As societies navigate the diverse landscapes of economic, sociocultural and educational domains, there is an array of perspectives that attempt to decipher the underlying causes of NEET prevalence. The first theory relates to the human capital theory. It provides an explanation for the returns on investments made in education and training, viewing them as strategic investments in human resources (Becker 1964). From this perspective, education and training are perceived as intentional endeavours aimed at equipping the workforce, thereby contributing positively to employee and organisational productivity (Fleischhauer 2007). These competencies not only play a crucial role in enhancing productivity within the labour market but also contribute substantially to the development of other essential skills. Thus, not having the necessary education and training may force young people into NEET (Mncayi 2020). Within the realm of human capital lies the issue of skills mismatch. Sala (2011) argues that holding other factors constant, young individuals may fall into NEET status because of disparities in their proficiency and acquisition of relevant skills, thus contributing to mismatch. This situation arises when there are discrepancies between the skills acquired by potential employees and those required by employers, leading to an inability to fill vacancies (Pitan & Adedeji 2012). Because of a skills mismatch, employers encounter difficulties in finding individuals possessing the required skills and education (Restrepo 2015).

The second perspective is grounded in the spatial mismatch theory, which suggests that individuals with lower skills often face unfavourable labour market outcomes because of their detachment from suburban job opportunities (Bunel, L’Horty & Petit 2016). The theory underscores the importance of proximity to specific types of low-skilled employment (Gobillon & Selod 2021). The physical distance between an individual’s residence and available employment hinders the job search and diminishes the likelihood of escaping unemployment (Kain 1968). Unemployed and less-educated young people residing farther from employment centres will likely encounter prolonged and expensive job searches, hence the increase in NEET numbers.

In conjunction with these insights, various scholars have also linked the prevalence of NEET among youth to the presence of economic deprivation, a situation where individuals struggle to fulfil their basic needs and perceive this lack as a denial of what is rightfully theirs (Ripamonti & Barberis 2021; Rahmani & Groot 2023; Su, Wong & To 2020). Within this context, economically deprived youth often find themselves in the category of NEET because of limited access to education, training and employment opportunities, hindering their path to economic independence. Consequently, in the face of formidable challenges posed by high unemployment rates, there is an increased risk that these young individuals may confront escalating financial hardships (Aassve, Cottini & Vitali 2013). Individuals who lack education, training and employment are more likely to experience poverty, leading to insufficient access to necessities such as food, clothing and shelter (Janlert & Hammarström 2009). Opposing these supply side assertions, the demand side explanations on the prevalence of NEETs have primarily looked at the quality of economic growth and how it influences youth economic opportunities, particularly in employment. Economic growth that is not employment friendly poses a substantial negative impact on youth, contributing to higher rates of unemployment and those classified as NEET. Notably, youth who have dropped out of school tend to fare worse in the labour market. During challenging economic periods, young people are usually laid off first, which impedes their capability to build skills and experiences, consequently inducing long-term scarring effects.

The last perspective is linked to the job search theory. The decision to embark on a job search is not solely an individual one, as structural and labour market issues often obscure this process, particularly for young people who find themselves in a vulnerable position compared to other demographic groups. They are frequently relegated to temporary employment, possess limited or no experience crucial for competitive job markets and face heightened challenges in the job search (Ylistö 2018). Often, young individuals lack adequate knowledge in their pursuit of employment because of unclear employment plans and a lack of understanding regarding careers. Insufficient work experience, coupled with a dearth of social capital and necessary skills, places them at a considerable disadvantage regarding employment prospects (Moore 2019). This group is prone to additional challenges, including negative perceptions from employers, a scarcity of entry-level positions and ineffective information channels. The significance of personal networks in the labour market is multifaceted, as contacts can facilitate job searches and contribute to various processes of social reproduction (Vacchiano and Spini 2021; Ylistö 2018). Young people’s access to social capital and their ability to navigate interactions with uneven resources are closely linked to individual tactics and social factors, with social background playing a significant role (De Jongh 2019). Social networks emerge as a primary source of extraordinary inequality where young people’s access or lack thereof frequently grants or denies them entry into qualified sectors, offering the potential for satisfying jobs (Vacchiano, Yepes-Cayuela & Martí 2021).

