The impact of the Great Recession on Generation X

Abstract

For people born in the 1960s and 1970s, sometimes known as Generation X, the Great Recession (GR) was the first major economic downturn that they have experienced. Using the 30+ year record of the Longitudinal Study of American Life (LSAL), this analysis examines the economic impact of the GR on an American cohort born in 1972–75 who were in their mid-40s in 2019. A confirmatory factor analysis index was constructed to summarise the economic experience of each LSAL participant in the period from 2007 (the eve of the GR) through 2014. Most of the LSAL participants did not experience negative economic consequences, but a significant subset of participants experienced substantial negative effects. A structural equation model (SEM) was used to estimate the relative influence of several educational and family variables on the nature and magnitude of the GR on employment and related economic issues. Educational attainment was the strongest predictor of the economic impact of the GR on individuals. The educational attainment of the parents of LSAL participants was the second strongest predictor, indicating the inherited advantages of social class. The impact of the GR on participants was unrelated to gender, but African-Americans were more likely to experience negative economic consequences from the GR than other young adults, holding constant differences in educational attainment, parent education attainment and other factors.

Introduction

Generation X is a label, at least in the US, for people born 1961–1980, between the Baby Boom, and those labelled Millennials (Strauss and Howe, 1991; Miller and Laspra, 2017). For this Generation X, the Great Recession (GR) that began in 2008 was the first major economic downturn that they had experienced in their lives. Most of these young adults in 2008 had spent their adolescence in the calm of the Reagan years and their college years during the economically prosperous Clinton Administration. For most Generation X adults, the turmoil of the Vietnam War years occurred before they were old enough to become politically aware and active and before they contemplated entering the workplace. Generation X is an important segment of the American population. This generation have completed the initial stages of their education, have entered into their occupational or profession track, have started a family if they plan to do so, and have become active in the political system if they intend to engage politically. Barack Obama was the first member of Generation X to be elected president and the number of members of the Congress – especially the House of Representatives – from Generation X is growing rapidly. A large number of newly elected House members in the 2018 US mid-term elections come from Generation X. The members of Generation X will dominate American society and politics for the next two or three decades.

Although some economic problems became apparent in 2007, the full impact of the economic downturn that we now call the Great Recession did not hit until the second half of calendar 2008. In the last months of 2008, the American economy was losing 800,000 jobs each month and this economic decline continued until the middle of 2009. The passage of the American Recovery and Reinvestment Act of 2009 (ARRA) in mid-February – about a month after President Obama took office – signalled the new administration’s intention to pump new spending into the economy but it took another four or five months until the loss of jobs stabilised. In the years since the beginning of the stimulus programme and related recovery measures, there has been substantial improvement in employment and economic activity, but the speed of recovery has been slower than many economists and political leaders would have liked and there are still groups and areas that have not experienced significant recovery (Pfeffer and Hertel, 2015; Kalleberg and von Wachter, 2017; Sironi, 2017).

A more precise look at the impact of the Great Recession

The preceding broad discussion is characteristic of journalistic or political campaign language, but it is useful to make a more precise analysis and description of the impact of the GR on Generation X. This paper follows a central cohort from the broader Generation X who are participants in the Longitudinal Study of American Life (LSAL), a national longitudinal study of seventh- and tenth-grade public school students selected in autumn 1987, who were in their early to mid-thirties when the GR struck. The LSAL includes important indicators of the scope and magnitude of the impact of the Great Recession on the lives of these young adults (Miller and Laspra, 2017). Readers should not see ‘young adults’ as referring to a specific age or age range, but to a cohort of participants as they move from secondary school into post-secondary institutions and into midlife work and family responsibilities

Although the negative impact of the Great Recession was substantial for a significant number of adults in Generation X, it is important to recognise that many members of this generation were untouched by the economic downturn. The broad journalistic statements about the impact of the GR and the slow recovery paint an incomplete picture. As in most generational analyses, it is important to recognise that there is a wide range of experiences and consequences within any generation.

