alike, and Turkey i s no exception, challenged the society's major tenets on which the social consensus was built. For example, there was the consensus that the emergingbourgeoisie in the Turkish Republic would take advantage o f the capital accumulation in the import-substitute industrializationunder the protection o f the state, keeping away from the potential rents in the urban1andmarket. Inthis w ay, i t w as possible t o distributeurbanrents to a larger population of middle and lower income groups through small-scale developers, addressing middleclasses andthrough squatter development, addressing lower classes, mainly rural-to- urbanmigrants(Pinarcioglu andIyk, 2001). However, this consensus was challenged by the economic crisis o f the mid-l970s, andas a result, radical changes instate and society relations took place, the state giving up its role as the referee o f society (Pinarcioglu and Igik, 2001). In 1980, by the January 24 Decisions, a new phase inTurkisheconomy opened up, which aimed at the liberalizationo f economy, and which adopted export-oriented policies instead o f import-substitute industrialization (Demir, 1993). The state, which is now presented inpublic discourse as a burden on the economy because of its too many employees, and as a corrupted institutionin which clientalist relations dominate, i s shrinking in terms o f both its size and fimctions in politics and economy. It affects lower classes in a double way: those who are employed in government jobs face the danger o f losing their jobs and experience decrease inreal wages (Boratav, 1990 and C elasun, 1989 inDemir, 1993), and suffer from the deterioration o f public services (especially ineducation, healthand social security). While upper classes are able to take advantage o f the high-tech, well-serviced and very expensive services (private schools, private hospitals, private insurance companies), the lower classeshave to be content with the services provided by the state, albeit how inadequate and sometimes inaccessible they might be. Furthermore, as the number o fjobs in the formal sector (both private and public sector) decreases due to the type of economic policies adopted, the informal sector expands (Demir, 1993), andthis means insecure andlow-payingjobs for the poor. With the efforts for global integrationofsociety, anincreasing numbero fyoung and educated people have begun to be employed in the producer services o f transnational/ multinational corporations that opened branches inTurkishmetropolises, mostly inIstanbul. For example, the number o f producer service firms established through the Foreign Investment Directorate o f Treasury and ForeignTrade (FDI) was only six before 1980, and only 10 between 1981 and 1984, while it reached 262 inthe post-1984, the two-year period 1989-1990 being the peak (Tokatli and Erkip, 1998). They have their distinct way o f life, and tend to exclude the poor from their lives and environment, for example, by building their 'gated communities' away from the city center, protectedbyhigh-techsecurity systems and private security personnel (Oncu, 1997; IQikand Pinarcioglu, 2001). Inthis way, the recent tendency o f the rich to separatethemselves from the poor physically through creating their own homogeneous (both culturally and economically) communities, which was well observed and described in the American context (e.g. Davis, 1992 about L.A.), has been replicating itself in the Turkish context, yet with some differences and variations. Here it should be acknowledged that this 'new urbanrich class,' which emerged as the result o f the 44 integration of Turkish society into the global capitalist economy, i s relatively small in numberandis quitevulnerable inthe face ofthe 'unexpected' movementso fglobal (foreign) capital, as the February2001 economic crisis has shown. Another important characteristics o fTurkish society, which enabled the reproduction o f the system, i s the significant role family plays in the lives of people. Turkish state, despite its intervention in economy (and politics) has never been a 'welfare state' as understood in the West. Even when compared to 'developing countries,' the amount o f money spent by the state on social services (mainly health, education and social security) remains quite low. For example, the ratio of social expenditures to the overall state expenditurein 1998was 63.6 inArgentina, 48.1 inMexico, 71.3 inChile, 26.2 inIndonesia, 26.5 in Philippines, 2 7.8 inKorea and 42.7 in Malaysia, whereas i t was 2 5.7 inTurkey (World Bank, 2001). And only 47.6 % o f the employedpopulation has membership in a social security fund (TUSIAD, 1997). Twenty percent of the population remains outside of any social security programme. Instead of the welfare state model, a welfare model based on family characterizes the Turkish case. However, this model is being challenged drastically by recent economic, social and demographic changes. The consecutive economic crises in Turkey are making it necessary to tighten the control over the limited financial resources, closing the door on 'permissive' policies and practices, for example, conniving at land occupation at city peripheries (BuBa, 2001). As a result, squatter (gecekondu)x'vfamilies, who have been maintaining themselves in the city by using their own resources andnetworks without any formal/ direct support by the state, can no longer do so. Changes inthe demographic structure are another challenge to the traditional family welfare model. Turkeyhas a 'young population': In1935, 50 % of the populationwas below the age o f 21.21, while in 1990 it was 22.21 (SIS, 1996). And the fertility rates are dropping: In1975, the number o f children bom alive permother was 4.06, while it was 3.70 in1990(SIS, 1996), whichpointsto the potentialofwomen's increasingparticipation inthe labor market. This means that more people are seeking employment in the market. Additionally, the number of people above the age o f 65 is increasing: In 1985, it was 2,125,908, while it was 2,417,363 in 1990, which challenges the prevailing practice o f elderly care giving within the family. All this makes it difficult to continue with the traditional welfare model based on family support. But under the present ideologies and realities, there may be no other alternative to this model, and this i s going to magnify the costs o f economic crises on people inTurkishsociety (Bugra, 2001). After having contextualized poverty (and wealth) inthe recent changes experienced in Turkish society as a result of economic restructuring policies and processes, let us consider the various groups that are prone to poverty. 45 UrbanPoverty (0 Gecekondu(Squatter)Peoplein the Cities: TheMass Urban Poor As mentionedearlier, from the early stages o f gecekondu formation inthe 1950sup to the 1980s, the urbanrent inperipheral landwas distributed among the urbanpoor, namely those who migrated from rural areas usually in search for economic opportunities. They were tolerated by governments when they built their houses on the landat the peripheries o f bigcities, largelyowned by the state. This tolerant attitude o f governments canberegarded as clientalism, that is, when the government lacks direct material resources to distribute in return for loyalty, it tends to offer 'non-material' rewards, namely, the "non-exercise o f legal controls or their selective relaxation" (Oncu, 1988; p.45). Today, a largenumber o f families own their gecekondus, and a small but recently increasing number o f families rent them. For example, ina research conducted in Ankara (METU, 2000), 87.3 % o f the respondents owned their houses/ apartments, while 23.3 rented them, and 16 % lived in the houses owned by their relatives without paying rent. InIstanbul'sgecekondu areas surrounding the city, high concentrations o f owner-occupiers were observed in 1990 (Giivenq and Iyk, 1996). In the earlier times of gecekondu formation, there was solidarity and cooperation among gecekondu residents regardless o f their sectarian, ethnic and regional differences since they needed each other in establishing their neighborhoods, negotiating with the government for services and infrastructure. However, in the 1980s, we observe hostility among gecekondu groups (Guneg-Ayata, 1990-91), and sharp competition among the urban poor for the gecekondu rent, which increased significantly as the result o f the laws passedin the mid-1980s regarding squatter housing. As claimed by some scholars (e.g. Senyapill, 1998), the government o f the time (the Ozal government), in order to compensate for the loses lower classes were experiencing as the result o f the neo-liberal policies practiced by the government, attempted to 'bribe' them through permittingthe construction ofup to four- story houses on the gecekondu land. Inthis way, it was aimed both to integrate gecekondu settlements (informal housing) into the formal housing market, and to take the support of lower classes in the implication o f liberal policies, or at least to prevent their uprising. Gecekondu owners have always been aware that their houses are their life-long security, and poor people until very recently carried the hope that one day they would build their own gecekondus. This prevented the urban poor from falling into pessimism and fatalism experienced by the urbanpoor inthe West ('culture o fpoverty' and'the underclass'). And by these laws, they firther came to view their gecekondus as an important means o f economic mobility and a source o fmuch profit. These changes in the laws promoted the commercialization o f gecekondu housing. The increasing rent on gecekondu land coupled with the decreasing role o f the state in economy intensifiedcompetition and increased the number o f actors involved init (Keyder, 2000). Not only the lower classes but also the upper andmiddle classes began to compete for the peripheral urban land. The cooperatives for the middle classes and the private sector oriented to the upper classes to build luxurious villa-type communities, wanted that land, as well as Municipalities to produce large-scale residential projects. And inthe meantime, the 46 traditional small-scale contractors were rapidly replacing single-story gecekondus with multi-story apartment buildings in retum for several apartments to the gecekondu owner. This paved the way to significant economic gains for gecekondu owners: they were once 'landless peasants in the city,' then became land and/or house owners, and finally they owned apartments, and insome cases, several apartments (Senyapili, 1998; Erman, 2001b). Thus, urban land has been a major means o f survival o f the mass urban poor.xv And throughout this process, gecekondu residents have been actively searching for any opportunity that would help them settle in the city and improve their conditions and positions, and have taken advantage o fvarious mechanisms andnetworks, such as clientalist relations withpoliticalparties andcommunity networks. The intensified competition for urban rent since the 1 9 8 0 ~and ~ the increasing passivity o f the state in the regulation o f economy, along with its decreasing role in the provision o f social services which have been traditionally provided by the state, have led gecekondu people to seek protectioninother relations, and through these relations, to obtain strategic positions intheir access to urbanrent (Senyapili, 1998; Pinarcioglu and Igik,2001). As a result, the already present divisions along ethnic and sectarian lines, which have been politicized in the identity politics o f the 1980s (Ayata, 1997), have begun to be used extensively as mechanisms o f unequal distribution o f urban rent. Thus, radical Islam, Alevism, Kurdish origin, common regional origin (hemJehriZik), all are acting today as collective identities on the basis o f which local networks develop and compete with one another aggressively for the rent ingecekondu land(Erder, 1997). Inthis competition, some gain and some lose. An empirical study carried out in a gecekondu district in Istanbul (Sultanbeyli, known as an Islamist district) demonstrates who winners and losers might be. Intheir research during which they conducted 611interviews, Igik and Pinarcioglu (2001) found that the better-off gecekondu households (winners, 13 %) were those who hadmoved into the settlement during its early stage o f development (20 years ago or earlier); the majority o f the 'heads o f the household' were employers (mainly in construction and real estate, and some in manufacturing), and those who were employees mostly worked in the Municipality (governed by the Islamist Party-- the Virtue Party; it has become the Felicity Party recently) or in the associations established by Islamist groups; they were all home owners, some owned other houses/ apartments inthe neighborhood, and some even owned apartments in other neighborhoods in Istanbul; and all had television sets, refrigerators, fully-automatic washing machines and telephones, and 79 % owned cars, 35 % owned dish washers and 35 % owned mobile phones.xv' They were mostly migrants from Trabzon and Giimughane (two major cities in the Black Sea) or those who moved to Sultanbeyli from other districts in Istanbul. On the other hand, inthe poorest gecekondu households (losers, 24 %), men mostly worked in daily construction jobs which they found in the local construction workers' market ('amelepazan') or as street vendors and sellers o f h i t s and vegetables in the local market. They were almost all tenants (90 %, and 6 % lived inthe houses o f their relatives without paying rent) and only 4 % owned their houses. While the majority owned television sets (86 %) and refrigerators (90 %), and some owned washing machines (57 %) and telephones (53 %), very few owned dish washers (4 %), mobile phones (8 %) and cars (8 %). The poorest 15 % o f the residents did not own telephones, washing machines, dish washers or cars. They w ere recent comers to the neighborhood (moved five years ago or later), and many were migrants o f Kurdish origin from the 47 Southeast, who moved directly from their villages mainly in Van, Elazig, Siirt, Tunceli, Hakkari and Diyarbakir to this neighborhood. Pinarcioglu and Igik (2001) interpret their data as 'poverty inturns' (n8betleJeyoksullull). It refers to the process inwhich established squatter residents are able to improve their economic conditions by exploiting the newcomers into the neighborhood. Through their involvement in land occupation and gecekondu construction inthe early stage o f the formation o f the gecekondu neighborhood, they find the opportunity to own land (although this land ownership i s illegal by definition, in practice it is usually recognized by the authorities as a result of intricate negotiation processes based on clientalism) and to build houses on it, and they make gains on the newcomers inthe later years, selling the landthey had plotted, renting their houses, andthe like. Inthis way, they handtheir poverty over to them. However, this optimistic picture o f urbanpoverty, which implies temporariness and the possibility o f getting rid o f it, is rapidly disappearing in the face o f