WPS7548 Policy Research Working Paper 7548 Out of School and Out of Work A Diagnostic of Ninis in Latin America Rafael de Hoyos Anna Popova Halsey Rogers Education Global Practice Group January 2016 Policy Research Working Paper 7548 Abstract Using all the household survey data available in Latin Amer- of ninis rose over the two-decade period, even as women’s ica during the period 1992 to 2013, this paper estimates education and employment rates were improving. Global that in 2015, 20 million youth ages 15 to 24 years in the comparisons show that Latin America is the region of the region were out of school and not working (making them world with the largest concentration of ninis among house- ninis, for “ni estudian ni trabajan”). The share of out-of- holds in the bottom 40 percent of the income distribution. school, out-of-work youth in Latin America, at about 19 Coupled with the long-lasting harm it causes to the youth’s percent, is roughly equal to the global average of 22 percent. future labor-market outcomes, the high incidence of ninis Although women make up over two-thirds of the ninis in among the poorest households tends to lock in income the region, the number of male ninis grew by 46 percent disparities from one generation to the next, obstruct- between 1992 and 2010. As a result, the absolute number ing social mobility and poverty reduction in the region. This paper is a product of the Education Global Practice Group. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at rdehoyos@ worldbank.org The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Out of School and Out of Work: A Diagnostic of Ninis in Latin America1 Rafael de Hoyos, Anna Popova, and Halsey Rogers The World Bank JEL classification: J13, O15, R23 Keywords: Latin America, Ninis, dropouts, youth unemployment 1 The present study is a background paper for the regional study “Out of school and out of work: Risk and opportunities for Latin America’s ninis.” (De Hoyos, Rogers, and Székely 2016). Earlier versions of this paper benefited substantially from comments and suggestions by Julian Messina, Reema Nayar, Kathleen Beegle, Michael Crawford, Harry Patrinos, Wendy Cunningham and participants of two workshops held at the World Bank in Washington DC in October 2013 and March 2014. The authors are especially indebted to Augusto de la Torre and Daniel Lederman for suggesting the introduction of global comparisons and overall guidance. We thank Israel Osorio for granting us access to the GIDD data and to Andrés Pérez for providing able research assistance. 1. Introduction Between the late 1990s and 2010, Latin America and the Caribbean (LAC) achieved substantial economic progress: GDP per capita grew by more than 50 percent in real terms (ECLAC), the proportion of poor individuals fell from 27.5 percent in 1992 to 12.6 percent in 2011 (SEDLAC, 2013), and inequality as measured by the Gini index fell by approximately 10 percent during the 2000s alone. Yet even as the region registered these gains, the attention of its public and policymakers has increasingly focused on an economic and social problem faced by countries throughout the region. This problem is the large number of youth who are neither in school nor at work—youth commonly labeled as ninis,2 from the Spanish phrase “ni estudian ni trabajan”—and it has proved persistent and seemingly structural. Unless it is understood better and tackled through policy interventions, the problem of ninis could limit the region’s ability to reap the dividend from its demographic transition, as its large youth cohort moves from dependency into its prime working years. To address the nini issue, policy makers in the region need to understand how broad the problem is, whether it is growing or subsiding, and who the ninis are. Yet until now, there has been little analysis of the nini phenomenon for the Latin America and Caribbean region as a whole.3 To provide such a diagnosis of the 15 to 24 year-old ninis, this study draws on micro data from 238 household surveys for 15 Latin American countries spanning the period 1992–2010. For these country-year combinations, we use harmonized data from SEDLAC, the Socio-Economic Database for Latin America and the Caribbean maintained by the Centro de Estudios Distributivos Laborales y Sociales (CEDLAS) and the World Bank, supplemented by surveys assembled in Cárdenas, de Hoyos, and Székely (2015), which include comparable variables for school attendance, economic activity, and socio-demographic characteristics. The resulting quantitative description of ninis in Latin America is enriched by a comparison with the prevalence and characteristics of ninis in other parts of the world. This newly expanded dataset allows an unprecedentedly complete portrait of the problem. Specifically, this study answers the following questions: What is the prevalence of ninis in LAC and how has it evolved over time? How do the share and characteristics of ninis in Latin America compare with those in other regions of the world? How does the likelihood of being a nini differ by age and gender? How many ninis are actively searching for work, versus choosing to be ninis or having disengaged entirely? And do these patterns vary with the decision to start a household? 2 In Brazil, they are known by the corresponding Portuguese term, nem-nem. 3 One exception is the recent analysis by Cárdenas, de Hoyos, and Székely (2015), on which this study builds. 2 The answers will take us closer to understanding the economic and social causes of the detachment of the youth from education and the labor market, while also highlighting the heterogeneity among different categories of youth, and thus the need for differentiated policy responses to support those youth. If the nini phenomenon reflects an inability to provide appropriate educational and employment opportunities for the youth, then there are policies to improve those opportunities that could help the region take greater advantage of its window of demographic opportunity. For most youth, these years of the life cycle are characterized by change and vulnerability, by the development of self-esteem and sense of belonging, by a need for support from health services, specialized supervision, and specialized orientation, and by exposure to risks. Youth who are divorced from the education system and the workforce in these years may find it particularly hard to fulfill their potential. The result could be lower societal productivity and growth (making it harder to support those over 65 when the window closes in the future), and a high nini rate could also contribute to crime, addiction, insecurity, disruptive behavior, and lower social cohesion. The paper is structured as follows. The next section describes the definition of ninis and the data used to quantify them in Latin America and elsewhere, and presents the overall estimate of the number of ninis in the region. Section 3 presents the evolution of ninis in Latin America over the period 1990 to 2010, and also shows how the composition of the nini population—broken down by gender, income, urban/rural location, and education level—has changed over time. Section 4 sets the Latin American experience in global perspective, by presenting new survey-based calculations of the number of out-of-work, out-of-school youth in each region of the world and comparing these to the Latin American case. To help target policy responses, Section 5 offers much more detail on the types of youth who are ninis, breaking down the total by gender, age, household status, and labor-market participation, as well as tracking how members of these different subgroups have become more or less likely to be ninis over time. Section 6 summarizes the key messages and their implications. 