Navigazione – Mappa del sito

HomeNumeri84- LXIVla società contemporanea / Transi...Is vocational education a safety ...

la società contemporanea / Transizioni scuola-lavoro: uno sguardo sociologico

Is vocational education a safety net?

The occupational attainment of upper secondary graduates from vocational and academic tracks in Italy
Is vocational education a safety net? The occupational attainment of upper secondary graduates from vocational and academic tracks in Italy
Carlo Barone e Moris Triventi
p. 59-89

Abstract

This article assesses the employment and occupational outcomes of upper secondary education graduates from academic and vocational tracks in Italy. In particular, we formulate and test the hypothesis that – contrary to some common expectations – academic graduates outperform vocational graduates at a stage of occupational maturity, even when considering individuals without a tertiary degree. Moreover, we explore differences between tracks by gender as well as across geographical areas and city sizes. Thanks to the detailed information available in the PLUS data, we assess labor market outcomes adjusting for a rich set of socio-demographic characteristics and for early academic performance. The results corroborate our hypothesis and indicate that the advantage of academic graduates holds across genders, areas and different city sizes.

Torna su

Note della redazione

The authors have equally contributed to this work and they are listed in alphabetic order. Corresponding author: Moris Triventi, Via G. Verdi 26, 38122 Trento; Email : moris.triventi@unitn.it.

Testo integrale

1. Introduction

1Stratification research has paid increasing attention to horizontal inequalities in education (Breen, Jonsson, 2000; Lucas, 2001). Several studies on higher education have investigated differences between fields of study, university institutions as well as academic and vocational programs, in terms of their social intake and their labor market rewards (Shavit et al., 2007; Triventi, 2013). The underlying hypothesis is that upper class families are overrepresented in the most rewarding educational paths, owing to their higher cultural, economic and social resources. However, research on horizontal inequalities in upper secondary education is much less developed. More precisely, it is well-known that upper class families are massively overrepresented in academic secondary tracks (Schizzerotto, Barone, 2006; Contini, Scagni, 2013; Panichella, Triventi, 2014), but the labor market prospects of academic and vocational diplomas have received much less attention. Still, in most OECD countries the majority of students leave the educational system with a secondary degree, even in recent cohorts (Oecd, 2019).

2This research gap has two explanations. First, it is typically assumed that academic tracks train students to succeed in higher education, rather than fostering their immediate employability. Academic diplomas are indeed associated with substantially higher rates of enrolment in and completion of tertiary education (Barone, Assirelli, Triventi, 2018). Hence, their labor market premia are largely driven by the profitability of tertiary degrees. In contrast, vocational tracks prepare for rapid labor market insertion by equipping students with practical skills that are directly applicable without pursuing further education. Hence, it is commonly assumed that vocational diplomas are more rewarding for students who do not continue to (or fail to complete) tertiary education. Indeed, the ‘safety net’ hypothesis postulates that vocational tracks are particularly attractive for working class families because these tracks do not suppose a long-term investment in higher education (Arum, Shavit, 1995; Shavit, Muller, 1998). At the same time, this supposed strength can be a trap, to the extent that working-class students with the potential to succeed in tertiary education are diverted from the academic path (Becker, Hecken, 2009). These students could perceive the vocational track as a safer option that minimizes the risks of educational failure and fosters occupational opportunities if they do not continue to higher education (Shavit, Müller, 2000; Becker, Hecken, 2009).

3However, a growing number of empirical studies challenges this narrative. In particular, the assumption that, among students without tertiary qualifications, vocational diplomas outperform academic diplomas in the labor market has been increasingly questioned (Hampf, Woessmann, 2017). To the contrary, the latter are found as rewarding as the former, if not more rewarding, particularly when considering mid- and late-career outcomes (van Loo, 2001). In particular, while vocational diplomas may facilitate transition to the first job and, in some countries, higher initial earnings, academic diplomas enjoy access to more skilled jobs, steeper earnings profiles, better career prospects and lower risks of job loss (Hanusek et al., 2017; Forster et al., 2016; Rozer, Bol, 2019).

4Two sets of theoretical arguments can explain why academic diplomas can be more rewarding at least in the long-run (Korber, Oesch, 2019). On one hand, from a human capital perspective, academic tracks foster reasoning, language and communication skills to a greater extent than vocational tracks. In contemporary service economies (Kamens et al., 1996; Van Houtte et al., 2012), these skills are highly demanded in white-collar occupations and other middle class jobs to which upper secondary graduates may aspire (Cedefop, 2013). Students of academic tracks are also more exposed to classical, highbrow culture and upper class conventions, thus developing cultural capital resources that could be rewarded by employers. Even when they do not complete tertiary education, they attend it more often than vocational graduates, thus enjoying additional opportunities to boost their skills and their social networks. Vocational tracks are more tailored to the transmission of manual skills and other content-specific skills that are more at risk of becoming rapidly obsolete. On the other hand, from a signaling perspective, academic tracks are often perceived as more demanding and selective than vocational tracks (Van Houtte et al., 2012). Hence, obtaining an academic diploma may be taken as a signal of higher ability and motivation. Conversely, vocational tracks are more and more negatively labeled as a ‘reservoir’ for low-performing students, typically with a working-class and immigrant background, or with behavioral problems. This negative ‘stigma’ may be detrimental to the labor market insertion of their graduates.

5The hypothesis that academic diplomas outperform vocational diplomas among mid-career workers has immediate relevance from a social stratification perspective: the overrepresentation of upper class students in academic tracks would foster the intergenerational reproduction of social inequalities not only via the attainment of tertiary degrees, but also even when diploma holders from the academic track fail to complete tertiary education, and thus access the labor market with a high school diploma. In other words, if this hypothesis is confirmed, a novel, under-researched channel of intergenerational reproduction is identified. This channel would be particularly important in Italy, where the share of tertiary graduates is comparatively low, so that a majority of youngsters leaves the educational system with an upper secondary certificate.

6Interestingly, the empirical evidence that academic diplomas outperform vocational diplomas concerns also countries such as Germany, Switzerland and the Netherlands characterized by a strong supply of vocational education (e.g. Korber, Oesch, 2019; Rozer, Bol, 2019; Lavrijsen, Nicaise, 2017). There are therefore even stronger reasons to expect that this is the case in countries like Italy, France or Spain, where vocational training is underdeveloped and vocational tracks in upper secondary education are weakly tied to employer organizations, make limited use of apprenticeships and are more disconnected from labor market demands (Isfol, 2016).

7However, the data needed to assess the aforementioned hypothesis are not easily available. This is the second reason why the issue of horizontal inequalities in secondary education is under-researched. A first problem is that students are not randomly assigned to secondary tracks. For instance, it is well-documented (Shavit, Müller, 1998; Correll, 2001) that gender, socio-economic and ethnic background, as well as academic performance and occupational aspirations affect both track choice and occupational attainment (regardless of track choice). Hence, if these confounders are not controlled for when assessing returns to educational qualifications, the resulting estimates are likely to be biased. Unfortunately, several data sources do not contain sufficient information to control for selection into tracks. In particular, in the Italian case, the Labor Force Surveys (Indagine sulle Forze Lavoro) as well as the main social mobility surveys carried out in Italy (Indagine sulla Mobilità Sociale in Italia, Indagine Longitudinale sulle Famiglie Italiane, Indagini Multiscopo Famiglia e Soggetti Sociali), which are routinely used to assess inequalities in occupational attainment, do not collect information on academic performance before track choice. Surveys on upper secondary graduates, such as Indagini Istat sui Diplomati or Almadiploma in Italy, collect this information, but they typically survey graduates only a few years after graduation. Such a short observation window is problematic, because the above-cited studies suggest that long-term returns to secondary qualifications can differ from those observed in the short run.