Young people are a great source of economic and social potential in any country (Rollnik-Sadowska 2016). However, the challenges that the youth encounter in the labour market are a tremendous threat to any potential gains that may be made from their participation in the economy. Unemployment rates among young people are high globally, exacerbating poverty and inequality (ILO 2022). First-time job seekers face challenges in securing employment and accessing the labour market (Leibbrandt et al. 2010). Although many countries have made significant strides to improve employment and economic growth, such improvements have not been able to keep up with the continually expanding labour market supply of young people (Mncayi & Meyer 2022). For example, according to the 2020 Global Youth Employment Trends Report, between 1999 and 2019, the number of young people’s participation in the workforce worldwide decreased from 568 to 497 billion, despite an increase in the number of young people from 1 to 1.3 billion (ILO 2020, 2022).

The prevailing global pattern indicates significant difficulties for young individuals in securing satisfactory employment, let alone any form of employment (Mncayi & Meyer 2022). As the onset of the global financial crisis, the employment rate for young people has decreased by over 15%, contrasting with the consistent rise in the employment rate for adults (United Nations 2019). While the diminishing participation of young people in the labour force may be largely attributed to an increased percentage attending school, another significant factor could be their inclusion in the NEET category (Canon, Debbaut & Kudlyak 2013; Wickremeratne & Dunusinghe 2018). The disproportionate representation of young individuals in NEET statistics is worrisome, posing a threat to social dissatisfaction and stability, as highlighted by Escudero and Mourelo (2013).

Generally, the labour force framework categorises people as employed, unemployed or inactive. One is considered employed if they are in paid work or self-employed, while those who are unemployed are actively looking for employment (Assirelli 2013). Those considered inactive are without work and not looking for any. The NEET, on the other hand, refers to those unemployed and inactive, whether by choice or not and remains an important research question (Carcillo et al. 2015; Chen 2011). The concept of NEET was first coined in the late 1980s in the United Kingdom, later gaining recognition in the mid-1990s (Holmes, Murphy & Mayhew 2021; Holte et al. 2019). At that time, it described young people in the 16–18-year-old age category who were not in employment or education. The importance of NEET lies in the fact that it not only speaks about the labour market but also education and training, essentially zooming in on school-to-work transitions. The exact reasons behind the NEET position are diverse and are connected to several and often entwined individual, household and economic-related aspects. As argued by De Lannoy and Muridizi (2019), most personal and household factors can arise because of an individual being NEET, at the same time also cause an individual to be NEET. The proportion of women in NEET statistics is even worse, especially since a large share of them are confronted with childcare and family responsibilities (Maguire 2018), including marriage, which prevents them from the labour market to engage in income-earning activities.

The NEET measure encompasses a broad categorisation of young individuals for whom finding employment is either not feasible or not deemed worthwhile (Mncayi & Meyer 2022). However, it specifically focuses on young people engaged in unproductive or peripheral jobs, as highlighted by Assaad and Levison (2013) and Meyer and Mncayi (2021). The increasing prevalence of informal employment among the youth has led to heightened concerns about well-being issues, particularly in the current scenario where securing stable employment is progressively challenging. Nevertheless, the number of young people in the NEET category remains very high and continues to rise. This implies that the youth’s full potential can be permanently impaired as they are not acquiring any working experience in the labour market, which further limits any chance they may have of finding employment (Kumo, Chulu & Minsat 2016) or getting any income from work or improving their learning (ILO 2020).

At a macroeconomic level, the NEETs are extremely problematic in the sense that the prolonged period of NEET hinders productivity and competitiveness, thereby being a major impediment to the realisation of economic growth (Schmillen & Umkehrer 2017). This consequently results in lost output and government revenue with increased government spending (Ebell & O’Higgins 2015). On the other hand, according to Dickens and Marx (2020), being NEET has a sturdy bearing on physical, emotional and psychosocial well-being. For most young people, being part of NEET in the early stages of their working life can bring about major economic and social consequences in the long term (Holmes et al. 2021), in the process trapping them in the vicious cycle of poverty and causing them to be socially excluded (Gutiérrez-García et al. 2017; Ralston et al. 2016). The latter can be even more pronounced among young males who commonly find themselves at the borders of society and are more susceptible to perilous and violent behaviour Allen (2016).

