The Role of the Welfare State for NEETs: Exploring the Association Between Public Social Spending and NEET in European Countries

Abstract

This study explores the role of the welfare state in reducing young people not being in education, employment, or training (NEET)s across 15 European countries. Using data from the Survey of Adult Skills in the Program for the International Assessment of Adult Competencies (PIAAC) in combination with the Social Expenditure Database, we conducted cross-national analysis to reveal whether the increase in public social spending moderates the risk of being NEET at a young adult age, especially for socially disadvantaged young people. Our results highlighted that the rise of one percentage point of public social spending per gross domestic product, including social expenditure on education, active labor market, and unemployment, is significantly associated with decreasing the odds of being NEET. Furthermore, these social expenditures appeared to lower the NEET risk given socially disadvantaged backgrounds suggesting that young people with low educational levels, whose parents have low educational attainment levels, non-immigrant families, and females are likely to benefit given the robust social protection system. These findings suggest that public social spending may be an effective investment in promoting the social involvement of young people from socially disadvantaged background.

Keywords

NEET public social spending education expenditure active labor market unemployment benefit

Introduction

The increasing number of young people not in education, employment, or training (NEET) has gained growing concerns across European countries (Eurofound, 2012; International Labour Organization [ILO], 2019). Within their borders, several European countries are rapidly expanding preventive intervention policies and programs to integrate young people into the job market and education to reduce economic and social consequences associated with NEETs (Carcillo et al., 2016; Eurofound, 2012; ILO, 2019). While these intervention programs may encourage young people to gain self-sufficiency, international organizations suggest the need to strengthen social safety nets to support young people under challenging circumstances (Carcillo et al., 2015; Eurofound, 2012). According to International Organisations’ reports, public social spending may avoid young people’s unemployment and poverty by providing benefits to meet their basic needs to engage in education and the labor market under constraint economic conditions (Carcillo et al., 2015; Eurofound, 2012; ILO, 2019). Nevertheless, a little empirical examination has been conducted to reveal the extent to which the welfare state is associated with NEET risk, especially for socially disadvantaged groups of young people.

To this end, this study examined the association between public social spending, including spending on education, public spending on labor and employment protection, and the NEET risk across 15 European countries. In addition, the moderating impact of public social expenditure on NEET risk given socially disadvantaged backgrounds was investigated to verify the role of social protection system in promoting the social involvement of young people under challenging circumstances. To address this objective, we constructed an international comparative data set by incorporating the Programme for the International Assessment of Adult Competencies (PIAAC) and the national level of information from OECD Social and Welfare statistics to address this objective. In doing so, we employed a rigorous conceptualization of NEET status for those youths who have not participated in education, employment, or training during the past year. Defining NEET status based on a relatively long period is expected to capture discouraged and inactive young people lost in the transition from education to labor.

Much of previous NEET studies predominantly focused on identifying risk factors that may lead youth to social disconnection within a country (Alfieri et al., 2015; Assmann & Broschinski, 2021; Bynner & Parsons, 2002; Tamesberger & Bacher, 2014). These studies contributed to developing preventive policies such as mentoring programs, school-work coordination, and mental health support programs to prevent social disengagement (Carcillo et al., 2015; Pemberton, 2008). Such intervention programs may indeed be valuable to assist young people in incorporating into education and the labor market under challenging circumstances. However, there is not much empirical evidence of whether the welfare state may effectively achieve reductions in NEETs, especially from socially disadvantaged family backgrounds. In this view, this study aimed to shed light on the importance of redistributive structural intervention in integrating a new generation into society.

Literature Review

Individual Factors and NEET Risk

Many studies suggest that structural constraints, such as economic recession and unstable labor conditions, affect young people’s social exclusion (Caroleo et al., 2020). For example, Gangl (2002) and Wolbers (2007) argue that a country’s unfavorable economic condition often translates into labor market uncertainty, increasing the unemployment rate and inactivity risks of young people. Indeed, The Great Recession in 2007 to 2008 caused labor market turmoil, which decreased family income, and left many young people without jobs (Caroleo et al., 2020). The recovery has been slow and fragile to bring young people back into employment (ILO, 2019). Accordingly, an OECD report (2016) showed that the unemployed NEET rate in most OECD countries had dramatically increased in the aftermath of the Great Recession, reaching a maximum of 7.5% in 2013. However, the average NEET rate across the OECD countries was not likely to fall more than 2% to 3% percentage points as the economy recovers, and socially disadvantaged young people’s employment was particularly slow. Furthermore, after the economic crisis, the NEET rate in lower-middle-income countries raised dramatically, indicating more than twice as high compared to high-income countries (12%) and nearly 50% more than low-income countries (ILO, 2019). Evidence from the analysis of European LFS (1992–2009) aligns with these findings showing that the GDP decrease is associated with a 0.18% increase in the NEETs rate (Eurostat, 2019).

Under challenging economic circumstances, socially disadvantaged young people are particularly vulnerable in the education and labor market (Alfieri et al., 2015; Kevelson et al., 2020). According to previous studies that revealed individual risk factors, socioeconomic status such as parental educational level are the critical risk factors that raise the likelihood of becoming NEET relative to childhood experiences, parental psychiatric disorder, and family structure, among others (Alfieri et al., 2015; Thompson, 2011). This pattern of results is even consistent in countries known for low socioeconomic inequality with a robust welfare system. For example, as found by a Finnish longitudinal study based on nearly 10,000 children born between 1986 and 1993, the youths of primary parental education showed 5.33 times higher the odds of being NEET at 18 compared to the youths of secondary parental education, as well as the parental education was more closely related to preventing social exclusion than parent-youth relationship, and family support (Pitkänen et al., 2021). Specifically, for men, whether the mother’s completing higher education had the highest marginal effect among all the variables, followed by the father’s completing higher education.

Moreover, both international studies and country-specific studies consistently confirmed that parental educational level displays intergenerational influences that raise young people’s chance of being NEET even after controlling for other family characteristics (Carcillo et al., 2015; Kevelson et al., 2020). Indeed, young people with low-educated parents were more likely to leave school earlier than their counterparts, which in turn increased the likelihood of entering a permanent NEET situation (Bynner & Parsons, 2002; Kelly & McGuinness, 2015; Mauro & Mitra, 2015; Rennison et al., 2005; Tamesberger & Bacher, 2014). Along this line, empirical evidence from European countries highlighted that youth with no or low educational qualifications at age 16 demonstrated six times higher NEET risk than those who completed secondary education or above, showing that the highest the magnitude of effect size (Bynner & Parsons, 2002; Eurofound, 2012; Rodriguez-Modroño, 2019). Moreover, in the U.S., high school graduation and academic experiences tend to maintain their impact throughout young adult age, whereas SES and other family characteristics appear to diminish over time (Millett & Kevelson, 2018). In many countries, “education first” has been the primary approach to reducing early school leavers to promote individual participation in the labor market.

