Does the date of birth affect educational achievements and well-being in youth? Possibly yes. This would not be due to astrological forces; rather, it could be due to the distance of the birthdate from the so called cut-off date, that is, the date that separates students in different grades and determines maturity differences within a class. For instance, in Austria the cut-off date is September 1st, which implies that in the same grade there are students born from September 1st of year t to August 31st of year t+1. Therefore, usually, these maturity differences between classmates consist of at most one year. There is large evidence that this maturity gap jeopardizes human capital accumulation. The age difference between classmates, called “relative age,” causes gaps in perceived ability and performance, called “relative age effects (RAEs).” These RAEs then reflect into worse educational paths for the youngest students in a class; moreover, it reflects into non-cognitive abilities and well-being gaps. It is natural to expect that the youngest students in a class experience a lower life-satisfaction as well, and the study of this effect of relative age on adolescents is economically relevant for at least two reasons. First, life-satisfaction in adolescence is the best predictor of adults’ life-satisfaction and emotional health. Second, life-satisfaction seems to affect adolescents’ success in education and labor market outcomes, both directly—through the effect on school performance, and indirectly—for instance, by affecting the big five personality traits (e.g. openness and neuroticism) and self-esteem. No study has so far investigated whether relative age affects life-satisfaction, however. The first contribution of our study fills this gap in the literature. We investigate a representative sample of adolescents from Europe (about 380,000 students from 32 countries), from the survey “Health Behaviour in School-Aged Children (HBSC),” which is currently the only survey to allow cross-country comparisons of adolescents’ life-satisfaction. This paper provides a second original contribution. There has been a recent upsurge of studies that investigate the reasons why life-satisfaction decreases between the age of 10 and 16. Currently, some potential explanations for this phenomenon are the increase in school related stress, the increase in depression, the innate optimism that turns into a more realistic world view, and body changes. The possible role of relative age in determining a worsening of life-satisfaction has so far been neglected, however. Our study poses to fill this gap in the literature. Why could relative age worsen life-satisfaction in adolescence, on average? The decrease in the importance of relative age, which occurs with the increase in absolute age, reduces various types of negative gaps owed to maturity gaps. For instance, with the increase in absolute age, maturity gaps reduce and so do performance gaps; thus, the oldest students in a class realize that their superior performance in earlier stages of their education was due—at least partly—to their greater maturity rather than greater talent; in turn, this might have a negative impact on their life-satisfaction. The youngest students in a class might have a diametric realization; thus, for them, the reduction in performance gaps might have a less negative effect on life-satisfaction. Based on this background, we test whether the interaction of relative and absolute age determine an improvement of life-satisfaction. Evidence of this effect would imply two things: (i) students who are among the oldest in their class suffer a larger decrease in life-satisfaction in adolescence; and (ii) life-satisfaction gaps may disappear in the long-term. We find results that partly align with our expectations. There is evidence that regular students who are one year younger than the oldest regular student in class are less satisfied with their life. Moreover, country-level analyses suggest that RAEs might be lower in countries with high age-compliance, where educational settings create fewer disadvantages to the youngest students. Finally, we do not find evidence that this negative gap in life satisfaction decreases with absolute age; this result is somehow alarming, because it suggests that this negative gap in life-satisfaction could prolong into adulthood. These results lead to an important policy implication. In order to improve the life-satisfaction of the youngest students in a class, the age-grouping system could be shortened, so that the largest possible within-class age difference was down to, e.g., 6-9 months. Although costly in the short-run, a reform of the age-grouping system promises positive long-term returns in terms of life-satisfaction and emotional health. For future studies, a logical next step might be to investigate (possible) RAEs in life-satisfaction among adults.