Analysis plan

This analysis will begin with the construction of a summary measure of the impact of the GR. In the 2008 and subsequent annual LSAL surveys, a number of questions were asked to determine both positive (continued full-time employment, bought or continued payments on a home, saved money from current income, bought a car, took a vacation) and negative (lost a job, lost one’s home to mortgage foreclosure, unable to pay medical bills, lost health insurance, depleted personal savings) economic experiences during the eight years following the onset of the Great Recession. Some questions were repeated annually; others were not. A confirmatory factor analysis is used to compute a summary scale that ranges from +10 to −10, reflecting the net impact of the Great Recession. A listing of the items used in this index are included in Appendix A.

The second step in the analysis will be the identification and measurement of factors that may have influenced or may predict the net impact of the Great Recession on the adults in the LSAL. This discussion will be set in the context of the literature on the impact of the Great Recession and will seek to use comparable measures whenever possible.

The final step in the analysis will be to construct a structural equation model (SEM) to obtain a more rigorous measure of the relative influence of each of the major factors on the impact of the Great Recession on the Generation X adults in the LSAL. Recognising that many readers may not be familiar with general models, this discussion will proceed on parallel levels designed to make the analysis accessible to all readers.

The concluding section will discuss the invaluable role of longitudinal studies in understanding the impact of major social and economic events on society or components of a society. Western economic systems appear to experience periodic economic cycles and it is important to examine the longer-term and formative influence of these experiences on Generation X adults.

Literature review

The GR is still not far behind us. The literature on the impact of the GR on Americans remains patchy; however, several papers provide important insights on the experiences of different groups (Fligstein and Goldstein, 2011; Grusky, Wesgtern, and Wimer, 2011; Hout, Levanon, Cumberworth, and Wimer, 2011; Zick, Mayer, and Glaubitz, 2012).

One of the most highly contested topics in the literature on the GR is the relevance of gender. Some analysts have argued that lower-income men were predominately affected in the GR, but this conclusion is not unanimous. Although the unemployment rate of women appeared to bounce back faster than that of men, one theory is that this is due to women’s greater rate of withdrawal from the workforce, rather than because they had some sort of advantage (Farber, 2011). Christensen (2015) supports this theory, arguing that as many fields of work in the United States are still largely gendered and women have less consistent workforce participation rates, a comparison of impact is not easy to draw. Although men lost more jobs than women early in the GR, unemployment figures for women are inaccurate because more women than men are underemployed or give up the work search and leave the workforce. Kalleberg and von Wachter (2017) argue that gender played no significant role in employment outcomes, but that African- and Hispanic-Americans suffered more than other groups during the GR. The most widely accepted explanation is that higher education levels were linked to lower rates of job loss and higher rates of job recovery (Farber, 2011).

In terms of recovery, some analysts conclude that Generation X experienced more economic losses than other groups. When savings and retirement investment behaviours are compared between generations in the aftermath of the GR, Generation X faced greater hardships due to a lack of good retirement investments. Generation X had slightly worse outcomes than the youngest strata of workers (Farber, 2011), but were less likely to have lost jobs in the first place than younger generations in the GR, possibly due to more work experience (Abel and Deitz, 2017).

Although the LSAL contains data only pertaining to Americans of Generation X, it is crucial to keep in mind that the GR was not limited to the United States. Hoffmann and Lemieux (2016) compare the US performance in the GR with Germany and Canada and find that the unemployment rates in the US rose faster and stayed high longer than in Canada and Germany. Sironi (2017) examined similar patterns in the US, the UK, Norway, Germany and Spain and found similar results.

Life in 2007

On the eve of the GR, 87% of LSAL respondents – the centre of Generation X – were employed and 11% were voluntarily out of the workforce. Employment includes full-time and part-time employment for pay, self-employment, and active duty service in the military. Respondents who reported that they were seeking a job or were on temporary layoff and hoping to return were classified as seeking work. Three per cent of these LSAL adults were seeking work, which translates into a federal unemployment rate of 3.1% (see Table 1). This rate of unemployment was slightly lower than the national rate, reflecting the higher level of educational attainment and the age of LSAL participants.

