LOW-INCOME EMPLOYEES IN THE UNITED STATES

LOW-INCOME EMPLOYEES IN THE UNITED STATES James T. Bond and Ellen Galinsky Families and Work Institute November 2012 This report is funded by the Ford Foundation as part of its efforts to understand and address the needs of low-income employees in the United States. The research findings presented here are drawn mainly from the 2008 National Study of the Changing Workforce (NSCW) conducted by Families and Work Institute (FWI). The survey sample was randomly selected to represent the entire U.S. workforce. This report looks only at wage and salaried employees (N=2769) who are employed by someone else. For purposes of this report, the sample is weighted to the Current Population Survey on various factors. 1 This is the first of two reports addressing this topic; the second is entitled What Difference Do Job Characteristics Make to Low- Income Employees. It is important to note that we do not use the terms income and earnings interchangeably in this report. Rather, we follow the formal distinction made by economists, defining earnings as the wages or salary earned by an individual from paid work and income as the total income from all sources of a family or household unit. HIGHLIGHTS Arguably, the most striking and important differences between low-income employees and their more economically advantaged counterparts are apart from income age, education and race or ethnicity: On average, low-income employees are much younger than others: 38% of low-income employees are young adults 18-24 years old, while only 8% of middle- income and 3% of high-income employees are. On average, low-income employees also have much less education than others: 64% have no more than a high school education, while only 37% of middle-income employees and 18% of high-income employees have no more than a high school education. Low-income employees are much more likely than others to belong to racial or ethnic minorities: only 50% are non-hispanic Whites versus 73% of middle-income and 79% of high-income employees. The ways in which age, education and racial/ethnic background interact to affect income are explored at greater length in following sections of this report. Below, we touch only on major highlights. 1 The March 2007 Current Population Survey provided control totals for calculating sample weights that is, sample proportions were adjusted to CPS proportions. The weighting algorithm included the following demographic factors: number of eligibles in household, gender, education level completed, race/ethnicity (non-hispanic White, non-hispanic Black, Hispanic, other) and age. 1

Age: Obviously age does not, in and of itself, directly determine how much one earns on the job or how much income one s family has. However, age is strongly associated with other factors that do have a direct bearing on wages and, in turn, income: Younger employees have less experience in the workforce than older employees, which reduces their employability and earnings power. On average, younger employees have worked fewer years for their current employers i.e., have less job tenure which is partly a matter of having spent less time in the workforce because of age and partly a matter of having frequently changed jobs something that is characteristic of young workers in general. 2 Longer job tenure is associated with higher earnings. Younger employees are also less well educated on average than older employees, in part because they have had less time to complete their educations, and less education means less earning power. Together, these age-related factors significantly depress earnings and thereby, income. In addition: Younger employees are also much less likely than their older peers to have a spouse or partner or, more importantly a spouse or partner who is also employed and contributes to family income. It goes without saying that having a second earner in the family will increase family income on average. Note that our study does not include employed youth under 18 years of age who are minors, typically living with their parents, going to high school and earning low wages in part-time jobs. Education: Having a good education has long been touted as the best way to succeed in life, and low-income employees have much less of it than others. Although it used to be that youth were encouraged to finish high school in order to be able to get a decent job, today they are urged to get a college education or more. 2 High job turnover is endemic among young workers both lower and higher income as they complete their schooling, change residence and explore different employment options. For example, the U.S. Department of Labor, Bureau of Labor Statistics reports that a representative national sample of young adults interviewed in 2009-10 had held an average of 5.4 jobs from age 18 to age 24 (Publication: USDL-12-0216, February 2012). 2

Only 9% of low-income employees have four-year college degrees or more versus 29% of middle-income employees and fully 61% of high-income employees. As noted above, this discrepancy in education level is due partly to age, since many young employees who are disproportionately represented among low-income employees have not yet had the opportunity to complete their schooling. This age effect is also reflected in current educational enrollment: Significantly more low-income employees than their more advantaged peers are currently enrolled in education and training programs. An important factor impeding the educational progress of low-income employees is affordability. As we are frequently reminded in the daily news, the costs of postsecondary education are soaring and financial aid remains limited. Consequently, many low-income employees must work in order to afford continuing education and training, and part-time employment is often a viable solution. Low-income employees are much more likely (41%) than others (12% of middle-income and 9% of high-income employees) to have part-time jobs. Fully 61% of low-income part-timers would like to be working more hours. Of these, 26% cannot find a job offering more hours and 23% cannot find a job that offers them the schedule flexibility they need. Although the survey did not ask specifically about the demands of schooling and training as obstacles to working more hours for the 28% who are in school or training programs, it seems very likely that this accounts for some part of the 23% who can t get the flexibility they need and the 50% whose reasons for not working more hours were classified as other by interviewers. Note that hourly employees (88% of low-income employees) earn less when they work less. Thus, parttime employment results in significantly lower earnings, which is associated with lower income and, of course, less money for continuing education. Race/ethnicity: Like age, belonging to a racial or ethnic minority group does not directly determine wage and income, but it is strongly related. Although outright discrimination based on racial/ethnic background (or for that matter, on gender) may no longer have the socioeconomic impact it did a generation or two ago because of broad cultural 3