The severe economic, political and social crisis has upset the EU labour market plunging the most part of member states in an employment collapse. In this context, skills development is highly important for building the virtuous circle in which the quality of education and training stimulates a new value of labour and a “new Deal” for the EU labour market. Innovation, investment, technological change, enterprise development, economic diversification and competitiveness surely boost the creation of additional and more profitable jobs. However, some imbalances due to increasing global competition, the skill--‐biased technological change, increasing migration from poor countries, the ageing of population still impede finding the right people for the appropriate jobs. Investigating the nature and consequences of this mismatch and measuring it appropriately is therefore essential. Occupational mismatch is usually defined as the discrepancy between the characteristics of employed workers and the requirements of the jobs that they occupy. It can be conceptualized either through education or individual’s skills. Education mismatch compares the formal education qualifications held by workers with the requirements of their jobs, while skill mismatch considers the discrepancy between the skills possessed by a worker and those required to perform his/her job. A large literature exists proposing various methodologies to measure mismatch in education and skills. The paper aims at estimating occupational mismatch using data from the first European Skills Survey collected on 48,676 adult employees across all the European Union’s 28 Member States. Taking into account the nature of our data, we apply a subjective approach that relies on information provided by the worker and which consists of using his/her opinion on whether the job matches with his/her level of both education and skills, either through direct questions or by asking employees about the requirements of their current job. Given the hierarchical structure of our data, we implement a multilevel logistic model and we assume that the probability of workers to be overeducated and overskilled is associated with two level--‐related factors – individual and country level. Indeed, our model includes several fixed effects predictors such as age, gender, education level of employees, sector and firm size, etc., at level--‐1, plus a fixed intercept and a random intercept for each level--‐2 component (e.g., one for each country). Our results show that, considering the individual characteristics of respondents, male workers who attained post--‐secondary and tertiary education have higher probability of being overeducated and of perceiving themselves overskilled, irrespective of the field of studies. Having a professional and manager position and a part--‐time contract, as well as accumulating training and experience throughout a worker's career, decrease the probability of being both overeducated and overskilled. The wage dimension seems to negatively influence the education mismatching measure: the higher is the wage penalty in terms of lower salary, the higher is the worker probability of self--‐perceiving as overeducated. Differences can be observed across firms: in the agriculture sector workers have higher probability to be overeducated, while workers in public sector report higher probability to be overskilled; concerning the firm size, we notice significant negative effect with respect to the probability of being overskilled in the small firms as expected, but without any other effect in the other cases. Our estimates indicate that the source of variability at the country level plays a non--‐negligible role in interpreting results. Including the country effect in our analysis, we observe that, in the case of overeducation, the highest effect is given for UK, France, Check Republic and Croatia, while the lowest effect is estimated in the Netherlands, Luxemburg and Romania. In the case of overskilling outcome, the largest impact is noticed in Austria, UK, Greece and Germany. On the other hand, Baltic countries show the lowest probability to be overskilled. Overeducation and overskilling continue to be problems characterising the actual workplace condition in Europe, affecting lower professions and higher educated employees. Further research can take account also of the OECD scores by countries such as PISA and PIAAC, in order to better differentiate overeducation and overskilling effects mediated by competence indicators.
Internationalization of higher education has become a priority in the European education policy. According to the strategic objectives of Europe 2020, “an EU average of at least 20% of higher education graduates should have had a period of higher education-related study or training abroad, representing a minimum of 15 ECTS credits.” (EU Council of Ministers of Education, 2011). However, learning mobility barriers still exist and the shares of study abroad participants vary widely across Member States.
Research in this area is expanding with the aim of understanding motivations and potential benefits of students’ international mobility. The aim of this paper is twofold. First, by making use of a rich dataset based on students enrolled in a medium size university of the North of Italy (the University of Bergamo), we empirically analyse the characteristics of the students who apply for a period of international mobility and we compare them with those who do not take this path. Differently from the existing studies, in which the intention to apply is analyzed, we consider those who have applied to the international mobility program.
The second aim is to investigate students’ motivations and concerns about the international mobility experience. We have conducted an on-line survey addressed to all the outgoing applicants in the a.y. 2015/16 and 2016/17 before the mobility experience. In the survey, we ask students to motivate their decision to study abroad, i.e. to enhance future employability, to enrich their CV, to live a new experience, to improve the knowledge of a foreign language, to get in touch with the culture of the host country. Furthermore, we analyse the factors that address the choice towards the host country and the university: i.e. compatibility of study programs and availability of scholarships, reputation of the host university, living costs of the host country. Finally, we ask students to express their concerns, i.e. about the different teaching methodology, in attending courses and taking exams in a foreign language, in living away from the home country.
