Increasing diversity, precarity and prolonged periods of education in the transition from school to work in Britain

Key takeaways

content: "@pelikhIncreasingDiversityPrecarity2024" -file:@pelikhIncreasingDiversityPrecarity2024

Reading notes

Imported on 2024-05-07 19:36

⭐ Important

• & The distinctive British early transition from school to work is still prevalent, although trajectories have become more complex and precarious with an increase in part‐time employment and prolonged stays in education among the youngest cohorts. Occupational outcomes of highly educated men and women were similar. However, women who did not continue education were more likely to experience turbulent transitions with longer spells of part‐time work and inactivity. The proportion of university graduates from lower socioeconomic backgrounds has increased, yet their chances of being in professional and managerial occupations remain significantly lower. (p. 1)
• & life course transitions during the early stage of adulthood have become less standardised, more turbulent, individualised, and ‘protracted’ (Billari & Liefbroer, 2010;Elzinga &Liefbroer,2007; Macmillan, 2005; Shanahan, 2000). (p. 1)
• & the British pattern is often viewed as being defined by social class (Bynner, 2001; Cavalli & Galland, 1995). (p. 1)
• & The British tradition is often linked to open labour market relationships based on free market forces and competition and a flexible education and training system that allows various pathways to obtaining necessary work (p. 1)
• & qualifications (Blossfeld et al., 2005; Bynner, 2001; Mills & Blossfeld, 2003; Raffe et al., 1998). (p. 2)
• & The shift to a service economy starting in the 1970s and rapid development of information technology, led to the restructuring of the labour market and a subsequent polarisation of jobs (Ashton et al., 1990; Goos & Manning, 2007; White & Lakey, 1992). (p. 2)
• & Young people who had positive perceptions regarding future career prospects without continuing education, faced a new reality of scarce employment opportunities without having specific qualifications (Bynner, 2001; Maguire & Maguire, 1997; Roberts et al., 1994). (p. 2)
• & the median age at starting a first full‐time job has increased from 17 years among men and women born in the 1950s to 19.3 and 20.2 years among men and women born in the 1980s, respectively (Pelikh, 2019). (p. 2)
• & Previous work has focused mainly on either describing school‐towork trajectories in the United Kingdom (Anders & Dorsett, 2017; Anderson & Nelson, 2021; Brzinsky‐Fay, 2007; Dorsett & Lucchino, 2014; McMunn et al., 2015; Quintini & Manfredi, 2009; Schoon & Lyons‐Amos, 2016, 2017; Schoon et al., 2012)oron studying early labour market outcomes and earnings with respect to obtained educational qualifications (e.g. Belfield et al., 2018; Blundell et al., 2000; Britton et al., 2020, 2021; Howieson & Iannelli, 2008; Smith et al., 2001). (p. 2)
• & these studies usually find that higher educated groups have better employment prospects, they have paid less attention to the long‐term association between school‐to‐work trajectories and occupational outcomes. (p. 2)
• & To be able to grasp how early employment conditions have changed over time, we turn to a more granular approach to defining ‘work’ by accounting for differences in the type of employment (full‐time, part‐time, self‐employed) and looking at a change across 5‐year birth cohorts: 1974–1979; 1980–1984; 1985–1990. Having come of age after the expansion of higher education and shift to a service economy, their life course trajectories may share similarities but also diverge due to factors like the housing and economic crises, and introduction of tuition fees. Examining 5‐year birth cohorts helps capture the interplay of change and continuity in life course patterns during this transformative period. (p. 2)
• & Socioeconomic and cultural changes dating back to the 1960s (and often associated with the Second Demographic Transition; Van de Kaa, 1987) have dramatically influenced young people's lives in industrialised countries, leading to the destandardisation of life trajectories (Buchmann, 1989; Liefbroer, 1999; Macmillan, 2005; Shanahan, 2000). (p. 2)
• & individualisation of the life course has led to a larger variety in the occurrence and sequences of life course events associated with the transition to adulthood, including school‐to‐work transitions, among cohorts born from the 1970s onwards (Huinink, 2013; Macmillan, 2005; Schoon & Lyons‐Amos, 2016; Shanahan, 2000). (p. 2)
• & Yet, an alternative view has emphasised the prevalence of structured trajectories defined by socioeconomic origins, despite the increased individualisation of the life course and expansion of higher education (Billari et al., 2019; Côté, 2002; Côté & Bynner, 2008; Furstenberg, 2008; Sironi et al., 2015). (p. 2)
• & These differences in British society are often referred to as the ‘youth divide’—the polarisation between the advantaged and the disadvantaged—and the existence of a so‐called ‘fast‐’ and ‘slowtrack’ in the transition to adulthood (Bynner, 2001, 2005; 2of17 | PELIKH and ROWE 1 5 4 4 8 4 5 2 , 0 , D o w n l o a d e d f r o m h t t p s : / / o n l i n e l i b r a r y . w i l e y . c o m / d o i / 1 0 . 1 0 0 2 / p s p . 2 7 7 1 b y T e s t , W i l e y O n l i n e L i b r a r y o n [ 1 8 / 0 4 / 2 0 2 4 ] . S e e t h e T e r m s a n d C o n d i t i o n s ( h t t p s : / / o n l i n e l i b r a r y . w i l e y . c o m / t e r m s a n d c o n d i t i o n s ) o n W i l e y O n l i n e L i b r a r y f o r r u l e s o f u s e ; O A a r t i c l e s a r e g o v e r n e d b y t h e a p p l i c a b l e C r e a t i v e C o m m o n s L i c e n s (p. 2)