changes as well as civil rights legislation, few would argue that discrimination has disappeared. To what extent persistent discrimination accounts for the current relationship between minority status and economic well-being cannot be determined using data from our study, but should be considered when addressing issues of the lowincome workforce. For example, recent longitudinal research strongly suggests that those who are born into low-income families, which are disproportionately ethnic and racial minorities, are significantly more likely to live in low-income families as adults than those who are born into families with greater economic means and vice versa. As discussed later in this report, educational achievement alone is not sufficient to overcome this disadvantage in many cases. Thus, even in the absence of overt racial/ethnic discrimination, the economic disadvantage of racial/ethnic minorities in one generation is, to an important degree, self-sustaining in the next. Although we might expect this relationship to continue to diminish over time, its dogged persistence is deeply disturbing. In addition, its negative impact may be exacerbated by structural economic factors that have fostered growing income disparity in the United States as measured by economic differences between the rich and the poor. DEFINITIONS Although this paper focuses on low-income employees, we touch on the issue of lowwage employees here because of the relationship between earnings and income. We define low-wage employees as those whose earnings in their main jobs fall in the bottom 25% of the earnings distribution $10.00 per hour or less in 2008 dollars when our study was conducted. In communities with higher costs of living, one might reasonably argue that the low-wage cut off should be significantly higher. Nonetheless, the definition used here seems appropriate since we are examining a random sample of employees from the entire U.S. population. In the summer of 2007, the federal minimum hourly wage that was increased from $5.15 to $5.85. Although low-wage employees by our definition earned up to $4.15 more than the minimum wage of $5.85 prevailing during the period of our data collection, a family of four that depended upon the income 4

of a single full-time low-wage earner in 2008 would be poor with maximum family income below the (100%) poverty threshold of $22,025. For Table 1, we turn to the Current Population Survey conducted by the U.S. Bureau of the Census to estimate the approximate numbers of the U.S. wage and salaried workforce in different wage/salary groups at the time of the 2008 NSCW survey. As one can see, there are substantial numbers of low-wage earners in the U.S. workforce. Table 1: Approximate Number of Low, Middle and High Earners 18 or Older in the U.S. Wage and Salaried Workforce Low Earners: bottom 25% of earnings distribution Middle Earners: middle 50% of the earnings distribution High Earners: top 25% of the earnings distribution 28,000,000 56,000,000 28,000,000 Estimates based on data from the 2007 March Current Population Survey. As important as individual earnings are, however, family income is a much more important measure of overall economic well-being. Thus, this report focuses on employees who are from low-income families. Although earnings are correlated with family income, the correlation is only moderate in the workforce as a whole. Family income, the focus of this report, includes all reported income of the employee and his or her spouse or partner from any source i.e., earnings from employment, tax credits, investment income, social security benefits, other public benefits and so forth. We define low-income as annual family income below 200% of the federal poverty threshold, which takes into account the number of persons in the family i.e., employee, spouse and children under 18. Families below 100% of the poverty threshold are considered poor, while those between 100% and 200% of the poverty threshold are sometimes referred to as near-poor. Two times, or 200% of, the 2008 federal poverty threshold for a family of four is $44,050. All low-income families are eligible for various state and federal benefits and programs including the federal Earned Income Tax Credit (EITC), which enhances economic security by increasing working families disposable incomes. Many low-income families are also eligible for other public benefits depending upon their state of residence e.g., health insurance, child care subsidies and so forth. 5

Families with annual incomes of 650% or more than the federal poverty threshold are considered high income for purposes of this report, while families in the middle range between 200% and 650% of the federal poverty level are defined as middle income, though middle-income families close to our threshold for low-income are just a step away from poverty. The median annual family incomes of low, middle and high-income families in the sample are presented in Table 2. As can be seen, the range of family incomes among employees in the United States is very broad, and the difference between those at the top of the income distribution and those at the bottom has increased steadily over recent decades resulting in greater income inequality. Table 2: Median Annual Family Incomes of Low, Middle and High-income Families Low Income: less than 200% of the $15,600 poverty threshold (N=601) Middle Income (N=1,482) $62,400 High Income: 650% or more of the poverty threshold (N=567) $140,400 Source: 2008 National Study of the Changing Workforce, Families and Work Institute. Family income data are missing for 4% of respondents, thus the Ns reported above do not add to the total wage and salaried sample size of 2,769. For Table 3, we return to the 2007 Current Population Survey conducted by the U.S. Bureau of the Census to estimate the approximate numbers of the U.S. wage and salaried workforce in different family income groups at the time of the 2008 NSCW survey. The low-income group represents approximately the bottom quintile (20%) of the income distribution, while the high-income group represents approximately the top quintile (20%). Comparisons of the top and bottom quintiles are often made to assess the degree of income inequality. Table 3: Approximate Number of U.S. Wage and Salaried Employees 18 or Older Living in Low, Middle and High-Income Families Low Income: less than 200% of the 26,000,000 poverty threshold Middle Income 63,000,000 High Income: 650% or more of the poverty threshold 6 24,000,000 Estimates based on data from the 2007 March Current Population Survey.

