Family (Dis)Advantage and Life Course Expectations

Abstract

Optimistic assessments of life chances can positively influence life outcomes, but conflicting theories suggest these assessments either reflect structural privilege or develop as a result of childhood hardship. In addition, competing hypotheses suggest that these assessments may matter differently depending on who holds them. We examine whether family socioeconomic status shapes adolescents’ expectations about how successful their lives will turn out. We distinguish generalized life expectations (GLE), capturing anticipated success in life across multiple domains, from intergenerational comparative expectations (ICE), which register expectations about improvement relative to observed success within the respondent’s family lineage. We find that adolescents from higher socioeconomic status families are simultaneously more optimistic about their likely success in life (GLE) but less likely to anticipate relative improvement in life success across generations (ICE). Holding high GLE in combination with low ICE predicted doing better in adulthood across a range of health, attainment, and well-being outcomes, though in most cases high GLE, regardless of ICE, was the key. These beneficial patterns are, for the most part, at least as beneficial for socioeconomically disadvantaged youth as they are for advantaged youth.

People’s beliefs about their futures matter in important ways. Adolescents’ answer to the question, “How far do you really think you’ll go in school” is a now taken-for-granted indicator for understanding status attainment, with the classic model emphasizing a key role for subjective beliefs. Sociology is also now seeing a fledgling movement across subfields emphasizing future-oriented action, including how beliefs about the future help direct and construct the life course across multiple domains (Frye 2012; Mische 2009; Tavory and Eliasoph 2013). Although most studies of future expectations focus on specific domains (e.g., education), recent research suggests that general expectations about one’s life affect later well-being and attainment Halleröd 2011; Hitlin and Johnson 2015; Schafer et al 2011). Expectations about the future form an important basis for self-assessments of life course agency (Hitlin and Johnson 2015), suggesting that aspects of a person’s worldview that transcend specific domains have importance for predicting a range of life course outcomes.

Important questions remain, however, about the development and operation of such agentic beliefs. We consider two such questions. First, who is likely to develop these favorable life expectations? Decades of stratification research demonstrate that adolescents look toward higher educational attainment and higher status occupations when they come from more advantaged families and when significant others in their lives, such as parents and teachers, hold high expectations for them (Alexander, Eckland and Griffin 1975; Bozick et al. 2010; Sewell and Shaw 1968). Yet, when it comes to optimism about life more generally, the few empirical studies to date suggest complex and somewhat contradictory effects about (dis)advantage potentially motivating such beliefs Halleröd 2011; Jessor, Donovan and Costa 1996; Schafer et al. 2011).

Our second question examines whether the impact of optimistic life course expectations on later outcomes varies by whether youth come from advantaged or disadvantaged origins. A resource substitution hypothesis would argue that high expectations matter more when other resources are lacking, whereas a cumulative disadvantage hypothesis would argue the opposite—that more privileged youth benefit the most from high expectations. Previous research supports both possibilities; as we note below, the type of expectation examined may explain which process occurs. Limited empirical evaluation of these ideas for life course expectations to date indicates greater support for the cumulative disadvantage hypothesis, though the issue is far from resolved Halleröd 2011; Schafer et al 2011).

In this paper, we assess links between adolescents’ expectations about their futures and their level of family disadvantage or advantage, indexed by multiple measures of socioeconomic status. We then assess the influence of these future expectations on a range of young adult outcomes, considering whether (dis)advantage moderates the impact of favorable life expectations. We employ the “Life Chances Scale” (Jessor, Donovan and Costa 1996) recently utilized by Hitlin and Johnson (2015) to demonstrate the influence of adolescents’ optimistic life course expectations on the subsequent life course. In considering specifically the role of (dis)advantage, however, we find it useful to differentiate expectations for one’s life by whether they are a) general expectations about one’s life in the future in absolute terms or b) comparative beliefs about one’s life in the future, anchored relative to observed circumstances. Drawing on data from 12^th grade respondents in the Youth Development Study, we examine the extent that expectations are shaped by family socioeconomic status, and whether socioeconomic status moderates the influence of future expectations on a range of life outcomes many years later, when respondents are in their mid-30s.

The Formation of Life Expectations

Most research on future expectations has focused on specific domains, such as educational, occupational, or fertility expectations, with less attention paid to the development of general life expectations that transcend or incorporate multiple life-course domains. We know less about how people develop broader ideas about their future lives, capturing a global sense of whether one’s life will ‘work out.’ Halleröd (2011) argues that generalized beliefs develop during childhood, indicating that people sharing a similar social location have similar life chances, and that these beliefs aid in the evaluation of one’s own future prospects (see also Young 2004). We develop understandings of the life opportunities and potentials for ‘people like us’ that shape our expectations and in turn our choices and life trajectories. Consistent with this idea, at least one study found that adolescents from higher socioeconomic status backgrounds anticipated better chances of success in life (Jessor, Donovan and Costa 1996). Likewise, young people who report very low expectations for their lives (not even expecting to live beyond age 35) are disproportionately from lower socioeconomic status backgrounds and tend to live in poorer neighborhoods (Swisher and Warner 2013). In this way, we might expect those growing up in higher socioeconomic status families to have higher expectations for their lives.

Somewhat paradoxically, other studies suggest that adversity in early life is associated with optimistic expectations. Schafer, Ferraro and Mustillo (2011) find that early life adversity, stemming from economic situations, family instability, abuse and poor health, can promote the sense that life is getting, and will continue to get, better. Reflecting on the past reveals the marks of adversity, however, with those experiencing more difficulty evaluating their pasts more negatively than others. Halleröd (2011) finds that young people who believe their lives will turn out better than their peers’ lives are a heterogeneous mix of those with privileged and surprisingly, working-class, backgrounds. It would seem that adversity, including lower socioeconomic status backgrounds, may also foster hope for the future.

We suggest that this apparent contradiction can be resolved by thinking more carefully about the referent people use when developing and expressing life expectations. We argue that life expectations can be distinguished by whether they are generalized – with an absolute standard – or whether they are anchored comparatively to one’s current circumstances or another referent – a relative standard. Generalized expectations tap into a sense of optimism about life, a general positive evaluative attitude about the future (Peterson 2000).¹ They capture the degree to which individuals think their lives will turn out well. Comparative life expectations, in contrast, are anchored differently–the future is conceived in terms relative to observed conditions, usually one’s own life (now or in the past), or compared to that of relevant others. This situates one’s hopes for the future within a more comparative framework measuring one’s own chances to specific, relevant reference groups or the self (depending on the measure). As such, comparative anticipated outcomes are not easily categorized as bad or good in objective terms; one can expect one’s life to be relatively worse in the future (compared to life now, or compared to the lives of a reference group) but yet still be a very good life, especially if one comes from a privileged background.

