Contents

• Introduction
• About This Report
• How We Define Families
• How We Treat Income
• Family Income and Poverty Status
• Parents' Work and Family Income Status
• Parents' Wages, Job Attachment, and Family Earnings Adequacy
• Wage Rates Relative to Family Poverty Income Levels
• Factors Affecting the Determination of Wages
• Parents' Wages and Educational Attainment
• Wage Rates by Parents' Age
• Parents' Age and Poverty Status by Family Type
• Parents' Wages by Occupation
• Parents' Work During the Year
• Parents' Annual Hours Worked
• Weeks Worked and Work Schedules by Family Type
• Economic Vulnerabilities and Some Policy Issues

Summary

Family and work structure most Americans' lives. Work provides the principal means by which most families support themselves, and public policies directed at low-income families with children have generally attempted to encourage and support work. Family structure also has been a focus of public policy because an increasing number of children live with a single parent, and poverty rates for such children are much higher than for those in married-couple families. Families with children, regardless of marital status, are at greater risk of poverty, with child poverty rates higher than those for either nonaged or aged adults.

Based on nuclear family income in 2004, 17.1% of families with children had total incomes below the poverty line; 21.5% had incomes of less than 125% of the poverty line (poor or near-poor); and 34% had incomes of less than 200% of poverty (low-income). Most low-income families had a parent who worked during the year.

Many parents work at jobs that produce low incomes for their families. In 2004, 13% of parents worked at a wage rate that would have produced below-poverty income for a family of three, and 45% of parents worked at a wage rate that would have produced incomes below 200% of the poverty level, even assuming full-year, full-time work—a level designated as low-income in this report. Thus, families with children headed by a single parent—with only one potential earner—are more likely to be poor than those headed by a married couple. However, married couples with children and one working parent sometimes struggle to attain even a modest income above 200% of poverty; it often takes both parents working for a family to exceed low-income levels. Moreover, for many families, if one parent's earnings are lost, the economic status of many married-couple families would suffer.

Parents' characteristics (e.g., education and work experience) and the occupations in which they are employed affect wage rates, and therefore the possibility that a family is poor or low-income. Those with lower levels of educational attainment are more likely to earn low wages and be in low-income families. We estimate that 17% of working parents with just a high school diploma earn wages that are too low to support a family of three at the poverty line—assuming full-time, full-year work—and 60% would be unable to support such a family at twice that level. Among working parents with a bachelor's degree, only about 6% earn wages too low to support a family of three at the poverty line, but 24% lack the earnings capacity to support such a family at twice the poverty line. Additionally, younger parents typically earn less than older parents, who have had more time to accumulate work experience. This puts the families of young parents, who tend to have young children, particularly at risk for poverty and low income.

Annual earnings and economic status also depend on number of hours worked per week and the number of weeks worked during the year. In married-couple families, strong attachment to the workforce of both parents is associated with higher income levels. However, among poor married-couple families, only half of all poor families had one earner working full-time all year. This report will not be updated.

---

Parents' Work and Family Economic Well-Being

Introduction

Family and work structure most Americans' lives. Work provides the principal means by which most families support themselves, and public policies directed at low-income families with children have generally attempted to encourage and support work. Family structure also has become a focus of public policy since an increasing number of children live with a single parent, and poverty rates for such children are much higher than for those in married-couple families. Families with children, regardless of marital status, are at risk of poverty; child poverty rates are higher than those for the aged or nonaged adults.¹ Moreover, research has shown that poverty has negative consequences on a child's development, which could affect the child's life chances as an adult.²

About This Report

This report examines the cash income and poverty status of families with children. The report specifically looks at the role parents play in supporting their families through work, with a focus on the adequacy of parental earnings. It examines parents' effectiveness as breadwinners, and seeks to address such questions as:

• How well are American families with children doing? How many families are poor, and how many are poor in spite of parents' work? If not poor, how many are near poor, or just getting by?
• Is the full-time work of one parent sufficient to support a family, or must other family members, most notably a spouse, also work to attain a modest level of income security relative to basic needs?
• Which categories of families are particularly struggling? Could they increase their income security by working more? How many families, while getting by, would be economically vulnerable if they lost a breadwinner's income or had an additional child?

The report presents data on income and poverty status of families with children; looks at parents' wages relative to poverty and factors affecting wages, including parents' education level, age, and occupation; and examines the amount of work provided by parents and the relationship of job attachment to earnings adequacy. The report concludes by highlighting those families most vulnerable to poverty or low income based on parents' earnings, and briefly discusses public policies and government programs most relevant to those families.

