98665 Understanding Out-of-Work and Out-of-School Youth in Europe and Central Asia Joseph Mauro and Sophie Mitra1 Fordham University July 30, 2015 1 The authors are very grateful for the insightful comments and help from Cesar A. Cancho, María E. Dávalos, Natasha De Andrade Falcao, Tu Chi Nguyen, Patrizia Luongo and Indhira Santos. This paper was done as background work for Europe and Central Asia: Jobs and shared prosperity (P148191), TTLs María E. Dávalos and Indhira Santos. . 1 Abstract Objectives: The objectives of this study are to describe and analyze the out-of-work and out-of- school youth (ages 15-24) in the Europe and Central Asia2 (ECA) region. People who are out-of- work and out-of-school are referred to as NEET (Not in Employment, Education or Training). This study attempts to characterize the NEETs by age, gender, education and their activity status. Methods: This study uses micro level data obtained from several surveys conducted from the early 2000s through 2011 to study the NEETs in three time periods: 2001-2002, pre-2009 and post- 2009. NEET rates for different age intervals, gender and educational attainment are investigated for the ECA region and countries within. Regression analysis is used. Results: The main findings of this study are listed below. First, we find that in the post-2009 period the youth NEET rate for the ECA region was 19.60%, which is higher than the OECD youth NEET rate of 16% in 2011 (OECD 2013). Second, this study finds that the NEET rate prior to the financial crisis in 2009 was on the decline, and increased in the post-2009 period. Third, this study finds that the NEET rate for ECA is higher for women than for men for all years. However, since the financial crisis in 2009, the gender gap has declined from 4.64 in pre-2009 to 2.75 percentage points in post-2009, suggesting that young men were more adversely affected by the recession than women. Forth, this study finds that in the ECA region youth males are more often classified as NEETs but active in the labor market, and youth females are more often classified as NEETs but inactive in the labor market. Fifth, using a linear probability model, this study finds that individuals who are 20-24 years of age, have a lower level of educational attainment and married females are more likely to be NEET. Also, individuals living in urban areas and with lower household sizes are less likely to be NEET. Sixth, another linear probability model was constructed using household budget surveys for six countries in ECA from 2009. The main finding from this model was that NEET youths tend to live in households with lower per capita consumption than their non-NEET counterparts. Seventh, there is an increase in the NEET (unemployed) rate after the crisis, while the NEET (inactive) rate stayed roughly constant. 2. The countries in the Europe and Central Asia (ECA) region include Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Estonia, Georgia, Hungary, Kazakhstan, Kosovo, Kyrgyz Republic, Latvia, Lithuania, the former Yugoslav Republic of Macedonia, Moldova, Montenegro, Poland, Romania, the Russian Federation, Serbia, the Slovak Republic, Slovenia, Tajikistan, Turkey, Turkmenistan, Ukraine, and Uzbekistan 2 1 Introduction and Background This report is an investigation of out-of-work and out-of-school youths in the Europe and Central Asia (ECA) region. We use the term out-of-work and out-of-school interchangeably with NEET (Not in Employment, Education or Training). A lack of data on the characteristics of out-of-work and out of school youths makes it difficult to develop appropriate policies. Furthermore, a high NEET population can have significant economic and social costs for a region due to lost productivity and misallocation of resources. With this motivation in mind, this report attempts to fill some of the knowledge gaps by analyzing the characteristics of the NEET population in ECA. Because limited research has been conducted on this issue in ECA so far, the background and literature review to follow also cover countries outside ECA, in Europe and OECD (Organization for Economic Co-operation and Development) countries in particular. 1.1 Employment Challenges Youth unemployment is a challenge in the ECA region, and more generally in Europe and in the OECD. Table 1 presents the youth unemployment rate (YUR), the general unemployment rate, and the ratio of the unemployment rate for youths compared to the general unemployment rate for countries in ECA as well as OECD countries outside the ECA region. Table 1 shows that the unemployment rates for youths are almost uniformly higher that the general unemployment rate, with the exception of Kazakhstan. The YUR is highest in FYR Macedonia at 53.9% and lowest in Kazakhstan at 3.9%. The YUR rates for ECA countries are almost always greater then OECD countries outside the ECA region, particularly those not in Europe. When comparing the ratio of the YUR to the unemployment rate for the whole labor force it can be seen that the YUR is almost double or more than double in all countries with the exception of Kazakhstan. This is not only the case within ECA but also for OECD countries outside of ECA, suggesting that youth unemployment is a serious problem globally but particularly within the ECA region. Youth unemployment has been of particular concern for countries in the ECA region. For example, in Bulgaria in 2000 the YUR was 33.7% compared to an unemployment rate (UR) of 14.5%. The problem was even worse in Croatia and Slovakia where the YUR were 37% and 37.3% respectively (Eurostat, 2013). Table A1 in the appendix gives the YUR and adult UR for select countries in ECA. Youth unemployment was seen to be declining slightly during the 2000s and did not start rising again until 2007 (ILO, 2012). Some credit this overall decline to be due to the excellent economic conditions in the middle of the last decade (Brenke, 2012). However, the financial crisis in 2009 and the recession that followed seem to have undone any progress that was seen earlier in the 2000s with youth unemployment and has had serious consequences on labor markets. This was a global crisis that hit all populations, but some argue that the group that was hurt the most by this was the youth population (15-24) (Verick, 2009). In 3 EU27, Figure 1 suggests that the YUR largely went up following the crisis and has continued to increase years afterwards, suggesting the recession has had a persisting negative effect on the youth labor market. Any prior gains made in reducing youth unemployment over the past decade vanished by the 2009 recession, particularly in ECA where the YUR has typically been higher than in the rest of Europe. The 2007 World Development Report puts the YUR for ECA at an average of 25% in 2006. Since the recession, a great amount of research has been done on the impact of the 2008 financial crisis on youths. The onset of the financial crisis in late 2008 had an immediate impact on young people (Verick, 2009). This is mainly due to their lack of work experience and high concentration in areas highly sensitive to business cycles (Caroleo and Pasotre, 2007, Eurofound, 2012). Furthermore, the job prospects were at an all-time low for new graduates. Even when the recession officially ended in much of Europe in 2009, youths were still feeling the effects. In 2011, the YUR in the EU was 33.6% for 2011, 22.4% for 2012, with rates greatly varying across Europe (Eurofound, 2012). There has also been some research on the impact of financial crises in general without limiting oneself to the 2008 financial crisis. Choudry et al. (2012) find that a financial crisis has a significant and robust impact on the YUR, even after controlling for GDP. Choudry et al. also find that the persistence of the negative effects for labor market will affect the YUR for five years after the onset of the crisis, with the worst period of youth unemployment occurring in the second and third years after the onset. Verick (2009) considers the impact of financial crises throughout time and finds that YUR for men and women are higher than the unemployment rate for prime aged men. According to Verick, it is the youth who are hit the hardest during and after financial crisis. Verick also finds that the YUR will continue to rise and remain so for some time to come, suggesting that the increased gap between the unemployed youth and unemployed adults will remain, something that has been seen. Finally, several studies attempt to find determinants of unemployment among youths in Europe and in selected ECA countries. Bell and Blanchflower (2011) used data on OECD countries, including some from the ECA region such as Turkey, Romania, Slovenia, and Poland to construct a probit model to estimate the probability of being unemployed. A key result from Bell and Blanchflower was that the probability of being unemployed for youths was significantly higher than that of being unemployed at an older age. Bell and Blanchflower use unemployment data for youths aged 15-24 and adults aged 25-64, from 1970-2009 in OECD countries to estimate a relationship between the YUR and the adult UR. They find that YUR changes by 1.79 percentage point for every percentage point change in the adult rate. This suggests that there is a higher degree of cyclical sensitivity for youths compared to their adult counter parts, a result confirmed by the OECD (Bell and Blanchflower, 2011, OECD, 2008). Also found by Bell and Blanchflower is that 4 unemployment for youths lasts longer and that young workers who experience unemployment spells at a young age are more likely to experience more periods of unemployment later in life. This result is consistent with that of Perugini and Signorelli (2010) who analyze youth labor market performance in the EU15 and four countries from Central and Eastern Europe (Poland, Hungary, Czech Republic, and Slovak Republic) using econometric techniques. They find that there is a strong persistence over time of youth labor market performance. This means that if weaknesses such as youth unemployment are left unsolved, they can endure for a long period of time. However, this result is symmetric in the sense that a successful youth labor market policy could also endure for a long period of time. Finally, limited research is available on the demographic characteristics of unemployed youths. While unemployment rates for young men were higher than for young women in the first quarter of 2008, they converged as time went on with men only slightly higher in the EU27 (Eurofound, 2012). Also, Bell and Blanchflower suggest that those youths with the lowest levels of education and skills are particularly at risk for unemployment, especially ethnic minorities. However as time has gone by, as per Eurofound (2011), it appears that in some Eastern European countries, having completed a tertiary level of education no longer lowers the risk of unemployment compared to lower levels of education (Eurofound, 2011). Employment challenges in the region do not end with unemployment, but extent toward inactivity as well. Table 2 presents inactivity rates for the total labor force, youths aged 15-24, and males and female aged 15-24 for the EU27 and select countries in ECA. The first thing to note is that the inactivity rate for youths is almost double that for the total labor force. It also should be noted that inactivity rates for ECA countries are strictly above the EU27 average. The final take away from table 2 is that females aged 15-24 always have a higher inactivity rate than their male counterparts. This suggests that youth inactivity is a major problem in the ECA region, especially for young females. 1.2 Idle Youths in ECA Using the traditional unemployment rate and inactivity rate as measures of labor market performance for youths does not provide all that insightful a picture of how youths are doing in the labor market. The traditional unemployment rate only identifies the share of unemployed out of total the number of employed and unemployed, where the unemployed are not employed but seeking work. Youths aged 15-24 are often in school or training and do not seek work, and thus are not counted as part of the labor force and in the unemployment rate. Instead of the unemployment rate, a more inclusive and perhaps more useful measure of how youths are doing in the economy is to count the share of those who are Not in Employment, Education or Training (NEET) out of the entire youth population. The NEET group has been studied primarily in Europe, 5 Japan and in South America. Different definitions have been used for the NEET depending upon the questions and countries being researched. Sometimes NEET only include those who are single, or do not own homes or have a particular educational background (Genda, 2007, Gaston and Kishi, 2005). Some include in NEET those who are ages 15-29 (Marshall, 2012, OECD, 2013). More often than not however, NEETs refer to those individuals who are 15-24, have all educational backgrounds, and can be single or married (Eurofound, 2011). This is also the definition used by the ILO and is the definition adopted in this paper. In 2011, 7.5 million people were considered NEET in the European Union, which increases by another 6.5 million if NEETs are expanded to include 25-29 year olds (Eurofound, 2012). Table 3 presents youth NEET rates using Eurostat data for select countries in ECA and in the EU27 region. On average, the provisional NEET rate for the EU27 in 2013 was 12.9 %. However, many countries in ECA are at or above this rate. The NEET rate is at its highest in Turkey and Macedonia at 28.7% and 24.8% respectively and is at a minimum of 9.1% in the Czech Republic. Like the YUR, the NEET rate in the EU was particularly high at the beginning of 2000 at about 13% for the EU (ILO, 2012). By 2008, NEET rates for Europe were 11% for 15-24 year olds and 17% for 25-29 year olds. However, according to Eurostat data, after the financial crisis rates raised to 13% and 20% for the two age groups, respectively. So the NEET rate in general increased by 2-3 percentage points, depending upon the age group under consideration. There is a lot of variation though in Europe itself, with rates as low as 7% in Luxembourg and the Netherlands to over 17% in countries like Bulgaria and Turkey. For comparison, the youth NEET rates in 2010 for 15-24 year olds in the US, Japan and New Zealand were 15.6%, 9.7% and 13.7% respectively (ILO, 2012). In 2011, the NEET rate on average was 12.8% for OECD (ILO, 2012). In 2011, the NEET rate in Canada was 13% for youth ages 15-293 (Marshall, 2012). Having a large share of idle youths poses several important economic and social challenges, which makes it an important issue to understand. The economic cost associated with a high youth NEET rate can be substantial. This economic cost includes, at least, the foregone earnings and human capital of having youths out of the labor market and out of school, as well as the resources that are sometimes dedicated to support income or living conditions for this group. In 2011, the conservative loss due to NEET Youths in Europe was about €153 billion – about 1.2% of GDP. The economic cost also varies greatly by country, and is above 2% of GDP in some ECA countries (Eurofound, 2012). There are also social ills associated with high youth NEET rates. Walther and Pohl (2005) provide numerous examples of this. In Bulgaria, Romania, the Slovak Republic and Slovenia for instance, youths who were interviewed reported a “feeling of abandonment by the government” and a lack 3 According to Eurofound (2012), the NEET rates for those aged 25-29 years and for those aged 15-24 years are highly correlated (0.88) and nations that have the highest level of NEETs ages 15-24 also have the highest level of NEETs ages 25-29. 