recent economic crises,xvii which have led, among other things, the private sector's aggressive attempts t o appropriate urbanrent. Furthermore, the networks based on common origin (hemphrilik),which acted as an important mechanism o f support and social mobility among rural migrants, have become selective and exclusionary (Erder, 1995, 1996). As winners and losers become crystallized inthe years spent in the city, andruralmigrantscompete for diminishing resourcesand opportunities, migrant networks tend to include those who have something to offer and to exclude those who will be a burden on the network. While the former group includes families inwhich there are adult men, the more the better (hence reproducing the dominant position o f men and adults in the family), the latter group includes the elderly (especially elderly widows), the disabled, those families in which there is chronic unemployment, alcoholism and drug abuse, andmore importantly, those families fkom the Southeast which are the latest comers to the city andwhich migrated as the result o f terrorism intheir region ('forced migration') (Erder, 1995, 1996). They are disadvantaged in the city compared to earlier migrant groups for a number o f reasons (Erder, 1998a; SPO, 2001a). First, due to their forced nature o f migration, they come to the city unprepared without having established their networks and support systems. This is not the case with voluntary migration, which i s usually chain migration. Except for the initial period o fmigrationinthe late 1940s and early 1 9 5 0 ~people have been migrating to cities by receiving help and ~ information from their relatives and fellow-villagers already settled in the city. Secondly, their relations with the village terminate upon their migration, which is not again the case with voluntary migrants who usually continue their relations with their villages, bringing from the village various food stuffs (flour, lentil, cracked wheat, tomato paste, yogurt), timber, and the like, which contribute to their economic survival inthe city. Thirdly, they may be discriminated against due to the stigma placed on the people from the Southeast as potential terrorists. A nd1ast but not 1east, they migrate as crowded families with many children, which may create further problems in the survival in the city, such as finding housing. In a research conducted with 28 women and 35 men who migrated from their village inMardin to Istanbul upon their abandoning their village when they had no option buttojointhe korucu (thecivilian armedgroups formed bythe Stateto fight against terrorist groups), it was observed that they lived in one-room shanties covered with oilcloth 48 (linoleum) on the periphery o f Istanbul, which the author calls the "tent village'' (Kayacan, 1998). In those nine shanties lived 27 children, and 14 o f them attended school. The families tended to keep their daughters at home, and this was because o f both their patriarchal tradition andtheir very limitedincomes that requiredthe families to select among their childrenthose who would be sent to school. Furthermore, the families expressed their concern for their children in school whose education lagged behind the usual educational period, for example, their sons o f 12 years o f age attending the second grade ando f 10years o f age attending the first grade. The children attended a primary school located at the 15 minutes o fwalking distance from the 'tent village,' andthey walked on a dirt pathto school, which got very muddy in the winter. The women also suffered very much fi-om the consequences o f their 'forced migration.' They had to do laundry in the open by using the water carried from the nearby forest, which meant cold andmudinthe winter; and they fed their families with the bread they made inthe tandzr (earth oven). None o fthe households had refrigerators or washing machines, yet each household had a television set and a small cassette player. Thirteen men and three women had paid jobs. Except for two men, they were employed temporarily and lacked social security. As a conclusion, we can say that the socio-economic system, which, with its clientalist politics, interventionist state, 'illegal' housing, and 'informal' job market, was able to provide channels to the urban poor to escape poverty until recently, and hence creating optimism in them and enabling them to developing their capabilities to struggle against poverty, cannot maintain itself any more. This means a new type o f poverty in urban Turkey, which is a 'no-way-out' situation. An example o f the new urbanpoor in the Turkish context is the 'scavengers' inthe city. In an empirical research, Ozgen (2001) investigated 'garbage' as a commodity and 'profit-making on garbage' as a hierarchical process inwhich various actors were involved, rangingfrom people collectinggarbage inthe streets and at garbagecollection sites to small, medium, and large-scale garbage dealers (hurdaci) and national and/or international buyers. Inthe research, itwas found out that, the 'garbagepeople' at the lower levels ofthe 'garbage hierarchy,' namely, garbagepickers, regardless o f how muchmoney they made, were bound to remain unintegrated into the society. For example, in Denizli, a rapidlygrowing city which had an organized garbage collection and re-cycling process, garbage pickers were mostly recent migrantso fKurdishorigin fi-omSoutheastern andEastem Anatolia who lived insquatterhousesontheperipheries ofthe city wheretheyhadhighlyinadequate education and health facilities. They regarded picking garbage as their permanentjobs ('professions') on which they would live also inthe hture (they said they made muchmoney in collecting garbage which they could never make in any other job available to them), and they took their young children out o f school to work in 'their family business' (many families sorted out garbage intheir homes). The author ends her article by stating that 'the new poverty in global times' cannot be measuredby the amount o f money people make: regardless o f how much they make, when their access to power i s closed andwhen they are renderedinvisible inthe city due to the (lack of) status of their jobs, they are excluded fi-om the mainstream society, andthis exclusion tends to reproduce itselfinthe next generations. 49 There i s a studyworth mentioning inthis context. Ina survey carried out inAnkara in 1992 (Morgol andGitmez, 1995), which aimed to identifythe different types ofthepoor, poverty was measured in terms o f both structural (income) and cultural/ psychological (individual locus o f control) dimensions o f living conditions. Among the variables used, income, family assets, neighborhood, education and housing conditions were used as indicators o f objective livingconditions, and locus o f control, life satisfaction andhappiness were used to measure subjective well being. The number o f years spent in Ankara was included in the analysis as the indicator o f experience in urban life. Interviews with 550 persons 1ivingin 14 different neighborhoods o f Ankara w ere c onducted, 438 out o f 5 50 living in lower income neighborhoods. The authors identified three major groups interms o f their objective poverty, namely, doers (those with highest level o f economic means), losers (those with lowest level o f economic means), and accommodators (those inbetween). Interms of subjective poverty (that is, how people evaluate their economic conditions in comparison to other people, including relatives and neighbors, and to their past, and how they perceive their future), doers had the highest level o f satisfaction with their lives, although their economic conditions lacked significantly behind the non-poor, and they felt themselves in charge o f their lives more than the others. They assessed their own living conditions as considerably better than five years ago and their living standards higher than their close relatives and neighbors. On the other hand, accommodators evaluated their economic conditions as the same compared to the past. And losers regarded themselves as worse-off when compared to the others, and they exhibited the sense o f loss o f control over their lives, and had the lowest scores on the variables life satisfaction and happiness. Interestingly, both doers and losers were relatively new inthe city, whereas accommodators had longer duration o furbanexperience. The authors explained their findings inthe context o f Turkey's economic development and social change. They stated that, while "the majority of Turkey's urban poor had been accommodators prior to the 1980s," as a result o f the structural adjustment policies in the early 1980s, which ''opened up a new phase o f economic development in the country" (p. 419), doers and losers have emerged as new categories o fthe urbanpoor: The doers and losers can be considered the products o f the post-1980 economic conditions and the accompanying cultural climate in the country. The effects o f lifting economic subsidies and the incremental moves toward market-oriented policies seem to be double edged, creating both upward and downward mobility and the associated psychological attitudes. Many slipped down on the hierarchy, some found ways to accommodate themselves at lower levels. Others lost control over their lives and became losers. On the other hand, some used the economic opportunities that arose from the new economic policies to lift themselves up somewhat on the hierarchy, or at least gained hopes that they could move upward. These are the doers (pp. 419-420). As policy recommendations, they mentioned the need "to establish safety nets for losers to prevent them from slipping further down" (p.420), and for accommodators, measures to induce behavioral changes in them (namely, to motivate them for economic mobility), such as education, should be used. 50 This study contributes to our knowledge on the urban poor by pointing out to the possibility o f economic improvement for some as a result o f economic restructuring. Yet, it fails to inform us about who this group o f doers are interms o ftheir economic activities, job status, family composition, ethnic background, and the like, as well as about their proportion in society. Furthermore, it fails to recognize the social mobility experiences o f accommodators throughout their lives in the city until they have reached their 'stable' economic conditions. Also, it is necessary to recognize the fact that following the recent financial crises, there have emerged threats to the 'stable' positions o f 'accommodators,' and the possibility o ftheir downwardmobility has increased. In another research (Sallan-Giil, 2001) conducted with 36 persons who received assistance from the 'Fak-Fuk-Fon' Fund, they were categorized into three groups, namely, those who were the victims o f the recent economic crises (temporary poverty), those who had been either unemployedor had inadequate incomes for some time [they lacked skills or education; majority worked in the informal sector without job security as porters, construction workers, peddlers, painters, scrap dealers, and the like (2/3 o f respondents lacked social security); and were tenants], and those (mostly elderly andwidows) who lived merely on assistancethey received from the Fundand help from their neighbors (some lived in deserted houses). Another reason which made especially women to apply to the Fund was their disabled children. R espondents1ived inuhealthy conditions; they spent their money only on food; they remained mostly insidetheir neighborhood to save transportation costs and shopped inlocal markets on credits; they had furniture and clothes given to them byother people. Despitethe hardshipthey lived in,they were not pessimistic. The majority defined the cause o f their poverty as lack o fjobs (1S), and only four persons had fatalistic views. Especially women emphasized the significance of social security in maintaining a decent life. Seventy-five percent o f respondents (Le., those receiving financial assistance from the State) wanted to work (others were old or sick). Majority said they thanked God for theirpresent situation becausethey knew some other people livinginworse conditions. (ii) Residentsof Inner CityAreas: TheNew Urban Poor and the Underclass Inadditionto urbanpoormasseswho liveinsquattersettlements andwho havebeen able to cope with poverty and even to improve their economic conditions, discussed inthe section above, there i s an increasing number o f the urbanpoor residingin inner city areas, who have come to represent the 'urban underclass' in the Turkish context. The most important characteristics o f the urbanunderclass inthe Turkish case i s their exclusion from social networks and solidarity ties o f migrants(Erder, 1995, 1996). Ina study conducted in inner city areas o f Izmir, the third biggest metropolitan city o f Turkey, 323 households residing in 12 different neighborhoods were interviewed (Sonmez, 2001). Seventy-six percent o f the household heads had migrated from another region to Izmir, and o f those, 44 % had come from Southeastem region (the majority coming from the province o f Mardin). Inthe study, astrong correlationwas foundbetweentheperiodofmigration andtheregion that contributed to migration: those from the Southeast had migrated after the 1980s, whereas those from the Aegean region, where Izmir is 1ocated, had migrated before the 1980s. Thus, the newcomers were mainly from the Southeast. They were young families with many children. The number o f persons in a household exceeding five was 38 % and 51 the median age o f couples was between 20-24. Among the early comers, many had improved their financial conditions and moved out to apartments in the formal housing market, rentingtheir low-qualityhousesto the newcomers, andthose who were still residing inthe inner city areawere the poorest, the elderly andthe lonely. Inbrief, inthe inner city areas o fIzmir, resided two major groups, namely,the lonely elderlyandyoung families with many children. The majority was tenants. In terms o f their employment, they were concentrated in the informal sector without job security and social security benefits (23 % employed in textile production, 10 % in leather industry, 19 % in service jobs as guards, nurses, waiters and servants, and 10 % worked as street vendors). They mostly worked in work places inthe central city close to their homes. Inthis research (Sonmez, 2001), the social ties andneighborly relations incentral city areas were found to be weak. The fact that the majority o f residents were tenants and displayed highresidential mobility (58 % o f respondents had moved at least twice) acted as a significant factor, preventingthe formation o f strong neighborly relations and solidarity ties, as well as a sense o f belonging to the neighborhood, and hence creating uncaring attitudes towards their neighborhood and emotional detachment from it (75 % o f respondents said they would leave the neighborhood as soon as they found the opportunity). This ledto increased crime rates (the area displayed the third highest crime rates). Residents complained about robbery