The global sense that things will work out well (Jessor et al. 1996)² is conceptually distinct from an evaluative comparison with one’s current life or with significant others’ lives in setting a benchmark for potential futures Halleröd 2011; Schafer et al. 2011). This distinction provides a plausible explanation for seemingly contradictory findings in past research. The Life Chances Scale that Hitlin and Johnson (2015) recently employed to consider adolescents’ future expectations primarily contains items tapping expectations of general success in life, but also a pair tapping relative success – expecting improvement in life success across generations (i.e. expectations anchored relative to family lineage). We label these generalized life expectations (GLE) and intergenerational comparative expectations (ICE), and suggest disentangling these items is important for properly exploring life course agency.

Consistent with domain specific research on attainment beliefs as well as research on survival expectations, we argue that socioeconomically disadvantaged youth are not likely to be very optimistic when it comes to expecting general success in life (GLE). Expectations may be generally optimistic (even inflated) on the whole, but advantaged adolescents’ GLE will be higher than disadvantaged adolescents’ GLE.

But might socioeconomically disadvantaged youth be hopeful in particular ways, as suggested by the adversity perspective? Studies in this vein, indicating a link between adversity and optimistic anticipated futures, examined comparative, not general, beliefs. Thus, people from disadvantaged circumstances demonstrate hope when referenced against their own past (Schafer et al., 2011), and may do so when compared to previous generations. Although it seems likely disadvantaged youth could be more optimistic in relative terms than they themselves are in absolute terms, it is much less clear that they could be more optimistic in relative terms than advantaged youth. In other words, will they demonstrate equal ICE or might they have higher ICE than adolescents coming from more advantaged family backgrounds? One possibility is that intergenerational comparative expectations are not particularly tied to socioeconomic (dis)advantage. The putative openness of the American stratification system may foster widespread belief in brighter futures across generations. Both advantaged and disadvantaged youth may expect to do as well or better than their parents and anticipate similar outcomes for their own offspring.

At the same time, there may be reason to think disadvantaged youth might anticipate comparatively successful lives across generations more so than advantaged youth. If individuals perceive they are relatively low in the stratification system, they might think they will at least be able to do similarly or better than their family of origin. There may be a “floor effect” operating, with youth thinking that things “certainly can’t get worse,” or “can only go up from here.” Socioeconomically disadvantaged youth may also perceive they can make use of education to achieve a better life, as schooling is widely believed to be the primary route to upward mobility. Socioeconomically advantaged youth, for their part, might feel they will not exceed the successes of their high-achieving parents, having a high benchmark for their own comparative expectations.

For much of American history, this was not likely an issue for advantaged youth. The American Dream has long been a foundation of optimism about intergenerational mobility in the U.S, with each generation expecting to do better than the previous one. Yet in the period since the 1970s, conditions of rising inequality, increasing global competitiveness, and growing awareness that it may not be possible for standards of living to rise indefinitely may have tempered everyone’s expectations, but particularly those of the advantaged. Time and place matter (Elder, Johnson, and Crosnoe 2003). Children of socioeconomically successful parents, particularly those with already highly educated parents, may not foresee that they will have greater success than their parents. We might expect such adolescents to be generally optimistic – their lives will turn out well – without expressing the traditional American belief that their (and their children’s) futures will surpass the success of their parents’ lives. In contrast, disadvantaged youth may be more likely to see room for improvement, and yet they may not expect “the good life” to the extent of advantaged youth.

In sum, we expect SES disadvantaged adolescents to have lower generalized life expectations (GLE) than SES advantaged adolescents; we examine competing expectations about whether these lower GLE are combined with intergenerational comparative expectations (ICE) that are similar to or exceed that of SES advantaged adolescents.

The Influence of Expectations on Later Outcomes

Classic status attainment theory (Alexander, Eckland and Griffin 1975; Sewell and Shaw 1968), as well as more recent explorations of educational expectations (Morgan 2004; Reynolds and Johnson 2011), posit that educational expectations affect status attainment. Likewise, fertility expectations are theorized to influence childbearing patterns (Schoen et al. 1999). But what about general expectations about one’s future as a whole? Sociological tendencies to focus on particular domains create partitioned versions that do not reflect the way people view their lives as a unified enterprise (Smith 2009).

General expectations do matter; the sense that life has gotten better and will get better in the future measured at one point in time predict higher life evaluations measured years later (Schafer et al. 2011). Halleröd (2011) found that controlling for background factors, children’s expectations of how their lives would turn out relative to other children their age predicted adult income and evaluations of pay, though not mortality or receipt of social assistance; on average, the most pessimistic children had poorer later life experiences. Hitlin and Johnson (2015) found that optimistic adolescent expectations about one’s future life predicted a range of attainment and well-being outcomes across the transition to adulthood, even when controlling for conventional structural variables as well as widely employed social psychological resources (educational expectations and competency beliefs). Even while expectations reflect one’s place within the stratification system, they potentially act as an independent force shaping life course outcomes.

Beliefs about the future matter because they inform everyday decisions about how one should work toward goals, direct long-term planning, and whether to engage in risky behavior (Clausen 1993; Cowan 2011; Schneider & Stevenson 1999; Soller and Haynie 2013). Positive expectations for the future foster intentional self-regulation abilities over time, enabling individuals to better select behaviors that facilitate goal attainment (Schmid, Phelps and Lerner 2011). Short-term steps taken with longer-term goals in mind can place people on positive life course trajectories. Beliefs about the future can also be motivational, increasing perseverance in the face of hardship (Frye 2012), and buffering individuals from setbacks (Oyserman et al. 2004). Indeed, for those facing structural disadvantage, social psychological factors can also contribute to what psychologists term ‘resilience’, experiencing good or better-than-expected outcomes despite the presence of significant risk (e.g., Goldstein and Brooks 2012). Masten (2001) argues, for example, that various protective systems and resources, among them positive views of self and motivation, facilitate resilience, helping to explain why most children with rough starts manage to do better-than-expected in adulthood.

