75469 State Planning Organization of the Republic of Turkey and World Bank Welfare and Social Policy Analytical Work Program Working Paper Number 6: Growth, Employment, Skills and Female Labor Force Erol Taymaz Department of Economics, Middle East Technical University Ankara, March 2010 State Planning Organization of the Republic of Turkey and World Bank Welfare and Social Policy Analytical Work Program Working Paper Number 6: Growth, Employment, Skills and Female Labor Force Erol Taymaz Department of Economics, Middle East Technical University Ankara, March 2010 State Planning Organization World Bank Copyright @ 2010 The International Bank for Reconstruction and Development The World Bank 1818 H Street, NW Washington, DC 20433, USA All rights reserved The World Bank enjoys copyright under protocol 2 of the Universal Copyright Convention. This material may nonetheless be copied for research, educational or scholarly purposes only in the member countries of The World Bank. Material in this report is subject to revision. Growth, Employment, Skills and Female Labor Force iii Growth, Employment, Skills and Female Labor Force Table of Contents 1. Introduction ......................................................................................................................................................... 1 2. Data Sources........................................................................................................................................................ 1 3. Economic growth and employment..................................................................................................................... 2 4. Patterns of employment generation ..................................................................................................................... 4 5. Urban employment and “good jobs” ................................................................................................................... 6 6. Labor market participation and wages................................................................................................................. 10 References ............................................................................................................................................................... 17 Tables Table 1 : Population and employment, 2000-2006 .............................................................................................. 18 Table 2 : Output and employment growth by sectors, 2001-2007 .......................................................................18 Table 3 : Employment elasticities, 2001Q1-2007Q3 ...........................................................................................19 Table 4 : Distribution of employment by region, 2000-2006...............................................................................19 Table 5 : Distribution of employment by sector, 2000-2006................................................................................20 Table 6 : Distribution of employment by education level, 2000-2006.................................................................20 Table 7 : Distribution of employment by establishment size, 2000-2006............................................................21 Table 8 : Distribution of employment by occupation, 2000-2006 .......................................................................21 Table 9 : Distribution of employment by status, 2000-2006................................................................................21 Table 10 : Distribution of employment by registration status, 2000-2006.............................................................22 Table 11 : Average monthly wage rates in urban areas, 2006 ................................................................................22 Table 12 : Sectoral composition of urban employment, 2000-2006 ......................................................................22 Table 13 : Share of formal employment in urban areas by sector, 2000-2006.......................................................23 Table 14 : Share of “good jobs” in urban areas by sector, 2000-2006 ...................................................................23 Table 15 : Composition of urban employment by establishment size, 2000-2006 ................................................24 Table 16 : Share of formal employment in urban areas by establishment size, 2000-2006 ...................................24 Table 17 : Share of “good jobs” in urban areas by establishment size, 2000-2006 ...............................................24 Table 18 : Distribution of urban working age (15+) population by education, 2000-2006....................................25 Table 19 : Share of employees by education, 2000-2006.......................................................................................25 Table 20 : Share of formal employees by education, 2000-2006 ...........................................................................26 Table 21 : Distribution of urban young (15-24 years old) population by education, 2000-2006...........................26 Table 22 : Share of young employees by education, 2000-2006............................................................................27 Table 23 : Share of formal young employees by education, 2000-2006 ................................................................27 Table 24 : Distribution of urban polulation by household size, 2000-2006 ...........................................................28 Table 25 : Share of people living in households with any employment, 2000-2006 .............................................28 Table 26 : Share of people living in households with any formal employment, 2000-2006..................................29 Table 27 : Share of people living in households with any “good jobs”, 2000-2006 ..............................................29 Table 28 : Descriptive statistics on variables used in the labor market participation model,.................................30 Table 29 : Estimated labor market outcome probabilities at mean values (percentage) ........................................30 Table 30 : Marginal effects of schooling on employment probability ...................................................................31 Table 31 : Marginal effects of schooling on informal manufacturing employment probability ............................31 Table 32 : Marginal effects of schooling on informal services employment probability.......................................31 Growth, Employment, Skills and Female Labor Force iv Table 33 : Marginal effects of schooling on formal manufacturing employment probability................................32 Table 34 : Marginal effects of schooling on formal services employment probability ..........................................32 Table 35 : Marginal effects of schooling on employer probability ........................................................................32 Table 36 : Marginal effects of schooling on self-employment probability ............................................................33 Table 37 : Marginal effects of household characteristics on employment probability...........................................33 Table 38 : Descriptive statistics for wage workers, urban regions, 2006 ...............................................................34 Table 39 : Determinants of urban wages, 2006...................................................................................................... 34 Table 40 : Determinants of urban wages, 2006...................................................................................................... 35 Table 41 : Determinants of urban wages, 2006...................................................................................................... 35 Growth, Employment, Skills and Female Labor Force 1 1. Introduction female labor, and the determinants of wages. We estimate a multinomial logit model for labor market participation 1. The Turkish economy has undergone trough a decisions of men and women living in urban regions dramatic change since the 2001 Crisis. The economy for each year since 2000, and estimate the wage equation achieved historically high growth rates in 6 years in a for 2006 by taking into account the labor market row after a period of turbulence and boom and bust participation decision (the sample selection effect). The cycles in the 1990s. In spite of unprecedented growth last section presents the main findings of the study. performance, the unemployment rate remained at a very high level against the background of declining labor 2. Data Sources participation rates. The weak employment performance of the Turkish economy after the 2001 Crisis calls for 4. Any study that aims at studying the links between a comprehensive analysis of the dynamics of economic growth and employment needs two types of 1 employment generation. The transformation of the data: output and employment. The output data at the economy and the reallocation of labor from rural sectoral level come from the national accounts. The activities towards urban activities have been rather fast Turkish Statistical Institute (Turkstat) revised the way in the last few years. This transformation process raises it calculates the national accounts and announced a new the issue of skill mismatch because the labor released GDP series for the 1998-2007 period in March 2008, from rural activities may not have skills necessary for The new GDP estimates have been compiled according productive employment in urban activities. Moreover, to the European System of Accounts (ESA-95), whereas labor participation rate is extremely low for urban former series were based on the System of National women. That is likely to be caused by the lack of Accounts (SNA 68). As a result of the revision process, employment opportunities for women in urban areas. the estimate for GDP (with 1998 as the base year) increased by 32 percent in current prices for the year 2. This study aims at providing information on i) 2006 compared to former estimates (with 1987 as the growth-employment link, ii) the patterns of employment base year). growth, the role of skills during the growth process, and the links between the pattern of growth and 5. Turkstat announced that the revision is needed to household welfare, and iii) the mechanisms that link “incorporate more comprehensive data sources, adoption economic growth to poverty reduction through changes of updated statistical standards and improved estimation in labor market participation decisions and wage techniques”. The new series exploited the results of the determination. This report discusses recent changes in 2002 Census of Industry and Business Establishments main data sources that will be used in the study, and that extended the coverage of economic activity in a summarizes a descriptive analysis of growth and number of areas, notably in manufacturing, as well as employment at the sectoral level. in mining, and services. In manufacturing only, the number of establishments with 10 or more employees 3. The report is organized as follows. After this more than doubled to 28,059 from its number covered introduction, the second section explains the main data in the 2001 Survey. New series are based on improved sources used in the study. The third section discusses estimates on “unrecorded” or the informal economy. the relationship between sectoral output growth and The difference in employment levels estimated by the employment generation, and presents basic estimates Labor Force Survey (LFS) and Structural Business on growth elasticities. The fourth and fifth sections Survey (SBS) is used to adjust the reported data for present detailed descriptive analyses on the patterns of production by using average production per worker in 2 employment growth. The sixth section summarizes the the small manufacturing enterprises. The increased findings of an econometric analysis on labor market sample size of the LFS and the increase in accuracy participation decisions, with a special emphasis on improved the method for measuring the nonobserved 1 Unless otherwise stated, “output” refers to real value added. 