Table 1:

LSAL employment status in 2007 (percentages)

	LSAL employment status 2007				Nb
	Employed	Seeking	Out of workforce	Unemployment ratea
All LSAL adults	87	3	11	3.1	2,453
Education
HS diploma or less	82	5	13	5.8	1,003
Associate degree	82	3	15	3.7	192
BA quintile 5 (lowest)	90	1	9	0.5	236
BA quintile 4	86	1	13	0.5	246
BA quintile 3	90	1		0.0	224
BA quintile 2	91	1	8	1.0	213
BA quintile 1 (highest)	90	1	9	1.4	227
Gender
Female	78	2	20	2.6	1,230
Male	94	3	3	3.3	1,223
Married
Not married	87	6	7	6.9	607
Married	86	1	13	1.6	1,843
Minor children at home
No minor children	86	8	6	8.2	534
One or more minor	86	1	13	1.4	1,917
Urban–rural residence in 2007
Central 1 million+	86	2	12	2.4	418
Suburban to 1 million+	87	2	11	2.2	711
500,000 to 999,999	85	3	12	3.4	133
50,000 to 499,999	84	4	12	4.3	429
Less than 50,000	85	3	12	3.3	755

^aThe unemployment rate is computed using the federal standard – the percentage of individuals in any group that are seeking work, excluding individuals that are not in the work force.

^bThe number of cases will not always equal the total due to rounding errors and a few cases of missing data.

An examination of the level of workforce engagement in various segments of this national sample of LSAL adults found some disparities in employment and unemployment. As expected, the level of educational attainment was positively related to employment (see Table 1). In this analysis, we employ a new measure of educational attainment that differentiates university degrees as a rough estimate of the quality of their undergraduate education. Approximately half of the LSAL cohorts have earned an undergraduate university degree (often referred to as a baccalaureate degree in the US) and there is broad recognition that there is significant variation in the quality of the educational experiences represented by those degrees. For more than 50 years, the US Office of Education has collected information from colleges and universities about the 75th, 50th and 25th percentile scores on the Scholastic Aptitude Test (SAT) or the equivalent score from the American College Test (ACT) for entering students from each institution of higher education. These scores reflect the quality of preparation of students admitted and institutions with more rigorous admission standards tend to graduate students who are more competitive in graduate and professional education and in the workplace. For this purpose, the 75th percentile SAT score for each LSAL participant’s university was used to separate university graduates into five quintiles, reflecting the rigor of their undergraduate college or university. On the eve of the GR, the results indicate that all university graduates tended to be employed at about the same rates and that only about 1% of university graduates were seeking employment in 2007. Later analyses will show that this measure of educational attainment is useful in predicting the net impact of the GR on LSAL adults.

Gender was strongly related to workforce engagement prior to the GR, with 94% of LSAL men reporting employment, compared to 78% of LSAL women (see Table 1). This difference reflects a larger proportion of women who were out of the workforce, often to have and care for children. Among those LSAL adults who were working or seeking work, the federal unemployment rate was slightly higher for men (3.3%) than for women (2.6%).

Although the rate of employment in 2007 was virtually the same for LSAL adults who were married and who were unmarried, the unemployment rate was significantly higher for unmarried adults than married adults – 6.9% to 1.6%. This differential reflects the larger number of married LSAL adults who are voluntarily out of the workforce (see Table 1). A similar pattern was found for LSAL adults with minor children at home and those adults without minor children at home – reflecting the differential rate of individuals voluntarily out of the workforce.

An examination of the impact of residential location found that young adults living in smaller cities or towns – urban areas with populations from 50,000 to 499,999 – had a significantly higher rate of unemployment prior to the GR than LSAL adults in central cities, suburbs, middle-sized cities or rural areas (see Table 1). The lowest rates of unemployment were found in suburban areas of major metropolitan areas.

Looking at the broad landscape of employment and workforce participation on the eve of the GR, the results suggest that a high proportion of LSAL adults were gainfully employed or self-employed and that the 11% of LSAL adults out of the workforce reflected personal decisions to not seek employment rather than non-participation forced by economic conditions. Individuals with a university degree were more successful in avoiding involuntary unemployment than LSAL adults with less post-secondary education.

