Psychosocial competencies and risky behaviours in Peru

Favara, Marta; Sanchez, Alan Article Psychosocial competencies and risky behaviours in Peru IZA Journal of Labor & Development Provided in Cooperation with: IZA – Institute of Labor Economics Suggested Citation: Favara, Marta; Sanchez, Alan (2017) : Psychosocial competencies and risky behaviours in Peru, IZA Journal of Labor & Development, ISSN 2193-9020, Springer, Heidelberg, Vol. 6, Iss. 3, pp. 1-40, https://doi.org/10.1186/s40175-016-0069-3 This Version is available at: https://hdl.handle.net/10419/169306 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 DOI 10.1186/s40175-016-0069-3 ORIGINAL ARTICLE Open Access Psychosocial competencies and risky behaviours in Peru Marta Favara1* and Alan Sanchez2 *Correspondence: [email protected] 1University of Oxford, Queen Elizabeth House, 3 Mansfield Road, OX1 3TB Oxford, UK Full list of author information is available at the end of the article Abstract We use a unique longitudinal dataset from Peru to investigate the relationship between psychosocial competencies related to the concepts of self-esteem, self-efficacy, and aspirations, and a number of risky behaviours at a crucial transition period between adolescence and early adulthood. First of all, we document a high prevalence of risky behaviours with 1 out of 2 individuals engaging in at least one risky activity by the age 19 with a dramatic increase between age 15 and 19. Second, we find a pronounced pro-male bias and some differences by area of residence particularly in drinking habits which are more prevalent in urban areas. Third, we find a negative correlation between early self-esteem and later risky behaviours which is robust to a number of specifications. Further, aspiring to higher education at the age of 15 is correlated to a lower probability of engaging in criminal behaviours at the age of 19. Similarly, aspirations protect girls from risky sexual behaviours. JEL classification: J24, J13, O15. Keywords: Teenage pregnancy, Risky behaviours, Psychosocial, Aspirations, Peru 1 Introduction Risky behaviours are associated with health problems, low productivity and more generally with a decline of individual and collective well-being in the short, medium and long run (see for example Parkes et al. 2010). The study of the determinants of risky and criminal activities is informed mainly by sociological and psychological literature establishing the link between cognitive skills, psychosocial competencies and risky behaviours (Agnew et al. 2002; Caspi et al. 1994; Pratt and Cullen 2000).1 The economic literature on crime and risky behaviours primarily adopts an opportunity cost framework. People choose to commit a crime or to engage in risky behaviours if their expected utility from engaging in that behaviour is greater than the expected utility from their outside options (for example in terms of labour market opportunities). Within this framework, more educated people or people with better cognitive abilities are less likely to be involved in risky behaviours (Lochner and Moretti 2004; Travis and Hindelang 1977). However, these models do not acknowledge the role of psychosocial competencies. More recently, economists have gained an interest in studying the role of soft skills (or non-cognitive skills) as predictors of economic outcomes, such as educational attainments, health and labour market outcomes (see for example Borghans et al. 2008b; Chiteji 2010; Cobb-Clark and Tan 2011; Dohmen et al. 2010; Heckman et al. 2006; © The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 2 of 40 Jaeger et al. 2010). Nevertheless, few economic papers analyse the role of soft skills on risky behaviours. The aim of this study is to get a better understanding of the link between psychosocial competencies and risky behaviours at a crucial transition period between adolescence and early adulthood. Specifically, our analysis has three objectives. First, to document the prevalence of risky behaviours in the context of Peru, and the heterogeneity of these outcomes by gender and area of location. Second, to test the hypothesis that dimensions related to the concepts of self-esteem, self-efficacy, and aspirations have an impact on the occurrence of risky behaviours during adolescence. Third, to test the robustness of this association by applying statistical methods that allow to control for unobservable cofounders. To our knowledge this is the first study that looks at the predictive role of soft skills on risky behaviours—among the youth population—in a developing country. While the challenges faced by the youth are numerous, engaging in risky behaviours is highly prevalent in developing countries (Wellings et al. 2006), and this can be associated with worse labour market and health outcomes later in life. By looking at how early skills predict early engagement in risky activities, our study contributes to the understanding of the different channels through which soft skills accumulated over the life cycle explain labour market outcomes (and other life outcomes) later in life. For this analysis, we exploit the longitudinal nature of the Young Lives data, a unique individual-level panel following a cohort of about 700 children in Peru over four rounds of data collection that took place between 2002 and 2013. The Young Lives data cover a critical phase of the life-cycle for human capital and skills accumulation following the same children between ages 8 and 19. Information on a number of risky behaviours are collected at the age of 15 and 19 which makes the Young Lives data particularly suitable for this analysis. Furthermore, rich information both at the household and individual level are collected which include children’s cognitive and psychosocial competencies, school history, parental and children’s aspirations and aspirations for education.2Based on the data available, we define indicators to measure the prevalence of (i) smoking behaviours; (ii) drinking behaviours; (iii) drinking and violence (engaging in violent or risky activities when drunk); (iv) consumption of illegal drugs; (v) criminal behaviours; (vi) possession of weapons; (vii) unprotected sex; and (viii) total number of risky and criminal behaviours. Evidence on psychosocial competencies as a predictor of criminality and delinquency invites questions about the ability to prevent risky behaviours by shaping those skills. Furthermore, while the most ‘sensitive’ (productive) periods for investment in both cognitive skills and psychosocial competencies occur earlier in people’s life, soft skills during adolescence are more malleable than cognitive skills (Carneiro and Heckman 2003; Cunha and Heckman 2007; 2008; Cunha et al. 2010; Knudsen et al. 2006). Of course, the differential plasticity of different skills by age has important implications for the design of effective policies. There are three recent studies that have looked at the determinants of risky behaviours at age 15 in Peru using the first three rounds of Young Lives data: Cueto et al. (2011), Crookston et al. (2014), and Lavado et al. (2015). The study by Cueto et al. (2011) and colleagues highlights the importance of parent–child relations and peer effects in predicting smoking habits and unprotected sexual relations at early ages. Crookston et al. (2014) document the association between children victimization at school on subsequent risky behaviours. Finally, the study by Lavado et al. (2015) looks at the relationship between Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 3 of 40 cognitive and non-cognitive skills and consumption of cigarettes and alcohol and the early initiation of sexual activity. Overall, their results suggest a negative relation between risky behaviours and cognitive and non-cognitive skills. However, the evidence these studies can provide is limited for two reasons. First, there is low prevalence of risky behaviours observed at age 15, which authors try to compensate by being very inclusive in the definitions used, particularly in the way smoking and drinking are defined. Second, there is an endogenous relationship between child characteristics and risky behaviour outcomes. There are reasons to think that psychosocial competencies and the outcomes of interest are jointly determined. Therefore, the main challenge is in assessing whether the effect of poor psychological resources on the probability in engaging in risky behaviours is due to potential endogeneity bias; either through reverse causality or uncontrolled confounding variables. In this analysis we try to overcome both challenges. First (low prevalence), we show that in most cases the frequency of risky behaviours has increased considerably between age 15 and 19 which makes the empirical study more viable. Furthermore, the use of the last round of data allows us to broaden the scope of risky behaviours observed (at age 19, individuals were asked to report about criminal behaviours in addition to the other risky behaviours collected in previous rounds). Second (potential endogeneity bias), although this paper does not claim any causal relation, we exploit the fact that the data was collected over multiple periods to implement strategies that minimize both sources of endogeneity. To deal with reverse causality, we use lagged values of the psychosocial variables of interest, measured 3 years before the realization of the risky behaviours. To deal with omitted variable bias, we estimate a child fixed effects model, which purges bias due to unobservables that are constant over time. These are our main findings. First, we find that the prevalence of risky behaviours is evident and increases significantly over time: by age 15, two out of 10 individuals had engaged in at least one risky behaviour, whereas by age 19 one out of two had. By age 19, the prevalence of smoking and drinking is 19 and 34%, respectively; 13% had consumed illegal drugs, 27% had had unprotected sex and 19% had engaged in criminal behaviours. Second, with the exception of unprotected sex, there is a notorious pro-male bias in the prevalence of most of these behaviours. There are also some differences by area of location, particularly in drinking habits which are more prevalent in urban areas. Third, and perhaps most importantly, we find a negative correlation between psychosocial competencies and risky behaviours. Keeping everything else constant, an improvement of 1 standard deviation in self-esteem at the age of 15 is associated with a reduction of 7, 6 and 8 percentage points respectively in the probability of smoking, drinking and engaging in violent behaviours while drinking at the age of 19. It is also associated with a reduction in the prevalence of criminal behaviours and in the possession of a weapon by 14 and 5 percentage points, respectively. No similar correlation is found with self-efficacy. These results are robust to a large set of controls at the child and household level, and to community characteristics that are fixed over time. Moreover, child fixed effects estimates show that these associations persist once controlling for time-invariant unobservable characteristics. We note further that early self-esteem, measured at the age of 12 is already a predictor of later drugs consumption, unprotected sex, criminal behaviours and the number of risky behaviours the adolescents engage with at the age of 19. Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 4 of 40 Finally, we find that aspiring to higher education at the age of 15 reduces the probability of engaging in criminal behaviours at the age of 19 by 23 percentage points. Furthermore, while on average girls are more at risk of unprotected sex, girls aspiring to higher education are less likely to have unprotected sex. Nevertheless, once we control for unobservable individual characteristics the correlation between aspirations and risky behaviours is no longer significant. The remaining of the paper is structured as the following: Section 2 provides a conceptual framework for our analysis, including key references from the economics literature as well as from the psychological literature; Section 3 documents recent patterns in risky behaviours in Peru using the Demographic and Health Survey; Section 4 describes the data and the core predictors of risky behaviours used in the present analyses together with some statistics on risky behaviours using the Young Lives data; Section 5 discusses the empirical strategy and specifications adopted and finally Sections 6 and 7 report and discuss our findings. 2 Conceptual framework The traditional economic approach to youth risk taking is, as mentioned, a utility maximization/opportunity-cost approach. Forward-looking individuals pursue a certain activity if the expected benefits of it exceeds the expected costs. One example of model using this approach is the “Theory of Rational Addiction” (TORA) developed by Becker and Murphy (1988). According to the TORA, the utility of an individual depends on the consumption of two goods, cand y. The difference between the two goods is that while the utility generated by the current consumption of yis completely independent of past choices, the present utility derived by the consumption of cdepends on the past consumption of c. This is what characterize habits or addiction. In other words, the TORA assumes that instantaneous utility depends on current consumption of the addictive good, the stock of past consumption of the addictive good, and current consumption of all other goods. Developmental psychology, although not necessarily in contrast to the traditional economic approach, considers a wider variety of factors determining youth decisions to engage in risky behaviours. As Fischhoff (1992) effectively summarizes, according to developmental psychologists, (risk) decision-making depends on three groups of factors: how people ‘think’ about the world, i.e. their capacity for thinking through problems, examining the alternative available and evaluating their implications (‘cognitive’ development); how people ‘feel’ about the world (‘affective’ development) and the roles that others play in people’s choices (‘social’ development). In this paper we conceptually integrate the psychology component into a more general economic model of decision making taking the inspiration from behavioural economics (O’Donoghue and Rabin 2001) and the economic literature on skills formation (Cunha and Heckman 2007). As argued by Borghans et al. (2008a), preferences are central to conventional economic choice models. Agents decide in a decision horizon T the bundle of good to consume based on their preferences and constraints (typically, information constraints and budget constraints). They also acknowledged the role of dynamic constraints connected to asset, skills and traits formation. Their model is consistent with a framework were individual preferences change over time, individual decisions are time inconsistent and Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 5 of 40 discount rates as well as preferences may vary with ‘age, mood, personality traits and cognition’. They argue that cognitive and personality traits can affect consumption choices through different mechanisms including risk aversion, inter-temporal preferences and the valuation of leisure. Insights from behavioural economics are hugely important to understand why young people might behave differently than adults. Empirical evidence suggests that young people are excessively myopic with respect to the future and therefore are more likely to have inconsistent preferences over time (Gruber and Koszegi 2001; O’Donoghue and Rabin 2001). More specifically, they have the tendency to have a higher discount rate in the short run than in the long run. Young people respond to the uncertainty about the future by reducing the importance of the future, an effect known as hyperbolic discounting. Furthermore, they tend to under-appreciate the effect of changes in their states and the extent to which their preferences may adapt over time. Because of that, they tend to inappropriately project the current preferences onto their future tastes (projection bias) (Loewenstein et al. 2003; O’Donoghue and Rabin 2001). For this reason, random changes to their current states affect their long-run decision making. Also, youth tend to be less risk averse which is consistent with the myopia and hyperbolic discounting features (Gruber and Koszegi 2001; O’Donoghue and Rabin 2001). Moreover, risky decisions are made in uncertain environments and for many risky activities, the cost is one-time and permanent. Uncertainty and one-time cost with longer term implications might increase risk-taking behaviours and a mistake made in the past becomes permanent in its consequences. Finally, younger teens tend to be both more impatient and subject to peer pressure (Lewis 1981). All these characteristics might help in explaining why risky behaviours are more prevalent among young people. On the other side, there are at least three factors which might counterbalance this: biology, income and law (Gruber 2001). Indeed, some risky activities (e.g. sexual intercourse) become desirable with age (biology). Moreover, some illegal activities for younger teens become legal at older ages (e.g. cigarettes consumption is illegal to under 19 in Peru) (law). Finally, older teens may have more money available to finance their risky activities (income). 2.1 Psychosocial competencies and cognitive skills as predictors of risky behaviours: evidence from policy and research Many studies in the economic literature find evidence of contemporaneous correlation between different risky behaviours (Chaloupka and Laixuthai 1997; Dee 1999; DiNardo and Lemieux 2001; DuRant et al. 1999; Farrelly et al. 2001; Model 1993; Wiefferink et al. 2006). Those evidence support the ‘bad seed’ hypothesis, as described by Gruber (2001). The hypothesis is that there is a certain segment of the youth population that is predisposed towards risky activities, while others are not. In that case, policies targeting the segment of population at risk should work effectively. An alternative hypothesis in psychological literature is that there is a certain amount of risk that youths have the tendency to take (‘conservation of risk’ hypothesis). Reducing risky activity in one area would have a substitution effect by increasing risky activities in another. To date, most intervention programmes have been targeting specific groups of the population considered at risk, mainly by targeting single risk behaviours. Most recently, there are examples of interventions taking a broader approach and target more than one risky behaviour at time. Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 6 of 40 More specifically, they aim to address some underlying determinants of risky behaviours which are believed to protect young people from, or predispose them to, distinct risky behaviours. Therefore, a better understanding of which childhood traits predict risky behaviours is crucial from a policy perspective. Empirical evidence suggest that interventions focusing on improving cognitive skills or aimed at improving soft skills are effective in reducing risky behaviours. An example of an intervention aimed at improving opportunities for children coming from poor backgrounds is the well-known Perry Preschool Programme, an intervention targeting a sample of 3–4-year-old African–American children living in poverty and assessed to be at high risk of school failure. Although the literature originally focused on the cognitive impact of the intervention, long-term effects have in fact been more persistent in noncognitive areas. Heckman et al. (2010) and Conti et al. (2015) show that Perry significantly enhanced adult outcomes including education, employment, earnings, marriage, participation in healthy behaviours, and reduced participation in crime teen pregnancy, and welfare dependency later in life. Interestingly, although the programme initially boosted the IQs of participants, this effect soon faded. A persistent effect of the programme has been found on improvements in personality skills (e.g. it reduces aggressive, antisocial, and rule-breaking behaviours). On the other side, Hill et al. (2011) show that several interventions that focus on personality rather than only on cognitive skills were effective at reducing delinquency and traits related to delinquency. Few economic papers analyse the role of personality traits and non-cognitive skills on criminal activities, or more generally, risky behaviours. Heckman et al. (2006) find that self-esteem and locus of control measured during adolescence are as powerful as cognitive abilities in predicting adult earnings. Moreover, they find that personality factors for men affect the probability of daily smoking more than cognitive factors and the opposite is true for women. Similarly, Cunha et al. (2010) show that personality traits are relatively more important in predicting criminal activity than cognitive traits are. Further, Conti and Heckman (2010) suggest that personality and health status measured during adolescence explain more than 50% of the difference in poor health, depression and obesity at age 30. For males, personality traits and health endowments are more predictive than cognitive skills while for women they are equally predictive. The role of self-efficacy and self-esteem as predictors of risky behaviours has been discussed in the psychological literature, particularly its role during the adolescence period. This is because during this stage individuals commonly start experimenting with risky activities (including alcohol abuse, smoking, drug use, and unprotected sex). Bandura et al. (2001) state that perceived self-efficacy (in the areas of academic, social, and selfregulatory efficacy) is important to resist peer pressure for transgressive activities, a view also shared by other authors (e.g. Wills 1994). Empirical evidence shows a negative relationship between self-efficacy and risky or delinquent behaviours, including use of alcohol and drugs, physical and verbal aggression, theft, cheating and lying (Bandura et al. 2001; Bandura et al. 2003). In addition, self-efficacy is thought to be important to change unhealthy behaviours, such as smoking (Schwarzer 2001). In the case of self-esteem, a negative relationship with risky behaviours is expected (Donnellan et al. 2005). First, people with low self-esteem perceive that they have less social ties (Rosenberg 1965), which in turn decrease conformity to social norms and increase delinquency. Second, it is theorized that aggression and antisocial behaviour Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 7 of 40 are motivated by feelings of inferiority rooted in early childhood experiences. In addition, it is thought that self-esteem mediates the impact of stress, which is of a subjective nature (Baumeister et al. 2003). People with high self-esteem are likely to experience less stress because they interpret negative events more benignly, are more optimistic about their coping abilities, and perceive they have more control compared to people with low self-esteem. Notwithstanding these arguments, others have argued that a positive relationship could arise, as noted by Baumeister et al. (2003). While it is true that young people with low self-esteem might be more prone to engage in risky behaviours—for solace when they feel bad about themselves, young people with arguably high self-esteem might have biases in their interpretation of events that allow them to feel better about themselves, either by minimizing their own vulnerability or by distorting how their parent will react. This is likely to be the case in particular for people with unrealistically high self-esteem, close to narcissism (Donnellan et al. 2005). Baumeister et al. (2003) provide a review of the