2 This is the so-called “labor input method” as described in Measuring the Nonobserved Economy – A Handbook jointly published by the OECD, IMF, ILO and the CIS in 2002. Growth, Employment, Skills and Female Labor Force 2 economy (NOE). Turkstat also introduced new and second period after the Second World War ended in better data sources such as the 2000 Building Census 1960 following a decline in growth rates. Import and a new commodity flow system, and new methods substituting industrialization policies were adopted in for measuring and imputing services according to ESA the third period that is also called the “planned economy” 95, and over-the-year chain-linking method for constant period. Although the average growth rate was quite price estimation. high in this period, it ended in 1980 by a serious balance of payment crisis. Outward-oriented policies were 6. The second major source of data for our analysis adopted in the fourth period. Although the economy is the Labor Force Surveys (LFS) conducted by Turkstat. achieved high growth rates immediately after 1980, the Turkstat introduced two major changes in the way the growth rates tended to decline and became very unstable in the 1990s. The fourth period ended in 2001 when survey was conducted, one in year 2000 (quarterly the economy experienced one of its worst crises since surveys), and the other one in 2005 (monthly surveys the establishment of the Turkish Republic. aggregated to quarterly series), and the survey questionnaire was modified to some extent over time. 8. Apparently, the Turkish economy is now in its fifth The LFS results are weighted to estimate national cycle of growth. GDP growth rate after the 2001 Crisis aggregates by using population projections. The last remained quite high for 6 years in a row, although it population census in Turkey was conducted in 2000. declined somewhat in recent years (Figure 2). The Turkstat estimated population projections (based on annual growth rate of GDP reached almost 10 percent age, gender and location) on the basis of the 2000 in 2004, but it declined gradually to 4.5 percent in 2007. Census. A new system of recording, the Address-based Employment did not respond rapidly after the crisis, Population Registration System (APRS), was introduced and remained almost at the same level (about 21.5 by a law enacted in 2006 (Population Services Law, million in 2000-2003), but started to increase at about No 5490), and Turkstat announced that there were 71 2% per year after 2003. The rate of increase in the million people living in Turkey at the end of 2007, number of employed is certainly quite low given the whereas the population estimate for the same year was high rate of population growth. Population increased about 74 million. Turkstat now estimates the labor force almost by 1 million each year since 2000 (average statistics by using the population data from APRS. annual growth rate was 1.6 percent). Since employment Although the revision from population projections to did not increase at the same rate, the employment ratio APRS is claimed to have no significant impact on (the share of employed people in population) declined distributions, the levels of all variables (such a during and after the 2001 crisis, from 32.5 percent in population, number of employees, number of 2000 to 30.6 percent in 2003 (see Table 1). 3 unemployed, etc) have been reduced by about 4 percent. 9. The aggregate data conceal substantial reallocation of labor from rural to urban areas and from agriculture 3. Economic growth and employment to industry and services. There is almost a stable decline in employment in rural areas whereas the urban 7. In order to understand the links between economic employment grew rapidly after the crisis. The net growth and employment since the 2001 Crisis, we need increase in the number of urban jobs from 2002 to 2006 to contextualize the period after 2001 in a historical amounted to 2 millions. Although it tends to increase setting. Figure 1 presents the data on GDP growth rates gradually, the employment rate is very low for urban for the Turkish economy since the establishment of the women. This is one of the most striking and problematic 4 Republic. As can bee seen in this figure, there are four aspects of employment issues in Turkey. clearly identifiable growth cycles in the Turkish economy. Each cycle takes about 20-years. The first 10. Since the employment structure changed rapidly after one from the beginning until the Second World War is the crisis, we need to look at sectoral changes as well (see 5 characterized by high and volatile growth rates. The Table 2). GDP increased almost 40 percent from 3 See the notes on LFS on the Turkstat web page, Hanehalký �þgücü Anketleri Hakkýnda Genel Açýklama, www.tuik.gov.tr. 4 5-year moving averages were used to reduce the effects of annual fluctuations. 5 Sectors were reclassified to have the same set of sectors for all LFS surveys and national accounts for the 2000-2007 period. Growth, Employment, Skills and Female Labor Force 3 2002 to 2007. Construction, transportation and utilities (EGW) also grew rapidly after the crisis (about communication, manufacturing, utilities (electricity / 60 percent from 2001 to 2007). Financial intermediation gas / water, EGW), and wholesale and retail trade and real estate services were badly affected by the crisis (including hotels and restaurants) achieved above- and started to recover only after 2004. Agriculture had average growth during the same period. The performance a volatile and stagnant output whereas mining, and other of agriculture was poor. services had low growth rates (on average, about 3 percent per year). 11. Table 2 provides the data on output and employment growth at the sectoral level for the 2001-2007 period. 14. On the employment side, finance and real estate Employment growth rates are given for urban and rural services achieved an outstanding performance after 2004 areas, and for men and women separately. The data (annual growth rate of employment was about 10 percent shows that the Turkish economy created a little more after 2004). Mining and utilities had erratic employment than 1 million new jobs from 2001 to 2007. Agriculture patterns with wild fluctuations. Wholesale and retail and fishing experienced a huge loss (25%, more than trade, hotels and restaurants, manufacturing, other 2 million jobs) which is partly compensated by new services, and transport, storage and communication jobs in wholesale and retail trade, hotels and restaurants services provided stable and continuous employment (1.1 million), other services (0.75 million), growth. Construction experienced a decline in manufacturing (0.66 million), and other sectors. It is employment until the late 2003, and it recovered up to interesting to observe that in all three major employment its 2001 level only in 2007. There is a striking decline creating sectors (wholesale and retail trade, hotels and in agricultural employment in 2005 and 2006. restaurants, other services, and manufacturing), the growth rate of female employment is higher than the 15. The slopes of the curves depicted in Figures 3 and growth rate of male employment. On average, male 4 reveal employment generation potential of the sectors. employment increased by 6.2 percent whereas female The sectors that have higher employment elasticities employment increased only by 1.6 percent in six years will have steeper employment-output curves. Moreover, (from 2001Q3 to 2007Q3). if the employment-output relationship is stable, the curve will get closer to a straight line. 12. In spite of the sharp decline in agricultural employment, agriculture is still the largest sector in 16. There seems to be a strong correlation between Turkey in terms of the number of employees (6 million employment and output growth in other services (the people at the end of 2007). Wholesale and retail trade, slope is almost equal to one). Finance and real estate hotels and restaurants (4.8 million), manufacturing (4.2 and wholesale and retail trade sectors have also steep million), and other services (3.8 million) are among the employment-output curves. These sectors are likely to main employment generating sectors. generate more employment if they grow rapidly. It seems that the relationship between employment and output 13. Sectoral data are presented in Figures 3 and 4 to growth is quite stable in manufacturing, and transportation provide visual evidence on the relationship between and communications, but the employment generation output growth and employment generation. In order to potential of these sectors seem to be lower. Agriculture have a comparable data across all sectors, the output is an obvious outlier with a sharp decline in employment and employment for each sector is normalized to 1 for and mediocre output growth. Construction experienced 6 2000Q4. As shown in these figures, construction, and a sharp decline in output and employment in 2001 and transport and communication services achieved highest 2002, but achieved a smooth transition towards a stable growth rates after the crisis. These two sectors achieved output and employment growth path. Mining and utilities more than 75 percent growth in 6 years. Wholesale and reveal erratic growth with almost no positive correlation 7 retail trade, hotels and restaurants, manufacturing and between employment and output. 6 In order to eliminate the effects of short term fluctuations and seasonal factors, we use 4-quarter moving averages. The first data point refers to the 2000Q1-2000Q4 average whereas the last point the 2006Q3-2007Q3 average. 