6 of eligibility for benefit entitlements as major factors for leaving the labor force. Similar sentiments were held in Spain, the UK and Poland. In Greece, Italy and Denmark, NEET rates were attributed to a lack of trust in the government or in government employment services. Pemberton (2008) notes that the lack of trust in public employment services was noted as the prime influence on NEET in Finland, Portugal and Austria. The 2013 World Development Report also addresses some social issues that are linked with youth unemployment. For all age groups, high rates of unemployment are linked with low levels of civic engagement and trust in government. In young people however, high rates of unemployment has also been linked with unrest and violence, such as the riots in the United Kingdom in 2011 (World Bank 2012). Social ills relate to more than just unemployment, but also inequality of job opportunities. The NEET population has been studied throughout the EU, and many OECD countries, but little research has been done on the ECA region and countries. There is limited knowledge beyond general NEET rates for ECA countries and no general profile for the ECA region as whole. This report will attempt to address this gap in the literature. 1.3 Objectives of the Study Given the limited literature on idle youths in ECA, this report aims to fill certain knowledge gaps. In particular, the objectives of this study are to: (i) Describe and analyze the NEET youth in ECA (i.e. Who are they? What are their characteristics? What are their families like? Where are they?); (ii) Identify policy lessons from international experience in incorporating youth into education or productive employment; and (iii) Highlight knowledge gaps in the literature. This report is structure as follows. Section 2 includes a detailed literature review so as to identify possible policy lessons based on international experience. Section 3 presents our data and methodology. Section 4 gives the results of the analysis. Section 5 includes a discussion of the results and recommendations for policy and research. 2 Literature Review This section reviews prior studies on NEET youths in general, and in ECA where possible. 2.1 Characteristics of NEET Youth This section attempts to summarize what is known on the characteristics of NEET rates in Europe and in ECA when possible. While the groups that compose the NEET population often vary greatly across countries, there are some traits of the NEET population that remain constant. Eurofound 7 (2012) reports that NEETs can be broken down into two main age groups, those ages 15-19 and those 20-24. In the EU, the NEET rate for those 20-24 years old was almost unanimously higher than that of 15-19 year olds for member nations. For the EU, the average NEET rate in 2011 was 6.9% for 15-19 year olds and 18.2% among 20-24 year olds. Also, countries that had the highest NEET rate for 15-19 year olds also had the highest rates for 20-24 year olds. NEETs can also be analyzed by educational attainment, although it is not as simple to do as with age. In most countries in the European Union (EU), more than 90% of NEETs ages 15-24 are composed of those who have less than a tertiary level of education (Eurofound, 2012). The 90% is equally split between those with a lower secondary level of education or less, on the one hand, and an upper secondary to post-secondary level of education, on the other. Those that have a tertiary level of education typically make up a very small percentage of the NEET population, except in the UK, Cyprus and Luxembourg where more than 10% of the NEET population has a tertiary level of education. So, the NEET population as a whole is composed of youth who have lower levels of educational attainment (Eurofound, 2012). Furthermore, the recent recession has increased the risk for individuals with higher levels of education to become NEET (Eurofound 2012, ILO 2012)4. Another way of trying to characterize the NEETs is through their labor force participation status. The ILO defines a person as inactive when they have not worked for a least one-hour in the survey period and were not available for work/ did not look for a job. For the NEET population in 2011, the NEET population in Europe is split about evenly between those who are unemployed (51.2%) and those who are inactive (48.8%). Countries such as Bulgaria, Hungary and Romania, with high general unemployment rates, have a NEET population that is predominantly inactive (Eurofound, 2012). To further breakdown the labor force status of the NEET population, one can also consider the percentage of discouraged workers among NEETs5. In 2011, the share of inactive NEETs in the EU who were discouraged was 57%. Also, the countries with the highest proportion of NEETs also have the greatest share of inactive NEETs and highest proportion of discouraged workers. Finally, the NEET population in Europe contains 52% who have never worked (Eurofound, 2011). NEET rates also vary by gender. According to Eurofound (2012), in 2000, in Europe the average NEET rate for young men was 11.5% compared to 14.9% for young women. This result does not hold throughout all of European countries, however. In the Northern and Baltic states, the NEET rate is higher for men than for women, while the opposite is generally true for continental and eastern European countries (Eurofound, 2012). 4 Characterizing idles youths by educational attainment has also been done in Latin America. Cardenas, Hoyos and Szekely (2011) find that among idle youths ages 15-18, 51% did not complete primary school, 33% did not finish secondary school and 13% never completed post-secondary school. 5 Discouraged workers are workers who are available for work but did not seek employment (so cannot be labeled as unemployed) because they felt there was no work available to them (ILO 2009). 8 Gender differences can also vary by age. For those ages 15-19, the male NEET rate is higher than the female NEET rate, 7.3% and 6.6% respectively in 2011. This was the average result for Europe, with exceptions in Bulgaria, Cyprus, Hungary, Latvia and Romania. For the 20-24 year old population, the NEET rate in Europe for women was higher than for men, 19.4% and 17.1% respectively. This did not hold in nine countries however, including Bulgaria, Latvia, Lithuania and Slovenia. So overall, trying to find gender patterns is very difficult as the relative NEET rates vary greatly throughout Europe. The gender gap in NEET rates in Europe was reduced from three percentage points in 2000 to less than one percentage point in 2011 (12.5% for men and 13.4% for women) (Eurofound, 2011). So, as time has passed the NEET rate for men in Europe has increased while it has decreased on average for women. The ILO (2012) also investigated gender trends for the NEET. They also find that the gender gap has converged greatly over the past decade, where the rise in NEET rates in 2010 for young men was 2.6% compared to 1.1% for young women. A significant take away from this shows the stronger impact the recent recession may have had on male youth, in comparison to female youth (ILO, 2012). There are some other important characteristics of NEET in Europe. Eurofound (2011) finds that young people with a disability are 40% more likely to become NEET compared to others. Also, if a young person is an immigrant, he or she is 70% more likely to become NEET. Pemberton (2008) also notes that there is a racial dimension to NEET but to our knowledge, no NEET rate is available by race in Europe. Finally, Eurofound (2012) states that living in remote areas increases a young person’s probability of becoming NEET by up to 1.5 times. In other words, youth living in rural areas are more likely to be NEET than their urban counterparts. It should also be noted that studies have also been done to characterize the NEET population outside Europe, for instance in Japan (Genda 2007)6, Canada (Marshall 2012)7, and Latin America (Cardenas, Hoyos and Szekely)8. Given the lack of evidence on NEETs characteristics in ECA, 6 Genda (2007) uses a multinomial logistic regression to describe the NEET population in Japan. Genda uses microdata available from the Employment Status Surveys of 1992, 1997 and 2002. Genda’s regression results showed that Japanese NEETs were more likely to be young people whose expected returns from working were low, mainly females and the less educated. Genda also found that young, less-educated men from poor families were becoming more likely to lose interest in work and become NEET compared to those in middle-income families. 7 Marshall (2012) deals with NEETs in Canada. Using the Labor Force Survey for 2011, Marshall finds that the Canadian NEET youths aged 15-29 is equally divided between men and women. Also most of the NEETs are older youths– ages 20 to 29. Marshall finds however, that the older male NEET population considers themselves unemployed, while older female NEETs are more likely to be out of the labor force. So the type of NEET inactivity varies by gender in the context of Canada. 8 Cardenas, Hoyos and Szekely (2011) attempt to characterize and understand idles youths in Latin America. In 2009, the average NEET rate of youths 15-24 in Latin America was 22.4% (Cardenas et al. 2011). Cardenas et al. also find that on average in Latin America that about 75% of the idle youth population is made up of the bottom three income quintiles. Estimating a probit model on the likelihood of being considered and idle (NEET) youth, they 9 one of the objectives of this study is to describe this group in this region and compare it to NEETs in other parts of the world. 2.2 On the underlying causes of the idle youth phenomenon It is important to try and identify the causes of the NEET phenomenon and in particular, how youths become NEET. By understanding factors and attributes that cause youths to become NEET, policy options may be developed to address specific factors. There are several different categories of causes of the NEET youth phenomenon at both micro and macro levels and in relation to education and employment. First, micro level factors are reviewed. A youth may become NEET due to their family socioeconomic situation. Youths in families where adults are unemployed or earn low income may be pressed to leave school and look for jobs. For instance, Eurofound (2011) found that young people who have a lower household income are more likely to become NEET in comparison to middle-income families. Also, having parents that have been unemployed at some point increases the likelihood of a young person being NEET by 17%. The marital status of young persons’ parents also comes into play, as a young person with divorced parents is 30% more likely to be NEET. Educational factors causing NEET can range from those dealing with school-dropouts, an individual or family’s prioritization of academics and the educational system as a whole. Bynner et al. (2000) and Bynner and Parsons (2002) used birth studies to find a set of education related factors that increase the probability of NEET status. They find that the educational level of the parents impacts the probability of being NEET, with those parents with a lower level of education being more likely to have children being NEET. Another key factor is the level of emphasis a parent places on their child’s education. Strelitz (2004) and Rennison et al. (2005) go deeper into the investigation of the role of the parents’ level of education and their interest in their child’s education. Strelitz finds that parents who have a higher education know how to better meet their child’s needs and guide them, thus reducing their probability of becoming NEET. Rennison et al. find that people who are NEET often have parents who know very little about the educational system. They also find an inverse relationship between parents’ attendance at parents’ nights at schools and NEET status. One of education’s primary purposes is to ready a young person for employment. Breen (2005) uses data on OECD countries from 1995-1999 to estimate a model of youth unemployment based on education signaling and employment protections. Breen finds that schools that offer vocational training have a lower level of youth unemployment than countries that do not have vocational find that household income, education of the household head, age of the household head, employment status of the household head, gender of the household head and household size are micro factors associated with the probability of belonging to idle youth. They find that with higher household income and with an employed household head, the risk a youth has for becoming idle is lower. 