The impact of the Great Recession

The GR began in 2008 and progressed rapidly into 2009. At the end of 2008, the American economy was losing 800,000 jobs each month. The incoming Obama Administration took office in January, 2009, and the Congress passed the American Recovery and Reinvestment Act of 2009 (ARRA) – known as the Stimulus Package – which included $862 billion in new federal spending to support grants for construction and other economic activities that could be initiated quickly. The unemployment rate continued to grow until the late summer of 2009 and began a gradual recovery. The first task of the recovery was to stem the loss of jobs and to begin to rebuild demand and confidence through stimulus spending and other monetary and fiscal policy initiatives.

To measure the initial impact of the GR, it is useful to look at self-reported employment during the period after the initial 2007 employment profile discussed above. In the 2014 LSAL survey, each participant was asked about the cumulative impacts of the Great Recession, including several items specifically related to employment. The reported impacts cover the six-year period from the onset of the GR in 2008 until the completion of the survey in 2014. Some individuals may have had one period of unemployment or reduced hours, and other individuals may have had two or three periods of unemployment. The reported impacts include the full range of employment impacts over this period.

Some 14% of LSAL adults who were employed in 2007 reported that they lost a job as a result of the GR and 17% of those employed in 2007 indicated that they had their hours of employment reduced due to the GR (see Table 2). Sixteen per cent of LSAL adults employed in 2007 lost benefits (including insurance) as a result of the GR. Nine per cent of LSAL adults employed prior to the GR experienced two or more of these negative consequences and 5% experienced all three of the negative consequences asked about in the 2014 LSAL survey.

Table 2:

Impact of the Great Recession on employment, 2008–14 (percentages)

Employment	Impacts of the Great Recession 2008–14					N
	Lost job	Reduced hours	Lost benefits	Maintained employment	Contributed to retirement
Employed in 2007	14	17	16	78	68	2,105
Seeking employment	31	42	30	35	25	64
Out of workforce	8	9	12	34	40	284
All LSAL young adults	14	17	16	72	63	2,452

Although the impact of the GR on adults who were employed in 2007 was substantial, it was devastating to those young adults already struggling to find employment. Among LSAL adults who were looking for work prior to the onset of the GR, 31% reporting getting and losing a job during the following six years and 42% reported getting a job with reduced hours – that is, part-time employment in most cases. Thirty per cent of these LSAL participants indicated that they obtained some employment during the six years after the onset of the GR, but lost or did not get health insurance or other employee benefits. Although our baseline measures found that only about 3% of LSAL participants were actively looking for work in 2007, it is important to recognise that Generation X includes approximately 84 million young Americans and the 2% would represent about 1,680,000 individuals.

The GR also impacted adults who were voluntarily out of the workforce in 2007 (a small number of young adults who were out of the workforce in 2007 were disabled and unable to work). Subsequent surveys of LSAL participants found that many of these young adults re-entered the workforce during the six years after the onset of the GR and experienced some negative employment consequences. Eight per cent of LSAL adults who were out of the workforce in 2007 reported that they re-entered the workforce and subsequently lost a job due to the GR (see Table 2). Nine per cent re-entered the workforce and were employed on a part-time basis due to economic conditions, and 12% of these adults reported losing or not being able to obtain employee benefits like insurance during the six years after the Great Recession for economic reasons.

It is important to put these numbers in context. Applying the LSAL results to the Generation X population, 71 million Generation X adults were working in 2007 and 78% (66 million) continued to be employed throughout the GR. Two thirds of Generation X adults (57 million) were able to continue making regular contributions to their retirement funds¹ throughout this period (see Table 2). In contrast, only 35% of Generation X adults who were unemployed and seeking work at the beginning of the GR were able to find and maintain employment during the six years after the onset of the GR.