and children's crimes, as well as underage girls belly dancing for money. Many respondents fixther complained about the lack o fhelp and support from their neighbors, especially inthe case o f care to children, the sick and the elderly. Moreover, the stigma placed on the area as the place o f criminals and the very poor prevented the residents from identifying with their neighborhood and fiom improving it. Inthe sections above, the focus hasbeenonbigcities inthe West which have been receiving migration since the 1950s. However, the cities inthe Southeast have their unique qualities, particularly due to the GAP Project (the Southeastern Anatolia Project, which includes the construction o f the Atatiirk dam), and they deserve special attention. Inthis context, Urfaas a case i s considered inthe following section. InUrfa, the city which is the locus ofthe 'GAPProject' and anexample ofcities in the Southeast where feudal relations which have dominated the area for centuries tend to dissolve, about 70-80 % o f the city population live inpoverty in squatter areas on the city's peripheries (that is, about 450- 500,000 out o f 600- 650,000), and there i s a strict spatial differentiation between the better-off and poor residential districts (the rich moving to suburbs lose to fashionable malls, such as YIMPAS) (Can, 2001). The poor are mostly migrantswho movedfromtheir villages t o the city eitherwhen their feudal 1ords (a@) released them from their ties and loyalty, or whenthey sold their small landwhich proved to be economically insufficient as a result o f the region's agricultural mechanization and integration into the capitalist economy by the GAP Project. There are some who move to other cities, such as Diyarbakir, Adana, Mersin and Antalya, and yet many move to their provincial city (Urfa). They usually continue their subsistence living in the city, growing their own crops, keeping husbandry, making their own bread, and the like. They are different from those villagers who hadto leave their villages because o f terrorism (e.g. those in Sirnak, Siirt, Hakkariand Diyarbakir) and who experience absolute poverty. Yet, their 52 economic conditions are very limited. Many work as seasonal agricultural workers, and some work as construction workers, street peddlers and porters, those jobs which lack reliable income. Those families in which one o f the persons has a permanent job i s very rare, and it i s considered as a prestige symbol (Can, 2001). The increase intheir numbersi s decreasing their bargaining power, causing wage drops. Most o f the houses lack running water and sewage system, and residents tend to use illicit electrical power. Yet, many families own television sets, which are the only high-tech products in their homes (Can, 2001). The migrant families inUrfa are very crowded: many o f them have more than five children. These families use their children's labor as a means o f survival in the city: Children, even those at the age o f seven or eight, usually work, many as shoe-shine boys or insmall workshops as apprentices. Onthe other hand,religious conservatism is preventing women from engagingin employment, except for the case when they work inthe fields as laborers together with the other members oftheir families. Marriedwomen are not allowed to attend courses given by CATOM (Multi-Purpose Social Centers) or by the Governor's Office (e.g. c arpet-weaving c ourses). R eligion and tradition w ork hand inhand to keep women out of schooling, although it seems that poverty i s the real cause behind it (Can, 2001). Religion i s also important inreproducing conformity and fatalism, suppressing any rebellious action. On the other hand, the recent economic crises have paved the way to reactions against the state, manifestedin street protests. Furthermore, the problemo f 'street children' and 'street gangs' i s becoming visible inUrfa although it i s still less common than inDiyarbakir andGaziantep (Can, 2001). InDiyarbalur,which is mentionedas the city ofthe poor, 23.8 YOofthe households had less than $53.3 annually in 1996, 41.2 % had incomes below the minimumwage, and 64 % lacked social security (Goktiirk, 1998). Diyarbakir i s much affected by 'forced migration' inthe Southeast. Ofthe 369 migranthouseholds (2,622 persons), 272 households (2,059 persons) migrated due to forced migration (TMMOB, 1998). Ina survey conducted inDiyarbalurwith689 households (TMMOB, 1998), one-third ofrespondents saidthat they did not receive any help fi-om their relatives. It points to the very difficult economic conditions o f migrants, preventing them to help each other despite the fact that in the Southeastbondsbetween relatives are very strong. InMersin, only 38.7 YOof the household headsworkedinthe formal sector, andamongthe migrants to the city, 43.1 % worked inthe informal sector and 12 % were unemployed (GoktiirkandKaygalak, 2001). Survival Strategies of the Urban Poor When families experience poverty due to a decreaseinthe amount o fmoneycoming into the household, one way to cope with the new situation i s to increase the number o f household members that participate in the labor force (Demir, 1993). Thus, women, childrenand the elderly start working outside (sometimes inside)the home, while mentake on a second, or a third, job. However, since the formal sector lacks enough employment positions to absorb the increase in labor supply, they end up working inthe informal sector for little paywithout ajob contract andjob security. 53 Women, through participating in money-earning activities, producing at home, and saving on consumption, play an important role in the survival strategies o f the urban poor (This issue is expanded inthe 'Women and Poverty' section). The gecekondu gardens the urbanpoor have, andthe h i t trees and vegetablesthey grow inthem, may contribute to the economic survival o f the poor. However, as time spent in the city increases, food production ingardens tends to decrease. For example, while 61.3 YOo f the households had gardens, only 24.5 % grew vegetables intheir gardens (METU, 2000). Yet, this tendency may be reserved in the face o f intensifylng poverty following the economic crises in 2000 and 2001. On the other hand, it should be acknowledged that the apartmentization process going o n inmany gecekondu s ettlements and the t endency t o use gecekondu 1and more intensivelyis eroding this option as a survival strategy for the urbanpoor. Usually poor families cut on their consumption expenses: 44.3 % o f the families cut on food (especially onmeat, milk and milk products) and 20.4 % cut their expenses o n clothing and fumiture, while 35.3 % cut on their expenses in general (METU, 2000). It i s interesting to note the tendency to cut more on food than clothing and fumiture. There may be two explanations. First, they may not be cutting on clothing and h i t u r e because they may not be spendingany money on them, and secondly while what they eat inside the home i s not known to others, hence may be not important, what they wear and have at home may beimportant since they are noticedbyothers as status symbols. Continuing relations w iththe v illage, bringingfood and other materials from the villages may be another survival strategy for the urbanpoor. Rural migrants' relations with their villages have beendecreasing over the years as they become more settled inthe city. For example, only 55.6 % o f those who migrated to the city 40 to 50 years ago continued their relationswiththe village, whereas 89.3 %o fthe recent migrants didso (METU, 2000). However, the reliance on the economic support received fiom the village may increase under the recent economic crises. Inan empirical study conducted inAnkara's gecekondu areas in 1999 (METU, 2000), 75.5 YOo f the respondents continued their relations with their villages, andreceiving economic support fiom the village had increased inrecent years. On the other hand, there are several factors that prevent it. First, some migrants have already sold their landinthe village and/or do not have relatives livinginthe village anymore. And secondly, the 'forced migration' fiom the villages inthe East and Southeast has beencutting off migrants from their rural properties and resources. The 'village-city project' can be one way o f solving this problem. Another survival strategy o f the urban poor is the solidarity and support they develop with 'their own people.' This refers to ethnic origin, religious sect, being from the same region (hemgehrilik-- c ommon o rigin), as well as political patronage. A s we have seen, the tendency o f people to differentiate themselves on the basis o f ethnicity, sect and hemgehrilik is getting stronger, and more political, under the influence o f the intensifylng competition inthe urbanjob market and for urbanland. Research has demonstrated that the majority o f migrants are members o f 'village solidarity associations' (koy dayanzgma dernekleri). Through this membership, they reach 'their own people,' and receive help and services. Yet, whether these organizations cater for the needs o fthe very poor i s disputable: they tend to include those relatively better-off migrants who can afford paying the 54 membership fee, and exclude the very poor (Erder, 1995, 1996; METU, 2000). These associations are usually organized on the basis o f shared ethnicity and sect, and fbrther exclude those having different ethnic and sectarian affiliations (METU, 2000). Furthermore, political clientalism i s often used as a survival strategy by the urbanpoor. For example, Islamist Municipalities distribute free coal and food to their supporters (direct help), and they give them larger shares in the legalization process of the gecekondu land (indirect help). RuralPoverty Poverty i s largely a matter o f the rural in the Turkish context. Many families have been migrating to cities to escapepoverty intheir villages. Rural households received 23.2 % o f the national income in 1987, whereas urbanhouseholds received 76.8 % (SIS, 1990). Thesenumbers were 39.9 % and60.1 % respectively in 1973. Sixty-five percent o fthe poor live inrural areas, whereas 35 percent live inurban areas; and while 30 persons out o f 100 are poor in rural areas, in urban areas 17 persons out o f 100 are poor (Dansun, 1997). In rural households, the view that as the size o f the household increases, it contributes to the household's economy interms o f unpaid family labor does not hold true any more. On the contrary, the largerthe household size, the higher i s the number o funemployed inthe family (Erdogan, 1997). InruralTurkey, wecantalk aboutthreemajortypes ofproduction, namely, capitalist production based on large-scale land ownership, production for the market by small-scale land owners (mostly uses unpaid family labor; 60 % o f the rural population is engaged in this type o f agriculture), and subsistence production. InEastern and Southeastem Anatolia, large-scale landownership has not tumedinto capitalist production, preservingits traditional (feudal) character to a large extent. Thus, we can conclude that the agricultural sector has failed to organize itself in capitalistic terms, contributing to capital accumulation only in limited measure. This has created a strong tendency in peasants to move to cities. Inthe 1950s, the structural changes in the agricultural sector, such as the introduction of tractors, fertilizers and irrigation systems, which aimed to increase productivity in the sector and to integrate it into market economy, pushed large numbers o fpeasantto cities. The mass rural- to-urban migrationhasbeen going on since then inan escalating manner. This has produced squatter settlements on the city peripheries that are the residential environments o fthe urban poor. However, still an important number o f peasants remain invillages. They are mostly small-scale land owners, and they survive on government subsidies, which might be viewed as the state's paying to peasants for their 'forced-residence,' as a type o f social security paid by the state to the rural poor (Dansuk, 1997). Inthis way, while poverty in rural areas is kept under control in the short run, in the long run, it is reproduced by keeping peasants away from the 'real forces o f the market,' creating an agricultural structure with low efficiency and limitedcapital accumulation (Dansuk, 1997). The recent attempts to cut back government subsidies (Bugra, 2001) means chronic rural-to-urban migration and persistent poverty both in rural and urban areas (Dansuk, 1997). In addition, the economic restructuring process that i s takingplace inthe Southeast inthe context o f the GAP Project (the Southeastem Anatolia Project), in which the prevailing feudal relations are to dissolve and to release a large number o f people from their agricultural engagement, is expected to 55 create a strong tendency inpeople to migrate to cities in search of employment. However, this new group ofmigrants willbe the most disadvantagedamongmigrants since 'poverty in turns' (Iyk and Pmarcioglu, 2001) period is over, competition over urban land and urban jobs (both in the formal and informal sector) i s intense, excluding newcomers fi-om the housing andjob markets. When we analyze rural poverty in terms o f regional distribution, we see that in the Aegean and Marmararegions, 5.4 % o fthe ruralpopulationi s below the poverty line, inthe Mediterranean region this numberi s 22 %, inCentral Anatolia 29 % (the average income o f these families i s 19 % lower than the amount o f income defined as the poverty line), inthe Black Sea 11.6 % (the poorest get only 3.56 % o fthe total income o fthe rural households in the region), and in the Eastern and Southeastern Anatolia 32 % (the household size in the poorest segments in this region i s 7.87, while it is 6.05 in the poor rural households in average, which means that the largefamily size acts as decreasing the amount of income per personand hence causingpoverty) (Dumanli, 1996). Interms o f inequality withinrural and urban areas, we can say that there is relatively less inequality in rural areas (the gini coefficient i s 0.41) than inurbanareas (the gini coefficient is 0.52). Thus, we can conclude that poverty is more commonly experienced inruralthanurbanareas (Dumanli, 1996). UnevenRegionalDevelopmentand Poverty InTurkey, there is unevendevelopment and highly unequal distribution of wealth among the regions, the West being highly advantaged and the East being highly disadvantaged. The metropolitan cities located in Western Anatolia receive much o f the national income. For example, Istanbul, the biggest metropolis in Turkey where extreme poverty and extreme wealth coexist in sharp contrast, received 21.8 YOo f the national income in 1995 (it was up to 27.5 % in 1998, which was 14 % 20 years ago), followed by Izmir(8 YO),Ankara, the capital (7.5 %), Kocaeli (4.4 YO),andBursa(4 YO)(Sonmez, 1998). On the other hand, many cities in Eastern Anatolia receive minimal or no shares fi-om the national income. For example, Bayburt, Hakkari and Ardahan received no shares fi-omthe national income, and Tunceli (0.1 %), Igdir (0.1 %), Bingo1(0.1 %), Simak (0.1. %), A@ (0.1 %), Bitlis (0.1 %), and Mu?