10 education. Similar results are in Eurofound (2012). Using a pooled OLS model with country fixed effects, Eurofound finds that increasing the proportion of upper-secondary schools that attended a dual education system by one percentage point decreased NEET rates by 0.04-0.09 percentage point. These results make sense, as having some type of vocational education signal employers about the greater qualifications and higher expected productivity of young people (Breen 2005). It should be noted that the success vocational education may be due to correlations and their impact has not been rigorously evaluated. Certain groups may face specific challenges in staying in school or accessing jobs. For instance, youths with disabilities or immigrant youths may not find school programs that are adapted to potential special needs or language skills. They may also face barriers to employment that the general population does not face. This may then explain why, as stated previously, those with disabilities or immigrants were found more likely to be NEETs (Eurofound 2011). Minority groups may also face special challenges staying in school and barriers to employment. Pemberton (2008) and Quintini and Manfredi (2009) both see race and ethnicity as influencing the probability of a young person becoming NEET. Inactivity accounts for a large part of the NEET perhaps due to discouraged workers, i.e. youths who stop looking for work because they think that their probability to find work is limited and thus drop out of the labor force (ILO 2012). Bell and Blanchflower (2011) find that for younger aged groups (16-24) participation rates are negatively correlated with unemployment rates during the Great Recession. This suggests a higher presence of discouraged workers in countries with already high youth unemployment. The larger inactivity rate for women may be due to child-rearing activities and other household responsibilities. Elder, Novkovska and Krsteva (2013) find that in FYR Macedonia, after educational reasons, the second most common reason for youths (15-24) to be considered inactive was due to family responsibilities or household chores. This was especially the case for young women in Macedonia where 22.7% of young women listed family responsibilities and household chores as their reason for being inactive compared to only 1.8% of young men. Also, young women in ECA have children at younger ages compared with women in developed countries which also helps to explain the low labor market participation rates for women ages 16-24 in the region (Sattar 2012). There are of course some macro-level causes for the NEET phenomenon. The size of the youth NEET group is first expected to be affected by population growth. Eurofound (2012) finds that the larger the youth cohort size is relative to the working-age population, the higher the NEET rate among young people. In addition, the size of the youth NEET group is expected to be affected by the labor market’s overall performance. In countries with high unemployment rates and low labor force participation rates, one expects to find larger youth NEET groups. 11 In addition, employment protection9 policies and legislations can affect youth integration into the labor force. The argument is that employers in countries with a high degree of employment protection will be weary of hiring workers who lack work experience – namely youth (Eurofound, 2012). In a study done by Breen (2005), it was found that countries that had a higher degree of employment protection had a higher level of youth unemployment. Likewise, Eurofound (2012), using a pooled OLS, random effects and fixed effects models, finds that countries that have a higher degree of employment protection have a higher youth NEET rate – especially when regulations apply to temporary jobs. As part of employment protection, some specific mechanisms can be identified as perhaps hindering or promoting youth employment. That is the case for minimum wage protection. Although one may first consider the minimum wage as benefiting young workers, Neumark and Wascher (2004) find that in OECD countries, higher levels of minimum wage negatively impact youth employment. They find that the level at which minimum wage effects youth employment differs depending upon the fixed effects they control for, but an increase in the minimum wage ratio by one percent can decrease youth employment by about 0.15 percentage points. Eurofound (2012) finds that increasing the relative minimum wage by one percentage point increases the youth NEET rate by 0.06-0.07 percentage point. Active labor market policies (ALMPs) can be used to lower rates of NEET. This is what a study by Eurofound (2012) found. More precisely they find that increasing ALMP expenditures per unemployed worker by one percentage point of GDP per member of the labor force lowers the overall youth NEET risk by 0.15-percentage points. 3 Data and Methodology The data used in this study is microlevel data taken from several surveys done in 20 countries in the ECA region 10 . In particular, the following surveys were used: the Living Standard Measurement Survey (LSMS), the Integrated Living Conditions Survey (ILCS), Statistic on Income and Living Conditions (SILC), the Labor Force Survey (LFS), and the Household Budget Survey (HBS). Due to the use of multiple surveys it must be noted that comparisons among numbers should be made cautiously. When possible, data for each country was used at three points in time: from the early 2000s, just prior to the financial crisis pre-2009 and the most recent year available after the financial crisis post-2009. Also, countries will be presented in the following sub-regions when possible Moldova and Ukraine (BMU), Central Asia, EU11, Russian Federation, South Caucasus, South Eastern Europe and Turkey11. This allows for comparison within the ECA 9 Employment protection refers to policies that limit an employer’s ability to dismiss an employee as well as minimum wage, unions and other labor market institutions aimed at protecting the rights of the employee. 10 Albania, Armenia, Bulgaria, Czech Republic, Estonia, Hungary, Kyrgyz Republic, Latvia, Lithuania, Macedonia, Moldova, Poland, Romania, Russia, Serbia, Slovak Republic, Slovenia, Tajikistan, Turkey, Ukraine. 11 Countries in the BMU are Moldova and Ukraine. Countries in Central Asia are Kyrgyz Republic and Tajikistan. Countries within the EU11 are Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic and Slovenia. The Russian Federation contains Russia. The South Caucasus includes Armenia. It is 12 region. With the HBS, additional household information such as per capita expenditures, receipt of public or private transfers was used for six selected countries in ECA12. A complete list of countries studied, along with their respective survey type and year, is available in Appendix Table 2. All estimates are calculated using sample weights from the included surveys. The data is used to determine NEET rates for youths for the selected countries in ECA, compared to NEET rates for other adult age groups. We then examine NEET rates by gender and educational attainment level. When data is available, we examine NEET rate trends for these groups. The labor force participation is also examined for the NEET populations in ECA. We then present descriptive statistics on the NEET youth, using the most recent year available for each country. Descriptive statistics for non-NEET youth are also given for comparison purposes. All ECA estimates are simple averages of the 20 country estimates for each statistic of interest. Finally, two linear probability models were used to investigate the probability of being NEET among youths. For the first linear probability model, the control variables include age, gender, marital status, household size, lives in urban area, number of people over 65 in the household and educational attainment. In particular, teen is a dummy variable equal to 1 if the individual is in the 15-19 years old age group, 0 otherwise. Female is a dummy variable equal to 1 if the individual is female. Married is a dummy variable equal to 1 if the individual is married. We also include an interaction term in order to capture the impact of being married on females compared to males. Urban is a dummy variable equal to 1 if the individual lives in an urban area. For education, dummy variables were created for each level of educational attainment and those with a primary or less level of education were omitted as a reference group13. The second linear probability model makes use of the HBS data, and controls for the characteristics included in the first model as well as several additional household level variables, including the gender and marital status of the household head, the log of total per capita expenditures (measured in USD), and whether the household receives public and private transfers14. In two countries (Poland and Tajikistan) we are able to control for the share of household members who are employed, aged 0-6, 7-15 and 65+. Interaction terms are included to capture the impact of being married (as in the first model) and of the share of household members who are ages 0-6 and 7-15 on females vs. men. 4 Results important to note that for some years data from certain countries are missing and thus are omitted from the sub- regional analysis. 12 Household information was available for Poland, Russia, Serbia, Tajikistan, Turkey and Ukraine, all from 2009. 13 It should be noted that educational attainment is associated with age. For example, tertiary education is irrelevant for the younger people in the 15-24 age group. However, the primary goal is to compare NEETs and non-NEETs, so this association has little impact on our findings. 14 Total per capita expenditures exclude health expenditures, rent and purchases of durables. 13 4.1 Understanding How Youth do Vis a Vis other Age Groups in Terms of Being NEET and non-NEET. Table 4 presents the rate of individuals classified as NEET in each country and an average NEET rate is also given for the ECA region. The following age grouping is considered: working age individuals (15-64 years old), youth (15-24) and prime aged workers (25-49). When possible, NEET rates for each group are given for three points in time: 2001-2002, pre-2009, post-2009. For the entire working age population, post-2009, the average NEET rate for the ECA region stood at 29.41%. Post-2009, the NEET rates for the entire working age population range from a low of 8.35% in Russia to a high of 43.45% in Macedonia. In the early 2000s, the average NEET rate for the region was 30.67%, which decreased to 28.45% pre- 2009 and rose to 29.41% post-2009. This dip in NEET rates prior to the financial crisis is seen in most but not all ECA countries. In Macedonia and Turkey, the NEET rates have declined steadily throughout. It should be noted that these two countries have among the highest NEET rates in ECA. For youths, post-2009, the average NEET rate for the ECA region was 19.60%. The NEET rate for youths ranged from 6.41% in Slovenia to a high of 39.76% in Tajikistan. The average NEET rate for the region decreased from 21.60% in 2001-2002 to 18.57% prior to 2009 and then increased to 19.60% post 2009. Again, this is not the case in Macedonia and Turkey. Post-2009, the average NEET rate for prime aged workers for the region was 26.42%. The NEET rate for prime aged workers is at its lowest in the post 2009 era in Russia at 9.24%, and at its highest in Tajikistan at 44.82%. This rate declined from 25.93% in 2001-2002 to 24.34% pre-2009, and increased to 26.42% post-2009. So we get a similar dip in NEET rates prior to 2009 and an increase afterward. Comparing the three different age groups presented in Table 4, the NEET rates for prime age workers are typically higher when compared to youth. This is expected because in general, youths tend to be in school, while prime age workers are not. Likewise, NEET rates for the entire working age population aged 15 to 64 are higher than those for prime aged workers given that the former may reflect early retirements among older workers aged 50 and above. Table 5 presents the rate of individuals classified as NEET in each sub-region in ECA and an average NEET rate is also given for the ECA region. This is done for the following age groupings: working age individuals (15-64 years old), youth (15-24) and prime aged workers (25-49). When possible, NEET rates for each group are given for three points in time: 2001-2002, pre-2009, post- 2009. For the entire working age population in the post-2009 period, the average NEET rate for the BMU, Central Asia, South Eastern Europe and Turkey were all higher than the ECA region 14 average with the following respective rates: 34.82%, 33.05%, 32.14% and 43.23%. A similar result is seen in the pre-2009 time period as well. This result does not only hold for the whole working age population but particularly for the youth population. Figure 2 presents Youth NEET rates for two age groups: youths aged 15-19 and 20-24, in 2001- 2002, pre-2009 and post-2009. The NEET rates for those aged 20-24 tend to be two to three times greater than those of youths aged 15-19. The average NEET rate for the ECA region in the post- 2009 period is 26.33% for those who are 20-24 and 11.16% for those who are 15-19, as seen in Appendix Table 3. The NEET rate for those aged 20-24 in the post-2009 period is highest in