Our survey involves about 600 students (64% females and 36% males) selected to spend one/two semesters abroad for an Erasmus+ or Extra EU program during the academic years 2015/2016 and 2016/2017. We consider the students belonging to all six departments: Engineering (10.0%), Foreign Languages (47.8%), Social Sciences (6.4%), Law (3.3%), Art and Philosophy (3.3%) and Economics (29.2%). Students study abroad mainly during the bachelor (63%) and the fall semester (53.5%). The Erasmus+ student spend their credit mobility mainly in Spain (27.5%), France (20.4%), Germany (19.7%) and United Kingdom (13.4%). After more than one reminder, the answer rate to the questionnaire is approximately 74%.
To assess the results of the survey, we apply a principal component analysis and we use the open-source R language. We consider 454 students and 22 items concerning motivations to apply to study abroad and concerns before the departure (4 increasing level for each item). In this very preliminary analysis, we identify three first principal components (pc):
Concerns for the academic career
The concerns of the student about his/her academic career are twofold. On the one hand, the concerns of a new teaching methodology and of having lectures and exams in a foreign language. On the other hand, the concern of extending the duration of academic career and eventually of worsening the average score.
More educated parents and previous experience abroad make students less concerned and males seem to be less worried than women about the mobility experience.
The second component seems to show the interest for a foreign university which better valorizes the experience, in particular in terms of employability. Such interest becomes more and more important for the master degree students.
Non academic motivations, Curiosity
The third component is related to the personal motivation, i.e. the curiosity, the will to live a new life experience, to meet a foreign country culture. Such type of interest is more relevant for young students and for who come from less educated parents and lower income families.
In conclusion, our preliminary results show that studying abroad is motivated both by the aim to develop academic skills and foreign language’s proficiency and by curiosity toward different cultures and countries and the desire to live a new experience. At the same time, students are concerned for their academic career. They are worried to have a negative impact of the mobility abroad on the duration and the performance of their study program in the home country, this is particularly true for females. Based on our results, the social and economic context and the educational background of the students have a very important role in encouraging toward the experience abroad.
Gender and racial disparities in STEM graduation and major choice are stark and studies have frequently found that women and racial minorities (hereafter called STEM minorities) leave STEM fields at higher rates than their counterparts (Anderson & Kim, 2006; Hill et al., 2010; Griffith, 2010; Huang, Taddese, & Walter, 2000; Kokkelenberg & Sinha, 2010; Shaw & Barbuti, 2010). Since STEM degrees pay substantially and increasingly more than other fields (Altonji et al., 2012; Altonji et al., 2014), the high attrition rates among STEM minorities are believed to be a driving factor behind the gender and race wage gap (Gerber & Cheung, 2008; Brown & Corcoran, 1996; Thomas, 1985).
This paper aims to advance research by providing the first direct examination of the effects of changing the grading scale on patterns of college grades in lower-division courses and STEM graduation and major choice. In the early 2000s, two universities in Florida, the University of Central Florida (UCF) and University of South Florida (USF), changed their grading scale from whole-letter to plus/minus grades. Previous research on changing the grading scale has been mostly descriptive and produced in the context of curbing grade inflation.
Using a restricted-use administrative database on Florida public college enrollees, I applied a difference-in-difference approach, comparing similar students whose grading differentials between STEM and non-STEM courses changed over time at institutions with or without the change to estimate the effect of this institutional grading policy on STEM outcomes. I contribute to the literature by providing the first quasi-experimental estimates of the effect of a change in the grading scale rather than on other academic and more policy-relevant outcomes such as major choices, persistence, and graduation rates. I also improve on this prior work by identifying students who take STEM and non-STEM courses and analyzing if the effect of this grading policy is differentiated across STEM and non-STEM departments (low-and high-grading departments) and lower-division courses.