(0.1 %) received only very minimalshares (Sonmez, 1998; pp. 260-292). In1996, the yearly income per person was $7,096 inKocaeli (this changed after the earthquake of August 17, 1999), $4,286 in Izmir, $4,286 in Istanbul, $4,242 in Kirklareli (all are cities in the West), and $3,976 in Ankara (the capital), whereas it was $654 inMug, $667 inAgn, $813 inBitlis, $863 inHakkari,$915 inBingol, $972 inSimak, and $997 in Van (all are cities in the East). In the Eastern and Southeastern Anatolia Regions, 47 out o f 100persons live below the poverty line, whereas this number i s down to 28 in the Mediterranean Region, 27 in the Central Anatolia Region, 22 in the Black Sea Region, and 5 inthe Aegean and Marmara Regions; and out o f 100poor persons, 40 live in the East and Southeast, 26 live in Central Anatolia, 18 live inthe Black Sea, 15 live inthe Mediterranean, and 6 live inthe Aegean and Marmara (Dansuk, 1997; p.44). Inthe Aegean region, the percentage o f persons living in absolute poverty i s much less (3 %) than in Eastem and Southeastem Anatolia (18 %) (Erdogan, 1997). On the other hand, interms o f the unequal distributiono f income, the Marmara Regionhas the highest inequality (the gini 56 coefficient i s 0.56), whereas Eastern and Southeastern Anatolia has the lowest inequality (the gini coefficient is 0.37 and 0.38 respectively) (SIS, 1994). Intems ofhealth services, in1995, while there were 329 persons per doctor in Ankara, 504 in Izmir, and 541 in Istanbul, there were 4,533 persons per doctor in vim&, 3,105 in Am, 3,067 in Sanliurfa, and 3,030 in Mu?(Sonmez, 1998). Interms o f schools, there i s no reliable statistics today, although it i s common knowledge that many schools in the East and Southeast suffer from lack o f or inadequate teaching personnel, as well as infrastructure. Additionally, in 1995, while there were 1,295 private car ownership per 10,000 persons inAnkara, 932 inIstanbul, and 912 inIzmir, this number was 28 in Sirnak, 42 inMug,44 inTunceli, 44 inHakkariand 44 inAEn (Sonmez, 1998). Additionally, the production levels o f cities vary significantly. In 1999, the Marmara Region, by its share of 36.5 %, made the highest contribution to the National Product, while the Eastern Anatolia Region, by its share of 4.3 %, made the lowest contribution (SIS, 1999). Inverybroadterms, this inequalityinthe distributionofwealthbetweenthe Western and Eastern regions is the outcome of the historical uneven development o f the country, including the economic policies favoring Western regions over the East and the Southeast in whichthe harshclimate andunproductive soil o fthe East, amongother factors, have played a role. In addition, the military confrontation between the Turkish armed forces and the P K K (the Kurdistan Workers Party) since the mid-1980s acted to accelerate poverty in the East and Southeast. Those villagers who could not survive in their villages any more because o f political and economic reasons had to move out o f their villages (since 1984 about 2,500 villages were evacuated) to cities, mainly to Istanbuland Izmir inthe West, to Adana, Mersin and Antalya in the South, and to Diyarbalur and Urfa in the Southeast, forming their 'ghettos' on the outskirts o f the cities, and in some cases, their 'tent cities' or 'shantytowns.' These people hadto leavebehindtheir landthat theyusedto liveon, coming to the city without any economic resources. This new wave o f migration brings the poorest rural families to the city which would have never migrated if there had not been such changesintheir region, and it t akes place as mass migration, dragging children and the elderly in the families to the city without the necessary preparations to accommodate them (Erder, 1998a). The highrates of unemployment in cities, especially in the Southeast (for example, inDiyarbakir the unemployment rate i s 70 YO,and it reaches 95 % inthe migrant group) hindersnewcomers' participationinthe labor force. Ina gecekondu neighborhood in Pendik (Istanbul), 32 persons shared one house (theyhadmovedfrom the Southeast), and in the family only five persons were working; the children worked inthe street, selling tissues, washing car windows when cars waited at the red traffic light, and the like (Erder, 1998b). Since their main concern was to bring home money needed for the basic daily survival of their families, the children from the recent migrant families moved from one marginal activity to another, and many lacked the opportunity to attend school or to get vocational training(Erder, 1998b). 57 To combat with severe poverty and political instability inthe East and Southeast, the government i s working on a new project, namely the 'Village-City Project' (Koy-Kent Projesi) aimed at relocating people back in their villages, and Mesudiye i s chosen for the pilot project. Women and Poverty At the Fourth World Conference on Women held in Beijing in 1995, the feminization o fpoverty was a major theme, which addressed women's worsening economic conditions as the result o f recession and adjustment policies in their response to the recent global economic crisis. It was emphasized that 70 % o f the persons below the poverty line were women. In an empirical research conducted for the World Bank (Ecevit, Erman, Kalaycioglu and Rittersberger-Tiliq, 1999) in four cities in Turkey (namely, Ankara, Istanbul, Izmir and Mersin), it was found that women's poverty inthe Turkish context was largely a family issue rather than an individual problem since the family occupied a central position in Turkish society, and divorce and women-headed or single-family households were few in the poor segments o f society compared, for example, to Latin American societies. For example, inan empirical study (METU, 2000), 281 women respondents were married and 13 were widows, and only one was divorced and one was single; and in 13 families out of 300, women were family heads. [The percentage o f women-headed households in urban Turkey in general is 11.1 YO(Hacettepe University Institute o f Population Studies, 1998)l. This should not, however, make us ignore women-headed households. They are those which usually face acute poverty. In women-headed households, in every socio-economic group, the annual income was found to be less than those headed by men, both in the urban and rural (Kasnakoglu, 1998), and the difference between men- and women-headed households increased as the socio-economic level o f the household decreased, that is, in the lowest 20 YOsegment, the income o f women-headed households was one-third o fthe men-headed households. And 80 % o fthe working women lacked social security, while 47 % o f the working men did not have social security (Kasnakoglu, 1998). Furthermore, women make 66.35 % o f the lowest paid labor, while men constitute 33.65 % (Dansuk, 1997), which points to women's difficulty o f economic survival without a man's income. The majority of women in Turkish society lack independent economic resources and control over financial matters (SPO, 2001). In a survey, those women who lacked any individual income was found to be above 77% in every income group (Karaduman, 2000). Inthe men-headed households, family's worsening economic conditions bringnew responsibilities andburdens for women. When husbands are unemployed (a rising number o f 'household heads' lose their economic means when factories andworkshops are closed, or employees are laid off when their employers (both in the private and public sector) go for labor reduction, or self-employed and small-scale employers g o bankrupt), or when they lack steady jobs, women in poor families participate increasingly in money-eaming activities, ranging from cleaning houses and offices to doing piecework at home, from working in clothing workshops to selecting beans, lentil, and the like in storage houses (Ecevit, Erman, Kalaycioglu and Rittersberger-Tiliq, 1999; White, 1994; Demir, 1993). Between 1980-1990, the number o f women who participated inthe labor force increased by 58 88.3 in Istanbul (Sonmez, 1996). Women fiu-ther knit, sew and do needlework at home to sell. Earnings women receive through their employnent in the informal sector enable the family's economic survival, andinthis way the decrease inreal wages inthe formal sector i s accommodated. Ina research conducted inMamak(Ankara) in 1990, inwhich 111families were interviewed(Demir, 1993), two-thirds of the women employed in the informal sector had startedworking after 1980, especially after 1985, and they were engaged in childcare, housecleaning, knittingand sewing (Demir, 1993). The jobs available to women in poor families are low-paying jobs without job security. They are seen by the families and usually by the women themselves as 'not-real' jobs done temporarily until the family secure back or improve its economic standing (Erman, 2001a). Peasant women prefer city to village because they expect to be 'housewives' in the city, living comfortable lives away from the burdensome work o f the village (Erman, 1997a). And they tend to stop working in the city when their economic conditions improve. Here it i s important to mentionthat, while the concem o f many rural migrant families for family honor (namus) tends to keep women inside the home, the increasing poverty makes women seek paid employment outside the home. When the concerns for the family honor are so strong as to preventwomen from working outside the home, or when the family suffers acute poverty, children may be placed injobs, sometimes part-time(e.g. working on weekends in their free time from school), and sometimes full- time, droppingout ofschool. Women's contributions to the survival o ftheir families facing poverty are not limited to their participation in the labor market. Women, as a survival strategy, try to economize on consumption (Ecevit, Erman, Kalaycioglu and Rittersberger-Tilig, 1999; METU, 2000). Inan empirical study conducted inthe gecekondu areas ofAnkara (METU, 2000), 82.3 YO o fthe women contributed to the familybudgetby engaging infood production at home. For example, they monthly made bread at home ('winter bread'/ In,~Zzkyu$z)insteadofbuyingit at the local grocery store. Ina research (METU, 2000), it was found out that 83.4 'YOof the households produced food at home, such as conserved and dried food (81.8 'YO), pickles (93.5 %), turhuna (dried yoghurt and dough to make soup) and eri@ (vermicelli) (63.2 %) and bread (73.3 %). And 56.2 % did needlework and knitted for their families (METU, 2000). Furthermore, while tend to spare money for their individual expenses, women tend to givepriority to the needs o ftheir families and childrenover theirs (Kog, 1997). Poverty affects women more than men regarding living conditions. Since it is women who spendmuch time at home anddo the housework, when, because ofpoverty, the house i s too cold inthe winters, or humid, or there i s no electricity and/or runningwater in the house, it is womenwho suffer the most from it. Inanin-depthresearchconducted with 50 housewives,xviii respondents stated that it was their husbands who made the decision about where to live, and the husbandstended to choose the cheapesthouse to rent because o f economic considerations. As a result, the women sufferedfrom rheumatism, bone diseases, allergy and respiratory problems, which were caused by the unhealthy living conditions o f their homes ($enol-Cantek, 2001). The women also suffered psychologically when they failed to satisfy the needs of their children, especially o f those attending school, and when they failed to entertain their guests 'appropriately' ($enol-Cantek, 2001). 59 In brief, poverty increases women's work load in the family, and women play a significant role inthe survival o ftheir families duringtimes o f economic crisis. Yet they are disadvantaged in terms o f access to power and decision making in the family. Thus, any international hnding agency should take this into consideration, recognizing the role women play in coping with poverty and the power men have in the family, controlling resources and limiting women's access to them. Credits (business, house and land credits) and funding should be made accessible to women, and poverty alleviation policies should give greater roles and initiatives to women, integrating them into the economic and social life. Any poverty alleviation policy should address gender inequality, recognizing it as a major obstacle inthe fight against poverty inthe long run(Akqar, 1998). Women's poverty and the burdensthey carry incoping with poverty increase as the family's income decreases due to husbands'unemployment or their low salaries and insecure jobs, and as the cost o f living rises due to highinflation and decrease in real wages. This picture describes mainlythe experiences o fwomen inurbanpoor families. The experiences o fwomen inthe Southeast, onthe other hand, havetheir specificities. Traditional patriarchy rulingthe region, makingwomen marryat a very young age, tends to prevent women from attaining schools. InEastern Anatolia 42 % o f the women and in Southeastern Anatolia 55 % o fthe women are illiterate, andinvim& this numberis upto 80 YO(Sevindi, 1998). As a result, many women whose mother tongue i s Kurdish cannot speak Turkish, which hinders their interaction in the public sphere, including hospitals and government offices (Ucan Haber, 2001). Communication and social interactionbecome a serious problem especially for those women who, once having lived intheir villages where people spoke Kurdish, have moved to cities where Turkish i s the prevalent language, and where different values and ways o f life exist. This impedes the women taking active roles in fighting poverty. Furthermore, manywomen inthe East do not practice birthcontrol andhave many children, andthis poses abigchallenge to children's education, especially inthe case o fthose families which are dislocated from their villages and try to survive on the peripheries o f cities. As we have seen in the example o f recent migrant families who make a living by garbage picking, children may be taken out o f school and placed invery low-status and undesirable (socially and health-wise) 'jobs' inorder to contribute to the family's economic survival. Children'sEmployment,Educationand Poverty The definition and measurement o f poverty is moving away "from a minimal, physical survival notion in the direction o f a relative, varying definition which puts increasing emphasis on social survival and starts to attach value to the quality of life that even the poorest in a community should be able to enjoy" (MacPherson and Silbum, 1998, p.2). And as the indicators o f quality o f life, school enrolment rates, along with life expectancy and infant mortality, are emphasized. Education i s one o f the three criteria used to develop the 'Human Development Index' (the others are health and income). Children are society's investment in its hture, and it i s crucial that they receive adequate education to develop their human capacity. And children's education turns out to be inadequate when mothers are not educated (Kagitqibagi, 1998). 