8In this work, we assess the hypothesis that academic diplomas outperform vocational diplomas in Italy among students with and without a tertiary degree. In section 2, we discuss the most relevant characteristics of the Italian context, stressing the gender and territorial segmentation of its labor market. In section 3 we illustrate our data and methods, in section 4 we present the results, and we discuss them in the concluding remarks of section 5.

2. The Italian case

  • 1 After lower secondary school, students can also opt for three-year vocational training courses (ist (...)
  • 2 Due to limited sample size, in the empirical part we cannot analyze the curricular differentiation (...)

9The analysis of labor market returns to secondary qualifications must be developed taking into account the characteristics of the Italian educational system and labor market. After having attended primary and lower secondary education, which are comprehensive in Italy, students make a choice, usually at the age of 14, between academic-oriented schools (licei), technical schools (istituti tecnici) and vocational schools (istituti professionali). They all take five years to complete and, upon successful completion, they afford a diploma that entitles students to access to higher education, which virtually coincides with university courses that comprise three-year bachelor and two-year master programmes since year 20011. Vocational schools are supposed to train students for specific occupations, but they are not organized around an apprenticeship system; practical skills are transmitted mainly via simulations and lab activities. Vocational schools display a marked overrepresentation of immigrant children and of low-performing students from low-educated families (Azzolini, Barone, 2012; Contini, Scagni, 2013). While they are not perceived as academically demanding, they display high drop-out rates. At the opposite extreme, licei are a fully academic track focusing on the transmission of general skills. Nine graduates out of then in these schools continue to university education (Barone, Triventi, Assirelli, 2018)2. Italian upper secondary education comprises three main pillars, rather than two: technical schools may be regarded as an intermediate option between fully academic and fully vocational tracks. While offering training for specific occupations like vocational schools, they have a stronger focus on the transmission of general skills (Gambetta, 1987). Their graduates display significantly higher probabilities of enrolling in university programs and of completing them. In terms of labor market outcomes, we would thus expect that their graduates enjoy more favorable prospects as regards employment status and that they are less disadvantaged than vocational graduates in terms of class attainment.

10As regards the Italian labor market, its most important characteristic is the marked territorial segmentation. The Northern regions are richer and more economically developed than the Southern regions, which display much lower employment rates and higher unemployment rates (Daniele, Malanima, 2014; Odoardi, Muratore, 2019). However, the class structure of Southern Italy does not display a lower share of upper or middle class jobs, owing to the hypertrophic development of liberal professions, small shops and white-collar jobs in the public sector. Instead, the demand for vocational skills is higher in the industrial regions of Northern Italy (Ballarino, Panichella, Triventi, 2014).

11Importantly, the economic structure of Northern regions displays significant sectoral specialization. The north-western regions are characterized by a stronger density of large firms and of high added-value service activities, such as finance and insurance services, while the regions of the Centre and of the North-East, the so-called ‘Terza Italia’, display a rich fabric of small firms and of economic districts (distretti economici) (Bagnasco, 1977). These regions could therefore be characterized by a larger demand for vocational skills relative to both North-western and Southern regions, which could boost the labor market prospects of graduates from vocational schools in the Terza Italia. The share of self-employed workers (23%) is comparatively high in Italy and these workers tend to be concentrated in the Northern regions (Istat, 2020).

12These territorial cleavages intersect the divide between large cities and small towns and villages. An emergent literature suggests a trend of increased polarization whereby knowledge, research and innovation tend to concentrate in cities, which disproportionately contribute to economic growth and productivity, while smaller agglomerations dominated by craft occupations and more traditional sectors are more exposed to the risks of the transition from an industrial to a service economy (Frank et al., 2019). Overall, the demand for the general, transferrable skills emphasized in academic curricula is higher in large towns and cities. These are more concentrated in the northwestern regions, where we find three of the five largest Italian cities in Italy (Milan, Turin and Genoa).

13The interplay between territorial divides and gender inequalities in the labor market is a key theme in the literature on the Italian case (Reyneri, 2017; Istat, 2020). The female activity rate is comparatively low, particularly among middle-aged women, while female unemployment rates are higher than those of males in virtually all regions, but these gender gaps are particularly large in Southern regions (Reyneri, 2017). In a context where employment opportunities are limited and women are strongly penalized, a low level of education entails particularly strong disadvantages for them in these regions as regards employment opportunities. At the same time, when considering the lower demand for vocational skills in the industrial sector in Southern Italy, men with vocational qualifications may be particularly penalized in terms of occupational attainment. Hence, women with academic secondary qualifications should enjoy the most favorable prospects.

14Based on the above arguments, we formulate the following hypotheses:

15H1: Graduates from the academic track outperform graduates from vocational schools in Italy with respect to employment and occupational outcomes assessed in the adult population;

16H2: Graduates from technical schools outperform graduates from vocational schools in Italy with respect to employment and occupational outcomes assessed in the adult population;

17H3: Differences between secondary tracks with respect to employment and occupational outcomes in the adult population persist even when restricting the analysis to graduates without tertiary qualifications;

18H4: Differences between secondary tracks with respect to employment and occupational outcomes in the adult population are weaker in the North-East and Centre of Italy;

19H5: Differences between secondary tracks with respect to employment and occupational outcomes in the adult population are stronger in large towns and cities;

20H6: The advantage of graduates from the academic track over graduates from vocational tracks and lower secondary graduates with respect to employment and occupational outcomes is stronger among women in Southern Italy.

3. Data, variables and models

  • 3 The ISFOL PLUS survey involves additional waves, which cannot be used because they lack some of the (...)

21For the analyses, we use the 2014 wave of the PLUS (Participation Labor Unemployment Survey) data, collected by the ISFOL research institute.3 This is a nationally representative survey containing information on family background, educational careers and occupational outcomes. This survey contains information on academic performance in lower secondary education, that is, before track choice. Moreover, it has reliable and detailed information on the employment and occupational status of the adult population. Finally, the large sample size allows us to analyze returns to educational qualifications across different geographic areas and city sizes. More information on the research design of this survey can be retrieved in Mandrone (2013).

22The analytical sample includes only Italian citizens, due to the small numbers of foreign interviewees. We select respondents aged 30 to 45 years old in order to analyze a relatively homogeneous sample in terms of school and labor market experiences. The exclusion of younger individuals reduces issues of selection into employment and the noise of early-career fluctuations, while the exclusion of older individuals attenuates issues of recall bias and of selective mortality.

23We consider four dependent variables measured at the time of the interview: i) being employed versus being either unemployed or inactive; ii) being unemployed versus being employed or inactive; iii) accessing the salariat class, which comprises large entrepreneurs, managers and professionals; iv) accessing unskilled manual jobs (routine manual occupations). The two occupational outcomes are operationalized according to the ESeC class schema (Harrison, Rose, 2010); for respondents not employed at the time of the interview, the last occupation is used if available. Following the ISFOL definitions, individuals with occasional jobs are not treated as “employed”. Unfortunately, due to data limitations, we cannot analyze earnings nor income.

