SUBJECTIVE WELL-BEING OF NEET BASED ON GENDER AND GENERATION SURVEY IN MOLDOVA

DOI: https://doi.org/10.36004/nier.es.2023.2-07

JEL Classification: C25, I30, J13, O52

UDC: 316.346.32-053.6(478)

Maxim SLAV

Scientific Researcher, National Institute for Economic Research, Academy of Economic Studies of Moldova

https://orcid.org/0000-0002-3061-1692

e-mail: [email protected]

Received 7 June 2023

Accepted for publication 29 October 2023

SUMMARY

NEET is an umbrella term for those who struggle with their school-to-work transition. It is a consensus today that NEET is an acutely heterogeneous group, and there are varying ways to classify its members. Some of them deliberately reject or postpone the school-to-work transition, for example, because they choose to fulfill their traditional gender role. In this article, the NEET classification was based on the individuals' well-being. Moldova suffers from simultaneous socio-economic crises, including high emigration and poverty rates. A standardized survey taken in Moldova thus would allow for greater flexibility when classifying NEET youth. The data were taken from the Gender and Generation Survey (GGS) conducted in 2020. The previous studies use different definitions of NEET; for this research, the 15 – 29 age group was chosen (298 observations comply with the NEET conditions and are complete). To conduct the classification, the author chose the questions on life satisfaction, happiness, and two depression sentiments. The variables are categorical and were analyzed with a Latent Class Analysis. There are substantial correspondences between the NEET well-being classes and the causes of NEET, as well as the SWB determinants. The results supplement the current knowledge on NEET heterogeneity from the side of self-assigned well-being rather than socio-economic inequalities.

Keywords: NEET, well-being, youth, Moldova, latent class analysis

INTRODUCTION

NEET is defined here as a group of people aged 15 – 29 who were not in employment, education, or training in the four weeks by the time of the poll. The concept attempts to highlight the most vulnerable part of the youth using the school-to-work transition framework (Contini et al., 2019). Since it is defined through negations, as Yates and Payne (2006) note, NEET was predetermined to unite various groups. The studies that followed the concept introduction showed various ways to dissect the group by socioeconomic status (e.g., UCW, 2013; Tamesberger & Bacher, 2014). Several research groups focused on the frequency and duration of the NEET state (Contini et al., 2019; Tamesberger & Bacher, 2014) NEET and the intra-NEET inequalities were studied in Moldova before. In their deep overview, Crismaru et al. (2017) took the differences in skills and education as the base of the classification.

Intuitively, the NEET status should be associated with subjective well-being (SWB). Those with a greater propensity to become and stay unemployed are likely to have an unprivileged position that unites other predictors and correlates of low SWB (i.e., low savings and consumption levels, low probability to get a high-paying job, etc. (Carver & Grimes, 2019)). At the same time, stay-at-home mothers are also part of NEET, and for some women, it is a desirable gendered position (Enneli & Enneli, 2017); fulfilling a gender role implies better life satisfaction and overall SWB. The causality in the relationship between SWB and NEET is generally hard to determine because poor SWB and psychological diseases are predictors of the acquisition of NEET status (Bonanomi & Rosina, 2022)

Thus, hypothetically, the NEET SWB classes should be associated with the socio-economic groups and have an ambiguous relationship with the class (i.e., primarily, with income and consumption) due to some additional factors. This study attempts to capture this ambiguity. Previously, researchers classified NEET via the causes of NEET and then studied the resulting clusters’ SWB. In this paper, the NEET is classified by SWB, and then the profiles of these SWB classes are described. While, to my knowledge, this is the first attempt to cluster NEET this way, classification of individuals by their SWB indicators was done before. The chief utility of this method is the differences between the SWB indicators. For example, unhappy and life-satisfied individuals differ from happy and unsatisfied ones. The classification is done via Latent Class Analysis based on the Gender and Generation Survey (GGS, MH, et al., 2020) conducted in Moldova in 2020. While the distribution of the classes throughout the pandemic year was unequal, the influence of the pandemic stage is counterintuitive (e.g., a widespread positive affect in the middle of the pandemic and a widespread negative one before it).