However, the empirical evidence is mixed depending on whether a family’s income is related to the likelihood of being NEETs. Studies based in Europe and the U.S. have often reported that young people from families with a lower household income are more likely to become NEET than those with higher-income families; those students receiving free lunch and food stamps were more likely to become NEET (Bynner & Parsons, 2002; ILO, 2019; Rodriguez-Modroño, 2019). Interestingly, household income increased the youth’s probability of becoming an inactive NEET in Japan, who neither expresses the desire to work nor seek employment. Specifically, the highest household income (about 15 million yen or more per year) was positively associated with inactive NEET risk, implying that individuals from wealthy families may not find necessary work and tend to depend on their families’ wealth (Genda, 2007).

Other demographic factors such as immigrant status, gender, and age were also highly associated with NEET risk. According to an OECD report, females were more often disconnected than men in employment, education, and training in most OECD countries. The risk of young women being NEET was more than three times higher than that for men, although the youth NEET rate between gender declined slightly between 2005 and 2018 (ILO, 2019; Organisation for Economic Cooperation and Development [OECD], 2016). Immigrant young people were also more often identified as NEET at a young adult age across European countries. Specifically, foreign-born youth showed 1.5 to 1.7 times higher chances of being NEET than native youth across the OECD and European countries (Eurofound, 2012). The labor market discrimination, disadvantaged socioeconomic status (SES), and language barriers, among others, appear to influence their participation in the job market and education despite the NEET rate of immigrant young people (Eurofound, 2012).

Based on these findings, related studies define the NEET youth as a consequence of a structural problem and suggest the need for policy intervention and fiscal input to reduce the NEET population across OECD countries (OECD, 2008). Accordingly, different countries actively implemented public policies and preventive intervention programs to keep youth from leaving education to become NEETs or reintegrating NEETs into schooling and the labor market (Barrett & McPeak, 2006; Eurofound, 2012). Nevertheless, the extent to which these institutional efforts, including the social protection system, reduce NEETs has not been widely documented (Assmann & Broschinski, 2021; Caroleo et al., 2020).

Public Social Spending and NEET Risk

International organizations such as OECD also suggest the need to expand public social spending to support young people’s living conditions (Carcillo et al., 2015). They claim that the social protection system that can directly support low-income youth may be the most effective strategy to avoid poverty and “meet the basic needs of young people to not withdraw from education and the labor market” (Barrett & McPeak, 2006, p. 37). Indeed, public social spending is documented as an effective redistributive approach to promote youth integration, and prevent poverty, economic inequality, and social exclusion (Lindert, 2003; Li et al., 2000; Ulu, 2018).

Of those public social spending, the expenditure on education, which is directly related to the quality of the educational system, emerged as the crucial factor that explains the variation in NEET population across European countries (Pastore & Zimmermann, 2019). According to these studies, educational quality is fundamental, given that school education is the primary institute that delivers the necessary skills required from the labor market (Caroleo et al., 2020). Furthermore, the educational system that can provide clear signals of job seekers’ skills in the labor market may facilitate the school-to-work transition, reducing youth unemployment (Breen, 2005). Caroleo et al. (2020), in this regard, revealed that education expenses relative to GDP appeared to be the most predictive indicator that reduced the likelihood of NEET risk. Findings from Eurofound (2012) align with this line of results, highlighting the importance of efficacy of an educational system that adequately supports school-to-work transition and work-based training as an important factor that may reduce NEET (Breen, 2005; Bynner & Parsons, 2002).

Alongside the quality of the educational system, there is a consensus in the literature that social expenditures on the active labor market and unemployment benefits are strong determinant factors that stimulate youth’s integration in the labor market and education (Caroleo et al., 2020; Marques & Hörisch, 2020; Russell & O’Connell, 2001). These studies claimed that active labor market policies and unemployment benefits induce young people into the labor market or education by providing adequate training, unemployment subsidies, hiring and direct job-creating subsidies, and public employment services. Indeed, a recent cross-national study added empirical evidence along this line, indicating that young people are more likely to make a rapid school-to-work transition from unemployment to employment in countries with higher spending on active labor market policy and unemployment benefits (Caroleo et al., 2020; Russell & O’Connell, 2001). Relevant evidence is also observed from the study of Assmann and Broschinski (2021), indicating a higher proportion of discouraged and unemployed NEETs in countries with a lack of active labor market policies, like Southern and Central Eastern European countries.

On the other hand, more controversial findings are suggested regarding whether the employment protections of the labor market are associated with NEET risk (Eurofound, 2012). This pattern of results is partially given that the consequence of employment protection and NEET risk vary whether the rigidity of the labor market is geared toward regular or temporary work. The existing theory thus distinguishes the employment protection legislation of laborers of individual and collective dismissal for regular and temporary contracts as it entails a varying impact on the labor market entrance of young people. While some argue that the rigidity of the regular labor market may impede the creation of new jobs for young people, making the school-to-work transition lengthy and uncertain (Piopiunik & Ryan, 2012), some studies indicate that the flexibility of temporary work may lower the labor market barrier providing more opportunities for young people. Concerning the risk of NEET, recent studies suggested that a flexible labor market protection for temporary workers, especially the deregulation of temporary contracts, has significantly reduced the NEET population, especially after an economic crisis (Assmann & Broschinski, 2021; Caroleo et al., 2020; Eurofound, 2012). There is still discussion and evidence, however, that the strictness of the use of temporary workers may create new jobs that otherwise would not have existed and thus raise the chance of labor market entrance of young people (Korpi & Levin, 2001).

In addition to these findings, recent studies on comparative social policy literature suggest the need to consider the directionality of social protection regimes of a country. As Lynch (2006) noted, the quantity of welfare spending and the age orientation of the welfare state needs to be considered given the directionality of the regimes may not necessarily accord with welfare regime type, levels of GDP per capita of general social spending (Comelli, 2021). In countries where the welfare regime concentrates on pension schemes and the elders, participants and students may create different youth unemployment rates compared to those countries that focus more on the active labor market measures. Comelli (2021), in this view, reported that a higher share of NEETs tends to appear in countries that have a welfare system that is elderly biased in conservative and Anglo-liberal countries.

Despite the empirical evidence of the importance of government public spending may function to integrate young people into the labor market and education, the question remains whether public social spending may benefit the most needed group of young people. While it may be true that the increase in social expenditure might lead to a larger redistributive budget, as welfare state research suggests, it does not imply that all young people may equally benefit within nations (Zwiers & Koster, 2015). The evidence concerning the distributional impact of government social spending is mixed, indeed. On the one hand, welfare state research argues that the socially disadvantaged group is more likely to benefit first due to being exposed more frequently to various social programs that seek to resolve inequality. Kevelson et al. (2020), in part, support this argument by suggesting that public expenditure on families, family leave policies, and childcare subsidies may moderate some of the impacts of parents’ educational level on the risk of becoming NEETs.