60 When we consider the relationship between education and poverty in the Turkish context, we see that 31.94 % o f the poor are illiterate, 24.42 % are literate but do not primary school diplomas, and 38.91 % are primary school graduates. Thus, illiterates and those who do not have primary school diplomas constitute 56.36 % o f the poor. This number reaches 67.43 % in the rural areas, while it i s 53.51 % in urban areas (Erdogan, 1997). On the basis o f this data, we can conclude that there is a direct correlation between poverty and educational level. At this point, it i s important to acknowledge the presence o f the informal sector where many o f these people are employed, which has as its major characteristics the lack o fjob security and social security benefits and unreliable payment, along with its easy entrance and the low educational levels o f employees. Among the working poor, those who were self-employed constituted 24.67 %, unpaid family laborers 49.86 YO,those working for daily wage 16.56 YO,salaried workers 8.61 YO,and employers 0.3 1 % (SPO, 2001a). Furthermore, increase in educational levels does not automatically bring employment and improved living conditions. Ina research inprogress conducted by the author in a gecekondu settlement in Ankara, people (rural-to-urban migrants; the majority was literate without primary school diplomas or primary school graduates) complained about unemployment, saying that they had made every effort to educate their children (many of the children were high-school graduates), yet they were unemployed. One woman said, "In the past, even primary school graduate could easily find jobs in government offices, but now our high-school graduate children cannot findjobs." In 1999, the unemployedyoungpopulationincities reached23 YO(SIS, 2001b). Furthermore,there is the gender dimension regardingeducation and poverty. Inthe illiterate poor group, the number ofwomen is twice that ofmen(SPO, 200al). On the other hand, inthe literate poor group who do not have primary school diplomas, there are more menthanwomen. This points to families' discriminatory behavior towards their daughters, providingthe chance to attend school selectively on the basis of gender. (For more detailed information on the subject, see pages 147-148, SPO, 2001a). Gender and regional disparities overlap to intensifywomen's disadvantagedpositions. The 'HumanDevelopment Index' for women in developed regions in Turkey is three times more than those in underdeveloped regions (the East and the Southeast). Women's access to education becomes more limited in Eastern regions, the women in the Southeast being the most disadvantaged. InEastem Anatolia 42 YOo f the women, and inSoutheastem Anatolia 55 % o f the women are illiterate, and in Simak this number i s up to 80 % (Sevindi, 1998). Traditional patriarchal values that disapprovewomen's participation inthe public realm play an important role inpreventinggirls from attending schools. This can be overcome bothby changingfamily's attitudes towards girls' education andby forming necessary institutions to encourage it, such as boarding schools for girls inthe region and providingtransportation to adjacent schools (Konta?, 1998). Recently, 7.83 % of the population who are at the age of schooling (age 6 and above) has moved into the poverty category (Erdogan, 1997). This numberrises upto 23.33 % when poverty is defined interms of basic needs instead of minimum food consumption (minimumcalorie neededfor biological survival). Whenpoverty increases, childrenare taken out o f s chool and placed in jobs (mostly inprecarious jobs in very disadvantaged positions). Although children's employment below the age o f 15 i s prohibited in 61 intemational agreements, in Turkey the economically active group i s defined as those between the ages 12-65, thus accepting the employment o f children above the age o f 11. Between 1980-1990, the number o f children ages 12-14 increased by 61 YOin Istanbul (it was about 50,000 in 1980 androse up to 80,000 in 1990), and those ages 15-19 increasedby 57 % (Sonmez, 1996). They were mostly employed in the manufacturing sector (Sonmez, 1996). In a research conducted in Mamak (Ankara) in 1990 (Demir, 1990), 35 % of the families hadtheir s ons and 9 % o f the families hadtheir daughters working outside the home. Inearlier studies (e.g. Senyapili, 1978; Sencer, 1979), the working children in rural migrant/gecekondu families were foundto bemuchless. InUrfawhere poor families work as seasonal agricultural laborers, children's school attendancedrops to 2/3 during the harvest season, and as a school principal in a gecekondu district stated, 1/3 o f the students work outside school hours, and this numberhas been increasing rapidly (Can, 2001). Ina study conducted in Izmir's residential areas where the very poor are clustered (namely, inner city areas) (Sonmez, 2001), it was found out that children started to participate inthe labor force when they were about 12. In the recent years, 14 % of the 12-19 age group became involvedinmoney-earning activities. And the younger the child was, the higher was the probability that he would lack social security. These working children were usually concentrated insmall-scale manufacturing. Moreover, when mothers start working outside the home because o f economic reasons, older children take care of their younger siblings and cany out housework. This was the case in a study carried out inthe inner city areas of Izmir where families were young with many children and where neighborly cooperation and help was absent (Sonmez, 2001). In brief, many children drop out of school to work outside the home (mostly boys) or to work inside the home (mostly girls). "Therefore povertyi s beingtransferred to the young generations'' (Sonmez, 2001; p.7). Although the level o f education in the society has been increasing (the 5-years- compulsory educationhas become 8 years), there are still significant inequalities interms of access to educational facilities between urban and rural areas, within urban areas, and between regions. InDiyarbalur, 86 % o f boys and 62 % o f girls attend primary school, and inurbanDiyarbakirthese numbers are 93 %and65 YO,andinruralDiyarbalur80 YOand49 %respectively (Kontq, 1998). And halfofthe women inDiyarbalurare illiterate, whereas it is 28 %nation-wide (Konta?, 1998). Only 42 % o fthe children between ages 7-18 attend school in Diyarbalur (Goktiirk, 1998). According to another study, only 52.4 % o f the children at the age of primary school education could attend school, and this number dropped to 34.3 % inthe case o fmiddle school (0 1998). The same report states Report, that 1,259 schools were closed inthe Region (CHP Report, 1998). The fact that many villages inthis region do not have schools and many schools are closed due to terror inthe region, and that the teaching personnel is inadequate in numbers, account for the larger illiterate group inthe InIstanbul, every year 60,000 childrenreach the school age due to population increase, and about 1500 class rooms have to be built to satisfy this demand. The annual gap inthe number o f classrooms is 1000 (Sonmez, 1996). InMersin, which receives mass migration from the Southeast, 23 YOo f the families fail to send their childrento school, and ofthis 18.5 YOwork inpaidjobs (GoktiirkandKaygalak, 2001). Not only the education of the poor, but also the poverty of teachers has become an importantissue inTurkey. Most of the teachers cannot live on their salaries (85.7 %), and 62 many carry out money-earning activities in addition to their teaching jobs in order to manage their lives economically (23 %), including street-vending(Egitim-Sen, 2001). 85.7 % o f teachers say that they do not eam enough money to support themselves, and 75 % cannot afford to buy daily newspapers. The inadequate pay o f teachers and the economic problems they have been facing have eroded the status o f teachers in society: 68 % o f respondents didnot want their childrento become teachers (Egitim-Sen, 2001a). The Elderly and Poverty As observed inthe cases o f women and children, we observe an increasing number o f older people (above the age o f 60) who start to engage in money-earning activities as a result o f intensifyingpoverty: InIstanbul, while there were 37,360 older people inthe labor marketin 1980, this number increasedto 58,000 in1990; and their numberdidnot decrease but remained stable inratio to the larger city population (Sonmez, 1996). This points to a very vulnerable population in the face of poverty. In a study conducted with persons receiving financial assistance fi-om the 'Fak-Fuk-Fon' Fund (Sallan-Giil), elderly persons expressedtheir need o f support from the State (food, fiel, cash); they were the ones among respondentswho hadthe worst living conditions. Poverty Eradication Policies The State Planning Organization has developed an action program to attack poverty (SPO 2001a). In the report, the target groups included the employed without adequate income; families which face poverty as feudal relations dissolve (mainly in the Southeast); women, the elderly, the disabled and children, including single women (divorced, widow); the exploited women inthe informal sector; earthquake victims; small-scale entrepreneurs; as well as rural and urban families living inpoverty. The report suggests concrete solutions to the problemo fpoverty. They include legalreforms (e.g. changes inthe Civil Code which would make it possible for women to keep some family assets after divorce; and changes in the social security laws to include those working in the informal sector); reforms in state policies and institutions ( e g in SHCEK-- Institute o f Social Services and Children's Protection, KSSGM-- Directorate General on the Status and Problems o f Women, YOK-- Higher Education Council); opening branches o f state welfare institutions in the East and Southeast; opening shelters for women both inurban and rural areas; prohibiting children's employment below the age 15, increasing the number o f vocational schools and providing credits to students in vocational schools; providing credits to small-scale entrepreneurs; supporting agricultural cooperatives, preventing unregistered labor in the rural sector, providing necessary financial and technical support to increase productivity, and introducing modem cultivation techniques (pilot projects should be carried out within the GAP Project in the Southeast); abolishing the requirement that in order to have insurance in the rural sector, one should be a family head, so that many women who work as unpaid family laborers can be insured; integrating the inactive housing stock in the city centers into the housing market addressing to lower-income groups; making birth control techniques more available to families; as well as providing direct financial aid to families. Here it i s crucial 63 that the money is to be paid to the woman in the family, and not to the family head (SPO, 2001a). Inaddition to the policies proposed to eradicate poverty, it is necessaryto mention the concrete actions o f the State in this direction. For example, the 'Social Assistance and Solidarity Fund' ('Sosyal YardimlaSma ve DayaniJma Tepik Fonu' known as the 'Fak-Fuk- Font-- Fakir, Fukara Fonu) was founded in 1987, which defined as its major goal to eradicate poverty (Sallan-Giil, 2001). This i s a new approach to cope with poverty since historicallypoverty has beenhandledwithin traditional mechanisms, such as the community and the family. By 2000, 13 % o f the population received financial assistance from the Fund, and inthe last four years, this numberreached 22 YO(15 millionpersons). And with the February2001 crisis, this number is expected to be higher. However, there are several problems with the practices o f the Foundations (933 in number) hctioning inthis institutionalframework. First, although the Fundwas originally established t o address the poor, its financial resources have beenused for other reasons, especially to pay internal debts. Secondly, there are no objective criteria to determine who the target population will be, and the distribution o f resources fi-om the Fund has been arbitrary and highly politized. And finally, the Fund offers only temporary solutions to poverty, for example, by providing food and fuel once in a while," and lacks long-term programs to integrate the poor into the labor force by teaching them skills and professions. Onthe other hand, the increasing share o f education from the Fundinrecent years has been a positive development? an attempt to eradicate poverty by equipping the young generation with necessary tools to reach jobs with satisfactory pays and social security benefits. The Fundhas beenbeneficialinhandlingthe increasing expensesineducation due to the recent eight-year compulsory schooling, particularly in the East and Southeast (Sallan-Giil, 2001). Another positive development has been to spend some o f the Fund's resources on the East and Southeast (20 % of the money given from the Fundwent to the people inthese two regions). NOTES 1. See the last section for informationon the Fund. 2. Gecekondu i s the name given to squatter houses in Turkey. Its literal translation means 'landed over night.' 3. See Erman (1996, 1997b) and Ayata (1989) for the preferences o f rural migrants regardingtheir gecekonddapartment preferences. 4. In another research conducted in four gecekondu areas of Ankara in which 300 households were interviewed (METU, 2000), it was found out that 96.33 YOowned refrigerators, 96.3 % owned color television sets, 90.33 % owned telephones, 67.3 % owned hlly automatic washing machines, 79 % owned vacuum cleaners, 10.67 % owned mobile phones, 10 % owneddishwashers, and 16 % ownedcars. 64 5. For example, in an empirical study (METU, 2000) conducted in Ankara's gecekondu areas, 51.8 % o f the respondents said that their present economic s ituations were worse whencomparedto thepast. 6. The researchwas conducted inMamak, Kegioren and Balgat (Ankara). The husbandsof 46 respondentswere employedinthe informal sector; 48 were migrants, and 40 were from the East and Southeast; their average age was 35; and except one, they were all mothers. 7. See Egitim-Sen(2001b) for information on the geographical distribution of classrooms, teachers and studentsinTurkey. 8. Between 1991-1994, the items provided to the poor from the Fund ranked as follows: food, health items, cashmoney, fuel, andclothing (Sallan-Giil, 2001). 9. 95 % o f those below the poverty line consists o f people with low educational levels (primary-school graduatesand illiterate) (SPO, 200lb). 65 REFERENCES Akcar, Sengiil. 1998. 'Kadin ve yoksulluk' (Women and poverty). In 'YoksullugU Onleme Stratejileri' (Poverty PreventionStrategies) (pp.197-202). Istanbul:TESEV. Ayata, Ay$e. 1997. The emergence of identity politics in Turkey. New Perspectives on Turkey, 17:59-73. Ayata, Sencer. 'Toplumsal qevre olarak gecekondu ve apartman' (Gecekondu housing and apartmenthousing as social environments). Toplum veBilim, 46/47: 101-127. Bugra, Ays. 2001. 