The rest of the paper is organized as follows. In the literature review, I overview the research on NEET socio-economic and SWB heterogeneity. I also cover the differences between the SWB indicators including life satisfaction, happiness, and depression symptoms. Next, I describe the data I use as well as the method, the Latent Class Analysis (LCA). Since I do not have hypotheses in this article and rather attempt to describe the individuals, the optimal number of classes is determined in this study via information criteria and cut-off points. Finally, I discuss the results: I describe the classes by their well-being and socio-economic position.

LITERATURE REVIEW

Socio-economic heterogeneities of NEET

Due to the general measurement, the group must contain those in natural unemployment and those unemployed long-term. The classical definition of NEET is based on the period of four weeks, but Bynner and Parsons (2002) used a definition of NEET based on six months rather than four weeks. Contini et al. (2019) define different NEET subgroups based on the duration of their status (e.g., up to three months, up to eleven months, etc.). There must also be the absence of some vulnerable youth (i.e., those that NEET is supposed to describe) with a low-paying precarious job.

Even with the problems of definition put aside, one can find heterogeneities inside the group. One of the main axes of such differentiation is the easiness of exit NEET and the factors that contribute to it. These include the education and skills of the individuals (Crismaru et al., 2017; De Luca et al., 2020) or their parents (Alfieri et al., 2015), the labor market status (Gagauz & Crismaru, 2024), age (Caroleo et al., 2020), and region (Özdemir et al., 2023). Universal rules are yet to be found, as the patterns differ from country to country and region to region. For example, women are more likely to be NEET in the South of Italy compared to the North (Contini et al., 2019). Tamesberger and Bacher (2014) dissect the Austrian NEET by the causes of NEET. They conclude with a quite complicated seven-group classification that includes, besides the others, “young married women with a migration background.” The studies on Moldova do not mention this group, which I would explain with the high level of emigration rather than immigration (Gagauz & Crismaru, 2024).

As Crismaru et al. (2017) reported before, the rates of NEET in Moldova are inflated due to the official statistics not accounting for immigration. The researchers estimated the NEET rate (for those aged 15 – 29 years) to be around 28%¹, which is still very high by European standards. The reported rates are higher only for Turkey and Macedonia. Gagauz and Crismaru (2024) dissect NEET by two axes: women vs. men and unemployed NEET vs. inactive NEET. The inactive NEETs are primarily women living in rural areas and with at least one child. The reverse is true: 75% of women aged 15 – 34 are NEET, and around 75% of them are homemakers (i.e., inactive). Most of the unemployed NEET are men who live in Chisinau. 40% of NEET men aged 15 – 24 and 61% of NEET men aged 25 – 34 work abroad or intend to leave Moldova. Those with secondary education or parents without tertiary education are more likely to be inactive or unemployed NEET. Finally, getting transfers from abroad is also a predictor for these two statuses.

SWB and NEET

This article focuses on four SWB indicators: life satisfaction, happiness, and two depression sentiments. Both life satisfaction and affective well-being indicators are assessments of one’s life (Van Praag et al., 2003; Vladisavljević & Mentus, 2019). However, the former results from self-reflection and life evaluation, and the second is based on pleasant and unpleasant feelings (positive and negative affective well-being). The difference manifests in the predictors of each: “Factors such as income, education, or marital status predict life satisfaction, while time use predicts affective balance more strongly” (Vladisavljević & Mentus, 2019, p.39). Such differences are more complicated if the affect is evaluated retrospectively (e.g., “Did you feel happy today?”). According to Ganzach et al. (2023), retrospective evaluation of positive affect is more based on the heuristics, e.g., self-image, while retrospective evaluation is closer to the experienced affect. Personality traits, e.g., neuroticism, also affect SWB, but their considerations are out of the scope of this paper.