24Our main independent variable of interest is the track of high school diploma, which differentiates between four categories: 1) lower secondary (those who did not attain any upper secondary diploma), 2) academic (scientific and classical lyceums, foreign languages and arts lyceums, teaching-training schools), 3) technical and 4) vocational track. This last category includes both the five-year diploma and the three-year qualifications (see footnote 2). We take individuals who attained an academic high school diploma as reference category. Unfortunately, owing to data limitations, we cannot further disaggregate this variable and, in particular, we cannot assess differences in labor market outcomes between different curricula offered within the same secondary track.

  • 4 City size refers to the place where respondents live, rather than to the place where they work. Thi (...)

25Following the arguments advanced in section 2, we consider two variables that could moderate returns to secondary qualifications: area of residence (North-East, North-West, Centre, South and islands) and city size (<20’000, 20’000-50’000, 50’000-250’000, >250’000 inhabitants4). Since we regard geographical area as a causal antecedent of city size, the statistical models assessing the heterogeneity by city size control for geographical area, while models on the heterogeneity by area do not control for city size.

  • 5 Unfortunately, the ISFOL plus data for parental occupation do not differentiate between self-employ (...)

26Control variables include gender, age and age squared, having grown up in a two-parent family (i.e., whether at the end of middle school, both parents were alive), the grade obtained at lower secondary school examinations (excellent, very good, good, pass), and the family background, which refers to parental education and social class. These two variables are built according to the dominance criterion, which selects the highest position among those of the two parents. Parental education has three categories: tertiary, upper secondary, lower secondary or less. Parental class is measured according to the ESeC schema in six categories (service class, intermediate occupations, small employers/self-employed, agricultural employment, skilled working class, unskilled working class5).

27Since all four outcome variables are binary, we specify logistic regression models and present average marginal effects and predicted probabilities. We thus focus on absolute returns to secondary qualifications, rather than on relative returns, which would be captured by logit parameters. Inter-group comparisons (e.g. between geographical areas) in the effects of educational qualifications based on logit parameters could be problematic, due to the scaling problem associated with unobserved heterogeneity (Allison, 1999).

4. Empirical results

4.1. Descriptive statistics

28Tables A1-A3 in the Online appendix report the main descriptive statistics. In the analytical sample, 35% of respondents is from the South, 26% from the North-West, 20% from the Centre, and 18% from the North-East. Among respondents aged 30 to 45, 33% did not complete high school, 26% completed an academic track, 30% a technical school, while around 11% obtained a vocational qualification. The incidence of individuals with no more than lower secondary education is largest in the Southern regions (37%), and lowest in the North-Eastern regions (17%).

29Looking at the distribution of the outcomes, we see that around 71% of the sample was employed at the time of the interview, while 17.5% was unemployed. As it is well-known, there are huge territorial differences in this respect: employment rates exceed 80% in the Northern regions, while they are lower than 55% in the Southern regions.

30Unemployment rates are around 10% in the Northern regions, whereas they are three times higher in the South. Slightly more than one-third of the sample attained a service class job (35.5%), whereas 7% ended up in an unskilled working class occupation. These figures reflect the fact that we focus on adults of relatively recent cohorts, born between the 1970s and the early 1980s, among whom around 29% attained a university degree. Variations across geographical areas are less pronounced when considering occupational attainment among employed individuals. For instance, the incidence of the service class is even larger in the Southern regions (around 38-40%) than in the North (34%).The share of the unskilled working class ranges from 5% (North-East) to 10% (Centre).

4.2. The lab market prospects of high school diplomas for individuals with and without a tertiary degree

31Our first research question refers to the long-term occupational outcomes associated with taking different tracks in secondary education. To this aim, we estimated a binomial logistic regression model where the main independent variable is having obtained a diploma in different upper secondary tracks or not having attained any upper secondary diploma, and the control variables are geographical area, socio-demographic characteristics, social background and academic performance before track choice. While this article focuses on differences between upper secondary tracks, it is important to incorporate also comparisons with individuals without an upper secondary degree, for two reasons. First, the literature on ‘diversion effects’ cited in the introduction stresses that vocational tracks can be seen by families as a safe option to minimize the risks of leaving education without any upper secondary degree, and it is therefore important to assess the labour market consequences of this risk. Second, the outcomes of lower secondary graduates provide a benchmark to appreciate whether differences between upper secondary tracks are large or small.

32We are first interested in the total effect of high school track and, for this reason, we omit to control for intermediate variables or mediators in our main specification (Morgan, Winship, 2015). For instance, we do not adjust for university degree attainment or the specific field of study attained in higher education, which are plausible mediators of this total effect. In a second specification, instead, we restrict the estimation only to individuals who did not earn any tertiary degree. We can thus assess whether and to what extent academic diplomas still convey lab market advantages compared to other qualifications among adults who did not attain any tertiary degree.

33Figure 1 reports the average partial effects from these two model specifications (the omitted reference category is represented by individuals who attained an academic diploma). Further estimates from the main model specification are reported in Table A4 in Appendix, where additional contrasts among tracks are presented. Beginning with the main specification (black dots), we see that achieving no more than a lower secondary degree is associated with remarkably low employment probabilities (20 percentage points, p.p.) and high risks of unemployment (14 p.p.), compared to obtaining an academic diploma.

Fig. 1

Fig. 1

Binomial logistic regression to analyse occupational outcomes: average partial effects of high school track of diploma and 95% confidence intervals

34Among comparable individuals in terms of socio-demographics, family background and prior academic performance, having attained a vocational or a technical diploma does not lead to any major disadvantage in terms of employment outcomes, compared to graduates of the academic track. Still, graduates with vocational qualifications face a moderately lower probability of being employed (4 p. p. , p= 0.003) than those with an academic diploma.

35From the second model specification (grey squares), we can see that, even among those without any tertiary degree, academic diploma holders enjoy an advantage of 16 p.p. (p<0.000) in terms of the probability of being employed and 12 p.p. (p<0.000) lower risks of being unemployed, compared to individuals who did not earn any high school diploma. The pattern of modest differences between upper secondary tracks is maintained also when focussing on the subsample of individuals without a university degree. Hence, the different types of upper secondary diplomas enjoy a similar, large advantage over lower secondary degrees in terms of employment outcomes and this gap is only modestly driven by their higher university degree attainment.

36Academic diploma holders enjoy much more visible advantages in terms of occupational attainment compared to all the other categories. From model 1 in figure 1 (bottom-left panel), we see that their advantage in accessing the service class is highest compared to those with a lower secondary certificate (40 p. p. , p<0.000), and it is also large over graduates with a vocational or technical qualification (21-22 p. p. , p<0.000). A substantial share of these advantages is due to university degree attainment, since the size of the coefficients shrinks in the second model specification reported in figure 1. Nonetheless, even among individuals without a university degree and with comparable social origin and prior school performance, attaining an academic diploma is associated with substantially higher chances of entering the service class, compared to technical and vocational diplomas. Hence, hypotheses 1 and 3 about the pay-offs of academic diplomas are confirmed at least with respect to occupational outcomes, while hypothesis 2 about differences between vocational and technical diplomas is only partially corroborated.