• & Jones, 2002). The ‘slow‐track’ is associated with prolonged periods spent in education and the postponement of labour market entry and family formation. ‘Slow‐track’ pathways were traditionally prevalent among young people from more advantaged backgrounds and among those whose parents have tertiary levels of education (Bynner & Joshi, 2002; Patiniotis & Holdsworth, 2005). ‘Fast‐track’ pathways, on the contrary, relate to young people from lower socioeconomic backgrounds who tend to leave school at a minimum age of 16 (raised to 17 in 2013 and to 18 in 2015 in England) and rapidly start work and family careers. The origin of the existence of ‘fast‐track’ pathways lies in the tradition of high demand for unskilled youth in labour‐intensive industries in Britain which allowed young people to enter the labour market straight after finishing compulsory school, without any further qualifications (Ashton et al., 1990; Maguire & Maguire, 1997). (p. 3)
• & The manufacturing sector, which provided jobs for a quarter of 16‐ to 24‐year‐olds in 1981, employed only 8% of workers by 2011 (Ashton et al., 1990; Sissons & Jones, 2012). (p. 3)
• & For instance, in 1986 around two‐thirds (63%) of part‐time jobs required no qualifications on entry, but by 2012 this had fallen to less than a third (30%) (Gallie et al., 2014). Some argue that employers have deliberately raised minimum education standards for recruiting young employees (often exceeding actual needs) thus propagating a culture of credentialism (Côté & Bynner, 2008; Goos & Manning, 2007). (p. 3)
• & Thus, the higher education participation rate increased from 12% in 1979 to 30% in the early 1990s and 49% in 20151 (Department for Education, 2019). (p. 3)
• & In 1992, women's higher education participation rates exceeded men's and reached 53% by 2015 compared to 43% in men (Department for Education, 2019). (p. 3)
• & 2004, the U.K. government raised tuition fees from £1000 to £3000 per academic year and introduced a new higher education funding system. Both changes were effective from the 2006–2007 academic year.2 (p. 3)
• & The ‘slow‐’ and ‘fast‐track’ division of youth has been criticised for not considering economic precarity, hindering rapid labour market entry for those not pursuing further education (Stone et al., 2011), as well as overlooking the existence of the ‘middling’ pathway (MacDonald, 2011; Roberts, 2011, 2013; Schoon, 2015). This pathway refers to those who manage to ‘successfully’ transition into the labour market, meaning they succeed in consistently remaining in employment and experience a steady career progression without pursuing higher education (p. 3)
• & Salary returns to education have been shown to vary with the level of qualification, parental socioeconomic background and gender (Card, 1999; Dearden et al., 2002; Friedman et al., 2017). (p. 3)
• & Indeed, it has been shown that, on average, highly educated individuals experience greater occupational advancement and income growth, compared to those without a degree (Blundell et al., 2016; Harkness & Machin, 1999; McIntosh, 2006; Walker & Zhu, 2008). (p. 4)
• & the proportion of young people with no qualifications who are ‘Not in Education, Employment or Training’ (NEETs) or experience ‘low‐pay, no‐pay cycles’ has increased (Bell & Blanchflower, 2010; Crawford et al., 2014; Howieson & Iannelli, 2008; Jongbloed & Giret, 2022; Shildrick et al., 2010; Sissons & Jones, 2012). (p. 4)
• & Although the numbers of highly educated young people from lower socioeconomic background have been consistently increasing in the United Kingdom, a gap in earnings and occupational outcomes among graduates from higher and lower socioeconomic backgrounds (SES) has persisted (Duta et al., 2021; Friedman et al., 2017) (p. 4)
• & The sample for the subsequent waves includes all adults reaching age 16 (‘rising 16s’) from the households recruited in Wave 1, (p. 5)
• & Only spells reported as primary economic activity were taken into consideration. Additionally, BHPS collected a job history module where individuals filled out their labour market activity history over the previous 12–18 months. However, these data do not contain the working hours for the additional employment spells and thus cannot be implemented in our analysis because working full‐ or part‐time contribute to two different states in our analysis. (p. 5)
• & In UKHLS, all economic activity states could be reported retrospectively since the last interview (up to nine spells in some waves) as part of the main questionnaire, providing the opportunity to create employment and education histories for individuals who might have missed some waves. We used nine waves of UKHLS to extend the observation window for the original BHPS sample and investigate employment and education careers of younger cohorts born in the 1980s (p. 5)
• & We focused our analysis on people who turned 16 between 1991 and 2008 (recruited as part of the BHPS sample) and followed them for as long as they remained in the study or until Wave 9 of UKHLS, if earlier. The sample is restricted to respondents for whom we can construct monthly employment and education histories between the years they turn 16 and 26. If the respondents missed some annual interviews entirely (including proxy interviews filled out by household members) without reporting any changes in employment retrospectively, but their economic activity status has changed in between, we imputed the start date at the middle of the interval between the two nearest interviews. (p. 5)
