The impact of the Great Recession on educational pursuits in adulthood in the US

Abstract

Economic downturns are known to spark periods of increased enrolment in traditional educational pursuits. The current study leverages 30-year longitudinal data from the Longitudinal Study of American Life (LSAL; N=1,556) to examine individual characteristics and experiences in adolescence, just prior to the Great Recession, and during it, to understand why some individuals chose to pursue new education or training in response to the recession whereas others did not. Indicators from adolescence include measures of self-esteem, locus of control, persistence, achievement in mathematics and achievement in science and were collected from 1987 to 1993. inclusive. Pre-recession indicators include level of education, occupational and marital status and were collected in 2007. Indicators of the impact of the recession were collected retrospectively in 2014 and include whether a job was lost, whether work hours were reduced, and whether there was difficulty making rent/mortgage payments. Binary logistic regression identified persistence in adolescence, pre-recession education level, reporting reduced hours and difficulty paying rent/mortgage during the recession as associated with the likelihood of pursuing new education during the recession. A follow-up analysis investigated whether the pursuit of additional education/training in response to the recession predicted the likelihood of being employed in 2017. Results indicate that obtaining new education during the recession was associated with later employment status, but the significance and direction of the effect depends on pre-recession education level. Implications of this longitudinal, life course analysis are discussed in addition to recommendation for future directions.

Introduction

The Great Recession, which technically lasted from December 2007 until June 2009 in the United States, created a range of occupational and financial challenges. As noted by Katz (2010) in his testimony to the US Congressional Joint Economic Committee, the Great Recession (GR) marked the strongest labour market deterioration since the Great Depression. The US unemployment rate was 4.8% in the last quarter of 2007 and jumped to 10.0% in two years, not taking underemployment into account. There is a lot work that illustrates the negative impact economic downturns have on individuals, such as increased alcohol-drinking behaviour associated with higher national unemployment rates (Dávalos et al, 2012) and sleep disturbances being linked to worry about financial difficulties (Hall et al, 2008; Dregan and Armstrong, 2009). Research by Cagney et al (2014) found that older adults living in neighbourhoods with higher foreclosure rates during the GR were more likely to experience the onset of depression. There is also evidence that an individual’s level of life satisfaction does not fully recover after experiencing unemployment (Lucas et al, 2004).

Economic downturns and education

Although economic downturns are rightfully viewed as a negative force in people’s lives, there is substantial evidence that adults are more likely to pursue education and training in these times than in those of economic prosperity (Taylor et al, 2010; Barr and Turner, 2013). Although the decision to obtain additional training due to the GR may result from challenging circumstances, it is possible that individuals will experience measurable benefits from their additional education in the long run. Research on level of education shows that it is a critical predictor for a wide range of developmental outcomes across the life course, including physical and mental health (Everson et al, 2002; Poulton et al, 2002; House et al, 2005), cognition (Plassman et al, 1995; Ardila et al, 2000; Stern, 2002; Langa et al, 2017), and occupational and financial status (Lynch, 2003; Autor, 2014). However, much of this work on level of education is focused on traditional degree acquisition and often fails to consider additional education pursued throughout adulthood.

Previous research has noted that education and training from multiple sources across the life course is necessary to maintain adequate workforce development in the current global, technology-driven labour market (Jacobs and Hawley, 2009; Katz, 2010). For example, Jacobs and Hawley (2009) report that sources of education and training for sufficient workforce development should come from schools and agencies before individuals enter or re-enter the labour market, from organisation-provided opportunities to improve skills, to organisation-level development to cope with changing market demands, and finally to adult education, which can assist individual workers as they age and transition in and out of the workforce. Jacobs and Hawley note that broad conceptualisations of education and training from a variety of sources are considered necessary to keep individuals competitive for jobs and to keep organisations functioning efficiently to meet society’s needs. In spite of the GR creating a financial and psychological strain for many adults in the United States, those for whom it was a catalyst to pursue new education/training may ultimately benefit in the long term.