37The bottom-right panel in figure 1 indicates that achieving a vocational or technical upper secondary degree protects from the risk of entering the unskilled working class, relative to comparable peers with only a lower secondary degree. The differences are non-negligible, since they range between 8 p.p. (vocational certificate) and 12 p.p. (technical certificate). In any case, an academic diploma still minimizes the risks of ending up in unskilled jobs compared to the other high school diplomas, although differences are less marked for this outcome, possibly due to floor effects (the baseline probability is low). Results from the second model specification are very similar, thus suggesting that the above pattern holds also for individuals without any university degree. Hence, academic diplomas enjoy the most favourable lab market prospects with and without a tertiary degree.

4.3. The lab market prospects of high school diplomas across geographical areas and genders

38We now explore the heterogeneity of labour market outcomes by area of residence. Of course, given the structural differences across Italian regions in terms of labb market opportunities, the average outcomes associated with upper secondary diplomas differ across areas, as we have already seen in the descriptive statistics, but we are interested here in the differences between areas in the lab market prospects of different high school diplomas. To answer this question, we developed a binomial logistic regression model that expands the previous model specification by incorporating a two-way interaction between high school track and geographical area. It should be noted that, while we can control for important confounders related to selection into track, we cannot take into account the potential endogeneity of the current workplace, namely, people can move to geographical areas where their educational degree is better valued in the labor market. In Italy, selective migration from Southern to Northern regions is well documented (Panichella, 2014).

39In figure 2 (and table A4 in the Online Appendix) we report the average predicted probabilities of being employed (upper-left graph), unemployed (upper-right graph), entering the service class (bottom-left graph) and ending up in the unskilled working class (bottom-right graph) across geographical areas (x-axis) and high school diplomas. We report in figure 2 the results for the whole sample, whereas those for the subsample of individuals without tertiary qualifications can be found in Figure 3. The corresponding average partial effects are presented in Figures B1 and B2 in the Online Appendix. Given that the results of these two analyses are very similar, we will only comment on the second when it differs from the result concerning the main sample.

Fig. 2

Fig. 2

Binomial logistic regression to analyse occupational outcomes: predicted probabilities of high school tracks across geographical areas

  • 6 In the North-East the absolute penalization amounts to 7 p. p. , while in the South to 25 p. p. The (...)

40Beginning with the analysis on the whole sample, our main finding is confirmed: the chances of being employed are similar among individuals with different high school diplomas, and this pattern holds across geographical areas. A notable peculiarity is found in the South, where the employment penalty of having attained a vocational qualification instead of an academic diploma (around 7 p.p.) is larger than in other areas, although the difference with the North-East is statistically significant only at the 90% confidence level. Having no more than a lower secondary certificate is also associated with lower employment chances compared to having an academic degree across all geographical areas, but again this penalty is stronger in Southern regions (28 p.p.) than in the North-West (19 p.p.) and Central-Eastern regions (12-13 p.p.). The results obtained when looking at the risk of unemployment are very similar and specular to what we have already commented. The most striking result across Italian regions is the penalization of those who attained no more than lower secondary education, against all other diploma holders. For instance, in the North-East the predicted unemployment risks of lower secondary certificate holders are around 14% and those of academic degree holders are around 7%, whereas in the Southern regions the same figures amount, respectively, to 49% and 25%. Overall, while there are remarkable differences across geographical areas in the absolute penalty of lower secondary certificates, the relative disadvantage is quite similar.6 An exception emerges, however, when we focus on individuals without a university degree (figure 3). In this group, the unemployment gap associated with lower secondary certificates (compared to the academic track) is significantly stronger in the South than in the other geographical areas.

41If we move to social class attainment (bottom panels in figure 2), the geographical patterns become more alike, both in terms of the overall distribution of occupational outcomes and of returns to educational qualifications. The probability of entering the service class for individuals with academic qualifications is high across all macro-areas, ranging between 53% (North-East) and 57% (South). Conversely, the chances of entering the service class for those with no more than lower secondary education are much lower, ranging between 13% (North-West) and 18% (South). Overall, we do not detect any major difference between areas with respect to returns to different types of secondary diplomas, measured by the absolute chances of access to upper class jobs. In particular, the favorable prospects of the graduates of the academic track hold similarly across areas.

42Furthermore, also the risk of ending up in the unskilled working class is similarly stratified according to the type of high school diploma across macro-areas. Individuals with lower secondary degrees display the strongest disadvantages, followed by those with vocational and technical education, while academic graduates enjoy once more the most favorable prospects, even when comparing individuals with similar social background and early school performance. The gap between lower and upper secondary degrees displays some fluctuations across areas, but the relatively large uncertainty around the point estimates prevents us from establishing any clear pattern. Hence, hypothesis 4 about differences between geographical areas is not supported. Overall, the advantage of academic diplomas over other high school diplomas holds with and without tertiary qualifications regardless of the area of residence.

Fig. 3

Fig. 3

Binomial logistic regression to analyse occupational outcomes: predictive margins of high school track of diploma and 95% confidence intervals for the subsample of individuals without a university degree

43Since we know that lab market outcomes vary greatly between men and women, we have further articulated our analysis by including in our baseline model a three-way interaction between high school diploma, area of residence and gender. In figure 4, we report the predicted probabilities derived from this model, in order to detect any gender-specific pattern of geographic heterogeneity in the labour market returns to upper secondary diplomas. In order to have enough statistical power for this analysis and to provide more interpretable results, we recoded area of residence into a dummy, contrasting northern regions (North-West and North-East) with central and southern regions. As for the previous models, we have replicated this analysis including only individuals without any university degree. The results reported in the Appendix (Figure C1) are very similar to the ones for the whole sample. Therefore, we will comment them in the text only when they significantly differ from results for the whole sample.

44First, we find the well-known patterns of higher employment rates among men than women. These employment probabilities are stratified according to the type of high school diploma in a similar way for men and women. With this overall finding in mind, we note that incorporating gender into the picture allows us to detect some patterns that were not visible in the previous aggregate analysis. In figure 4 (and Table A6 in the Appendix), we report the predicted probabilities of each of the four outcomes for all the combinations of high school track, geographical area and gender. We see that having an academic diploma conveys an advantage in terms of employment chances especially to women living in the South (around 33 p. p. against those without any diploma). This advantage is less pronounced for women in the North (23 p.p.) and men in the South (21 p.p.), and it is much lower for men in the North (6 p.p.). Moreover, the unemployment risks associated with lower secondary education are particularly marked for women in the South (55%), but the gender difference is similar in the Northern and Southern regions. Differently, women’s penalty in unemployment risks are larger in the South (around 15 p.p.) than in the North (around 6 p.p.), but with limited variations across high school types.

Fig. 4

Fig. 4

Binomial logistic regression to analyse occupational outcomes: predicted probabilities of high school tracks across geographical areas (North vs South) and gender

45The chances of entering the service class are particularly low for individuals without any high school diploma, but women living in northern regions are more disadvantaged in this respect. Differently, women in the South with a vocational qualification have comparatively higher chances of entering the service class compared to women in the North as well as to men from both macro-regions. The returns to technical diplomas are clearly stratified by gender and less by geographical area: the chances to enter the salariat class with this qualification are higher for men than for women, both in the North and in the South. The chances to enter the service class with an academic diploma are instead higher for women than for men, with slightly larger distances between macro-areas among the former. Overall, women benefit more than men from academic diplomas, both in terms of employment chances, particularly if they live in Southern and Central Italy, and of access to the upper class.