Although employees who live in low-income families are most at risk, employees and their families who fall into higher-income groups are also vulnerable to life events such as job loss, health problems, divorce and so forth that can push them into lower income groups or even into poverty. During the deep recession when this survey was conducted, the number of unemployed people with no earnings at all, as well as the number of underemployed workers with substantially reduced earnings, increased dramatically. Concomitantly, the incomes of many families declined dramatically. While our cross-sectional survey does not record these changes for individual employees and their families, it does provide a great deal of concurrent information about low income employees whose lives on and off the job have been affected by the troubled U.S. economy. Economic Update In the period surrounding 3 and following data collection for the 2008 NSCW (November 2007 March 2008), the U.S. economy has experienced economic challenges rivaled only by the Great Depression of some 80 years ago including doggedly persistent long-term unemployment, widespread underemployment and growing income inequality. The U.S. wage and salaried workforce actually shrank by nearly 4% from December 2007 to December 2011. 4 Consequently, the actual number of low-wage earners, defined as the bottom 25% of the earnings distribution, has shrunk accordingly. The U.S. unemployment rate increased from 5% in December 2007 to 8.5% in December of 2011. Though the rate of unemployment has gradually improved from 10% in October of 2009, it remains high and continues to impede economic recovery. Over the same period from December 2007 to December 2011, the proportion of the unemployed without employment for 52 weeks or more increased from 9.1% to 30.6%. Coincidentally, underemployment among wage and salaried employees due to involuntary part-time work 5 increased from 4.53 million (3.4%) to 7.95 million (6.1%) 3 On September 20, 2010, the National Bureau of Economic Research declared that the so-called great recession from which our economy is still recovering began in December 2007 and ended in June 2009 [http://www.nber.org/cycles.html]. 4 Data extracted from the Current Population Survey conducted monthly by the U.S. Bureau of the Census and made available by the U.S. Department of Labor, Bureau of Labor Statistics for public use. 5 Part-time is defined as fewer than 35 hours per week often far fewer. 7

from December 2007 to December 2012. 6 Not only did real earnings per hour (the value of earnings after inflation) stall during this period 7, but because growing numbers of employees have worked reduced hours, the incomes of many working families have declined. At the same time, the rich have grown richer: Inequality first began to rise in the late 1970s and early 1980s in some Anglophone countries, notably in the United Kingdom and the United States, followed by a more widespread increase [among OECD nations] from the late 1980s on. The most recent trends show a widening gap between poor and rich in some of the already high-inequality countries, such as Israel and the United States. 8 All of this is to say that the economic situations of many employees and their families have deteriorated since our survey was conducted. 6 All unemployment and underemployment data extracted from the Current Population Survey conducted monthly by the U.S. Bureau of the Census and made available by the U.S. Department of Labor, Bureau of Labor Statistics for public use. 7 Based on a comparison of the median weekly earnings of wage and salaried employees expressed in constant dollars in the fourth quarters of 2007 and 2011. (Data extracted from the Current Population Survey conducted monthly by the U.S. Bureau of the Census and made available by the U.S. Department of Labor, Bureau of Labor Statistics for public use.) 8 Organisation for Economic Cooperation and Development (May 2011). Growing Income Inequality in OECD Countries: What Drives It and How Can Policy Tackle It? Paris: OECD. 8

FINDINGS 9 The low-wage workforce is generally disadvantaged in many ways, but it is also quite diverse and many of these employees defy common stereotypes. For example, it may come as a surprise to learn that: Although the majority of low-wage employees (six in ten) live in lowincome families, four in ten do not. Thus, we should not assume that all low-wage employees are at immediate and extreme economic risk. However, among low-wage employees living in families above 200% of the federal poverty threshold, many are but a step away from poverty. Indeed, a large portion (43%) live in families below 300% of poverty families that are often referred to as lower-middle class or might even be considered low-income when they live in areas with higher costs of living. Furthermore, since low-wage employees in these families on average contribute half or more of family earnings, income from their jobs may lift their families above the poverty line. How Do Low Income Employees Differ from Their Higher Income Counterparts? 10 The 2008 National Study of the Changing Workforce (NSCW) provides a wealth of data about the lives of U.S. employees both on and off the job. Here we focus on a broad range of measures that suggest the causes and document the consequences of low income status. Demographic Characteristics Table 4 provides an overview of the basic demographic characteristics of employees in different income groups, highlights of which are described below. 9 Percentage distributions do not always add to 100 because of rounding error. Because of the relatively large sample sizes of the income groups and the large number of statistical tests conducted for this report, findings are only considered statistically significant when they reach significance at p <.01 or better that is, when the finding would occur by chance less than one in 100 times. 10 Column percentages do not always add to 100% because of rounding errors, and group sizes (Ns) reflect a small amount of missing data due to occasional refusals to respond and claims of I don t know. Statistical significance is reported at two levels p <.01 (less than 1 time in 100 being due to chance) and p <.001 (less than 1 time in 1000 being due to chance). 9