Early life characteristics and educational and occupational outcomes

Most research that addresses questions about education during economic downturns tends to focus on information that is temporally proximal to the downturn, such as individual’s pre-recession education level or the impact of recent policy changes (see, for example, Barr and Turner, 2015). However, it is known that individuals do not enter adulthood free of important influences from early life. For example, locus of control – the extent to which individuals believe they can exert an influence through their own behaviour on the events that happen around them – measured at age ten by Gale et al (2008) was significantly associated with a variety of health outcomes 20 years later. Similarly, evidence suggests that outcomes linked to education and employment are likely to have important roots in childhood and adolescence. Using data from the 1970 British Cohort Study, Flouri (2006) found that locus of control in childhood was significantly associated with educational attainment at age 26, highlighting the potential influence of early life characteristics on later educational outcomes. Looking within adolescence, Ross and Broh (2000) found that locus of control was a significant predictor of later scholastic achievement, but that the same was not true for self-esteem. The effects of individual characteristic measures have also been shown to matter within adulthood, such as work by Judge and Bono (2001) who conducted a meta-analysis and found that both job performance and job satisfaction are significantly associated with locus of control and self-esteem.

Recent research on the concept of grit, driven in large part by Duckworth (Duckworth et al, 2007; Duckworth and Quinn, 2009), has illustrated the potential impact of trait persistence and passion in pursuing goals and overcoming obstacles. In the context of education, Wolters and Hussain (2015) examined facets of grit as they relate to self-regulated learning and academic achievement in college students. They found several aspects of grit were associated with self-regulated learning, which was then linked to achievement. Interesting questions remain about how grit, and particularly the facet of grit related to persistence (Duckworth and Quinn, 2009), affected decisions made in response to the GR, an event that probably created substantial challenges. Recent work by Rimfeld et al (2016) examined grit and found that it has extensive phenotypic and genetic overlap with conscientiousness, which potentially explains why grit has been associated with goal achievement and may probably play an important role in outcomes across the life course. In general, there is evidence that early life experiences and traits may play important roles in adulthood when faced with challenges. Longitudinal data that captures early life experiences paired with information about the recent GR provides a unique opportunity to investigate these issues.

Multilevel impacts on lifelong learning

The literature on adult education has focused on identifying critical barriers to adult learning and investigating approaches to address these barriers (Cross, 1981; Darkenwald and Merriam, 1982; Merriam, 1984; Boeren, 2016, 2017). Boeren (2017) proposed a multilevel framework of influences on adult learning that has synthesised much of this work. The micro level includes individual factors such as previous educational experiences and personality characteristics (such as persistence or perseverance) that may differentially affect motivation to pursue new education in adulthood. Mesolevel influences are structural, such as the extent to which a learning environment allows for flexibility, which is often needed for adult learners who may have caregiving responsibilities and less free-time. The macro level is grounded at the policy level, and within the US can have impacts from both the state and federal levels (such as due to funding decisions, whether educational activities are made compulsory). Elman and Weiss (2014) note that within the US the adult education system is fairly ‘age-blind’, as shown by the gains in midlife adults reporting participation in education pursuits over the last several decades. This suggests a culture where pursuing new education in adulthood is fairly normative, and where a major event such as the GR may provide an unexpected opportunity for individuals to obtain new training or degrees.

Based on Boeren’s multilevel model of lifelong learning, the extent to which individuals actually made the decision to obtain new education in or after the GR may be related to a variety of individual, structural and policy-related forces. These forces may work for or against educational pursuits. If individuals are newly unemployed or underemployed, they may have more time available for training and education, but also fewer financial resources to support these goals. It is also possible that during the GR employers cut back on work-sponsored education and training, an impact from the mesolevel that would limit opportunities for education.