Socioeconomic (dis)advantage may not only facilitate the development of different life expectations, but may condition their impact on future outcomes as well. Expectations for the future may be more strongly linked to outcomes among those with the resources to pursue their goals, consistent with a cumulative advantage process. Educational goals, for example, are more predictive of school performance among children from higher socioeconomic status families because their goals better reflect performance feedback (Alexander, Entwisle, and Bedinger 1994). Degree expectations are more predictive of attainment among more advantaged youth as well, for whom expectations tend to be more stable (Bozick et al. 2010; Johnson and Reynolds 2013). Mastery, another social psychological resource reflective of socioeconomic advantage, facilitates meeting occupational expectations (Reynolds et al. 2007), though more for men than women due to men’s fewer structural barriers in achieving occupational goals. With respect to more general expectations, Halleröd (2011) found that children who thought they would do “much better” than their peers, but were from disadvantaged backgrounds and scored lower on standardized tests, were more likely to experience a year without income, much like their similarly situated peers who held very low comparative expectations for their futures. Expectations of doing “much better” among more privileged children were still advantageous (though not more so than anticipating doing “a little better”). Likewise, adults who suffer more misfortune as children benefit less from anticipating better futures for themselves (i.e. rating one’s life overall in the future more positively) than those with more privileged childhoods (Schafer et al. 2011). These studies suggest a group of “unrealistic optimists” Halleröd 2011), or at least suggest “limits to wishful thinking” (Schafer et al. 2011:1079) wherein positive expectations do not hold much power.

In contrast, a resource substitution process suggests the opposite—that social psychological resources matter more for those with fewer alternative resources. Shanahan et al. (2014) found, for example, that valued personality characteristics were stronger predictors of attainment among children of less educated parents than they were among children of more educated parents. Resource substitution hypotheses have also been supported with respect to mapping the education benefits for health. Education provides greater return to women than men (Ross and Mirowsky 2006) and to those from lower educated families than higher educated families (Ross and Mirowsky 2011).

We test both the cumulative disadvantage and resource substitution hypotheses. From the discussion above, it would seem that expectations matter less for young people when they are not grounded in some way (e.g. realism, access to resources, etc.). Such grounding may be evident in the relative optimism of life course expectations by socioeconomic status and in how expectations co-develop. For adolescents from socioeconomically disadvantaged families, holding high GLE and high ICE is internally consistent—achieving the “good life” would likely come with improvement in life outcomes across generations. Levels of realism may vary across adolescents, but these expectations could represent the sort of optimism that directs behavior and sustains effort to promote resilience (see Andersson 2012). Expecting some intergenerational improvement in life could also be consistent with more moderate expectations for success in life (i.e. high ICE but lower GLE; things will get better but still not with high chances of broad success). Both configurations could help set youth on paths leading to better health, attainment, and well-being. However high expectations for life generally with no expectation of comparative improvement suggests, for disadvantaged youth, life expectations are not well thought out and may do little to facilitate maintaining paths toward good outcomes. In contrast, for more advantaged youth, some sense that (upward) intergenerational mobility has a limit but that life success is still likely may reflect optimism combined with realism. We examine these GLE/ICE configurations to capture their interrelated development and associated meanings.

Because we argue that life expectations have wide-ranging implications, we examine a range of outcomes, including indicators of economic well-being, job quality, self-rated health, depressive affect, self-esteem, social support and self-evaluations of success in work, in family and in balancing work and family. In taking such a broad approach, we have not addressed specific literatures on each of these outcomes. Rather, we assess whether life expectations predict a wide range of outcomes (and for whom), controlling for adolescent achievement and other social psychological resources widely acknowledged to shape life paths. Specifically, in addition to educational expectations, which we have discussed as having important impacts, we also control academic achievement and mastery beliefs. Mastery beliefs (or related phenomena of self-efficacy or personal control) have broad effects on life success and well-being (Lewis, Ross and Mirowsky 1999; Pearlin and Pioli 2003). We believe that considering a range of outcomes provides an important window into the general processes outlined above and minimizes the chance that significant effects are picked up only in certain life domains or only when employing certain strategies of measurement. Constructing a contemporary life course involves handling multiple domains simultaneously.

Data and Measures

The data come from the Youth Development Study (YDS), a panel study beginning with randomly selected ninth graders from the St. Paul, Minnesota public high schools in 1988 (N=1,010). We selected this data set because it includes a perceived life chances scale, which contains both generalized and intergenerational comparative life expectation items. In addition, respondents in this study have been followed into adulthood, allowing an assessment of whether life expectations matter for a wide range of later outcomes differentially by family (dis)advantage status. Although a community-based sample, U.S. Census data indicate that St. Paul was comparable to the nation as a whole with respect to several economic and sociodemographic indicators (Mortimer 2003, 2012). We know of no nationally representative data that has included a suitable measure of life expectations with which to test the ideas presented here.

We draw on data from the fourth wave, collected via questionnaires administered in the adolescents’ classrooms in 1991 (when most had progressed to 12^th grade) to assess life expectations. Those who were not present on either of the two administration days, were no longer attending school, or had moved from the area completed questionnaires by mail. Parents were also surveyed by mail and our measures of family socioeconomic status are derived from parents’ reports. The senior year of high school is an especially opportune time to assess orientations toward the future, as adolescents are exerting agency (Elder et al., 2003) by making key decisions about further education, work, and family formation. We also draw on data from the 18^th wave, collected via mailed questionnaires, to assess whether life expectations shape outcomes related to economic well-being, job quality, health, social support and self-evaluations of life success. The 18^th wave, collected in 2009, retained 67% of the original sample (69% of the wave 4 sample). Retention was higher among women, whites, and those with at least one employed parent in the household, but was not found to be associated with many other factors assessed in adolescence, including achievement, mental health, delinquency, and parental income, education and marital structure (Mortimer 2003). Neither ICE or GLE predicted participation in wave 18.

Life Expectations

Our life expectations measures are derived from the Perceived Life Chances Scale (Jessor, Donovan and Costa 1988). Respondents were asked “Thinking about how you see your future, what are the chances that…” and the ten response items include, for example, “You will have a job that pays well,” “You will have a happy family life,” “You will be respected in your community”, and “Life will turn out better for you than it has for your parents,” rated on a 5 point scale (1 = very low, 2 = low, 3 = about fifty-fifty, 4 = high, and 5 = very high). The wording for the full set of items appears in Table 1. Factor analysis of the 10 items (principal components analysis, promax rotation) supported two factors, one with strong loadings for eight of the items, capturing expectations about widely valued life outcomes in a variety of domains, or generalized success (GLE, α=.84). The other, with strong loadings for two items, captures intergenerational comparative expectations (one item reads, “Your children will have a better life than you’ve had” and the other reads, “Life will turn out better for you than it has for your parents; ICE, two items correlated at .53).³ Factor loadings are presented in Table 1. The Perceived Life Chances Scale was not measured again after wave 4.