• & Sequence analysis represents each individual life course by a string of states and aims to describe and visualise sequences, compare individual sequences and identify the common types of sequences among populations of interest (Abbott, 1995). (p. 5)
• & As time passes, we distinguished between the seven economic activity states that young people could go through: employed full‐time (≥30 h7), employed part‐time (<30 h), full‐time student (including a small proportion of those in governmental training), unemployed, economically inactive (involved in family care or sick or disabled), taking parental leave and self‐employed (including a fraction of cases in UKHLS where working hours were not reported). (p. 5)
• & The algorithm of transforming one sequence into another includes three operations: substitution (one state is substituted with another), insertion (an additional state can be added at any place in the sequence) and deletion (any state can be deleted to make sequences more similar). All operations come at a ‘cost’ which the researcher defines arbitrarily based on theory or empirical estimations. (p. 5)
• & The distance between two sequences is defined by the (p. 5)
• & minimum ‘cost’ of operation that could be undertaken to transform one sequence into another (Abbot & Tsay, 2000). (p. 6)
• & To measure the dissimilarity between individual trajectories, we used the specification of Dynamic Hamming Distance (DHD) measure. DHD compares sequences element‐wise based on a substitution matrix. Substitution costs are not fixed by the researcher, but based on transition rates for each time point which are derived from the data (Lesnard, 2010). By taking into account the timing of transitions, DHD differs crucially to the widely used Optimal Matching technique (OM), which allows for insertion and deletion operations which shifts substantially the timing and keep the substitution costs fixed. As ‘costs’ can vary over time, for example, the transition rate from being a student to entering full‐time employment might be different if we compare these transitions at ages 16, 18 and 22, for example; and they should not be assigned as equal. After applying DHD, we obtain estimates of dissimilarities between individual educational and employment sequences. (p. 6)
• & Dissimilarity estimates are then used in a clustering algorithm to define a typology of representative trajectories. We used the partitioning around medoids algorithm (PAM). A medoid is an object of the cluster for which the average dissimilarity to all other objects in the cluster is minimal. The K‐medoid method is more robust towards outliers compared to the k‐means method as it minimises the sum of dissimilarities as opposed to the sum of squared interval‐scaled distances (Kaufman & Rousseeuw, 2009). It selects k representative medoids to split the data into k final clusters. (p. 6)
• & To select an appropriate number of k clusters, we followed a three‐stage approach. First, we analysed dendrograms produced from applying Ward's hierarchical clustering algorithm to identify natural breaks in the data. (p. 6)
• & Second, we computed the Studer et al. (2011) discrepancy measures of a set of sequences—pseudo F and pseudo R2 to compare the goodness of cluster solutions (Table A2 in the Appendix). Based on the distance, size and discrepancy parameters of cluster solutions, six and seven cluster solutions were chosen as the number of splits for the PAM algorithm. (p. 6)
• & Third, we compared the silhouettes of six and seven cluster solutions.8 The six‐cluster solution is presented in the paper as it produced more distinct clusters with higher silhouette width parameters. (p. 6)
• & We also controlled for whether a person had a child by age 26 as earlier childbearing can be associated with lower socioeconomic prospects (the mean age at childbearing for the cohorts born post‐1974 was over 30, Office for National Statistics, 2022). (p. 7)
• & TABLE 1 Mean time (in months) spent in each labour market activity state and mean number of spells by cohort. (p. 7)
• & Our study has some limitations. First, less than half of the chosen sample members transitioned from the BHPS to the UKHLS study. This poses a lesser challenge for the older cohorts, given that most of their 10‐year sequences fall within the BHPS lifespan. However, estimates for the youngest cohort may be affected by panel attrition as they experience more turbulent transitions and, on average, unemployed or inactive individuals, or those in precarious employment are more likely to discontinue their participation in the study. (p. 14)
• & Additionally, we recognise a lower representation of young people from less advantaged background in our sample. Together, these factors may result in an underestimation of the proportion of young individuals experiencing turbulent and precarious labour market sequences, particularly within the youngest cohorts. (p. 14)
• & Second, limited representativeness of ethnic minorities in BHPS prevented us from exploring differences in the transition from school to work by subgroups. Third, small sample sizes prevented us from analysing important characteristics of employment, such as the number of job changes within the trajectories, types of contracts or sector and industry. (p. 14)

⛔ Weaknesses and caveats

• ! The restructuring of the labour market in the United Kingdom has significantly disadvantaged the employment prospects of young people from lower socioeconomic backgrounds (Schoon, 2020). (p. 4)