7 Since these two sector employ about 100-150 thousands people, they are likely to exhibit wide fluctuations in employment due to random sampling. Growth, Employment, Skills and Female Labor Force 4 17. The quality of employment is as important as the 21. Male employment growth is positively correlated quantity of employment. As may be expected, those with output growth in other services (0.861), finance sectors that had the worst employment performance are and real estate services (0.350), manufacturing (0.343), among the sectors that achieved the highest growth in and wholesale and retail trade (0.212). In the construction labor productivity. These are construction, agriculture, sector where almost all employees are male, the output and utilities. It seems that either labor hoarding was elasticity is also quite high (0.609). Female employment prevalent in these sectors (which could be the case for growth is positively correlated with output growth only agriculture), or these sectors relied heavily on in wholesale and retail trade at the 10 percent level. outsourcing their activities to informal firms (which The female employment elasticity of output in this could be the case for construction). sector is 0.376. The lack of correlation between female employment growth and output growth in other sectors 18. Transportation and communication services sector may indicate the problems in providing jobs for female was able to increase both productivity (by 45 percent) workers in urban areas. We estimated output elasticities and employment (by 15 percent) from 2001 to 2007. for skilled (vocational school and university graduates) Manufacturing and wholesale and retail trade, hotels and unskilled (others) female workers but with the and restaurants had moderate increases in labor exception of unskilled female employees in wholesale productivity (3.8 percent and 1.9 percent per year, and retail trade, non of the estimated elastcities were respectively), whereas other services, finance and real found to be significant. estate services, and mining were not able to increase labor productivity at all in spite of rapid growth in the 22. Employment elasticities provide valuable infor- economy. mation about employment generation potential of the main sectors of the economy. The findings, however, 19. What is the employment elasticity of output growth? need to be interpreted cautiously because of the way We calculated annual changes in output and employment the output data are “estimated” for services. For other (from one quarter to the same quarter in the next year) services sector that includes education, health, and for all sectors, and looked at the correlations between public services, the output data are difficult to collect employment growth rates and output growth rates. If at thigh frequency. Therefore, Turkstat uses the growth the share of female employment in that sector is sizable, rate of private employment data from LFS to calculate we looked at the growth rates of male and female the real private output growth for other services sector. employment separately. “Employment elasticity” is Similarly, employment data from LFS are used in the defined as the estimated coefficient of output growth calculation of the output of the “domestic restaurants” rate variable in the model where employment growth sector which accounts more than half of the “hotels and rate is regressed on output growth rate. restaurants” sector. It is thus no surprise that the “other services” sector has the highest employment elasticity, 20. The employment elasticities for the 2001Q1-2007Q3 and lowest productivity growth. period are presented in . Other services has a very high employment elasticity (0.979). A one-percent increase 4. Patterns of employment generation in the output of other services sector leads to almost one percent increase in employment. Construction, and 23. In this section, the patterns of employment finance and real estate have also high employment generation in Turkey since 2000 are analyzed in detail elasticities (0.609 and 0.462, respectively). Employment to understand the dynamics of employment. elasticities of manufacturing (0.268) and wholesale and retail trade (0.231) are also positive and statistically 24. The data on the distribution of employment by significant. Agriculture, mining, utilities (EGW), and region (urban vs rural) are presented in (Table 4). As transportation and communications do not have mentioned before, there is a substantial reallocation of 8 statistically significant employment elasticities. labor from rural areas (mainly from agricultural 8 There is no correlation between agricultural output and employment but the intercept (constant) term is negative: agricultural employment tends to decline by 3.5 percent per year irrespective of the growth rate achieved in agriculture and fishing. Growth, Employment, Skills and Female Labor Force 5 activities) to urban areas (mainly towards industrial of vocational school and college (2- and 4-year higher activities and services). Total employment increased education and higher degrees) categories such that the only 3.7 percent from 2000 to 2006. The share of rural vocational school and college graduates accounted for areas in total employment declined rapidly, from 48.1 9.9 percent and 12.4 percent of all employees in 2006. percent (31.2 percent male, 16.9 percent female) to 41.4 Female employees have a more polarized education percent (27.5 percent male, 13.9 percent female). level than male employees. The shares of illiterate, Although there are some minor fluctuations, the share literate without any diploma, and college graduates are of men remained around 75 percent throughout the higher among female than male employees. period. 28. There seems to be a gradual shift towards working 25. The sectoral distribution of employment is shown in larger establishments (Table 7). The distribution of in (Table 5). Agriculture is the only sector with a decline employment by establishment size shows that the share in employment shares for both men and women. There of micro- and small establishments (those that employ is more than 8 percentage points decline in the share of less than 25 people) declined from 78.7 percent in 2000 agricultural employment. Wholesale and retail trade, to 70.7 percent in 2006. It is interesting to observe that hotels and restaurants increased its share in total the increase in the share of medium-sized and large employment by 3.3 percentage points, followed by other establishments cannot be explained only by urbanization, services (2.1 percentage points), and manufacturing because it is observed in both rural and urban areas for (1.9 percentage points). male employees, and in rural areas for female employees. The most rapid decline in the share of small 26. The share of female employees in mining, and establishments is observed in 2002 that could be due 9 construction is almost nill. Women employees are to the effects of economic crisis. heavily underrepresented in utilities (EGW), and transportation and communication, whereas they have 29. The changes in the occupational distribution of a relatively higher share in agriculture, other services, employment are dominated by the decline in agricultural and finance and real estate. Although other services, employment (Table 8). There is a sharp decline in the and finance and real estate generated a large number share of skilled agricultural workers (13.3 percentage of new jobs, they failed to compensate for the decline points from 2001 to 2006). There is an increase in the in agricultural employment, and the share of female shares of all other occupations but craft workers that employees in total employment declined 3 percentage experienced 1 percent point decline. Machine operators points from 2002 to 2006. and elementary occupations had the highest growth rates for both men and women. The share of female 27. The composition of employment by educational service workers had a relatively high growth as well. level has also changed significantly since 2000 (Table 6). The share of illiterate employees declined gradually, 30. There is a rapid increase in the share of regular from 3.1 to 1.7 percent for men and from 5.5 to 3.7 employment since the 2001 crisis due mainly to the percent for women, from 2000 to 2006. The most decline in the share of unpaid family workers (Table significant change is observed in the shares of primary 9). The share of casual workers declined among men and secondary education categories. As a result of the while it increased among women. These changes are law extending the duration of compulsory education to partly due to the decline in rural population, because 8 years in 1997, we observe a rapid decline in the share female unpaid family workers and male self-employed of primary school level, and an increase in the share of are the dominant categories of employment status in secondary school level. However, in spite of these rural areas. However, similar trends (the increase in changes, primary school graduates still constitute the regular employment, and the decrease in male casual largest group of employees. The fastest increase for employment are also observed in urban employment. both male and female employees is observed in the case Although the share of regular employment increased 8 There is no correlation between agricultural output and employment but the intercept (constant) term is negative: agricultural employment tends to decline by 3.5 percent per year irrespective of the growth rate achieved in agriculture and fishing. 9 This could also be due to a change in survey design because there seems to be a significant shift from small (10-24 people) to medium-sized (25-49 people) establishments. Growth, Employment, Skills and Female Labor Force 6 substantially, there is not a significant increase in the higher average wages whereas construction, 10 share of formal labor (Table 10). On the contrary, there trade, manufacturing and agriculture tend to pay was an increase in the share of informal employment lower wages. from 2000 to 2004, especially in urban areas. The share of informal employment declined about 5 percentage ii) Informal workers get much lower wages than points between 2004 and 2006 mostly due to the decline their counterparts working in the formal sectors. in rural employment where informal employment It is apparent that informality pays lower wages relationship is dominant. It is interesting to observe that for both men and women. although informal employment among urban men declined slightly after 2004, there is no similar decline iii) There are substantial wage differences between among urban women in the same period. men and women working in the informal sector, whereas, there is almost no gender wage 5. Urban employment and “good jobs” differential in the formal sector. This finding supports our earlier observations about the 31. The analysis in the preceding section shows that polarization of skills among female employees. Turkey experienced a rapid reallocation of labor from rural to urban areas, and from agriculture to industry iv) Finally, “good jobs” pay higher wages: those and services, since 2000. The huge decline in agricultural workers employed in “good jobs” get about 20 employment was not compensated for by the new jobs percent more than formal workers do. generated in urban regions. Although most of the employees in agriculture are self-employed and unpaid 34. Given the information about wages, we will now family workers (65 percent of rural employment in analyze the composition of employment and its quality 2006), these two categories account for a relatively in urban areas. The data on sectoral distribution in urban small part of employment in urban areas (only 19 areas are presented in Table 12. There are about 2.6 percent). Therefore, in order to understand the million new jobs created in 5 years following the 2001 relationship between growth, employment and poverty crisis. The share of female employees in urban reduction, we need to analyze the dynamics of urban employment increased slightly during the economic employment. crisis (2001 and 2002) due to the decline in the number of male employees (mainly in manufacturing and 32. There is an influential literature on skill traps caused construction) in 2001 and 2002, and an increase in the by skill-investment or skill-R&D complementarities number of female employees (in manufacturing, trade, (see, for example, Snower, 1994; Redding, 1996; and other services) in 2002. In other words, the crisis Acemoglu, 2001; Burdett and Smith, 2002). These had, on average, a weaker effect on female than on studies indicate the importance of high skill-good jobs male employment. The gender distribution remained that pay high wages for economic performance. almost constant after 2002. Although there are some Following this literature, we defined “good jobs” in minor fluctuations, the share of female employment this context as those formal (registered with a social has a tendency to increase only in wholesale and retail security institution) jobs for vocational school and trade, hotels and restaurants sector. Male employment college graduates. experienced a slight increase in the finance and real estate sector after 2001. 