Table 4: Basic Demographic Characteristics of Employees in Different Income Categories Basic Demographics Age: 18 24 years old 25 34 years old 35 44 years old 45 54 years old 55 or more years old Median age in years Gender Male Female Highest level of education completed: Less than high school High school or GED Some post-secondary College degree Higher degree % currently in school/training Race/ethnicity: White, non-hispanic Black, non-hispanic Hispanic Other % Living with spouse or partner % of total sample living with an employed spouse or partner % of total sample with any child under 18 at home % of total sample with any child under 3 at home Among parents: % women who are single mothers Among parents: % men who are single fathers Low-Income (n=596) 38% 25 20 12 6 (n=596) 28 (n=601) 45% 55 (n=600) 14% 50 27 8 1 (n=601) 28% (n=599) 50% 21 19 10 (n=600) 36% (n=600) 21% (n=600) 51% (n=600) 18% (n=186) 59% (n=122) 21% Income Group Middle- Income (n=1472) 8% 23 27 23 18 (n=1472) 41 (n=1482) 55% 45 (n=1482) 3% 34 34 20 9 (n=1482) 20% (n=1480) 73% 9 13 5 (n=1480) 70% (n=1480) 56% (n=1481) 46% (n=1481) 11% (n=316) 20% (n=366) 6% High-Income (n=566) 3% 11 27 33 26 (n=566) 48 (n=567) 54% 46 (n=567) <1% 18 21 35 26 (n=567) 19% (n=567) 79% 6 10 6 (n=560) 87% (n=560) 80% (n=566) 33% (n=566) 5% (n=91) 7% (n=96) 3% Sig. ** Source: 2008 National Study of the Changing Workforce, Families and Work Institute. Statistical significance: ** = p <.01; = p <.001; ns = not statistically significant. 10

Low-income employees are substantially younger than others (Figure 1). Figure 1: Income Groups by Age 40% 35% 38% 33% 30% 25% 20% 25% 23% 20% 27% 27% 23% 18% 26% 15% 10% 5% 8% 3% 11% 12% 6% 0% 18-24 years old 25-34 35-44 45-54 55 + years old Low-income Middle-income High-income Source: 2008 National Study of the Changing Workforce, Families and Work Institute Fully 38% of low-income employees are 18 24 years old, while only 6% are 55 or older, suggesting that the socioeconomic status of many younger employees will improve as they mature. This interpretation is supported by the findings that only 8% of middleincome and 3% of high-income employees are under 25 compared with 18% of middleincome and 26% of high-income employees who are 55 or more years old. Of course, mere association does not prove causation, and it is not plausible that age in and of itself determines income status. Rather the relationship between age and family income appears to be mediated by other factors described in Table 4. Low-income employees have much less education than other employees, but more than one in four are working to improve their educational credentials. As shown in Figure 2, 64% of low-income employees have a high school education or less, compared with only 18% of high-income employees who have a high school education or less. Looking at the other end of the educational attainment continuum, only 9% of low-income employees have four-year college degrees or more, compared with 61% of their high-income counterparts. Level of education attained is arguably the strongest single predictor of earnings, both early in one s work career and over time. 11

Figure 2: Income Groups by Education Level 70% 60% 50% 64% 61% 40% 30% 20% 37% 18% 27% 34% 21% 29% 10% 9% 0% High school or less Some post-secondary 4-year college or more Low-income Middle-income High-income Source: 2008 National Study of the Changing Workforce, Families and Work Institute As young employees mature, many continue their educations and, ultimately, attain higher levels. Among all 18 24 year olds (across all income groups), only 9% have a four-year college degree or more. However, the incidence of college graduation increases to 32% among all employees 25-34 years old and stabilizes at approximately that level through the life cycle. It is important to note that 44% of all employees 18 24 years old are currently enrolled in school or training versus only 17% of older employees. The same pattern holds true for low-income employees: 46% of low-income employees 18 24 years old are taking classes or training courses versus only 16% of older low-income employees. And, for all ages as shown in Table 4, significantly more low-income employees (28%) than others (20% and 19%) are currently involved in school or training. Although this section of the report examines demographic characteristics of the workforce, it seems appropriate at this juncture to interject a finding about employee aspirations. Supplemental analyses of survey data revealed that: Twice as many low-income employees (60%) as high-income employees (31%) want to move into jobs with greater responsibility, and among low-income employees under 25 years old, the proportion wanting more responsibility on the job rises to 76%. 12