Table 1.

Factor Loadings for Generalized Life Expectations and Intergenerational Comparative Expectations, Youth Development Study

	Factor 1 (GLE)	Factor 2 (ICE)
You will have a job that pays well	.75	.16
You will be able to own your own home	.76	.13
You will have a job that you enjoy doing	.76	.17
You will have a happy family life	.67	.21
You will be in good health most of the time	.61	.13
You will be able to live wherever you want to in the country	.65	.11
You will be respected in your community	.67	.20
You will have good friends you can count on	.48	.22
Life will turn out better for you than it has for your parents	.30	.80
Your children will have a better life than you’ve had	.12	.90

N=913. Questions read: “Thinking about how you see your future, what are the chances that…” (1 = very low; 2 = low; 3 = about fifty-fifty; 4 = high; 5 = very high)

Our arguments about (dis)advantage and life course expectations not only contrast high and low expectations, but hold that groups of youth will be more or less likely to develop high expectations on one dimension and low expectations on the other and that these work together in shaping life course outcomes over time. As such we classified respondents into four groups based on their placement on both scales (low-low, low-high, high-low and high-high) using an average of 4 (labeled “high” on the questionnaire item) to distinguish more positive expectations from lower ones.⁴

Family (Dis)advantage

We examine three indicators of parents’ socioeconomic status: family income, parents’ educational attainment, and the extent that parents’ jobs were intrinsically rewarding and self-directed. Family income was measured with an ordinal scale, from 1= under $5,000 to 13= $100,000 or more; we set respondents’ values to the midpoint of each category. Parental educational attainment references the most highly educated parent, measured ordinally by highest degree obtained (1 = less than high school degree to 7 = PhD or professional degree). Preliminary analyses explored threshold and nonlinear effects of family income and parental education on the two sets of life expectations, but found little evidence for them. Parental occupational characteristics, indicating intrinsically rewarding and self-directed work (Kohn and Schooler 1983; Parcel and Menaghan 1994), were measured by six items on which parents rated their jobs (see Appendix A). Items were standardized before taking the mean, as the first four items were rated on five-point scales and the last two items on four-point scales (α=.81). If both parents participated in the study and reported employment, we used the higher score of the two for the final value on this scale. If no parents were working, we coded this scale as zero; we included a dichotomous indicator for no parent employed.

Life Outcomes

Outcomes assessed at age 35–36 included educational attainment (in years), home ownership, unemployment (in past 12 months), and for those employed, logged hourly earnings, ratings of job security (1= not at all secure; 4=very secure), and whether current jobs were intrinsically rewarding (see Appendix A for scale items). Outcomes also included self-rated health (5-point scale from poor to excellent), ratings of how successful the respondents feel they are in their work lives, their personal or family lives, and in balancing the two (for each, 1=not successful at all; 4=very successful), experience with financial problems, level of social support, self-esteem (Rosenberg 1995), and depressive affect (Ware et al. 1979; scales for all of which appear in Appendix A).

Control variables

In parts of the analysis we control gender, family structure (two biological or adoptive parents, single parent, and those in other living arrangements, primarily step-families), and race/ethnicity. With limited racial/ethnic variation in the panel (minorities were not oversampled), we distinguish whites from non-whites.⁵ We also control wave 4 measures of educational expectations (1 = less than high school to 6 = Ph.D. or professional degree), average grades (0 = F to 11 = A), and mastery beliefs (Pearlin and Schooler 1978; see Appendix A).

Missing data

In all multivariate analyses, we employ multiple imputation techniques (with 20 imputed datasets) to address missing data on the independent variables. In predicting life expectations, the percent of cases imputed primarily ranged from zero (for gender and parents’ work quality among those employed) to 4 percent (for race and parents’ education), however 23.12 percent of cases were imputed for family income because participation among parents was lower in this wave than at the study’s initiation. In predicting life outcomes in 2009, the sample size varies by outcome, but in each case less than 5 percent of cases are missing on all independent variables, again excepting family income.

Findings

Descriptive statistics appear in Table 2. The sample is just under half male and predominantly (73%) white. Means on the two life expectations scales, at 3.94 for generalized life expectations (GLE) and 3.78 for intergenerational comparative expectations (ICE; each on five-point scales), indicate fairly positive expectations for the future among this sample. The four groups we defined to classify adolescents’ expectations reflect this generally positive outlook. Recall we employed a cut-off such that respondents were answering “high” or “very high” chances on the items of the scale on average to be considered high overall. Figure 1 shows the distribution of respondents across groups. We refer to them descriptively (high and low on the dimensions) as well as by quadrant location in Figure 1, to aid the reader in keeping track of key contrasts in the findings. It is important to note that having high expectations on both dimensions or low expectations on both dimensions was common (65% of the sample), consistent with the positive correlation between the two scales (r=.44).

Table 2.

Descriptive Statistics on Study Variables, Youth Development Study

	Percent	Range	Mean (sd)
Adolescent Life Expectations, 1991
Generalized Life Expectations (GLE)		1 – 5	3.94 (.61)
Intergenerational Comparative Expectations (ICE)		1 – 5	3.78 (.79)
Adult Outcomes, 2009
Social Support		1 – 4	3.35 (.82)
Self-rated Health		1 – 5	3.60 (.90)
Self-Esteem		1 – 4	3.16 (.54)
Depressive Affect		1 – 5	2.47 (.63)
Perceived Financial Problems		1 – 7	4.02 (1.76)
Success at Family/Personal Life		1 – 4	3.12 (.78)
Success at Work Life		1 – 4	2.90 (.84)
Success at Balancing Family and Work Life		1 – 4	2.96 (.74)
Educational Attainment (years)		10 – 20	14.50 (2.03)
Owns Home	62.03
Unemployment in past year	16.59
Logged Hourly Earnings (among employed)		1–5.66	3.01 (.70)
Job Security (among employed)		1 – 4	2.80 (.87)
Intrinsic Work Rewards (among employed)		1–5	3.65 (.70)
Family Socioeconomic (Dis)advantage, 1991
Parents’ Education		1 – 7	3.28 (1.62)
Household Income ($1,000s)		2.5 – 115	39.82 (26.50)
Parents’ Intrinsically Rewarding and Self-Directed Work (among employed)	−2.45 – 1.18		.14 (.66)
No parents employed	11.79
Controls, 1991
Lives with two biological/adoptive parents	47.75
Lives with single parent	28.66
Alternative family structures	23.42
Male	47.72
White	73.38
Educational expectations		1 – 6	3.77 (1.17)
Mastery		1 – 4	3.00 (.46)
Grades		0 – 11	7.06 (2.36)

N=1,010 except 930–932 for adolescent life expectations, 655–669 for 2009 outcomes, and 514–536 for 2009 outcomes limited to employed respondents.