Tajikistan and Turkey, where the NEET rates are 54.68% and 41.02% respectively. Bulgaria has the highest NEET rates for those ages 15-19 with the rate equal to 26.65%. For the 15-19 and 20- 24 age groups, the NEET rate is lowest in Slovenia, where it stands at 3.21% and 8.39% respectively. For those aged 15-19, the average NEET rate in ECA was 13.70% in the 2001-2002 period, with the rate decreasing to 12.21% pre-2009 and declining again to 11.16% post-2009. For youths aged 20-24, the average NEET rate in ECA was 29.95% in the 2001-2002 period, which decreased to 24.17% pre-2009, and then increased again to 26.33% post-2009. So, for those ages 15-19, the NEET rate has on average been declining since 2001. This result is found in 5 countries in the ECA region (Albania, Hungary, Macedonia, Turkey and Ukraine) and 10 countries have a dip in their rates pre-2009 and then an increase afterwards. The NEET rate for 20-24 year olds exhibits a dip prior to the recession and an increase post-2009. This result holds in 12 out of 17 countries. Table 6 gives the NEET Youth Rates for 15-19 year olds and 20-24 year olds by ECA sub-region over time. It can be seen that in every sub-region and time period the NEET rates are almost double for 20-24 year olds when compared to 15-19 year olds. When comparing across regions, it can be seen that in the post-2009 period, that rates are highest in Turkey and Central Asia for both 15-19 year olds (21.68% for Turkey and 17.70% for Central Asia) and 20-24 year olds (41.02% for Turkey and 39.09% for Central Asia). Across all time periods Central Asia, South Caucasus, South Eastern Europe and Turkey all have higher NEET rates then the ECA region average for both age groups. The EU11 and Russian Federation sub-regions have the lowest rates in ECA for both groups, both below ECA average in every time period. Figure 3 presents the post-2009 period NEET rates separately for men and women. Appendix Table 4 gives country data by gender for 2001-2002 and pre-2009. In the post-2009 era, the average NEET rates for ECA are 18.18% for men and 20.93% for women. The gender gap between women and men in Turkey and Tajikistan is particularly large, with the NEET rate for women at 43.25% and 17.23% for men in Turkey and the NEET rate for women at 50.45% and 27.48% for men in Tajikistan. The NEET rate for women in the post-2009 period is highest in Tajikistan and Turkey. The NEET rate for men is highest in Moldova at 31.01%. Young men have a higher NEET rate 15 than young women in the Czech Republic, Estonia, Lithuania, Moldova, Russia, Serbia, the Slovak Republic and Slovenia. Figures 4 and 5 presents NEET rates separately for men and women over the sample period. The average NEET rates for men in ECA were 19.38% during 2001-2002, 16.23% pre-2009 and then increased to 18.18% post-2009. For women, the average NEET rates in ECA were 23.07% in 2001-2002, 20.87% pre-2009 and 20.93% post-2009. Thus, over time, NEET rates for young men exhibit a U-shaped trend, with the rates declining from the early 2000s to right before 2009, and then increasing after 2009. This is seen in 11 countries in ECA. The average women’s NEET rate also has a u-shaped trend since 2000, but the increase in women’s NEET rates was much smaller compared to that of men. This u-shaped trend is found in seven countries. It is interesting to note that the gender gap seems to have declined after the recession, with NEET rates for men increasing more than the NEET rates for women. Table 7 gives the Youth (15-24) NEET rates for males and females over time, analyzed at the sub- region level. When comparing gender rates across sub-regions, some interesting results arise. First off, in the BMU region the male youth NEET rate is higher than the female youth NEET rate for each time period. This is also true for the Russian Federation, EU11 and South Eastern Europe in the Post-2009 period and the South Caucasus sub-region in the Pre-2009 period. Both the BMU and EU11 sub-regions exhibit the U-shaped trend over time for both males and females. Rates for men and women are in constant decline over time in both the South Eastern Europe and Turkey sub-regions of ECA. Tables 8 and 9 continue to analyze NEET rates in the ECA region by gender, but by breaking down the youth population into 15-19 year olds and 20-24 year olds. In 2001-2002, the ECA region average for 15-19 year old male and females was 12.78% and 14.62% respectively. The male youth (15-19) rate in 2001-2002 was at a maximum in Albania at 23.14% and was at a minimum in Slovenia at 4.01%. The female youth (15-19) rate in 2001-2002 was highest in Turkey at 47.38% and lowest in Slovenia at 3.60%. In eight of 14 countries, the female youth (15-24) NEET rate was higher then the male youth NEET rate. In the pre-2009 period, the ECA region average for 15-19 year old males and females was 12.01% and 12.55%. The male youth (15-19) rate in the pre-2009 period was highest in Armenia at 46.24% and lowest in Slovenia at 2.10%. The female youth (15- 19) rate in the pre-2009 period was at a maximum in Turkey at 40.80% and was at a minimum in Lithuania at 0.62%. In nine out of 17 countries, the female youth (15-19) NEET rate was greater than the male rate in the pre-2009 period. In the post-2009 period, the ECA region average male and female youth (15-19) rates were 11.28% and 11.37%. The male youth (15-19) NEET rate was highest in Bulgaria at 23.32% and lowest in Slovenia at 4.47%. The female youth (15-19) NEET rate was greatest in Turkey at 30.67% and lowest in Slovenia at 1.88%. In five out of 15 countries the female youth (15-19) NEET rate was higher than the male youth (15-19) rate. Looking at the average ECA numbers, the NEET rates for male and female youths (15-19) are in decline over the 16 period. Also, NEET rates between males and females are very close and there are only large gender gaps in a few countries such as Turkey. Table 9 looks at the 20-24 year old youth population. In the 2001-2002 period, the average ECA NEET rate for 20-24 year old males and females was 26.48% and 33.10% respectively. In 12 of 14 countries, the female youth (20-24) NEET rate was higher than the male youth (20-24) rate. In the pre-2009 period, the ECA region average NEET rate for 20-24 year old males and females was 19.91% and 28.28% respectively. In the post-2009 period, ECA’s average NEET rate for male and female youths (20-24) was 23.60% and 28.73% respectively. In 11 out of 15 countries, the female youth (20-24) NEET rate was greater than the male youth (20-24) NEET rate. On average, the u-shaped trend is much more exemplified by both males and females among the 20-24 year old youths. This u-shaped trend is seen in 13 countries for 20-24 year old males and in 8 countries for 20-24 year old females. When comparing the NEET male and female rates between 15-19 year olds and 20-24 year olds, it can be seen that the NEET rates for both genders are higher for 20-24 year olds across all time periods. It is also interesting to note that there is not much variation in the NEET rate for either gender over time for 15-19 year olds and there is a much higher degree of variation for 20-24 year olds. Table 10 presents the youth NEET15 rates across four levels of educational attainment: Primary or Less, Lower Secondary, Upper Secondary General & Technical and Post- Secondary/Tertiary/Above16. In the post-2009 period, rates for the ECA region on average are 30.24% for primary or less, 14.93% for lower secondary, 19.21% for upper secondary and 23.41% for post-secondary and beyond. The NEET rate is highest for those with a primary or less level. Interestingly, the next highest NEET rate is for those who have a post-secondary, tertiary and above. In the 2001-2002 period, the average ECA NEET rate for those with a primary or less level of education was 26.76%, 15.52% for lower secondary, 25.63% for upper secondary and 27.23% for tertiary and above. Prior to the recession, the pre-2009 NEET rates are typically higher for those with a primary level but lower for other educational attainment levels. For the ECA region on average in the pre-2009 period the NEET rate for 32.91% for primary or less, 15.20% for lower secondary, 17.73% for upper secondary and 22.50% for those with a post-secondary and beyond level of education. When trends within each education level are examined, rates for those with a primary or less level of education on average in ECA increased from 2000 to pre-2009 and then declined in the post 2009 period. Rates for those with a lower secondary level of education seem 15 Russia, Serbia and Tajikistan are measured using different levels of attainment because the necessary disaggregated levels of educational attainment were unavailable in these countries. 16 Education groups were defined this way as defining population by finer levels of educational attainments resulted in sample size that was too small. 17 to hover around 15% for ECA throughout the sample periods, with a slight decline. The average NEET rates for those with an upper-secondary level of education seem to drop pre-2009 and then increase post-2009. A similar result is found for those with a post-secondary or tertiary level of education. Figures 6 and 7 present a breakdown of enrollment, employment, NEET (inactive) and NEET (unemployed) rates for youths who are ages 15-24. The difference between the two NEET rates is the NEET (inactive) rate measure youths who are considered NEET and are not actively participating in the labor market and NEET (unemployed) rate captures youths who are considered NEET and are labor market participants. In the pre-2009 period, ECA’s average enrollment rate is 58%, the employment rate is 20.56%, the NEET rate (unemployed) is 6.16% and the NEET rate (inactive) is 9.29%. In six countries in ECA, the NEET (inactive) rate is higher than the NEET (unemployed) rate with the opposite being true in the other eight countries. When moving to the post-2009 period, the average enrollment rate for ECA was 59.07%, the employment rate was 25.11%, the NEET (inactive) rate was 9.45% and the NEET (unemployed) rate was 9.99%. So when comparing with the pre-2009 period, there is a rise in labor force participation among youths in the post – 2009 period. The NEET (unemployed) rate is greater than the NEET (inactive) rate in 12 countries. This increase in the NEET (unemployed) rate shows that there was an overall increase in the labor force participation of youth’s aged 15-24. Figures 8-11 look at youth activity by analyzing the youth population in different sub groups, using the most recent data available for each year. Figures 8 and 9 present activity in ECA for ages 15- 19 and 20-24 respectively. For 15-19 year olds, the average enrollment for ECA is 81.18%. Given the age group this is not an unsurprising number. The average ECA employment rate is 8.35%. Moving onto the NEET rates, the average ECA NEET (inactive) rate was 6.81% compared to the NEET (unemployed) rate of 5.01%. In seven of the countries, the NEET (inactive) rate was greater than the NEET (unemployed) rate, with the opposite result true in the 12 other countries. So even though the average rate for ECA suggests that for 15-19 year olds the NEET (inactive) rate is greater than the NEET (unemployed) rate, in the majority of countries in ECA more 15-19 year olds are classified as NEET (unemployed) rather than NEET (inactive). When shifting focus to Figure 9 to analyze 20-24 years, it can be seen that average enrollment for ECA is 40.55%. This is not an unsurprising result given age group. The employment rate for 20-24 year olds in ECA was 38.16%. The average ECA NEET (inactive) rate was 12.90% was less than the NEET (unemployed) rate of 13.41%, an opposite results compared to 15-19 year olds. In seven of the countries, the NEET (inactive) rate was higher than the NEET (unemployed) rate, with the opposite result true in the 12 other countries. This is the same as youths aged 15-19. The difference between the two age groups is that both the NEET (inactive) and NEET (unemployed) rates are much higher for 20-24 year olds than 15-19 year olds. Also, for the ECA region average as a whole, the NEET (inactive) rate was higher than the NEET (unemployed) rate for 15-19 year olds and the 18 opposite being true for 20-24 year olds. This means that in terms of the NEET population, 20-24 years olds are more likely to be active participants in the labor market. Figures 10 and 11 examine youth (15-24) activity for males and females respectively. ECA’s average enrollment and employment for male youths (15-24) was 55.36% and 29.48% respectively. The average ECA NEET (inactive) rate was 6.45% and the NEET (unemployed) rate was 11.91%. So on average, NEET male youths are more likely to be active participants in the labor market. In six of the countries in ECA, the NEET (inactive) rate exceeds the NEET (unemployed) rate, suggesting that in the majority of the countries in ECA, male youth NEETs are active participants in the labor market. The ECA region average enrollment and employment rates for female youths were 61.52% and 21.13% respectively. So the average ECA female youth enrollment rate is higher than the male enrollment rate and the average ECA female youth employment rate was lower than the male employment rate. The ECA region average NEET (inactive) rate for female youths was 13.7%, which was higher than the NEET (unemployed) rate of 7.61%. In 10 countries the NEET (inactive) rate is greater than the NEET (unemployed) rate for women. This suggests that, as opposed to their male counterparts, female youth NEETs in ECA are mainly inactive. Tables corresponding to figures 8-11 are located in the appendix. 