Our analysis is based primarily on data from the U.S. Census Bureau's Annual Social and Economic Supplement to the Current Population Survey (CPS/ASEC) for 2005. The survey collects information on family composition at the time of the survey (February through April of 2005) and income and labor force participation in the previous year (i.e., 2004). While shedding much light on the economic status of families with children, the CPS cannot address certain questions, particularly those relating to compensation other than earnings, such as sick and/or family leave, or other benefits, such as health and retirement benefits.

The major findings of the analysis in this report are:

• Based on nuclear family income, an estimated 6.8 million families with children were poor in 2004, accounting for 17.1% of families headed by a parent with one or more children. At an income standard of twice the poverty line, deemed to be "low-income" in this report, 14.4 million families with children qualified as low-income, accounting for 36.4% of all families with children.
• Many parents work at jobs with wage rates that translate into low incomes for their families. In 2004, an estimated 13% of all parents worked at a wage rate that would have produced below-poverty income (for a family of three) even if they worked full-time, all year; an estimated 45% of all parents worked at a wage rate that would have produced incomes of less than 200% of poverty.
• In married-couple families with children, it often takes the work of both parents to earn a modest income. In 2004, almost 40% of all married-couple families with children would have had incomes equal to less than 200% of the poverty threshold, based on the primary worker's earnings.
• Many single-parent families, having only one earner, have earnings that leave the family either poor or near poor. In 2004, more than half of mother-only families would have been poor based on the mother's earnings alone; more than three-quarters of these mothers had earnings that were less than 200% of the poverty threshold.
• Our economy has a division of labor and specialization. Some occupations pay relatively low wages, and a substantial number of parents are employed in these occupations. For example, in March 2005, 1.2 million parents worked as "cashiers," an occupation with a median wage of $8.00 an hour. This wage rate would produce an income of less than 125% of poverty for a family of three.
• Parents' earnings depend on many factors, the most prominent being their level of educational attainment. Those with college degrees earn more than those without them.
• Another factor affecting parents' earnings is their age—older workers who have had more time to acquire on-the-job experience and job skills are paid higher wages than less experienced, younger workers. This has important implications for parents in their 20s and early 30s (a period when many families are formed and people commonly have children), who earn less than older parents.

How We Define Families

In this report, we define families somewhat differently than the Census Bureau in its family income and poverty measures. Under the official Census Bureau measures, all persons related by blood, marriage, or adoption living in the same household (i.e., living quarters) comprise a family. As such, a married couple or a single parent with dependent (biological or adopted) children in their household constitutes a family, but so do more complex households containing other family members, such as grandparents, aunts and uncles, nephews, nieces, cousins, and grandchildren.

We employ a more restrictive definition of family than that used by the Census Bureau because we wish to gauge parents' effectiveness in providing for their immediate family's needs—that is, the needs of the parent or parents (if a spouse is present) and their dependent children. For this purpose, we define families as units composed of married couples or single parents with at least one child under the age of 18 (i.e., minor children). All minor children and all adult children living in the household are included in the family unit unless those children have a child of their own, or are married with a spouse residing in the household. In such cases, where a child or another relative who has a child lives in the household, those subfamily units are included in the analysis but are identified as separate family units. The family units constructed through this process are identified as "nuclear families," which are at times identical to the Census Bureau's definition of family, but in more complex family households are a subset of the Census Bureau's definition. (See Appendix A for a discussion of family living arrangements and treatment of income in defining poverty.)

How We Treat Income

As noted above, we define nuclear families as the basic unit of analysis in this report. Much of our analysis compares family income, or components of family income (e.g., parents' earned income), relative to Census Bureau poverty income thresholds. We use Census Bureau poverty thresholds as a measure of families' basic needs. These poverty thresholds vary by family size and composition (e.g., number of related children under the age of 18) and are the most often used standard for measuring basic income needs.³

In our analysis, we examine nuclear family income and poverty status based on the unit's income compared to poverty thresholds that reflect the unit's size and composition. This is because, as stated above, we are trying to gauge parents' effectiveness in meeting their immediate family's income needs. Unless otherwise stated, in cases where there are extended family members and/or related subfamily units present in the household, their income and needs are excluded from the nuclear family income and poverty definitions used in this report. However, when reference is made to the "official Census Bureau income and poverty measure," the income and needs of extended family members are included.

In this report, we construct income to poverty ratios by dividing the family's earnings or income by the applicable poverty threshold, based on the family's size and composition. A family with income below 1.00 times the poverty line is considered poor. We also assess parents' earnings and family income at other income standards. For example, we identify families with income below 1.25 times the poverty line as "near poor," below 1.5 times poverty as "economically fragile," and below 2 times poverty as "low-income." These labels, "near poor, economically fragile, and low-income," are used to facilitate the narrative and presentation and should not be construed as official government definitions or designations of relative income adequacy. The body of the report focuses primarily on married-couple families and families headed by lone mothers, with only limited discussion of lone-father families, in large part due to their comparatively small numbers. However, data for all three family types are presented in a data appendix (Appendix B).