literature about the role of self-esteem on several life outcomes, including smoking, alcohol and drug abuse, and unprotected sex. They conclude that evidence linking low self-esteem to risky behaviours during the adolescence is mixed and inconclusive, with positive, negative and zero effects found, particularly in the case of alcohol use, whereas in the case of smoking, the relation is mainly negative. On the other hand, Donnellan et al. (2005) use data from three different datasets which strongly support the notion that low self-esteem is related to aggressive behaviour. An important aspect is whether self-esteem and self-efficacy can be measuring similar dimensions of a person self-concept. In fact, some authors (Dercon and Krishnan 2009; Epstein et al. 2004) suggest that self-efficacy can be treated as a determinant of selfesteem. Wills (1994) shows empirical evidence that supports the notion that self-efficacy might be a more important factor than self-esteem, and suggests that, in absence of a control for self-efficacy, previous studies might have overstated the importance of selfesteem. Overall, what this seems to suggest is that it is important to control for both psychosocial dimensions in order to estimate the individual contribution of each. At the heart of the traditional opportunity cost approach to risky behaviours and of intertemporal choice models described above, are individual expectations. As mentioned, people make decisions taking into account the present utility, their expectations about future utility. Present-biased time preferences are likely to be more frequent among people who are pessimistic about their future. Consistently with the ‘opportunity cost’ argument in the risky behaviour literature, if an outcome is perceived as inaccessible, people might believe that they have little to lose by engaging in risky behaviours. There is a considerable body of economic literature investigating the role of aspirations and subjective expectations for contraceptive choices (Delavande 2008), (sexual) risky behaviour (De Paula et al. 2013; Shapira 2013) and non-marital childbearing choices (Wolfe et al. 2007). As Dalton et al. (2016) argue, how far people aspire depends on their own beliefs about what they can achieve with effort, i.e. their own expectations. People would not aspire to an outcome that is perceived as inaccessible. However, given the endogenous nature of aspirations, the empirical distinction between aspirations and expectations is hard to achieve in a non-experimental setting and often aspirations are used interchangeably with expectations. Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 8 of 40 Finally, the decision making model described above, yields several important implications regarding the role that cognition plays for the probability to engage in risky behaviours. There are a number of mechanisms through with cognitive skills might affect individual decision making, some of which can be amplified by the interaction between cognitive skills and schooling. First, individuals with higher cognitive skills might be more able to access information and more efficient at interpreting it. Second, cognitive skills are likely to shape preferences. As argued by Dohmen et al. (2010), people with better cognition appear to be more patient. They are also more willing to take risks. One potential explanation is that they are better able to envision future consequences and somehow reduce ambiguity about the future. In this sense, increased cognitive ability favourably influences behaviours, particularly when information is limited or idiosyncratic. Schooling is also considered a protective factor against risky behaviours (see for example Cutler and Lleras-Muney 2010). First, education promotes the accumulation of both cognitive and socio-emotional skills, which affects the way individuals process information and behave. Second, education shapes the nature of the social network available to the individual, which can have either a positive or a negative effect (Behrman 2015; Peters et al. 2010). Third, education might shape time preferences, e.g. because schooling focuses students’ attention on the future (Becker and Mulligan 1997; Fuchs 1982). Fourth, people with more education might be better informed about negative health consequences, either because they learned about these consequences in school, or because better educated people find it easier to obtain and evaluate such information (De Walque 2007; Kenkel 1991). Fourth, education could also influence behavior by increasing the opportunity cost of engaging in risky behaviours, i.e. by increasing future income. 2.2 Other predictors of risky behaviours The importance of family environment is well recognized by developmental research. Numerous studies show that children who grow up in single-parent families are more at risk of engaging in risky behaviours (see for example Evans et al. 1992). Adolescents from intact two-parent families tend delay the start of sexual activity relative to those in disrupted families (see for example Meschke and Silbereisen 1997). Clark and Loheac (2007) examine the consumption of tobacco, alcohol and marijuana in the USA and find that marijuana use is more widespread in single-parent families. They also find that smoking is more frequent amongst recent movers. Migration indeed might be potential source of instability. Gaviria and Raphael (2001) using secondary school data from the USA suggest that recent movers may be more susceptible to peer group pressure, at least with respect to the consumption of marijuana and cocaine. Similarly, children who have older siblings have a higher probability of engaging in risky behaviours (Averett et al. 2011), and there might be a number of plausible explanations for that. It might be that older siblings affect their younger siblings’ behaviours indirectly, by being a role model to them and directly by proving them more opportunities to interact with a different group of older friends. An alternative explanation might be that parents spend less time in supervising their younger offspring (Aizer 2004). It is worth to highlight that single parenthood as well as some other socio-economic characteristics frequently associated with poverty are some of the stronger predictors of risky behaviours. Risky sexual behaviours are often a manifestation of lack of opportunities, deprivation and poverty. Nevertheless, although risky behaviours are generally Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 15 of 40 Similarly, the dummy variable for weapon possession is equal to 1 for those who during thelast30dayscarriedaweaponatleastonce. Overall, looking at the incidence of risky behaviours by age, we notice that risky behaviours increase significantly between the age of 15 and the age of 19, in correspondence with the transition from childhood to adolescence (Table 3). It is worth to note that by the age of 19 male engagement in risky behaviours is about two to three times that of females in smoking, drinking and taking drugs and criminal behaviours. There is also an urban-rural difference in drinking, where adolescents living in urban areas drink more (10% by the age of 15 and 38% by the age of 19) relative to those in rural areas (5% by the age of 15 and 24% by the age of 19) and by the age of 19 are more likely to engage in risky behaviours while drinking. However, not only the prevalence but also the intensity of risky behaviours increases over time. We define a variable counting the number of risky activities the young people have been involved in by the age of 15 and 19.7By the age of 15, about 22% of young people have engaged with at least one risky behaviour. By the age of 19, slightly more than one out of two had engaged in at least one type of risky behaviour, with a distinctive promale bias (67% among males, 43% among females). While 26% of the population engaged in only one risky activity, a consistent segment of the youth population (29%), mainly male population, undertakes more than one of these activities. Our data reports quite a remarkable diffusion of risky behaviours among Peruvian adolescents and a worrisome predisposition towards risky activities for the relevant part of them. Although our data do not provide full support to either the ‘bad seed’ or the ‘conservation of risk’ hypothesis, it is worth to note that there is evidence of a certain persistence (or recidivism) in risky behaviours. Those who engage in risky behaviours at the age of 15 are indeed more likely to engage in risky behaviours at the age of 19. The average ‘number of risky behaviours’ at the age 15 is strongly correlated with the same measured 4 years later (standardized correlation coefficient of 0.6). Recidivism is more evident in some risky behaviours than others, particularly in drug consumption, drinking and smoking. In fact, adolescents who consume drugs at age 15 are 64 percentage points more likely to consume drugs at age 19. Similarly, drinking (smoking) at age 15 increases the probability of smoking (drinking) at age 19 by 38 percentage points (39 percentage points). In the next section we characterize further who are these young people, what is their history, their past experience, their ability and psycho-social well-being and where do they live using a multivariate approach. 4.3 Psycho-social competencies and cognitive skills in Young Lives data In this section we briefly define the core predictors of risky behaviours used in the analysis. As discussed in Section 2, psychosocial competencies have been identified as important factors in predicting risky behaviours. In our data, we capture them through two indicators that have been administered in the last three rounds of the Young Lives survey: the self-esteem scale and the self-efficacy scale. In the Young Lives database, these scales are referred as the pride index and the agency index, respectively. The self-esteem scale builds on the self-esteem concept by Rosenberg (1965). The objective is to measure a child’s overall evaluation of his or her own worth. In turn, the self-efficacy scale builds on the concept of locus of control by Rotter (1966) and self-efficacy by Bandura (1993). Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 16 of 40 In this case, the objective is to measures a child’s sense of agency or mastery over his/her own life. Each scale is measured based on respondents’ degree of agreement or disagreement with a number of positive and negative statements measured on a 4-point Likert scale (the full list of statements are reported in Table 2). Statements are adapted to measure specific dimensions of the children’s living circumstances. In order to calculate each scale, all statements were recoded to be positive outcomes, standardized (normalized to z-scores), and then averaged. The internal consistency of these scales is shown in (Dercon and Krishnan 2009).8 Another core predictor for risky behaviours investigated in this analysis is individual aspirations. The measure of aspirations considered in this study reflects a combination of aspirations and beliefs about the likelihood of achieving the aspired outcomes. Specifically, Young Lives collects information about educational aspirations by asking the child the following question: ‘Imagine you had no constraints and could study for as long as you liked, or go back to school if you have already left. What level of formal education would you like to complete?’