In short, a substantial majority of Generation X adults (as represented in the LSAL) were not negatively impacted by the GR and 72% of these young adults were able to sustain steady and adequate employment throughout the GR years. However, a significant minority of LSAL participants and Generation X adults – about one in four or more than 20 million individuals – experienced a loss of employment or a reduction in the hours of employment during the GR. Other analyses will examine the impact of these negative employment experiences on the health, education and ability to support the education of one’s children, but we extrapolate that this was a significant life event for an estimated 20 million Generation X adults.

The net effect of the Great Recession on Generation X

Although the negative consequences of the GR on Generation X (as reflected in the LSAL longitudinal sample) were significant and influenced the lives of many adults, an analysis of the responses from all LSAL participants indicates that the majority of LSAL adults did not suffer any negative consequences from this economic downturn. The LSAL collected a number of measures of negative consequences during the GR and a parallel set of measures of positive social and economic outcomes (see Appendix A for the wording of the 2014 summary questions). It is important to understand the pattern of these effects across the full spectrum of Generation X young adults.

To explore the structure of these impacts, a confirmatory factor analysis² was conducted, using LISREL. The results show that the experience of the GR by LSAL adults can be summarised along two dimensions – or factors – that reflect generally negative consequences or positive consequences (see Table 3). The factor loadings vary as a reflection of the magnitude of each kind of impact on LSAL participants. The strong negative correlation between the two factors indicates that LSAL participants who experienced some level of positive outcomes during the GR years were unlikely to have experienced negative effects at the same time, and that LSAL adults who experienced significantly negative consequences were unlikely to have experienced many positive effects in the same period.

Table 3:

A confirmatory factor analysis of the impact of the Great Recession, 2014

	Factor loadings
	Negative impact	Positive impact
Difficulty in making rent or mortgage payments	0.77	–
Deferred or postponed medical visits or procedures	0.67	–
Used savings to meet regular living expenses	0.53	–
Experienced a significant reduction in hours of work	0.48	–
Lost health insurance or other benefits from job	0.42	–
Lost job	0.40	–
Had a mortgage foreclosed	0.23	–
Continued to contribute to retirement plan beyond SS	–	0.70
Able to pay for health insurance or services as needed	–	0.68
Took personal or family vacations	–	0.61
Bought or made mortgage payments on a home	–	0.59
Maintained steady work at the level wanted	–	0.57
Saved money for college expenses for children	–	0.56
Bought a new automobile or similar vehicle	–	0.48
Correlation between the factors =	−0.68

Fit statistics: Chi-squares = 132.6; degrees of freedom = 54; Root Mean Square Error of Approximation (RMSEA) = 0.023; upper 10% confidence interval of RMSEA = 0.028; N=2,793.

It is possible to compute a simple zero-to-ten index for the level of negative consequences experiences and a similar index for the level of positive consequences experienced. By combining these two scales into a summary scale, we can describe the impact of the GR quantitatively as ranging from very negative (−10) to very positive (+10). The mean score on the combined index is 3.9, indicating a slightly positive impact of the GR across all Generation X adults in the US. The median score on the index was 5.0, which means that half of the respondents experienced a positive result of at least 5.0 on this −10 to +10 scale (see Figure 1).

Figure 1:

Distribution of scores on the index of the Great Recession impact, 2014.

Given the mixed impact of the GR on LSAL participants, it is important to ask which factors or life course variables predict a positive or negative experience during the GR. For this purpose, we will utilise a structural equation model (Jöreskog and Sörbom, 1993; Cudeck et al, 2001).

A structural equation model (SEM) is a set of equations that utilise the known chronological or logical order of variables that are used to predict an outcome and utilises that information in estimating the model. In principle, prior occurrence or influence flows from left to right. Variables that are in the same column are assumed to have occurred within the same chronological or logical frame. The path coefficients estimate the strength of each relationship, holding constant all variables to the left or in the same column – that is, not subsequent to the variables.