33. Average monthly wage rates for various groups of workers in 2006 are presented in (Table 11) . The data 35. The share of formal employment (as a percentage on wages reveal four facts on wage differences. of total employment in the sector) exhibits significant differences across sectors and gender (Table 13). Formal i) There are sizable inter-industry wage differentials. employment is dominant for men and women in utilities, Utilities (EGW), transportation and communica- mining, other services, and finance and real estate. The tions, mining, and other services pay relatively dominance of formal employment in all these sectors 10 We define “formal” and “informal” employment by the registration in a social security organization which is obligatory by law for all employees (for workers, civil servants, and employers/self-employed). Growth, Employment, Skills and Female Labor Force 7 (except finance and real estate) can be explained by the share for educated women than for educated men large share of public companies and institutions. The among those who are employed. Other services, finance share of formal female employees is much higher than and real estate, and utilities are the sectors that offer the share of male employees in construction and proportionally more “good jobs” for men and women. transportation. These two sectors are characterized by very low female employment. It seems that female 38. The changes in the distribution of employment by employees in these sectors work more in highly skilled establishment size have important policy implications. positions so that they benefit more from formality. In As we have seen before, small firms provide the bulk the case of manufacturing where a large number of of employment in Turkey. It is the case in urban female workers are employed, the formality rate is employment as well (Table 15). About half of all much higher for men (73.8 percent on average) than employees work in micro-establishments (those that for women (57.2 percent). This finding reflects the employ less than 10 people), whereas large precautious position of female workers in manufacturing. establishments (those that employ 50 or more people) As may be expected, formality is extremely low in account for less than 30 percent of urban employment. agriculture. The size distribution of employment did not change much from 2000 to 2006. There is a slight increase in 36. There is a very sharp decline in the extent of the share of female employment in micro-establishments, formality from 2000 to 2004, especially in the case of and male employment in medium-sized and large female employees. The shares of formal female and establishments, and a decline in the share of male male employment declined 9 and 5 percentage points employment in micro- and small establishments. in that period. It seems that the increasing share of female employment during the economic crisis is 39. There is a sharp difference between micro- and achieved at the expense of formal employment. other establishments in terms of formality of the Informality increased at a faster rate in services (trade employment relationship (Table 16). The share of formal and other services for female employees, and trade, employees is very small in micro establishments (48 transportation and communication, and finance and real percent for men, and only 28 percent for women on estate for male employees). There has been a minor average). The extent of formality increases mono- increase in the share of formal employment in 2006 but tonically by establishment size and exceeds 90 percent the data are not sufficient to suggest if this is the for both men and women in large establishments. There beginning of a new upward trend in formality. is a significant decline in the share of formal employees in micro-, small and medium-sized establishments from 37. Although the extent of formality declined during 2000 to 2004. The decline in formality in large and after the economic crisis in 2001, the share of “good establishments remained at low levels, but it is worrying jobs” (those formal jobs that employ vocational school to observe that although the average level of formality and college graduates) exhibits almost a continuous increased in 2006, it continued to decline in large 11 increase over time, including the crisis years. It seems establishments for male and female employees. that less-skilled workers are more vulnerable during the economic crisis, and they tend either to lose their 40. There is a strong positive correlation between jobs, or to shift to informal employment, probably under establishment size and the share of “good jobs”: large poor working conditions. As may be expected, the share firms tend to offer proportionately more “good jobs” 12 of “good jobs” in almost all sectors is higher for female than small establishments do (Table 17). Interestingly, employees than male employees. The high share of the share of “good jobs” for female employees is higher “good jobs” among women is an outcome of the labor that that for male employees in all but micro- market selection process: since the differential between establishment categories. There is not much difference labor market participation propensity between women between the shares of “good jobs” for men and wo- and men is reduced by education, we expect a higher men (9.5 and 11.7 percent, respectively) in micro- 11 There is an abrupt decline in 2004 that may be caused by changes in sampling methods in LFS in that year. 12 The only exception is agriculture. Growth, Employment, Skills and Female Labor Force 8 establishments. It seems that educated urban women the share for men (Table 20). The share of formality could find more jobs in relatively large establishments. and gender differentials depend monotonically on The share of “good jobs” increased slightly in small educational level. There seems to be no gender difference and medium-sized establishments, and significantly in for high school (77 percent), vocational school (79-81 large establishments especially during the time of percent) and college graduates (more than 90 percent), economic crisis in 2001. The share of “good jobs” whereas there is a substantial formality differential offered by micro-establishments is not only very low, between less educated women and men. Moreover, the but also stable without any long term improvement. share of formality is extremely low among less educated men and women. The economic crisis in 2001 had a 41. The distribution of urban working age population large unfavorable impact on formality. The share of (people 15 years of age or older) by gender and formal employees declined sharply irrespective of educational level is shown in Table 18. Because of educational level and gender, but the decline was deeper declining birth rates and rising life expectancy, the share among less educated and women. There was a slight of working age population in total population tends to improvement in the extent of formality in urban areas increase (the last row in Table 18). In terms of the in 2006. It seems that the changes in formality are pro- educational level of the working age population, we cyclical. observe a sharp decline in the share of primary school graduates, and a lesser degree increase in the share of 44. In order to check the position of new entrants in secondary school graduates. This is the expected outcome the labor market, tables 18-20 are reproduced for the of the new law extending the duration of compulsory young people aged between 15 and 24 (see Table 21- education to 8, as mentioned above. Moreover, the share Table 23). There is a slight decrease in the share of of college graduates grows continuously (2 percentage young in total population, from 19.5 percent in 2000 to points in 6 years). It is interesting that, in spite of the 17.3 percent in 2006 (the last row of Table 21). The popular rhetoric about the ineffectiveness of vocational composition of young by educational level shows that, schools, the share of vocational school graduates rises for both men and women, the share of more educated rapidly for both men and women (3 percentage points is higher among young than the older population thanks from 2000 to 2006), and the share of high school to improved access to schooling in recent decades. The graduates declines (1.6 percentage points in the same share of college graduates is lower among young than period). Finally, the share of educated women is much the older population simply because of the fact that lower than the share of educated men. There is a sharp most of the young people are at the college education reduction in the probability for women to continue after age, i.e., there are many young people currently enrolled compulsory education so that the share of educated in higher education. women (those with at least high school diploma) was only 19.4 percent in 2006 whereas the same ratio was 45. The share of employed among young people by 28.7 percent for men. educational level and gender is much lower than the share of employed adults because of continuing 42. The employment ratio (the employment/working schooling. However, the crisis in 2001 seems to have age population ratio) is extremely low for urban women, a strong negative impact on the employment prospects and it is increasing gradually (from 15.2 percent in 2000 of the young people who experienced a deep decline in to 16.7 percent in 2006, see Table 19). The share of the employment ratio from 2000 to 2003. The decline employed men declined sharply from 2000-2003, and is especially noticeable for male college graduates. This it increased somewhat until 2006. The employment finding shows that even the educated young people face ratio increases rapidly by education for men, but the with serious problems in finding jobs during an economic effect of education on employment rate is much lower downturn. for women with the exception of college education that boosts the employment prospects for women. 