'Ekonomik kriz kargisindaTurkiye'nin gelenekselrefahrejimi' (Turkey's traditional welfare regime inrelationto economic crisis). Toplum veBilim, 89: 22-30. Can, Kemal. 2001. 'Sanliurfa'dayoksulluk manzaralari' (Portrayals of poverty in Sanliurfa). Toplum veBilim, 89: 205-217. CHP Report. 1998. 'Do@ ve Giineydogu'dazorunlu go$' (The forced migration inthe East andSoutheast). In0.Baydar (ed.), 75 YzldaKoylerden Zehirlere (From Villages to Cities in 75 Years) (pp.333-341). Istanbul:Tarih Vakfi. Cinar, Mine. 1994. Unskilled urban migrant women and disguised employment: Home workingwomen inIstanbul, Turkey. WorldDevelopment, 22(3): 369-380. Dansuk, Ercan. 1997. 'Turkiye'de Yoksullu@n Olqiilmesi ve Sosyo-Ekonomik Yapzlarla Olqiilmesi' (Measurement of Poverty in Relation to Socio-Economic Structures in Turkey). ExpertnessThesis. Ankara: StatePlanning Institute, PublicationNo: DPT 2472. Davis, Mike. 1992. City of Quartz: Excavating the Future in Los Angeles. New York: Vintage. Demir, Erol. 1993. 'Ekonomi politikalar ve kentemekgi aileleri' (Economic policies and urbanworker families). Birikim,48: 68-77. Dumanli, Recep. 1996. 'Yoksulluk ve Tiirkiye'deki Boyutlarz' (Poverty and Its Amplitude in Turkey). ExpertnessThesis. Ankara: State Planning Institute, PublicationNo: DPT 2449. Ecevit, Yildiz, Erman, Tahire, Kalaycioglu, Sibel and Rittersberger-Tilig, Helga. Migrant women's participation in the labor force in urban Turkey. Report prepared for the World Bank,WashingtonD.C. Egitim-Sen. 2001a. http://www.egitimsen,org.tr ['5 Ekim 2001: Ogretmenler Yoksulluk Kiskacinda' (October 5,200 1:Teachers Are Caught inPoverty)]. -------. 2001b. http://www,egitimsen,org.tr ['Egitimde Aci Tablo: Rakamlarla Tiirkiye Gergegi' (The Worrying Picture inEducation:Turkey by Numbers)]. 66 Erder, Sema. 1995. 'Yeni kentliler ve kentin yeni yoksullm' (New urbanites and the new urbanpoor). ToplumveBilim, 66: 106-119. -------. 1996.'Istanbul'u Bir Kent Kondu: Umraniye' (A City LandedinIstanbul:Umraniye). Istanbul: Iletigim. -------. 1997.'KentselGerilim' (UrbanTension). Ankara: Um. -------. 1998a. 'Koysiiz "koylii" go@' (Migration of villagers without villages). GBriig, March: 24-26. -------. 1998b. 'Gogve kentlegme sorunlari' (Migration and the problemsof urbanization). In 'Yoksullu~uOnleme Stratejileri' (Poverty Prevention Strategies) (pp.233-236). Istanbul: TESEV. Erdogan, Guzin. 1997. 'Tiirkiye'de Y oksulluk: Boyutu v e Profili' (Poverty inTurkey: Its Amplitude andProfile). Ankara: StatePlanningInstitute Publication. Erman, Tahire. 1996. Women and the housing environment: The experiences of Turkish migrantwomen insquatter (gecekondu) andapartment housing.Environment and Behavior, 28(6): 764-798. -------. 1997a. The meaning of city living for rural migrant women and their role in migration: The case ofTurkey. Women'sStudiesInternational Forum, 20(2): 263-273. -------. 1997b. Squatter (gecekondu) housing versus apartment housing: Turkish rural-to- urbanmigrantresidents' perspectives. HabitatInternational, 28(1): 91-105. -------, 2001a. Rural migrants and patriarchy in Turkish cities. International Journal of Urban and RegionalResearch,25(1): 118-133. -------. 2001b. The politics of gecekondu (squatter) studies in Turkey: The changing representationsofrural migrants inthe academic discourse. Urban Studies,38(7): 983-1002. Goktiirk, Atilla. 1998. 'Aragtirma bulgulari ve giiniimiizdeki Diyarbalur'(Researchdata and Diyarbakir today). In 'Bolgeqci Zorunlu Goqten Kaynaklanan Toplumsal Sorunlarzn Diyarbalnr Kenti Olqeg'indeAragtzrzlmasz' (The Investigation of Social Problems Arising from Intra-Regional Forced Migration in the Context o f the City Diyarbakir) (pp.15-64). Ankara: TMMOB. Goktiirk, Atilla and Kaygalak, Sevilay. 2001. 'Gog, sosyal refah ve kentlegme: Mersin ornegi' (Migration, social welfare and urbanization: The case of Mersin). In 'Bolgesel Kalkinma Siirecinde SosyalHizmetler SempozyumuBildirileri' (Social Work inthe Process of Regional Development Conference Proceedings) (pp. 237-254). Ankara: Hacettepe University. 67 Guneg-Ayata, Ayge. 1990/91. 'Gecekondularda kimlik sorunu, dayanigma oriintiileri ve hemgehrilik'(The identity problem, solidarity networks and hemgehrilik). Toplum ve Bilim, 51/52: 89-101. Guveng, Murat and Igik, Oguz. 1996. 'Istanbul'u okumak: Statu-konut mulkiyeti farklilqmasina iligkin bir goziimleme denemesi' (Reading Istanbul: An essay on spatial differentiation with respect to employment status and housing tenure types). Toplum ve Bilim, 71: 6-60. Hacettepe University, Institute o f Population Studies. 1998. Population and Health Suwey. Ankara. Igik, Oguz and Pinarcioglu, Melih.2001. 'NobetlegeYoksulluld (Poverty inTurns). Istanbul: Iletigim. Jessop, B. 1994. The transition to post-fordism and the Schumpeterian workfare state. In R.Bwrows and B.Loader (eds.), Toward a Post-Fordist WelfareState? London: Routledge. Kagitqibagi, Cigdem. 1998. 'Tiirkiye'de aile ve gocuk egitimi' (Family and children's education in Turkey). In 'Yoksullugu Onleme Stratejileri' (Poverty PreventionStrategies) (pp.207-210). Istanbul: TESEV. Kamu-Sen. 2001. http://www.kamusen.org.tr/gecimeylul2OOl.htm. Karaduman, T.A. 2000. UnpublishedWorking Paper. State Institute o f Statistics. Kasnakoglu, Zehra. 1998. 'Tiirkiye'de gelir dagilimi' (The income distribution inTurkey). In 'Yoksullugu Onleme Stratejileri' (Poverty Prevention Strategies) (pp.123-127). Istanbul: TESEV. Kayacan, Gulay. 1998. 'IOziltepe'den Istanbul'a: Kemerburgaz qadir koyii' (From IOziltepe to Istanbul:The Kemerburgaz tent village). In0.Baydar (ed.), 75 YzldaKdylerden Sehirlere (From Villages to Cities in75 Years) (pp.353-359). Istanbul:TarihVakfi. Keyder, Caglar. 2000. 'Enformel konut piyasasindan kiiresel konut piyasasina' (From the informal housing market to the global housingmarket). InC.Keyder (ed.), 'Istanbul: Kuresel ile YerelArasznda' (Istanbul:Between the Local and Global) (pp. 171-191). Istanbul:Metis. Kog, Ismet. 1997. Female-headedhouseholds inTurkey andtheir income and consumption patterns. Paperpresented at the ThirdNational Demography Conference, Ankara: Hacettepe University,December2-5. Kontag, Mehmet. 'Turkiye'de aile ve gocuk egitimi' (Family and children's education in Turkey). In 'Yoksullugu Onleme Stratejileri' (Poverty Prevention Strategies) (pp.203-206). Istanbul: TESEV. Kiimbetoglu, Bellus. 1992. Women's informal sector contribution to the household survival inurbanTurkey. Unpublisheddissertation. Istanbul:MarmaraUniversity. 68 MacPherson, Steward and Silbum, Richard. 1998. The meaning and measurement of poverty. InJ.Dixon and D.Macarow (eds.), Poverty: A Persistent Global Reality (pp. 1-19). London: Routledge. METU. 2000. 'Kentsel Yoksulluk ve Gelipne Stratejileri: Ankara Ornegi' (Urban Poverty and Survival Strategies: The Case ofAnkara). Ankara: METUPublications. Mingione, Enzo. 1991. Fragmented Societies: A Sociology of Economic Life Beyond the Market. Oxford: Blackwell. -------. 1996. Urban poverty in the advanced industrial world: Concepts, analysis and debates. In E.Mingione (ed.), Urban Poverty and the Underclass: A Reader (pp.3-40). Oxford: Blackwell. Morqol, Goktug and Gitmez, Ali S. 1995.A typology of the urbanpoor inTurkey. Journal of UrbanAffairs, 17(4): 413-422. Musterd, Sako. 1994.A rising Europeanunderclass.Built Environment.20: 185-191. Myrdal, Gunnar. 1963. The Challenge ofAfluence. NewYork: PantheonBooks. O'Loughlin, John and Friedrichs, Jurgen. 1996. (eds.) Social Polarization in Post-Industrial Metropolises. NewYork: Walter de Gruyter. Oncu, Ayge. 1988. The politics o f the urban land market in Turkey: 1950-1980. InternationalJournal of Urban and Regional Research, 12(1): 38-63. -------. 1997. The myth of the 'ideal home' travels across borders to Istanbul.InA. Oncu and P. Weyland (eds.), Space, Culture and Power: New Identities in Globalizing Cities (pp. 56- 72). LondonandNew Jersey: Zed. Ozgen, Nege H.2001. 'Kentte yeni yoksulluk ve cop insanlari' (The new urbanpoverty and 'scavengers'). Toplum veBilim, 89: 88-101. Sallan-Giil, Songiil. 2001. 'Turkiye'de yoksulluk ve devletin yoksullukla mucadelesinin sosyolojik degerlendirilmesi: Fak-Fuk-Fon' (Poverty in Turkey and the sociological evaluation ofthe State's combat with poverty). Thesis for the Associate Professorship. Parkin, F. 2000. Marxism and class theory: A bourgeois critique. Sociological Research Online, 1,1, http://www. socresonline.org.uWl/l/3.html. Pinarcioglu, Melihand Ipk, O@z. 2001. '1980 sonrasi donemde kent yoksullari arasinda guce dayali ag iligkileri: Sultanbeyli ornegi' (The power-based networks among the urban poor inthe post 1980period: The case of Sultanbeyli).Toplum veBilim, 89: 31-61. Sassen, Saskia. 1991. The Global City: New York, London, Tokyo. Princeton, NJ: Princeton University Press. 69 Sencer, Yakup. 1979. 'Tiirkiye'de Kentlegme' (Urbanization inTurkey). Ankara: Publication ofMinistryof Cultural Affairs. Sevindi, Newal. 1998. 'Kadin ve yoksulluk' (Women and poverty). In 'Yoksullu@ Onleme Stratejileri' (Poverty Prevention Strategies) (pp.193-196). Istanbul:TESEV. Sonmez, Ipek. 2001. Spatially included-Socially excluded: The inner areas o f metropolitan cities of developing countries: The Case of Izmir, Turkey. Paper presented at the 'Social Inequality, Redistributive Justice and the City' Conference organized by the Intemational Sociological Association, Amsterdam, Netherlands, June 15-17. Sonmez, Mustafa. 1996. 'Istanbul'unIki Yiizii: 1980'den2000'e DegiJim' (TheTwo Faces of Istanbul: Transformationfrom 1980to 2000). Ankara: Arkadag. -------. 1998. 'Bolgesel Egitsizlik: Tiirkiye'de Do@-Bat1 Uqurumu' (Uneven Regional Development: The Gap Between the East and the West in Turkey). Istanbul: Alan Publications. State Institute of Statistics (SIS). 1990. 1987 Household Survey. Ankara: SIS Printing Division. -------. 1994.HouseholdIncomeDistributionSurvey. Ankara: SIS Printing Division. -------. 1996. Statistical Yearbook of Turkey. Publication no.1985. Ankara: SIS Printing Division. -------. 2001a. The results of the gross national product in 1999. The News Bulletin (23.05.2001). -------,2001b. TheHouseholdLabor ForceSurvey.Ankara: SIS Printing Division. State Planning Organization (SPO). 2001a. 'Gelir Dag'zlzmmzn Iyilegtirilmesi ve Yolcsullukla Miicadele' (The Amelioration in Income Distribution and Combat Against Poverty). Ankara: SPO Publication. -------. 2001b. '8. Bes Yillik Kallunma Plani 2001-2005, 2001 Yili Programi' (The Eighth Five-Year DevelopmentPlan2001-2005).Ankara: SPO Publication. $en, Seher. Spatial aspect of survival strategies of the urban poor. Unpublished Master's thesis. Ankara: Middle EastTechnical University. $enol-Cantek, FundaL.2001. 'Fakirkaneler:Yoksullugun 'lev hali"' (Poor households: The domestic picture ofpoverty). Toplum veBilim, 89: 102-131. 70 fjenyapili, Tansi. 1978. 'Biitiinlegmemig Kentli Nz&s Sorunu' (The Unintegrated Urban PopulationProblem). Ankara: METUPublications. ----___1998. 'Cumhuriyet'in 75. yili, gecekondunun 50. yili' (The 75th year of the Republic, . the 50th year of the gecekondu). InY.Sey (ed.), '75 YzldaDegigen Kent ve Mimarlill (The Changing City and Architecture in the 75th year of the Republic) (pp.301-316). Istanbul: Tarih Vakfi. TMMOB, 1998. 'Zorunlu go@n Diyarbakir orneginde aragtirilmasi' (Forced migration and Diyarbakir as an example). In0.Baydar (ed.), 75 YzldaKoylerden Sehirlere (From Villages to Cities in75 Years) (pp.342-352). Istanbul:Tarih Vakfi. Tokatli, Nebahat and Erkip, Feyzan. 1998. Foreign investment in producer services. Third WorldPlanning Review,20(1): 87-106. Towsend, Peter. 1993. Underclass and overclass: The widening gulf between social classes inthe 1980s.InG.Payne andM.Cross(eds.), Sociologyin Action. Basingstoke:MacMillan. TUSIAD. 1997. 'Turk Sosyal Giivenlik Sisteminde Yeniden Yapzlanma' (Restructuring in Turkish Social Security System). Istanbul. Tuziin, Sezgin. 2001. 'Kentsel Tiirkiye'de gelir dagilimi iizerine Veri-Sgt 2000 aragtirmasi bulgulari' (The data of the Veri-Sgt 2000 survey on the income distribution in urban Turkey). Unpublishedmanuscript.Ankara: Veri ResearchCorporation. Upn Haber, 2001. "'Giineydogu'da kadin: Sorunlar, kwuluglar, igbirligi ve iletigim" Toplantisi' (The Meeting on 'Women inthe Southeast: Problems, organizations, cooperation andcommunication'). 12:6-7. White, Jenny B. 1994. Money Makes Us Relatives: Women's Labor in Urban Turkey. Austin:University ofTexasPress. World Bank. 2001. World Development Report 2000/2001. New York: Oxford University Press. 71 4. "Measuring Poverty and Inequality in Turkey" pusufZiya Ozcan) 1. Introduction 1.1. DefiningPoverty Poverty can be understood and defined either as a concept deriving from low income o r as a c onsequence o f insufficient disposable income as reflected inpeople's consumption patterns or general well-being. What i s important i s the basis on which poverty i s evaluated. The concept o f `indirect poverty' i s measured according to the disposable income o f households, whereas the concept o f `direct poverty' measures how low income influences consumption, lifestyle or the accumulation o f other factors that cause personal deprivation. Three different approaches can be used in defining poverty: basic needs, relative poverty and multiple deprivation factors. Depending on the approach, poverty can be definedinthree different ways: 1) The poverty line is defined according to the minimum income required to satisfy a person's basic needs. This basic needs approach i s generally based on the cost o f a basket o f everyday consumer commodities. 2) The poverty line i s defined individually by the society concerned according to its income distribution. The classic method i s to draw the line at 50 per cent o f median income or consumption. 3) The third model defines as poor those who, in addition to having low incomes, suffer severe deprivation in other sub-factors o f material well-being. Typical such additional factors are unemployment, poor housing, poor health and inadequate education (Orshansky 1965b). In the case o f deprivation, the problem is often the lack o f key factors necessary for well-being (Eurostat 2000). 