I specifically test three related hypotheses. First, all else equal, a more continuous and refined grading scale will lead to reduce grade differentials between STEM and non-STEM courses. I find that a change in the grading scale from whole-grade grading scale to plus/minus grading has a substantial effect on first year grade differentials between STEM and non-STEM lower-division courses. Although not obvious a priori, this policy had the effect of substantially increasing grades in STEM fields. Students who attended institutions that changed their grading scale experienced a smaller difference between their STEM and non-STEM GPAs during the first year of enrollment than similar students attending institutions that did not implement any chance in their grading scale. I use these results to introduce my second hypothesis and to frame a discussion of the mechanisms potentially responsible for reducing STEM attrition.
Second, a reduction in grading disparities will make students more attracted to STEM, which in turn will improve STEM graduation and major choice. I find that after the grading policy change, students attending treated institutions are significantly more likely to graduate in STEM in 6 years or less, or to choose a STEM major at the beginning of their second and third year of enrollment. These positive effects on STEM outcomes seem to be equally explained by a reduction in the probability of majoring in non-STEM and dropping out of college without earning a BA. One possible reason to explain these results is that students might be vulnerable to grades differentials due to differences in the cost of study and grading standards between STEM and non-STEM fields (Arcidiacono, 2011; Barnes, Bull, Campbell, & Perry, 2001; K. Rask, 2010; Johnson, 2003).
Finally, a reduction in grading disparities will lead to an improvement in the gender and race gap in STEM attrition, based on the empirical evidence that suggests that STEM minorities value grades more than their counterparts (Rask & Bailey, 2002; Rask & Tiefenthaler,2008). I find significant impacts by gender and race. This policy has larger effects for men than women and for racial minorities than non- minorities. However, for women the effects on STEM outcomes seem to be explained by a reduction in the probability of leaving STEM by switching into non-STEM, whereas for men these effects are paired with a larger decrease in the probability of dropping out of college. In fact, men were relatively more affected at the bottom of their grade distribution which might explain why overall impact estimates are higher for men.
Recent empirical research has shown that social interactions at school can affect academic achievement and ultimately labor market outcomes. Two measures of school peer characteristics that have attracted attention are the share of girls and average parental background in the class/grade/school attended by an individual.
A potentially relevant question that has attracted little attention so far is whether the benefits and costs of peer interactions at school are affected by earlier interactions occurring within the family and involving both parents and siblings. For example, it is important to know whether the benefits from interacting with schoolmates having well educated parents accrue mainly to those coming from a similar “privileged” background or to the “disadvantaged”, as this may affect policies targeted at desegregating schools. Since peer interactions among differently gendered individuals start in the family, shaping the goals and expectations of girls and boys, we would also like to know whether having interacted with at least one sister in the family affects the benefits and costs of having many female schoolmates at school.
In this paper we provide evidence on the effects of school-family interactions on educational attainment and longer term labor market outcomes. We use Danish register data to investigate whether and how the effects of peer characteristics at school vary with the type of family interactions involving parents and siblings.
There are several economic and social mechanisms explaining why peer characteristics affect individual performance. A higher share of girls in the class or school can improve the learning environment by reducing disruption. Individual behavior may also change. For instance, pupils with more female schoolmates – or with more sisters at home - may change their attitude toward risk and competitiveness. On the other hand, peers with a better parental background may act as positive role models in education and facilitate access to economic and social networks.
Understanding the effects of school peer characteristics is important for the design of school admission and class formation policies. On the one hand, these effects help informing the ongoing debate about single-sex schools. On the other hand, they shed light on the benefits of de-segregating schools and – assuming that parental education is a relevant measure of family background – they also inform about the intergenerational social returns to education, about which relatively little is known to date.
When we do not distinguish males and females according to their interactions in the family, we find that the share of female schoolmates affects only the educational attainment of females and has no statistically significant effect on labor market outcomes. We also find that average parental background in the school increases both educational attainment and lifetime earnings for males but not for females. The statistically significant effects that we find are always very small.