The fact that 46% of these young low-income employees are also enrolled in school or training programs demonstrates that they are not simply wishing for a better life, but are actually doing something to achieve their goals despite the demands of work and very limited financial resources. Together, these findings suggest that many younger low-income employees will eventually attain higher levels of education and improve their economic circumstances, and early low-wage jobs may even enable continuing education, thus contributing to their upward mobility. With more education, young employees can command better jobs and higher wages allowing some degree of upward socioeconomic mobility within their lifetimes. Longitudinal research confirms this observation: Education especially beyond high school is a primary driver of upward intragenerational mobility. 11 Improving educational opportunities has long been viewed as perhaps the single most important thing that government can do to increase upward socioeconomic mobility and reduce economic inequality. There remains, however, plenty of room for improvement as well as significant challenges. Recent research on children s educational achievement (measured by standardized achievement tests in large-scale national studies conducted over five decades) offers some very good news along with a large dose of bad. Sean Reardon of Stanford University found that while the black-white achievement gap has decreased significantly over the past 50 years very good news, the gap in educational achievement between children in low- and high-income families has widened substantially. 12 Indeed, the lowhigh income achievement gap is now nearly twice as large as the black-white achievement gap, while 50 years ago the black-white achievement gap was one and one half to two times as large as the income gap. Efforts to explain this phenomenon are still tentative and incomplete. Although the growing income achievement gap has 11 Acs, G. & Zimmerman, S. (2008). U.S. Intragenerational Economic Mobility from 1984 to 2004: Trends and Implications. Published by the Economic Mobility Project: An Initiative of the Pew Charitable Trusts. 12 Reardon, S. (2011). The widening academic achievement gap between the rich and the poor: New evidence and possible explanations. In Greg Duncan & Richard Murnane (Eds.), Whither Opportunity? Rising Inequality, Schools, and Children s Life Chances. New York: Russell Sage Foundation. 13

paralleled growing income inequality in the U.S., it is not simply explained by that correlation. Other factors are at play. High-income parents are on average much better educated than low-income parents. Moreover, parental education is now, and always has been, a strong predictor of children s achievement, and the correlation between parental education and family income has increased over time. There is also evidence that parents investments (as measured by both time and money) in their children s cognitive development have increased significantly during the past half-century particularly among higher income and better educated parents differential investments that may well help to explain the growing income achievement gap among children. 13 Growing residential segregation by income that is associated with school quality may also play a role. Whatever the full explanation, Reardon concludes: The combination of these trends creates a feedback mechanism that may decrease intergenerational mobility. As the children of the rich do better in school, and those who do better in school are more likely to become rich, we risk producing an even more unequal and economically polarized society. 14 Just as Reardon finds an income achievement gap among children in elementary grades through high school, Martha Bailey and Susan Dynarski find a significant income gap in college completion. 15 Many low-income youth embark upon college careers illprepared for the academic demands they will encounter, and their educational achievement deficits, in conjunction with limited financial resources that often must be supplemented by employment that competes with studies, is associated with high rates of college dropout and longer times to college completion. The view that low-income employees are disproportionately members of minority groups is supported by NSCW data as shown by Table 4, above. 50% of low-income employees classify themselves as White-non- Hispanic, while 21% classify themselves as Black-non-Hispanic, 19% as Hispanic and 10% as Other. 16 13 Ibid. 14 Ibid. 15 Bailey, M.J. & Dynarski, S.M. (2011). Inequality in postsecondary education.. In Greg Duncan & Richard Murnane (Eds.), Whither Opportunity? Rising Inequality, Schools, and Children s Life Chances. New York: Russell Sage Foundation. 16 The other category includes Asian, Pacific-Islander, Native American, Mixed, and so forth. 14

In stark contrast, 79% of high-income employees classify themselves as White-non-Hispanic, with roughly equal proportions allocated across the three minority groups. Middle-income employees fall in the middle. Figure 3 looks at these same data from a different perspective that of the percentage of low-income employees within different racial/ethnic groups. Overall, 42% of Blacknon-Hispanic employees live in low-income households compared with only 16% of White-non-Hispanic employees. Hispanic employees (32%) and employees belonging to other racial/ethnic groups (36%) are also much more likely to be low-income than Whites. Figure 3: Income Status within Racial/Ethnic Groups 70% 60% 50% 40% 30% 20% 10% 16% 59% 24% 42% 47% 11% 32% 52% 16% 36% 44% 20% 0% White-non-Hispanic Black-non-Hispanic Hispanic Other Low-income Middle-income High-income Source: 2008 National Study of the Changing Workforce, Families and Work Institute According to longitudinal research conducted by Gregory Acs and Seth Zimmerman of the Urban Institute, race/ethnicity has become less predictive of upward intragenerational mobility in recent years. 17 However, the fact that one s race/ethnicity is less important to upward mobility today than a generation ago, does not mean that it is unimportant. Indeed, recent longitudinal research by Bhashkar Mazumder of the Federal Reserve Bank of Chicago (2008) reveals that upward economic mobility from 17 Acs, G. & Zimmerman, S. (2008). U.S. Intragenerational Economic Mobility from 1984 to 2004: Trends and Implications. Published by the Economic Mobility Project: An Initiative of the Pew Charitable Trusts. 15