Family Income and Poverty Status

Under the definition of nuclear family poverty used throughout much of this report, in 2004, an estimated 6.8 million families with children were poor, accounting for 17.1% of nuclear families, or family units composed of just parents and their dependent children. In all, we estimate there were approximately 39.7 million nuclear family units with children in 2004. Most of these families (31.6 million, or 80%) live in independent households, with no other family members or other unrelated household members. Under the official Census Bureau poverty definition, taking other family members into account, both in terms of their income and needs, reduces the number of poor families with children to an estimated 5.9 million, and their poverty rate to 14.9%.

Figure 1 shows families with children by their incomes, relative to the poverty line. The lines represent the cumulative percentage of families with children with incomes below a specified level, by family type. The labeled brackets in the figure indicate the range at which families are considered poor (below 100% of poverty), poor or near poor (below 125% of poverty), economically fragile (below 150% of poverty), or low-income (below 200% of poverty). The line for all families shows that 17.1% of families with children have nuclear family incomes below 100% of poverty. About one-fifth of all families with children (21.5%) were either poor or near poor (with incomes below 125% of poverty); and well over one-third (36.4%) were low-income (i.e., had incomes below twice the poverty line).

Most children live in families headed by either a married couple or a single mother. It is well known that the likelihood that a child will be impoverished depends greatly on which of these two family types he or she lives in. In 2004, just over two-fifths (41.8%) of nuclear families with children headed by a single mother were poor, while only 7.1% of families with children headed by a married couple were poor. Over two-thirds of mother-only families (68.8%) had incomes below twice the poverty line, while just over one-fifth of married-couple families with children (22.5%) fell within this low-income category.

Figure 1. Nuclear Family Income Relative to Poverty Among Families with Children, by Family Type, 2004

Though the likelihood that a family with children will be poor is much greater for those headed by a single mother than those headed by a married couple, married couples with children account for a nontrivial share of poor and low-income families with children. Figure 2 shows that, in 2004, more than one-in-four poor families with children (27.8%) were headed by a married couple, while married couples accounted for more than four of every ten (41.8%) low-income families with children. Being headed by a married couple is not sufficient for a family with children to escape poverty or low income status. In 2004, about 6 million (22.5%) married-couple families with children were low-income, having incomes below twice the poverty line.⁴

Figure 2. Composition of Nuclear Families with Children, by Family Type and Income Relative to Poverty, 2004

Parents' Work and Family Income Status

Earnings are the primary means by which most parents support their families. In 2004, 92% of families with children had at least one parent with earnings. In married-couple families, the earnings of a single breadwinner may provide income deemed by the couple to be sufficient. In other cases, both parents may choose to work to improve the family's economic standard of living. For families headed by single parents, choices between work and family obligations may be more difficult.

In what follows, we compare family income and components of family income to poverty in several ways. At times, we count only the primary earner's earnings against the family's poverty threshold to gauge the effect of one parent's labor market engagement in providing for his/her family's needs. To this measure, we separately add earnings of a second parent, if present, to judge the effect of combining the labor income of both parents in supporting their family. In addition to parents' earnings, in a separate step we add all other sources of cash income in the immediate (i.e., nuclear) family to provide a fuller assessment of their level of economic well-being. In a subsequent final step, the income and needs of extended family members living in the household are included in the assessment of family income and poverty status.

Considering earnings alone, of course, ignores the impact of other forms of income—particularly government transfer payments such as food stamps and the refundable Earned Income Tax Credit (EITC)—on parents' behavior and family economic well-being. Since the focus of this report is on low-income families with children, the impact of these other forms of income on economic well-being is obviously important. Considering the effect of earnings alone on family poverty status addresses the question: "What would the economic well-being of families with children be if they relied only on their income from work and received no other income, including no government help?"⁵

Figure 3 shows the role of parents' earnings and other income relative to the poverty threshold by family type. In 2004, more than half (52.6%) of families with children headed by a single mother would have been poor based on her earnings alone. Other income sources, such as alimony and child support, social insurance payments (e.g., social security), and government cash assistance, reduces the poverty rate of these families to 41.8%. Approximately one-in-four families headed by lone mothers live in household settings with extended family members (see Appendix A). If the income and needs of these additional family members is factored in, the poverty rate of lone-mother families drops to 35.0%—their official poverty rate.