Figure 1.

Distribution of Generalized Life Expectations and Intergenerational Comparative Expectations

The panel as a whole was doing well in 2009, with means on social support, self-rated health, self-esteem and perceived success in their personal lives toward the upper end of the scale and depressive affect closer to the lower end of the scale. Means on perceived success at work and in work-life balance were a little lower than for personal life, though still above the midpoint. Perceived financial problems were above the midpoint, and over 16 percent had experienced unemployment in the past year.

The multivariate model predicting the expectation groups appear in Table 3.⁶ The group with high generalized life expectations but low intergenerational comparative expectations (quadrant 2) is contrasted with the other three groups in Table 3 (this reference was rotated in supplemental models and additional significant contrasts are noted in the text).

Table 3.

Multinomial Logistic Regression Models of Cross-Classified Life Expectations, Youth Development Study

	Low ICE-Low GLE vs. Low ICE – High GLE (quadrant 1 vs. 2)	High ICE – Low GLE vs. Low ICE – High GLE (quadrant 3 vs. 2)	High ICE – High GLE vs. Low ICE – High GLE (quadrant 4 vs. 2)
	b (se)	b (se)	b (se)
Parent Education	−.15 (.07)*	−.25 (.09)**	−.18 (.07)*
Household income	−.01 (.005)*	−.02 (.01)***a	−.01 (.00)*
Parents’ Intrinsically Rewarding and Self-Directed Work	−.31 (.20)	−.45 (.23)+	−.33 (.20)+
No Parents Employed	.45 (.51)	.51 (.56)	.26 (.54)
Intercept	1.77 (.29)***	1.91 (.33)***	1.92 (.29)***

N=930 (missing data on independent variables imputed; see text).

⁺p<.10;

^*p<.05;

^**p<.01;

^***p<.001

^aCoefficient for family income is significantly different (p<.05) compared to quadrants 1 and 4 as well.

Parental education levels and family income predicted group membership in much the same way. Adolescents with high GLE and low ICE (quadrant 2 from Figure 1) came from families with higher parent education levels and higher family income levels than all three other groups. Although for both parent education and family income the magnitude of the differences was largest for the comparison to the high ICE-low GLE group (quadrant 3; see middle column), only family income differentiated group membership further. Adolescents with low GLE and high ICE (quadrant 3) came from even lower income families than did adolescents with expectations high on both dimensions or low on both dimensions. Parents’ job quality was associated with group membership in a similar way as parental income and education, but the pattern was not as strong and did not reach conventional levels of significance.

Findings with respect to our first research question are most consistent with the idea that advantage promotes high expectations of general life success, but without strong prediction of relative improvement in life success across generations (quadrant 2). Adolescents with less advantaged families develop other combinations of beliefs about their futures; those from lower income families are specifically more likely to anticipate relative improvement across generations without high chances of general life success (quadrant 3). It is important to note that parents’ education, family income, and parents’ job quality did not distinguish between those with low expectations on both dimensions (quadrant 1) and those with high expectations on both dimensions (quadrant 4) (contrast not shown, but see the similar coefficients in columns 1 and 3 as each are compared to the low ICE – high GLE group).

We turn next to our examination of whether these sets of life expectations predict status attainment and well-being markers in adulthood, measured 18 years later. The coefficients of primary interest for the life expectations variables from models of these outcomes appear in Table 4.⁷ All models controlled socioeconomic background, gender, race/ethnicity and family structure, as well as educational expectations, grades, and mastery in adolescence (full tables with coefficients for all the control variables available from the authors). The adolescent life expectations groups did not predict earnings, nor perceived job security or the level of intrinsic job rewards (among those currently employed) in these multivariate models and so are not shown.⁸

Table 4.

Select Coefficients, Models Regressing 2009 Outcomes on Adolescent Life Expectations, Youth Development Study

	Social Support	Self-Rated Health	Self-Esteem	Depressive Affect	Financial Problems
Low GLE Low ICE (quadrant 1)	−.26 (.10)**	−.26 (.11)*~	−.14 (.06)*~	.16 (.07)*	.42 (.21)*
High GLE Low ICE (REF; quadrant 2)
Low GLE High ICE (quadrant 3)	−.30 (.11)**	−.08 (.13)	−.11 (.07)	.12 (.09)	.59 (.24)*
High GLE High ICE (quadrant 4)	−.11 (.10)	−.01 (.11)	−.01 (.06)	.07 (.07)	.58 (.21)**
N	665	665	662	665	669
	Success at Work	Success at Personal/Family	Success at Work-Personal/Family Balance	Owns Homeˆ	Unemploymentˆ
Low GLE Low ICE (quadrant 1)	−.21 (.10)*~	−.31 (.10)**~	−.20 (.09)*	−.78 (.29)**	.85 (.42)*
High GLE Low ICE (REF; quadrant 2)
Low GLE High ICE (quadrant 3)	−.21 (.12)	−.29 (.11)**	−.16 (.11)	−.29 (.33)	.86 (.46)
High GLE High ICE (quadrant 4)	−.01 (.10)	−.14 (.09)	−.14 (.09)	−.43 (.29)	.91 (.41)*
N	657	662	657	668	668

^*p<.05;

^**p<.01;

^***p<.001.

^ˆDichotomous outcome modeled as logit.

^~Low GLE low ICE significant difference with High GLE High ICE. All models control socioeconomic status, family structure, race, gender, grades, mastery and educational expectations. Models for depressive affect and self-esteem also control a lagged measure of the outcome measured in wave 4 (contemporaneously with life expectations). Missing data on all independent variables was imputed (see text).

Holding high generalized life expectations but low intergenerational comparative expectations (quadrant 2) is the reference group in these models, and this group clearly does well in adulthood across a range of outcomes—as well or better that the other three groups on all ten outcomes in Table 4. For six of those outcomes (self-rated health, self-esteem, depressive affect, success at work, success at work-life balance, and home ownership), this group (quadrant 2) does better than those who held low expectations in both dimensions (quadrant 1), but not better than the other two groups. For two additional outcomes (social support and success in personal life), this group (quadrant 2) does better than those who held low expectations on both dimensions (quadrant 1) and those who had high comparative expectations and low generalized expectations (quadrant 3). This group (quadrant 2) also does better than all other groups with respect to perceived financial problems, and does better than the group with high expectations on both dimensions (quadrant 4) with respect to recent unemployment.