46These empirical patterns are replicated when focusing on the subsample of individuals without tertiary qualifications. However, the probability of accessing the service class among men in the North and in the South is more similar than in the whole sample (see Figure C2 in the Appendix). Moreover, the risk of ending up in the unskilled working class for low-educated individuals is particularly high for women, irrespective of geographical area, while for upper secondary graduates with technical and academic diplomas gender and geographical differences are much less pronounced. Finally, let us comment on graduates with vocational qualifications: men in the South with vocational diplomas are clearly the most disadvantaged group, which leads to larger gender differences in Southern regions than in the North, where men and women face similar risks of being employed in unskilled occupations. Overall, our results indicate that men benefit most from technical qualifications, while women from academic diplomas, but there are some specific patterns related to the different outcomes as well as to the territorial segmentation of the Italian labor market.

4.4. Local lab markets: the moderating role of city size

47We have further investigated the issue of territorial differences in the lab market returns to high school diplomas by comparing more or less densely populated contexts. The rationale for this analysis is that metropolitan areas, medium-large cities, medium-small cities, and small towns display different labor market structures, which could offer heterogeneous sets of opportunities to individuals with different skills and credentials. We have addressed this issue by adding to the baseline model an interaction term between secondary qualifications and city size. In this model, we control for area of residence in order to see whether city size moderates the profitability of high school diplomas. Figure 5 presents the predicted probabilities of each of the four outcomes across demographic contexts for different types of high school diplomas. Figure 6 reports the same estimates but for the subsample of individuals without a university degree.

Fig. 5

Fig. 5

Binomial logistic regression to analyse occupational outcomes: predicted probabilities of high school tracks across city sizes

48Overall, city size plays a marginal role as moderator of returns to secondary qualifications, with one notable difference, namely, individuals who did not attain any upper secondary diploma are particularly disadvantaged: they display lower chances of being employed, higher unemployment risks and higher probabilities of entering the unskilled working class in metropolitan areas and big cities than in small towns. For instance, while the difference in the probability of entering the unskilled working class between those with lower secondary education and those with an academic diploma is around 11 p. p. in small towns, it is more than 33 p. p. in large cities (i.e. cities with more than 250,000 inhabitants). The difference across the two contexts is both substantial (22 p.p.) and statistically significant at the 95% confidence level (p=0.004). Hence, the occupational disadvantages of low-skilled workers are exacerbated in large cities. Once more, this result is confirmed also when comparing them to academic diploma holders without a university degree (see Figure 6). Moreover, also when restricting the analysis to this sub-sample, differences between upper secondary tracks in terms of employment and class attainment are invariant across contexts with different population size, thus contradicting hypothesis 5.

Fig. 6

Fig. 6

Binomial logistic regression to analyse occupational outcomes: predicted probabilities of high school tracks across city sizes for the subsample of individuals without a university degree

5. Concluding remarks

49The most striking conclusion that emerges from our analyses is that academic diplomas are overall more rewarding than any other upper and lower secondary qualification regardless of university completion for whatever sub-group defined by geographic area, gender or demographic size. The advantage of the academic track thus looks remarkably stable in Italy.

50More specifically, academic diplomas perform as well as other upper secondary diplomas in terms of employment outcomes and they outperform them with regard to class attainment, measured either by the chances of accessing the upper class or by the risks of entering unskilled occupations, with and without a university degree. Academic diplomas are particularly rewarding for women in terms of both employment outcomes, particularly if they live in the South, and of class attainment. This is possibly because academic diplomas foster access to white-collar jobs and other occupations demanding general, soft skills, where women are overrepresented. Differences by area of residence and by demographic size in the labor market prospects of different upper secondary qualifications are generally modest, but in Southern regions vocational diplomas face a particularly strong disadvantage in terms of employment outcomes. This can be related to the low demand for skilled manual workers in these regions, traditionally characterized by a weak development of the industrial sector. Moreover, individuals without a high school diploma face a large, systematic disadvantage relative to graduates from all upper secondary tracks, but this gap is particularly marked in large cities, where the relative demand of skilled workers is higher.

51The advantage of academic diplomas documented in this work has important theoretical and policy implications. Since upper class children are massively overrepresented in the academic track, our results imply that they enjoy better lab market prospects than working class children even if they fail to complete university education. Moreover, in this work we have not explored returns to upper secondary qualifications by family background, but in separate work we found that upper class children without tertiary degrees reap higher benefits from academic diplomas than working class children (Barone, Triventi, Facchini 2021). At the same time, our results suggest that the overrepresentation of working class children in the vocational track cannot be explained by the safety net hypothesis. Indeed, vocational and technical tracks do not protect from unemployment or access to unskilled jobs more than academic tracks, even when students leave education without a tertiary degree.

52From a policy perspective, these results call into question the role of vocational and technical tracks in the educational system. Of course, these tracks protect low-performing children from early school leaving and ensure much better prospects than lower secondary degrees. However, our results indicate that these applied-oriented tracks have no edge over the academic track in the lab market, and it is well-known that the academic track offers a much better training for university education. If the academic track is then a win-win option for those students with the skills to complete it, the devaluation of more vocationally-oriented tracks is a concrete risk.

53Let us reiterate that in this work we have analyzed only the labour market prospects of graduates aged 30 to 45, thus ignoring early career outcomes. The prospects of vocational and technical diplomas could be more favorable at labor market entry, as well as with respect to outcomes that we could not consider in this work, such as earnings. However, the surveys on the labor market insertion of upper secondary graduates point to modest differences between upper secondary diplomas in terms of unemployment rates and of earnings among graduates without a tertiary degree who have left the educational system (Almadiploma, 2020; Istat, 2016). It should be noted also that these surveys consider only the five-year graduates of the vocational track, thus excluding three-year graduates (formazione professionale), who tend to perform quite well in some northern regions (Isfol, 2015). Unfortunately, in this work we could not disaggregate long and short vocational diplomas due to data limitations, nor could we explore curricular differences within tracks. These limitations represent important avenues for future research.

  • 7 The direction of potential biases in unclear when considering graduates without a tertiary degree, (...)

54Another limitation of this study is that we could not fully control for selection into tracks. For instance, achievement and effort orientations may drive unobserved heterogeneity in labour market outcomes that is not randomly distributed across tracks7. Still, this work used a rich set of control variables, including the important control for academic performance before tracking that had been omitted in previous social mobility analyses of the Italian case. Further research moving closer to causality is an another important avenue for future research in this domain.

Appendix A. Main estimates in table format

Table A1 Descriptive Statistics (N = 10,190)

Percentage/Mean

Employed

70.7

Unemployed

17.5

Salariat

35.5

Unskilled working class

7.3

Track of diploma

LowerSec.