2. Quantifying the Ninis The first step of this analysis is to document the extent of the nini problem in Latin America. The growth in media and policy attention to ninis in recent years need not necessarily reflect the true magnitude of the issue, which is an empirical question that must be carefully addressed. In this section, we present the most comprehensive quantification to date of ninis in Latin America and the Caribbean. 3 Data and methodology To carry out this analysis, we draw on micro data from 238 household surveys for 15 Latin American countries spanning the period from 1992 to 2013.4 For these country-years, we use harmonized data from SEDLAC, the Socio-Economic Database for Latin America and the Caribbean (CEDLAS and the World Bank 2015), which include homogenous variables for school attendance, economic activity, and socio-demographic characteristics. 5 We use these data to produce internationally comparable statistics on ninis, defined as those individuals aged 15 to 24 who were neither enrolled in school nor working at the time when the relevant household survey was conducted. Two qualifications deserve mention before we present the results. First, like the underlying variables from the original household surveys, nini status is self-reported. Second, not all countries are equally represented in the data used for this analysis, since household surveys are conducted with varying frequency across countries and the quality of harmonization varies across country-years. We exclude countries for which the number of years between survey data available is so great that interpolating would produce unreliable results. The table in Annex 1 specifies the particular surveys processed for each country. To construct the averages, we began with the full period for which SEDLAC data are available— 1980 to 2013 for the region as a whole. However, because several countries have missing data from the beginning and end of this period, particularly for the disaggregated analysis, we confine most of our analysis to 1992-2010, which is the longest time period with reasonably complete data. This year range provides nearly two decades of evolution of the nini phenomenon across the region, while also minimizing the need to interpolate missing years and to extrapolate back for countries that lack survey data in the early years. The regional figures presented in this note are the unweighted averages for the same 15 Latin American countries throughout. We focus on unweighted averages because they capture more accurately the experience of the average country in the region and avoid letting the largest countries 4 Argentina, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador, Honduras, Mexico, Panama, Paraguay, Peru, Uruguay, and República Bolivariana de Venezuela. Because of data availability, this report focuses primarily on Latin America and covers only one country in the Caribbean, the Dominican Republic. 5 The SEDLAC database was developed in a collaboration between the Centro de Estudios Distributivos, Laborales y Sociales, Universidad Nacional de La Plata and the World Bank. In our analysis we use an augmented SEDLAC database; of the 238 surveys used in our analysis, 24 are original country household surveys for which we manually harmonized variables of interest to make them consistent with SEDLAC data definitions. Please refer to highlighted cells in the table in Annex 1 for the specific country-years that rely on these additional surveys. 4 in the region, notably Brazil and Mexico, drive the results. As it turns out, in practice the weighted averages are quite similar to the unweighted averages that we present. Annex 2 compares the unweighted and weighted results for the evolution of the proportion of ninis aged 15 to 24 years between 1992 and 2010 and shows very similar levels and trends between the two. The same is true for the more disaggregated measures presented in this paper (such as the distribution of ninis by gender and education level), though they are not reported here. Defining and quantifying the nini population We begin our analysis by categorizing the population aged 15 to 24 in Latin America in 2010 according to their education and work status to establish the order of magnitude of the nini phenomenon in the region. Here ninis are the residual category after those youth studying or working (or both) have been accounted for. That is, we define ninis as 15 to 24 year-olds who are neither enrolled in formal schooling (whether public or private) nor working at the time they are surveyed. Working youth are defined—following the SEDLAC methodology—as those individuals who have worked at least one hour in the reference period of the given survey (typically the past week), as well as those who are employed but have not worked during the reference period due to extraordinary circumstances (such as illness, strike, or vacation). One critical point worth noting here is that there exists no official academic or political consensus on what constitutes “work” in this context. This is particularly problematic given that we are defining a segment of the population not by what they do, but by what they do not do. Incorrectly categorizing individuals’ activities— for example, by labelling female homemakers as “not working”—may lead us to jump to the incorrect conclusion that all ninis are unproductive and perhaps even irrational. Putting aside these limitations and the heterogeneity of sub-populations for the moment, Figure 1 offers a first look at the activities of the youth of Latin America and the Caribbean according to the above definitions. It shows that of all youth aged 15 to 24 in 2010, 34 percent were only studying, 33 percent were only working, 12 percent were both studying and working, and 19 percent—or a total of over 18 million individuals in these 15 SECDLAC countries—were neither studying nor working. Scaling up our survey-based estimate of 18.2 million ninis to cover all countries in the region yields an estimated 20.8 million ninis for the Latin American and Caribbean region as a whole in 2015. 