1.2. Measuring Poverty Traditionally, there have been two approaches to measure the poverty (Saunders 1998). The first i s absolute poverty and the second is relative poverty. Recently, a new approach i s added which i s termed as subjective poverty. In absolute poverty approach, some criteria relating to income or consumption i s set and households (or individuals) that fall under this criterion are classified as poor -- inability to meet very basic needs. In relative poverty, households are ordered by their income or consumption, and households falling a certain point are identified as poor- distance from the community norm. In subjective poverty (or felt poverty), the criterion i s obtained from the individuals by asking at what income level they meet all their needs. (Gordon et a1 2000, Bradshaw 2001, Forster 1994) Any attempt to measure the adequacy o f resources or the prevalence o f poverty requires that a poverty line (or threshold) be established against which the living standards o f an individual, family or household can be measured (Bradshaw and Finch 2001). That is, whichever approach i s used, if the aim i s to describe persons living in 72 poverty, we should define who i s poor and who is not. Defining a poverty line i s necessary ifwe are going to analyze not only the incidence o f poverty, but also the depth o f poverty (how poor are the poor?). Such an understanding necessitates that the analysis must, o fcourse, bedone for different family types, persons o fdifferent ages, andthe like. Drawing any poverty line i s arbitrary to some degree, even if there i s conceptual agreement on what poverty means. But in this case, even conceptual agreement i s lacking. Popular and expert opinions apparently clash on whether and to what extent poverty should be defined in absolute terms - inability to meet very basic needs - or in relative terms - distance from the community norm. In fact, upon closer examination, virtually all measures o f low income are relative (except the most extreme), and they differ mainly intheir underlying assumptions about how wide a gap should exist between "the poor" and fellow citizens within the same society. This gap can be defined interms o f income or interms of consumption. One exampleo f an absolute poverty line i s the World Bank's extreme poverty line o f $1U.S. per day, which i s applied to developing countries as a proxy for a minimum survival budget. Although it may be instrumental measuring and comparing poverty in developing countries, it i s basically irrelevant to understanding poverty in advanced, industrialized countries, since virtually all their citizens have incomes above that minimal level. So-called "absolute" poverty lines do, in fact, embody community norms and standards, at least to a limited degree. Indeed, it i s acknowledged that "the basic needs poverty line i s not absolute.. .to be meaningful, a poverty line has to be connected to the society inwhich people live." (Measuring Poverty in Canada, 2001) Most social policy advocates have strongly defended a relative definition o f poverty, arguing that to be poor is to be distant from the mainstream o f society and to be excluded from the resources, opportunities and sources o f subjective and objective well- being which are readily available to others. Adam Smith himself closely embraced a relative definition o fpoverty, arguing that to be poor was to have to go without what was needed to be a "creditable" member o f society. H e noted that, in his day and age, a man needed a linen shirt if he was to appear in public "without shame." What defines being poor i s not just being without a shirt (any shirt), but being without the means to be minimally respected ina particular society at a particular time. Similarly, the Organization for Economic Co-operation and Development (OECD) has argued that absolute poverty lines have little meaning in advanced industrialized societies, and that poverty should be seen conceptually not as deprivation o f very basic needs, but as exclusion from the standards o f living broadly available to others inthe same society. As they note, '!inorder to participate fully inthe social life o f a c ommunity, individuals may need a 1eve1o f resources that is n o t t 00 inferior t o the norm o f that community." They add that, "from a normative perspective, it may be considered unfair for members o f a community to benefit unequally from a general increase inprosperity."(OECD Employment Outlook, June 2001, p. 41) 73 It is very difficult to arguethat relative well-being does not matter to people. How many o f us would prefer to have an income of 300,000 TL/month in a society where the average was 500,000 TL/month, compared to having an income o f 290,000 TWmonth in a society where the average was 250,000 TL/month? It i s simply untrue to say that people are indifferent to where they sit inthe total income distribution compared to others, even provided their basic needs are met. Research also demonstrates that many key social outcomes such as health and literacy are closely linked to the relative incomes o f people, rather than the extent to which they suffer absolute deprivation. The Nobel-prize-winning economist, Amartya Sen, uses as a telling example the fact that blacks in the United States have much lower life expectancy than do citizens of the state o f Kerala in India, despite having much higher absolute incomes (1983). In advanced industrialized countries, the link between health outcomes and average GDP per person breaks down beyond a threshold level and it is most closely linked to income gaps. The central point is that virtually all measures of poverty are relative. Whether we define poverty in terms o f a given level o f income compared to the average (an income line) or in terms o f the cost o f a basket of goods and services compared to the average, we still have to settle on how great a distance we want to exist between ``the poor" and others who live inthe same society. This decision i s explicit inrelative income lines such as the Low Income Measure (LIM), but more implicit inthe c onsumption- based approach. All measures o f poverty, whether they are based on income or a basket o f goods and services, are also arbitrary, at least to some degree. It i s really a matter o f values how great a distance we are prepared to accept between the "poor" and the rest o f society. Most social welfare advocates support a relatively generous poverty line, because we recognize needs as social as well as physical. To be poor is to experience a significant degree o f exclusion from the wider society, and not just to be deprived o f very basic needs. It is unfortunate that the development literature generally takes either point of view but rarely both. It i s the case that relative poverty i s often found to be related to or synonymous o f `inequality' and therefore of no primary relevance for `actual', that is `absolute' poverty issues and policy. Infact, there i s no reason why both concepts should not be considered simultaneously'. In the past, too much priority has been given to absolute poverty over relative deprivation, or equivalently social exclusion. It i s important to reestablish some parity betweenboth points o f view inall cases where this i s justified. (For such an effort refer to Atkinson and Bourguignon 1999) "There are countries inAfrica and Asia where the dominant problem i s absolute poverty and where social exclusion or relative deprivation i s presently o f lesser importance. But there are countries where both physical and social poverty are of concern, as in many parts o f Latin America." (Bourgignon 1999) 74 2. Sample and Data Two sets o f data are collected in this research. The first set comprises the information collected through a survey and the second set includes 68 in-dept interviews conducted with 38 households indifferent parts o f the country. 2.1. Survey The sample used inthis research represents Turkey and is drawnby State Institute o f Statistics (SIS) in2001. It i s a multi-stage stratified cluster sample which allows comparisons byregion and rural-urbanplaces. The unit o f analysis inthis research i s household. The sampling design requires stratifying the country into 7 regions at first stage. At the second stage, residential units ineach region is divided into population strata such as places with 0-2000 population, 2000-5000, 5000-10000, 10000-20000, 20000-50000, 50000-100000, 100000-150000 and places with more than 150000 population. Clusters are formed by combining 30 households within each population strata. At final stage, clusters are selected within each population stratum independently by using random selection technique. Clusters obtained this way are proportionate to their size in the population. The initial size o f the sample selected with this technique was 8000 households. Giventhe budget o f the research, the sample size reduced to 4000 by way o f resampling which was done the following way. Clusters are isolated from their respective stratum information and assigned random numbers. 50 percent of them were selected randomly with the help o f computer. Later, they were identified by combining with their respective stratum information. Considering the possible loses due to non-response, the sample size was determined as 4300. Each household has a substitute in case no one was found at home at the first address. Addresses from 1997 population count administrative division were used inthe sample. The following table provides information to compare sample and population distributions o fregions, rural urban places. Table 1. Sample and PopulationDistributionofRegionsand RuralUrbanPlaces,% Sample Population (2000') Urban' 83.1 70.6 Rural 16.93 29.4 Mediterranean 12.1 12,86 Aegean 12.7 13,44 Marmara 30.7 25,74 Southeast 7.9 9,75 East 7.3 8,93 Central 17.5 16,83 Blacksea 11.7 12,48 ''Estimationdefined by the end ofthe year. Urban is asplaces 20.000 and morepopulation Includes villages only Source: SLY, SPO 75 There i s fair match between the sample and the population distributions. Seemingly serious difference between urban and rural distributions stems from the fact that rural inthe sample includes only v illages whereas rural is defined as places with population less than 20000 in SIS statistics. When the households are reclassified according to the SIS criterion, the sample portion o f rural households becomes 18 percent. It should be noted that this exercise o f reclassification carries certain amount o f error stemming from the difference between population figures for 2000 census and 1997 population count. The sample includes 62 provinces. There are 7 provinces inMediterranean, 8 in Aegean, 10 inMarmara, 7 inSoutheast, 8 inEast, 11in Central and 12inBlacksea. Data collection instrument: Data collection instrument was questionnaire. There were seven modules inthe questionnaire each o fwhich probed a different issue. These modules are: A. HouseholdRoster which has two sub modules. B. Characteristics ofhouseholdwhichhas two submodules. C. Consumptionwhich has two sub modules. D. Income and employmentwhich hastwo sub modules. E. Savings F. Access to services which has four sub modules. G. Perceptiono fwealth andincome The questionnaire was pre-testedby the research team in five different provinces and necessary corrections were made before giving it to the contracted company. Detailed nature o f the information sought made the questionnaire longer, administration o f which took, on the average, one and a half hour in the field. A copy o f the questionnaire i s given inAppendix 1. Fieldwork: Fieldwork is contracted to a company, which specializes inlarge scale data collection. Region supervisors o f the company were trained before the fieldwork started. In fact, two training sessions were organized in Ankara. A manual to be used by the interviewers w hich was also used duringtraining o f supervisors w as prepared and givento all interviewers. Interviews took place between 18 July and 25 September, 2001. Face to face technique was used to interview respondents. Responses were recorded by using paper and pencil technique. Initially five teams took part indata collection process. Two more teams were added later to expedite the data collection process. Non-response: The total number o f questionnaires completed in the field was 4307. For 80 percent o f households, someone was found for interview inthe first address and substitutes were used for only 20 percent households. Non-response rate was 76 negligible despite the fact that no incentives provided for the respondents. However, 296 cases were eliminated during data editing and cleaning phase. Most o f the eliminated questionnaires had the problem o f too many missing answers and inconsistent information given by the respondents. After cleaning and editing, 4119 questionnaires remained for the analysis, which i s little over than the initially targeted sample size o f 4000. Limitations o f the data: Household surveys usually fail to pick up individuals who do not have residence, which also create serious measurement problems. However, estimates o f the number o f homeless suggest that this would not make a substantial difference to the results in terms o f the overall distribution although it might somewhat affect measures o f poverty. Despite daily observations that indicate increased number o f the poor, which i s apparent in the number o f street beggars, homeless families and the rampant incidence o f theft, they are under- represented in the sample since they do not have fixed addresses and records to take into account in sample selection. Infact, similar problems are valid for the other end o f the distribution. Richpeople often are hard to find and interview. This studyhadbothlimitations to certain degree, the extent o fwhich is difficult to determine. Moreover, although few in numbers, some difficulty has also been experienced in locating some o f the respondents at the addresses provided by the State Institute o f Statistics. Despite the short time span between the last population count (1997) and the fieldwork (2001), there have been changes in addresses o f some respondents. Some families move out due to construction o f new housing in their neighborhood while somejust changed their residence for various reasons. It is also the case that there were problems with the reporting o f some important variables such as income. In general, there is a certain extend o f under-reporting o f incomes which i s commonly encountered in household surveys. Under-reporting o f capital and property incomes i s often significant. The degree to which this is the case is difficult to judge and likely that it i s concentrated in certain groups. It is observed that some o f the families under the poverty line did not report their consumption correctly which created a false picture o f positive saving for the poor household. Although it i s shown that under-reporting o f household incomes i s small (Burniaux et al. 1998) there i s no study reporting the impact o f under-reporting for consumption on the calculation o f poverty lines. In the light o f this information, household income was used both in determination o fthe poverty lines and measuring inequalities inthis research. 2.2. In-dept Interviews Total o f 122 interviewswere conducted between October 2001 and January 2002 by two experienced interviewers. It took 4 months to complete the in-depth interviews with the poor families. Eight o f the interviews were done with widowed women who were household heads. 