Although several studies suggest that public social spending may reduce social inequality, the idea has been well established that the consequences of social protection may vary within a country. In particular, International research showed that welfare policies significantly reduced absolute poverty, whereas the largest share, about 50% of the poor, was not reached by social assistance (Gaentzsch, 2018; Zwiers & Koster, 2015). Gaentzsch (2018), in this vein, discussed that “targeted programs may reach to poor, but universal programs can benefit the middle and higher classes,” indicating that the effects of social spending are unequally distributed by regions and other demographic characteristics (Goodin & LeGrand, 1987; LeGrand, 1982; Tullock, 1983). This research suggests that the excessively marginalized group of people may still be left behind, given their lack of access to the information or specific spatial pattern, among others, despite the overall expansion of social spending (Zwiers & Koster, 2015).

In sum, previous studies related to NEET have mainly focused on revealing individual risk factors based on a specific region and country-specific data. Although these studies provided substantial implications in developing youth policies, little is known about the redistributive structural intervention promoted to integrate youths into the education or labor market. In particular, not many international comparative studies have been conducted examining the extent to which public social spending may attenuate the risk of being NEET, especially for socially disadvantaged groups of young people. The lack of attention concerning the moderating impact of public spending may be partly because most previous studies have analyzed aggregated national-level data sets without incorporating information at the individual level (Zwiers & Koster, 2015). Nevertheless, this question remains vital as it examines whether government spending on social protection systems is equally shared across young people within a country. To this end, using PIAAC combined with the Social Expenditure Database, the present study aims to expand existing NEET literature by exploring the role of the welfare state in reducing NEET risk for socially disadvantaged groups of young people.

Research Question

Previous studies have concentrated on identifying individual risk factors and evaluating intervention programs at the national level and suggested the importance of preventive intervention policies and programs to motivate and help young people gain self-sufficiency. Although these studies provided substantial implications for developing NEET preventive policies across countries, not much attention has been devoted to the welfare state’s role in supporting socially vulnerable groups of young people to engage in employment, education, and training. To this end, this study aims to merge welfare state and NEET studies by investigating the relationship between public social spending and NEET risk, especially for socially disadvantaged groups of young people. In doing so, this study aims to broaden the existing literature by investigating whether the public expenditure on education, unemployment, and labor market, as well as other institutional characteristics such as employment protection and welfare directionality, among others, to explain the cross-national variation of NEETs holding individual and institutional attributes at constant. To meet this research objective, we examined the following three questions using PIAAC and Social Expenditure Database: (a)What are significant individual and institutional factors that significantly explain NEET risk across European countries? (b) to what extent is the degree of public social spending, alongside government spending on education, employment protection, and the national labor market, associated with the risk of becoming NEET? (c) to what extent does public social spending contributes to lowering the risk of being NEET for young people from a socially disadvantaged background? These questions will provide empirical evidence for identifying the importance of different spheres of public social spending for the vulnerable group of young people across European countries.

Data and Method

Data

The data for this study comes from the Programme for the International Assessment of Adult Competencies (PIIAC), which was designed to assess a variety of information on adults’ work and life, including literacy skills such as information processing, literacy numeracy skills (Kirsch & Thorn, 2013). PIAAC performed a four-stage stratified random sampling design receiving national representative cases among non-institutionalized person ages 16 to 65. One of the main advantages of using PIAAC is that the participating countries of PIAAC shared and monitored the data collecting process to ensure data comparability across countries (Kirsch & Thorn, 2013). Furthermore, it includes various information on respondents’ backgrounds, such as educational level, NEET status, and other demographic characteristics, which enable the researcher to account for risk factors that may affect NEET risk. We also constructed a national-level data set that comprises government public social spending, national Gross Domestic Product (GDP), population data from the OECD Database, and the OECD Social Expenditure Database of 2011 (OECD, 2020). This national-level data set was then merged into the PIAAC data, consisting of individual and family information to examine the impact of public social spending, holding other individual and structural determinants constant. To maximize international comparability, the OECD’s approach (2013) to remarking NEET risks between 15 and 29 was used. Finally, this study used the PIAAC data surveyed in 2011 and 2012, primarily focusing on the age group of 15 to 29 across 16 European countries that have all the data needed for analysis.

Measures

The dependent variable of this study is NEET status, defined as whether the youth has participated in education, employment, or training during the past year. Specifically, the participation in education or training includes only those who attempted part-time or full-time participation and excludes those who have engaged in the activities for a very short period. Besides, employment and education participation includes those who have worked more than 1 hr of paid work during the past year. The advantage of using such rigorous identification of NEET status is that it excludes those who have temporarily opted out from the labor force or education, capturing discouraged and inactive NEETs (Kirsch & Thorn, 2013). Furthermore, defining NEET status based on a relatively long period is known to improve the reliability and validity of NEET status (Eurofound, 2012). The NEET status is coded as a binary variable.

The primary independent variables are the total expenditure of social spending, spending on education, spending on employment protection, and labor markets. These are numerical variables of the proportion of public social spending per GDP, representing the national economy scale’s relative level of public social spending. The total public social spending is a financial flow that the general government controls, and it is not included social benefits provided by the private sector. It generally comprises cash benefits, tax benefits for social purposes, and financial support for social services that may target socially vulnerable groups such as low-income and unemployed persons. Public spending on education comprises the expenditure of the primary to non-tertiary education, and educational services provided to families and related public subsidies as a percentage of GDP. Public unemployment spending is the benefits for those unemployed measured in percentage of GDP, which includes cash subsidies to unemployed workers, payments from public funds, and payments of pensions for those before they reach pensionable age. Public spending on the labor market is the percentage of GDP spending on training, hiring, direct job-creating subsidies, unemployment benefits, and public employment services. This information is taken from the OECD Data and the OECD Social Expenditure Statistics, and merged into PIAAC data.