parent to child is substantially lower for Blacks than Whites. 18 All in all, this is a dismal assessment of the economic opportunity available to Black Americans. Although public policies that improve access to high quality education at all levels and reduce discrimination have certainly improved outcomes for recent generations of Black Americans, greater understanding of the issue and more effective public and private sector action will be required to achieve true equality of opportunity and optimize outcomes. Three additional findings from Table 4 (above) also merit special attention: Low-income employees are much less likely to live with a spouse or partner (36% versus 70% and 87% for the other income groups) or to live with an employed spouse or partner (21% versus 56% and 80% for the other income groups) who supplements family income. Low-income mothers of children under 18 are much more likely to be single mothers 59% versus 20% and 7% and thus, live in families with a single earner. Similarly, low-income fathers are more likely (21%) to be raising children alone and relying on a single source of earnings than their higherincome counterparts (6% and 3%). Although two-earner families are not necessarily well off, they are much more likely to avoid poverty or low-income status. It has become increasingly difficult for families to make ends meet when they depend on a single earning stream, and it is all the more difficult when there are children in the family. Moreover, it is well established that parents who are able to invest more in their children have children who achieve greater socioeconomic success on average. Thus, two-earner families are more able to break the cycle of poverty that leads to persistent economic inequality. This is not to say that low-income single parents can t succeed magnificently with their children. History is replete with such examples. Additionally, it is also very important to note that the likelihood of living in a two-earner couple increases dramatically, and that being a single mother decreases, as employees grow older. Among all employees, only 23% of 18-24 year olds live in two-earner families versus 58% of those 25 and older, and though 56% 18 Mazumder, B. (2008) Upward Intergenerational Economic Mobility in the United States. Published by the Economic Mobility Project: An Initiative of the Pew Charitable Trusts. 16

of all mothers under 25 years old are single, only 29% of mothers 25 and older are single. Not only are low-income employees more likely than others to have children under 18 at home, as might be expected given their age they are also more likely to have very young children: 18% of low-income employees have children under three years old versus 11% of middle-income and only 5% of high-income employees. Finding appropriate and affordable care for very young children is a challenge for many working parents, and particularly so for low-income parents. Public subsidies for child care in the U.S. are not currently sufficient in either magnitude or availability to remove this constraint on employment among low-income families with young children too many of whom quite literally cannot afford to take a job. Table 5 compares the three income groups with respect to their work histories and current employment, factors that have very important bearing on individual earnings and, by way of earnings, family income. 17

Table 5: Work History and Current Employment Work History and Current Employment Years in the labor force since 18 years old: Less than 12 years 12 21 years 22 30 years More than 30 years Low-Income (n=600) 60% 19 13 9 18 Income Group Middle- Income (n=1481) 22% 28 25 25 High-Income (n=567) 9% 20 33 39 Median years in the labor force 9 yrs 22 yrs 28 yrs Years of tenure with current employer: Less than 2 years 2 4 years 5 10 years More than 10 years (n=601) 52% 27 14 7 (n=1482) 20% 25 29 27 (n=568) 17% 17 28 38 Median years of tenure with current employer 1 yr 5 yrs 8 yrs Usual hours (paid/unpaid) worked per week in main job including all hours doing jobrelated work at any time of day, on any day, and in any place: Less than 35 hours 35 40 hours 41 50 hours More than 50 hours % whose main job is considered part-time by their employer % of part-timers who would prefer to work more paid hours Usual hours doing any work related to all jobs (or only job): Less than 35 hours 35 40 hours 41 50 hours More than 50 hours % with more than one job % who would like to work more paid hours % of those who would like to work more paid hours who cannot find a job offering more paid hours Work schedule: Regular daytime shift Evening/night/rotating/split shift or on call % who usually work on Saturday or Sunday % who are hourly (nonexempt) employees Occupation: Manager or professional Service occupation Other occupation (n=597) 41% 43 14 5 (n=596) 41% (n=243) 61% (n=597) 38% 43 12 8 (n=601) 12% (n=594) 44% (n=258) 33% (n=599) 56% 44 (n=601) 38% (n=595) 88% (n=591) 13% 33 54 Source: 2008 National Study of the Changing Workforce, Families and Work Institute. Statistical significance: ** = p <.01; = p <.001; ns = not statistically significant. (n=1465) 12% 42 33 13 (n=1479) 12% (n=169) 36% (n=1454) 10% 39 31 20 (n=1482) 14% (n=1441) 15% (n=211) 46% (n=1480) 78% 22 (n=1481) 23% (n=1465) 70% (n=1455) 34% 11 55 (n=564) 9% 27 41 23 (n=567) 9% (n=44) 39% (n=558) 8% 23 40 30 (n=567) 16% (n=555) 6% (n=33) 18% (n=557) 81% 19 (n=567) 18% (n=557) 42% (n=552) 62% 5 33 Sig. ns