5 Figure 1: Breakdown of youth population aged 15 to 24 in LAC, 2010 Study only 34% Work only 33% Youth (aged 15 - 24) 92.2 million = 100% Study & Work 12% Ninis 19% Source: authors’ own computation with data from CEDLAS and The World Bank (2015) 3. Stylized facts: The evolution of ninis in Latin America6 Having identified that 1 in 5 Latin American youth are both out of school and out of work, we analyze how this rate has changed over the past 20 years, to understand whether the problem is becoming more or less severe, as well as how individual countries’ trends vary around this mean trend. Figure 2 presents the evolution of ninis aged 15 to 24 years in Latin America between 1992 and 2010, in both absolute terms and as a proportion of the relevant age demographic. The figure shows that the unweighted average share of ninis has decreased modestly over that period, from 23.4 percent in 1992 to 19.3 percent in 2010, which amounts to a decline of 4.1 percentage points. Over the same period, however, the number of ninis rose slightly from 16.4 million to 18.2 million, amounting to an increase of 14 percent (or 1.8 million individuals) in spite of their decline relative to the 15 to 24 year-old population. This trend of a declining share but rising number of ninis is driven by high population growth for this age group during the 1990s. 6 Although for most of the analysis in this report the latest year included is 2010, more household surveys became available as the study was being completed. To confirm the report’s findings, the key statistics presented here were updated to circa 2013, and all of the messages remain unchanged. For instance, the long term trend in the decline in the share of ninis remains reaching 19 percent in 2013, while the absolute number of ninis declines only marginally to less than 18 million in that same year. The most recent statistics are included in country-specific briefs and are available from https://openknowledge.worldbank.org/handle/10986/22349. 6 Figure 2: Number of ninis aged 15-24 and their share of age group in LAC, 1992-2010 100.0 20.0 18.0 18.2 18.2 Ninis (% relevant age group) 17.1 16.4 80.0 15.0 Ninis (millions) 60.0 10.0 40.0 23.4 23.2 22.6 20.5 19.3 5.0 20.0 0.0 0.0 1990 1995 2000 2005 2010 Year Number of ninis (right axis) Percentage of ninis in relevant age group (left axis) Source: authors’ own computation with data from CEDLAS and The World Bank (2015) A 14 percent increase in the number of ninis over a 20 year period may seem small, but disaggregating by gender reveals that the increase for young men was much more dramatic. The number of male ninis jumped 46 percent between 1992 and 2010, with the additional males accounting for the entire increase in ninis throughout the period (1.8 million individuals). This is an important distinction, since the number of male ninis is correlated with crime and violence in contexts where violence is already a serious problem (de Hoyos, Gutiérrez, and Vargas 2015). The combination of this surge in male ninis, the link with crime and violence, and the very emergence of the term “ninis” may explain the increased media attention to the phenomenon in recent years.7 Not all countries show the same trend, of course. The change in the national percentage of ninis between 1992 and 2010 ranges from a decrease of 12.3 percentage points in República Bolivariana de Venezuela to an increase of 1.6 percentage points in Argentina. For each country, Annex 4 provides graphs showing how the total number and percentage of ninis aged 15 to 24 have evolved over that period, while Figure 3 and Annex 3 summarize these national levels and trends using maps. Using data for 2010, Figure 3 shows that Peru and República Bolivariana de Venezuela have the lowest shares of ninis (at 10.9 and 13.0 percent, respectively), while the highest shares are concentrated around Central America. The highest numbers of ninis are found in Brazil, Colombia, 7 According to an analysis using Google Trends, searches from Mexico using the term “ninis” surged in late 2010, after having been non-existent before that year. Moreover, this was not simply due to economic circumstances: searches using “desempleado” (unemployed) had already risen sharply some two years earlier, just after the global financial crisis hit. 7 and Mexico (Annex 4), primarily because of the larger populations of those countries. In terms of the evolution of ninis since 1992, Peru and República Bolivariana de Venezuela have experienced the greatest reductions in ninis in both percentage and absolute terms (Annex 4). Figure 3: Percentage of ninis in LAC among those aged 15-24, circa 2010 Source: authors’ own computation with data from CEDLAS and The World Bank (2015) To probe exactly where among the nini population our greatest concern should lie, we next analyze the socio-demographic characteristics of the nini population, by urban/rural location, income, and education levels, as well as how that composition has changed over the past two decades. We analyze the incidence of ninis by rural/urban location and by income decile calculated using household income per capita, as defined by the SEDLAC project. Rural/urban: Figure 4 shows the incidence of ninis in rural and urban areas and how this has changed since 1992. A greater proportion of rural than urban youth are ninis: 21.3 percent in rural areas were ninis in 2010, compared to 16.8 percent in urban areas. This rural-urban gap has actually grown over the past 18 years, even as the proportion of ninis in both categories has decreased. Although as a share of population, ninis are more prevalent in rural areas than in urban areas, given the high urbanization rate in the region, the vast majority of ninis—close to 13 million of the 18 million total—live in cities. 8 Figure 4: Incidence of ninis by urban/rural status (ninis as % of youth aged 15-24) 30.0% 26.2% 25.0% 23.7% 21.3% 20.0% 16.8% 15.0% 10.0% 5.0% 0.0% Rural Urban Rural Urban 1992 2010 Source: authors’ own computation with data from CEDLAS and The World Bank (2015) Income: Figure 5 shows the incidence and distribution of ninis across the income distribution, as measured by deciles of household income per capita. As might be expected, youth from poor households are more likely to be ninis than their better-off peers. While the percentage of ninis in every decile has decreased over the past 18 years, there remains a huge disparity across income deciles: in 2010, 30.5 percent of ninis in the poorest deciles were neither working nor studying, compared with only 6.6 percent of those in the richest decile. Moreover, considered from the perspective of the distribution of the nini population across deciles, this disparity is growing. In 1992, an average of 36.8 percent of nini youth were in households in the three poorest deciles of the population, while 18.9 percent were in the three richest deciles. By 2010, the share of the three poorest deciles had grown to 45.1 percent of ninis, versus only 13.5 percent in the three richest deciles. Thus on average across the region, the nini problem has become increasingly concentrated among the poor over the past two decades. 