4.2 Comparing NEET youth vs. other youth in terms of household and individual characteristics. Tables 11 and 12 present descriptive statistics for the NEET youth population and the non-NEET youth population. To compute the descriptive statistics for both groups the most recent year available in the data is used. We find that the average age for the ECA NEET population is 21, with about 75% being between 20-24 years of age. This is a big difference compared to the non- NEET population whose age distribution is much more evenly split between 15-19 and 20-24 years of age. This difference between the NEET and non-NEET population is statistically significant in every country, except Bulgaria, in ECA. The NEET population is slightly more likely to be female, with the average split of ECA being 49.14% male. The NEET population is statistically more likely to be females in four countries in ECA (Ukraine, Kyrgyz Republic, Albania and Turkey). The NEET population is statistically more likely to be male in five countries (Moldova, Estonia, Lithuania, Slovenia, and Armenia). In Kyrgyz Republic and Turkey, the NEET population is predominantly women, with their respective NEET male populations only 30.66% and 25.55%. The Czech Republic, Lithuania, Estonia, Moldova, the Slovak Republic and Slovenia have a higher proportion of men in their NEET populations. The non-NEET populations are more evenly distributed across gender. More often than not, fewer NEETs inhabit urban areas. This result is statistically significant in nine countries. The ECA region on average has about 44% living in urban areas, compared to about 52% for the non-NEET. Turkey and Ukraine both have a majority of their NEET population in urban areas, but this holds for the non-NEET population as well and the difference is only statistically significant in Turkey. On average, most NEETs also have either 19 a lower secondary or upper secondary education level, and this is similar finding in non-NEET youths. It is important to note, that individuals classified as NEET tend to be older than their non- NEET counter parts yet have higher percentages at the lowest levels of educational attainment. On average, 18.2% of the NEET population is married compared to 4.56% in the non-NEET population. The mean household size in ECA for youth NEETs is marginally larger than that of non-NEET youths, 5.37 and 4.93 respectively. This result is statistically meaningful in 14 countries in ECA. The mean number of people over the age of 65 in the youth NEET community was 1.84, only slightly higher than in the youth non-NEET population where it is 1.70. In countries where the data is available, 2.85 of the youth NEET population are the household head in ECA and 3.59% for youth non-NEETs. A linear probability model is used to investigate determinants of the probability of being NEET among youths aged 15 -24. Table 13 gives the results of the regression analysis. In 15 out of 18 countries, if you are in the 15-19 year old age group you are significantly less likely to be NEET compared to individuals who are 20-24 years old. The female variable is significant in 11 out of the 18 countries, and is positive in two of them (Kyrgyz Republic and Turkey). The interactive term female and married is significant in 13 of 18 countries, and is positive in all them. This suggests that if an individual is a female and married she is more likely to be NEET, 42.8% on average. When household size is significant in 13 out of 20 countries, the higher the household size the greater the probability an individual is NEET. In 11 countries, we find that a person with an upper secondary or tertiary level of education is significantly less likely to be NEET compared to someone with a primary or less level of education. This is the case of only nine countries for lower secondary level of education. In Armenia and Ukraine, having a higher level of education makes an individual significantly more likely to be NEET. The regression results from this study show that individuals who are 20-24 years of age are much more likely to be NEET than those who are 14-19. This study also finds that generally, the lower the level of educational attainment the more likely the individual will be NEET. However it is important to note that in some countries higher level of education increase the likelihood that an individual is considered NEET. This suggests that educational achievement is not necessarily a shield against the idle youth phenomenon. Individuals who are both married and females are also much more at risk of being NEET. Finally, those youths who have larger households are also more likely to be NEET. All key findings from the LPM were also found when using a probit model. Tables 14 and 15 present descriptive statistics for NEET and non-NEET youths for countries where household budget information is available. Descriptive statistics are consistent with those in tables 11 and 12. The NEET youths have a higher age and are less likely to live in urban areas in comparison to the non-NEET youths. This finding is significant in all six countries. Computing the ECA region average of these six countries, 73.39% of the NEET youth population is 20-24 years of age compared to 45.25% in the non-NEET population. This corresponds to a higher mean 20 age of 20.94 in the NEET population. NEET and non-NEET have similar educational attainment and household size, although the marginally higher household size of NEETs compared to non- NEET youths is significant all six countries. In terms of marital status, 43.34% of the NEET population is married compared to 30.92% in the non-NEET group, and the NEET population has significantly more married people than non-NEETs in four countries. There are also fewer NEETs in urban areas in comparison to non-NEETs (53.89% vs. 59.73%). Tables 14 and 15 have additional household information compared to Tables 11 and 12. The average total per capita consumption for households containing the NEET population is USD 41,496.68, about USD 8,500 lower than that of the non-NEET households. The lower levels of consumption by households containing NEETs are seen in all six countries and are significantly different in each county. The NEETs also have a lower average share of household members employed than non-NEETs (19% vs. 26%). NEETs and non-NEETs are similar in terms of the gender of the household head, the marital status of the household head, and the share of households receiving public and private transfers. Table 16 presents the results from a linear probability model using the HBS data. In all six countries, it was found that individuals who were 15-19 years old were significantly less likely to be NEET compared to those who were 20-24 years of age, about 12.8% less likely on average. The female dummy variable was found to be significant in six countries, and was positive in three of them showing that in these countries female individuals were more likely to be NEET than their male counterparts. The married variable was only significant in three countries: it was negative in two of them and positive in one. The interaction term between female and married was significant in four out of six countries, and was positive in three of them, which is consistent with previous results in Table 8. The log of total per capita consumption per household is statistically significant in all six countries, and is negative in all six of them. Household size is significant in four countries, and positive in two of them implying that an individual with a larger household size is more likely to be NEET. For the two countries where the information is available, the share of the household that is employed is significant in both countries and is negative. Also, the share of the household that is between the ages of seven and 15 is significantly negative in Poland, Tajikistan and Ukraine. This is only the case for Tajikistan and Turkey for the share of the household between zero and six years of age. Interacting the female dummy variable with the share of household between zero and six is significant in Tajikistan and Turkey where it is negative. Interacting the female dummy variable with the share of household between seven and 15 is significant in Poland, Russia, Tajikistan and Ukraine. This variable is statistically negative in all these countries except for Russia. Having a basic level of education lowers the risk of being NEET compared to those with no or less than basic levels of education in four countries. This is the case in only three countries for those with a general secondary level of education, one country for special secondary and two for those with a tertiary level of education. In Russia, having a basic, general secondary, special secondary or tertiary level of education raises an individual’s probability of being NEET compared 21 to those with a less than basic level of education. The variable measuring the effect of having a female household head is only significant in Turkey and Ukraine, where it is positive in both. Receiving public transfers is only significant in two countries and is positive in one of them. Private transfers are only significant in Turkey and Ukraine, where they have a negative effect in Turkey and a positive effect in Ukraine. 5 Discussion and Policy Recommendations 5.1 Summary of Key Findings The primary objective of this report was to describe and analyze the NEET youth in ECA. In this respect, this report has seven main findings. First, on average, in the ECA countries under study about one in five youths are not in school, nor employment. Out-of-work and out-of-school youths thus make up a sizable group within the ECA region. Second, on average in ECA, the NEET (inactive) rate for youths aged 15-24 was 9.45% and the NEET (unemployed) rate of 9.99%. Looking at these rates from pre-2009 to post-2009 there was a rise in both NEET rates, but a larger rise in the NEET (unemployed) rate. This shows that in post-2009 more youths (15-24) considered themselves as part of the labor force. Examining activity rates of youths in ECA in different subgroups reveals some other interesting results. Using the most recent data available for each country in ECA finds that the average NEET (inactive) rate for 15-19 year olds was 6.81% compared to the NEET (unemployed) rate of 5.01%. When looking at 20-24 year olds, the ECA average NEET (inactive) rate was 12.90% and the NEET (unemployed) rate was 13.41%. Having a higher NEET (unemployed) rate compared to the NEET (inactive) rate is a result that is found in 12 countries. So the majority of NEETs ages 15-19 consider themselves inactive in the labor market whereas the majority of NEETs ages 20-24 consider themselves as active in the labor market but simply cannot find jobs. This is important for policy actions as policies that target 20-24 year olds should be aimed at getting them employed as they clearly want to work but simply cannot find jobs. Looking at activity rates across gender reveals first that on average in ECA females ages 15-24 have a higher enrollment rate than males ages 15-24 (61.52% vs. 55.36%). For males 15-24, the ECA average NEET (inactive) rate was 6.45% and the NEET (unemployed) rate was 11.91%. The ECA average NEET (inactive) rate for females 15-24 was 13.47% and the NEET (unemployed) rate was 7.61%. So more than the majority of male NEETs considered themselves as part of the labor force, whereas the majority of female NEETs considered themselves as not part of the labor force. Third, there is a lot of heterogeneity across countries in the characteristics of NEET youths compared to non-NEET youths. However, in order to draw a general profile of NEETs in ECA, one can say that in a majority of countries, compared to other youths, NEET youths more often are female, older, married, have lower educational attainment, live less often in urban areas and 22 have a larger household size. NEETs also tend to have lower per capita total household expenditures than their non-NEET counterparts. Females who are married are found more likely to be NEET in 15 out of 20 countries. Fourth, no consistent picture emerged on gender and NEET youth (15-24) rates for the entire ECA region. While, the ECA average has the NEET rate for males 15-24 to be 18.18% and for females 15-24 to be 20.93% this is not a consistent result across sub-regions. In the EU11 and BMU regions males (15-24) in the post-2009 period males have a higher NEET rate than females. Also over time this study finds that NEET rates for males (15-24) exhibit a u-shaped trend over the three time periods, whereas female (15-24) NEET rates show more of a decline from 2001-2002 to pre-2009 and a slight increase in post-2009. This can be explained by the fact that more males consider themselves to be part of the labor force than women, so the 2009 recession may have hit men harder than women. Looking at NEET rates for 15-19 year olds by gender finds very little difference over all time periods between males and females, with the respective ECA average NEET rates being 11.28% and 11.37%. When looking at 20-24 year olds, the NEET gender gap is much larger. The ECA average male (20-24) NEET rate was 23.60% compared to 28.73% for females (20-24). Fifth, in terms of trend, in nine out of 17 countries, the NEET rate for youths declined over the 2000s prior to the financial crisis. When the crisis hit however, the NEET rate increased in 11 countries in ECA. Sixth, the gender gap due to a higher NEET rate for women in several countries seems to have declined in recent years. In 2001-2002, this study finds that the average ECA NEET rate for males was 19.38% and for women was 23.70% resulting in a gender gap of more than four percentage points. In the post-2009 period, ECA had male NEET rates of 18.18% and female NEET rates of 20.93%, giving a gender gap of under three percentage points. Seventh, after the 2009 financial crisis there was a rise in the NEET (unemployed) rate for ECA. However, the same rise did not occur with the NEET (inactive) rate. This suggests that the primary reason for the increased NEET rates for ECA can be attributed more to a lack of employment among youths and not inactivity. This could have potential ramifications when implementing NEET based policies. 