Previous studies and preliminary analyses mainly guided the selection of individual and institutional covariates. Following covariates were selected. At the institutional level, (a) the national GDP, transformed into a natural logarithm to minimize the fluctuations in the data, was employed as a proxy variable to measure the volume of economy and inequality of a country (Eurostat, 2012). (b) Each country’s 15 to 29 year population was selected as the covariate, given it represents the labor market competitiveness that may engender difficulty for inexperienced young people to participate in the labor market (Müller, 2005). (c) The unemployment rate is the number of people without work as a percentage of the labor force and is seasonally weighted. (d) The intergenerational inequality index is designed to reflect whether the welfare state is focused on a pro-elderly or active population to estimate whether the directionality of government public spending was integrated to control the directionality of welfare regimes (Comelli, 2021; Lynch, 2006). It is a measured value of the elderly-focused spending divided by the non-elderly-focused spending among public social spending. To be specific, the elderly-focused public social spending includes pension and other benefits for the elderly, welfare benefits for the disabled, and survivors, while non-elderly focused spending includes support for family (birth, childcare services, etc.), active labor market programs, unemployment protection, housing, other social policy areas. All information is derived from the OECD Social Expenditure Statistics. (f) Employment protection legislation (EPL) for fixed and part-time is the evaluation of the individual regulations and collective dismissal of regular and temporary contract workers. These indicators were derived from the OECD Directorate for Employment, Labour, and Social Affairs database.

The individual level, age, gender, individual and parental education level, degree of social trust, and immigration status were selected. (a) Gender: a dummy variable coded with female as the reference group; (b) age: two dummy variables coded with the age group of 15 to 19 as the reference group; (c) individual and parental educational attainment level: two sets of dummy variables of upper-secondary education, and college or beyond with lower-secondary or below as the reference group was created based on the classification of the International Standard Classification of Education (ISCED); (d) social trust: a continuous measure of how much the respondents disagreed with the statement “there are only a few people you can trust completely” ranging from 1 to 5; (e) immigration status; a dummy coded variable with non-immigrant family as the reference group. The specific descriptive statistics for all variables by country are displayed in Appendix A.

Analytic Plan

The following model building was carried out in steps consistent with research questions to address the research objective, whether the increase in public social spending reduces NEET risk. Using logistic regression, first, we predicted parameter estimates of the relationship between the total public social spending as a percentage of GDP and NEET risk after adjusting for individual and institutional covariates across 15 European countries. Then, in the specified model, we incorporated public social expenditure on education, labor market, and unemployment to identify how the specific type of public social spending contributes to lowering NEET risk. Lastly, retaining the model, we specified the interaction terms between each type of public spending and individual background characteristics: (1) public social spending × individual education level, (2) public social spending × parental educational attainment, (3) public social spending × immigrant status, and (4) public social spending × gender to investigate the moderating impact of public social spending in reducing the impact of risk factors. Alongside this, we further analyzed the interactions between public expenditure on education, labor market, and unemployment and each individual variable. This set of analyses enables us to answer the question about the distributional impact of public social spending. It suggests who may benefit the most given the expansion of the welfare state.

The use of replication methods is critical given its ability to reflect the differential likelihood of sample selection at each stage and thus to make a valid statistical inference. Moreover, BRR adjusts for clustering error caused by the hierarchical data structure, non-response, and post-stratification, among others. Following the recommendation for the proper use of PIAAC data, our model specifications were adjusted by the Balanced Repeated Replication (BRR) weighting method using 80 replicate sampling variables (Kirsch & Thorn, 2013). The replicate method involves creating multiple replicate subsamples, which reflect the design of the full sample. Estimates are then calculated for each of the subsamples and the full sample. The final estimate is produced as the sum of squared deviations between each subsample and the full sample.

Results

Descriptive Analysis

Figure 1 displays a descriptive portrait of differences in NEET rate across the 15 European countries. In this study, the average NEET rate across 15 European countries was 10%, and the median was 5.9%. The NEET rate variation ranged from the lowest of 2.2% in the Netherlands to the highest of 16.8% in the Slovak Republic. Interestingly, all Scandinavian countries known for a relatively strong social welfare system indicated a below-average NEET rate, indicating 3.94%. All other European countries showed an above-average level of 12.16%, which is about three times higher than Scandinavian countries.

Figure 1.

The proportion of NEET among 15 to 29 years old youths across 15 countries.

It is noteworthy that there is nearly an eight percentage point gap between our result from PIAAC and the findings suggested by the OECD report (2014). This difference can be attributed to the varying definitions of NEET status adopted in the PIAAC data and OECD survey. The OECD report identifies NEET as when a youth does not participate in education, employment, or training for more than 4 weeks. In contrast, PIAAC identified NEET status based on those who have not experienced education, labor, and training during the past year (Tamesberger & Bacher, 2014).

The Relationship Between Public Social Spending and NEET Risk

In this section, we present the logistic regression results that reveal how public social expenditure is associated with the probability that the youths belong to the NEET group as a function of individual and institutional characteristics. We fitted a logistic regression model that nominated non-NEET as a reference group and calculated log odds for NEET. In Table 1, the association between the extent to which public social spending, including spending on education, labor, unemployment, and the Intergenerational inequality index, are associated with the NEET rate in the Full and Specified models.

Table 1.

Parameter Estimates Predicting the Association Between Public Social Spending and NEET Rate After Controlling for Individual and Institutional Characteristics Across 16 European Countries.

Variables	Total model		Specified model
Variables	Beta	O.R.^a	Beta	O.R.
Country-level variable
Public social spending % of GDP (PSS)	−0.04*** (0.00)	0.96
Public spending on education			−0.44*** (0.04)	0.64
Public spending on the labor market			−0.12*** (0.01)	0.88
Public spending on unemployment			−0.36*** (0.02)	0.70
Intergenerational inequality index	0.12*** (0.01)	1.13	0.01^† (0.01)	1.01
EPL^b for regular job	−0.57*** (0.02)	0.57	−0.70*** (0.04)	0.50
EPL for part-time job	0.18*** (0.02)	1.20	0.35*** (0.02)	1.42
GDP (log)	0.81*** (0.07)	2.26	0.60*** (0.07)	1.81
Population 15–29 (log)	0.11*** (0.01)	1.12	0.02* (0.01)	1.02
Unemployment rate	−0.02*** (0.00)	0.98	0.03*** (0.00)	1.03
Individual-level variable
Individual Education (Reference = Lower secondary or below)
Upper secondary	−0.69*** (0.02)	0.50	−0.71*** (0.02)	0.49
Tertiary	−1.72*** (0.03)	0.18	−1.73*** (0.03)	0.18
Parental Education (Reference = Lower secondary or below)
Upper secondary	−0.42*** (0.02)	0.66	−0.48*** (0.02)	0.62
Tertiary	−0.97*** (0.03)	0.38	−1.00*** (0.03)	0.37
Age Group (reference=15-19 years old)
20–24 years old	2.12*** (0.04)	8.33	2.11*** (0.04)	8.29
25–29 years old	2.81*** (0.04)	16.56	2.80*** (0.04)	16.49
Female	0.56*** (0.02)	1.76	0.56*** (0.02)	1.76
Paid Work Experience (reference = yes)	1.12*** (0.02)	3.07	1.11*** (0.02)	3.03
Immigrant Status of Parent (reference = no)	−0.11*** (0.02)	0.90	−0.09*** (0.02)	0.91
Social Trust	−0.21*** (0.01)	0.81	−0.20*** (0.01)	0.82
Book possessions (reference: 0–10)
11–25	−0.40*** (0.03)	0.67	−0.41*** (0.03)	0.66
26–100	−0.45*** (0.03)	0.63	−0.46*** (0.03)	0.63
101–200	−0.76*** (0.03)	0.47	−0.77*** (0.03)	0.46
202–500	−1.12*** (0.04)	0.33	−1.12*** (0.04)	0.32
More than 500	−0.74*** (0.05)	0.48	−0.74*** (0.05)	0.48
Constant	−11.45*** (0.64)	0.00	−6.74*** (0.73)	0.00

Note. ^aOdds ratio.