Typically, the SWB of NEET was contrasted with the SWB of non-NEET. On average, NEET is associated with lower happiness and life satisfaction (Bonanomi & Rosina, 2022; Felaco & Parola, 2022) and a higher probability of mental health problems (Bonanomi & Rosina, 2022; Goldman-Mellor et al., 2016). There are reasons for NEET status to deteriorate the SWB of individuals, including lower incomes and worse prospects. However, those with mental health problems in childhood and adolescence (ADHD, depression, etc.) do have significantly and substantially higher chances of becoming NEET (Goldman-Mellor et al., 2016), which suggests a causal loop.

If one accounts for socio-economic aspects alone, the well-being of NEET should be heterogeneous for several reasons. One of the reasons is the socio-economical NEET heterogeneity. The lack of employment (or enrollment in an education institution as a socially approved alternative to employment) is natural for some situations, e.g., for the younger part of the youth (Caroleo et al., 2020). Second, NEET is constructed to target specifically those who do not work or have no socially accepted substitutes to work (training or education). A great part of the association is non-meeting the expectations from oneself, disappointment, and marginalization (Crismaru et al., 2017). Some believe a young person must work or study to get a job. The correct position of a young person, one can claim, is to work or to study to get a job. Theoretically, such claims echo the concept of an ‘out of place’ position (Douglas, 2001) and lead individuals to fatalism (Thompson et al., 2018). Researchers on SWB of NEET note the present-fatalistic orientation of the NEET youth (Felaco & Parola, 2022).

The aspired position varies. For example, Enneli and Enneli (2017) studied young lower-class Turkish women who were discontent² with their low-income jobs and dreamed of being a traditional housewife (i.e., of a NEET position). There are traces of these divisions in Moldova. 80% of NEET women desire to have a happy family, and only 15% want to earn “a lot of money” (the figures are 40% and 48%, respectively, for men (Crismaru et al., 2017). Enneli and Enneli (2017) argue that two groups criticize the gender roles: the higher-educated women and the men who cannot fulfill the breadwinner role. Sobotka (2008) observed the same phenomenon in the post-communist countries. Low education levels are associated there with both traditional values and wide adoption of cohabitation and extra-marital childbearing. In other words, the lower class may strive towards an old family ideal with a hierarchy and fixed boundaries and be unable to afford it.

There is evidence that the NEET’s subjective well-being is affected the way mentioned above: the NEET women with children do not suffer from the same effects as the childless women and men (Bonanomi & Rosina, 2022).

Gender and Generation survey (MH, et al., 2020) was used to study the well-being of the NEET. NEET was formed from two questions. First, dem06, a question on employment status. The following replies were accepted as NEET: unemployed, homemaker, ill or disabled. Second, dem08a, a question on training in the past four weeks. As a result, 298 individuals remained in the sample.

To a limited extent, the survey allowed for following the methodology of the other well-being researcher (Bonanomi & Rosina, 2022). Specifically, the questions used were on satisfaction with life (wel01), happiness (wel08), and two depression sentiments: feeling depressed (wel11b) and feeling life was a failure (wel11c). The scale for the wel01 question was framed as follows: 0 means “absolutely not satisfied”, 10 is “absolutely satisfied”, and 5 is “neither satisfied nor unsatisfied” (wel08, the question on happiness, is scaled accordingly). For the analysis simplicity and due to the software limitations, the scales for these two variables were recoded into four- and five-level scales, respectively. Cronbach’s alphas equal 0.924 and 0.949, respectively. Since alphas greater than 0.9 are considered excellent, no substantial distortions due to the recoding are expected. Finally, the two depressive symptoms have four-point scales with “Never” being recoded as “1”, and “Most of the time” being recoded as “4”. Thus, greater values refer to stronger sentiments. The descriptive statistics for the [pre-recoded] variables used as well as for gender and age can be found in Table 1.

It was mentioned above that heterogeneities in NEET’s SWB must exist because of the socio-economic heterogeneities: there is a finite number of statuses that individuals in NEET have that affect their socio-economic position (Tamesberger & Bacher, 2014). Another reason to use clusterization is the multimodality of the SWB indicators (see Figure 1).