5.2 Compare findings with relevant benchmarks from EU15 or OECD Countries In this study, in the post-2009 year, the average youth NEET rate for ECA was 19.60%. The average youth NEET rate for the OECD was 16% in 2011 (OECD 2013). So on average, the ECA region has a youth NEET rate that is close to 3-percentage point higher than that in the OECD. A common result across the EU was a declining NEET rate prior to crisis and an increase afterwards (Eurofound 2012): this study finds a similar trend for the ECA region. 23 In the ECA region, this study finds that the youth NEET rate is higher among women than men, although this does not hold for eight countries in ECA. This is consistent with results for the EU and OECD. On average, the youth NEET rate among women is higher than that of men in the EU (Eurofound 2012) and the OECD (OECD 2013). For example, the OECD average youth NEET rate for women aged 15-29 is 18%, and only 13% for men (OECD 2013). However, this finding is not found systematically in all EU and OECD countries (Eurofound, 2012). Another notable comparison is in terms of the NEET gender gap through time. The declining gender gap found in this study is consistent with a similar trend observed in the EU (Eurofound, 2012) and with research that suggests that that young men were hit harder than young women during the financial crisis (Verick 2009, ILO 2012, Eurofound 2012). This study finds that in the ECA region in the post-2009 period, the NEET rate for 20-24 year olds is more than double that of 15-19 year olds. This is the case in every country in ECA. This finding is concurrent with findings in OECD, where the NEET rate for 15-19 year olds in 2010 was 7.67% and for 20-24 year olds was 17.97% (OECD 2013b). This is also in line with the EU, where in 2011 the NEET rate for 20-24 years old was 18.2% and for 15-19 year olds was 6.9% (Eurofound 2012). The average ECA NEET rate for the 15-19 age group was about four percentage points higher than the OECD and EU average. The average ECA NEET rate for the 20-24 age group was about eight percentage points higher than the OECD and EU average. 5.3 ECA-relevant possible policy actions to incorporate youth into productive employment or education, drawing on the international evidence Out-of-work and out-of-school youths represent a significant issue in the ECA region and relevant policies are needed in this region. In general, policies implemented to address the NEET problem are two-facetted: they aim to keep/ reintegrate youths in school or they help youths get jobs. The following sections will: (i) Review education based policies aiming at keeping or reintegrating youths in school (ii) Review employment based policies (iii) Propose policy recommendations for ECA 5.3.1 Education based NEET Policies There are a variety of programs that have been used in Europe to address early school leaving. Policies that curb dropout rates in schools have become more and more prevalent across Europe and are usually implemented at the country level and are “umbrella policies”. For example, the Netherlands in 2002 started their ‘Drive to reduce dropout rates’ program, which focuses on working at the local level with contracts, additional funds and financial incentives to prevent early school leaving. The ‘Drive to reduce dropout rates’ includes compulsory school attendance as well 24 as digital attendance programs that help to identify student who are at risk of leaving school early. The Ministry of Education, Culture and Science (2012) found that the program resulted in a reduction in early school leaving by almost 20% from 2005/2006 to 2008/2009. Some countries have adopted more consequence-based policies to keep youths in school. For example, in Hungary in 2010 families can lose a portion of their state benefits if their children of required school age do not attend school. The Czech Republic, in 2005, said it would remove unemployment benefits for those who left school early. This policy did have a positive effect, as the unemployment rate for early school leavers dropped after this program was implemented (Hawley et al. 2012). Other educational programs are aimed at getting youths who have left school to return. Eurofound (2012) highlights policies that fall under the umbrella of reintegration. Tracking programs are used by Denmark, Luxembourg, the Netherlands and England with the purpose of keeping tabs on early school leavers and contacting and supporting them in searching for employment or training. There are also ‘second chance schools’ which range from evening schools to certification exams. In Cyprus, Portugal and Spain, many of these programs are vocational in nature. Many of these second chance schools also have alternative learning methods aimed at giving those who left education a more positive experience. Hawley et al. (2012) describes a program in Slovenia called ‘Project Learning for Young Adults’ where participants are meant to achieve a positive learning experience and learn to define career and life goals. The program has had massive success and about 60% of participants re-enter education or find employment after completing the voluntary program. Finally, some countries have recently begun using financial incentives to motivate young people to return to school. Sweden began a policy in 2011 that gave unemployed young people, ages 20- 24, who did not have an upper second-level qualification a higher level of student aid. Activity Allowance, a UK pilot program run between 2006-2011, gave NEETs aged 16—17 years £30 who took part in a personalized plan session aiming to reintegrate them into education (Eurofound, 2012). In Macedonia, a financially motivated incentive program known as the Conditional Cash Transfer (CCT) program was implemented. The CCT program provides cash transfers to poor households (i.e. those eligible for Social Financial Assistance) if they have children of secondary school age and these children attend secondary school at least 85% of the time. An initial impact evaluation of the program was done and revealed that secondary school enrollment for target children increased by about 10 percentage points as a result of the CCT program, a huge increase since average secondary school attendance for targeted children is about 60% (Armand and Carneiro 2013). These finding are concurrent with Ralwings and Rubio (2005), who found that CCT programs similar to the one on Macedonia were successful at increasing school enrollment rates in Colombia, Mexico and Nicaragua. 25 5.3.2 Employment Based Policies Other types of policies implemented around Europe are focused on easing the transition into employment for youths and increasing the employability of youths. Some policies aim to reduce NEET rates in Europe focus on the school-to-work transition period, because for many young Europeans it takes two years or more to find their first job after finishing school (Hawley et al. 2012). Compared to the United States, Quintini and Manfredi (2009) found that school-to-work transitions involved significantly more time spent in unemployment. Programs that have been implemented in Europe to reduce the school-to-work transition are youth guarantee programs, counseling/advising programs and entrepreneurship programs. Youth guarantee policies have been put in place in several countries in Europe – Finland, the Netherlands, Poland, Sweden to name a few. The European Parliament also proposed the European Youth Guarantee in 2010, which would give any young person out of work four months the right to a job, apprenticeship and further training. The EU Council of Ministers adopted this proposal as the Youth Guarantee Recommendation in April 2013, and the European Commission is currently urging Member States to make the youth guarantee a reality (European Commission, 2013). In Finland, the youth guarantee program targets unemployed youths less than 25 years of age and aims to help them develop a plan, help them assess what they need to do to gain employment and find them a job or some type of training to increase employability within three months of registration. The program saw 83.5% of its job seekers get a job or receive some type of academic or vocational training placement within three months of registering as unemployed (Eurofound, 2012b). In Sweden, the job guarantee for young people targets unemployed youths ages 16-24 who have been registered as unemployed over three months. It uses active labor market activities to help the individual find employment or training. In 2008, the Institute for Labour Market Policy Evaluations (IFAU) evaluated the Swedish youth guarantee program and found that unemployed 24-year olds who participated in the program were able to find employment quicker than older people who were registered with public employment services. It should be noted though that both programs in Sweden and Finland were found to be less successful in times of economic crisis (Eurofound 2012b). In order to reduce the school-to-work transition period, some countries have enacted policies aimed at providing youths with guidance and counseling, to help them manage expectations and offer career advice. Advice can come from the education sector and from specific public counseling and support programs. A key feature of these programs is to reduce search frictions between employers and youths. One way in which this is done is through matching services. Countries like Bulgaria, the Czech Republic, Latvia and Sweden have begun to use electronic matching programs in an attempt to help youth and employers connect more easily. Other countries such as Estonia, Greece, Lithuania and Slovenia emphasize Job Fairs and career days as a way to connect youths with employers (Eurofound, 2012). Although in theory such programs are expected 26 to help reduce youth unemployment and inactivity, to our knowledge, there is no empirical evidence on the effectiveness of such programs Entrepreneurship policies have also been introduced in Bulgaria, Cyprus, Greece, Slovenia and Romania. These programs offer advice and support to young people who wish to start their own businesses. Some countries such as Slovenia target unemployed youths and others like Greece target any youths who wish to start their own businesses. There is no information on the effectiveness of entrepreneurship policies in reducing the YUR or the youth NEET rate. Other employment-based schemes that have been used in Europe generally aim at increasing the employability of youths, either through work-study programs, apprenticeships or financial incentives to employers. Countries like Austria, Belgium, Czech Republic, the Slovak Republic, and the UK have work-study programs. Work-study programs offer vocational education at the upper secondary and post-secondary non-tertiary levels of education. According to OECD (2013), the transition between school and work has been easier in countries which have work-study programs that those without, albeit less so in times of economic recession. Furthermore, countries with work-study programs did better than the OECD average with respect to some, but not all measures of labor market performance, and had a 14% NEET rate in comparison to 16% for the OECD (OECD, 2013). In our view, it is uncertain however how much of this gap in the NEET rate can be allocated to the work-study programs alone. Other programs involve apprenticeships aimed at giving youths skills and experience, which will make them more employable. According to Eurofound (2012), apprenticeship programs have been very successful in easing the transition from school to work for young people. Austria and Germany have amongst the best apprenticeship programs, which helped keep unemployment down during the crisis (OECD, 2010). Italy also has an apprenticeship program targeted at 15-29 year olds aimed at providing individuals with off-the-job training and paid employment. The program was conducted in two pilot phases (2004-2008 and 2010) with 1,000 participants. There was no experimental design. In 2008, the program was evaluated based on the employment status of the participants over time (ISFOL 2011). The ISFOL’s evaluation found that 71% were in employment with the same firm, 21% elsewhere (usually to better, higher paid positions) and the remaining 8% dropped out. This deemed the program a success (ISFOL 2011). Overall, both apprenticeship and work-study programs find support in the literature as effective ways at reducing the NEET rate by aiding youths in their search for employment (Eurofound, 2012, OECD, 2013). The evidence in support of such programs however remains limited. Finally, some programs are aimed at employers and making youths more attractive job candidates and thus increasing the demand for young workers and help them enter the labor market. These programs involve tax breaks and in some countries wage subsidies for youths to increase the demand for young workers. Employer incentive programs have been shown to have positive effects in the short run, but the long run effects has shown to be less effective than training 27 programs (Kluve 2006). An example of a wage subsidy program is the Chances Card policy in Finland, introduced in May 2010. It had as many as 18,500 participants by January 2011. This program was targeted at youths aged 18-30 who were unemployed and had higher levels of educational attainment. Employers who hired these ‘card’ carriers received a subsidy and thus could lower their labor costs. An evaluation done by Pitkänen, Aho and Syrjä in 2012 found that 22.1% of card carriers found employment with Chances Card while 40% of card carriers did not use the cards when applying for jobs. Furthermore, 40% of employers stated they would have hired the young person even without the subsidy and only half remained employed after the subsidized period ended (Eurofound, 2012). Several research studies adopt a bird’s eye view of the relative effectiveness of different Active Labor Market Policy Measures (ALPMs). ALPMs have been used to target people of all ages, but some have been aimed directly at the youth population. Bell and Blanchflower (2011) use data from the US and UK to analyze youth unemployment. In their review of ALPMs, it is found that ALPMs that are successful in increasing the probability of gaining employment usually target particular subgroups, such as groups of a certain level of educational attainment or certain age subgroups. They also find that ALPM programs are usually more successful for women than for men. More importantly, ALPMs directed at young people are less successful than those that target older age groups (Bell and Blanchflower 2011, Kluve 2006). Bell and Blanchflower state that even though the effectiveness of current policies for youth unemployment are uncertain, there is still a need for policy intervention to address youth unemployment. Pemberton (2008) notes that perhaps the overall limited success of policies aimed at NEETs has been because of the variety of influences on a young person’s NEET status. Pemberton suggests that a greater emphasis should be placed on intergenerational issues such as positive parenting, addressing family breakdown and raising hopes. Pemberton also speculates that combining an emphasis on intergenerational issues with policies that offer alternative forms of education and increase employability of youths should be effective. 