9 Figure 5: Incidence and distribution of ninis by income decile Panel A: Incidence of ninis by income per capita decile (ninis as % of youth aged 15-24) 40.0% 36.6% 34.7% 35.0% 30.3% 30.5% 30.0% 27.9% 28.2% 25.9% 25.9% 24.2% 23.2% 25.0% 20.2% 21.2% 20.0% 17.3% 18.1% 14.8% 15.0% 13.1% 12.0% 9.8% 9.4% 10.0% 6.6% 5.0% 0.0% D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 1992 2010 Panel B: Distribution of ninis by income decile (ninis in given income decile as % of ninis aged 15-24) D1 D2 D3 D4 1992 D5 12.5% 11.5% 12.8% 11.7% 11.3% 11.4% 9.9% 8.5% 6.5% 4.0% D6 D7 D8 D9 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% D10 D1 D2 D3 D4 2010 D5 16.5% 14.7% 13.8% 13.0% 11.3% 9.8% 7.4% 6.3% 4.5% 2.8% D6 D7 D8 D9 D10 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Source: authors’ own computation with data from CEDLAS and The World Bank (2015) 10 Education: Similarly, we analyze the profile of ninis in terms of educational attainment and how this has changed over the past two decades. Figure 6 shows the breakdown of the ninis by highest level of education completed. We present this analysis disaggregated into two subgroups, 15 to 18 year-olds and 19 to 24 year-olds, because ninis in the younger group are unlikely to have any higher education. While there are some differences in levels, education patterns of both groups have evolved similarly over the past two decades. For both age groups, the proportion of ninis with incomplete primary education has fallen since 1992; nevertheless, in 2010 over 15 percent of ninis in both groups still had not completed primary school, and over 20 percent had not finished secondary. Moreover, the incomplete-secondary group has grown substantially since 1992 for ninis aged 15 to 18 years old. This figure suggests that being a 15 to 18 year-old nini in LAC is increasingly associated with dropping out of school, and with incomplete secondary education in particular, which increases the probability of joining this vulnerable group. For 19 to 24 year-old ninis, Panel B in Figure 6 shows that an increasing share of them have complete secondary, suggesting that among those who manage to finish middle school, the quality of education received is not enough for them to successfully insert themselves into the labor market. Figure 6: Percentage of ninis in LAC by education level and age, 1992-2010 Panel A: 15-18 years Panel B: 19-24 years Source: authors’ own computation with data from CEDLAS and The World Bank (2015) 11 4. Latin American ninis in global perspective How do the dimensions of Latin America’s nini problem compare to those of other regions? Using a collection of household surveys at the global level standardized by the World Bank’s Global Income Distribution Dynamics (GIDD) project,8 we are able to quantify the total number of youth out of school and not working at the global level. To compute the total number of ninis at the global level, we rely on two main assumptions. First, we assume that the share of ninis in countries for which we did not have a reliable household survey close to 2010 is the same as the average for the region to which they belong. For instance, the share of ninis in Bolivia is assumed to be 19.3 percent, which is the average share for Latin America and the Caribbean. Second, we assume that the share of ninis remained constant between 2010—the last year for which a reliable household survey is available for a large number of countries—and 2015. Our calculations thus take into account the total global youth population of 1.19 billion in 2015 and assume that the share of ninis is the same as that observed circa 2010. In 2015 there were some 260 million youth ages 15 to 24 out of school and not working.9 More than 35 percent of the world’s ninis live in South Asia, and more than one fifth of them in East Asia and the Pacific. Figure 7 shows the share of youth aged 15 to 24 who are ninis in all countries in the world where household data is available (circa 2010). The countries with the highest incidence of ninis are Iraq (51 percent), Guyana (44 percent), Nigeria (43 percent), and Bangladesh (41 percent), each with shares of more than 40 percent. At the other end of the spectrum, Sweden (4.2 percent), Slovenia (4.7 percent) and Luxembourg (4.8 percent) have the lowest shares of ninis. The region with the highest share of ninis is Middle East and North Africa, where one in three youth are out of school and out of work, while the lowest share is in high-income countries (11 percent). The share of ninis in Latin America is close to the global average of 22.4 percent. The GIDD data also reveal that the gender breakdown of the nini problem varies greatly across country groups: in East Asia and the Pacific, in Europe and Central Asia, and in high-income countries, woman are no more likely to be ninis than are men. Meanwhile, in South Asia women account for 82 percent of total ninis, and in the Middle East and North Africa and Sub-Saharan Africa the female share of ninis is similar to that of Latin America (two-thirds). 8 For details on the GIDD data and methods, see: http://go.worldbank.org/YADEAFEJ30. 9 Note that this global estimate is broadly consistent with the figure of 224 million calculated in a recent unpublished paper by Newhouse (2013). 12 Figure 7: Percentage of ninis in all countries in the world, circa 2010 Source: authors’ own computation with data from World Bank (2015) The global benchmarking also reveals that in all developing regions except Europe and Central Asia, more than two-thirds of ninis have not completed secondary school, and that less than 10 percent have any post-secondary education.10 In terms of rural/urban breakdown, Latin America is the region with the highest proportion of ninis located in urban areas. Among other developing regions, only the Middle East and North Africa has a majority of ninis in urban areas; in the remaining regions, the share is less than 40 percent. In terms of nini incidence in the bottom 40 percent of the income distribution, Latin America registers the largest share with 64 percent, followed by East Asia and the Pacific (58 percent), Sub-Saharan Africa (49 percent), Europe and Central Asia (44 percent), and high-income countries (32 percent). 5. Who are ninis, how has this changed, and why does this matter? Having acquired a sense of the scale of the nini problem and where they are situated—in terms of geography, income, and education—and a comparison with other regions in the world, we turn now to a decomposition of the nini population that highlights their heterogeneity. To differentiate between types of ninis and what is driving their condition, we disaggregate ninis by gender, age group, and economic activity, as well as by the decision to start their own household, both with and without children. In terms of economic activity, we zoom in on ninis not searching for work (defined as those ninis who are not classified as unemployed by CEDLAS, because they have not actively 10 In Europe and Central Asia, only 48 percent of the total youth classified as ninis does not have complete secondary education. 