Employment Protection Legislation.

***

p < .001. **p < .01. *p < .05. ^†p < .10. All values are weighted by Balanced Repeated Replication method.

As the Full model in Table 1 displays, at the individual level, females (0.56, p < .001) were more likely to be NEET at young adult age, whereas NEETs demonstrated a lower educational attainment level, lower social trust, and a higher ratio of a family from an immigrant background. Besides, the older youth cohort of 20 to 24 (2.12, p < .001) and 25 to 29 (2.81, p < .001) showed a significantly higher chance of being NEET than the younger age cohort groups. The family background characteristics were also significantly associated with the likelihood of being NEET, indicating that having more books at home or coming from a family with a higher parental education level led youth to be less likely to be NEET as young adults. In addition, those who had paid work experience also tend to have a lower risk of becoming NEET.

At the institutional level, Institutional characteristics such as national GDP (0.81, p < .001), population (age 15–29) (0.11, p < .001), EPL for fixed-job (−0.057, p < .001), and part-time job (0.18, p < .001), and unemployment percentage (−0.02 < 0.001) appeared to be associated with NEET risk across European countries. Moreover, the intergenerational inequality also tends to increase the risk of being NEET (Exp(0.12) = 1.13, p < .001), suggesting that countries with elderly biased welfare regimes are more likely to increase the risk of becoming NEET.

After taking into account these individual and institutional covariates, our result showed that the expansion of government social expenditure as a percentage of GDP was significantly related to a decrease (Exp(−0.04) = 0.96, p < .001) in the chances of being NEET across European countries. A similar pattern of findings was found when we specified that public expenditure on education, labor, and unemployment explain NEET variance across countries. As it can be seen in the specified model, our results revealed that countries with a higher level of spending on education (Exp(−0.44) = 0.64, p < .001), expenditures on labor (Exp(−0.12) = 0.88, p < .001), and unemployment (Exp(−0.36) = 0.70, p < .001) are likely to lower the risk of being NEET.

Figure 2 illustrates the relationship between government social spending and the risk of becoming NEET, reported in the Full Model in Table 1. The Unconditional model depicts the relationship between public social spending and the national average NEET rate without any covariates. The Conditional model illustrates the association between public social expenditure and the predicted NEET rate after adjusting for the national and individual characteristics. In this model, the Y-axis reflects the probabilities of being NEET, assuming individual and institutional factors are held at their means. The X-axis is public social spending relative to the national GDP. As is depicted in Figure 2, in the Conditional model, countries that spend a more significant proportion of public social expenditure appeared to demonstrate a lower NEET rate, which suggests that a social protection system may substantially decrease the risk of being NEET for young people in European countries.

Figure 2.

The association between public social spending and predicted probability to be NEET holding all covariates at their means.

The Interaction Between Parent Education and Public Social Spending on NEET Risk

This section provides the empirical evidence related to the third set of research questions, whether public social spending may have a moderating impact on NEET risk depending on individual risk factors. To this end, we specified a series of cross-level interaction models (1) each social spending × individual education level, (2) each social spending × parental education attainment, (3) each social spending × immigrant status, and (4) each social spending × gender, to determine the moderating impact of public social spending on reducing NEET rate after holding individual, and country characteristics constant. These results are presented in Table 2, retaining the models shown in Table 1.

Table 2.

Parameter Estimates Predicting the Role of Public Social Spending (PSS) in Moderating the Impact of Risk Factors After Controlling for Individual and Institutional Characteristics Across 15 European Countries.

Interaction term	Total public social spending		Spending on education		Spending on the labor market		Spending on unemployment
	Beta	O.R.^a	Beta	O.R.	Beta	O.R.	Beta	O.R.
× Individual education
× Upper secondary	−.00 (.00)	1.00	−.15*** (.04)	.86	−.06** (.02)	.94	−.13*** (.02)	.88
× Post-secondary	.02** (.01)	1.02	−.24*** (.06)	.79	.05** (.02)	1.06	−.00 (.02)	1.00
× Parental education
× Upper secondary	−.02*** (.00)	.98	−.22*** (.04)	.81	.13*** (.02)	1.14	.16*** (.02)	1.17
× Post secondary	.02** (.01)	1.02	−.22*** (.06)	.81	.28*** (.02)	1.32	.33*** (.02)	1.40
× Immigrant status	.02*** (.01)	1.02	−.13** (.04)	.88	.27*** (.02)	1.31	.30*** (.02)	1.35
× Female	−.07*** (.00)	.93	.16*** (.03)	1.17	−.36*** (.01)	.70	−.34*** (.02)	.71

Note. ***p < .001. **p < .01, *p < .05. The results of other variables are omitted. Standard errors are adjusted by Balanced Repeated Replication (BRR) method.

Odds ratio.

In Figure 3, we visualize the predicted probability of being NEET by the individual educational level, parental educational attainment, immigration status, and gender relative to public social spending. All covariates are held to their mean values. As shown in Table 2 and Figure 3, our findings suggest the moderating impact of public social expenditure on reducing the social exclusion of young people with a low level of education and coming from parents with low educational levels. Specifically, our interaction estimates of parental educational level and public social spending revealed that the expansion of public social expenditures may reduce the NEET risk of young people whose parents have a lower secondary educational level (−0.02, p < .001) and upper secondary educational attainment (0.02, p < .001) compared to those young people coming from highly educated parents. Moreover, the interaction term of individual educational level and total public social spending was also partly significant (0.02, p < .001). Our interaction model of public social expenditure and females also indicated a negative relationship (−0.07, p < .001), suggesting that the females were more likely to benefit by incrementing public social spending. Although the effect size was not very large in the case of immigrant status, our finding showed that the increase in public social expenditure is significantly associated with reducing the NEET risk of non-immigrant young people (0.02, p < .001) more than those of immigrant families.

Figure 3.

The predicted probabilities of being NEET by the parental education attainment level, immigrant status, and gender given the increase of total and each specified social spending holding covariates at their means.