In practical terms, an SEM takes advantage of our prior knowledge that some things must precede others, thus reducing the number of equations or relationships that need to be estimated to understand the flow of influence over time, or the life course, in this analysis. We know, for example, that parent education generally precedes the education of LSAL participants, although a small number of parents may continue their education at the same time their child is in school. We know that a respondent’s gender is largely set at birth, although some adolescents or adults may change their gender later in the life course. The structure of the variables in Figure 2 reflects a broad understanding of the nature of the life course and the flow of life experiences and decisions (see Appendix B for a description of the construction of each variable in the model).

Figure 2:

A model to predict in the impact of the Great Recession.

We can estimate the total effect of each variable in the model on the outcome variable – our measure of the net impact of the GR on each individual – by computing the product of all of the path coefficients connected directly and indirectly to the outcome variable. The table included in Figure 2 indicates that the strongest predictor of the net impact of the Great Recession was the level of educational attainment (0.37). In this analysis, the educational attainment of LSAL adults was measured using a new measure that takes into account an indicator of the quality of the educational experience at the university level, and this measure demonstrates the impact of the attainment of a university degree and the quality of that degree.

The second strongest indicator was the level of each individual’s parents’ education (0.19), reflecting the class structure of educational advantage and inherited economic security (see Figure 1). Parents with a higher level of educational attainment provide a home environment that fosters complex cognition, encourage their children to seek and value higher levels of education, place them in better pre-college school systems, and encourage them to seek admission to more competitive and rigorous university programmes. It is a system of cumulative advantage (DiPrete and Eirich, 2006).

Women were slightly more likely to experience negative GR consequences than men (−0.04). African-Americans were substantially more likely to experience negative GR consequences than others (−0.16), but Hispanic-Americans were less likely to experience negative consequences than African-Americans (−0.05).

Individuals who were married or living with a partner in 2007 or 2008 were more likely to experience a positive effect from the GR than were unmarried LSAL adults (0.18). This result suggests that households with two incomes were less likely to experience strong negative consequences than single-person households – a sharing of the risk. A respondent’s time in job was positively related to a better net effect (0.21). This result indicates that individuals who worked in the same job or firm for some years prior to the GR were less likely to experience negative consequences, holding constant differences in educational attainment, race, ethnicity and other prior factors. One of the characteristics of an SEM is that each path between two variables is the residual or net effect, holding constant all of the other variables in the same column or to the left of the variable that is the source of the path.

This model has good fit parameters and accounts for 24% of the total covariance in the model. Given the diffuse pattern of impact of the Great Recession (note distribution in Figure 1), a model that accounts for a quarter of the total covariance is a satisfactory analytic outcome. It would be possible to increase the R² of the model by collapsing the scale scores into fewer ordinal categories, but we believe that it is important and instructive to retain the actual variance and covariance found in the data. Life course outcomes are rarely neat and it is important for social analysts to deal with the actual dispersion of impacts to be able to better understand the relative influence of selected variables on these outcomes.

Discussion

The impact of the GR on Generation X – as represented by the LSAL – was mixed. Some LSAL adults experienced substantial negative effects, ranging from the loss of a job to the loss of their home. Given the range of economic and social consequences that flowed from the GR, it is important to examine the factors that were associated with the net impact of the GR on these LSAL adults.

The results of both descriptive and model-based analyses of the relationship of major demographic and educational factors to job loss and economic insecurity pointed to the importance of educational attainment and possibly to length of time in the job held at the onset of the GR. A subsequent analysis of a structural equation model indicated that the strongest predictor of avoiding negative consequence from the GR was a higher level of educational attainment. This is consistent with a good deal of the literature concerning the GR (Abel and Deitz, 2017; Kalleberg and von Wachter, 2017) and with our general knowledge of the influence of education on life outcomes (Sexton, 1961; Hyman et al, 1975; Alexander et al, 2014).

The second strongest predictor of job loss in the GR was the level of education of one’s parents, which should be seen as the persistence of social class influence and a product of a system of cumulative advantage/cumulative disadvantage that permeates the American educational system and social structure. The social class predictor was only half as strong as an individual’s own educational achievement, suggesting some room for generational mobility, but it should remind us of the persistent influence of educational and social class in the US.