However, our data indicate that the effects of school peer characteristics vary with individual parental education. In particular, we find that, for males, a higher share of “privileged” schoolmates increases years of education and lifetime earnings for the “disadvantaged”, but have no statistically significant effect on the “privileged”. For females, we find instead that having a sister in the family significantly affects the impact of the share of female schoolmates on years of completed education and employment at age 31 to 40.
In particular, we find that females (males) with at least one sister are more (less) likely to be employed after experiencing a higher share of female schoolmates. We believe that, for females, the combination of many female schoolmates and the presence of sisters in the family may attenuate feminine stereotypes and positively affect competitiveness and risk taking, with positive effects on long term employment prospects. On the other hand, males with sisters may be not only less disruptive but also acquire more feminine traits, for instance in terms of lower competitiveness and higher risk aversion. Interacting with a higher share of girls at school could foster these traits, with negative labor market consequences.
With some extrapolation, our results point out that the benefits and costs of single schools are likely to be small and to vary not only with individual gender but also with the gender composition of siblings. They also indicate that attracting students from “privileged” background can produce small benefits for (male) “disadvantaged” students in terms of education and lifetime earnings, without having statistically significant consequences on the “privileged”. Thus, they provide an argument in favor of mixing students from different backgrounds.
The role of doctoral recipients in promoting research and driving innovation is stressed in the literature (Stuen et al. 2012). In particular, the European Union policy agenda emphasizes the role of doctoral recipients since PhD education and training can help achieve the Lisbon goal of developing a competitive-knowledge based economy. However, in order to achieve these goals it is imperative that the doctoral recipients accept jobs that exploit their skills and training. A doctoral recipient would typically search for a job matching his/her education in their geographical area. When they cannot achieve this they could either accept a job requiring less education or skills; remain unemployed or outside the labor force, or widen the job search area (Büchel & van Ham 2003). If a doctoral recipient accepts a job that requires less education or skills than what they have acquired then they are mismatched for the job thus resulting in overeducation or overskilling. This labor market mismatch could be potentially costly both for the PhD recipient and society due to lower levels of productivity, lower job satisfaction and the inefficient use of investment made in education. (McGuinness 2006). Consequently, this has led to a growing literature documenting the incidence of labor market mismatch in the early career of young workers (Dolton & Vignoles 2000).
Using data from the Professional Integration Survey of PhDs administered by the Italian National Institute of Statistics (ISTAT) on Italian PhD graduates, this study contributes to the existing literature on job mismatch in several ways. First, we quantify the extent to which spatial mobility represents a strategy that reduces the risk of being overeducated or overskilled. Second, we use a new measure of overskilling where we first rely on standard measures based on workers' self-assessment and then exploit information on the R&D content of the job in order to capture the mismatch between the academic research training received before graduation and the actual job duties. Third, this study is important from the host country perspective since this group of tertiary education embodies high-skill labor and knowledge that could lead to higher innovative activities and knowledge creation in the host country (Bosetti et al. 2015; Chellaraj et al. 2008).
A clear pattern emerges from our analysis. Spatial mobility is found to reduce the probability of mismatch. Our estimates indicate that if a random individual from the population of PhDs chooses to migrate, he/she would face a 15.8 percent lower probability of getting a job for which the degree was formally required or at least useful at the hiring stage, a 20% lower probability to be overskilled and a 11.7% lower probability of being in a non-research-oriented job. These pieces of evidence broadly confirm what already found in the literature focusing on the role of spatial mobility in reducing the education-job mismatch of high-skill individuals and imply that PhD recipients enlarging their job-search area are able to transfer abroad, at least partially, their human capital. Since PhD holders possess a set of occupation-specific skills whose returns occur only in a limited set of activities, and since the geographic distribution of high-skill occupations requires individuals to move to the areas in which the job opportunities are located, individual mobility turns out to pay off in terms of lower chances to be mismatched in the labor market. Moreover, the estimated effects suggest that there are sizeable differences in the effects of migration on overskilling when we move from a self-assessment measure (20%) to our objective measure (11.7%). First, this result challenges the appropriateness of the use of measures for overskilling based on a subjective evaluation. Second, the result highlights a tendency of the Italian economy of being unable to employ all the potential scientific workforce trained in domestic institutions.