It is obvious that greater work experience in general and tenure with one s employer in particular would generally have a positive impact on earnings and indirectly, by way of earnings, on family income: Low-income employees have spent significantly fewer years in the labor force (Figure 4) and, therefore, have much less work experience. Figure 4: Income Status by Years of Work Experience Since 18 70% 60% 60% 50% 40% 30% 20% 10% 22% 9% 19% 28% 20% 13% 33% 25% 25% 9% 39% 0% Fewer than 12 Years 12-21 yrs 22-30 yrs More than 30 years Low-income Middle-income High-income Source: 2008 National Study of the Changing Workforce, Families and Work Institute Low-income employees have also spent significantly less time working for their current employers in large part because they are younger than more advantaged employees. Not only are work experience and job tenure predictive of earnings, both are also strongly associated with age, helping to explain the relationship between age and income status. Returning to the issue of intragenerational mobility touched on above, as young employees mature they gain work experience and have more opportunities to find jobs that fit, which results in longer tenure. Both experience and tenure are likely to lead to increased earnings and socioeconomic well-being over the life span. Other aspects of employment are also rather obviously important to earnings and, thereby, to family income: 19

Low-income employees work significantly fewer hours than others, and consequently, as predominantly hourly employees, they tend to earn less. The prevalence of part-time employment among low-income employees makes this point (Figure 5). Figure 5: Family Income Status by Part-Time Employment 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 41% 12% Part-time 9% 59% 88% Full-time 91% Low-income Middle-income High-income Source: 2008 National Study of the Changing Workforce, Families and Work Institute 41% of employees in low-income families have part-time jobs versus only 12% of those in middle-income and 9% of those in high-income families, and 61% of low-income part-timers would like to work more paid hours. Importantly, low-income employees as a group (both part- and fulltimers) are much more likely (44%) than others (15% and 6%) to want to work more paid hours than they currently work, but many are unable to find jobs offering more hours, while others are unable to arrange the schedule flexibility they need to add additional work hours or to find affordable and appropriate child care. This is to say, employees in low-income families are more likely than others to be underemployed and involuntary part-time workers. Although having less work experience and lower educational qualifications than their more advantaged peers make it more difficult to secure good, full-time employment, another factor contributing to part- 20

time employment appears to be their greater participation in education and training programs while also working for pay. Among all employees, those enrolled in education or training programs are twice as likely to work part-time as those not in school or training, and low-income employees are more likely to be in school or training. It is also the case that among all employees, those under 25 years old are much more likely to be enrolled in school or training (50%) than those over 25 (13%), and low-income employees are also much more likely than others to be under 25. Although majorities of employees in all income groups work regular daytime schedules, low-income employees are much less likely to do so: 44% of low-income employees work evening, night, rotating or split shifts or are on call versus 22% of middle-income employees and 19% of highincome employees. Moreover, low-income employees are more likely (38%) than others (23% and 18%) to work weekends. Non-daylight shifts and weekend hours pose particular challenges for working parents dependent upon child care centers, most of which operate during daylight hours from Monday through Friday. Variable work schedules are also be difficult for child care centers to accommodate. And, as noted in Table 4 (above), low-income employees are more likely than others to have children under three for whom care is most difficult to arrange and most expensive. Moreover, they are more likely to be single parents both mothers and fathers unable to call upon their spouse/partner for child care assistance. Employees in low-income families typically work in lower status and lower paid jobs only 13% work in managerial/professional jobs versus 62% for high-income employees, while 33% work in service sector jobs versus 5% for those in high-income families. More work experience, longer job tenure, and more education improve one s employment options and the odds of being upwardly mobile. The findings presented in Tables 4 and 5 (above) would appear to explain much of the earnings and income gap between low-income employees and more advantaged employees. 21

Briefly Revisiting the Relationship between Age and Income: We have made much some would say way too much of the positive correlation between employees age and income in the preceding pages. Obviously, the risk of low-income status does not somehow evaporate with age. At the same time as we linked income with age, however, we have tried to explain what accounts for this relationship. For example, as employees grow older, they tend to complete more education, gain more work experience, work longer for the same employer, get married to someone who is also employed, work longer hours and earn more all of which contribute to higher income. Nonetheless, these outcomes are not the inevitable consequences of growing older. Indeed, about 10% of all employees who are 45 years old or older live in lowincome families/households. By comparison with higher income employees of the same ages, they not only earn less at their jobs, but also have much less education; are much less likely to live in dual-earner households; are more likely to be never-married, divorced or widowed; are more likely to be employed part time; and are much more likely to want to work more paid hours. Importantly, older low-income employees also experience poorer physical health and much worse mental health on average than their higher-income peers conditions that constrain economic success and frequently compromise social relationships and general well-being as well. On a particularly discouraging note, we find that African-American employees are nearly four times as likely as others to be low-income as they move into and through the so-called prime earning years. Among employees 45 and older, opportunities for upward socioeconomic mobility are typically quite limited, even diminishing. Either they ve already made it into the middleclass or it becomes increasingly unlikely they ever will. And risks of downward mobility remain, if not increase, with age among those who have made it due to divorce, death of an employed partner, ill health or job loss perhaps followed by re-employment in a lower-paying position. America s economic safety net may help break one s fall, but it does not prevent serious injury or relative impoverishment as witnessed by the current extent of unemployment, food stamp receipt, home foreclosures and erosion of retirement savings in the wake of the last recession. 22