Another significant challenge brought by NEET is the decline in human capital that arises in that idle state, which can pose more challenges, especially for low-educated young people with no work experience. Education attained at some point in the past, without being complemented by practical work experience or on-the-job training, can result in the depletion of human capital or skills, akin to the depreciation of physical capital when left unused (Amadeo 2018). This situation can have enduring repercussions on career advancement, earning potential and the accumulation of retirement income, as highlighted by Li, Duncan and Miranti (2015). Furthermore, the ramifications of the NEET phenomenon are underscored by crime statistics revealing that a significant proportion of house burglaries, particularly in countries like South Africa, involve individuals aged 19–25 years (Meyer & Mncayi 2021). Approximately 90% of those apprehended in connection with these crimes either did not complete grade 12 or are unemployed, as indicated by research conducted by Cloete and Butler-Adam (2012). Many of them discontinued their education because of various factors, including familial responsibilities, financial constraints and other challenges (Mncayi 2020). The subsequent section provides a summary of the findings relating to previous studies on NEETs.

The NEET measure is not widely accessible for the majority of developing nations, as indicated by Assaad and Levison (2013) and Mncayi (2020). Regardless, a growing body of literature has been emerging, investigating different aspects relating to NEETs. Using information from two British birth cohorts, Bynner and Parsons (2002) found several risk indicators for being NEET in the United Kingdom. According to their research, the socioeconomic position of the family, the education of the parents as well as their level of involvement in the child’s schooling, residential area and children’s educational attainment are all reliable indicators of NEET status in the later years of their lives. A study by Chen (2011), which examined 15–20-year-old Taiwanese youth who were not employed, in education or in vocational training found economic reasons to largely explain the participants’ NEET status, suggesting that most of the respondents did not become NEETs by choice.

In a study by Bardak, Maseda and Rosso (2015) on young people who are NEET in the 29 partner countries of the European Training Foundation, it was discovered that family and individual characteristics such as gender, education level, age and socioeconomic background play a significant role in determining how well young people transition from school to work and are therefore some of the main causes of young people becoming NEETs. However, there are also significant differences between the country samples, which can be attributed to the various social or cultural norms, the various structures and the efficacy of their educational and vocational education and training (VET) systems, as well as the operation of regional labour markets and economies in general.

Using a longitudinal survey of Australian youths between the ages of 15 years and 24 years, Stanwick, Forrest and Skujins (2017) discovered that prolonged NEET states are largely determined by not finishing grade 12, having children and coming from a lower socioeconomic background. The study also discovered that having a period of persistent NEET status between the ages of 15 and 19 years is linked to a higher likelihood of persistent NEET status between the ages of 20 years and 24 years as well as a lower likelihood of beginning or finishing a higher-level qualification by the age of 24 years. These young individuals appear to have also been affected by the global financial crisis.

Using the Poverty Monitoring Survey in Senegal, Cabral (2018) found that young people’s NEET status is determined by the existence of physical and mental disabilities, as well as the age and gender of the respondent. In addition, the gender, education status, employment, marital status of the head of household and household income were other significant variables explaining the NEET status of the sample. In a study that looked at the youth NEET in Sri Lanka, Wickremeratne and Dunusinghe (2018) discovered that age, gender, education, residential sector and marital status were the key drivers of youth NEET in Sri Lanka. Quintano, Mazzocchi and Rocca (2018) found that the portion of NEETs in Italy is more pronounced for women and immigrants regardless of their level of education or field of study.

In their review of factors associated with being NEET among black youths in South Africa, De Lannoy and Mudirizi (2019) found that age, gender, health and education fuelled the likelihood of being NEET among this group. Further results from De Lannoy and Mudirizi’s study showed that living in high-income households with an employed person significantly lessens the chances of being NEET in the sample, regardless of gender. O’Higgins (2019) observed that individuals with higher levels of education and those residing in rural areas are more susceptible to NEET status. However, in terms of numerical representation, there is a higher count of NEETs in urban areas than in rural ones.