Table 1. Descriptive Statistics of the Independent Variables.

It is supposed that there are links between the above-mentioned variables, and hence the Latent Variable approach is used. It implies that the values of the variables are determined by an unknown variable. The previous research shows there are subgroups in the NEET group; the variables offered by GGS are categorical. This leads to the Latent Class model (LCA).

To report the results of LCA, I follow the guidelines of Weller et al. (2020) and Nylund-Gibson and Choi (2018). To run LCA, I use the poLCA package for R (Linzer & Lewis, 2011). When the optimal number of classes was determined, every model calculation was repeated 200 times with 5000 iterations for each model. The rules to establish the optimal number of classes can be divided into three groups. First, several indicators have cut-off points; if a model does not satisfy them, it is usually not considered. The first measure with such a criterion is entropy, the degree of separation between the classes, and the second one is the minimal posterior probability, the accuracy with which the model determines the class of an observation. For both, the desired value is 1, but the cut-off points vary being 0.6 for entropy and 0.8 for the posterior probability (Weller et al., 2020). Finally, generally, no class should contain less than 5% of a sample and less than 50 observations, although samples of 30 observations are sometimes enough for the whole LCA (Nylund-Gibson and Choi, 2018).

Second, the optimal number of classes is determined through information criteria. For all three provided below, AIC, BIC, and log-likelihood, it is optimal to be the minimal. AIC and BIC are based on log-likelihood and, unlike log-likelihood, they punish for the number of parameters used. It was reported before that BIC tends to underestimate the correct number of classes while AIC tends to overestimate it. Nylund-Gibson and Choi (2018). suggest considering AIC as the upper bound of the optimal number of classes and BIC as the lower bound.

The third criterion for the determination of the number of classes is the prior expectations of the number of classes. I expect at least two: those comfortable with their NEET position and the others. A greater number of classes would mean that there are either levels of (dis)satisfaction or varying types of dissatisfaction.

MAIN RESULTS

Finding the optimal number of classes.

Table 2. Fit indices for Six LCA Models.

Table 3. Classes' Frequencies and Proportions (%).

The overview of the model

The statistics on the SWB indicators can be found in Appendix 1. Overall, the positive affect and life satisfaction of the classes can be ranged linearly with Class 2 having the highest life satisfaction and happiness levels, Class 3 having the lowest, and Class 1 having the intermediate position. The negative affect (sadness and disappointment in life) also can be ranged with Class 1 being slightly better off compared to Class 2 and substantially better off compared to Class 3. The Class 3 position may be intuitively clear; after all, a class with low SWB was expected. Apart from the pure description of the classes, the existence of two classes with (different) high SWB indicators should be explained. The next two subsections are focused on the socio-economic characteristics of the classes and their values with the potential solution being suggested in the Discussion section.

An exact Fisher test suggested that the distribution of the classes was uneven throughout 2020 (a link especially important during the pandemic; the significance level for the Fisher test is 5%). At the same time, it is hard to explain the association. 30% of Class 3, the one with the lowest SWB, was in February, more than a month before the COVID regulations took place in Moldova. 32% of Class 1 were interviewed in August. The distribution of Classes by the month of the interview can be found in Table 4.

Table 4. Distribution of Classes Across the Months of 2020.

Classes’ socio-economic characteristics.

Socio-demographically, the classes have similar demographic status distribution. Class 1 has the highest proportion of women (60.5%), but it is only slightly higher than the lowest share (53.7%, Class 3). The classes differ more regarding the age composition with Class 2 being younger and Class 3 being older. Still, every age group exists in every class; those aged 20-24 take a third in each class, and those aged 25-29 take up between one-half and two-thirds of each class. Finally, every second representative of each class has a child. See Tables 5 and 6 for statistics on the classes’ gender, urban/rural, age, employment, partnership, marriage, and parenthood status composition. The only significant differences³ between classes are in their employment and marriage status. Only 27% of Class 3 have a spouse, and 45% of Class 1 are married; Class 2 has an intermediate position. There are also significantly and substantially more disabled and unemployed people in Class 3 compared to the other two classes, in which homemakers are a majority.