How Is Socioeconomic Background Related to Current Income Status and Upward Mobility? Although information about employees backgrounds from the cross-sectional NSCW survey is limited, it certainly suggests (Table 6) that low-income employees come from more socioeconomically disadvantaged backgrounds than others: By comparison with employees in middle- and high-income families, those in low-income families were raised by parents with significantly less education and in families that were, in their view, somewhat less-well-off than the average American family. Employees in low-income families are also less likely to feel that they have the financial support they need from their families. Table 6: Socioeconomic Background Socioeconomic Background Mother s education: Less than high school High school or GED Some post-secondary education Four-year college or more Father s education: Less than high school High school or GED Some post-secondary education Four-year college or more Respondent assessment of the financial situation of family in which they grew up by comparison with the average American family: Much better off Somewhat better off About the same Somewhat worse off Much worse off Agree that they have the financial support needed from family members? Strongly agree Somewhat agree Somewhat disagree Strongly disagree Respondent expectation of own future financial situation (if under 40) or assessment of current situation (if 40 or older) by comparison with family in which they grew up: Much better off Somewhat better off About the same Somewhat worse off Much worse off Low-Income (n=551) 24% 44 13 19 (n=506) 24% 45 12 20 (n=601) 7% 29 32 20 11 (n=591) 43% 27 11 19 (n=601) 25% 42 18 11 4 Income Group Middle- Income (n=1409) 18% 48 15 19 (n=1352) 22% 42 12 24 (n=1483) 8% 25 39 20 7 (n=1448) 51% 28 9 12 (n=1483) 29% 41 21 8 1 High-Income (n=550) 14% 47 14 26 (n=552) 16% 38 14 32 (n=567) 10% 28 34 20 7 (n=541) 60% 26 5 9 (n=567) 46% 36 14 4 <1 Sig. ** 23

Source: 2008 National Study of the Changing Workforce, Families and Work Institute. Statistical significance: ** = p <.01; = p <.001; ns = not statistically significant. Respondents were also asked whether they thought they would eventually be better off (if under 40 years old) or are currently better off financially (if 40 years old or more) than the family in which they grew up. Figure 6 summarizes the findings. Figure 6: Expect to Be or Are Better Off Financially Than Family in Which Grew Up 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 82% 70% 67% 21% 18% 15% 14% 9% 4% Worse off About the same Better off Low-income Middle-income High-income Source: 2008 National Study of the Changing Workforce, Families and Work Institute Although some respondents may be engaging in wishful thinking, the majority of respondents in all three wage and income groups (67%-82%) believe they already are or eventually will be better off financially than the families in which they grew up an enduring tenet of the American dream, while relatively few (15%-4%) think they will be worse off. As the country plunged into deep recession as this study was being conducted, such optimism may have eroded. Of course, the actual extent of intergenerational socioeconomic mobility can only be satisfactorily measured with longitudinal data. The Issue of Socioeconomic Mobility in America We will not attempt a comprehensive review and discussion of published longitudinal research regarding economic mobility across generations i.e., from parent to child in 24

this paper. Given the limitations of the cross-sectional research presented here, however, brief attention to important recent longitudinal research conducted by others is warranted. Americans tend to be extremely tolerant of large income differentials in society when they believe that upward socioeconomic mobility is widespread and largely a matter of hard work and playing by the rules. But just how mobile are Americans within the space of their own lifetimes and across generations? We have already addressed the issue of intragenerational mobility related to education, using both our cross-sectional data and findings from longitudinal research. We repeat: Education particularly schooling beyond high school is a primary, consistent driver of upward [intragenerational] mobility. [In contrast] getting a high school diploma rather than dropping out has become less important to upward mobility. 19 There appears to be substantial upward intragenerational mobility within at least the first decade of adult lives (related to continuing education, increasing job experience, longer job tenure, and other factors previously discussed). Recent research has also shed important new light on the issue of intergenerational mobility. For example, according to a review by Ron Haskins of the Brookings Institution, 23% of adult children (40 years old) whose parents (at the same age) were in the top 20% (top quintile) of the income distribution end up in the top 20% themselves even when they do not have a college degree. In contrast, only 19% of adult children of parents from the bottom 20% (bottom quintile) of the income distribution end up in the top 20% even when they have college degrees. In short, even after controlling for education levels, children from wealthy families are likely to have a substantial advantage over children from poor families. Although education does not eliminate the effects of dramatically different family backgrounds, it does contribute to relative mobility. Specifically, adult children of parents from all levels of the income distribution 19 Acs, G. & Zimmerman, S. (2008). U.S. Intragenerational Economic Mobility from 1984 to 2004: Trends and Implications. Published by the Economic Mobility Project: An Initiative of the Pew Charitable Trusts. 25