46. The extent of formality among young employees is shown in Table 23. The composition of and the trend 43. Among those employed in urban areas, the share in formality among the young is similar to the ones of formal employees for women is slightly lower than observed among the working age population. Informal Growth, Employment, Skills and Female Labor Force 9 employment is dominant among less educated young, live in small (1-3 people), and 12.1 percent in large (7 and it increased sharply across all education categories and more people) households. There is a very modest from 2000 to 2004, followed by a minor decline in 2005 shift in the size distribution: the share of middle-size and 2006. There is a discernible difference between the households (4-6 people) declined slightly (2.5 percentage patterns of formality among young men and women. points), and the shares of small (1.5 percentage points) Although the shares of formality among young and and large (1 percentage point) households increased adult women are almost the same for all education from 2000 to 2006. categories, the share of formality among young men is much lower than the share among adult men to the 50. The data on the shares of people living in a extent that and educated young man is less likely to household with at least one employed person reveal have a formal employment than an educated young information about the most vulnerable groups (Table woman. 25). The share of employed among women living alone (women living in one-person households) is extremely 47. These finding show that, i) less educated young low, only 12.4 percent on average. Half of men living women start their job careers with informal jobs, and alone are also unemployed. The share increases by they are not likely to move to formal jobs over time, ii) household size, but there seems quite a large number less educated men also start with informal jobs, but a of people living in large households without any large number of them (about one quarter of men without employed household member. About 13-14 of all women any diploma and more than half with at least primary and 11-13 percent of all men live in large households school diploma) are eventually employed formally, iii) (with 7 or more people) without any employed member, educated young women are more likely to start their i.e., about one million people live in large households job careers with formal, secure jobs (almost 90 percent), without any employment. iv) while educated young men have initially lower formality ratio than educated young women (on average, 51. The crisis in 2001 led to an increase in the share of 78.4 percent for young men vs 88.7 percent for young people living in households with no employment (about women), but they tend to have more formal employment 5 percentage points from 2000 to 2003). Interestingly, over time. In other words, a young man is more likely large households were affected more by the crisis, and to start with informal employment and to move to formal the share of people living in large households with no employment, whereas a young woman is more likely employment increased by 9 percentage points in the to start with and to stay in informal employment if she same period (Table 25). is less educated, and in formal employment if she is more educated. 52. The extent of formal employment changes by household size. There is an inverted-U type relationship 48. We have analyzed so far changes in employment between formality and household size. The share of patterns at the individual level. However, what is people living in households with at least one formal important for welfare and poverty analysis is the changes employment is very low in small and large households, in employment at the household level, because the and it reaches its maximum value in households with household is the economic unit in which the decisions 4 people (Table 26). About 70 percent of people living on employment and consumption are made. The in households with 4 people have at least one formal calculations on poverty measurement are also made by employee in the household (so that all household member at the household level by using the data on household are likely to benefit from social security coverage), but income and expenditures. the same rate is less than 50 percent in very small (1- 2 people), and large (7 or more people) households. As 49. We classify households by household size (the expected, the share of people living in households with number of people living in the household). Table 24 at least one formal (registered) employee declined presents the data on the distribution of urban population sharply from 2000 to 2004 (about 7.3 percentage points), by household size and gender. The bulk of the urban and the decline is larger among large households. population lives in households with 4-6 people (58.5 percent on average), whereas the remaining 29.3 percent 53. The share of people living in households with at Growth, Employment, Skills and Female Labor Force 10 least one “good job” (a formal job for vocational school and the type of job. In this specification, we assume and college graduates) has a distribution similar to the that the person will produce qh units of services by one observed for formal employment. It is very low in home production if she opts to stay at home. If she very small and large households, whereas it is higher works at job j, she will receive a certain wage (wij) and in households with 3-4 people (about 23-26 percent). a benefits package (sj). She will pay for the services However, the share of people living in households with that would otherwise be provided by home production at least one “good job” did not experience any decline (for example, child care, dining, etc.), and the rest of during the economic crisis. On the contrary, it increased her wage (wij – pihqih) will be spent for market-based almost continuously during and after the crisis across products. all households and gender categories (almost 5 percentage points from 2000 to 2006). The increase in 56. We assume that there are six types of jobs available the share of people living in households with at least for an individual: formal jobs in manufacturing (fm) one “good job” is likely to contribute to reduce the and services (fs), informal jobs in manufacturing (im) 13 extent of poverty in urban Turkey. and services (is), entrepreneur (employer, e) and self- employed (se), i.e., j {fm, fs, im, is, e, se}. If the person stays at home, she is considered to be at the “non- 6. Labor market participation and 14 employment” status (h). The person will chose the wages state that maximizes her utility. 54. The analysis in the preceding section describes the 57. There are a number of critical variables that changes in employment patterns at the aggregate level. determine an individual’s labor market participation In order to shed light on the determinants of these decision. First, the most important variable that changes, it would be helpful to look at labor market determines an individual’s decision is the level of participation decisions and the determination of wages education (or the level of human capital) the individual at the individual level. We assume that a person at the has acquired, because the level of education determines working age has two options. First, she may prefer to the wage rate. If the level of education has the same stay home and to participate in non-market home impact on the wage rate in all types of jobs, the production such as caring children and elderly, cooking, probability of employment will increase by education. home renovations, etc. Leisure can also be interpreted If there is a difference in the impact of education in as home production. Second, she can participate in the different types of jobs (for example, in formal vs informal labor market and accepts a job in the formal or informal jobs, or in manufacturing vs services), then the sector. probability of getting employed in a certain type of job will change by education. 55. The utility the person will derive from these options could be defined as follows: 58. Second, the quantity of services provided by home Uih = Uih(qih) [1] production is important. If an individual is required to Uij = Uij(sij, qim,qih) = U(sij, (wij – pihqih)/pim, qih) [2] provide more services by staying at home, she is more where qh is the quantity of home production (for likely not to participate in the labor market because if example, number of hours worked to provide home she works in a workplace she will spend a larger part services), qm the quantity of products and services of her wage to buy these services on the market. consumed, w the wage rate, ph the price of services provided by home production, pm the price of other 59. Third, the value of non-wage benefits obtained (market-based) products and services, and sj the non- through a (formal) employment will have a significant wage benefits such as unemployment benefit, health impact on labor market decisions. If, for example, a insurance, etc., the person will receive for being person enjoys health benefits thanks to a formally employed at job j. Subscripts i and j refer to the person employed person in the household, she will not receive 13 For an employer, w includes the profits as well. 14 “Manufacturing” includes manufacturing proper, mining, and utilities. “Services” include construction, trade, transportation and communications, finance and real estate, and other services. Growth, Employment, Skills and Female Labor Force 11 any additional health benefit by participating in the 64. The effects of educational level are captured by five labor market. In such a case, she will be inclined to stay dummy variables: Primary for literates and primary 17 at home (non-employment), or, to accept informal jobs. school graduates , Secondary for secondary school However, if there is no formally employed person in graduates, High school for high school graduates, the household, the incentives to get a formal job would Vocational for vocational high school graduates, and be much higher because that job would bring forth College for 2- and 4- year higher education graduates. additional benefits (to all household members). The omitted variable is the Illiterate category, i.e., the educational level dummy variables measure the effects 60. Fourth, relative prices of market-based products of relevant levels of education relative to illiteracy. and services provided by home production are important in determining the labor market participation decision. 65. There are two dummy variables for marital status: If it gets cheaper to buy services provided by home Single for never-married singles, and Divorced for the production (for example, if the child care gets cheaper), divorced and widowed. The omitted variable is the then the incentives for participating in the labor market Married category. will get stronger. 