Table 5. Gender, Rural/Urban, and Age Composition of Class, %.

Table 6. Employment Status and Place of Residence, by Class, %.

Class 1 respondents (46.2% of the sample), the moderate and tranquil NEET, have low levels of negative affect and moderate levels of positive affect and life satisfaction. For example, only 12,3% gave the highest scores of happiness, and nobody gave the highest satisfaction scores, but only 98% said they felt their life was a failure. The low negative affect levels also reveal themselves by other indicators, including various loneliness symptoms, e.g., 77% have enough people to rely on, and 80% have enough people they feel close to.

Socio-economically, they are the most educated class with 10% having higher education and 37% having a professional education. As for their income, 27% of the class make ends meet “with difficulty” or “with great difficulty”, and 29% do not report any hardships at all. By consumption possibilities, they have an intermediate position with basic needs almost fully covered and having no or almost no budget for luxuries. For example, 75% of the classes can buy new clothes and 73% can afford two pairs of shoes, 69 can keep their houses warm. At the same time, only 15% can afford a one-week holiday outside of their residence, 19% can have leisure activities, and 29% can buy a car. Finally, 86% of Class 1 mostly speak Romanian or Moldovan at home, and only 8.2% speak Russian.

Class 2, affluent and happy NEET, (32.5% of the sample) shows the greatest happiness and life satisfaction: 96% have the highest satisfaction score, and 75% have the highest happiness score. At the same time, 14% feel disappointment in life from time to time. They also have low levels of loneliness and depression symptoms that are not substantially different from those of Class 1. For example, 82% never feel emptiness, 82% have people they can rely on, etc.

Socio-economically, two-thirds of the class only went to school, and only 6% got higher education. By their consumption patterns and possibilities, Class 2 is the most affluent. 26% of them report they make their ends meet “with difficulty” or “with great difficulty”. 40% of them report no difficulties at all. A majority can entertain their friends or family (89%), keep up with their credit payments (83%), keep their houses warm (80%), and buy clothes (79%) or two pairs of shoes (76%). 57% can afford unexpected payments, 42% can replace their furniture, and 39% can buy a car. An overwhelming majority of Class 2 speak Romanian (46%) or Moldovan (43%) at home.

Class 3, poor and unhappy NEET, (21.3% of the sample) contains all those completely unsatisfied (7%) and unhappy (45%), as well as those who feel depressed (6%) or disappointed in themselves (6%). They have the worst levels of other SWB indicators: the frequencies of sadness (each third feels it ‘often’), fearfulness (each fifth), or inability “to shake off the blues” (each fifth) as well as the spread of the sense of emptiness (every forth feels empty). They also have the worst social deprivation by various measurements: only 50% have people they feel close to, less than 50% have people to rely on, and more than 25% feel rejected. At the same time, this is also the class with the highest share of those helping others with childcare (37.5% vs. 20% for other NEET groups). Their physical well-being is also the worst in the group. It was mentioned before that 10% of the Class 3 representatives are ill or disabled; it can be added that 48% of the class consider their health state as “acceptable” or “bad” (20% for the other two classes).

Socio-economically, Class 3 is the least affluent of the classes. An overwhelming majority reports some levels of difficulty in making ends meet. This translates into their low consumption possibilities on virtually every good mentioned above. 59% of them can afford to buy clothes, 55% can afford shoes, and 60% keep their houses warm. However, 14% of them can afford a vacation and 19% afford leisure activities. 36% cannot keep up with their credit payments. Finally, only 21% of them can afford to replace furniture. Finally, almost 40% of them do not have home internet (compared to 23% for other classes). This disadvantaged financial situation is coupled with the lowest education among the classes: only 5% have higher education and 21% have professional education.