13 looked for work in the period of reference). To investigate whether youth who decide to start their own household are more likely to be ninis, we examine those ninis who are either a household head, the partner of the household head, or married or in a civil partnership (to anyone, not necessarily the household head). We delve further into the relationship between nininess and family characteristics by looking at how having children affects the likelihood of being a nini.11 Figure 8 summarizes this decomposition of the nini population for those aged 15 to 24 in LAC in 2010. The various categories of the nini population are presented in percentage terms, allowing us to trace the problem through this cross-section to identify which categories—each with differing motivations for nininess attached to them—the majority of ninis fall into. At any vertical point in the flowchart, the proportions of ninis in each category sum to 100 percent of ninis aged 15 to 24 years old (a total of 18.2 million youth in 2010 for the countries for which we can reliably disaggregate data). This exercise can provide a better sense of where the problem is concentrated, whether it is in fact a problem at all, and if so, which specific areas policies aimed at decreasing so- called “idleness” could focus on. The most striking lesson from the cross-sectional breakdown in Figure 8 is that over two-thirds of ninis (68.6 percent) are women—the large majority of whom either live with their parents and are not looking for work (35.5 percent of all ninis) or have started a new household with children (17.6 percent of all ninis). In contrast, men make up only 31.4 percent of the nini population, and almost all of them (29.7 percent of all ninis) are young males still living with their parents, with over half of that group not actively searching for work. These patterns suggest that it may be wise to target anti-dropout/employment programs (depending on the age profile and individual context) at both young male and female ninis who are living with their parents and are not searching for work, perhaps because they lack skills, information, or motivation. In contrast, the nature of the nini phenomenon among women starting a household with children is intuitively very different: it could reflect a rational personal choice, and reducing the number of ninis in this category will likely involve different policies, including child care and perhaps teenage pregnancy prevention programs. 11 One caveat on the issue of children: due to a limitation of the SEDLAC harmonized data, we are unfortunately unable to identify the parentage of children unless they are those of the household head. We attempt to overcome this limitation by identifying those ninis who live with at least one child aged 5 years or under (the age range most likely relevant in prompting parents to stay at home to care for children) as an approximation for those that have children in this age range. Thus, for example, when we analyze trends for ninis who decide to start their own household with children, we are actually including any nini who is either a household head, partner of the household head, or married or in a civil partnership, and who is living with at least one child age 5 or under. 14 Figure 8: Breakdown of the ninis population aged 15 to 24 in LAC, 2010 Look for work Remain in HH 13.2% 29.7% Do not look for work 16.5% Look for work 0.4% Start new HH Male without children Do not look for work 31.4% 0.7% 0.3% Look for work Start new HH with 0.7% children 1% Do not look for work Ninis 0.3% (aged 15 - 24) 100% Look for work Remain in HH 11% 45.5% Do not look for work 35.5% Start new HH Look for work Female without children 0.6% 68.6% 4.4% Do not look for work 3.8% Look for work Start new HH with children 1.1% 18.7% Do not look for work 17.6% Source: authors’ own computation with data from CEDLAS and The World Bank (2015) In the remainder of this analysis, we develop the cross-sectional breakdown from Figure 8 into a more detailed, longitudinal analysis to observe how the composition of ninis across these dimensions has changed over the past two decades. Here again, we analyze trends for the 15 to 18 and 19 to 24 age groups separately, to reflect the possibility that school dropout and labor market participation decisions differ in nature across these subgroups. From a policy perspective, being a nini between the ages of 15 and 18 is particularly worrisome, because at this stage in the life cycle many societies have affirmed that formal education is the most socially productive activity in which youth can be engaged. Youth under 18 years of age are still of school age and are still developing physically, mentally, and emotionally. During these critical years, being in school in a protected and constructive environment contributes to the development of cognitive and socio-emotional skills, and it can also be a mechanism for becoming integrated into the community, acquiring social values, and developing trust in institutions and the rule of law. In contrast, the story for 19 to 24 year-old ninis may be more one of youth unemployment; nevertheless, we present results for both subgroups both to understand the heterogeneity of ninis as a group, and for comparison purposes. 15 Age: Figure 9 disaggregates the total ninis trend by age subgroup, splitting ninis into those aged 15 to 18 and those aged 19 to 24. The central result that emerges from this picture is that while the share of ninis is declining across age groups in LAC (an average decrease of 2.9 and 3.9 percentage points for those aged 15 to 18 and 19 to 24, respectively, between 1992 and 2010, and a consistently downward trend across countries), the rising total number of ninis in LAC has been driven by a slight increase in ninis in the 19 to 24 age bracket. While the number of ninis aged 15 to 18 in LAC remained roughly constant from 1992 to 2010 at between 5.8 and 5.9 million, the number of those aged 19 to 24 increased by 1.7 million over the same period. Moreover, the majority of ninis are concentrated in the 19 to 24 age group (which accounted for 68 percent of total ninis in 2010), both because of the wider age bracket it covers and because this older group is not covered by compulsory schooling laws; as a result, the increasing number of ninis within this subgroup has more effect on aggregate nini trends. Figure 9: Total number and percentage of youth who are ninis in LAC by age, 1992-2010 Panel A: Percentage of youth who are ninis Panel B: Number of ninis in LAC by age in LAC by age group group 30.0 20.0 Ninis (% relevant age group) 18.0 18.2 18.2 27.3 27.0 17.1 26.5 16.4 25.0 15.0 Ninis (Millions) 24.3 23.4 23.2 23.4 22.6 12.3 12.1 12.3 11.1 20.5 10.6 20.0 10.0 19.3 18.4 18.3 17.6 15.6 15.5 5.8 6.0 5.9 6.0 5.9 15.0 5.0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Year Year % ninis 15-18 % ninis 19-24 Ninis 15-18 Ninis 