66. In order to test the effects of household size, we 61. We estimate a multinomial logit model to understand include to the model the Parent*household size and the determinants of labor market decision for men and Child*household size interactions. The household size is measured by the (log) number of people in the women aged 15 or more living in urban areas. In order household. It is interacted with the Parent and Child to observe changes over time, the model is estimated dummy variables because the effects of household size for each year separately. There are seven labor market 15 on parents and children are likely to differ. We expect outcomes: the base outcome is non-employment. The that the Parent*household size variable may have a next four outcomes are about wage employment in negative effect on the probability of non-employment formal/informal manufacturing and services. The last for men, but it may have a positive effect on the two outcomes, the “employer” and “self-employed” probability of non-employment for women, because categories are not classified into sectors because of the women are more likely to be involved in home lack of sufficient number of observations. The following production, and the need for home production will variables are used as explanatory variables in the increase by the household size. In other words, the value 16 multinomial logit model. of household production will increase for women by household size, whereas the need for workplace 62. The age of a person has a significant impact on employment will increase for men. labor market decisions. We added (log) age and its square as explanatory variable to allow for non- 67. We use a dummy variable, any formal, to test if the monotonic effects of the age variable. We expect that availability of social security benefits provided by another there would be a U-shaped relationship between age formally employed person in the household increases and the probability of non-employment outcomes. the non-employment probability and decreases the formal employment probability of other persons in the household 63. The status in the household is captured by the Child (the effect of the s term in Equation 2). We expect that dummy that takes the value 1 if the person is if there is a formally employed person in the household, “daughter/son”, “daughter-/son-in-law”, “grand other members of the household are likely to benefit daughter/son”, or “other relative/non relative” aged less from social security coverage (health insurance, etc.) so than 30, and 0 otherwise. The omitted variable is the that if they get a formal job, the value of additional non- “Parent” category that includes all other people not wage benefits will be low. This may discourage other 18 included in the Child category. household members to get a (formal) job. 15 This outcome includes non-participation in the labor market, and unemployment. We experimented with separate “agriculture” and “unpaid family worker” outcomes as well, but the estimation results for these outcomes were not significantly different from non-employment in most of the cases. Therefore, we added “agricultural employment” and “unpaid family workers” to the “non-employment” category. 16 The descriptive statistics on the variables used in estimating the labor market participation model are presented in Table 28. 17 The “literate without any diploma” and “primary school diploma” categories are merged together because the number of people in the former category was too low. 18 Although the employment decision, as discussed here, is likely to be made at the household level, we model it at the individual level because of the lack of panel dimension in our data. Growth, Employment, Skills and Female Labor Force 12 68. Finally, we use a dummy variable for persons in a points). This finding is contrary to the popular household whose “head” is unemployed, Unemployed understanding that considers self-employment as a HH. This variable takes the value 1 for a person whose means of survival in the last resort during severe 19 21 household head is not employed, and 0 otherwise. The economic crisis. household head is expected to earn the bread and butter for the household. If the household head is unemployed, 72. The effects of education on labor market outcomes 22 the incentives for other household members may change. are summarized in - Table 30-Table 36. There are major differences between men and women in terms of 69. The estimated coefficients from the multinomial the marginal effects of education on labor market logit model are difficult to interpret quantitatively. Thus, outcomes. First, primary and secondary schooling has 23 we calculated the marginal effects of each variable on insignificant effect on the employment probability for labor market outcomes. For continuous explanatory women although its effect for men is quite important variables, the marginal effect is the change in the (on average, 13.6 percent for primary school and 35.2 probability of the relevant outcome’s realization in percent for secondary school). Second, higher education response to a slight change in the dependent variable (College) has a much higher impact on women’s i.e., the marginal effect is defined as Pr(j)/ xk where employability than on men’s employability (73.4 percent Pr(j) is the probability that the labor market outcome vs 47.1 percent) partly because of the fact that an j will be chosen, and xk is the kth explanatory variable. illiterate woman has almost no chance to be employed For dummy variables, the marginal effect shows the whereas the employment probability of an illiterate man change in the probability Pr(j) induced when the dummy is more than 50 percent. Third, the impact of education variable changes from 0 to 1. on women’s employment probability declined to a small extent during the crisis but the marginal effect of 70. Estimated labor market probabilities for all outcomes education for men increased rapidly in the same period. 20 for an average illiterate married parent without any formal employee in the household are presented in 73. The differences on the effects of education on men’s Table 29 . The probability of non-employment was and women’s employment prospects become even more extremely high for women, and it declined slightly (less informative when the effects on various labor market than 1 percentage point) since 2004. It was 36.0 percent outcomes are analyzed. Education (with the exception for men in 2000, and increased significantly in 2001 of higher education) improves women’s employability and 2002, and reached 49.5 percent in 2004. The in informal jobs in the industry and services, but probability of non-employment for men declined only education beyond secondary school has a significant after 2004 following 4-years of rapid growth in the negative impact on men’s employment probability in economy. informal jobs, i.e., more educated women (but not the most educated ones) tend to work in the informal sector, 71. It is interesting to observe that the probability of but more educated men afford not to accept informal informal wage employment (both in the industry and jobs (Table 31 and Table 32). services) increased for women during the economic crisis while the probability of formal wage employment 74. The effect of education is stronger in the case of decreased. The probability of formal wage employment formal employment in manufacturing, and, especially, for men experienced even a larger decline, from 33.6 in services (Table 33 and Table 34). Educated men and percent in 2000 to 22.9 percent in 2004. Moreover, the women are more likely to have a formal job. The probability of self-employment for men also declined marginal effect of education on the employment significantly from 2000 to 2002 (almost by 3 percentage probability in formal manufacturing declined in 2005 19 By definition, the value of this variable is zero for household heads because we test the effect for other household members. 20 Sample means are used for all continuous variables. 21 Cárdenas and Villarreal (2007) suggest that “self-employment has been an escape to inflation and staggered wages bringing as a consequence reduced inequality” in Mexico. The experience in Turkey seems to be different. 22 The marginal effects for women are not reliably calculated for 2000 and 2004 because of the lack of observations for certain categories. Thus, they are not presented in these tables. 23 The employment probability is equal to (1– non-employment probability), which is also equal to Pr(j), j fm, fs, im, is, e, se}. Growth, Employment, Skills and Female Labor Force 13 for men, but that decline was compensated by the 79. The presence of a formal employee in the household increase in employment probability in the formal service has a very large and significant negative impact on the sector. employment probability for men and women. It seems that a formally employed person in the household 75. Education beyond secondary school for women has generates substantial (social security) benefits to other a positive impact on the probability of being an employer members of the household so that it reduces the utility (Table 35). The impact is similar for men but the of getting a job in the labor market. The presence of a magnitude of marginal effects are relatively higher for formally employed person in the household has the men than for women. It is interesting that the marginal biggest negative impact on the probability of informal effect of vocational education on the probability of service employment for females (more than 20 percent), being an employer is slightly lower than the marginal and no impact on the probability of being a female effect of high school education. It seems that vocational employer. In the case of men, its biggest negative impact education is better suited to enhance the probability for is observed in the case of self-employment (close to 20 wage labor. percent). 76. Finally, the effect of education on the probability 80. Contrary to our a priori expectations, the probability of self-employment is in opposite directions for men of employment declines for women and especially for and women (Table 26). There is a monotonic relationship men if the household head is not employed. This could between the level of education and the probability of be due to the role of social networks in finding a job in self-employment for women, but there is a monotonic Turkey. negative relationship between these two variables for men. 81. The age-employment probability profiles are depicted in (for women) and (for men). Age profiles 77. Our findings show that education improves the are usually steeper for men than for women, i.e., the participation of women in all types of employment, but employment probabilities change faster by age for men the strongest effect is observed in the case of formal than they do for women. In spite of this difference, the services. However, more educated men are more likely age profile becomes negative at almost the same age to be employed as wage workers in formal manufacturing for men and women employed as wage worker in formal and services, and as employers, and they are less likely manufacturing (around 45-47 years) and formal services to be employed as wage workers in informal (around 52-53 years). In other words, there is no manufacturing and services, and as self-employed. difference in the age of “retirement” between women and men working in the formal sector, and those who 78. The effects of household characteristics on labor work in formal services are likely to work longer than market decisions are summarized in Table 37. As may those working in formal manufacturing. The marginal be expected, the effect of household size on the effect of age on self-employment and entrepreneurship employment probability of female parents is negative (employer) probabilities remain positive even for older possibly because of the bigger need for home production people. The probability of employment in informal in larger households. The household size has a very sectors for men is decreasing by age for men, i.e., older small positive impact on the employment probability men are less likely to work in the informal sector at all of female children. Men, both parents and children, in age levels. larger households have a stronger tendency to participate in the labor market. These findings indicate that parent 82. Finally, we estimated Mincerian wage equations to women are either more productive in home production shed light on the poverty implications of labor market than men, or there are cultural factors that consider participation decisions. Since the wage rate is observed home production as a feminine activity, so that parent for only those who work as wage workers, we used women tend to be involved in home production (tend multinomial logit sample selection model in estimating 24 to stay at home) in large households. the wage equations to eliminate the selection bias. 24 In estimating the wage equation, the sample for labor market participation is used, i.e., the data for all individuals living in urban areas with a positive wage. We did not estimate seperate wage equations for full time and part time workers because it would require modelling full time and part time jobs separately and the estimation results would suffer from the lack of sufficient number of observations for a number of labor market outcomes. Note that full time workers (individuals working at least 35 hours per week) account more than 90 percent of all workers in our sample. 