25% of Class 3 speak Russian at home, 27% speak Romanian, and 42% speak Moldovan. The highest proportion of Russian speakers and the lowest proportion of Romanian speakers may indicate their ethnic background or political views; it should be specifically highlighted that the share of Gagauz, Bulgarian, or Romani speakers in the group is not substantially different for Class 3 compared to the other two.

Classes’ opinions on gender roles

The GGS dataset offers eleven questions on gender roles. There were statistically significant associations between the class and five of them: two on marriage (if marriage is outdated and if unmarried cohabitation is normal) and three on childbearing (if women (men) need children to be fulfilled and if a working mother can secure warm relationship with her child). Overall, one can attempt to put them on the traditional values – modern values scale, but a deeper analysis suggests the picture is more complicated.

First, the Class 3 representatives value the marriage institution the least, and Class 2 values it the most. Only 48% of Class 3 believe that marriage is not outdated, and 20% believe that a partnership without a marriage is not normal. The values are 63% and 35%, respectively, for Class 2. Class 1 has a mixed position, believing that marriage is not outdated (65%) while also tolerating unofficial partnerships (they are not normal for 19%). Second, more than 75% “absolutely” or “partially” agree that both men and women need children to be fulfilled in their lives. The Class 2 members are more rigid in this regard than the others: 84% “absolutely believe” that women need children whereas only 76% of Class 1 and 58% of Class 3 do. Similarly, more than 75% of each class assured that the bond between a child and a working mother can be as strong as a bond between a child and a non-working one.

It can be summed up as follows: Class 2 has more traditional views, and Class 3 has the more diverse ones. However, the answers to the other questions do not show statistical significance between the classes. Among these questions is another question on working mothers, fatherless families, and LGBT+ rights. Coupled with the previously mentioned financial struggles of Class 3 members, one can suggest that these are examples of traditional values becoming blurred when people cannot afford to people who cannot afford to have an ideal traditional family. The analysis becomes more complicated when one analyses the genders separately. For example, there are statistical differences between women (but not men) in different classes regarding the question of marriage for life: Class 2 is the most radical in their stance (69% “absolutely” agree), and Class 3 has more individuals with modern views (28% "partially” or “absolutely” disagree compared to 7% for the other classes.

DISCUSSION AND CONCLUSIONS

While previously, the NEET was first classified through socio-economic characteristics and then the groups’ SWB was studied, this article undertook a different route: from the SWB to class. NEET includes individuals with varying opportunities and statuses. The literature suggested that the two major groups of NEET in Moldova are inactive women and unemployed men. It was thus hypothesized that those chronically NEET should have worse SWB compared with those casually NEET. Values are one of the mediators of the NEET status on SWB. For example, men who accept the traditional breadwinner role but do not comply with it must have worse SWB. At the same time, as the previous studies show, the inability to act on the gender role also weakens the role.

In this study, NEET youth were differentiated with four SWB indicators: life satisfaction, happiness, feeling depressed, and being disappointed in their life. It was hypothesized that two classes exist: those with low affluence and SWB and those with high affluence and SWB. With Latent Class Analysis, three classes were found. Class 3 (21% of the sample) has the worst SWB with the lowest life satisfaction and positive affect and the highest depression symptoms. Class 2 (33%) has the highest positive affect and life satisfaction with quite low levels of negative affect. Class 1 (46%) has moderate positive affect and life satisfaction levels and very low negative affect. The mere fact that Class 2 has moderate (and not the lowest) levels of negative affect bars one from putting the classes on a “worse SWB – better SWB” scale.

Several explanations could be suggested. Perhaps, the simplest one is the fact that many people report having no negative affect (Schimmack, 2006). However, the question should be reformulated then: is there a reason why the share of those who [generally] report no negative affect is so high in Class 1? The sections on the socio-economic characteristics of the classes and their values show that Class 1 generally holds an intermediate position between the other two classes. Class 1 only has a more affluent position than Class 2 in education. A greater model complication may also be an answer. While the differences between classes may be pronounced for one gender, this may not be true for the other and vice versa.