Life course developmental timing of the Great Recession

Another important factor to consider is the extent to which individuals’ age at the time of the recession may have an impact on the likelihood that they chose to pursue new education or training. As noted by the seminal work of Elder (1998), a careful consideration of the intersection between period effects (the recession, for example) and cohort experiences can provide important understandings of developmental phenomena. For example, when it comes to making a decision about whether to go back to college for a new degree or certificate, the age and previous experiences of an individual probably plays an important role. Younger adults in their twenties are more proximal to traditional education ages and also have a potentially long period of work years in their future. For these individuals, the decision to go back to college may have fewer perceived barriers compared to potential future benefits. Alternately, older adults in late midlife or in early late life are perhaps decades away from their last formal educational experiences, are decades older than traditional education ages, and have presumably fewer future work years remaining. For this group, the decision to pursue additional education during the GR may be less likely to be due to strong perceived barriers and diminished long-term occupational benefits. Generation X adults who were in their thirties during the Great Recession (that is, born in the 1970s) are likely 10 to 20 years away from their last formal educational experiences as well as from traditional college ages, but they also have roughly half of their work lives ahead of them. For this group, there may be important perceived barriers to going back to college, but there are also potentially important long-term benefits.

Current study

The current study, as well as two others in this special issue (Miller and Cepuran, 2019; Tang and Miller, 2019), capitalises on unique data from the Longitudinal Study of American Life (LSAL). It spans 30 years from adolescence to midlife, used here to address two primary research questions. First, what life course factors are associated with the likelihood that an individual engaged in new educational pursuits during the GR? Specifically, sets of factors from adolescence, pre-recession (2007), and during the recession (2008–09) are considered as potential predictors of the likelihood that an individual in their thirties chose to pursue additional education or training during the recession. This study will be one of the first to incorporate an extended set of life course predictors from adolescence through to midlife to investigate impacts of the GR. Importantly, the current study is also inclusive of any form of educational pursuit, rather than just focusing on post-secondary and graduate degree work. Second, did those who pursued additional education and/or training during the GR have an increased likelihood of being employed several years later? After adjusting for the life course factors found to be linked to educational pursuits in the first research question, this study then tests whether this additional training is associated with being employed in 2017.

Method

Sample

The current study used longitudinal data from the LSAL study, which was first collected in 1987 when participants were age 13 or 17. The study was originally conceived to longitudinally track the impact of education and achievement in seventh- and tenth-graders in 104 public schools across the United States. The original sample included 12 study strata in four geographic regions and identified three urban development levels (central city, suburban, non-metropolitan). The sample was followed annually through to 1993, then again annually from 2007 to 2011, and most recently annually from 2014 to 2017. Academic achievement and test measures were collected by self-report paper questionnaires given to the student participant, teachers and parents. Response rates for the study have been relatively high, ranging from 89% during the school year waves to 78% through 2011. Online surveys have been introduced to reflect participant preferences and to sustain the response rate.

To include pre-recession constructs in the analysis, the current study selected those individuals for whom their 2007 wave responses were prior to the official onset of the Great Recession. As such, a recession-based weight was applied to all analyses where a non-zero value indicated that the 2007 interview was completed by the first quarter of 2008 (N=3,168). The weight used in the current analyses adjusted for the complex sample design and reflects the national distribution of these birth cohorts in the United States. Additional selection criteria included having a valid response on the question about pursuing education/training since the recession (N=2,216), and non-missing values on other predictors and covariates included in the models resulting in a final analytic sample of N=1,556. Selection checks found that the analytic sample did not significantly differ from the full sample in terms of gender distribution (x²(2, N=3,168) = .19, p = .66), but did have a significantly higher level of education in 2007 (x²(6, N=3,168) = 118.78, p < .001) as well as higher academic achievement in science (t(3140) = 13.97, p < .001) and mathematics (t(3133) = 14.74, p < .001) in high school.

Table 1 summarises the unweighted descriptive characteristics of the final analytic sample.

Table 1:

Unweighted sample descriptives (N=1,556)

% Women	53.4
% Married pre-recession	70.9
% 2010 household children under age 6	35.2
Highest parental education
% Less than high school	3.6
% High school	44.7
% Some college	13.6
% Four-year college degree	20.4
% Advanced degree	17.5
Adolescent achievement and traits (mean (SD))
Self-esteem	7.3 (1.5)
Science achievement	67.5 (10.6)
Maths achievement	70.6 (12.3)
Locus of control in tenth grade	3.3 (0.6)
Persistence in tenth grade	3.6 (0.7)
% Pre-recession health limitation	3.9
Pre-recession education
% Less than high school	1.9
% High school	33.7
% AA-AS	7.7
% Baccalaureate	35.7
% Master’s	15.7
% Doctorate/professional	5.5
Pre-recession occupational status
% Management, business and finance	22.5
% Professional	32.5
% Service	7.5
% Sales and office	16.9
% Natural resource, construction and maintenance	2.9
% Production and transport	7.5
% Out of workforce	10.4
Recession context
% Lost job	11.9
% Fewer work hours	14.5
% Difficulty paying rent/mortgage	29.0
% Pursued education/training	19.3