. In this study, we define a dummy variable equal to 1 for individuals with high aspirations, i.e. for those children who aspire to go to university, and 0 otherwise. It is important to note that aspirations are understood to be shaped to a large extent by self-efficacy (Bandura et al. 2001). Finally, we look at a set of predictors relating to education, namely school enrollment and cognitive development. Related to the latter, we include an indicator of literacy measured by the Peabody Picture Vocabulary Test (PPVT), a test of receptive vocabulary. The task of the test taker is to select the picture that best represents the meaning of a stimulus word presented orally by the examiner. The items used were validated independently for local teams in each country and are age-standardized. The PPVT was administered since round 2, when the children were 8 years old. To measure early age cognitive skills, in round 1 Young Lives administer the Raven test, a widely used test of abstract reasoning and regarded as a non-verbal estimate of fluid intelligence. Besides the PPVT and the Raven test, a numeracy test was administered. Given that Young Lives is not a school-linked study, numeracy assessments are not aligned with school curricula; however, the contents of the tests could be linked with learning that should be emphasized in schools. In order to account for wide variations in the grade and skill levels of individuals both within and across countries, the tests incorporated questions with differing levels of difficulty: at the basic level the tests included questions assessing basic number identification and quantity discrimination; at the intermediate level, questions on calculation and measurement; and at the advanced level, questions related to problem solving embedded in hypothetical contexts that simulate real-life situations (e.g. tables in newspapers). The numeracy skills indicator used in this analysis is the number of correct answers in the Math test. Notably, the cognitive tests have been collected for all children regardless whether they are attending school or not. This feature of the data avoids any selection problem which commonly arises using school-based data. A validation of the psychomethric properties of the PPVT and Math scores can be found in Cueto and Leon (2012) in and Cueto et al. (2009). 5 Empirical strategy In this section we define a multivariate set-up, estimating linear probability (OLS) models. Our dependent variables are the risky behaviours as defined in the previous sections. Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 17 of 40 With the exception of the intensity variable (number of risky behaviours), the dependent variable is a variable equal to 1 if the young person engages in risky behaviours at the age of 19, and 0 otherwise. First of all, we investigate the predictors of risky behaviours looking at the association between risky behaviours measured at the age of 19 and psychosocial competencies measured at the age of 15, controlling for schooling achievement and a broad set of early (or time-invariant) individual and household-level characteristics as follows: Yij,19 =β0+αi+β1self −efficacyi,15 +β2self −esteemi,15 +Xi,15+ωij,19 +i,19 (1) In this model Yij,19 denotes risky behaviour outcomes of individual iliving in the community jobserved at age 19; self −efficacyi,15 and self −esteemi,15 are measured at the age of 15; Xi,15 is a vector of pre-determined characteristics of individual irecognized as potential predictors of risky behaviours.9 In light of the results of past research, Xi,15 includes a number of indicators of household socio-economic status; information about family structure (number of siblings, whether he is living only with one biological parent, whether the young person has an older sibling); child demographic characteristics (gender and age at the time of the 2013/14 survey round); a dummy variable equal to 1 whether the child at the age of 19 is living in the same community as when he/she was 15 years old and 0 otherwise; and individual schooling and cognitive skills. The term αireflects unobserved individual characteristics that are constant over time. Finally, i,19 is an idiosyncratic error and we approximate the socioeconomic status of the natal household by using mother’s education level, an indicator for the rural/urban location where the household resides, and the tercile of wealth index, a composite measure of living standards including housing quality, access to service and a consumer durable index as defined in Table 2. Finally, we look at a set of predictors relating to education. More specifically we look at school enrolment, delayed enrolment and school achievement. School achievement, measured either by the Raven test score or the Peabody Picture Vocabulary Test (PPVT) and a Math test, can also be considered as a proxy of the child’s cognitive skills. Notably, the two tests have been collected for all children regardless whether they are attending school or not. This feature of the data avoids any selection problem which commonly arises using school-based data. Similarly, we investigate the correlation between educational aspirations measured at the age of 15 and risky behaviours at the age of 19. According to the ‘opportunity cost’ argument, we would expect to find a negative correlation between aspirations and risky behaviours if the perceived cost of engaging in risky behaviours increases with aspirations. The descriptive statistics presented in Table 4 indeed shows that adolescents engaging in at least one risky behaviour at the age of 19 have lower aspirations than ‘not at risk’ adolescents. Given that aspirations is likely to feed into the child’s self-efficacy and self-esteem, we estimate a separate model similar to the one discussed above but including a dummy variable equal to 1 for those children that at the age of 15 aspire to complete higher education (university), and 0 otherwise: Yij,19 =θ0+αi+θ1aspirationsi,15 +Xi,15+ωij,19 +i,19 (2) Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 18 of 40 Table 4 Descriptive statistics Total Not at risk At risk ttest Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. pvalue Child is male 0.53 0.500 0.40 0.491 0.63 0.483 0.000 Age in R4 18.41 0.582 18.36 0.540 18.46 0.610 0.052 Mother’s education - primary school or none 0.33 0.472 0.32 0.468 0.34 0.476 0.600 Mother’s education - secondary school 0.42 0.493 0.40 0.490 0.43 0.496 0.404 Mother’s education - higher education 0.16 0.365 0.19 0.391 0.14 0.343 0.114 Descriptives, age 15 Type site - rural, age 15 0.22 0.418 0.24 0.431 0.21 0.407 0.351 Migrated between 15 and 19 0.08 0.268 0.07 0.254 0.08 0.278 0.535 First tercile of wealth 0.32 0.469 0.30 0.461 0.34 0.475 0.392 Second tercile of wealth 0.34 0.473 0.35 0.477 0.33 0.471 0.711 Third tercile of wealth 0.34 0.474 0.35 0.478 0.33 0.471 0.634 Single parent, age 15 0.26 0.440 0.21 0.406 0.30 0.461 0.016 Child has older siblings 0.35 0.478 0.35 0.478 0.35 0.478 0.974 Number of siblings 1.92 1.281 1.98 1.303 1.87 1.264 0.367 Self-efficacy 0.02 0.521 0.08 0.525 −0.03 0.514 0.015 Self-esteem 0.00 0.594 0.04 0.601 −0.02 0.589 0.229 Child aspirations 0.92 0.26 0.95 0.220 0.90 0.294 0.064 Mother’s aspirations 0.93 0.247 0.95 0.210 0.92 0.273 0.125 Child is enrolled 0.94 0.232 0.96 0.189 0.93 0.261 0.085 PPVT (standardized) −0.02 1.009 0.03 1.015 −0.06 1.005 0.337 Math (standardized) −0.01 0.998 0.09 1.062 −0.08 0.940 0.069 Observations 490 217 273 Descriptives, age 12 Migrated between age 7 and 15 0.13 0.338 0.12 0.323 0.14 0.351 0.393 Self-efficacy 0.01 0.515 0.04 0.468 −0.02 0.549 0.192 Self-esteem 0.04 0.666 0.07 0.641 0.01 0.686 0.264 Child aspirations 0.92 0.272 0.94 0.240 0.90 0.294 0.151 Mother’s aspirations 0.95 0.213 0.95 0.223 0.96 0.206 0.672 Raven (standardized) 0.02 1.009 0.05 1.053 −0.01 0.975 0.512 Observations 524 230 294 In both Eqs. 1 and 2, self-efficacy, self-esteem and aspirations are measured at the age of 15. An empirical question is whether the psychosocial competencies measured at younger ages predict later risky behaviours. Young Lives collect self-efficacy, self-esteem and aspirations at both age 12 and 15 which allow us to look at the long-term association with risky behaviours. We also report results for this long-term specification. In this case, all control variables are either time invariant or measured as early as possible (at age 8). In this case, the Raven score measured at age 8 is used as indicator of school achievement. Although informative, an estimation of the risky behaviour equations using crosssectional data would be unbiased only under very strong assumption about the role of unobservable variables. In absence of plausibly exogenous variations in the regressors, their estimation raise endogeneity concerns and might lead to biased interpretations. Therefore, our intention is not to identify causal effects. Rather, the estimated parameters should be interpreted as partial correlations which may be revelatory about potential drivers of risky behaviours at different ages and the channels through which such effects may be mediated. Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 19 of 40 Further, we exploit the fact that we have repeated measures of risky behaviours and we estimate the outcome of interest using a child fixed effects model, as follows: Yij,19−15 =β1self −efficacyi,19−15 +β2self −esteemi,19−15 +Xi,19−15+ωij,19−15 +i,19−15 (3) and similarly, Yij,19−15 =θ1aspirationsi,19−15 +Xi,19−15+ωij,19−15 +i,19−15 (4) In this specification, the role of self-efficacy and self-esteem is identified by exploiting changes between ages 12 and 15 that in turn lead to changes in risky behaviours between ages 15 and 19. In doing so, we implicitly assume the relevant coefficients are ageindependent. This strategy has the advantage that it controls for individual unobservable characteristics that are constant over time. 6 Understanding risky behaviours As an initial exploration of factors that might affect the probability of engaging in risky behaviours at the age of 19, we compare the mean characteristics of the predictors listed above for adolescents ‘at risk’ (engaging in at least one risky behaviour by the age of 19) and adolescents ‘not at risk’. All predictors are measured when the adolescent was 12 and 15 years old. These differences are presented in Table 4 alongside tests for statistical significance. Looking first at the individual characteristics, young people engaging in risky behaviours by the age of 19 are more likely to be boys and slightly older than those who are not at risk. Young people ‘at risk’ have lower self-efficacy (slightly lower self-esteem) and are less likely to aspire to university at the age of 15. Furthermore, risky behaviours are more prevalent among young people having lower cognitive skills (performing worse in the Math test) and those who have already dropped out of school by the age of 15. Interestingly, risky behaviours are not necessarily a phenomenon prevalent among young people living in poverty. Indeed, young people living in poverty are as likely as young people living in less poor households to engage in risky behaviours. Additionally, there is no difference in the prevalence of risky behaviours in rural and urban areas and the level of parental education is the same among young people at risk and not at risk. Notably, risky behaviours are more prevalent in single-parent households. While the differences in mean characteristics across young people engaging in risky behaviours and their peers are instructive, Sections 6.1 and 6.2 go a step forward in the identification of the potential predictors of young people’s engagement in risky behaviours within a multivariate set-up, as described in Section 5. 6.1 Main results The main results of the analysis are reported in Tables 5 and 6. Outcomes are measured at age 19 whereas, unless otherwise expressed, predictors are measured at age 15. Smoking, drinking and drinking and violence are the outcomes for which the highest proportion of the variance is explained by the selected predictors, with an R-squared of around 20%. In contrast, for drug consumption, unprotected sex and criminal related outcomes between 10 and 13% of the variance is explained. Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 20 of 40 Table 5 Psychosocial competencies at age 15 on participation in risky behaviours at age 19 Drinking Drugs Unprotected Criminal Carried No. of Smoking Drinking & violence consumption sex. beh. a weapon risky beh. Self-efficacy, age 15 −0.008 −0.041 −0.026 −0.019 −0.027 0.004 −0.022 −0.117 (0.799) (0.360) (0.616) (0.672) (0.550) (0.942) (0.390) (0.361) Self-esteem, age 15 −0.065** −0.062* −0.076** −0.017 −0.020 −0.144* −0.045** −0.210** (0.011) (0.063) (0.023) (0.532) (0.648) (0.060) (0.019) (0.015) Child is male 0.221*** 0.225*** 0.251*** 0.128*** −0.034 0.130*** 0.022 0.562*** (0.000) (0.000) (0.000) (0.000) (0.455) (0.004) (0.235) (0.000) Age in R4 0.092*** 0.100** 0.099** 0.036 −0.004 −0.004 −0.024 0.201* (0.006) (0.036) (0.021) (0.121) (0.943) (0.920) (0.400) (0.059) Type site - rural, age 15 −0.079 −0.270*** −0.224** 0.025 0.062 0.123 0.033 −0.228 (0.408) (0.000) (0.010) (0.703) (0.389) (0.236) (0.375) (0.220) Migrated between 15 and 18 0.061 0.137 0.162 0.024 0.132 −0.143 −0.088** 0.265 (0.422) (0.117) (0.115) (0.763) (0.170) (0.303) (0.024) (0.350) Second tercile of wealth, age 15 −0.010 0.079 0.079 −0.007 −0.006 −0.064 −0.005 0.051 (0.888) (0.265) (0.275) (0.854) (0.903) (0.280) (0.870) (0.766) Third tercile of wealth, age 15 −0.097 0.071 0.044 −0.024 0.054 −0.030 −0.027 −0.024 (0.176) (0.379) (0.655) (0.667) (0.434) (0.635) (0.259) (0.899) Mother’s education - secondary school 0.040 −0.025 0.029 0.026 0.038 0.003 −0.018 0.062 (0.317) (0.547) (0.516) (0.471) (0.459) (0.929) (0.489) (0.521) Mother’s education - higher education 0.059 0.006 0.062 0.039 −0.066 0.093 0.020 0.057 (0.377) (0.944) (0.464) (0.572) (0.244) (0.242) (0.491) (0.762) Single parent, age 15 0.071 0.038 0.089 0.072 0.140*** 0.167** 0.047* 0.367** (0.285) (0.490) (0.113) (0.197) (0.006) (0.024) (0.099) (0.021) Child has older siblings 0.004 0.048 0.057 0.028 −0.037 0.019 0.009 0.051 (0.903) (0.281) (0.157) (0.429) (0.267) (0.800) (0.676) (0.604) Number of siblings, age 15 −0.012 −0.013 −0.004 0.010 0.001 0.048* 0.020 0.006 (0.384) (0.462) (0.858) (0.300) (0.964) (0.069) (0.137) (0.854) Child is enrolled, age 15 −0.072 −0.098 −0.134 0.007 −0.048 −0.269** −0.049 −0.260 (0.395) (0.339) (0.127) (0.937) (0.494) (0.010) (0.547) (0.136) PPVT z-score, age 15 −0.051* −0.026 −0.003 −0.045 0.037 −0.024 0.015 −0.069 (0.063) (0.551) (0.926) (0.188) (0.254) (0.667) (0.375) (0.478) Math z-score, age 15 0.009 −0.001 −0.024 −0.015 −0.016 −0.016 −0.020 −0.043 (0.703) (0.954) (0.216) (0.464) (0.561) (0.514) (0.118) (0.290) Number of observations 490 490 490 490 490 490 490 490 R-squared 0.207 0.193 0.195 0.114 0.104 0.130 0.099 0.182 Note: The table reports the estimates of the linear probability model with standard errors (in parentheses) clustered at cluster level, *p<0.1; **p<0.05; ***p<0.01. All controls were included as reported together with dummy variables for the cluster that individuals were recruited in the 2002 round; coefficients for these are not reported Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 21 of 40 Table 6 Educational Aspirations at age 15 on participation in risky behaviours at age 19 Drinking Drugs Unprotected Criminal Carried No. of Smoking Drinking & violence consumption sex. beh. a weapon risky beh. Child aspired for higher education, age 15 −0.090 −0.119 −0.093 −0.123 −0.070 −0.226** −0.044 −0.447* (0.391) (0.121) (0.310) (0.178) (0.274) (0.042) (0.461) (0.081) Child is male 0.221*** 0.229*** 0.254*** 0.128*** −0.032 0.128*** 0.025 0.571*** (0.000) (0.000) (0.000) (0.000) (0.510) (0.003) (0.177) (0.000) Age in R4 0.091*** 0.099** 0.098** 0.035 −0.004 −0.005 −0.025 0.195* (0.009) (0.040) (0.026) (0.124) (0.925) (0.886) (0.365) (0.063) Type site - rural, age 15 −0.084 −0.277*** −0.229*** 0.018 0.058 0.110 0.030 −0.255 (0.379) (0.000) (0.010) (0.769) (0.411) (0.276) (0.343) (0.145) Migrated between 15 and 18 0.070 0.149* 0.173* 0.031 0.138 −0.122 −0.082** 0.305 (0.308) (0.077) (0.078) (0.668) (0.139) (0.314) (0.038) (0.226) Second tercile of wealth, age 15 −0.015 0.073 0.071 −0.008 −0.008 −0.073 −0.011 0.031 (0.821) (0.277) (0.307) (0.834) (0.860) (0.162) (0.731) (0.840) Third tercile of wealth, age 15 −0.106 0.060 0.032 −0.024 0.049 −0.045 −0.036 −0.056 (0.136) (0.444) (0.746) (0.661) (0.451) (0.507) (0.138) (0.753) Mother’s education - secondary school 0.037 −0.031 0.022 0.027 0.036 −0.002 −0.023 0.045 (0.383) (0.449) (0.594) (0.439) (0.483) (0.939) (0.322) (0.556) Mother’s education - higher education 0.043 −0.010 0.042 0.037 −0.071 0.060 0.008 0.006 (0.520) (0.891) (0.604) (0.568) (0.201) (0.512) (0.808) (0.973) Single parent, age 15 0.068 0.036 0.087 0.070 0.139*** 0.161** 0.045 0.358** (0.293) (0.511) (0.117) (0.192) (0.007) (0.024) (0.110) (0.022) Child has older siblings −0.001 0.039 0.049 0.026 −0.041 0.009 0.003 0.026 (0.966) (0.359) (0.221) (0.459) (0.165) (0.899) (0.876) (0.780) Number of siblings, age 15 −0.015 −0.017 −0.008 0.008 −0.001 0.041 0.017 −0.007 (0.284) (0.345) (0.709) (0.402) (0.949) (0.107) (0.209) (0.834) Child is enrolled, age 15 −0.042 −0.062 −0.106 0.046 −0.027 −0.192 −0.036 −0.122 (0.682) (0.575) (0.268) (0.620) (0.717) (0.112) (0.695) (0.552) PPVT z-score, age 15 −0.052** −0.028 −0.007 −0.041 0.036 −0.023 0.011 −0.074 (0.049) (0.517) (0.850) (0.178) (0.253) (0.661) (0.473) (0.424) Math z-score, age 15 0.013 0.001 −0.021 −0.015 −0.016 −0.007 −0.018 −0.036 (0.543) (0.946) (0.304) (0.457) (0.575) (0.757) (0.160) (0.374) Number of observations 490 490 490 490 490 490 490 490 R-squared 0.201 0.188 0.188 0.119 0.104 0.117 0.085 0.174 Note: The table reports the estimates of the linear probability model with standard errors (in parentheses) clustered at cluster level, *p<0.1 **p<0.05 ***p<0.01. All controls were included as reported together with dummy variables for the cluster that individuals were recruited in the 2002 round; coefficients for these are not reported Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 22 of 40 The four most consistent predictors of risky and criminal behaviours are gender, age, self-esteem and whether the individual comes from a single-parent household. The fact that there are differential patterns by gender and age was already evident in the descriptive statistics. The probability of smoking, drinking and engaging in drinking and violence increases respectively by 22, 23 and 25 percentage points for males compared to females. Similarly, males are 13 percentage points more likely than females of consuming drugs and engaging in criminal behaviours. Although the average age is 19, many individual were aged 18 at the moment of the interview. We find that moving from 18 to 19 years old increases the likelihood of smoking, drinking and drinking and violence by around 10 percentage points in all cases. Beyond the role of gender and age, our main finding is related to the association between self-esteem and risky and criminal behaviours. Given that the estimation controls for demographic and socio-economic characteristics, as well as for schooling achievement and time-invariant community characteristics, among other aspects, the estimated parameter can be interpreted as a robust association. Keeping other factors constant, a 1 standard deviation increase in self-esteem at age 15 reduces the likelihood of engaging in smoking, drinking, and drinking and violence by 7, 6 and 8 percentage points (respectively); it also reduces the likelihood of criminal behaviours and carrying a weapon by 14 percentage points and 4 percentage points. In contrast, the association with self-efficacy is not statistically significant, though it is interesting to observe that the estimated coefficients have the expected (negative) sign. About the role of family structure, a specific dimension that plays a role is whether the individual comes from a single-parent household, which increases the likelihood of engaging in risky sex, in criminal behaviours and carrying a weapon by 14, 17 and 5 percentage points, respectively. In addition, the number of siblings is positively associated with criminal behaviours. It is interesting to observe that living in a rural area reduces drinking by 27 percentage points and drinking and violence by 23 percentage points. Finally, keeping other factors constant, we find a strong negative association between school enrollment and criminal behaviours. Also, an increase of one standard deviation in the PPVT test is associated with a reduction in smoking of 5 percentage points. From the factors previously mentioned, gender, age, self-esteem and living in a singleparent household stand out as factors that systematically predict risky and criminal behaviours. These are also the factors that predict the (overall) number of risky behaviours in which the individual has engaged. Also, it is interesting to observe that psychosocial competences do not play any role in predicting risky sexual behaviours. This is quite surprising given that previous literature suggest self-efficacy (or self-confidence) to be one of the key factors for contraceptive uses and particularly for the use of condom which, particularly for girls, requires negotiating its use with the partner (see for example Salazar et al. 2005). More generally, unprotected sex is the behaviour for which fewer predictors turn out to be statistically significant (only one, coming from a single-parent household) which suggest that other important predictors might have been neglected. Factors such as being born to a teenage mother, knowledge on sexual and reproductive health, access to contraceptive methods, age of the sexual debut and relationship status are some of the factors commonly correlated with teenage pregnancy and motherhood (see for example Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 23 of 40 Azevedo et al. 2013; Ermisch and Pevalin 2003). Furthermore, being married or in a stable relationship might influence the decision to use of using contraceptive methods. Nevertheless, these factors have been not included in the analysis mainly for two reasons: first, to preserve comparability across the different risky behaviours considered; second, some of those variables are only collected at Round 4.10 In Table 6 we report the results for the risky behaviours models including educational aspirations. Keeping everything else constant, in this model we observe that aspiring for higher education reduced the likelihood of engaging in criminal behaviours by 23 percentage points. Higher aspirations are also negatively correlated with the total number of risky behaviours. The role played by the other predictors (the same as in the previous model) remains very similar. One noticeable difference is that, once aspirations are controlled for, school enrollment does not predict criminal behaviours which suggest that aspirations measured and school enrollment both measured at the age of 15 are strongly correlated. To further explore the possible differential correlation of psychosocial competencies to risky behaviours by gender, in Tables 7 and 8 we replicate the same results adding an interaction between male gender, self-esteem and self-efficacy, and between male gender and aspirations, respectively. There are two noticeable results: first, while on average boys are more likely than girls to smoke, an increase in boys’ self-esteem by one standard deviation reduce boys’ probability to smoke by 11 percentage point more than for girls. Second, girls aspiring to higher education are relatively less likely to engage in unprotected sex (by 20 percentage points). So far we show that psychosocial competencies and aspirations measured at the age of 15 predict many risky behaviours that occur at the age of 19. An empirical questions is whether this correlation is constant over time and psychosocial competencies and aspirations measured earlier in life similarly predict later behaviours. In Tables 9 and 10 we report the estimates for the risky behaviour model where early psychosocial competencies and aspirations are measured at the age of 12. Analogous to previous results, early self-esteem is negatively correlated with a number of risky behaviours: drinking and engaging in violent behaviours, drugs consumption, unprotected sex, carrying a weapon. Similarly, children aspiring to higher education at the age of 12 are less likely to engage in criminal behaviours and to carry weapons at the age of 19. In both cases, higher self-esteem and aspiring to higher education is negatively correlated with the intensity of engagement in risky behaviours more generally. This seems to suggest that higher self-esteem during childhood and throughout adolescence might play a protective role against risky behaviours later on in life. On the contrary, self-efficacy at the age of 12 (as well as at age 15) is not associated to risky behaviours at age 19. 