According to a study on NEET in Morocco by Alfani et al. (2020), young men and women with less education are likelier to become NEET. This is especially true for married and/or parenting women. Compared to large cities or rural farming communities, medium-sized towns are likewise more inclined to have a greater prevalence of NEETs. In general, better economic situations and more educated parents tend to reduce the likelihood that young members of the household may become NEETs. In a qualitative study examining the existing literature on young people in NEET in sub-Saharan Africa, Cieslik et al. (2022) discovered that NEET rates tend to be more persistent for young women compared to young men. The study also underscores the challenges associated with the scarcity of data in this region. The empirical procedures used in the study will be explained in more detail in the next section.

The focal purpose of the study was to determine the factors that influence the likelihood of young working-aged individuals being classified as NEETs. In doing so, the study employed a quantitative approach using secondary data from Stats South Africa. The data were obtained from the fourth Quarterly Labour Force Survey (QLFS) of 2021. The QLFS is a household-based sample survey that aims to collect data on the national labour market activity on individuals of working age (15 years and older). To achieve the primary objective, a cross-sectional research design was used with the sample consisting of young working-age individuals between the ages of 15 and 34 years residing in all nine provinces of the country. Based on this criterion, the sample consisted of a total of 14 338 individuals. The study comprised both qualitative and quantitative elements. The former included a literature review, while the latter included the use of various techniques in which the collected data were analysed.

The analysis of the data was carried out using the Statistical Package for Social Sciences (SPSS) software version 27. The process included various techniques and was divided into three distinct steps. The first step comprised a descriptive analysis with the use of frequency and percentage tables to provide a demographic and geographic overview of the sample’s composition. In the second step, more inferential techniques were employed, including the use of cross-tabulations, chi-square tests and Cramer’s V-test statistics to identify possible associations and strengths in the relationships between the categorical variables that were used. Finally, to identify the possible predictors of NEET status for youths in South Africa, the study employed a binary logistic regression model. Possible predictors were identified using the literature review. These are shown in Table 1 together with the coding criteria that were included in the modelling process.

In line with determining the influence of the predictor variables discussed earlier on the NEET status of the working-aged youth, the dependent variable (NEET status) was coded dichotomously. In this instance, 1 indicates young people not being employed, in education or any form of training, while 0 shows that these individuals were employed, currently attending some form of educational institution, or participated in training courses. Considering this, the study estimated the binary logistic regression as follows: NEETi=θ0+θ1AGEi+θ2GENi+θ3MASi+θ4RACEi+    θ5LOEi+θ6GEOi+θ7METi+εt[Eqn 1]

Here, NEET_i represents the dependent variable, the NEET status of the participants, θ₀ shows the constant while θ₁,θ₂,θ₃…θ₇ shows the coefficients of each of the included explanatory variables in the model. Finally, to account for the statistical noise, the error term is included, denoted in Equation 1 as ε_t. To ensure the reliability of the results, the modelling included the use of various diagnostic tests in order to ascertain whether the data were a good fit for the model used as well as to ensure that no serious form of multicollinearity was present between the explanatory and dependent variables. These tests included the use of the omnibus test of model coefficients, and the Hosmer and Lemeshow (2000) test, while both variance inflator factors and tolerance values were employed, respectively. In addition to this, both the Nagelkerke R-square (Nagelkerke 1991) and Cox and Snell R-square (Cox & Snell 1989) statistics were used to estimate the percentage of variation in the dependent variable as explained by the model.

Ethical clearance to conduct this study was obtained from the North-West University Economic and Management Sciences Research Ethics Committee (EMS-REC) (No. NWU-01820-23-A4).