Note: AA-AS refers to an Associate of Arts or Associate of Science degree.

Measures

Adolescent achievement and traits

Several constructs related to achievement and individual traits were created based on student and teacher self-report data collected while the participants were in high school. In terms of achievement, scores of overall maths and science achievement are included which represent the achievement level from an individual’s final year of high school, which was grade 12 for those who graduated, or from an earlier grade if they did not complete high school. The achievement scores are based on percentile scores on tests of mathematics and science using item sets from the US National Assessment of Educational Progress (NAEP). Considering trait characteristics which were assessed in adolescence, indicators of self-esteem, locus of control and persistence were included. The self-esteem measure was adapted from Rosenberg (1979) and was assessed via seven items on a five-point Likert scale. An example item is, ‘I take a positive attitude toward myself.’ Respondents were asked the extent to which they agreed with this statement from 1 = strongly agree to 5 = strongly disagree. This scale was asked in each year of high school and an average from these individual sum scores across those years was used in the current study with a resulting range of 0–10. In tenth grade, the students also completed scales measuring locus of control and persistence, in much the same way as self-esteem. Locus of control was assessed via five items on a five-point Likert agreement scale. An example item is, ‘Good luck is more important than hard work for success.’ Persistence was assessed by four items on the same five-point Likert agreement scale. An example item is, ‘I would rather keep struggling with a problem than give up on it before I get the answer right.’ Items were scored such that higher scores indicated higher persistence and external locus of control. Persistence and control were averaged across items, with ranges of 1–5 for both scales.

Pre-recession context

To assess the impact of individual’s lives just before the GR on their later decisions to pursue education during the recession, information from the 2007 wave on occupational status, marital status and highest educational degree was included. Specifically, occupational status was self-reported in an open-ended field in the survey which was then coded down into seven possible categories. The categories include: management, business and finance; professional; service; sales and office; natural resource, construction and maintenance; production and transportation; and out of the workforce. For analyses, out of the workforce was used as the referent category. Marital status was coded as 1 = married/partnered and 0 = not married/partnered. Highest educational degree included six categories: less than high school; high school; associate’s degree; baccalaureate degree; master’s degree; and doctorate or other professional degree. All analyses used high school as the reference category.

Recession context

To assess the impact of the GR, participants were asked in the 2014 wave to assess retrospectively the extent to which they experienced a variety of challenges. Before answering the questions, the survey instruction first introduced the concept of the GR with the following language:

In recent years, we have asked you a series of questions about your educational activities and your employment. One of the major issues concerning Generation X is the impact of the Great Recession on the lives of young adults in our generation. The following sets of questions concern some possible negative effects of the Great Recession. We need your answers to these questions even if you did not personally suffer major negative effects. Please note that the Great Recession started in December of 2007 and technically ended in June of 2009, but many of the effects have been longer lasting for individuals, such as the loss of a job or mortgage. We realize that we are asking you to recall events from the last six years and urge you to make your best estimate.

The current study considered four variables derived from this section of questions. First, the primary outcome variable for this study is whether an individual pursued additional education or training since the start of the GR. The question asked, ‘Since the beginning of 2008, have you enrolled in additional education/training courses to improve your job skills?’ Responses of ‘Yes’ were coded as 1 and ‘No’ were coded as 0. In the final analytic sample, 19.3% of respondents indicated they did enrol in some form of education or training. To assess whether additional recession experiences were associated with the choice to pursue additional education, the current study also included whether a person reported losing a job due to the GR, whether they experienced a significant reduction in hours of work due to the GR, and whether the individual experienced difficulty in making rent or mortgage payments due to the GR. All items were coded as 1 = yes and 0 = no.