6.2 Fixed effects estimates In order to obtain a better identification of the relationship between psychosocial competencies and the outcomes of interest, we report individual fixed effects estimates obtained by differencing risky and criminal behaviours at ages 19 and 15 on differences in psychosocial competencies at ages 15 and 12, as well as on differences in all the other control variables at ages 15 and 12. These results are reported in Tables 11 and 12. For this part of the analysis, the criminal behaviours variable is dropped because it is not observed at age 15. Gender and maternal education do not vary over time and age varies uniformly across all children between survey waves, thus they are also dropped. Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 24 of 40 Table 7 Psychosocial competencies at age 15 on Risky Behaviours at age 19, with gender interactions Drinking Drugs Unprotected Criminal Carried No. of Smoking Drinking & violence consumption sex. beh. a weapon risky beh. Self-efficacy , age 15 −0.057 −0.002 0.013 −0.057 −0.066 0.002 −0.042 −0.225 (0.273) (0.963) (0.796) (0.153) (0.400) (0.982) (0.159) (0.106) Self-esteem, age 15 −0.012 −0.053 −0.040 −0.018 −0.026 −0.099* −0.039* −0.148 (0.655) (0.207) (0.437) (0.598) (0.646) (0.098) (0.052) (0.112) Male x self-efficacy, age 15 0.085 −0.067 −0.067 0.065 0.067 0.004 0.035 0.185 (0.339) (0.319) (0.284) (0.364) (0.519) (0.967) (0.349) (0.429) Male x self-esteem, age 15 −0.111* −0.020 −0.076 0.003 0.012 −0.094 −0.012 −0.129 (0.087) (0.750) (0.361) (0.950) (0.894) (0.360) (0.799) (0.424) Child is male 0.218*** 0.228*** 0.253*** 0.125*** −0.037 0.130*** 0.021 0.556*** (0.000) (0.000) (0.000) (0.001) (0.428) (0.005) (0.259) (0.000) Age in R4 0.098*** 0.098** 0.098** 0.039* −0.001 −0.002 −0.022 0.212** (0.003) (0.043) (0.027) (0.074) (0.991) (0.962) (0.431) (0.046) Type site −rural, age 15 −0.066 −0.262*** −0.207** 0.021 0.056 0.138 0.033 −0.218 (0.505) (0.000) (0.023) (0.749) (0.469) (0.165) (0.348) (0.260) Migrated between 15 and 18 0.052 0.133 0.152 0.026 0.136 −0.152 −0.089** 0.258 (0.503) (0.141) (0.157) (0.739) (0.154) (0.274) (0.017) (0.366) Second tercile of wealth, age 15 −0.004 0.081 0.083 −0.008 −0.007 −0.059 −0.005 0.057 (0.949) (0.262) (0.250) (0.845) (0.887) (0.316) (0.882) (0.743) Third tercile of wealth, age 15 −0.086 0.070 0.048 −0.022 0.055 −0.023 −0.025 −0.008 (0.219) (0.380) (0.623) (0.691) (0.419) (0.706) (0.310) (0.966) Mother’s education - secondary school 0.039 −0.025 0.028 0.027 0.039 0.002 −0.018 0.061 (0.332) (0.539) (0.525) (0.465) (0.462) (0.960) (0.486) (0.524) Mother’s education - higher education 0.059 0.003 0.058 0.041 −0.064 0.091 0.021 0.060 (0.363) (0.966) (0.477) (0.561) (0.248) (0.240) (0.464) (0.751) Single parent, age 15 0.071 0.039 0.090 0.071 0.139*** 0.168** 0.047* 0.367** (0.284) (0.492) (0.114) (0.190) (0.007) (0.024) (0.098) (0.021) Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 31 of 40 Table 12 Fixed effects estimation of educational aspirations on participation in risky behaviours at age 19 Drinking Drugs Unprotected Carried No. of Smoking Drinking & violence consumption sex. a weapon risky beh. Child aspired for higher education 0.071 0.060 0.041 0.028 −0.035 0.035 0.160 (0.367) (0.459) (0.641) (0.682) (0.629) (0.598) (0.431) Type site - rural −0.237*** −0.527*** −0.665*** −0.235*** −0.352*** 0.040 −1.312*** (0.002) (0.000) (0.000) (0.000) (0.000) (0.520) (0.000) Migrated between rounds 0.141** 0.248*** 0.331*** 0.091* 0.138* −0.038 0.579*** (0.025) (0.002) (0.000) (0.068) (0.090) (0.501) (0.003) Wealth tercile : middle −0.059 0.032 −0.004 −0.024 0.048 0.008 0.005 (0.308) (0.606) (0.952) (0.461) (0.455) (0.848) (0.972) Wealth tercile: top −0.145** 0.070 0.001 −0.089** 0.024 −0.017 −0.157 (0.038) (0.379) (0.995) (0.046) (0.779) (0.749) (0.379) Single parent 0.073 0.135 0.191** 0.044 0.133* 0.023 0.409** (0.314) (0.124) (0.026) (0.275) (0.090) (0.354) (0.023) Number of siblings −0.012 −0.013 −0.011 0.001 −0.042* 0.000 −0.065 (0.519) (0.538) (0.640) (0.936) (0.069) (0.974) (0.227) Child is enrolled −0.128 −0.229** −0.170 −0.151* −0.255** 0.081 −0.682*** (0.116) (0.045) (0.171) (0.079) (0.013) (0.386) (0.004) PPVT z-score −0.026 −0.088*** −0.050 −0.051*** 0.003 0.005 −0.157** (0.307) (0.006) (0.157) (0.005) (0.935) (0.789) (0.019) Math z-score −0.021 −0.058** −0.062** 0.019 −0.029 −0.015 −0.104 (0.399) (0.045) (0.037) (0.346) (0.324) (0.400) (0.138) Number of observations 872 872 872 872 872 872 872 R-squared 0.055 0.123 0.141 0.074 0.096 0.011 0.156 Note: The table reports the estimates for the individual fixed effects model, *p<0.1 **p<0.05 ***p<0.01. All controls were included as reported Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 32 of 40 In Table 11, we report the results for the individual fixed effects estimates using self-esteem and self-efficacy as predictors of risky behaviours. The results are qualitatively similar to the ones discussed above. An increase in self-esteem is negatively correlated to the the prevalence of risky behaviours over time. More specifically, one standard deviation increase in self-esteem reduces smoking, drinking and engaging in drinking and violence by 5, 10 and 9 percentage points respectively; it does not predict the likelihood of carrying a weapon, but the point estimate is very similar (3 percentage points). In contrast to the results in the previous model, self-efficacy is predictive of carrying a weapon. One standard deviation increase in self-efficacy reduces the probability of carrying a weapon by 5 percentage points. Besides this, in this set of estimations schooling achievement is found to play a more prominent role. School enrollment reduces the likelihood of drinking, drugs consumption, and risky sex. A similar role is played by vocabulary and math achievement. In addition, coming from a singleparent household, area of location and migration remain as important predictors of risky behaviours. In Table 11, we report the results for the individual fixed effect model including aspirations among the predictors. However, in this case we are not able to detect a relationship between aspirations and the outcomes of interest. To summarize, the fixed effects estimates show that the relationship between selfesteem and risky behaviours is very robust, whereas the relationship between self-efficacy, aspirations and risky behaviours is not. In addition, there seems to be a lot of meaningful variation over time in the control variables, which allows us to show that coming from a single-parent household, area of location, migration and schooling achievement are also important factors that play a role in the determination of risky and criminal behaviours. 7 Conclusions and discussion There is a growing concern about the prevalence of risky behaviours among the youth population, which ultimately leads to worse outcomes later in life, including lower salaries and worse socio-economic and life outcomes. On the other hand, there is little evidence about the prevalence of these behaviours and their determinants in the context of developing countries. Our aim is to try to fill this gap using a unique individual-level panel data from Peru following a cohort of children for over a decade between the ages of 8 and 19. We constructed indicators to measure the prevalence of smoking and drinking; engaging in risky behaviours when drunk; consumption of illegal drugs; unprotected sex; criminal behaviours; possession of weapons; and total number of risky behaviours. While we do not claim any causal relation, the methods used allow us to deal with bias arising from reverse causality and omitted variables that are constant over time. From this analysis we identify a number of drivers of risky behaviours. In particular, there is a specific group of the youth at risk; boys, living in urban areas and growing up in single-parents households. In the case of girls, they are more likely to be exposed to unprotected sex. Although these groups are identified for the Peruvian context, similar patterns are likely to be observed in countries with similar characteristics (middleincome countries with relatively high levels of poverty and low levels of secondary school attainment). We also observe a dramatic increases in risky behaviours between age 15 and 19 which suggests that policy interventions aiming at preventing risky behaviour should be put Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 33 of 40 in place at age 15 or earlier, when risky behaviours only manifest in a small part of the population. Although the present analysis is not sufficient to claim any causal relation between socio-emotional competencies and risky behaviours provides some interesting hints. Our results suggest that psychosocial competencies, and self-esteem and high aspirations in particular, might play a role in reducing risky behaviours.This connects well with evidence from the psychological literature that finds a similar relationship in developed countries. To our knowledge this evidence is unique in the developing countries context, and provides an important message: policies aimed at promoting soft skills during childhood and adolescence can play an important role as a mechanism to reduce risky and criminal activities among the youth. From a policy perspective, considering the age range analysed as well as the fact that, by age 15, most Peruvian adolescents are still attending school, we argue that it is worth to explore whether interventions designed to take place at secondary-level schools can reduce the engagement of adolescents in risky behaviours. In terms of more comprehensive interventions, the Minister of Education in Peru is currently implementing an Extended School Day Program (Jornada Escolar Completa, JEC). This initiative seeks both to extend the length of the school-day and to provide better services to students at the secondary level in urban areas. Theoretically, JEC and similar initiatives can have direct as well as indirect effects on the prevalence of risky behaviours. First of all, longer school hours implies that students spend a greater number of hours per day under adult supervision, limiting the possibility to engage in risky behaviours (Bellei 2009; Aguero and Beleche 2013). Further, inasmuch as extended school days have been found to improve academic achievement in middleincome countries, this type of programme can be expected to reduce the prevalence of risky behaviours by increasing the opportunity cost of engaging in them (indirect effect). Finally, as part of the JEC programme in Peru a full-time psychologist has been incorporated into every JEC school to improve students’ psycho-social well-being. Our results suggest improving psychological competencies might be an additional mechanisms through which the JEC might reduce the prevalence of risky behaviours. Similar programme are currently being implemented in the Latin American region (in Chile, Colombia, Mexico and Uruguay). In the case of Chile, a nation-wide education reform extended the school day from 32 to 39 hours per week. Berthelon and Kruger (2011) find teens living in municipalities with greater access to full-day high schools had a lower probability of becoming mothers during their adolescence. An increase of 20 percentage points in the municipal share of full-day high schools reduces the probability of motherhood in adolescence by 3.3%. This encouraging findings from Chile suggests that it is worthwhile to explore the potential effects of this type of reforms and risky behaviours. Further research on JEC in Peru and its effect on risky behaviours will be done using the next round of data. Endnotes 1There is an ongoing debate, and little agreement, on how to refer to those skills which represent the “patterns of thought, feelings and behaviour” Borghans et al. (2008b) and that encompass those traits that are not directly represented by cognitive skills or by formal conceptual understanding. The current list includes such terms as behavioural Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 34 of 40 skills, soft skills, personality traits, non-cognitive skills or abilities, character, life-skills, socio-emotional and psychosocial skills or competencies. In this paper, we use the term “soft skills” and “psychosocial competencies” interchangeably. 