To summarize across this range of outcomes, it appears that high adolescent generalized life expectations are beneficial in adulthood when combined with low intergenerational comparative expectations (all contrasts between low GLE – low ICE [quadrant 1] and high GLE – low ICE [quadrant 2] are significant). When combined with high ICE they are generally benign (many of the contrasts between high ICE – low GLE [quadrant 3] and high ICE – high GLE [quadrant 4] trend in a positive direction but none are statistically significant), though such respondents perceived greater financial problems in adulthood and were more likely to have experienced unemployment than those who held high generalized expectations in combination with low intergenerational comparative expectations (i.e. quadrant 4 vs. quadrant 2). They were also better off, much like our reference group, compared to those with low expectations on both dimensions with respect to self-rated health, self-esteem, and evaluations of work success (i.e. quadrant 4 vs. quadrant 1). ICE appear to have very little independent predictive power for these outcomes in adulthood (low GLE –high ICE [quadrant 3] is never significantly different from low GLE – low ICE [quadrant 1]), again with the exception that perceived financial problems are worse and unemployment more common for those with high ICE and GLE compared to those with high GLE but lower ICE (quadrant 4 vs. 2). But in the context of low GLE, whether ICE are higher or lower show no significant differences across any outcome.

The last aspect of our analysis examines whether optimistic life expectations matter equally with respect to life course outcomes regardless of whether youth come from socioeconomically advantaged or disadvantaged families. In order to test this hypothesis, we estimated models for each outcome including interaction terms with each SES indicator and expectation group membership, considering one indicator at a time. Most estimates indicated no conditional effects of the future expectations by SES, indicating that their benefits were shared. Though adolescents from socioeconomically disadvantaged families were less likely to form the most beneficial combinations of future expectations, when they did do so they were equally predictive of better outcomes in adulthood. A limited set of estimated conditional effects were significant, however, with expectations mattering differentially by parent education level for two outcomes in adulthood. Although this is not overwhelming evidence, the patterns are revealing.

For ease of interpretation, we re-estimated the two models indicating statistically significant conditional effects by parent education separately by whether parents a) left school with a high school degree or less or b) whether they had earned any postsecondary education.⁹ These models are presented in Table 5. For both self-esteem and home ownership, it is clear that the positive association of holding high GLE but low ICE (quadrant 2) is limited to adolescents whose parents had low levels of education. Indeed, among those with less educated parents, this particular combination of expectations is statistically significantly different compared to all three other groups (because in this set of analyses sample size is reduced considerably within group, we consider statistical significance beginning at the p < .10 level).

Table 5.

Select Coefficients, Models Regressing 2009 Outcomes on Adolescent Life Expectations by Parent Education Level, Youth Development Study

	Self-Esteem		Owns Homeˆ
	Parent HS or less	Parent postsecondary	Parent HS or less	Parent postsecondary
Low GLE Low ICE (quadrant 1)	−.23 (.12)+	−.10 (.08)	−1.92 (.68)**	−.34 (.34)
High GLE Low ICE (REF; quadrant 2)
Low GLE High ICE (quadrant 3)	−.31 (.12)*	.01 (.10)	−1.37 (.71)+	.23 (.43)
High GLE High ICE (quadrant 4)	−.19 (.12)+	.08 (.07)	−1.76 (.70)*	.17 (.34)
N	266	396	269	400

Note: Because parent education is among the independent variables for which missing data was imputed, categorization into low and high parent education varied slightly across imputed datasets. N listed is for imputation 1. All models control other measures of socioeconomic status, family structure, race, gender, grades, mastery and educational expectations. Models for self-esteem also control self-esteem measured in wave 4 (contemporaneously with life expectations). Missing data on all independent variables was imputed (see text).

Considering the full range of outcomes we assessed, this is poor support for the resource substitution hypothesis. Our findings support the notion that when such positive future expectations are developed, they matter just as much (and potentially more) for disadvantaged youth than advantaged youth.

Discussion

Our findings offer strong support for the idea that young people’s future expectations about the life course are marked by structural location and independently shape subsequent life outcomes. The distinction between two types of future oriented beliefs, however, provides a more nuanced picture of the ways adolescents think about their lives differently based on their disadvantaged and advantaged family positions, as well as which expectations represent agentic forces that can contribute to bettering their lives across a range of life course outcomes.

Our first objective addressed the problem that we know little about how these life course orientations are structurally shaped. Our findings about generalized life expectations (GLE) and intergenerational comparative expectations (ICE) were consistent with the idea that SES advantaged youth tended to hold high expectations for general success in life, but without a strong sense that they would have more successful lives than their parents and that their kids would stay on that upwards trajectory (quadrant 2). The more privileged one’s SES background, the more they saw high chances of building a “good life,” but did not see high chances of relative improvement of life success across generations. These youth present confidence that their lives will reflect the advantaged positions of their family of origin.

In contrast, lower SES youth develop a range of future expectations, with some expecting high success (GLE) and intergenerational improvement (ICE), but with many being not particularly optimistic about their chances of success (low GLE; occurring both with and without high ICE). Adolescents from lower income families were especially likely to hold low generalized life expectations but high intergenerational comparative expectations (quadrant 3). Recall that these differences exist among relatively positive expectations, with the average on both scales indicating adolescents’ belief of better than 50-50 odds of achieving these desired ends (and 50-50 classified as “low” in our analyses). The supposedly meritocratic ideals of ‘hard work’ leading to success are widely suffused throughout this population.

Future work should develop more complex models for understanding the development of life course expectations beyond socioeconomic status, including mediational processes linking socioeconomic status to life course expectations. In order to get a basic handle on how SES was related to different types of life course expectations, we set aside consideration of how other factors (such as school achievement) mediate SES effects on expectations and the possibly reinforcing relationship between such factors and expectations across the adolescent years. In assessing whether life expectations influence outcomes we controlled for many of these potential factors as a strategy to evaluate whether expectations matter above and beyond them, but this simplifies key processes deserving of future research attention.

Overall our findings on socioeconomic status help untangle the seemingly contradictory findings of previous studies suggesting life course optimism can stem from both adversity as well as privilege Halleröd 2011; Jessor, Donovan and Costa 1996; Schafer et al. 2011). Consistent with our argument, we found links between levels of adversity/privilege that differed across the types of future expectations under examination. Distinguishing between generalized expectations and comparative expectations allows us to begin to isolate different referents that respondents may be using to anchor their beliefs about the future. Asking about generalized successes captures an understanding of one’s future possibilities of achieving widely valued outcomes with shared measuring sticks.