19-24 % ninis 15-24 Ninis 15-24 Source: authors’ own computation with data from CEDLAS and The World Bank (2015) Gender: Figure 10 presents a breakdown of nini trends by gender in both percentage and absolute terms, and for both the 15 to 18 and the 19 to 24 age groups. The results show that while detachment from both education and the labor market is still more common among women than among men, female ninis have decreased both as a share of young women and in absolute terms over the period from 1992 to 2010. Moreover, this reduction in female ninis drove the overall reduction in the share 16 of ninis over this period; in fact, the share of male ninis increased marginally during those years (for both age groups). And over the same period, the increase of 14 percent in the total number of ninis (1.8 million individuals) is entirely explained by the increase in the number of male ninis. Simultaneously, the number of female ninis has fallen in the case of those aged 15 to 18 (decreasing by 5.6 percentage points, or over 400,000 individuals) and stayed relatively constant for those aged 19 to 24. These trends, which are likely explained by the general increase in women’s education levels and labor-force participation rates in the region, combine to produce a shift in the gender composition of ninis towards men. Figure 10: Total number and percentage of youth who are ninis in LAC by gender and age, 1992-2010 Panel A: Percentage of youth who are ninis in Panel B: Number of ninis in LAC by LAC by gender, (15-18 years) gender, (15-18 years) 25.4 6.0 25.0 24.9 6.0 6.0 5.8 5.9 5.9 Ninis (% relevant population) 23.3 5.0 20.5 Ninis (Millions) 20.0 19.8 18.4 18.3 4.1 4.1 4.0 4.0 17.6 3.8 3.7 15.6 15.5 15.0 3.0 11.7 12.0 2.1 2.2 11.1 11.0 11.5 2.0 1.9 1.9 1.7 10.0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Year Year Total Men Total Men Women Women Panel C: Percentage of youth who are ninis in LAC by Panel D: Number of ninis in LAC by gender, (19-24 years) gender, (19-24 years) 41.1 41.2 12.1 12.3 12.3 6.0 8.0 10.0 12.0 40.0 Ninis (% relevant population) 37.7 11.1 35.6 10.6 33.5 9.1 9.0 Ninis (Millions) 8.8 8.8 30.0 8.5 27.3 27.0 26.5 24.3 23.4 20.0 4.0 14.7 3.2 3.5 12.6 12.9 13.1 3.0 12.2 2.2 2.3 10.0 2.0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Year Year Total Men Total Men Women Women Source: authors’ own computation with data from CEDLAS and The World Bank (2015) 17 Job search: Figure 11 characterizes the ninis according to their economic activity—specifically, whether or not they are searching for work. The clear picture is that the overwhelming majority of ninis across ages do not look for work; the only exception among the demographic groups is the male ninis aged 19 to 24, 55% of whom are looking for work. Job search is less common among younger ninis (15 to 18 years old), whose nini status may stem more from educational than labor- market factors. The overall low rate of job search among ninis is driven predominantly by a lower rate among women. At the same time, the absolute number of young female ninis who are not looking for jobs has decreased since 1992, perhaps because of women’s increased education and labor-force participation. During the same period, the absolute number of male ninis not looking for a job has increased. These trends highlight one element that may drive heterogeneity across ninis: the decision to start and look after a new household, and the differing demands that this decision places on the time of young men and women. The number of ninis not actively searching for work remained fairly constant from 1992 to 2010 for both age ranges, as slight decreases for women were offset by slight increases for men. At the same time, the proportion increased slightly for 15 to 18 year-olds and fell for 19 to 24 year-olds, among both sexes. 18 Figure 11: Total number and percentage of ninis not searching for work in LAC by gender and age, 1992-2010 Panel A: Percentage of ninis not searching Panel B: Number of ninis not searching for 50.0 60.0 70.0 80.0 90.0 for work in LAC by gender (15-18 years) work in LAC by gender (15-18 years) 5.0 87.6 87.4 (% relevant Ninis population) Ninis not searching for work Ninis not searching for work 85.1 84.8 4.5 4.5 83.2 4.4 4.3 4.4 81.1 79.7 4.0 76.9 77.5 3.5 3.5 74.0 3.3 (Millions) 3.1 3.0 69.7 3.0 63.3 63.5 60.6 2.0 55.0 1.4 1.2 1.0 1.0 1.1 1.0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Year Year Total Men Total Men Women Women Panel C: Percentage of ninis not searching Panel D: Number of ninis not searching for for work in LAC by gender (19-24 years) work in LAC by gender (19-24 years) 40.0 50.0 60.0 70.0 80.0 90.0 86.8 8.2 8.3 8.2 8.0 8.0 8.0 (% relevant Ninis population) Ninis not searching for work Ninis not searching for work 81.2 80.1 80.2 7.2 7.1 7.1 78.4 79.0 6.7 6.6 6.0 71.7 70.7 69.6 69.3 (Millions) 2.0 4.0 48.5 1.5 44.8 1.3 1.3 1.0 0.9 40.4 41.2 38.7 0.0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Year Year Total Men Total Men Women Women Source: authors’ own computation with data from CEDLAS and The World Bank (2015) Household status: Figure 12 and Figure 13 present the breakdown of ninis who decide to start their own households (defined as those ninis who are either a household head, partner of a household head, or married or in a civil partnership), both overall and with children. In both cases, the patterns are similar for both age groups, with a much greater percentage of female ninis starting households and starting households with children. While for 19 to 24 year-olds the percentage of ninis—be they male or female—starting their own households with or without children has risen over the past 18 19 years, the rise has been steeper among ninis aged 15 to 18 years, especially in the case of women. The increasing share of young women—particularly below 18 years of age—starting a new household, suggests that a potential driver of the high rate of female ninis is marriage before age 18, compounded by teenage pregnancy. Figure 12: Percentage of ninis that decide to start their own household in LAC by gender and age, 1992-2010 Panel A: 15-18 years Panel B: 19-24 years 30.0 52.7 52.9 27.7 10.0 20.0 30.0 40.0 50.0 48.5 Ninis starting own household (% relevant Ninis population) Ninis starting own household (% relevant Ninis population) 25.3 46.2 47.1 22.3 41.9 41.3 38.6 38.4 20.0 19.0 19.9 38.0 18.4 17.1 14.0 15.1 12.9 10.0 1.7 2.2 2.5 11.4 11.4 1.0 1.2 9.7 10.6 9.1 0.0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Year Year Total Men Total Men Women Women Source: authors’ own computation with data from CEDLAS and The World Bank (2015) Figure 13: Percentage of ninis that decide to start their own household with children in LAC by gender and age, 1992-2010 Panel A: 15-18 years Panel B: 19-24 years 10.0 15.0 20.0 0.0 10.0 20.0 30.0 40.0 50.0 18.4 (% relevant Ninis population) (% relevant Ninis population) 17.5 43.3 43.6 40.3 40.9 household with children household with children 15.8 39.2 Ninis starting own Ninis starting