5.3.3 Policy Recommendations for the ECA Region As seen in the previous two sections, for ECA NEET policies to effective they must combat both keeping/reintegrating youths into school as well as helping youths gain employment. Since the NEET population in ECA is mainly comprised of 20-24 year old, in general, it appears that policies aimed specifically at curbing youth inactivity should prioritize transitions to employment over school dropout prevention programs. Many of the education based programs reviewed seem theoretically sound but have not been rigorously evaluated -e.g. second chance programs, financial incentive programs. National umbrella policies to prevent early school leaving such as the Netherlands ‘Drive to reduce dropouts’ have been found to be successful at reducing dropout rates (Ministry of Education, 28 Culture, and Science 2012, Hawley et al. 2012). Financially driven policies to curb school dropout rates have been implemented in Sweden and the Czech Republic and have shown that the threat of loss of unemployment benefits to those that leave school early is effective (Hawley et al.) As for employment policies, they should strive to reduce the school-to-work transition period for individuals, which tend to be long in Europe (Quintini and Manfredi 2009). Youth guarantee programs have been seen to work in both Finland and Sweden except in times of economic downturn (Eurofound 2012b). Public employment services for youths are other options for reducing the frictions between youths and employers. Programs such as electronic matching services, while untested, are theoretically sound and should be implemented in countries that have the necessary infrastructure and have significant frictional unemployment. Other employment related policies could work to increase the employability of young people in ECA. Programs that have shown to be successful at increasing the qualifications of young people are work-study programs and apprenticeships. Germany and Austria have had great success with apprenticeship programs, and they also have some of the lowest NEET rates in the EU (Eurofound, 2012). Italy also has implemented an apprenticeship program for higher educated young people, which has also been successful (ISFOL 2011). The lack of vocational training in the educational system has also been shown to be a cause of the NEET phenomenon (Breen 2005, Eurofound 2012). For these reasons, programs such as apprenticeships that offer vocational training may be implemented in some countries in the ECA region. Wage subsidy programs have also been shown to be effective in the short run (Kluve 2006, Eurofound 2012). These programs are expensive, but give employers an incentive to hire young workers. In general, countries such as Germany, Austria, the Netherlands, the UK, Finland and Sweden may be looked to as possible guides for policy as these countries have the lowest NEET rates in Europe. However, many of the programs mentioned earlier and used in countries in Europe and the OECD are expensive. For example, public employment services programs, and financially based programs such as financial aid for early school leavers and wage subsidies/tax breaks for employers who hire youths, come with significant costs. Cost is not the exclusive barrier to adopting such policies and programs. Training programs such as apprenticeships and work-study programs require a working relationship between educational institutions and employers, which may not always be in place. Furthermore, schools that offer vocational training to their students are not found in every country (Eurofound 2012). This would require an overhaul of the educational infrastructure. Finally, ALMPs have been shown to be much more successful when they target specific subgroups of the population (Bell and Blanchflower 2011, Pemberton 2008). Due to the great differences in economy, institutional infrastructure and the heterogeneous makeup of the NEETs across ECA 29 found in this study, it would be unwise to implement policies that blanket the region. Each country should carefully investigate its own NEET population to identify key subgroups that can be targeted with national programs, as well as determine its own institutional capacities in order to successfully combat its own NEET problem. 5.4 Knowledge gaps and a possible research and policy agenda on NEET youth in ECA. Limited research is available on the NEET youth population in the ECA region. This report attempted to give information on the NEET youth population in the ECA region. NEET rates are determined for age groups, males and females as well as different levels of educational attainment. A linear probability model is also used to determine groups that are more at risk for becoming NEET. However, this report has several limitations that could be addressed in future research with improved data. In particular, this study drew a general profile of youth NEETS in ECA. It might be the case that there are other characteristics that help characterize youth NEET in ECA and understand the determinants of becoming an out-of-work and out-of-school youth. For example, the ethnic and family backgrounds and the health of the individual were not available for the purpose of this report, while in other countries (Pemberton 2008, Eurofound 2012, Quintini and Manfredi 2009), these have been shown to give insights into the NEET populations, especially the income and educational level of a NEET’s parents. Further research is needed in ECA countries to draw more detailed profiles of youth NEETs in countries when data is available on demographic, socioeconomic and health characteristics of youths and their families. Further studies on NEETs in ECA should also address issues regarding the impact of labor regulations governing part-time work and apprenticeships. It would also be of interest if longitudinal data were used to measure the duration an individual is classified as NEET. Less harmonized versions of SILC, LSMS, ICLS and the HBS may be of use to further research some of these fields. Finally, surveys specifically aimed at NEET youths could be used to determine the if their inactivity is a choice or involuntary. The results on the determinants of being a youth NEET presented in this report uses a simple linear probability model with cross sectional data. It suffers from omitted variable bias and measurement error. Research is needed using longitudinal data following youths over time and their transition to NEET status or work. A model based on longitudinal data would address omitted variable bias and systematic measurement error. It would also help in order to identify factors that help or hinder. Longitudinal data is available with SILC (e.g. Bulgaria, Estonia, Hungary, Poland) and the LSMS for certain years and countries (i.e. Tajikistan LSMS 2007 and 2009, Albania LSMS 2002- and 2004 and Bosnia and Herzegovina LSMS 2001 and 2004). This report also reviewed policies and programs aimed at curbing inactivity among youths. There is a lot research done on countries that are in the OECD and the EU. Results from such studies may not be transferrable to ECA countries due to different school, labor market and institutional 30 contexts. Minimal information is available about government policies and programs aimed at reducing their NEET rates in ECA. In addition, to the best of the authors’ knowledge, no evaluation of such program is available. Program evaluations of school dropout prevention programs as well as school-to-work transition programs and policies aimed at increasing the employability of the region’s youth population are needed in the ECA region. 31 6 Bibliography Armand, A., & Carneiro, P. (2013). Impact Evaluation of the Conditional Cash Transfer Program for Secondary School Attendance in Macedonia. Working Paper. Bell, D., & Blanchflower, D. (2011). Young People and the Great Recession. Oxford Review of Economic Policy , 27 (2), 241-267. Breen, R. (2005). Explaining Cross-national Variation in Youth Employment. European Sociological Review , 21 (2), 125-134. Brenke, K. (2012). Unemployment in Europe: Young People Affected Much Harder Than Adults. DIW Economic Bulletin (9), 15-23. Bynner, J. (2005). Rethinking the Youth Phase of the Life-course: The Case for Emerging Adulthood. Journal of Youth Studies , 8 (4), 367-384. Bynner, J., & Parsons, S. (2002). Social exclusion and the transition from school to work: the case of young people not in education, employment. Journal of Vocational Behavior , 8, 289-309. Bynner, J., Joshi, H., & Tsaysas, M. (2000). Obstacles and Opportunities on the Route to Adulthood: Evidence from Rural and Urban Britain. Smith Institute . Cardenas, M., Rafael de Hoyos, & Szekely, M. (2011). Idel Youth in Latin America: A Persistent Problem in a Decade of Prosperity. Caroleo, F., & Pastore, F. (2007). The Youth Experience Gap: Explaining Differences Across EU Countries. Quad Dip Econ Fin Stat 41, University of Perugia . Choudhry, M., Marelli, E., & Signorelli, M. (2012). Youth unempolyment rate and impact of financial crises. International Journal of Manpower , 33 (1), 76-95. Contini, B. (2012). Youth employment in europe: do institutions and social capital explaine better than mainstream economics? The European Journal of Comparative Economics , 9 (2), 247-277. Elder, S., Nobkovska, B., & Krsteva, V. (2013). Labour Market Transitions of Young Women and Men in the former Yugoslav Republic of Macedonia. Work4Youth Publication Series . Eurofound. (2012). NEETs – Young people not in employment, education or training: Characteristics, costs and policy responses in Europe. Luxembourg: Publications Office of the European Union. Eurofound. (2011). Young People and NEETs in Europe: First Findings. Dublin: Eurofound. Eurofound. (2012b). Youth Guarantee: Experiences from Finland and Sweden. European Commission. (2013, May 28). EU Measures to tackle youth unemployment (MEMO). Brussels: European Commission. Eurostat. (2013, May). Statistics on Employment and Social Policy. Retrieved June 2013, from Eurostat: 32 http://epp.eurostat.ec.europa.eu/portal/page/portal/employment_social_policy_equality/in troduction Fares, J., Montenegro, C., & Orazem, P. (2006). How are Youth Faring in the Labor Market? Evidence from Around the World. World Bank Policy Research Working Paper No. 4071 . Gaston, N., & Kishi, T. (2005). Labour Market Policy Developments in Japan: Following an Australian Lead? The Australian Economic Review , 38 (4), 389-404. Genda, Y. (2007). Jobless Youths and the NEET Problem in Japan. Social Science Japan Journal , 10 (1), 23-40. Hawely, J., Nevala, A., & Weber, T. (2012). Recent policy developments related to those not in employment, education and training (NEETs). Dublin: Eurofound. ILO. (2009). School-to-Work Transition Survey: A methodological guide. Geneva. International Labour Office. (2012). Global Employment Trends for Youth 2012. Geneva: International Labour Organization. ISFOL. (2012). Italian apprenticeships through a continuous reforming process. Third International Seminar of the National Apprenticeships Service, National Apprenticeship Week . Kluve, J. (2006). The Effectiveness of European Active Labor Market Policy. IZA Discussion Paper No. 2018 . Marshall, K. (2012). Youth neither enrolled nor employed. Perspectives on Labour and Income , 24 (2), 1-15. Ministry of Education, Culture and Science. (2012). The approach to Early School Leaving Policy in the Netherlands and the provisional figures of the 2010-2011 performance agreements. Netherlands. Neumark, D., & Wascher, W. (2004). Minimum wages, labor market institutions, and youth employment: A cross-national analysis . Industrial and Labor Relations Review , 57 (2), 223-248. OECD. (2012). Employment and Labour Markets: Key Tables from OECD. OECD. OECD. (2013). How Difficult is it to move from School to Work. Education Indicators in Focus No 13 . OECD. (2013b). Improving the Economic Situtuation of Young People. In OECD Economic Surveys: France 2013. OECD. OECD. (2010). Off to a good start? Jobs for Youth. OECD. (2008). Off to a Good Start? Youth Labour Market Tansitions in OECD Countries. Pemberton, S. (2008). Tackling the NEET generation and the ability of policy to generate a 'NEET' solution - evidence from the UK. Environment and Planning C: Government and Policy , 26 (1), 243-259. Perugini, C., & Signorelli, M. (2010). Youth labour market performance in European regions. Economic Change and Restructuring , 43 (2), 151-185. 