This belief that one’s life will be generally successful, however, is conceptually and empirically distinct from a belief about the chances of relative improvement in life success across generations. For children of relative privilege, modern economic life does not foster the same beliefs of previous generations that they will necessarily do better than their parents. Their lives will be successful if they simply maintain familial levels of success in economic, health, and family/social domains. For the socioeconomically disadvantaged, however, a potentially accurate understanding of the probability of financial successes starting from that position may temper their generalized beliefs. Yet some may view a path toward an upward intergenerational trajectory as feasible, and in rare cases such beliefs may be motivational, though in general we suggest they are less so. Although adolescents from lower socioeconomic status were more likely to fall in all three other groups (compared to the high GLE – low ICE group characteristic of high SES adolescents), those from lower income families were especially likely to fall into the low GLE – high ICE category (quadrant 3) in which they were not particularly optimistic about their likely success in life, but anticipated some relative improvement across generations.

The two intergenerational comparison items we examined here reference a ‘better’ life, but unlike our other scale, which averaged expectations across life domains (friends, social standing, health, material success, fulfillment at work), we do not know in what ways our respondents are thinking life will be better or not in comparing generations. Future research needs to unpack what adolescents are thinking when they say they or their children can have a “better” life, and whether that varies systematically across adolescents.

Our second objective was to better understand the influence of life expectations on the subsequent life course, testing whether the benefits of positive expectations were shared equally across levels of (dis)advantage. Classic status attainment literature as well as more recent work have established the utility of future beliefs for contributing to life course outcomes; we build on this literature through focusing on the constellations of life course expectations adolescents hold and for whom they are beneficial. We first consider the main effects of life expectations. At a basic level, our findings strongly support the established claim that expectations matter across a range of life course domains; for 10 of the 14 outcomes we selected to examine, it was better to be in at least one higher expectation combination group than it was to hold low expectations on both dimensions.

With respect to the main effects of life course expectations, the “best” outcomes were found among those with high expectations for success in life but lower expectations for relative improvement across generations (quadrant 2), although in all but two instances the group with high expectations on both dimensions did equally well (quadrant 4). Overall, then, higher expectations for life generally predict positive life course outcomes; the key force appears to be generalized life expectations (GLE), not intergenerational comparative expectations (ICE), at least with these available measures. The two instances where those with high expectations on both dimensions did not do as well were with respect to financial problems and recent experience with unemployment. Perhaps these outcomes share something in relation to risk taking, representing instances where intergenerational comparative expectations fostered higher risk strategies for pursuing success, or perhaps they are tied to overly-optimistic self-assessments in adolescence.

In contrast to generalized expectations, intergenerational comparative expectations do not seem to predict better life course outcomes. Both groups with higher ICE (quadrants 3 and 4) fail to show statistically significant differences in the outcomes compared to their counterparts with lower ICE (quadrants 1 and 2, respectively), except in the two instances noted above where having high comparative expectations was harmful (financial problems and unemployment). In three instances, self-rated health, self-esteem, and ratings of success at work, high ICE expectations with high GLE expectations (quadrant 4) is better than low expectations on both (quadrant 1), but that is not definitive evidence for the power of ICE, as high ICE also predicted worse outcomes in the two instances noted above.

For the most part, our tests of the cumulative disadvantage and resource substitution hypotheses produced nonsignificant estimates of conditional effects, indicating that the effects of life expectations were shared equally across levels of socioeconomic (dis)advantage. Our analysis of 14 aspects of attainment, health, and well-being found significant differences in the effects of expectations in only two instances, both for parent education levels. Although there were only two, both were consistent with a resource substitution hypothesis. Life expectations mattered more for lower SES youth with respect to adult levels of self-esteem and home ownership.

At the same time the findings offer limited support to the resource substitution hypothesis, the specific pattern was different than we anticipated, in that it was disadvantaged youth in the high GLE-low ICE group (quadrant 2) and not those anticipating intergenerational improvement (quadrants 3 and 4) that achieved the most, mirroring the hierarchy across groups evident in the main effects models. We had reasoned that the high GLE-low ICE group might reflect a sense of confidence-yet-realism when held by higher SES youth, but the same constellation would represent life expectations that were not well thought out for lower SES youth. For disadvantaged youth, anticipating high chances of success across many valued life domains (as represented in GLE) ought to come with an expectation of intergenerational mobility; this high GLE-low ICE quadrant suggests some misunderstanding of the nature of either societal structure or personal potential trajectories. The answer may lie in the evaluative comparison adolescents are making—in other words how successful do they see their lower socioeconomic status families? This may be a key aspect of unpacking what different meanings adolescents hold in anticipating a ‘better’ life, and so we reiterate our earlier call to further develop an understanding of what adolescents are thinking in their intergenerational comparative expectations and whether it systematically varies.

In the end, with such a limited pattern of conditional effects, we cautiously conclude that disadvantaged youth benefit at least as much, and possibly more, from developing high expectations for success in life, apart from whether they expect their lives to improve across generations. As such, even though disadvantaged youth are less likely to develop this combination, when they do it is at least as much of an asset for their lives as it is for more advantaged youth. Being more common among high SES youth, holding this combination of expectations serves to reproduce inequality across generations. At the same time, because it is beneficial among disadvantaged youth, it represents the type of resource that can facilitate what life span psychologists refer to as resilience (Luthar et al., 2000). Addressing childhood disadvantage directly would, according to our findings, facilitate agentic beliefs associated with long-term benefits and reduce the need for resilience.

Our longitudinal data do not, unfortunately, allow us to trace the development of these orientations over the transition to adulthood. Referents for optimistic beliefs may change during the life course. For example, Galatzer-Levy et al., (2010) find that, prior to unemployment, people used social comparisons to judge success, but after unemployment they ceased making these comparisons (Galatzer-Levy, Bonanno and Mancini 2010). This idea that adolescents can, in general, ‘bounce back’ from hardship is a subject of debate in psychology; finding resilience within a population appears to be sensitive to statistical model specifications (Infurna and Luthar 2016). Overcoming disadvantage, in our sample, is aided – but statistically rarer – by developing higher GLE, more typically found among the advantaged adolescents and seemingly more important than locating one’s future success within an intergenerational comparative framework.