14 Growth, Employment, Skills and Female Labor Force The wage equation is estimated, by taking into account of explanatory variables show the percentage change the selection process, for four categories of wage emp- in the wage rate when a dummy explanatory variable loyment (formal/informal and manufacturing/services) change from 0 to 1. for men and women separately. Since the reliable wage data are available for only 2006, the wage equations 85. There is an inverted U-shape relationship between were estimated for that year. The following variables, 26 wages and age. The wage rate increases up to a certain in addition to the sample selection correction terms, are age (on average, around 40s), than tends to decline. included into the model. There is a monotonic increase in the wage rate by firm size in services. In manufacturing, micro-firms pay 83. The (log) age and its square are used to estimate lower wages, but firm-size differential between small, the age-wage profile and to estimate the effects of medium-sized and large firms is negligible for women. age/experience on wages. All educational level dummies Those who work full time receive higher wages, but are included to estimate the effects of education on the effect of working time, after controlling for the full wages. There are a number of dummy variables for time status, is ambiguous. In some cases, those who occupations (Managers, Professionals, Technicians, work longer hours get lower wages. Clerks, Service workers, Skilled agricultural workers, Craft workers, Machine operators) where “Elementary 86. The effects of educational level on wages reveal a occupations” is the omitted occupation variable. The 27 number of policy-relevant phenomena. First, there occupation variables are likely to be endogenous in the seems to be no significant wage differential between wage equation. For this reason, all wage equations are illiterates and primary and secondary school graduates estimated without these variables as well. Dummy after controlling for the selection (labor market variables for firm size (small, 10-24 employees; medium- participation) effect. The importance of correcting for sized, 25-49 employees; large, 50 and more employees) selection becomes obvious when we compare the are added to test the effect of firm size. The micro-firm estimation results with selection control (Table 39) with (less than 10 employees) dummy is the omitted firm the OLS estimates without any selection control (Table size variable. The firm size is included in the model as 41). The OLS results suggest a monotonic increase in a proxy for unobserved labor and product market the wage rate by education, and significant returns to conditions. For example, workers in large firms are all, including primary and secondary, education. more likely to be organized in trade unions, and bargain for higher wages, i.e. firm size could be a proxy for However, once the selection effect is controlled for, the unionization. Moreover, large firms are more likely to positive effect of primary and secondary education on have stronger market power, and if they share their wages disappears, i.e., the estimation of the wage higher profits with their workers as a result of (Nash) equation without selection correction tends to bargaining, we would expect a positive correlation overestimate returns to education especially for primary between firm size and wages. The (log) working time and secondary education. per week and a dummy variable for full time employees are used to control for working time and employment 87. Second, vocational high school graduates earn much status on wages. The full time dummy is used to capture higher wages than illiterates, and primary and secondary the effects of compensating wage differentials between school graduates, and, even in many cases, high school full time and part time employees. The working time graduates. The informal sector does not pay higher is included to control for over time wage premium. 25 wages for men graduated from vocational schools than men graduated from high schools. 84. The descriptive statistics and estimation results are presented in Table 38-41. Since the dependent variable, 88. Third, the estimated returns to education increase the wage rate, is defined in log form, the coefficients when the occupation variables are omitted (cf. Table 25 The working time and firm size variables are likely to be endogenous. We estimated the wage models by dropping these variables, but there was no qualitative change in the estimation results for other variables. 26 The only exception is the formal service sector for men. In this sector, there is an unexpected U-shape relationship. Note that some of other estimation results for this sector (for example, the effects of education) are not reasonable. 27 The results indicate that there are not significant returns to education for informal females working in industry (column 1 of Table 39 and 40). However, this result is not reliable because of the lack of sufficient number of observations in high skill categories. There are only 25 (80) observations for “college” (“vocational school”) category. Growth, Employment, Skills and Female Labor Force 15 39 and Table 40). This is an expected outcome because rapidly after the 2001 Crisis. The average annual growth high wage paying occupations like managers, rate of urban employment was about 4 percent from professionals and technicians are expected to be filled 2002 to 2006, well above the population growth rate 29 by more educated people. which was 1.6 percent in the same period. There were almost 2 million new jobs created in the urban regions 7. Conclusions and policy implications in 4 years, from 2002 to 2006. These data show that the manufacturing industry and services in Turkey were 89. The Turkish economy achieved very high growth able to generate new jobs in the urban areas, but they rates after the 2001 Crisis. The average annual growth could not compensate for the rapid loss of employment 28 rate in GDP from 2002 to 2006 was about 7.5 percent. in agriculture. The estimates on employment elasticities The economy failed to generate any new job in 2001 suggest that the manufacturing industry and a number and 2002, and in spite of rapid growth in output from of service sectors have significant employment 2002 to 2006, the average annual growth rate of generation potential if they sustain high growth rates. employment remained at very low levels (about 1 percent) in the same period. Thus, the growth rate of 94. Although the employment generation performance employment lagged behind the growth rate of population, of the urban regions is respectable, our analysis point and, consequently, the employment ratio declined from to a number of problem areas that need to be tackled. 32.5 percent in 2000 to 30.6 percent in 2003, and The first major problem in urban employment is the recovered only a little afterwards (30.8 percent in 2006). fact that the employment ratio for urban women is extremely low. The service sector, especially “other 90. Although the aggregate data present a rather bleak services” (public administration, education, health, etc.), view on the employment performance of the Turkish provided a limited number of formal jobs for more economy since the 2001 Crisis, an analysis at the sectoral educated (college graduate) women, and manufacturing level offers a different view, and sheds light on potential and wholesale and retail trade, hotels and restaurants problem areas and relevant policy issues. were the sources of informal jobs for less educated women. An assessment for the changes in employment 91. Turkey had a very high rural population compared patterns of urban women and the labor market to other countries at the same level of development, participation decisions indicates that the “under- and the share of rural regions accounted almost half of total employment in 2000. Agriculture was the largest participation trap” could be a real problem for urban employment generating sector: the share of agriculture women in Turkey. As Booth and Coles (2007) suggest, in total employment was 36 percent in 2000. an imperfectly competitive labor market leads to under- participation in the labor market. Those people who 92. Turkey has experienced a very rapid reallocation have high home workplace ability and home productivity of labor from rural to urban regions. Agricultural will prefer to stay at home because they will not receive employment declined by 22 percent in only 6 years the full return to their investment in education if they from 2000 to 2006, i.e., 1.7 million people lost their work in the market. These workers, who are in the jobs in agriculture. Agricultural output, however, did under-participation trap, will further lower their ex ante not collapse. On the contrary, there was an increase in investment in education because the education they get agricultural output albeit at a very low level. This shows could be useful at the workplace, but not so much for 30 that there was a substantial underemployment in home production. agriculture and the rapid decline in rural population and agricultural employment did not cause a serious fall in 95. The educational system in Turkey is likely to agricultural output. exacerbate the under-participation trap because only a small minority of high school graduates can get access 93. Employment in urban regions increased rather to higher education. The value of education beyond 28 The annual growth rate of the economy dropped to 4.5 percent in 2007. 