The purpose of reporting on the demographic and socioeconomic composition of the sample is twofold. Firstly, when showcasing the responses to the survey, it assists in determining that the collected information on the population parameters is accurate (Groves 2006). More importantly, though, is that it provides additional context, which assists in better comprehending the inherent dynamics of the study topic. As such, Table 2 indicates the results pertaining to the demographic composition of the sample. As shown, the sample consisted of an even distribution between males (47.8%) and females (52.2%). Considering the population group representation, more than four out of the five (89.5%) samples were African or black people, followed by 6.0% who indicated that they were mixed race. Only a small portion of the sample indicated they were white people (3.1%), with Asians or Indian people accounting for only 1.4% of the total 14 338 participants. Given the national scope of the QLFS survey, this distribution does reflect the composition of the population of South Africa (Stats SA 2022b).

Given the focus on young working-aged individuals, the study made use of the national definition of youth, which sees young people between the ages of 15 years and 34 years (Khuluvhe & Negogogo 2021). As can be seen from Table 2, within this age category, the sample was approximately evenly spread between younger groupings, 15–19 years (29.8%), 20–24 years (24.3%) and those in the older groups, that is 25–29 years (23.2%) and 30–34 years (22.7%) of the cohort.

Upon review of marital status, the distribution seems to reflect the life stage of the sample. A much lower percentage of the participants (11.6%) indicated they were indeed married or living with someone, while the majority (88.4%) indicated otherwise. This also points to the fact that with continued struggles in the labour market, many young working-age individuals delay family or couple formation or cohabitation (De Jongh 2019). In addition, participants indicated the province and geographic location of their residence. Results from Table 2 show that participants from all nine provinces in South Africa were included, with Gauteng (22.7%), KwaZulu-Natal (21.7%) and Limpopo (13.0%) accounting for most of the young working-aged individuals. Moreover, from a geographical perspective, 60% resided in urban regions. However, the participants were not all situated within metropolitan municipal areas (non-metro = 60.7%; metro = 39.3%). In fact, upon further review (as shown in Appendix 1, Table 1-A1), females, mixed race youth and African youth were more likely to be situated in non-metropolitan areas.

Given the importance of education in labour market outcomes, it was critical to identify the present level of education within the sample. Responses in Table 2 indicated that most of the participants attributed some form of secondary education (84.7%), while only 9.7% (6.9% + 2.8%) indicated that they have acquired a qualification higher than this. Finally, Table 2 likewise shows the current level of young NEETs in the South African labour market. Of the 14 338 sampled participants, 42.4% of young people of working age indicated that they were not in any form of employment, education or training. Compared to similar countries such as India (30.7%), Malaysia (12.6%) and Brazil (23.5%), even according to the international definition of youth (15–24 years), South Africa attributes relatively higher NEET rates (ILO 2021).

After providing some insight into the demographic and socioeconomic background of the participants, the analysis ensued with the use of cross-tabulations, chi-square tests and effect size estimators. This was done with the purpose of identifying the influence of the possible predictors on the NEET status of young working-aged individuals. Table 3 reports the results for these techniques, which showed that all possible predictors identified in the literature review attributed statistically significant different distributions (at a 1% significance level) with varying degrees of strength of association with the likelihood of being classified as NEET. From these results, statistics showed that younger ages, especially within the 15–19-year age categories, attributed the lowest distributions of NEETs (12.2%). The highest distributions were recorded for those aged 25–29 years, which contributed a 61.3% NEET rate. In addition to this, from a gender perspective, females (46%) attributed more young NEETs when compared to males (38.4%).

In relation to their marital status, being married or living with someone, results showed that 47.1% were classified as NEET compared to a 41.8% NEET rate for those who were not married or living on their own. Similar results were also reported by De Lannoy and Mudiriza (2019), which suggests that these might point to gendered aspects and cultural norms of marriage, especially relating to young females. Given the socioeconomic and historical significance of race within the country, it was important to determine the influence of this characteristic on the NEET status of young working-aged individuals. Results in Table 3 in this regard seem to suggest that the highest NEET rates were reported for black people (43.8%). This was followed by the mixed-race cohort (38.5%) and Indian or Asian people (29.4%), while the white people cohort attributed the lowest NEET rates, at 15.9%, respectively. These findings are in line with those reported by Holte et al. (2019).