While there are statistically significant differences in labor market status between the classes (i.e., unemployed vs. inactive), other factors seem to be more substantial. SWB also has a pronounced link with health, income, and consumption patterns. The ‘out of place’ state, i.e., the impossibility of confirming with the traditional roles, was suggested as a major factor of mental health deterioration. The methods of this study did not allow to confirm or to reject the hypothesis, but it is tangentially supported by the fact that the class with the highest proportion of the unemployed and men has the lowest SWB indicators. At the same time, the link between the SWB classes and the gender roles opinions (while holding the gender fixed) should be explored more.

REFERENCES

Alfieri, S., Sironi, E., Marta, E., Rosina, A., & Marzana, D. (2015). Young Italian NEETs (Not in Employment, Education, or Training) and the Influence of Their Family Background. Europe’s Journal of Psychology, 11(2), 311-322. https://doi.org/10.5964/ejop.v11i2.901

Bonanomi, A., & Rosina, A. (2022). Employment Status and Well-Being: A Longitudinal Study on Young Italian People. Social Indicators Research, 161(2-3), 581-598. https://doi.org/10.1007/s11205-020-02376-x

Bynner, J., & Parsons, S. (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

Caroleo, F. E., Rocca, A., Mazzocchi, P., & Quintano, C. (2020). Being NEET in Europe Before and After the Economic Crisis: An Analysis of the Micro and Macro Determinants. Social Indicators Research, 149(3), 991-1024. https://doi.org/10.1007/s11205-020-02270-6

Carver, T., & Grimes, A. (2019). Income or Consumption: Which Better Predicts Subjective Well-Being? Review of Income and Wealth, 65, S1, S256-S280. https://doi.org/10.1111/roiw.12414

Contini, D., Filandri, M., & Pacelli, L. (2019). Persistency in the NEET state: A longitudinal analysis. Journal of Youth Studies, 22(7), 959-980. https://doi.org/10.1080/13676261.2018.1562161

Crismaru, M., Gagauz, O., & Buciuceanu-Vrabie, M. (2017). Inclusion of Youth Not in Employment, Education or Training (NEET Youth). NIER. http://dspace.ince.md/xmlui/handle/123456789/1821

De Luca, G., Mazzocchi, P., Quintano, C., & Rocca, A. (2020). Going Behind the High Rates of NEETs in Italy and Spain: The Role of Early School Leavers. Social Indicators Research, 151(1), 345-363. https://doi.org/10.1007/s11205-020-02370-3

Douglas, M. (2001). Purity and Danger. Routledge. https://monoskop.org/images/7/7d/Douglas_Mary_Purity_and_Danger_An_Analysis_of_Concepts_of_Pollution_and_Taboo_2001.pdf

Enneli, H. Ç., & Enneli, P. (2017). Reinforcement and erosion of traditional gender roles among young people in a poor metropolitan area of Turkey. Journal of Youth Studies, 20(3), 349-365. http://dx.doi.org/10.1080/13676261.2016.1217320

Felaco, C., & Parola, A. (2022). Subjective Well-Being and Future Orientation of NEETs: Evidence from the Italian Sample of the European Social Survey. Social Sciences, 11(10), 482. https://doi.org/10.3390/socsci11100482

Gagauz, O., & Crîșmaru, M. (2024). Factorii de risc pentru plasarea tinerilor în categoria NEET: cazul Republicii Moldova. Calitatea Vieții, 35(1), 1-19 https://revistacalitateavietii.ro/journal/issue/view/2024-1

Ganzach, Y., Bulmash, B., & Pazy, A. (2023). Top-down versus bottom-up processes in the formation of positive and negative retrospective affect. Cognition and Emotion, 37(1), 86-97. https://doi.org/10.1080/02699931.2022.2146658

Goldman-Mellor, S., Caspi, A., Arseneault, L., Ajala, N., Ambler, A., Danese, A., Fisher, H., Hucker, A., Odgers, C., Williams, T., Wong, C., & Moffitt, T. E. (2016). Committed to work but vulnerable: Self‐perceptions and mental health in NEET 18‐year olds from a contemporary British cohort. Journal of Child Psychology and Psychiatry, 57(2), 196-203. https://doi.org/10.1111/jcpp.12459