2017 employment status

To address research question 2, the current study includes a measure of employment status in 2017 as an indicator of the long-term impact of recession-related educational pursuits. This variable was coded as 1 = currently employed and 0 = not employed.

Additional covariates

In addition to the variables already described, all analyses also included gender and highest parental education level. Gender was coded as 0 = man and 1 = woman. Highest parental education level is a five-category variable that represents the highest degree attained by either parent, including categories for less than high school, high school, some college, a four-year degree and advanced degree. All analyses use the high school category as the reference category. Presence of children in the household under age six, assessed in 2010, was included as having a young child/children at home may have impacted an individual’s ability to pursue education during the GR. An indicator that chronic health problems limited an individual’s ability to work prior to the GR (in terms of the type or amount of work) was also included, as these health limitations may have also impacted the ability to pursue education. Specifically, individuals were coded as having a pre-recession health limitation if they responded yes to either of the questions, ‘Are you limited in the kind of work you can do on a job because of your health?’ or ‘Are you limited in the amount of work you can do because of your health?’

Analysis plan

Binary logistic regression was used to test research question 1. Specifically, proc surveylogistic was used with SAS version 9.4 to model the likelihood that an individual reported engaging in new education/training since the recession. Several sets of predictors were included: covariates (gender, parental education), adolescent achievement and traits, pre-recession context and the recession context. Results are reported in odds ratios and upper and lower 95% confidence limits. The C-statistic is also reported as an indicator of the overall model predictive value and is a measure of the area under the Receiver Operating Characteristic (ROC) curve (Hosmer and Lemeshow, 2000). A C-statistic of .5 indicates a binary logistic model that predicts no better than chance. Binary logistic regression was also applied to address research question 2, whether pursuing additional education/training during/since the recession is associated with the likelihood of employment in 2017. This analysis was done on all cases that had a valid response in the 2017 wave regarding their employment, with a resulting N=1,360. As a follow-up analysis to understand whether the effect of recession education differed by an individual’s pre-recession educational status, a follow-up analysis was tested with proc surveylogistic that included an interaction term between Recession Education Pursuits and Pre-recession Education, both of which were treated as class variables. An LSMEANS statement was included to obtain individual odds ratios testing the difference across combinations of the class variables with a Bonferroni adjustment for multiple comparisons.

Results

Likelihood of pursuing education/training

Results from the logistic regression investigating the likelihood an individual engaged in new education/training since the recession are reported in Table 2. Overall, the analysis identified four factors associated with the choice to engage in additional education. Although there was little evidence that adolescent scholastic achievement in maths and science played a role, nor did trait indicators of adolescent self-esteem and locus of control. A measure of persistence (that is, grit) from tenth grade was a significant predictor. Specifically, an individual had 1.3 higher odds of pursing new education or training during the recession with every increased unit¹ of persistence assessed in adolescence (Odds Ratio = 1.35 (1.04–1.75); p < .05). Pre-recession occupational status was not significantly associated with the likelihood of pursuing education in the current study, however pre-recession educational status was a significant predictor. Specifically, compared to those who had a high school education prior to the Great Recession, those with a doctorate or other professional degree were significantly less likely to pursue additional education since the recession (Odds Ratio = 0.16 (0.05–0.56); p < .01). Finally, two recession-specific contextual factors were significantly associated with the likelihood that an individual obtained new education. Individuals who reported fewer work hours (Odds Ratio = 1.67 (1.04–2.67); p < .05) or who reported difficulty paying their rent or mortgage due to the recession (Odds Ratio = 1.57 (1.09–2.71); p < .05) were significantly more likely to choose to get more education or training. Overall, the model had a C-statistic of 0.64, reflecting the area under the ROC curve.²

Table 2:

Binary logistic regression results on the likelihood of pursuing education during the Great Recession (N=1,556)