29 When the crisis years included, the average annual growth rate of urban employment becomes 2.6 percent from 2000 to 2006, still higher than the population growth rate. 30 It must be emphasized that the valuation of home production performed by female members of the family is based on economic as well as cultural factors. Growth, Employment, Skills and Female Labor Force 16 compulsory level for a woman in Turkey is a weighted construction, manufacturing, and wholesale and retail average of expected returns to high school and college trade, hotels and restaurants. Moreover, informality education, where weights are probabilities to complete seems to be a permanent status for less educated women a college education, and to find a job after education. (“informality trap”), and it substitutes for formal jobs Since the probability to complete a college education during economic crisis. There are substantial wage is low, the expected value of high school or vocational differentials between formal and informal employees, school education could be lower than the value of home and these wage differentials are possibly due to production for many women. For those women whose productivity differentials. Moreover, informal female home production is more valuable than a job in the employees are paid much lower than informal male market, there will not be any incentive to invest in employees, but the gender wage differentials diminish education beyond compulsory level because that in the formal sector. The government could pursue education is likely to have almost no contribution to policies to encourage firms to employ formally the home production. Thus, we will observe low investment young people, and to help less educated women, through in education and low participation in the market, as is active labor market policies, to move to formal jobs. the case in Turkey. Moreover, policies that could reduce the barriers against small firm growth would encourage not only the 96. There are a number of policies that may help urban generation of new jobs, but also the expansion of the women to overcome the under-participation trap. As formal sector. we have seen in equations 1 and 2, the relative prices of market products and services provided by home 98. Finally, the quality of jobs is as important as the production are important for the labor market quantity of jobs for poverty reduction. The share of participation decision. A reduction in the price of services “good jobs” in the urban sector has increased even provided by home production may increase the during the crisis years in Turkey. The increase in the 31 participation rates. In this context, Booth and Coles share of “good jobs” across all household size categories (2007) propose that optimal policy is to subsidize labor may partly explain the decline in poverty rates in Turkey market participation which might be efficiently targeted in the last 5 years. In spite of these encouraging as state-provided childcare support. Moreover, the extent developments, the share of “good jobs” is still extremely of informality reinforces the under-participation trap low (about 26 percent in 2006). One of the priorities of because informal jobs, by paying low wages, avoiding the public policy has to raise the share of “good jobs” employment security, and offering poor working that encourage higher participation rates, afford higher conditions, make workplace jobs less valuable than wages, and possibly stimulate the demand for skilled home production. labor through pecuniary externalities between human capital investment and R&D (Redding, 1996). The 97. The second major problem in urban employment analysis presented in this paper show that vocational is the extent of informality. Informality is widespread education in Turkey, in spite of its existing problems i) among less educated employees, especially among and shortcomings, helps to generate “good jobs” because less educated female employees, ii) among young those people with vocational training are more likely to employees, iii) in micro-establishments, and iv) in participate in the labor market and to get higher wages. 31 There was a decline only in the largest household category (those households with 10 or more people). The share of this category in total population was 2.6 percent. Growth, Employment, Skills and Female Labor Force 17 References Acemoglu, D. (2001), “Good Jobs versus Bad Jobs”, y Políticas Públicas, Campus Monterrey, Working Journal of Labor Economics , (19): 1-21. Papers No. 20072. Booth, A.L. and Coles, M. (2007), “A Microfoundation Engström, P., Holmlund, B. and Kolm, A. (2005), ” Tax for Increasing Returns in Human Capital Differentiation, Search Unemployment, and Home Accumulation and the Under-Participation Trap”, Production”, Oxford Economic Papers (57): 610- European Economic Review (51): 1661–1681. 633. Burdett, K. and Smith, E. (2002), “The Low Skill Trap”, Redding, S. (1996), “The Low-Skill, Low-Quality Trap: European Economic Review (46): 1439- 1451. Strategic Complementarities between Human Capital and R&D”, Economic Journal (106): 458- Cárdenas, M. and Villarreal, H.J. (2007), Inequality 470. Reduction through Self-employment under High Inflation Periods: The Mexican Experience, Snower, D.J. (1994), The Low-Skill, Bad-Job Trap, Escuela de Graduados en Administración Pública IMF Working Paper WP/94/83. Growth, Employment, Skills and Female Labor Force 18 Table 1: Population and employment, 2000-2006 Table 2 : Output and employment growth by sectors, 2001-2007 Growth, Employment, Skills and Female Labor Force 19 Table 3: Employment elasticities, 2001Q1-2007Q3 Table 4: Distribution of employment by region, 2000-2006 (percentage of total employment) Growth, Employment, Skills and Female Labor Force 20 Table 5: Distribution of employment by sector, 2000-2006 (percentage of total employment) Table 6: Distribution of employment by education level, 2000-2006 (percentage of total employment) Growth, Employment, Skills and Female Labor Force 21 Table 7: Distribution of employment by establishment size, 2000-2006 (percentage of total employment) Table 8: Distribution of employment by occupation, 2000-2006 Table 9: Distribution of employment by status, 2000-2006 (percentage of total employment) Growth, Employment, Skills and Female Labor Force 22 Table 10: Distribution of employment by registration status, 2000-2006 (percentage of total employment) Table 11: Average monthly wage rates in urban areas, 2006 Table 12: Sectoral composition of urban employment, 2000-2006 (percentage of total urban employment) Growth, Employment, Skills and Female Labor Force 23 Table 13: Share of formal employment in urban areas by sector, 2000-2006 (percentage of sectoral employment by gender) Table 14: Share of “good jobs” in urban areas by sector, 2000-2006 (percentage of sectoral employment by gender) Growth, Employment, Skills and Female Labor Force 24 Table 15: Composition of urban employment by establishment size, 2000-2006 (percentage of total urban employment) Table 16: Share of formal employment in urban areas by establishment size, 2000-2006 (percentage of employment by establishment size and gender) Table 17: Share of “good jobs” in urban areas by establishment size, 2000-2006 (percentage of employment by establishment size and gender) Growth, Employment, Skills and Female Labor Force 25 Table 18: Distribution of urban working age (15+) population by education, 2000-2006 (percentage of total working age population in urban areas) Table 19: Share of employees by education, 2000-2006 (percentage of population of working age in urban areas by education and gender) Growth, Employment, Skills and Female Labor Force 26 Table 20: Share of formal employees by education, 2000-2006 (percentage of employment in urban areas by education and gender) Table 21: Distribution of urban young (15-24 years old) population by education, 2000-2006 (percentage of total young population in urban areas) Growth, Employment, Skills and Female Labor Force 27 Table 22: Share of young (15-24 years old) employees by education, 2000-2006 (percentage of young population in urban areas by education and gender) Table 23: Share of formal young employees by education, 2000-2006 (percentage of young employment in urban areas by education and gender) Growth, Employment, Skills and Female Labor Force 28 Table 24: Distribution of urban polulation by household size, 2000-2006 (percentage of total urban population) Table 25: Share of people living in households with any employment, 2000-2006 (percentage of population by gender and household size, urban areas) Growth, Employment, Skills and Female Labor Force 29 Table 26: Share of people living in households with any formal employment, 2000-2006 (percentage of population by gender and household size, urban areas) Table 27: Share of people living in households with any “good jobs”, 2000-2006 (percentage of population by gender and household size, urban areas) Growth, Employment, Skills and Female Labor Force 30 Table 28: Descriptive statistics on variables used in the labor market participation model, (mean values, urban areas, working age population) Table 29: Estimated labor market outcome probabilities at mean values (percentage) (for an average illiterate married parent without any formal employee in the household) Growth, Employment, Skills and Female Labor Force 31 Table 30: Marginal effects of schooling on employment probability (base: illiterate) Table 31: Marginal effects of schooling on informal manufacturing employment probability (base: illiterate) Table 32: Marginal effects of schooling on informal services employment probability (base: illiterate) Growth, Employment, Skills and Female Labor Force 32 Table 33: Marginal effects of schooling on formal manufacturing employment probability (base: illiterate) Table 34: Marginal effects of schooling on formal services employment probability (base: illiterate) Table 35: Marginal effects of schooling on employer probability (base: illiterate) Growth, Employment, Skills and Female Labor Force 33 Table 36: Marginal effects of schooling on self-employment probability (base: illiterate) Table 37: Marginal effects of household characteristics on employment probability Growth, Employment, Skills and Female Labor Force 34 Table 38: Descriptive statistics for wage workers, urban regions, 2006 (mean values) Table 39: Determinants of urban wages, 2006 (multinomial logit selection model) Growth, Employment, Skills and Female Labor Force 35 Table 40: Determinants of urban wages, 2006 (multinomial logit selection model, no occupation variables) Table 41: Determinants of urban wages, 2006 (OLS estimates with no sample selection) Growth, Employment, Skills and Female Labor Force 36 Figure 1: Long-term economic growth cycles in Turkey (5-year moving average GDP growth rates) Source: Turkstat (new GDP series after 1998) Figure 2: GDP and employment growth rates, 1999-2007 Source: Turkstat (new GDP series) Growth, Employment, Skills and Female Labor Force 37 Figure 3: Output and employment growth, selected sectors, 2000Q4-2007Q3 Figure 4: Output and employment growth, selected sectors, 2000Q4-2007Q3 Growth, Employment, Skills and Female Labor Force 38 Figure 5: Age-employment probability profiles, female employees (percentage) Figure 6: Age-employment probability profiles, male employees (percentage) World Bank Copyright @ 2010 The International Bank for Reconstruction and Development The World Bank 1818 H Street, NW Washington, DC 20433, USA All rights reserved