own 14.2 13.1 32.6 33.8 33.3 31.7 31.9 11.6 12.0 10.6 9.8 8.9 5.0 7.3 6.8 6.7 6.0 5.7 0.9 1.1 1.2 0.6 0.5 0.0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Year Year Total Men Total Men Women Women Source: authors’ own computation with data from CEDLAS and The World Bank (2015) 20 Household status and job search: Figure 14 and Figure 15 present those ninis who have decided to start their own household and who are not searching for work in two different ways: (1) as a proportion of the relevant nini population (in terms of age group and gender), and (2) as a proportion of ninis not searching for work (for corresponding age group and gender). Once again, this status is much more common among female than male ninis, with the female ninis in this category forming 25.7 percent of the population aged 15 to 18 and 47.2 percent of that aged 19 to 24, by 2010. The disparity between these rates for men and women has increased over time, especially among ninis aged 15 to 18 years old: the proportion of female ninis who have started their own household and who are not searching for work has increased substantially for both age groups, while the proportion among male ninis has remained fairly constant or even slightly decreased over time. This demographic has also increased as a proportion of female ninis not searching for work for both age groups. This may support the hypothesis that nini status among young women in particular is driven by a potentially rational desire to partake in domestic work; if this is the case, it may call for a specialized policy response. Figures 16 and 17 present the information in these same two ways, but zooming in on those ninis who decide to start a family with children. Starting a household with children and not searching for work is again much more common among female than male ninis, with female ninis in this condition forming 18 percent of the population aged 15 to 18 and 39.7 percent of that aged 19 to 24, by 2010, compared to rates of just 0.6 and 1.8 percent, respectively, among men. This gender disparity has increased slightly over time among ninis aged 15 to 18 years old, while remaining fairly constant for 19 to 24 year-olds. This demographic too has increased as a proportion of female ninis not searching for work for both age groups, reinforcing the hypothesis that young women in particular may choose to neither work nor study for rational reasons including a desire to care for their children. This reason may be very different from the factors driving young men to become ninis, and each may require equally tailored policy approaches. 21 Figure 14: Percentage of ninis who decide to start their own household and are not searching for work in LAC by gender and age, 1992-2010 Panel A: 15-18 years Panel B: 19-24 years 0.0 10.0 20.0 30.0 40.0 50.0 25.7 0.0 5.0 10.0 15.0 20.0 25.0 46.9 47.2 23.5 (% relevant Ninis population) (% relevant Ninis population) Ninis starting own household Ninis starting own household 43.5 43.5 41.6 & not searching for work & not searching for work 20.4 18.2 34.7 35.7 35.1 17.8 33.0 32.7 16.7 15.4 13.6 12.2 12.4 4.0 3.5 3.3 0.8 0.8 1.1 2.4 2.5 0.3 0.3 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Year Year Total Men Total Men Women Women Figure 15: Ninis who decide to start their own household and are not searching for work in LAC by gender and age, 1992-2010 (as % of ninis not searching for work) Panel A: 15-18 years Panel B: 19-24 years 60.0 30.0 29.2 58.7 58.2 (% Ninis in relevant population (% Ninis in relevant population 27.4 54.1 Ninis starting own household Ninis starting own household 50.6 51.9 50.9 & not searching for work & not searching for work 49.3 not searching for work) not searching for work) 23.8 46.2 46.3 21.8 44.5 21.0 20.5 40.0 20.0 19.7 17.5 16.6 15.4 20.0 10.0 7.6 8.7 7.9 5.8 6.2 1.3 1.2 1.6 0.4 0.5 0.0 0.0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Year Year Total Men Total Men Women Women Source: authors’ own computation with data from CEDLAS and The World Bank (2015) 22 Figure 16: Percentage of ninis who decide to start their own household with children and do not search for work in LAC, by gender and age, 1992-2010 Panel A: 15-18 years Panel B: 19-24 years with children & not searching for work with children & not searching for work 10.0 15.0 20.0 10.0 20.0 30.0 40.0 39.2 39.7 37.3 37.3 36.0 (% relevant Ninis population) (% relevant Ninis population) Ninis starting own household Ninis starting own household 17.5 16.2 14.7 29.5 28.5 29.6 29.3 27.9 12.5 12.8 11.2 10.5 9.8 8.4 8.8 5.0 2.4 1.4 1.7 2.0 1.8 0.1 0.2 0.4 0.4 0.6 0.0 0.0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Year Year Total Men Total Men Women Women Figure 17: Ninis who decide to start their own household with children and do not search for work in LAC, by gender and age, 1992-2010 (as % of ninis not searching for work) Panel A: 15-18 years Panel B: 19-24 years with children & not searching for work with children & not searching for work 0.0 10.0 20.0 30.0 40.0 50.0 10.0 15.0 20.0 19.9 48.9 49.0 (% Ninis in relevant population (% Ninis in relevant population 18.8 46.2 Ninis starting own household Ninis starting own household 43.2 45.0 17.2 42.1 41.1 not searching for work) not searching for work) 15.6 39.8 39.4 37.7 14.2 13.8 13.4 12.5 11.8 10.4 5.0 4.5 3.4 4.1 5.1 4.1 0.3 0.7 0.6 0.9 0.1 0.0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Year Year Total Men Total Men Women Women Source: authors’ own computation with data from CEDLAS and The World Bank (2015) 6. Conclusion This note has offered a descriptive diagnostic of the nini phenomenon in Latin America and the Caribbean—the patterns at a regional level and within individual countries, as well as the evolution of the phenomenon between 1992 and 2010. The analysis draws on data from 238 household surveys collected together and harmonized in the SEDLAC database supplemented by surveys assembled in Cárdenas, de Hoyos and Székely (2015). While this analysis is just a first cut at the data, it reveals 23 some interesting patterns and trends that could yield insights for education and labor policies in the region in several areas. First, the note provides new information on the size of the nini problem. In the 15 countries for which we have harmonized household survey data through the extended SEDLAC database, 18.2 million youth between the ages of 15 and 24 were living as ninis in 2010. Scaling up this survey- based estimate to cover all the countries in the region yields an estimated 20.8 million ninis for the Latin American and Caribbean region as a whole in 2015. Latin America is not an outlier when it comes to the issue of ninis, however. Globally, there are close to 260 million youth both out of school and out of work, spread across regions. The share of ninis in Latin America, at about 19 percent, is roughly equal to the global average of 22 percent. The region with the highest share of ninis is the Middle East and North Africa, where one in three youth is out of school and out of