As is clearly illustrated in Figure 3, a similar pattern of moderating relationships was reported when the interaction terms between public expenditure on education and individual characteristics were added to the model specification. Specifically, the expansion of public social spending on education, labor market, and unemployment benefits attenuates the risk of becoming NEET given low educational attainment level, females, and coming from a non-immigrant family with low educated parents. These findings suggest that young people from disadvantaged backgrounds are more likely to benefit given the increased expenditure on education and the labor market, which supports the existing theory that the redistributive structural intervention may compensate for their disadvantaged background by attenuating their risk of being social exclusion.

Discussion

Using PIAAC combined with international data sets, this study aimed to reveal whether public social spending moderates NEET risk, especially for young people from socially disadvantaged backgrounds. Unlike previous studies that have mainly focused on identifying the individual and institutional risk factors, this study aimed to provide empirical evidence on whether the increase in public social spending would significantly reduce the risk of NEET and benefit socially disadvantaged groups of young people. Our main findings suggest that the increase in total public social expenditures as a percentage of GDP is significantly related to reducing the likelihood of being NEET, holding individual and institutional characteristics constant. This result was consistent when we specified social expenditure on education, labor market, and unemployment, indicating a significant decrease in NEET risk independent from national GDP, welfare directionality, and employment protection legislation. Moreover, public expenditure on school-to-work transition and labor market appeared to attenuate the risk of being NEET given disadvantaged background characteristics and benefit young people with low educational level, whose parents have low educational attainment, and females. Nevertheless, the magnitude of effect sizes was not noticeably different between different dimensions of public spending.

This pattern of findings confirms previous literature that suggested countries with a well-developed welfare system, with higher expenditure on education and labor market, present a lower share of the NEET population by providing young people with necessary opportunity and supports for social integration in education, training, and labor market (Assmann & Broschinski, 2021; Carcillo et al., 2015). Indeed, the empirical evidence that building a more robust social protection system may help young people engage in education labor is consistent with the OECD report (2016) that discussed a more comprehensive social protection system that can widely shield people from risk and vulnerability may ultimately benefit young people in challenging circumstances. From this perspective, our findings provide the policy implication that expanding the redistributive social protection system may be an effective investment in reducing the exclusion of new generations (Thompson, 2011).

Our findings also align with existing literature that stresses the importance of expenditure on education for young people under unfavorable conditions. According to previous studies, the quality of the educational system determines the competency and skills of young people and influences early school dropout, which is a critical factor that leads to NEET. In particular, countries that provide more comprehensive educational subsidies and supports may benefit socially disadvantaged young people to stay longer in school, prevent early school dropout, and acquire better qualifications for a successful labor market entry. In addition, public spending on the active labor market may also help a smooth transition from school to work for vulnerable groups of youths since they often encounter limited access to the information and opportunities to participate in training. Active labor market policies, in this regard, may also serve as important resources for job search assistance, training programs, or coaching facilitating to acquire of job-related skills that can promote labor market entry. Similarly, the government support may be essential for young workers left without work to have the subsidies and resources to re-educate themselves or reintegrate into education that may facilitate their re-entry into the labor market.

Other noteworthy findings are the extent to which employment protection is associated with NEET risk across European countries. Concerning the flexibility of the labor market, our findings confirm previous studies that showed the deregulation of part-time jobs might stimulate the young population into the labor market. In contrast, employment protection of regular job functions lowers the NEET rate (Eurofound, 2012). This pattern of results leads to the policy implication that lowering the labor market entry barrier may be an effective approach to inducing young people into the labor market. However, such flexibility of the part-time regular labor market may need to be accompanied by the institutional guarantee with the prospective conversion into regular jobs shortly with robust employment protection. While the complementary labor market protection system between regular and part-time employment may prevent employers from capitalizing on young people for a pure cost reduction, it may ease young people’s labor market entry and provide the necessary security to stay in the labor market.

Caution is required, however, in overestimating the role of the social safety net, expecting all young people may be equally better off from the expansion of social spending. As our findings have shown, young people from immigrant families may not equally benefit from the social protection system. This pattern of results is consistent with previous studies that suggested in countries with a limited public social spending, the most marginalized group of people are often excluded from the social benefits, despite the average population being better off given the social protection system (Barrett & McPeak, 2006; LeGrand, 1982; Tullock, 1983). Gaentzsch (2018), in this vein, discussed that unless allocations to unprivileged young people are extended meaningfully in volume and coverage, public social spending may be limited in its ability to tackle this vulnerability.

Conclusion

From this view, this study stresses the importance of the total public social expenditures alongside the social spending in specific spheres that can reach those most needed groups of young people and their families, which suggests that a larger redistributive fiscal budget and greater effort to reach people and families that have so far been excluded are necessary to reduce the unequal distribution of public social spending within countries. This study, however, does not undermine the importance of youth-targeted intervention programs that can promote the self-sufficiency of young people. In particular, intervention programs specifically designed to serve youth in different living circumstances are needed to reach far excluded populations. For example, early school leavers or young people from immigrant families may lack social resources such as lack of information channels or access to an adequate program. Thus they might encounter multiple barriers that prevent them from benefiting from the social support system. Intervention programs that can specifically tackle such underprivileged groups and efforts to find ways to effectively deliver such programs, alongside the more robust social protection system, may be critical in mitigating NEET risk across countries.

Several limitations of this study should be noted. First, our explanatory indicators were limited to public social spending directly related to the labor market and school-to-work transition, leaving behind some relevant indicators such as family benefits spending, and qualitative welfare system. Our approach was partly given that the primary interest of this study lies in addressing the distributional impact of public social expenditure on socially disadvantaged groups of young people. As a result, for analytical purposes, the scope of main indicators needed to be restricted to those indexes that can directly influence school-to-work transition. If any, however, future study studies may extend this line of inquiry by assessing another type of public social spending dedicated to young people relative to its effect on school-to-work public social spending. In addition, future studies should consider other alternative datasets of the welfare states, such as the Comparative Welfare Entitlements Dataset (CWED2), which measured welfare approaches and structures in various aspects, including qualifications for subsidies, benefit periods, administrative time, etc. While CWED2 does not provide data from several Eastern European countries that are included in our analysis, future studies may extend the inquiry by analyzing such nuanced data.

Second, this study cannot rule out the endogenous issue associated with social expenditure and the youth population that is not fully accounted for in our statistical method. Thus, the design of this study is correlational, which is limited to drawing causal inferences. However, this study is intended to provide insights into the welfare state’s role in preventing the social exclusion of youth and may be a preliminary step that lays the foundation for more rigorous research that addresses the effective prevention program’s causality.