Future work might profitably compare cohorts on intergenerational comparative expectations; one limitation of our study is that it does not necessarily represent contemporary adolescents’ views of the future. Analysis of this earlier cohort offered the key advantage of being able to assess the impact of adolescents’ expectations on subsequent life outcomes, but what contemporary adolescents expect and the extent to which future outlooks continue to be patterned by socioeconomic status cannot be evaluated with these data. Among the range of historical changes documented over the past several decades are a rise in inequality and in adolescents’ educational and occupational ambitions. The rise in expectations for college degrees has come disproportionately from adolescents with lower socioeconomic backgrounds (Goyette 2008; Schneider and Stevenson 1999), and education is a key anticipated mechanism for intergenerational mobility. Such changes could mean higher GLE among lower SES youth and a weakened association between SES and beliefs. Additionally, future work might explore whether a wider range of ICE comparisons finds a stronger effect for this set of orientations, considering intergenerational comparators, but also the self and relevant peers. Moreover, there is a need to distinguish both beliefs about the stratification system and belief in the American Dream as an ideology from expectations applied to the self. Future studies might distinguish between beliefs in the meritocratic system, itself, contrasted with one’s belief in being able to successfully navigate that system.

Individuals direct their lives within the bounds of social structure (Elder et al. 2003), in part through the beliefs they hold about themselves, their circumstances, and their futures (Gecas 2003; Hitlin and Johnson 2015). We build on sociological work that highlights the power of future-oriented beliefs, including high expectations, aspirations, or a general sense of optimism, for the achievement of stratified outcomes and general well-being. Ultimately, this study supports the argument that generalized future beliefs are an important social force across the life course, more so than intergenerational comparisons about potential life course trajectories. Although intergenerational comparative expectations played an important role in understanding how structural origins related to the development of expectations about the future in adolescence (research question 1), they had a much less robust role in predicting outcomes in adulthood (research question 2). As such, our findings offer a refinement of the measure Hitlin and Johnson (2015) propose as a future-oriented dimension of agency, suggesting the eight items we label generalized life expectations (GLE) were likely driving those patterns. In addition, we find that while life course expectations are influenced by structural position, their advantages accrue to whichever adolescents are able to develop them, especially generalized beliefs about likely future life successes. Given their grounding in structural advantage, however, the greatest potential gains would come from reducing structural disadvantage.

Biographies

Monica Kirkpatrick Johnson is Professor of Sociology at Washington State University. Her research focuses on education and work-related processes during adolescence and the transition to adulthood, and particularly social psychological experiences therein. Her recent research examines families’ and young people’s adaptations to the changing nature of the transition to adulthood. She is co-editor (with Michael Shanahan and Jeylan Mortimer) of the 2^nd volume of the Handbook of the Life Course (Springer 2016).

Steve Hitlin is a professor of sociology at the University of Iowa. His work covers areas of morality, social psychology, and the life course. His recent publications have appeared in the American Journal of Sociology (with Monica Johnson), Society and Mental Health, and the Annual Review of Sociology.

Appendix A. Items Included in Scales, Youth Development Study

Adolescence

Parents’ Intrinsically Rewarding and Self-Directed Work

“How much control do you have over the way you spend your time at work – over when and how long you work on the various parts of your job?” (1 = almost no control at all, 2 = some, but not much control, 3 = a fair amount of control, 4 = a great deal of control, 5 = complete control over the way I spend my time at work)

“Overall, how much freedom do you have to make important decisions about what you do at work and how you do it?” (1 = almost none at all, 2 = some, but not much freedom, 3 = a fair amount of freedom, 4 = a great deal of freedom, 5 = complete freedom in deciding what I do and how I do it)

“Do you have to think of new ways of doing things or solving problems on your job?” (1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = almost always)

“How often are you interested enough to do more work than your job requires?” (1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = almost always)

“Overall, how challenging would you consider your present job?” (1 = not at all challenging, 2 = only a little bit challenging, 3 = somewhat challenging, 4 = very challenging)

“How interesting is your job?” (1 = not at all interesting, 2 = slightly interesting, 3 = fairly interesting, 4 = very interesting)

Mastery

“How strongly to you agree or disagree with the following statements” (1 = strongly disagree; 2 = disagree; 3 = agree; 4 = strongly agree)

There is really no way I can solve some of the problems I have (reversed)

Sometimes I feel that I’m being pushed around in life (reversed)

I have little control over the things that happen to me (reversed)

I can do just about anything I really set my mind to do

What happens to me in the future mostly depends on me

I often feel helpless in dealing with the problems of life (reversed)

There is little I can do to change many of the important things in my life (reversed)

Adulthood

Social Support

“When things get rough, do you have a friend (or friends) who you can really talk to, someone you can turn to for support and understanding?” (1=no, I don’t; 4=yes, I am sure I do)

“How close do you feel to your closest friend?” (1 = not at all close; 4 = extremely close)

“Do you think there will always be people you can count on in the future?” (1 = I doubt there will be; 4 = I am certain there will be)

Self-Esteem

“How strongly to you agree or disagree with the following statements” (1 = strongly disagree; 2 = disagree; 3 = agree; 4 = strongly agree)

I feel I have a number of good qualities

I certainly feel useless at times (reversed)

I feel I do not have much to be proud of (reversed)

I take a positive attitude toward myself

On the whole, I am satisfied with myself

At times I think I am no good at all (reversed)

I wish I could have more respect for myself (reversed)

Depressive Affect

“During the past month, how much of the time” (1 = not at all; 5 = all of the time)

Have you felt that the future looks hopeful and promising? (reversed)

Have you been under any strain, stress, or pressure?

Have you generally enjoyed the things you do? (reversed)

Have you felt calm and peaceful? (reversed)

Have you felt downhearted and blue?

Have you felt cheerful or lighthearted? (reversed)

Have you been moody or brooded about things?

Have you been in low or very low spirits?

Perceived Financial Problems

“How much stress have you felt in meeting your financial obligations during the past year?” (1 = not at all stressful; 7 = extremely stressful)

“How difficult is it for you to pay your bills on time? These bills might include insurance, rent, mortgages, car payments, credit cards, etc.” (1 = not at all difficult; 7 = extremely difficult)

“How much of a burden do you feel from debt (from credit cards, mortgages, personal loans, etc.)? (1 = no burden at all; 7 = extremely high burden)

Intrinsic Work Rewards

“Do you have to think of new ways of doing things or solving problems on your job?” (1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = almost always)

“How often do you feel bored at work, or that time is dragging?” (1=always; 5=almost never)

“How often do you feel that your work is meaningful and important?” (1 = almost never; 5 = always)

“How true is each statement” (1 = not at all true; 4 = very true)

My job gives me a chance to learn a lot of new things.

My job uses my skills and abilities.

My job gives me a chance to be helpful to others.