32.8

Vocational

10.9

Technical

30.1

Academic

26.2

Geographical area

North-West

26.2

North-East

18.5

Center

20.2

South

35.0

City size

>=20k

49.6

20k-50k

17.4

50k-250k

17.7

>250k

15.3

Sex

52.2

Parental education

Tertiary

8.1

Upper Secondary

25.0

Lower secondary or less

66.9

Parental Class

Service class

16.7

Intermediate occupation

22.4

Small employers and self-employed

21.3

Agricultural employment

7.1

Lower technical

13.4

Routine

19.2

Lower Secondary Final Mark

Excellent

21.9

Very good

25.3

Good

31.3

Pass

21.5

Age

42.1

Intact families at the end of middle school

86.1

Note: estimates are weighted by sample weights provided by ISFOL.

Table A2 Percentage distribution of key variables by geographical area

Geographical area

North-West

North-East

Center

South

Total

%

%

%

%

%

Employed

No

18.8

18.0

22.6

47.0

29.3

Yes

81.2

82.0

77.4

53.0

70.7

Total

100

100

100

100

100

Unemployed

No

89.5

90.7

86.2

68.7

82.5

Yes

10.5

9.3

13.8

31.3

17.5

Total

100

100

100

100

100

Salariat

No

65.7

66.5

65.0

61.0

64.5

Yes

34.3

33.5

35.0

39.0

35.5

Total

100

100

100

100

100

Unskilled working class

No

93.5

94.7

90.5

92.1

92.7

Yes

6.5

5.3

9.5

7.9

7.3

Total

100

100

100

100

100

Track of diploma

LowerSec.

32.2

30.6

31.8

35.1

32.8

Vocational

12.2

13.6

9.7

9.1

10.9

Technical

30.4

30.9

28.8

30.1

30.1

Academic

25.2

24.9

29.7

25.6

26.2

Total

100

100

100

100

100

N

2,093

1,790

2,019

4,288

10,190

Table A3 Percentage of individuals with a university degree by high school track, macro-area and city size

Lower Sec.

Vocational

Technical

Academic

Total

Macro-area

North-West

0

5.9

13.8

52.0

18.0

North-East

0

7.3

18.7

52.4

19.8

Center

0

9.7

17.6

54.2

22.1

South

0

10.7

18.9

50.7

19.7

City size

>=20k

0

6.0

14.4

47.8

15.6

20k-50k

0

8.5

21.3

51.8

19.7

50k-250k

0

13.2

20.0

54.7

23.2

>250k

0

10.9

19.2

57.8

29.4

Total

0

8.3

17.3

52.1

19.8

Table A4 Average partial effects of high school track referred to additional contrasts among categories ccompared to those reported in Figure 1

95% confidence intervals

APE

SE

p-value

Lower bound

Upper bound

Employment

(LowerSec. vs Vocational)

-0.868

0.087

0.000

-1.039

-0.698

(Vocational vs Technical)

-0.239

0.078

0.002

-0.392

-0.086

(Technical vs Academic)

-0.013

0.062

0.839

-0.134

0.109

Unemployment

(LowerSec. vs Vocational)

0.791

0.107

0.000

0.582

1.001

(Vocational vs Technical)

0.203

0.096

0.034

0.015

0.392

(Technical vs Academic)

-0.006

0.075

0.941

-0.153

0.142

Salariat

(LowerSec. vs Vocational)

-1.078

0.187

0.000

-1.445

-0.711

(Vocational vs Technical)

0.032

0.118

0.784

-0.198

0.263

(Technical vs Academic)

-0.938

0.086

0.000

-1.105

-0.770

Unskilled working class

(LowerSec. vs Vocational)

0.824

0.216

0.000

0.401

1.248

(Vocational vs Technical)

0.598

0.225

0.008

0.158

1.039

(Technical vs Academic)