Despite these limitations, using PIAAC with the international nation information, the current study aimed to explore whether the public social expenditure is associated with the NEET population and whether such effects may vary among different groups of young people. This study suggests that a nation’s public social spending is highly associated with the probability of that nation’s youth becoming NEET. In particular, the expenditure on the education system, active labor market, and unemployment appeared to moderate the NEET risk of socially disadvantaged groups of young people. Based on these findings, we discussed that strengthening the social safety net by expanding public social spending may be an effective structural intervention that may promote the social integration of youths living under challenging circumstances.

Footnotes

Appendix

Appendix A.

The mean values of variables by country.

Country	NEET	Individual education level(reference = Lower-secondary or below)		Parental education level (reference = Lower-secondary or below)		Years old (reference:15–19)		Female	work experience (1 = yes)	Immigrated parents (1 = yes)	Social trust	Book possessions (reference:0–10)					GDP (log)	Population (log)	Unemployment rate	Social expenditure	Spending on education	Spending on the labor market	Spending on unemployment	Intergenerational inequality index	EPL^a for regular job	EPL for a part-time job
Country	NEET	Upper secondary	Tertiary	Upper secondary	Tertiary	20–24	25–29	Female	work experience (1 = yes)	Immigrated parents (1 = yes)	Social trust	11–25	26–100	101–200	202–500	More than 500	GDP (log)	Population (log)	Unemployment rate	Social expenditure	Spending on education	Spending on the labor market	Spending on unemployment	Intergenerational inequality index	EPL^a for regular job	EPL for a part-time job
Austria	0.04	0.45	0.24	0.60	0.28	0.35	0.38	0.49	0.17	0.26	2.25	0.12	0.34	0.18	0.16	0.11	1.69	15.94	46.60	26.70	3.49	0.81	0.90	3.21	2.29	1.31
Belgium	0.04	0.45	0.31	0.42	0.45	0.34	0.34	0.51	0.37	0.17	2.24	0.19	0.28	0.13	0.10	0.05	1.63	16.22	74.80	28.90	4.13	2.78	3.50	1.59	1.88	2.25
Czech Republic	0.08	0.54	0.16	0.74	0.23	0.37	0.35	0.48	0.32	0.08	1.95	0.08	0.34	0.24	0.22	0.09	1.26	16.17	22.30	19.60	2.45	0.52	0.65	3.13	3.38	1.31
Denmark	0.03	0.38	0.18	0.39	0.49	0.34	0.32	0.50	0.08	0.22	2.77	0.11	0.28	0.21	0.21	0.11	1.68	15.53	145.10	29.80	4.13	3.61	0.00	1.91	1.53	1.38
Finland	0.04	0.51	0.17	0.51	0.40	0.35	0.35	0.49	0.16	0.08	2.66	0.09	0.33	0.24	0.22	0.10	1.60	15.50	92.20	27.00	3.86	2.38	1.63	2.10	2.08	1.56
France	0.10	0.47	0.27	0.48	0.33	0.33	0.35	0.51	0.35	0.25	1.81	0.15	0.31	0.19	0.16	0.08	1.53	17.96	44.90	3.80	3.47	2.76	1.50	2.23	2.50	3.13
Germany	0.04	0.41	0.23	0.52	0.42	0.33	0.37	0.48	0.21	0.26	2.40	0.09	0.31	0.22	0.19	0.12	1.65	18.22	58.80	24.70	2.79	1.76	1.13	2.55	2.60	1.00
Ireland	0.11	0.32	0.43	0.38	0.36	0.32	0.37	0.52	0.24	0.32	2.14	0.17	0.32	0.18	0.13	0.06	10.71	15.34	129.70	23.00	4.16	3.41	2.80	1.02	1.10	0.63
Italy	0.15	0.41	0.12	0.43	0.14	0.28	0.37	0.47	0.46	0.14	1.90	0.20	0.29	0.17	0.13	0.06	10.48	17.92	59.20	26.60	2.82	1.64	0.79	6.25	3.02	2.00
Netherlands	0.02	0.47	0.20	0.32	0.40	0.36	0.34	0.51	0.12	0.22	2.55	0.12	0.27	0.18	0.20	0.12	10.74	16.63	163.30	17.50	3.19	2.36	1.26	1.72	3.24	0.94
Norway	0.03	0.35	0.24	0.39	0.53	0.34	0.34	0.49	0.13	0.21	2.44	0.07	0.23	0.24	0.24	0.18	11.05	15.42	124.80	21.60	4.74	0.96	0.41	2.26	2.33	3.00
Poland	0.10	0.44	0.32	0.72	0.23	0.35	0.42	0.49	0.29	0.01	2.39	0.15	0.39	0.19	0.15	0.06	10.00	17.47	6.00	19.60	3.16	0.70	0.23	6.10	2.33	1.63
Slovak Republic	0.17	0.53	0.17	0.68	0.22	0.35	0.39	0.49	0.41	0.05	2.02	0.15	0.37	0.22	0.12	0.04	10.13	15.50	27.50	17.10	2.38	0.77	0.54	3.07	2.89	1.75
Spain	0.12	0.29	0.26	0.28	0.23	0.31	0.45	0.50	0.31	0.17	2.32	0.16	0.35	0.19	0.14	0.09	10.39	17.65	9.80	23.50	2.90	3.79	3.36	2.13	2.21	2.56
United Kingdom	0.10	0.50	0.32	0.50	0.35	0.37	0.39	0.50	0.17	0.29	2.25	0.16	0.31	0.19	0.13	0.11	10.51	17.96	108.70	23.10	3.72	0.53	0.39	1.37	1.51	0.38

Note. All values are weighted by Balanced Repeated Replication method.

Employment Protection Legislation.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Minjong Youn

Chungseo Kang

References

Alfieri

Sironi

Marta

Rosina

Marzana

(2015). Young Italian NEETs (not in employment, education, or training) and the influence of their family background. Europe’s Journal of Psychology, 11, 311–322.

Assmann

M. L.

Broschinski

(2021). Mapping young NEETs across Europe: Exploring the institutional configurations promoting youth disengagement from education and employment. Journal of Applied Youth Studies, 4, 95–117.

Barrett

C. B.

McPeak

J. G.

(2006). Poverty traps and safety nets. In de Janvry

Kanbur

(Eds.), Poverty, inequality and development. Economic studies in inequality, social exclusion and well-being (Vol. 1, pp. 131–154). Springer.

Breen

(2005). Explaining cross-national variation in youth unemployment. European Sociological Review, 21(2), 125–134.