2It is important to note that information about cognitive and psychosocial competencies are collected for all children regardless of their school enrollment status which avoids any selection problem commonly arising using school-based tests. 3These include 3 clusters in the department of Lima, and 17 in Amazonas, Ancash, Apurimac, Arequipa, Ayacucho, Cajamarca, Huanuco, Junin, La Libertad, Piura, Puno, SanMartinandTumbes. 4For more details about the sampling design see (Escobal and Flores 2008). 5In other words they could be defined as “social drinkers”. “Social drinking” refers to casual drinking in a social setting without necessarily an intent to get drunk. 6Unfortunately, Young Lives collects information only about the use of contraceptive methods in the last sexual relationship. 7The intensity variable includes all the risky behaviours variables as defined above. With respect to alcohol consumption we include the “drinking” variable only. 8It is worth noting that the correlation between these scales is 0.25. This suggests that the two scales capture different dimensions of the child. 9Table 2 documents the indicators used in the analysis, and their definitions or procedure of computation. 10 It is important to note that including a dummy for marital/cohabiting status and an indicator for the child’s knowledge about sexual reproductive the estimated coefficients for self-efficacy and self-esteem do not qualitatively change. however, the inclusion of those variables improve the statistical fit of our model and the R-squared increases from 0.09 to 0.16. Appendix Table 13 Consumption of cigarettes, alcohols and drugs Age 15 Age 19 %n%n Alcohol consumption How often do you drink alcohol? Everyday 0.5 3 0.7 4 At least once a week 1.6 10 3.0 18 At least once a month 3.6 23 5.9 35 Only on special occasions 16.2 104 31.8 190 Hardly ever 13.1 84 29.1 174 I never drink alcohol 65.1 417 29.6 177 How much do you usually drink per day? I never drink alcohol 69.5 417 35.5 177 1 cup/glass or less 18.3 152 28.8 213 2 cups/glasses 6.5 38 13.2 76 3 cups/glasses or more 5.7 34 22.6 132 Have you ever been drunk for too much alcohol? Yes 11.5 68 35.2 211 No 88.5 522 64.8 388 Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 35 of 40 Table 13 Consumption of cigarettes, alcohols and drugs (Continued) Cigarettes consumption How old were you when you tried a cigarette for the first time? Average age NA 16.0 How often do you smoke cigarettes now? Everyday 0.6 4 1.0 6 At least once a week 3.0 19 6.8 40 At least once a month 3.7 24 12.2 72 Hardly ever 14.0 90 27.1 160 I never smoke cigarettes 78.7 505 53.0 313 How many cigarettes do you usually smoke per day? I never smoke cigarettes 78.7 505 67.3 313 1 cigarette or less per day 18.5 119 27.1 248 2 to 5 cigarettes per day 2.3 15 5.0 27 6 or more per day 0.5 3 0.5 3 Drugs consumption Have you ever tried drugs? Yes 3.1 20 14.2 84 No 96.7 617 85.8 508 Sexual behaviours How old were you when you had sex for the first time? Average age NA 16 Ever had sex? Yes 19.4 109 67.2 391 No 80.6 453 32.8 191 Used condom on last sexual relation Yes 12.6 71 40.5 236 No 6.8 38 26.5 155 Never had sex 80.6 453 32.8 191 Criminal behaviours During the last 30 days, on how many days did you carry a weapon? Never 91.9 588 3.2 567 1 day 5.6 36 0.7 19 2 to 3 days 0.8 5 1.5 4 More than 4 days 1.7 11 94.7 9 Have you ever been member of a gang? Yes NA 5.5 33 No NA 94.5 565 Have you ever been arrested by the police for illegal behaviour? Yes NA 5.8 35 No NA 94.2 567 Have you ever been sentenced to spend time in a corrections institution? Yes NA 6.7 10 No NA 93.4 591 Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 36 of 40 Table 14 Consumption of cigarrettes, alcohols and drugs at Age 18 by gender and location Female Male Urban Rural %n %n test %n %n test Alcohol consumption How often do you drink alcohol? Everyday 0.4 1 0.7 2 −0.3 0.5 2 0.7 1 −0.3 At least once a week 0.8 2 5.3 16 −4.5*** 3.9 17 0.7 1 3.1* At least once a month 3.3 9 7.9 24 −4.6** 5.7 25 5.8 8 −0.1 Only on special occasions 30.6 83 32.9 100 −2.3 33.5 148 26.8 37 6.7 Hardly ever 24.0 65 34.5 105 −10.55*** 30.8 136 25.4 35 5.4 I never drink alcohol 41.0 111 18.8 57 22.21*** 25.8 114 40.6 56 −14.8*** How much do you usually drink per day? I never drink alcohol 41.0 111 18.8 57 22.21*** 31.2 114 46.3 56 −14.8*** 1 cup/glass or less 38.4 104 32.9 100 5.48 29.5 160 27.2 46 2.87 2 cups/glasses 60 8.9 24 16.5 50 −7.59*** 14.2 60 10.3 14 3.43 3 cups/glasses or more 11.8 32 31.9 97 −20.10*** 25.1 108 16.2 22 8.49** Have you ever been drunk for too much alcohol? Yes 22.4 60 44.4 134 −22.0*** 37.7 165 22.6 31 15.0*** No 77.6 208 55.6 168 62.3 273 77.4 106 Cigarettes consumption How old were you when you tried a cigarette for the first time? Average age 16.1 16.0 14.1 16.5 How often do you smoke cigarettes now? Everyday 0.7 2 1.3 4 −0.6 1.4 6 0.0 0 1.4 At least once a week 2.6 7 10.7 32 −8.1*** 7.5 33 5.1 7 2.4 At least once a month 4.4 12 18.4 55 −14.0*** 11.6 51 11.7 16 −0.0 Hardly ever 18.8 51 35.1 105 −16.3*** 27.2 119 28.5 39 −1.3 I never smoke cigarettes 73.4 199 34.5 103 39.0*** 52.3 229 54.7 75 −2.5 How many cigarettes do you usually smoke per day? I never smoke cigarettes 73.4 199 34.4 103 39.0*** 52.3 229 54.7 75 −2.5 1 cigarette or less per day 25.1 68 57.5 172 −32.4*** 42.0 184 42.3 58 −0.3 2 to 5 cigarettes per day 1.5 4 7.0 21 −5.6*** 5.3 23 2.2 3 3.6 6 or more per day 0.0 0 1.0 3 −1.0* 0.5 2 0.7 1 −0.3 Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 37 of 40 Table 14 Consumption of cigarettes, alcohols and drugs at Age 18 by gender and location (Continued) Drugs consumption Have you ever tried drugs? Yes 7.6 20 18.8 57 −11.23*** 14.0 61 12.4 17 1.6 No 92.4 244 81.2 246 86.0 374 87.6 120 Sexual behaviours How old were you when you had sex for the first time? Average age 16.6 16.0 16.1 16.5 Ever had sex? Yes 55.1 146 78.8 231 −23.7*** 66.7 288 70.2 92 −3.56 No 44.9 119 21.2 62 33.3 144 29.8 39 Used condom on last sexual relation Yes 24.2 64 55.6 163 −31.5 *** 39.8 172 43.5 57 −3.7 No 30.9 82 23.2 68 7.7 ** 26.9 116 26.7 35 0.1 Never had sex 44.9 119 21.2 62 23.7 *** 33.3 144 29.8 39 3.6 Criminal behaviours During the last 30 days, on how many days did you carry a weapon? Never 96.0 262 93.7 284 2.2 95.3 422 93.5 129 1.8 1 day 2.2 6 4.0 12 −1.8 3.2 14 2.9 4 0.3 2 to 3 days 0.8 2 0.7 2 0.1 0.5 2 1.5 2 −1.0 More than 4 days 1.1 3 1.7 5 0.6 1.1 5 2.2 3 −1.0 Have you ever been member of a gang? Yes 2.9 8 7.6 23 −4.7** 5.9 26 3.6 5 2.3 No 97.1 264 92.4 280 94.1 416 96.4 133 Have you ever been arrested by the police for illegal behaviour? Yes 2.6 7 8.2 25 −5.6*** 5.4 24 5.8 8 −0.4 No 97.4 267 91.8 280 94.6 421 94.2 131 Have you ever been sentenced to spend time in a corrections institution? Yes 4.0118.526 −4.5** 5.9 26 7.9 11 −2.1 No 96.0 262 91.5 279 94.1 418 92.1 128 Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 38 of 40 Abbreviations DEVIDA: The national committee for a life without drugs; JEC: Extended school day programme (Jornada Escolar completa); OLS: Ordinary least squares; PPVT: Peabody picture vocabulary test; R2: Round 2; R3: Round 3; R4: Round 4; STDs: Sexual transmitted disesases; TORA: Theory of rational addiction; US: United States Acknowledgements Thanks to Maria Gracia Rodriguez and Grace Chang for excellent research assistance. We would also like to thank the anonymous referees and the editor for the useful remarks. Responsible editor: David Lam. Funding Young Lives is an international study of childhood poverty, following the lives of 12,000 children in 4 countries (Ethiopia, India, Peru and Vietnam) over 15 years. www.younglives.org.uk. Young Lives is core-funded from 2001 to 2017 by UK aid from the Department for International Development (DFID), and co-funded by IrishAid from 2014 to 2015. The William and Flora Hewlett Foundation fund aspects of Young Lives gender research (2014-2016), including the research carried out for this paper. The views expressed are those of the authors. They are not necessarily those of, or endorsed by, Young Lives, the University of Oxford, DFID or other funders. Competing interests The IZA Journal of Labor & Development is committed to the IZA Guiding Principles of Research Integrity. The authors declare that they have observed these principles. Author details 1University of Oxford, Queen Elizabeth House, 3 Mansfield Road, OX1 3TB Oxford, UK. 2Grupo de Análisis para el Desarrollo, GRADE, Av. Almte. Miguel Grau 915, Lima, Distrito de Lima 15063, Peru. Received: 14 July 2016 Accepted: 19 December 2016 References Agnew R, Brezina T, Wright JP, Cullen FT (2002) Strain, personality traits, and delinquency: Extending general strain theory. Criminology 40(1):43–72 Aguero JM, Beleche T (2013) Test-mex: Estimating the effects of school year length on student performance in mexico. J Dev Econ 103(C):353–361 Aizer A (2004) Home alone: supervision after school and child behavior. J Public Econ 88(9–10):1835–1848 Amato P, Ochiltree G (1986) Family resources and the development of child competence. J Marriage Fam 48(1):47–56 Averett S, Argys L, Rees D (2011) Older siblings and adolescent risky behavior: does parenting play a role? J Popul Econ 24(3):957–978 Azevedo JP, Favara M, Haddock SE, Lopez-Calva LF, Muller M, Perova E (2013) Teenage pregnancy and opportunities Latin America and the Caribbean on early child bearing, poverty, and economic achievement. World Bank Group, Washington, DC. http://documents.worldbank.org/curated/en/983641468238477531/Embarazoadolescenteresumen-2013 Bandura A (1993) Perceived self-efficacy in cognitive development and functioning. Educ Psychol 28(2):117–148 Bandura A, Barbaranelli C, Caprara G, Pastorelli C (2001) Self-efficacy beliefs as shapers of children’s aspirations and career trajectories. Child Dev 72(1):187–206 Bandura A, Caprara G, Barbaranelli C, Gerbino M, Pastorelli C (2003) Role of affective self-regulatory efficacy in diverse spheres of psychosocial functioning. Child Dev 74(3):769–782 Baumeister R, Campbell J, Krueger J, Vohs K (2003) Does high self-esteem cause better performance, interpersonal success, happiness, or healthier lifestyles? Psychol Sci Public Interes 4(1):1–44 Becker G, Mulligan C (1997) The endogenous determination of time preference. Q J Econ 112(3):729–58 Becker GS, Murphy KM (1988) A theory of rational addiction. J Polit Eco 96(4):675–700 Behrman J (2015) The effect of increased primary schooling on adult women’s hiv status in malawi and uganda: Universal primary education as a natural experiment. Soc Sci Med 127:108–115 Bellei C (2009) Does lengthening the school day increase students academic achievement? results from a natural experiment in chile. Econ Educ Rev 28(5):629–640 Berthelon M, Kruger D (2011) Risky behavior among youth: Incapacitation effects of school on adolescent motherhood and crime in Chile. J Public Econ 