77 Selection o f cities as interview sites has been done carefully to reflect the social- economic characteristics o f the country. Attention i s paid to include poorest households who live in various disadvantaged environments such as areas subjected to long lasting terrorism, places where out migration is high, places attract poor people along with seasonal/temporary workers and earthquake-ridden areas. Households from both urban and rural areas are included inthe selection. Inthe lightoftheseconsiderations, thefollowing cities are selectedfor interviews; City No. o fhouseholds Place o f interview Ankara 6 urban Corum 6 rural Kutahya 6 rura1 Siirt 5 rural Malatya 6 urban Adana 6 urban Giresun 5 urban Istanbul 5 urban Mersin 11 urban Bingo1 11 urban U8ak 11 urban Sanliurfa 11 urban Aksaray 11 urban Samsun 11 urban Bursa 11 urban Inthe field, the selection of the households is done with the help of The Prime Ministry's Social Solidarity Foundation(Bagbakanlik Sosyal Yardimlagma ve Dayanigma Vakfi). According to the prevailing procedure, poor families apply to this foundation and their applications are kept infiles. Upon investigation o f the Foundation personnel, some are granted help while some are denied. During the selection o f households, these files were searched to find the most suitable families to interview. In case there were doubts about the household's well-being and validity o f the information provided, muhtars and residents o f the respective communities were resorted for additional information. One o f the interviewers who was female interviewed the wives and probed the issues relating to the management o f house, food and non-food consumptions in the house, healthproblems inthe family, education o f t he children. Inaddition, questions were asked to understand how they have been reacting to economic changes in the last decade and coping strategies they have developed to ease the blows o f the crises. The second interviewer who was male conducted interviews with the husbands andprobedthe issues concerning employment o f the household membersand household income Interview time varied between 60-90 minutes. Wife and husband were interviewedseparately preferably in separate rooms. Ifthe house has only one room, the husband was interviewed outside o f the house. 78 Every household selected responded to interview, which rendered the non- response rate to 0. 3. Poverty Line Multidimensionality o f poverty and various measures and indices developed to capture as many dimensions as possible created a situation where many measures are available to tab the same phenomenon (World Bank 2001; Senses 2001). Moreover, each measure employing a different rationale produces a poverty line which i s different than that o f others. Choosing one or few measures among the existingones becomes a difficult task for the researcher. Even a researcher, who employs all o f the measures ends with many poverty lines which are different from each other. There are mainly two reasons for s electing relative measures over the absolute ones. First, the qualitative research conducted to complement this survey suggested that use o f the absolute poverty i s not possible in case o f Turkey. The household defined as the poorest lost its meaning when even poorer household was found which, was the case repeatedly encountered in the qualitative research. Second, most absolute poverty measures impose an external criterion to be used in the calculation o f the poverty line. We feel that using external criterion in the case o f Turkey will basically reveal how Turkey stands vis a vis that criterion which does not do justice to mirror her structural characteristics. If, for example, Turkish families consume 40 percent o f their income on food on the average, cutoff should be 40, not 50 or 60 percent as suggested and used by some authors. This does not hamper comparative edge o f the research in any way. Just to the contrary, it helps to compare countries by taking their true mean food consumption as the base. Priority in selecting measures i s given to those criterion o f which was obtained from the data as opposed to ones that use external criteria such as 2450 calorie per day for each household member. The criteria used inthe measurement o f the poverty 1ine originated from the data itself. One other difficulty in case o f Turkey is the absence o f studies that use data that reflect the true degree o f poverty in the country. Almost all o f the previous studies on poverty have usedthe 1994 Income and Consumption data collected by the State Institute o f Statistics. It i s crucial to remember that Turkey had the most serious economic crises in 1994 and the 1994 data were collected aftermath o f that crisis. Ironically, the present data were collected right after Turkey experienced two big economic crises inNovember 2000 and February 2001. Neither o f the data can reflect the true nature o f the poverty in the country since both carry the imprints o f economic crises. Besides hiding true nature o f the poverty, the absence o f data collected at times o f no crisis wipes out the possibility o f comparison. In fact, during the analysis o f data, the research team felt helplessness at times, which stemmed from the lack o f solid ground to compare the results o f this study. The selection o f poverty and inequality measures i s also guided by the practical reason which can be stated as to choose ones that were employed in the previous poverty research on Turkey (Akder 1999; Dagdemir 1999; Dansuk 1996; Dumanli 1996; Uygur and Kasnakoglu 1998; Dumanli ve Bulutay 2000; Erdogan 1996, 2000; Ozer 2001; 79 Pamuk 2000; World Bank 2000; Sonmez 2001; DPT 2000) Two studies which seemed to have wider coverage in terms of measurements were taken as reference in particular. These are Erdogan(1996,2000) and Pamuk (2000) 3.1. TheLow IncomeMeasure (LIM) The first relative poverty line measure uses total household income and takes the halfmedian value as the poverty line. This measure will be termed as `The Low Income Measure' (LIM) Likeany other measure, LIMhas pros and cons. LIMis apurely relativepoverty measure which i s used inmany international comparisons. It explicitly defines low income as being much worse off than average, andit i s drawn at one-half the median income o f an equivalent household. Pros o f LIMcan be listed as 0 LIMis simpleto understand andcaptures theimportant inequalitydimensionof low income. 0 LIMfacilitates international comparisons andmakes clear how Turkey is doingin comparisonto other countries. needs - however they are defined - but it does tell us that they are significantly LIMdoesnottell us directly ifthe poor havesufficient income to meettheirbasic distant from the social norm interms o f income available to buy goods and services inthe market economy. Cons o f LIMare 0 Itis conceptually unclear what kindso fregionaldifferences shouldbe allowed for inLIMs. Thus itrequires calculationofmedian for different regions, provinces andthe like. Still, ahousehold identified as poor inone regionmaynot be poor in another region. 0 LIMtells usnothingdirectly about what a low-income householdcanbuy. 0 L I M s do not vary with the business cycle. Poverty using the LIMtends not to increase inrecessions, even though income falls for the working poor. That i s because unemployment reduces median incomes and notjust the incomes o f low- income households. The income unit i s the household, which i s defined as a group o f persons sharing a set o f common resources, not necessarily related by blood or marriage. Household disposable income i s defined as total market income (income from labor and capital), plus income transfers from government, monetary and equivalents o f non monetary assistance from persons and other institutions (from private charity and solidarity institutions and NGOs), less income taxes. Although household income i s self-reported, it i s checked by adding all monetary incomes generated by the family members and equivalents of all non-monetary income received by the household. Incase the latter was bigger than the first, the later was used as the total household income. It should be noted that choice o f income unit, households as opposed to individuals, is important since it 80 affects the level o f income inequality. An increase in the size of income unit lowers the degree of income dispersion. In order to calculate one-half of the median, data were broken down by 7-fold region variable and the half median was calculated for each region. The half median value was taken as the poverty line. Table 2 presents poverty line for regions, urban and rural places (will be called locations interchangeably hereafter) and the country along with other descriptive statistics such as mean, standard deviation, percent of the total income, total income, percent o fthe total households and the number o f households. Table 2 Total HHIncome(1000 TL) by Regions,UrbanRuralPlaces andPoverty Line REGION Descriptive Statistics Total HHincome Poverty line 1 Mediterranean Mean 409267.75 Median 300000.00 150000 Std. Deviation 494594.43 YOof Total Sum 13.4% Sum 201769000 % of Total N 12.4% N 493 2 Aegean Mean 320886.60 Median 250000.00 125000 Std. Deviation 278686.36 % of Total Sum 10.9% Sum 163973054 YOof Total N 12.8% N 511 3 Marmara Mean 458130.10 Median 350000.00 175000 Std. Deviation 503599.23 % ofTotal Sum 36.7% Sum 554337418 % ofTotal N 30.4% N 1210 4 South-East Mean 240989.90 Median 200000.00 100000 Std. Deviation 135156.45 YOof Total Sum 4.7% Sum 71574000 % of Total N 7.5% N 297 5 East Mean 320003.50 Median 250000.00 125000 Std. Deviation 337877.63 % of Total Sum 6.1% Sum 91521000 % of Total N 7.2% N 286 81 6 Central Mean 378682.70 Median 300000.00 150000 Std. Deviation 293431.89 % ofTotal Sum 17.9% Sum 270000764 % of Total N 17.9% N 713 7 Blacksea Mean 331908.90 Median 300000.00 150000 Std. Deviation 204401.02 % of Total Sum 10.4% Sum 156661000 % ofTotalN 11.9% N 472 Urban Mean 397905.15 Median 300000.00 150000 Std. Deviation 407847.81 YOof Total Sum 87.2% Sum 1316270236 YOof Total N 83.1% N 3308 Rural Mean 287189.91 Median 200000.00 100000 Std. Deviation 261472.29 % of Total Sum 12.8% Sum 193566000 YOof TotalN 16.9% N 674 Turkey Mean 379165.30 Median 300000.00 150000 Std. Deviation 389178.13 YOof Total Sum 100.0% Sum 1509836236 % of Total N 100.0% N 3982 Ifthe regions arecompared interms ofmeanincome, Marmararanks first which i s followed by Mediterranean, Central, Blacksea, East, Aegean and Southeast. Despite considerable differences inthe mean incomes, there i s less variation intheir medians. 82 500000 1 400000 300000 200000 MEAN 100000 MEDlAN REGION Figure1: Mean andMedianIncomeby Region Taking the halfmedian value as the cutoffpoint, households were grouped into 5 categories as 1. households that hadincome less than halfo fthe medianincome 2. householdthat had income more than the half-median but less than the median 3. households that had income betweenthe median (average income)and twice medianincome 4. households that had income between twice and three times the median income 5. households that had income more than three times the medianincome. Table 3 providesjoint distribution o f respondents below and above the poverty line within each region. The percentages inthe first column are the poverty rates, also called head count index expressed inpercentages, for corresponding locations. Table3 Percentof HHBelow the PovertyLineBy RegionandLocation Total Region 1.oo Total HHincome* 2.00 3.00 4.00 5.OO 1 Mediterranean 115 164 140 49 25 493 23.3% 33.3% 28.4% 9.9% 5.1% 100.0% 2 Aegean 146 193 140 19 11 509 28.7% 37.9% 27.5% 3.7% 2.2% 100.0% 3 Marmara 164 385 471 118 64 1202 83 13.6% 32.0% 39.2% 9.8% 100.0% 4 South-East 102 136 53 3 -5.3% 294 34.7% 46.3% 18.0% 1.O% - 100.0% 5 East 70 111 89 12 3 285 24.6% 38.9% 31.2% 4.2% 1.1% 100.0% 6 Central 156 238 218 68 30 710 22.0% 33.5% 30.7% 9.6% 4.2% 100.0% 7 Blacksea 98 182 155 27 7 469 20.9% 38.8% 33.0% 5.8% 1.5% 100.0% Location Urban 626 1139 1131 267 129 3292 19.0% 34.6% 34.4% 8.1% 3.9% 100.0% Rural 225 270 135 29 11 670 33.6% 40.3% 20.1% 4.3% 1.6% 100.0% Turkey 851 1409 1266 296 140 3962 * 21.5% 35.6 32.0 7.5 3.5 100.0 1 Below thepoverty line 2 Between thepoverw line and the median. 3 Between the median and two medians 4 Between the two medians and the three medians 5 More than the three medians Southeast i s the region with the highest concentration o f poor people whereas Marmara has the least. Interestingly enough, East seems to have less number o f poor people than Aegean which may be related to the fact that the large portion o fpoor people who do not have anything at home have been migrating to Southeast region and Western cities and notably to Imir along with other cities in the Aegean region. It seems that poverty has also migrated with the poor migrants. When people who have incomes above the poverty line but less than the region average are added to those under the poverty line, Southeast ranks first with 81 percent which i s followed by Aegean (66.6 %), East (63.5 %), Blacksea (59.7 %), Mediterranean (56.6 %), Central (55.5 %) and Marmara (45.7 %). On both accounts, Marmara seems better off than all the other regions despite two devastating earthquakes lived in the region. Figure 2 pictures number o f households under (group 1 and 2 combined) and above the medianby region. 84 600 500 400 300 0 % 200 f0 100 =Under the median 0 20 0 Above the median Region Figure2 NumberofHouseholdsUnder andAbove the Medianby Region As for urban rural differences inpoverty, urbanplaces seem to have less poverty than rural places (19 % vs 33.6 %, respectively). 3.2 Food Ratio Measure Engel claims that the poor households spend larger portion o f their incomes on food. In Turkey, households spend on the average 40.4 percent o f their income on food which makes up 36.4 percent o ftheir total consumption. The ratio o f food expendituresto total income was formed and 40 percent cutoff point was used as the poverty line. Of course, one can perform the same analysis by forming ratio o f food consumption to total household consumption. In order to allow detailed analysis, percent o f food in the total household income was calculated and recoded the following way: 1. Lessthan40 percent o ftotal monthlyincome spent on food 2. 41 - 60 percent o ftotal monthlyincome spent on food 3. 61 - 80 percent o f total monthly income spent on food 4. 81 - 100 percent o f total monthly income spent on food. Here, respondents who were in the second through fourth categories are defined as poor. For the detailed analysis where gradations among poor i s needed, the second group can be termedpoor, the thirdpoorer and the fourth the poorest. The following procedure was used in the estimation o f poverty lines for household o f different size (For a similar analysis for 1994 SIS data refer to Erdogan 1996). Ifthe food expenditure i s denoted with c and total household income yt, the food ratio can be formed as 85 Ifthemaximumfoodratio isdenotedwithcpo (which is 40percent inthis case) When c / yt > cpo household i s poor and when c / yt < cpo household is not poor. It i s assumed that yt# 0. The relationship between food consumption and the household income is captured by the Engelfunction (Orshansky 1965, 1969) which canbewritten as Similarly, for the poverty line, one can write, Incpo =In(c / yt) =Inc -Inyt replace Inc with (1) =( a0 +a1 Inyt)-Inyt (2) Leaving the yt alone in (2) poverty line (yp) can be obtained. Inyp=(Incpo - ao)/ (a1- 1) (3) Foodratio i s also the function o fthe family size. Assumingthat food consumption has log linear distribution with the household size, then Inc (yt,hhs) = a0 + 011Inhhs ) +a2 Inyt (4) By estimating (4) parameters = a0 ,a1 anda2 are obtained. Here, assuming that yt# 0, If[c(yt,hhs)/yt]>cp,thehouseholdispoor,and If[c(yt,hhs)/yt]