0.800

0.296

0.007

0.219

1.381

Table A5 Predictive margins presented in figure 2 in the main article

Employment

Unemployment

Estimate

95%CI: Min

95%CI: Max

Estimate

95%CI: Min

95%CI: Max

Compulsory#North-West

0.685

0.641

0.729

0.182

0.141

0.223

Compulsory#North-East

0.724

0.680

0.769

0.140

0.102

0.178

Compulsory#Center

0.681

0.632

0.730

0.186

0.142

0.230

Compulsory#South

0.363

0.320

0.406

0.491

0.436

0.546

Vocational#North-West

0.840

0.807

0.873

0.090

0.063

0.116

Vocational#North-East

0.856

0.825

0.886

0.072

0.049

0.094

Vocational#Center

0.782

0.735

0.829

0.141

0.100

0.181

Vocational#South

0.574

0.526

0.623

0.276

0.227

0.326

Technical#North-West

0.872

0.853

0.891

0.070

0.055

0.085

Technical#North-East

0.864

0.843

0.884

0.075

0.060

0.091

Technical#Center

0.818

0.793

0.844

0.111

0.090

0.131

Technical#South

0.635

0.610

0.660

0.239

0.215

0.263

Academic#North-West

0.874

0.854

0.894

0.065

0.050

0.080

Academic#North-East

0.864

0.843

0.885

0.074

0.057

0.090

Academic#Center

0.808

0.783

0.833

0.118

0.098

0.138

Academic#South

0.646

0.620

0.672

0.240

0.214

0.267

Salariat

Unskilled working class

Compulsory#North-West

0.129

0.064

0.195

0.157

0.099

0.215

Compulsory#North-East

0.165

0.077

0.252

0.098

0.041

0.154

Compulsory#Center

0.140

0.061

0.220

0.202

0.124

0.281

Compulsory#South

0.178

0.095

0.261

0.163

0.102

0.223

Vocational#North-West

0.333

0.254

0.412

0.040

0.013

0.067

Vocational#North-East

0.294

0.220

0.368

0.072

0.031

0.112

Vocational#Center

0.354

0.248

0.460

0.062

0.010

0.115

Vocational#South

0.376

0.290

0.461

0.141

0.073

0.210

Technical#North-West

0.332

0.284

0.379

0.030

0.012

0.049

Technical#North-East

0.324

0.273

0.376

0.033

0.012

0.054

Technical#Center

0.319

0.266

0.373

0.059

0.029

0.090

Technical#South

0.345

0.303

0.388

0.058

0.034

0.081

Academic#North-West

0.552

0.499

0.605

0.012

-0.002

0.027

Academic#North-East

0.533

0.474

0.592

0.007

0.000

0.015

Academic#Center

0.538

0.485

0.591

0.032

0.008

0.055

Academic#South

0.568

0.521

0.616

0.031

0.008

0.054

Table A6 Predictive margins presented in figure 4 in the main article

Employment

Unemployment

Estimate

95%CI: Min

95%CI: Max

Estimate

95%CI: Min

95%CI: Max

LowerSec.#North#Men

0.841

0.810

0.873

0.129

0.099

0.158

LowerSec.#North#Women

0.555

0.514

0.596

0.222

0.186

0.259

LowerSec.#South#Men

0.523

0.444

0.603

0.436

0.355

0.517

LowerSec.#South#Women

0.221

0.181

0.261

0.546

0.482

0.610

Vocational#North#Men

0.919

0.896

0.943

0.063

0.042

0.084

Vocational#North#Women

0.748

0.715

0.782

0.136

0.109

0.163

Vocational#South#Men

0.761

0.693

0.830

0.201

0.136

0.266

Vocational#South#Women

0.404

0.339

0.469

0.367

0.294

0.440

Technical#North#Men

0.929

0.916

0.941

0.054

0.043

0.065

Technical#North#Women

0.783

0.761

0.804

0.121

0.103

0.138

Technical#South#Men

0.796

0.765

0.826

0.161

0.134

0.188

Technical#South#Women

0.481

0.442

0.519

0.345

0.305

0.386

Academic#North#Men

0.901

0.881

0.921

0.068

0.051

0.084

Academic#North#Women

0.787

0.766

0.807

0.111

0.096

0.127

Academic#South#Men

0.733

0.685

0.782

0.192

0.149

0.235

Academic#South#Women

0.536

0.504

0.568

0.304

0.270

0.337

Salariat

Unskilled working class

LowerSec.#North#Men

0.176

0.110

0.241

0.129

0.099

0.158

LowerSec.#North#Women

0.088

0.049

0.127

0.222

0.186

0.259

LowerSec.#South#Men

0.192

0.067

0.316

0.436

0.355

0.517

LowerSec.#South#Women

0.164

0.075

0.253

0.546

0.482

0.610

Vocational#North#Men

0.299

0.217

0.381

0.063

0.042

0.084

Vocational#North#Women

0.349

0.293

0.404

0.136

0.109

0.163

Vocational#South#Men

0.347

0.222

0.471

0.201

0.136

0.266

Vocational#South#Women

0.426

0.321

0.531

0.367

0.294

0.440

Technical#North#Men

0.362

0.318

0.407

0.054

0.043

0.065

Technical#North#Women

0.277

0.243

0.311

0.121

0.103

0.138

Technical#South#Men

0.374

0.317

0.431

0.161

0.134

0.188

Technical#South#Women

0.294

0.240

0.348

0.345

0.305

0.386

Academic#North#Men

0.512

0.451

0.573

0.068

0.051

0.084

Academic#North#Women

0.549

0.515

0.584

0.111

0.096

0.127

Academic#South#Men

0.500

0.416

0.584

0.192

0.149

0.235

Academic#South#Women

0.602

0.554

0.651

0.304

0.270

0.337

Appendix B. Average partial effects of main analysis

Fig. B1

Fig. B1

Binomial logistic regression to analyse occupational outcomes: average partial effects of high school track of diploma and 95% confidence intervals across geographical areas. Note: reference category is academic track (the zero line)

Fig B2

Fig B2

Binomial logistic regression to analyse occupational outcomes: average partial effects of high school track of diploma and 95% confidence intervals across geographical areas for the subsample of individuals without a university degree.
Note: reference category is academic track (the zero line).

Appendix C. Additional results for individuals without a university degree

Fig. C1

Fig. C1

Binomial logistic regression to analyse occupational outcomes: predicted probabilities of high school tracks across geographical areas (North vs South) and gender for the subsample of individuals without a university degree

Torna su

Bibliografia

Almadiploma (2020), Le scelte dei diplomati 2016 a tre anni dal titolo, Retrieved on 7th April 2021 at: https://www.almadiploma.it/info/pdf/scuole/occupazione2019/Schede_a_tre_anni.pdf.

Arum R., Shavit Y. (1995), Secondary vocational education and the transition from school to work, «Sociology of Education», 68, 3, pp. 187-204.

Azzolini D., Barone C. (2012), Tra vecchie e nuove disuguaglianze: la partecipazione scolastica degli studenti immigrati nelle scuole secondarie superiori in Italia, «Rassegna italiana di sociologia», 53, 4, pp. 687-718.

Bagnasco A. (1977), Tre Italie: la problematica dello sviluppo territoriale in Italia, Bologna, il Mulino.

Ballarino G., Panichella N., Triventi M. (2014), School expansion and uneven modernization. Comparing educational inequality in Northern and Southern Italy, «Research in Social Stratification and Mobility», 36, pp. 69-86.

Barone C., Triventi M., Assirelli G. (2018), Explaining social inequalities in access to university: a test of rational choice mechanisms in Italy, «European Sociological Review», 34, 5, pp. 554-569.

Barone C., Triventi M., Facchini M. (2021), Social origins, tracking and occupational attainment in contemporary Italy, «Longitudinal and Life Course Studies», published online April 2021, online ISSN 1757-9597, https://doi.org/10.1332/175795921X16137571163021.

Becker R., Hecken A. E. (2009), Why are working-class children diverted from universities? — An empirical assessment of the diversion thesis, «European Sociological Review», 25, 2, pp. 233-250.

Breen R., Jonsson J. O. (2000), Analyzing educational careers: A multinomial transition model, «American Sociological Review», 65, 5, pp. 754-772.

Cedefop (2013), Piloting a European survey on skill needs, «Cedefop Working paper», 36.

Contini D., Scagni A. (2013), Social origin inequalities in educational careers in Italy: performance or decision effects?, in Jackson M. (ed.), Determined to succeed? Performance versus choice in educational attainment, Stanford, Stanford University Press. pp. 149-184.

Contini D., Triventi M. (2016), Between formal openness and stratification in secondary education: Implications for social inequalities in Italy, in Blossfeld H.-P., Buchholz S., Skopek J., Triventi M. (eds.), Models of secondary education and social inequality, Cheltenham, Edward Elgar Publishing, pp. 305-322.

Correll S. (2001), Gender and the career choice process: the role of biased beliefs, «American Journal of Sociology», 106, 6, pp. 1691-1730.

Daniele V., Malanima P. (2014), Falling disparities and persisting dualism: Regional development and industrialisation in Italy, 1891-2001, «Investigaciones de Historia Económica-Economic History Research», 10, 3, pp. 165-176.

Frank M. R., Sun L., Cebrian M., Youn H., Rahwan Y. (2019), Small cities face greater impact from automation, retrieved on 7th April 2021 at: https://arxiv.org/pdf/1705.05875.pdf.

Forster A. G., Bol T., Van de Werfhorst H. G. (2016), Vocational education and employment over the life cycle, «Sociological Science», 3, pp. 473-494.

Hampf F., Woessmann L. (2017), Vocational vs. general education and employment over the life cycle: New evidence from PIAAC, «CESifo Economic Studies», 63, 3, pp. 255-269.

Hanushek E. A., Schwerdt G., Woessmann L., Zhang L. (2017), General education, vocational education, and labor-market outcomes over the lifecycle, «Journal of human resources», 52, 1, pp. 48-87.

Isfol (2016), Rapporto su istruzione e formazione professionale, Roma, Isfol.

Istat (2016), I percorsi di studio e di lavoro dei diplomati, Roma, Istat.

Istat (2020), Il mercato del lavoro 2019: una lettura integrata, Roma, Istat.

Lucas S. R. (2001), Effectively maintained inequality: Education transitions, track mobility, and social background effects, «American Journal of Sociology», 106, 6, pp. 1642-1690.

Kamens D., Meyer J., Benavot A. (1996), Worldwide Patterns in Academic Secondary Education Curricula, «Comparative Education Review», 40, 2, pp. 116-138.

Korber M., Oesch D. (2019), Vocational versus general education: Employment and earnings over the life course in Switzerland, «Advances in Life Course Research», 40, pp. 1-13.

Mandrone E. (2013), Indagine Isfol-Plus. Panel biennali 2010-2011, Roma, Isfol.