	Odds Ratio	Lower CL	Upper CL
Women	0.80	0.58	1.12
Married pre-recession	0.86	0.60	1.23
2010 household children under age 6	0.74	0.51	1.08
Highest parental education
Less than high school	1.20	0.52	2.79
Some college	1.33	0.83	2.13
Four-year college degree	1.16	0.74	1.81
Advanced degree	1.12	0.67	1.83
Adolescent achievement and traits
Self-esteem	1.03	0.92	1.16
Science achievement	1.01	0.99	1.04
Maths achievement	0.98	0.96	1.00
Locus of control in tenth grade	0.94	0.72	1.24
Persistence in tenth grade	1.35*	1.04	1.75
Pre-recession health limitation	0.55	0.24	1.26
Pre-recession education
Less than high school	0.27	0.05	1.47
AA-AS	1.68	0.96	2.93
Baccalaureate	1.36	0.87	2.14
Master’s	1.67	0.92	3.03
Doctorate/professional	0.16**	0.05	0.56
Pre-recession occupational status
Management, business and finance	0.73	0.40	1.33
Professional	0.82	0.45	1.50
Service	1.13	0.55	2.31
Sales and office	0.90	0.49	1.66
Natural resource, construction and maintenance	0.72	0.26	1.96
Production and transport	0.70	0.31	1.55
Recession context
Lost job	0.82	0.51	1.34
Fewer work hours	1.67*	1.04	2.67
Difficulty paying rent/mortgage	1.57*	1.09	2.71
C-statistic	0.64

Notes: Reporting odds ratios (95% confidence intervals).

Highest Parental Education referent group is high school degree; 2007 Pre-recession Education referent group is High School degree; Pre-recession Occupational Status referent group is Out of Workforce.

C-Statistic is an indicator of model predictive value, where levels over 0.5 indicate some predictive value.

^*p < .05,

^**p < .01.

AA-AS refers to an Associate of Arts or Associate of Science degree.

To address potential bias due to missing data, a follow-up analysis applying multiple imputation for missing predictor variables was run with SAS proc MI with a multivariate normal distribution. Specifically, the base sample included those who had a valid response to the primary outcome variable regarding the pursuit of education since the recession (N=2,721). The multiple imputation model specified ten imputations to generate estimates for missing data on the predictor variables, which were then analysed with logistic regression. Proc mianalyze was then run to pool the parameter estimates across the logistic regression analyses of the ten imputed datasets. The pooled results revealed a similar pattern of results as those reported in Table 2, with the only difference being that working fewer hours during the GR was no longer significantly associated with the likelihood that an individual pursued education during the recession.

Likelihood of employment in 2017

Results from the logistic regression analysis modelling the likelihood that an individual was employed in 2017 are reported in Table 3. Based on the previous analysis, the logistic regression analysis dropped all adolescent achievement and trait variables aside from persistence, and also did not include the recession context variable indicating a job was lost. While gender was not associated with the likelihood that an individual chose to purse new education since the GR, the current analysis found that women were significantly less likely to be employed in 2017 compared to men (Odds Ratio = 0.43 (0.23–0.80); p < .05). Although there were no significant main effects of adolescent persistence, pre-recession education, or any of the recession context variables including whether new education was pursued, significant effects of pre-recession occupational status were identified. Specifically, individuals who reported working in natural resource, construction and maintenance jobs (Odds Ratio = 0.15 (0.05–0.43), p < .01) as well as those who worked in production and transportation (Odds Ratio = 0.14 (0.06–0.30), p < .01) were significantly less likely to be employed in 2017 compared to those who were out of the workforce pre-recession. The C-statistic for this mode was 0.74.

Table 3:

Binary logistic regression results on the likelihood of being employed in 2017 (N=1,268)