Linzer, D. A., & Lewis, J. B. (2011). poLCA: An R Package for Polytomous Variable Latent Class Analysis. Journal of Statistical Software, 42(10), 1-29. https://doi.org/10.18637/jss.v042.i10

Ministry of Health, Labour and Social Protection of the Republic of Moldova, National Bureau of Statistics (sample development), UNFPA & NIDI-GGP (partner and data distributor). (2020). Republic of Moldova - Generations and Gender Survey. https://moldova.unfpa.org/sites/default/files/pub-pdf/summary_of_generations_and_gender_survey_1.pdf

Nylund-Gibson, K., & Choi, A. Y. (2018). Ten frequently asked questions about latent class analysis. Translational Issues in Psychological Science, 4(4), 440-461. https://doi.org/10.1037/tps0000176

Özdemir, M. O., Ozcan, K. M., & Üçdoğruk, Ş. (2023). Youth not in Employment, Education, or Training (NEET) in Turkey: A Regional Analysis. Ege Academic Review, 23(3), 409-426. https://doi.org/10.21121/eab.1178984

Schimmack, U. (2006). Methodological Issues in the Assessment of the Affective Component of Subjective Well-Being. In: A. D. Ong, M. H. M. van Dulmen (Eds.), Oxford Handbook of Methods in Positive Psychology (pp. 96-110). New York: Oxford University Press. https://doi.org/10.1093/oso/9780195172188.003.0007

Sobotka, T. (2008). The diverse faces of the Second Demographic Transition in Europe. Overview Chapter 6. Demographic Research, 19, july - december, 171-224. https://doi.org/10.4054/DemRes.2008.19.8

Tamesberger, D., & Bacher, J. (2014). NEET youth in Austria: A typology including socio-demography, labour market behaviour and permanence. Journal of Youth Studies, 17(9), 1239-1259. https://doi.org/10.1080/13676261.2014.901492

Thompson, M., Ellis, R., & Wildavsky, A. (2018). Cultural Theory. Routledge: Taylor & Francis Group.

Understanding Children’s Work (UCW) Programme. (2013). NEET youth dynamics in Indonesia and Brazil. A cohort analysis. https://thedocs.worldbank.org/en/doc/81b760637f22dc1f26465c9397efdac2-0050022022/original/NEET-Youth-cohort-effects-Indonesia-Brazil.pdf

Van Praag, B. M. S., Frijters, P., & Ferrer-i-Carbonell, A. (2003). The anatomy of subjective well-being. Journal of Economic Behavior & Organization, 51(1), 29-49. https://doi.org/10.1016/S0167-2681(02)00140-3

Vladisavljević, M., & Mentus, V. (2019). The Structure of Subjective Well-Being and Its Relation to Objective Well-Being Indicators: Evidence from EU-SILC for Serbia. Psychological Reports, 122(1), 36-60. https://doi.org/10.1177/0033294118756335

Weller, B. E., Bowen, N. K., & Faubert, S. J. (2020). Latent Class Analysis: A Guide to Best Practice. Journal of Black Psychology, 46(4), 287-311. https://doi.org/10.1177/0095798420930932

Yates, S., & Payne, M. (2006). Not so NEET? A Critique of the Use of ‘NEET’ in Setting Targets for Interventions with Young People. Journal of Youth Studies, 9(3), 329-344. https://doi.org/10.1080/13676260600805671

Appendix 1. SWB and Socio-Demographic Characteristics of the Classes.

Table 1. The Share of Individuals with Respective Life Satisfaction Levels, by Class, %

Table 2. The Share of Individuals with Respective Happiness Levels, by Class, %

Table 3. The Share of Individuals Reporting the Respective Frequency of Feelings Depressed, by Class, %

Table 4. The Share of Individuals Reporting the Respective Frequency of Disappointment in Life, by Class, %