33 Quintini, G., & Manfredi, T. (2009). Going Separate Ways? School-to-Work Transitions in the United States and Europe. OECD Social, Employment and Migration Working Papers No. 90 . Quintini, G., & Martin, S. (2006). Starting Well or Losing their Way? The Position of Youth in the Labour Market in OECD Countries. OECD Social, Employment and Migration Working Papers No. 39 . Rawlings, L. B., & Rubio, G. M. (2005). Evaluating the Impact of Conditional Cash Transfer Programs. The World Bank Research Observer , 20 (1), 29-55. Rennison, J., Maguire, S., Middleton, S., & Ashworth, K. (2005). Young people not in education, employment or training: evidence from the education maintenance allowance pilots database. Centre for Research in Social Policy, Department for Education and Skills , RB 628. Salazar-Xirinachs, J. (2012). Generation Ni/Ni: Latin America's Lost Youth. Americas Quarterly , 110-113. Sattar, S. (2012). Women in the Labor Market. Washington DC: The World Bank. Scarpetta, S., Sonnet, A., & Manfredi, T. (2010). Rising Youth Unemployment During the Crisis: How to Prevent Negative Long-Term Consequences on a Generation? OECD Social, Employment and Migration Working Papers No. 106 . Social Exclusion Unit (SEU). (1999). Bridging the Gap: New Opportunites for 16-18 Year Olds Not in Education, Employment or Training. CM4405 . Strelitz, J. (2004). Tackling disadvantage: education. Tackling UK Poverty and Disadvantage in the 21st Century: An Exploration of the Issues Eds D Darton, J Strelitz (Joseph Rowntree Foundation, York) , 57-75. Verick, S. (2009). Who is hit hardest during a Financial Crisis? The Vulnerability of Young Men and Women to Unemployment in an Economic Downturn. IZA Discussion Paper No. 4359 . Walther, A., & Pohl, A. (2005). Thematic Study on Policy Measures Concerning Disadvantaged Youth. Study commissioned by the European Commission: Final Report . World Bank. (2013). Data: Europe & Central Asia. Retrieved June 2013, from The World Bank: http://data.worldbank.org/region/ECA World Bank. World Development Report 2007: Development and the Next Generation. Washington DC: The International Bank for Reconstruction and Development/The World Bank. World Bank. (2012). World Development Report 2013: Jobs. International Bank for Reconstruction and Development / The World Bank. 34 7 Tables Table 1: Labor Market Indicators for ECA and OECD Countries outside ECA ECA Countries Youth (15-24) General Ratio of OECD Countries Youth (15-24) General Ratio of Unemployment Unemployment Unemployment Outside of ECA Unemployment Unemployment Unemployment Rate (%) -2012 Rate (%) - 2012 For Youths (15- Rate (%) -2012 Rate (%) - 2012 For Youths (15- 24) to General 24) to General Unemployment Unemployment Rate - 2012 Rate - 2012 Albania 27.9 13.4 2.08 Canada 14.3 7.2 1.99 1 Armenia 42.6 21.6 1.97 France 23.9 9.8 2.44 Azerbaijan 14.2 5.2 2.73 Germany 8.1 5.5 1.47 Bulgaria 28.1 12.3 2.28 Japan 7.9 4.4 1.80 Croatia 43 15.8 2.72 United Kingdom 21 7.9 2.66 Czech Republic 19.5 7 2.79 United States 16.5 8.1 2.04 Estonia 20.9 10 2.09 Georgia 33.3 15 2.22 Hungary 28.1 10.9 2.58 Kazakhstan 3.9 5.3 0.74 Kyrgyz Republic 17.6 8.4 2.10 Latvia 28.5 15 1.90 Lithuania 26.7 13.4 1.99 FYR Macedonia 53.9 31 1.74 Moldova 13.1 5.6 2.34 Poland 26.5 10.1 2.62 35 Romania 22.7 7 3.24 Russian 14.8 5.5 2.69 Federation Serbia 51 23.9 2.13 Slovak Republic 34 14 2.43 Slovenia 20.6 8.8 2.34 2 Tajikistan 16.7 11.5 1.45 Turkey 15.7 8.2 1.91 Ukraine 17.3 7.5 2.31 Source : ILOSTAT - LFS 2012 1: LFS 2011; 2: LFS 2009 36 Table 2: Inactivity Rates for the Total Population, Youths (15-24), Males (15-24) and Females (15-24) in select ECA countries and EU27 Region Country Total Pop (15-64) Youths (15-24) Males (15-24) Females (15-24) Bulgaria 32 70 66 75 Czech Republic 27 69 63 74 Croatia 36 70 65 75 Estonia 25 60 59 62 FYR Macedonia 35 66 60 73 Hungary 35 73 70 76 Latvia 26 61 57 64 Lithuania 28 69 64 73 Poland 33 67 62 72 Romania 35 69 64 74 Slovenia 30 66 63 70 Slovak Republic 30 69 62 76 Turkey 46 61 49 73 EU27 28 58 55 61 Source: Eurostat - EU LFS 2013 37 Table 3: Rates of Not in Employment, Education or Training (NEET) among Youths (15-24) in select ECA Countries and EU27 Region Country NEET (%) Bulgaria 21.6 Czech 9.1 Republic Croatia 18.6 Estonia 11.3 FYR 24.8 Macedonia1 Hungary 15.4 Latvia 13 Lithuania 11.1 Poland 14.2 Romania 17.2 Slovenia 9.2 Slovak 13.7 Republic Turkey1 28.7 EU27 12.9 Source: Eurostat - EU LFS 2013 1: EU LFS 2012 38 Table 4: % NEET by Age Group NEET - working age (15-64) NEET - Youth (15-24) NEET - Prime Aged Workers (25- 49) ECA Sub- Country 2001- Pre - Post - 2001- Pre - Post - 2001- Pre - Post - region 2002 2009 2009 2002 2009 2009 2002 2009 2009 ECA - All1 30.7 28.5 29.4 21.6 18.6 19.6 25.9 24.3 26.4 BMU Moldova 29.1 37.8 41.3 25.6 26.7 27.7 27.4 40.0 43.4 (0.2) (0.2) (0.2) (0.4) (0.3) (0.3) (0.3) (0.2) (0.2) BMU Ukraine 30.4 26.3 28.4 20.7 16.9 19.0 23.6 21.7 23.8 (0.1) (0.1) (0.1) (0.2) (0.2) (0.2) (0.1) (0.1) (0.1) C.A Kyrgyz Republic NA NA 25.2 NA NA 16.7 NA NA 26.6 (0.2) (0.4) (0.3) C.A Tajikistan NA NA 40.9 NA NA 39.8 NA NA 44.8 (0.7) (1.1) (1.0) EU11 Bulgaria 38.2 25.5 27.1 29.5 22.2 27.3 31.0 18.8 19.4 (0.2) (0.6) (0.5) (0.5) (1.3) (1.2) (0.3) (0.7) (0.6) EU11 Czech Republic 23.4 23.9 24.7 13.9 11.0 12.6 16.9 17.3 17.9 (0.2) (0.4) (0.4) (0.4) (0.9) (0.9) (0.2) (0.5) (0.5) EU11 Estonia 24.6 17.8 27.5 13.1 10.6 18.7 22.2 15.4 25.1 (0.8) (0.5) (0.6) (1.4) (0.9) (1.1) (1.1) (0.6) (0.8) EU11 Hungary 31.9 28.3 30.4 14.8 13.8 16.0 26.1 20.4 23.8 (0.2) (0.4) (0.4) (0.4) (0.8) (0.8) (0.3) (0.5) (0.5) EU11 Latvia 28.2 20.5 33.4 19.1 12.4 21.9 23.2 17.2 31.2 (0.7) (0.5) (0.5) (1.4) (0.9) (1.1) (0.9) (0.6) (0.7) EU11 Lithuania 25.9 18.6 28.2 13.8 5.9 17.8 22.3 15.9 25.4 39 (0.6) (0.6) (0.8) (0.9) (0.9) (1.5) (0.7) (0.8) (1.0) EU11 Poland 35.5 27.5 29.0 23.3 10.5 13.5 32.2 21.4 20.9 (0.3) (0.3) (0.4) (0.4) (0.5) (0.7) (0.3) (0.4) (0.4) EU11 Romania 30.6 25.1 27.4 24.7 12.4 15.2 25.5 19.7 20.9 (0.3) (0.5) (0.5) (0.7) (0.9) (1.1) (0.4) (0.6) (0.6) EU11 Slovak Republic 31.0 19.6 24.6 22.6 7.4 13.5 24.8 11.8 18.2 (0.3) (0.4) (0.4) (0.6) (0.5) (0.7) (0.4) (0.4) (0.5) EU11 Slovenia 22.5 22.9 25.2 8.5 5.1 6.4 14.5 13.8 15.7 (0.4) (0.4) (0.4) (0.5) (0.4) (0.5) (0.4) (0.4) (0.4) R.F. Russia NA NA 8.4 NA NA 9.5 NA NA 9.2 (0.1) (0.2) (0.2) S.C. Armenia NA 40.4 NA NA 47.0 NA NA 39.1 NA (0.4) (0.9) (0.6) S.E.E. Albania 32.6 29.7 NA 31.9 27.2 NA 28.4 26.5 NA (0.6) (0.6) (1.1) (1.1) (0.7) (0.7) S.E.E. Macedonia NA 45.3 41.7 NA 31.9 25.3 NA 44.7 41.5 (0.3) (0.3) (0.6) (0.6) (0.4) (0.4) S.E.E. Serbia NA NA 22.6 NA NA 21.4 NA NA 24.5 (0.5) (1.0) (0.7) Turkey Turkey 45.6 46.1 43.2 41.0 36.0 30.5 44.9 45.9 43.3 (0.1) (0.1) (0.1) (0.2) (0.2) (0.2) (0.2) (0.1) (0.1) Source: Author's Calculations Standard Errors In Parentheses 1: ECA Total and Sub-regions are country averages 40 Table 5: % NEET for Age Groups by ECA Sub-region NEET - working age (15-64) NEET - Youth (15-24) NEET - Prime Aged Workers (25-49) Region/Subregion1 2001- Pre - Post - 2001- Pre - Post - 2001- Pre - Post - 2002 2009 2009 2002 2009 2009 2002 2009 2009 ECA - All1 30.7 28.5 29.4 21.6 18.6 19.6 25.9 24.3 26.4 BMU 29.7 32.0 34.8 23.1 21.8 23.3 25.5 30.9 33.6 Central Asia (C.A.) NA NA 33.0 NA NA 28.2 NA NA 35.7 EU11 29.2 23.0 27.8 18.3 11.1 16.3 23.9 17.2 21.9 Russian Federation (R.F.) NA NA 8.4 NA NA 9.5 NA NA 9.2 South Caucasus (S.C.) NA 40.4 NA NA 47.0 NA NA 39.1 NA South Eastern Europe 32.6 37.5 32.1 31.9 29.6 23.4 28.4 35.6 33.0 (S.E.E.) Turkey 45.6 46.1 43.2 41.0 36.0 30.5 44.9 45.9 43.3 1: ECA Total and Sub-regions are country averages Source: Author's Calculations 41 Table 6: NEET Youth Rates for 15-19 and 20-24 Year Olds Over Time - By Sub-region Region/Subregion 2001-2002 Pre-2009 Post - 2009 15-19 (%) 20-24 (%) 15-19 (%) 20-24 (%) 15-19 (%) 20-24 (%) ECA - All 13.7 30.0 12.2 24.2 11.3 26.3 BMU 12.9 35.3 11.5 31.9 11.8 33.6 Central Asia (C.A.) NA NA NA NA 17.7 39.1 EU11 10.4 26.2 6.6 14.2 9.1 21.0 Russian Federation (R.F.) NA NA NA NA 5.1 13.7 South Caucasus (S.C.) NA NA 36.1 57.8 NA NA South Eastern Europe (S.E.E.) 27.4 38.5 21.1 39.2 13.2 31.9 Turkey 34.3 48.1 28.5 44.5 21.7 41.0 Source: Author's Calculations Table 7: Youth (15-24) NEET Rates by Gender over Time - By Sub-region 2001-2002 Pre-2009 Post - 2009 Region/Subregion Male (%) Female (%) Male (%) Female (%) Male (%) Female (%) ECA - All 19.4 23.7 16.2 20.9 18.2 20.9 BMU 23.5 22.7 22.2 21.5 23.7 22.9 Central Asia (C.A.) NA NA NA NA 18.8 36.9 EU11 17.2 19.5 9.4 12.9 16.7 15.9 Russian Federation (R.F.) NA NA NA NA 10.6 8.5 South Caucasus (S.C.) NA NA 49.1 45.0 NA NA 42 South Eastern Europe (S.E.E.) 26.6 35.9 26.2 32.9 23.4 23.3 Turkey 25.7 55.7 19.8 51.4 17.2 43.3 Source: Author's Calculations 43 Table 8: Youth (15-19) NEET Rates by Gender 44 Table 9: Youth (15-24) NEET Rates by Gender Over Time 45 Table 10: Youth (15-24) NEET Rates by Educational Attainment 46 Table 11: Descriptive Statistics of the Youth (15-24) NEET 47 Table 12: Descriptive Statistics of the Youth (15-24) non-NEET 48 Table 13: Regression Analysis for Probability of being NEET among Youths (15-24) 49 Table 14: Descriptive Statistics for NEET Youths (15-24) using the Household Budget Survey in Six Countries 50 Table 15: Descriptive Statistics for non-NEET Youths (15-24) using the Household Budget Survey in Six Countries 51 Table 16: Regression Analysis for probability of being NEET among Youths (15-24) using the Household Budget Survey 52 8 Figures Figure 1: Youth (15-24) and Adult Unemployment Rates for the EU27 (Source: Eurostat) 25 20 15 UR (ages 25-74) 10 YUR (ages 15-24) 5 0 53 NEET Rate NEET Rate 10 20 30 40 60 70 10 20 30 40 50 60 0 50 0 ECA ECA and Post-2009 Moldova Moldova Ukraine Ukraine Kyrgyz Republic Kyrgyz Republic Tajikistan Tajikistan Bulgaria Bulgaria Czech Republic Czech Republic Estonia Estonia Hungary Hungary Latvia Latvia Lithuania Lithuania Country Country Poland Poland 54 Romania Romania Slovak Republic Slovak Republic Pre-2009 2001-2002 Slovenia Slovenia Russia Russia Armenia Armenia Albania Albania Macedonia Macedonia Serbia Serbia Turkey Turkey 20-24 (%) 20-24 (%) 15-19 (%) 15-19 (%) Figure 2: Youth NEET Rates for 15-19 and 20-24 year olds in 2001-2002, Pre-2009 NEET Rate NEET Rate 10 20 30 40 50 10 20 30 40 50 60 0 60 0 ECA ECA Moldova Moldova Ukraine Ukraine Kyrgyz Republic Kyrgyz Republic Tajikistan Tajikistan Bulgaria Bulgaria Czech Republic Czech Republic Estonia Estonia Hungary Hungary Latvia Latvia Lithuania Lithuania Poland Poland 55 Romania Romania Slovak Republic Slovak Republic Post-2009 Slovenia Slovenia Russia Russia Armenia Armenia Albania Figure 3: Youth (15-24) NEET Rates by Gender - Post 2009 Albania Macedonia Macedonia Serbia Serbia Turkey Turkey Male (%) 20-24 (%) 15-19 (%) Female (%) NEET Rate 0.00 10.00 20.00 30.00 40.00 50.00 60.00 10.00 20.00 30.00 40.00 50.00 60.00 0.00 ECA Moldova ECA Moldova Ukraine Ukraine Kyrgyz Republic Kyrgyz Republic Tajikistan Tajikistan Bulgaria Bulgaria Czech Republic Czech Republic Estonia Estonia Hungary Hungary Latvia Latvia Lithuania Lithuania Poland 56 Country Poland Romania Romania Slovak Republic Slovak Republic Slovenia Slovenia Russia Figure 4: Male Youth (15-24) NEET Rates Over Time Russia Figure 5: Female Youth (15-24) NEET Rates Over Time Armenia Armenia Albania Albania Macedonia Macedonia Serbia Serbia Turkey Turkey Pre-2009 Pre-2009 2001-2002 2001-2002 Post - 2009 Post - 2009 Figure 6: Youth (15-24) Activity Breakdown for Pre-2009 period 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Country Enrolled in School (%) Employed (%) NEET (inactive) (%) NEET (unemployed) (%) 57 Figure 7: Youth (15-24) Activity Breakdown - Post 2009 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Country Enrolled in School (%) Employed (%) NEET (inactive) (%) NEET (unemployed) (%) 58 Figure 8: Youth (15-19) Activity Breakdown - Most Recent Data 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Enrolled in School (%) Employed (%) NEET (inactive) (%) NEET (unemployed) (%) 59 Figure 9: Youth (20-24) Activity Breakdown - Most Recent Data 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Country Enrolled in School (%) Employed (%) NEET (inactive) (%) NEET (unemployed) (%) 60 Figure 10: Male Youth (15-24) Activity Breakdown - Most Recent Data 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Country Enrolled in School (%) Employed (%) NEET (inactive) (%) NEET (unemployed) (%) Figure 11: Female Youth (15-24) Activity Breakdown - Most Recent Data 61 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Country Enrolled in School (%) Employed (%) NEET (inactive) (%) NEET (unemployed) (%) 62 9 Appendix Table A 1: Youth and Adult Unemployment Rates for select countries in the ECA 10 63 Table A 2: List of Years and Surveys Used for Each Time Period 64 Table A 3: Youth NEET Rates for those ages 15-19 & 20-24 65 Table A 4: Youth NEET Rates by Gender 66 Table A 5: Youth (15-24) by Enrollment, NEET (inactive) and NEET (unemployed) 67 Table A 6: Youth (15-19) by Enrollment, NEET (inactive) and NEET (unemployed) 68 Table A 7: Youth (20-24) by Enrollment, NEET (inactive) and NEET (unemployed) 69 Table A 8: Male Youth (15-24) by Enrollment, NEET (inactive) and NEET (unemployed) 70 Table A 9: Female Youth (15-24) by Enrollment, NEET (inactive) and NEET (unemployed) 71 Table A 10: List of Public Transfers in each Country using the Household Budget Survey in Six Countries. Poland Russia Serbia Tajikistan Turkey Old Age Pension Old Age Pension Disability Insurance Old Age Pension Old Age Pension Disability Insurance Social Assistance Health & Life Insurance Disability Insurance Other Pensions Health & Life Insurance Social Insurance Survivor Insurance Social Assistance Maternity Other Pensions Special Merit Pension Scholarship Social Insurance Social Assistance Social Assistance Housing & Utilities Other Pensions Child & Family Child & Family Unemployment Benefits Social Assistance Scholarship Subsidy for Veterans Child & Family Housing & Utilities Scholarship Unemployment Benefits Housing & Utilities Unemployment Benefits 72