This model indicates that gender made very little difference on the impact of the GR on young adults in the LSAL, with a total effect of −0.04. This result does not mean that all gender discrimination has ended, but young women have earned more university degrees than young men in the US for more than 20 years and these results show the cumulative impact of the levelling of educational opportunity in American society.

Our model also indicates that African-American LSAL participants were more likely to experience negative consequences from the GR than other LSAL adults (−0.16), holding constant parent education and gender. African-American participants in this sample were likely to have a slightly lower level of educational attainment (−0.12) than other LSAL adults, holding constant parent education and gender. This is a complex issue reflecting generations of overt discrimination and social and economic disadvantage which cannot be fully reflected in even a 30-year longitudinal study.

Hispanic-American LSAL participants were slightly more likely to experience a negative effect from the GR than other LSAL adults (−0.05), holding constant parent education and gender. This is about half of the negative impact experienced by African-American LSAL participants. It is interesting to note that the Hispanic-American disadvantage in educational attainment was about half of the disadvantage associated with African-American participants.

The results from both African-American and Hispanic-American LSAL participants should remind us that the attainment of equal opportunity is not complete in the US. While we have made substantial progress in the post-war years, there is a good deal of work yet to be done.

Finally, we note the potential longer-term effects of the GR experience on political and social attitudes. As we observed in our introduction to this analysis, the GR was the first major economic downturn experienced by Generation X. Our analysis found that there was substantial variation among LSAL participants in the positive or negative impact of the GR and we were able to provide refined estimates of the relative impact for women and for racial and ethnic minorities. LSAL young adults who experienced little or no negative effects from the GR may have no lingering scars from the experience and may even reduce their concern about public policies that produce short-term economic downturns. Conversely, LSAL young adults who lost a job or a home or experienced other negative consequences from the GR may hold altered views of economic stability or economic policy that will influence their political and social views in the decades ahead. We have no current data relevant to these longer-term concerns, but one of the values of longitudinal studies over several decades is that we will be able to revisit these questions in the years ahead.

Appendices

Appendix A

Items included in the measurement of the net impact of the Great Recession

Appendix B

Variables included in the structural equation model in Figure 2

Gender	Dichotomous variable entered in the model as a dummy variable. Female is coded as 1 and male as 0, so path coefficients and total effects describe women’s characteristics.
Parent education during high school	Five-category ordinal variable: (1) includes all individuals who did not complete secondary school or obtain a GED, (2) high school graduates and GED holders, (3) respondents with an associate degree or some college, (4) individuals who earned a baccalaureate but not a graduate or professional degree, (5) individuals with a graduate or professional degree.
Respondent educational attainment in 2007/08	Six-category ordinal variable: (1) includes all individuals who did not complete secondary school or obtain a GED, (2) high school graduates and GED holders, (3) respondents with an associate degree, (4) individuals who earned a baccalaureate or four-year university degree, (5) individuals with a masters’ degree, (6) individuals with a doctorate or professional degree (law, medicine, architecture and similar).
African-American	Dichotomous variable entered into the model as a dummy variable. African-American is coded as 1 and all others are coded as 0.
Hispanic-American	Dichotomous variable entered into the model as a dummy variable. Hispanic-American is coded as 1 and all others are coded as 0.
Married in 2007/08	Dichotomous variable enter into the model as a dummy variable. Married or with partner is coded as 1 and others are coded as 0.
Minor children at home 2007/08	Dichotomous variable entered in the model as a dummy variable indicating whether a respondent has one or more children under the age of 18 living in their home.
Rural 2007/08	Dichotomous variable entered in the model as a dummy variable, where 1 means that the individual lives in a community of 50,000 or fewer population and 0 means that the individual lives in a community larger than 50,000.
Time in job 2007/08	Dichotomous variable entered in the model as a continuous dummy variable, where 1 means that the individual had held his or her 2008 job for more five years of more.
GR impact 2014	This summary measure of the impact is a continuous variable that ranges from −10 to +10 and is treated as a continuous variable in the model. The construction of this index is described in the text of the article and Table 3 displays the factor loadings.