work, while the lowest share is in high-income countries (11 percent). Second, we provide information on the time trends of the nini phenomenon in Latin America. Increased prosperity over the past two decades has not solved the problem of ninis. For LAC as a whole, the percentage of youth who are ninis declined gradually between 1992 and 2010, but because the region’s youth population has increased, the number of ninis has actually grown slightly. Unsurprisingly, nini trends have varied across the region, with some countries showing sharp declines in the percentages of ninis over the 1992-2010 period and others seeing small increases. Third, we describe the characteristics of ninis. In gender terms, the nini phenomenon in LAC is largely a female phenomenon, but it is becoming more of a male problem over time. Women make up over two-thirds of ninis in LAC, but the number of male ninis jumped 46 percent between 1992 and 2010. As a result, males account for the entire regional increase in the absolute number of ninis over this period (1.8 million individuals). This surge, along with the documented correlation between male ninis and violence in some high-crime settings (De Hoyos, Gutierrez and Vargas 2015), may help explain the increased media and policy attention to Latin America’s ninis in recent years. On average, ninis are both less educated and poorer than their peers. As the education level of the general population has risen, so has that of ninis. Nevertheless, around 60 percent of even the 19 to 24 year-old ninis have failed to complete secondary school, pointing to a problem of dropout or inadequate progression to upper secondary. In addition, ninis are increasingly concentrated among the poorest households. A priori, this result is not obvious—it could be that only better-off youth can afford to be ninis—but the trend is unmistakable. In 1992, 36.8 percent of nini youth were in households in the three poorest deciles of the population, while half that number (18.9 percent) were in the three richest deciles. By 2010, the share of the three poorest deciles had grown to 45.1 percent of ninis, while the number in the three richest deciles had dropped to 13.5 percent. This high 24 incidence of ninis among poor households makes Latin America the region in the world with the largest concentration of ninis among households in the bottom 40 percent of the income distribution. Coupled with long-lasting harm to labor-market performance, these imbalances tend to lock in income disparities from one generation to the next, obstructing social mobility and poverty reduction in the region. Finally, we describe the multiplicity of possible causes of the nini phenomenon and what this implies for policy. The factors driving youth to become ninis are likely quite heterogeneous. As one example, women who have started a household and have children account for around 40 percent of 19 to 24 year-old female ninis, and half of these women are not looking for work. Intuitively, they are likely to be in a very different situation from younger female or male ninis who have not yet started their own household, and their situation will call for different policy responses (or perhaps for none at all). More generally, the findings of even this initial analysis reveal substantial heterogeneity within the nini population, and suggest that any policy approach to the nini phenomenon will need to be appropriately targeted and therefore multipronged. This descriptive diagnostic is only a first step toward facing the nini problem in the region, of course. For more on the topic, see De Hoyos, Rogers, and Székely (2016), which summarizes an extensive body of new research (including this paper) to flesh out the correlates, causes, and consequences of the nini problem and to suggest policy responses. 25 References Cárdenas, M., de Hoyos, R., and Székely, M. 2015. "Out of School and Out of Work Youth in Latin America: A Persistent Problem in a Decade of Prosperity." Economia, 16 (Fall 2016): 1 - 40 CEDLAS and The World Bank. 2015. Socio-Economic Database for Latin America and the Caribbean Centro de Estudios Distributivos, Laborales y Sociales. Universidad Nacional de La Plata and The World Bank. De Hoyos, Rafael, Carlos Gutierrez, and Vicente Vargas. 2015. “The Harmful Interaction between Economic Crisis, Violence, and Ninis in Mexico.” Background paper for the “Out of School, Out of Work” study. Washington, DC: World Bank. Available at https://openknowledge.worldbank.org/handle/10986/22349. De Hoyos, Rafael, Halsey Rogers, and Miguel Székely. 2016. Out of School and Out of Work: Risk and Opportunities for Latin America’s Ninis. Washington, DC: World Bank. Available at https://openknowledge.worldbank.org/handle/10986/22349. ECLAC. 2015. CEPALSTAT Databases. United Nations Economic Commission for Latin America and the Caribbean. http://estadisticas.cepal.org/cepalstat/WEB_CEPALSTAT/Portada.asp?idioma=i Newhouse, David. 2013. “New Estimates of Youth Idleness and Employment Outcomes in Developing Countries.” Unpublished manuscript, Washington, DC: World Bank. SEDLAC. 2013. CEDLAS Poverty Statistics. Socio-Economic Database for Latin America and the Caribbean. http://sedlac.econo.unlp.edu.ar/eng/statistics-detalle.php?idE=34 World Bank. 2015. GIDD database. Global Income Distribution Dynamics Project. Washington, DC: World Bank. http://go.worldbank.org/YADEAFEJ30. 26 ANNEX Annex 1: List of SEDLAC Harmonized Household Surveys Used Country 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 ARG X X X X X X X X X X X X X X X X X X X BRA X X X X X X X X X X X X X X X X CHL X X X X X X X X X X COL X X X X X X X X X X X X X X X X CRI X X X X X X X X X X X X X X X X X X X DOM X X X X X X X X X X X X X X ECU X X X X X X X X X X X X X X HND X X X X X X X X X X X X X X X X X X MEX X X X X X X X X X X X PAN X X X X X X X X X X X X X X X X PER X X X X X X X X X X X X X X X X PRY X X X X X X X X X X X X X X X X X X X SLV X X X X X X X X X X X X X X X X X URY X X X X X X X X X X X X X X X VEN X X X X X X X X X X X X X X X X X X # surveys 11 6 8 11 13 12 14 11 13 13 13 14 14 14 15 14 15 14 13 27 Annex 2: Ninis as a percentage of the population aged 15 to 24 years old Weighted vs. Unweighted Results % Ninis 15-24 % Ninis 15-24 Year (Weighted) (Unweighted) 1992 23.32 23.41 1993 22.91 23.29 1994 22.54 23.08 1995 22.41 23.15 1996 22.75 23.75 1997 22.06 23.16 1998 21.87 23.04 1999 21.28 22.32 2000 21.23 22.61 2001 21.35 22.58 2002 21.08 22.72 2003 20.95 22.33 2004 20.43 21.03 2005 20.16 20.48 2006 19.59 20.17 2007 19.43 19.44 2008 19.14 19.17 2009 19.56 19.52 2010 19.69 19.34 28 Annex 3: Heterogeneity in the number and evolution of ninis across Latin America Number of ninis aged 15-24 in LAC, 2010 (Thousands) 29 Change in percentage of ninis in LAC among those aged 15-24, 1992-2010 (p.p.) Change in number of ninis in LAC among those aged 15-24, 1992-2010 (%) 30 Annex 4: Number of ninis and their percentage of youth aged 15-24 by LAC country, 1992- 2010 31 32 33 34 35 36 37 38