In conjunction with the aforementioned results, Table 3 also shows the distributions according to the different levels of education among the participants and the prevalence of being classified as NEET. The distributions suggest that higher levels of education seem to offset the probability of not being employed, in education or in some form of training. Another important consideration for the study was to determine whether the location of working-age individuals contributed an influence on their ability to gain employment or to actively improve their skill sets. From this point of view, residing in either urban (40.2%) or metropolitan municipal areas (39.7%) contributed lower NEET rates compared to those situated in rural (45.6%) or non-metropolitan areas (44.1%), respectively.

Given the nature of specific household dynamics in which members participated in the production of non-market goods (i.e. subsistence farming, collection of water, general household work), it was important to determine the influence of this allocation of time on young working-aged individuals’ NEET status. Results show that actively allocating time to these activities did, in fact, contribute to higher NEET rates (yes = 48.6%; no = 41.0%). Finally, using Cramer’s V effect size estimator, the study determined the strength of association between each of the predictors and the NEET status of the participants. Based on the estimates shown in Table 3, gender (V = 0.077), marital status (V = 0.034), level of education (V = 0.071), geographic location (V = 0.054), municipal area (V = 0.044) and involvement in non-market activities (NMAs) (V = 0.060) all contributed weak associations. However, race (V = 0.105) and age (V = 0.405) contributed moderate and strong associations, respectively.

Following the cross-tabulation analysis, the study proceeded with the logistic regression analysis. This was done with the purpose of identifying the determinants of the sample’s NEET status. The results for the analysis are reported in Table 4. Upon review of the results associated with age, it seems that the younger groupings (15–19 years) were less likely to be NEET compared to the reference category (30–34 years). In fact, the odds ratio of 0.103 suggests that individuals within this specific age category were 89.7% (0.103–1) less likely to be classified as NEET compared to their older counterparts. Contrariwise, older age groupings, particularly those between the ages of 25 and 29 years, contributed a 47.6% (1.476–1) higher likelihood of being NEETs. These findings are in line with those reported by Bardak et al. (2015).

For South Africa, its legacy of the apartheid era and its racially discriminative legislature have had enduring effects on various aspects of its labour market (Beukes et al. 2017). In this regard, it was important to include race as a possible predictor of young working-aged individuals’ NEET status. In doing so, being classified as a white person was used as the reference category. Results from Table 3 for this aspect show that all included categories of race contributed positive and statistically significant (at 1% level of significance) coefficients. From this, it can be concluded that a notable determinant of NEET status for young South Africans relates back to their racial characteristics.

In line with the importance of education within the NEET framework, the study included five different levels of education within the regression model. Attributing a tertiary education within this aspect was used as the reference category. Results show that all levels of education, apart from the post-secondary level, had statistically significant (sig. = 0.000) positive coefficients. As expected with higher levels of education, the likelihood of being classified as NEET does seem to diminish. When compared to those who had obtained a tertiary education, those with no completed primary education contributed the highest likelihood of being classified as NEET (253.4%), which was followed by those with a completed primary education (101.4%) and those with a secondary education (51.0%).

In addition to these variables, the model included two predictor variables that served as proxies towards the measurement of location and distance and their influence on the NEET status of young working-aged South Africans. In this regard, results show that the coefficients for both geographic types (non-urban = reference category; sig. = 0.004) and the municipal area of residence (non-metro = reference category; sig. = 0.001) were negative and statistically significant. The odds ratios of 0.868 and 0.850 show that being situated in urban and metropolitan areas reduces the likelihood of being classified as a young NEET by 13.2% and 15.0%, respectively. These findings correspond to those shown by Stanwick et al. (2017) and Alfani et al. (2020). Finally, upon understanding the dynamics of involvement in non-market activities that come to characterise various developing regions, this characteristic was included in the model. Results from Table 3 indicate that allocation of time towards these activities does in fact seem to increase the likelihood (odds ratio = 1.324; sig. = 0.000) of not being employed, nor being registered for any form of education or training.

After the estimation of the model, various diagnostic tests were employed to ascertain the reliability of the results. All VIF and tolerance values from this perspective were within the required ranges, signifying that there was no presence of multicollinearity between the predictor variables. Furthermore, both the Hosmer and Lemeshow (sig. = 0.161) and the Omnibus test of model coefficients (sig. = 0.000) confirmed good model fit.