Bynner

Parsons

(2002). Social exclusion and the transition from school to work: The case of young people not in education, employment, or training (NEET). Journal of Vocational Behavior, 60(2), 289–309. https://doi.org/10.1006/jvbe.2001.1868

Carcillo

Fernandez

Konigs

Minea

(2015). NEET youth in the aftermath of the crisis: Challenges and policies (OECD Social, Employment, and Migration Working Papers, No. 164). OECD Publishing. https://doi.org/10.1787/5js6363503f6-en

Carcillo

Fron

Hyee

Keane

Konigs

Ladaique

(2016). Society at a glance 2016: OECD social indicators: A spotlight on youth. Organisation for Economic Co-operation and Development. https://www.oecd-ilibrary.org/social-issues-migration-health/society-at-a-glance-2016/the-neet-challenge-what-can-be-done-for-jobless-and-disengaged-youth_soc_glance-2016-4-en

Caroleo

F. E.

Rocca

Mazzocchi

Quintano

(2020). Being NEET in Europe before and after the economic crisis: An analysis of the micro and Macro determinants. Social Indicators Research, 149, 991–1024.

Comelli

(2021). The impact of welfare on household debt. Sociological Spectrum, 41(2), 154–176.

10.

Eurofound. (2012). NEETs – Young people not in employment, education or training: Characteristics, costs and policy responses in Europe. Publications Office of the European Union.

11.

Eurostat. (2012). Eurostat statistics. Eurostat.

12.

Eurostat. (2019). Young people neither in employment nor in education and training (NEET). Author.

13.

Gaentzsch

(2018). The distributional impact of social spending in Peru (discussion paper no. 7). School of Business and Economics, Freie Universität. Berlin.

14.

Gangl

(2002). Changing labour markets and early career outcomes: Labour market entry in Europe over the past decade. Work Employment & Society, 16, 67–90.

15.

Genda

(2007). Jobless youths and the NEET problem in Japan. Social Science Japan Journal, 10(1), 23–40.

16.

Goodin

R. E.

LeGrand

(1987). Not only the poor: The middle classes and the welfare state. Allen and Unwin.

17.

ILO. (2019). World employment social outlook 2019: Trends 2019. Author.

18.

Kelly

McGuinness

(2015). Impact of the Great Recession on unemployed and NEET individuals’ labour market transitions in Ireland. Economic Systems, 39(1), 59–71.

19.

Kevelson

M. J.

Marconi

Millett

Zhelyazkova

(2020). College-educated yet disconnected: Exploring disconnection from education and employment in OECD countries, with a comparative focus on the U.S. PERC Report and ETS Research Report Series.

20.

Kirsch

Thorn

(Eds.) (2013). Technical report of the Survey of Adult Skills (PIAAC). OECD Publishing.

21.

Korpi

Levin

(2001). Precarious footing: Temporary employment as a stepping stone out of unemployment in Sweden. Work, Employment and Society, 15(1), 127–148.

22.

LeGrand

(1982). The strategy of equality. Redistribution and the Social Services. George Allen and Unwin.

23.

Xie

Zou

H. F.

(2000). Dynamics of income distribution. Canadian Journal of Economics, 33, 937–961.

24.

Lindert

(2003). Why the welfare state looks like a free lunch (NBER Working Paper). http://www.nber.org/papers/w9869

25.

Lynch

(2006). Age in the welfare: The origins of social spending on pensioners, workers, and children. Cambridge University Press.

26.

Marques

Hörisch

(2020). Understanding massive youth unemployment during the E.U. Sovereign debt crisis: A configurational study. Comparative European Politics, 18, 233–255. https://doi.org/10.1057/s41295-019-00184-3

27.

Mauro

J. A.

Mitra

(2015). Understanding out-of-work and out-of-school youth in Europe and Central Asia (World Bank Other Operational Studies 22806). The World Bank.

28.

Millett

C. M.

Kevelson

M. J. C.

(2018). Doesn’t get better with age: Predicting millennials’ disconnection (Research Report Series No. RR-18-42). Educational Testing Service. https://doi.org/10.1002/ets2.12219

29.

Müller

(2005). Education and youth integration into European labour markets. International Journal of Comparative Sociology, 46, 461–485.

30.

Organisation for Economic Co-operation and Development. (2008). Off to a good start? Youth labour market transitions in OECD countries. Employment Outlook, 26, 25–78. https://www.oecd.org/els/offtoagoodstartjobsforyouth.htm.

31.

Organisation for Economic Co-operation and Development. (2014). OECD employment outlook 2014. OECD Publishing.

32.

Organisation for Economic Co-operation and Development. (2016). Society at a glance: OECD social indicators. https://www.oecd-ilibrary.org/

33.

Organisation for Economic Co-operation and Development. (2020). Social expenditure (SOCX) update 2020: Social spending makes up 20% of OECD GDP. http://www.oecd.org/social/expenditure.htm.

34.

Pastore

Zimmermann

K. F.

(2019). Understanding school-to-work transitions. International Journal of Manpower, 40(3), 374–378.

35.

Pemberton

(2008). Tackling the NEET generation and the ability of policy to generate a ‘NEET’ solution—Evidence from the UK. Environment and Planning A, 26(1), 243–259.

36.

Piopiunik

Ryan

(2012). Improving the transition between education/training and the labour market: What can we learn from various national approaches? Report for the European Commission. RENEE. Analytical Report 13.

37.

Pitkänen

Remes

Moustgaard

Martikainen

(2021). Parental socioeconomic resources and adverse childhood experiences as predictors of not in education, employment, or training: A Finnish register-based longitudinal study. Journal of Youth Studies, 24(1), 1–18.

38.

Rennison

Maguire

Middleton

Ashworth

(2005). Young people not in education, employment or training: Evidence from the Education Maintenance Allowance Pilots Database. DfES Publications.

39.

Rodriguez-Modroño

(2019). Youth unemployment, NEETs and structural inequality in Spain. International Journal of Manpower, 40(3), 433–448.

40.

Russell

O’Connell

P. J.

(2001). Getting a job in Europe: The transition from unemployment to work among young people in nine European countries. Work Employment & Society, 15(1), 1–24.

41.

Tamesberger

Bacher

(2014). NEET youth in Austria: A typology including socio-demography, labour market behaviour and permanence. Journal of Youth Studies, 17(9), 1239–1259.

42.

Thompson

(2011). Individualisation and social exclusion: The case of young people not in education, employment or training. Oxford Review of Education, 37(6), 785–802.

43.

Tullock

(1983). Economics of income redistribution. Kluwer-Nijhoff Publishing.

44.

Ulu

(2018). The effect of government social spending on income inequality in OECD: A panel data analysis. Uluslararası Ekonomi Siyaset İnsan ve Toplum Bilimleri Dergisi, 1(3), 184–202.

45.

Wolbers

M. H. J.

(2007). Patterns of labour market entry: A comparative perspective on school-to-work transitions in 11 European countries. Acta Sociologica, 50, 189–210.

46.

Zwiers

Koster

(2015). The local structure of the welfare state: Uneven effects of social spending on poverty within countries. Urban Studies, 52(1), 87–102.