95(1-2):41–53 Borghans L, Duckworth A, Heckman J, Weel BT (2008a) The economics and psychology of personality traits. J Hum Resour 43(4):972–1059 Borghans L, Meijers H, ter Weel B (2008b) The role of noncognitive skills in explaining cognitive test scores. Econ Inq 46(1):2–12 Carneiro PM, Heckman JJ (2003) Human capital policy. IZA Discussion Paper Series, No. 821, Institute for the Study of Labor (IZA). http://ftp.iza.org/dp821.pdf Caspi A, Moffitt TE, Silva PA, Stouthamer-Loeber M, Krueger RF, Schmutte PS (1994) Are some people crime–prone? replications of the personality-crime relationship across countries, genders, races and methods. Criminology 32(2):163–196 Chaloupka FJ, Laixuthai A (1997) Do youths substitute alcohol and marijuana? some econometric evidence. East Econ J 23(3):253–276 Chiteji N (2010) Time preference, noncognitive skills and well being across the life course: Do noncognitive skills encourage healthy behavior? Am Econ Rev 100(2):200–204. doi:10.1257/aer.100.2.200 Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 39 of 40 Clark AE, Loheac Y (2007) “it wasn’t me, it was them!” social influence in risky behavior by adolescents. J Health Econ 26(4):763–784 Cobb-Clark DA, Tan M (2011) Noncognitive skills, occupational attainment, and relative wages. Labour Econ 18(1):1–13 Conti G, Heckman JJ (2010) Understanding the early origins of the education–health gradient. Perspect Psychol Sci 5(5):585–605 Conti G, Heckman JJ, Moon SH, Pinto R (2015) Long-term health effects of early childhood interventions. IZA Discussion Paper Series, No. 9247, Institute for the Study of Labor (IZA). http://ftp.iza.org/dp9247.pdf Crookston BT, Merrill RM, Hedges S, Lister C, West JH, Hall PC (2014) Victimization of peruvian adolescents and health risk behaviors: Young Lives cohort. BMC Public Health 14(1):1–7 Cueto S, Leon J (2012) Psychometric characteristics of cognitive development and achievement instruments in round 3 of Young Lives. Young Lives Technical Note No. 25. http://www.younglives.org.uk/content/psychometriccharacteristicscognitive-development-and-achievement-instruments-round-3-young Cueto S, Leon J, Guerrero G, Muñoz I (2009) Psychometric characteristics of cognitive development and achievement instruments in round 2 of Young Lives. Young Lives Technical Note No. 15. http://www.younglives.org.uk/content/ psychometric-characteristicscognitive-development-and-achievement-instruments-round-2-young Cueto S, Saldarriaga V, Munoz IG (2011) Conductas de riesgo entre adolescentes peruanos: un enfoque longitudinal. Salud, interculturalidad y comportamiento de riesgo, Lima: GRADE:119–58. http://www.grade.org.pe/publicaciones/ 1052-conductas-de-riesgo-entreadolescentes-peruanos-un-enfoque-longitudinal/ Cunha F, Heckman JJ (2007) The technology of skill formation. Am Econ Rev 97(2):31–47. doi:10.1257/aer.97.2.31 Cunha F, Heckman, JJ (2008) Formulating, identifying and estimating the technology of cognitive and noncognitive skill formation. J Hum Resour 43(4):738–782 Cunha, F, Heckman JJ, Schennach SM (2010) Estimating the technology of cognitive and noncognitive skill formation. Econometrica 78(3):883–931. doi:10.3982/ECTA6551 Cutler D, Lleras-Muney A (2010) Understanding differences in health behaviors by education. J Health Econ 29(1):1–28 Dalton PS, Ghosal S, Mani A (2016) Poverty and aspirations failure. Econ J 126(590):165–188 De Paula A, Shapira G, Todd PE (2013) How beliefs about hiv status affect risky behaviours: Evidence from malawi. J Appl Econom 29(6):944–964. doi:10.1002/jae.2342 De Walque D (2007) Does education affect smoking behaviors? evidence using the vietnam draft as an instrument for college education. J Health Econ 27(5):877–895 Dee T (1999) The complementarity of teen smoking and drinking. J Health Econ 18(6):769–793 Delavande A (2008) Pill, patch or shot? subjective expectations and birth control choice. Int Econ Rev 49(3):999–1042. doi:10.1111/j.1468-2354.2008.00504.x Dercon S, Krishnan P (2009) Poverty and the psychosocial competencies of children: evidence from the young lives sample in four developing countries. Childre, Youth Environ 19(2):138–163 DEVIDA (2013) Comisión nacional para el desarrollo y vida sin drogas - iv estudio nacional prevencion y consumo de drogas en estudiantes de secundaria 2012. Technical report. http://www.devida.gob.pe/wpcontent/uploads/2014/ 12/Informe-Ejecutivo_-IV-Estudio-Nacional-Prevenci%C3%B3n-y-Consumo-de-Drogas-en-Estudiantes-deSecundaria-2012.pdf DiNardo J, Lemieux T (2001) Alcohol, marijuana, and American youth: The unintended effects of government regulation. Journal of Health Economics. 991–1010 Dohmen, T, Falk A, Huffman D, Sunde U (2010) Are risk aversion and impatience related to cognitive ability? Am Econ Rev 100(3):1238–60. doi:10.1257/aer.100.3.1238 Donnellan M, Trzesniewski K, Robins R, Moffitt T, Caspi A (2005) Low self-esteem is related to aggression, antisocial behavior, and delinquency. Psychol Sci 16(4):328–335 DuRant R, Smith J, Kreiter S, Krowchuk D (1999) The relationship between early age of onset of initial substance use and engaging in multiple health risk behaviours among young adolescents. Archives of PediatricsArch Pediatr Adolesc Med 153(3):286–291. doi:10.1001/archpedi.153.3.286 Epstein J, Griffin K, Botvin G (2004) Efficacy, self-derogation, and alcohol use among inner-city adolescents: Gender matters. J Youth Adolesc 33(2):159–166 Ermisch J, Pevalin DJ (2003) Who has a child as a teenager? ISER Working Paper Series:30. https://www.iser.essex.ac.uk/ research/publications/working-papers/iser/2003-30 Escobal J, Flores E (2008) An assessment of the young lives sampling approach in Peru. Young Lives Technical Note (3). http://www.younglives.org.uk/content/assessment-young-lives-samplingapproach-peru Evans W, Oates W, Schwab R (1992) Measuring peer group effects: A study of teenage behavior. J Polit Econ 100(5):966–991 Farrelly MC, Bray JW, Zarkin GA, Wendling BW (2001) The joint demand for cigarettes and marijuana: Evidence from the national household surveys on drug abuse. J Health Econ 20(1):51–68 Fischhoff B (1992) Risk taking: A developmental perspective. In: Yates JF (ed). Risk Taking Behavior. John Wiley & Sons, New York. pp 133–162 Fuchs V (1982) Time preference and health: An explorative study. In: Victor R Fuchs (ed). Economic Aspects of Health. University of Chicago Press, Chicago. pp 93–120 Gaviria A, Raphael S (2001) School-based peer effects and juvenile behavior. Rev Econ Stat 83(2):257–268 Goodman E (1999) The role of socioeconomic status gradients in explaining differences in us adolescents’ health. Am J Public Health 89:1522–28 Gruber J (2001) Risky Behavior among Youths: An Economic Analysis, National Bureau of Economic Research, Chicago: University of Chicago Press. http://www.nber.org/books/grub01-1 Gruber J, Koszegi B (2001) Is addiction ‘rational’? theory and evidence. Q J Econ 116(4):1261–1303 Heckman J, Stixrud J, Urzua S (2006) The effects of cognitive and noncognitive abilities on labor market outcomes and social behavior. J Lab Econ 24(3):411–482 Heckman JJ, Moon SH, Pinto R, Sayelyev PA, Yavitz A (2010) Analyzing social experiments as implemented: A reexamination of the evidence from the highscope Perry Preschool Program. Quant Econ. 1(1):1–46 Favara and Sanchez IZA Journal of Labor & Development (2017) 6:3 Page 40 of 40 Hill PL, Roberts BW, Grogger JT, Guryan J, Sixkiller K (2011) Decreasing delinquency, criminal behavior, and recidivism by intervening on psychological factors other than cognitive ability: A review of the intervention literature. NBER 1030 Working Paper (16698). http://www.nber.org/papers/w16698 Jaeger D, Dohmen T, Falk A, Human D, Sunde U, Bonin H (2010) Direct evidence on risk attitudes and migration. Rev Econ Stat 92(3):684–689 Kenkel D (1991) Health behavior, health knowledge and schooling. J Polit Econ 99(2):287–305 Knudsen E, Heckman J, Cameron J, Shonkoff J (2006) Economic, neurobiological, and behavioral perspectives on building america’s future workforce. Proc Natl Acad Sci 103(27):10,155–10,162 Lavado P, Aragón C, Gonzales M (2015) Cuál es la relación entre las habilidades cognitivas y no cognitivas y la adopción de comportamientos de riesgo un estudio para el Perú. Apuntes: Revista de Ciencias Sociales 42(76) Lewis CC (1981) How adolescents approach decisions: Changes over grades seven to twelve and policy implications. Child Dev 52:538–44 Lochner L, Moretti E (2004) The effect of education on crime: Evidence from prison inmates, arrests, and self-reports. Am Econ Rev 94(1):155–189. doi:10.1257/000282804322970751 Loewenstein G, O’Donoghue T, Rabin M (2003) Projection bias in predicting future utility. Q J Econ (2003) 118(4):1209–1248 Meschke L, Silbereisen R (1997) The influence of puberty, family processes, and leisure activities on the timing of first sexual experience. J Adolesc 20(4):403–418 Model KE (1993) The effect of marijuana decriminalization on hospital emergency room drug episodes: 1975–1978. J Am Stat Assoc 88(423):737–747 O’Donoghue T, Rabin M (2001) Risky Behavior among Youths: Some Issues from Behavioral Economics. In: Risky Behavior among Youths: An Economic Analysis. University of Chicago Press. pp 29–68. http://www.nber.org/chapters/c10686 Parkes A, Wight D, Henderson M, West P (2010) Does early sexual debut reduce teenagers’ participation in tertiary education? evidence from the share longitudinal study. J Adolesc 33(5):741–754. doi:10.1016/j.adolescence.2009.10.006 Peters E, Baker D, Dieckmann N, Leon J, Collins J (2010) Explaining the effect of education on health: A field study in ghana. Psychol Sci 21(10):1369–1376 Pratt TC, Cullen FT (2000) The empirical status of gottfredson and hirschi’s general theory of crime: a meta-analysis. Criminology 38(3):931–964. doi:10.1111/j.1745-9125.2000.tb00911.x Rosenberg M (1965) Society and the adolescent self-image. Princeton University Press, Princeton, NJ Rotter JB (1966) Generalized expectancies for internal versus external control of reinforcement. Psychol Monogr Gen Appl 80(1):1–28 Salazar LF, Crosby RA, DiClemente RJ, Wingood GM, Lescano CM, Brown LK, Harrington K, Davies S (2005) Self-esteem and theoretical mediators of safer sex among african american female adolescents: Implications for sexual risk reduction interventions. Health Educ Behav 32(413) Schwarzer R (2001) Social-cognitive factors in changing health-related behaviors. Curr Dir Psychol Sci 10(2):47–51 Shapira G (2013) How subjective beliefs about HIV infection affect life-cycle fertility: evidence from rural Malawi. Policy Research Working Paper, The World Bank (6343). http://elibrary.worldbank.org/doi/abs/10.1596/1813-9450-6343 Steinberg L (1987) Single parents, stepparents, and the susceptibility of adolescents to antisocial peer pressure. Child Dev 58(1):269–275 Travis H, Hindelang MJ (1977) Intelligence and delinquency: A revisionist review. Am Sociol Rev 42(4):571–587 Wellings K, Collumbien M, Slaymaker E, Singh S, Hodges Z, Patel D, Bajos N (2006) Sexual behaviour in context: a global perspective. Lancet 368(9548):1706–1728 Wiefferink CH, Peters L, Hoekstra F, Dam GT, Buijs GJ, Paulussen TMGW (2006) Clustering of health-related behaviors and their determinants: Possible consequences for school health interventions. Prev Sci 7(2):127–149. doi:10.1007/s11121-005-0021-2 Wills T (1994) Self-esteem and perceived control in adolescent substance use: Comparative tests in concurrent and prospective analyses. Psychol Addict Behav 8(4):223–234 Wolfe B, Haveman R, Pence K, Schwabish JA (2007) Do youth nonmarital childbearing choices reflect income and relationship expectations? J Popul Econ 20(1):73–100 Submit your manuscript to a journal and benefi t from: 7 Convenient online submission 7 Rigorous peer review 7 Immediate publication on acceptance 7 Open access: articles freely available online 7 High visibility within the fi eld 7 Retaining the copyright to your article Submit your next manuscript at 7 springeropen.com