	Odds Ratio	Lower CL	Upper CL
Women	0.43*	0.23	0.80
Married pre-recession	1.19	0.69	2.05
Persistence in tenth grade	0.87	0.60	1.26
Pre-recession education
Less than high school	0.97	0.24	3.41
AA-AS	1.58	0.44	5.75
Baccalaureate	1.00	0.56	1.79
Master’s	2.15	0.89	5.18
Doctorate/professional	1.14	0.27	4.88
Pre-recession health limitation	0.21***	0.09	0.49
Pre-recession occupational status
Management, business and finance	0.63	0.25	1.58
Professional	0.71	0.20	2.54
Service	0.93	0.37	2.34
Sales and office	0.49	0.09	2.84
Natural resource, construction and maintenance	0.15**	0.05	0.43
Production and transport	0.14**	0.06	0.30
2010 household children under age 6	1.31	0.75	2.27
Recession context
Fewer work hours	1.11	0.52	2.39
Difficulty paying rent/mortgage	1.04	0.59	1.82
Pursued education	0.64	0.31	1.33
C-statistic	0.74

Notes: Reporting odds ratios (95% confidence intervals).

2007 Pre-recession Education referent group is High School degree; Pre-recession Occupational Status referent group is Out of Workforce.

C-Statistic is an indicator of model predictive value, where levels over 0.5 indicate some predictive value.

^*p < .05,

^**p < .01,

^***p < .001.

AA-AS refers to an Associate of Arts or Associate of Science degree.

A follow-up analysis tested whether there was a significant interaction between recession educational pursuits and pre-recession education level, since pre-recession education was a significant predictor for research question 1. Results found that the interaction was indeed significant (p < .001). Those who had less than a high school degree before the recession and did not pursue new education/training during the GR were significantly less likely to be employed compared to those who did get additional education (Odds Ratio = < .001; p < .001). For those with a high school diploma before the recession, pursuing education during the recession was not associated with the likelihood of employment in 2017 (Odds Ratio = 0.64; p = 0.34). The likelihood of employment in 2017 did significantly differ among those with an associate’s degree prior to the recession based on whether they chose to get new education or not. Counter to expectations, those who chose not to get additional training were significantly more likely to be employed in 2017 compared to those who did pursue more education (Odds Ratio = 16.37; p < .05). The choice to get additional education/training since the GR did not affect the likelihood that individuals with a baccalaureate degree prior to the GR were employed in 2017 (Odds Ratio = 0.18; p = 0.12). Finally, both those who had a master’s degree and those with a doctoral/professional degree prior to the recession were significantly less likely to be employed in 2017 if they did not pursue additional education compared to their counterparts who did (master’s Odds Ratio = < .001; p < .001; doctoral Odds Ratio = < .001; p < .001).

Discussion

The first research question of current study investigated factors associated with the likelihood that individuals in their thirties chose to pursue new education and/or training as a result of struggles during the GR. This sample is part of a unique 30-year longitudinal study of the Generation X cohort, which provided the opportunity to examine potential predictors from different phases of the life course. Specifically, scholastic achievement as well as trait measures of self-esteem, locus of control and persistence were assessed when the respondents were in high school. In adulthood, measures of 2007 pre-recession education level, occupational status and marital status were included, as well as retrospective indicators of the impact of the recession (reported in 2014) which measured whether a job was lost, whether work hours were limited, and if there was difficulty making rent/mortgage payments. Generation X adults in their thirties during the time of the GR are particularly interesting to consider as they are probably many years from their most recent traditional educational experiences, but also have many years of work ahead that can benefit from additional skills and degrees.

The results indicated that pursuing additional education in or after to the recession was driven by factors linked to various points in the life course. First, the trait of persistence measured in tenth grade was a significant predictor of whether individuals decided to pursue education since the GR while aged in their thirties. This result is somewhat remarkable in that it was significant over and above other educational and occupational indicators. Persistence is very similar to the concept of grit, which has recently been identified as an important predictor of goal achievement, over and above intelligence and trait conscientiousness (Duckworth et al, 2007). Within the context of the current study, an economic downturn such as the Great Recession certainly places a strain on individuals and their families. The results of the current paper suggest that Generation Xers experiencing economic difficulty were substantially more likely to use the GR as an opportunity to obtain new skills via education and training if they were persistent as adolescents. New work finds too much persistence can be bad in cases where letting go of a goal is healthier than refusing to give up (Lucas et al, 2015), which may be an interesting avenue for future research on recession-linked coping behaviour.