Morgan S. L., Winship C. (2015), Counterfactuals and causal inference, Cambridge, Cambridge University Press.

Odoardi I., Muratore F. (2019), The North–South Divergence in Italy during the Great Recession, «The Manchester School», 87, 1, pp. 1-23.

Panichella N. (2014), Meridionali al Nord, Bologna, il Mulino.

Panichella N., Triventi M. (2014), Social inequalities in the choice of secondary school: Long-term trends during educational expansion and reforms in Italy, «European Societies», 16, 5, pp. 666-693.

Reyneri E. (2017), Sociologia del mercato del lavoro, Bologna, il Mulino.

Rose D., Harrison E. (2007), The European socio-economic classification: a new social class schema for comparative European research, «European Societies», 9, 3, pp. 459-490.

Rose D., Harrison E. (2010), Social class in Europe, London, Routledge.

Rözer J. J., Bol T. (2019), Labour market effects of general and vocational education over the life-cycle and across time: accounting for age, period, and cohort effects, «European Sociological Review», 35, 5, pp. 701-717.

Scalmato V. (a cura di) (2008), La domanda di istruzione e formazione degli allievi in diritto-dovere all’istruzione e formazione: i risultati dell’indagine Isfol, Roma, Isfol.

Schizzerotto A., Barone, C. (2006), Sociologia dell’istruzione, Bologna, il Mulino.

Shavit Y., Arum R., Gamoran A. (ed.) (2007), Stratification in Higher Education: A Comparative Study, Stanford, Stanford University Press.

Shavit Y., Müller W. (1998), From School to Work. A Comparative Study of Educational Qualifications and Occupational Destinations, Oxford, Oxford University Press.

Id. (2000), Vocational secondary education, «European Societies», 2,1, pp. 29-50.

Triventi M. (2013), The role of higher education stratification in the reproduction of social inequality in the labor market, «Research in Social Stratification and Mobility», 32, pp. 45-63.

Van Houtte M., Demanet J., Stevens P. (2012), Self-esteem of academic and vocational students: Does within-school tracking sharpen the difference?, «Acta Sociologica», 55, 1, pp. 11-24.

Torna su

Note

1 After lower secondary school, students can also opt for three-year vocational training courses (istruzione e formazione professionale), which do not afford access to university education. However, these courses are quantitatively marginal and they are often offered by the same schools that deliver the full vocational certificates. In the analyses, we aggregate vocational training certificates to vocational diplomas.

2 Due to limited sample size, in the empirical part we cannot analyze the curricular differentiation of these tracks, but it should be noted that, within licei, two curricula (classical and scientific studies) are more academic-oriented than the others (foreign languages and teacher education). For a discussion of this point, see Scalmato (2008).

3 The ISFOL PLUS survey involves additional waves, which cannot be used because they lack some of the variables needed for the analyses.

4 City size refers to the place where respondents live, rather than to the place where they work. This is a significant limitation, given that mobility for work reasons is not rare in Italy.

5 Unfortunately, the ISFOL plus data for parental occupation do not differentiate between self-employed and manual workers. Moreover, firm size is not available for parental occupation.

6 In the North-East the absolute penalization amounts to 7 p. p. , while in the South to 25 p. p. The relative penalization, expressed in terms of risk ratios is thus around 2 in both macro-areas.

7 The direction of potential biases in unclear when considering graduates without a tertiary degree, because students enrolled in academic tracks tend to be positively selected, but only a minority of them do not achieve any tertiary degree.

Torna su

Indice delle illustrazioni

Titolo Fig. 1
Legenda Binomial logistic regression to analyse occupational outcomes: average partial effects of high school track of diploma and 95% confidence intervals
URL http://journals.openedition.org/qds/docannexe/image/4168/img-1.jpg
File image/jpeg, 163k
Titolo Fig. 2
Legenda Binomial logistic regression to analyse occupational outcomes: predicted probabilities of high school tracks across geographical areas
URL http://journals.openedition.org/qds/docannexe/image/4168/img-2.jpg
File image/jpeg, 204k
Titolo Fig. 3
Legenda Binomial logistic regression to analyse occupational outcomes: predictive margins of high school track of diploma and 95% confidence intervals for the subsample of individuals without a university degree
URL http://journals.openedition.org/qds/docannexe/image/4168/img-3.jpg
File image/jpeg, 216k
Titolo Fig. 4
Legenda Binomial logistic regression to analyse occupational outcomes: predicted probabilities of high school tracks across geographical areas (North vs South) and gender
URL http://journals.openedition.org/qds/docannexe/image/4168/img-4.jpg
File image/jpeg, 194k
Titolo Fig. 5
Legenda Binomial logistic regression to analyse occupational outcomes: predicted probabilities of high school tracks across city sizes
URL http://journals.openedition.org/qds/docannexe/image/4168/img-5.jpg
File image/jpeg, 254k
Titolo Fig. 6
Legenda Binomial logistic regression to analyse occupational outcomes: predicted probabilities of high school tracks across city sizes for the subsample of individuals without a university degree
URL http://journals.openedition.org/qds/docannexe/image/4168/img-6.jpg
File image/jpeg, 233k
Titolo Fig. B1
Legenda Binomial logistic regression to analyse occupational outcomes: average partial effects of high school track of diploma and 95% confidence intervals across geographical areas. Note: reference category is academic track (the zero line)
URL http://journals.openedition.org/qds/docannexe/image/4168/img-7.jpg
File image/jpeg, 206k
Titolo Fig B2
Legenda Binomial logistic regression to analyse occupational outcomes: average partial effects of high school track of diploma and 95% confidence intervals across geographical areas for the subsample of individuals without a university degree. Note: reference category is academic track (the zero line).
URL http://journals.openedition.org/qds/docannexe/image/4168/img-8.jpg
File image/jpeg, 184k
Titolo Fig. C1
Legenda Binomial logistic regression to analyse occupational outcomes: predicted probabilities of high school tracks across geographical areas (North vs South) and gender for the subsample of individuals without a university degree
URL http://journals.openedition.org/qds/docannexe/image/4168/img-9.jpg
File image/jpeg, 172k
Torna su

Per citare questo articolo

Notizia bibliografica

Carlo Barone e Moris Triventi, «Is vocational education a safety net?»Quaderni di Sociologia, 84- LXIV | 2020, 59-89.

Notizia bibliografica digitale

Carlo Barone e Moris Triventi, «Is vocational education a safety net?»Quaderni di Sociologia [Online], 84- LXIV | 2020, online dal 01 septembre 2021, consultato il 13 février 2025. URL: http://journals.openedition.org/qds/4168; DOI: https://doi.org/10.4000/qds.4168

Torna su

Autori

Carlo Barone

Observatoire sociologique du Changement, Sciences Po, Paris

Articoli dello stesso autore

Moris Triventi

Dipartimento di Sociologia e Ricerca Sociale, Università di Trento

Articoli dello stesso autore

Torna su

Diritti d'autore

CC-BY-NC-ND-4.0

Solamente il testo è utilizzabile con licenza CC BY-NC-ND 4.0. Salvo diversa indicazione, per tutti agli altri elementi (illustrazioni, allegati importati) la copia non è autorizzata ("Tutti i diritti riservati").

Torna su
Cerca su OpenEdition Search

Sarai reindirizzato su OpenEdition Search