Unsurprisingly, the current study found that individuals who had a doctoral or other professional degree prior to the recession were significantly less likely to pursue additional education at or after the recession compared to those with a high school degree. In many ways, this could be considered a ceiling effect, although even those with terminal degrees in their fields often have opportunities for additional training and certifications. The current study also identified that individuals who reported reduced hours due to the GR and those who struggled to pay their rent or mortgage were significantly more likely to pursue additional education and training. These recession context indicators relate to financial hardship and suggest that individuals who were most financially impacted by the Great Recession are those most likely to search for new skills to improve their future occupational prospects.

The second research question asked whether the decision to obtain additional education/training at or after the GR was associated with a higher likelihood of being employed several years after the recession for those who are part of the labour force. The analysis identified that education obtained during the recession was linked with later employment in 2017 depending on one’s pre-recession education level. There was no main effect of GR educational pursuits on 2017 employment, unless the estimate was interacted with pre-recession schooling. Those who did not pursue more education had lower likelihoods of later employment compared to those who did if their pre-recession educational status was less than a high school diploma, a master’s degree or a doctoral/professional degree. Those with an associate’s degree (7.7% of the sample) had an effect in the opposite direction, such that those who did not pursue additional education at or after the GR were more likely to be employed in 2017. The likelihood of employment in 2017 did not significantly differ among individuals with a high school diploma or a baccalaureate degree prior to the GR, based on their choice to pursue education/training during the recession. These results paint a complicated picture regarding the long-term impact of recession-based education.

While the current study benefited from an inclusive definition of educational activities, which could include formal degree work, vocational and certification work, as well as less formal training opportunities, one limitation is that independent effects of the various education pursuit types were not tested. It is possible that the effects of these educational experiences on later employment status differ. Further, these effects may differ based on an individual’s current occupation and/or educational status. Knowledge gained from additional research to understand how the different sorts of educational pursuits uniquely contribute to employment outcomes would help to deconstruct the results from the current study. For example, a follow-up study could examine whether completing a formal college degree has the same impact on later employment as getting a trade certification. Another limitation was that the only future outcome tested was employment status in 2017. While this is a useful outcome to determine if recession-linked education was beneficial to long-term employment, there are many other indices that would also be worth investigating. For example, perhaps different results would be found if we looked at income or earnings in 2017 beyond simple employment status. It would also be worth investigating the extent to which GR educational pursuits predict later physical and mental health outcomes, as it was noted previously that in general education is positively associated with these domains across the life course. Future studies that prospectively track the impact of traditional education levels as well as education attained later in adulthood may help to answer questions about the processes by which education affects health and quality of life indicators.

Another question about the educational pursuits individuals undertook during and immediately following the recession is whether they were done in traditional in-person classroom settings versus the wide variety of online courses now available. An experimental study by Deming et al (2016) found that employers are significantly less likely to show interest in job applicants who have a business bachelor’s degree from a for-profit online institution compared to those with the same degree from a nonselective public institution. These suggested biases against online programmes may be specific to degree programmes and not impact employer’s views on vocational training work or skill trainings. However, not enough is known about these diverse educational options at this point to make recommendations to individuals interested in becoming more appealing to future employers. What is clear is that not all educational pursuits are considered equal in the labour market which is an issue future research should address. The detailed information collected in the LSAL study regarding education and training provides a unique data source for this future work.

The current study supports previous findings that economic downturns are often an opportunity for educational pursuits. And although this can be considered a positive outcome in the face of adversity, the long-term impact is yet to be fully understood. The current paper found mixed evidence that obtaining additional skills since the recession was beneficial for later employment. As noted earlier, income may be a better marker of the impact of these pursuits. There is also some concern that, while pursuing additional education is in general beneficial to individuals, there are potential costs worth consideration. Barr and Turner (2013) note that enrolment in post-secondary education increased across many age groups during the GR at a time when financial resources at public and non-profit colleges and universities are extremely strained due to diminished state support. In response to this, many colleges and universities chose to raise tuition fees to combat financial deficits, which adds to the probable debt burden of individuals who enrol at these institutions. A careful consideration of the benefits and costs of post-secondary education during economic downturns for individuals and institutions is necessary before suggestions of policy adaptations are made.