Figure 3. Earnings and Income Relative to Poverty for Married-Couple and Mother-Only Families with Children, 2004

The inability of single mothers to provide their families with a basic—let alone modest—level of income security may be due to low wages, few hours worked, or a combination of the two. When only full-time, full-year working single mothers are considered, 23% are unable to provide a standard of living above poverty based on their earnings alone; 33% are unable to secure an earnings-based standard of living that is higher than near-poverty; and 62% are unable to attain a modest earnings-based level of income security above twice the poverty line (not shown in figure). It is unlikely that these mothers can work more, as they already are fully engaged in market-based work. As a single mother, there may be little time for in-home production (i.e., preparing meals, doing laundry, cleaning), if other family members are not available to chip in, and income may be limited to hire out these activities. Additionally, children must be cared for and nurtured. Low-cost or free child care (e.g. provided by a relative) may be essential for these mothers to work to support their families.

Among married-couple families with children, the poverty rate based on the earnings of the primary earner⁶ alone is estimated at 12.9%. In the majority of these families (77%), the husband is the primary earner (not shown in figure). When the earnings of both parents are considered, the poverty rate for married couples with children falls to 10.0%. All other nuclear and extended family income (if present) reduces the poverty rate of married couples with children to 6.9%—the official level under the Census Bureau definition.

Among married-couple families with children, earnings of a second working parent are particularly important in attaining a higher family income standard. Although earnings of a second parent have only about a three percentage point effect on the poverty rate, the importance of these earnings is seen higher in the income distribution. Based on the earnings of a married couple's primary earner alone, nearly 40% of married couple families with children would be classified as low-income. However, when the earnings of the second parent are added to family income, that percentage drops to 27%. Thus, it often takes two working parents for a married-couple family to attain a modest income.

In 2004, among primary earners in married-couple families with children who worked full-time full-year, 9% were unable to provide their families with a standard of living above poverty based on their earnings alone; 15% were unable to secure an earnings-based standard of living beyond near poverty; and 36% were unable to attain a modest level of income security above twice the poverty line (not shown in figure). In many cases, both parents work, presumably at least in part to attain a higher standard of living. Married-couple families face a trade-off relating to whether a parent should stay home to care for the children, or go to work to secure additional income. The decision may be weighed by such factors as the secondary worker's earnings potential and career ambitions, suitability and cost of child care arrangements, and other work-related costs.

Parents' Wages, Job Attachment, and Family Earnings Adequacy

A worker's ability to support a family depends upon hourly wages earned, number of hours worked, and family size. Under the official U.S. poverty definition, the poverty income cutoffs, or thresholds by which families are considered to be poor, vary by family size and composition. Full-time, full-year work may not be sufficient, given the parent's wage rate, for a parent to support a family of a given size above poverty, or to secure a modest standard of living at an income level of twice the poverty line.

Wage Rates Relative to Family Poverty Income Levels

Table 1 shows annual poverty income thresholds and hourly poverty wage rates for families of varying size and type, in 2004. The table shows, for example, that the poverty income threshold is higher for a four-person family ($19,157, in 2004), consisting of two adults and two children, than for a three-person family ($15,205, in 2004), consisting of two adults and one child.⁷ Hourly poverty wage rates shown in the table are calculated by dividing the corresponding annual poverty threshold by 2,080 hours—the equivalent of 40 hours of work per week for 52 weeks per year. Thus, for the four-person family, a single earner working full-time, full-year (i.e., 52 weeks, 40 hours per week, or 2,080 hours) would need to earn $9.21 per hour to support his/her family at a poverty-line income, based on earnings alone. For the three-person family depicted above, the parent would need to earn $7.31 per hour, working full-time, full-year, to support his/her family at a poverty-line income. A single parent with one child would need to earn about $6.26 per hour to bring her/his family income up to the poverty line. From the table, it is apparent that full-time, full-year work of one parent, at the current federal minimum wage of $5.15 per hour, would be insufficient to bring a family's income up to the poverty line under any family configuration.

As shown in Table 1, for a married-couple family with one child, it takes one parent earning a wage rate above $14.62 per hour to support the family at twice the poverty line. In many families, the principal wage earner works for less than $14.62 per hour, and in some families, the principal wage earner works less than full-time for all 52 weeks in the year. In many families, both parents must work in order to attain a modest level of income security, above twice the poverty line. Both hourly wages and the amount of work family members provide contribute to families' income living standards.

Factors Affecting the Determination of Wages

Employers do not set wages based on the size of their workers' families. Rather, workers are theoretically compensated according to their presumed productivity—output per worker. Thus, the wage rates determined by the forces of the labor market will not necessarily bear any relationship to the poverty threshold or presumed needs of families with children.

Workers' wages are related to the market valuation of their productivity. A worker's productivity is related to the knowledge, skills and abilities brought to the labor market. As shorthand, we will refer to these worker traits simply as skills or human capital. Employers' need for specific skills (demand), along with the relative scarcity of those skills in the labor force (supply), help determine the market value, or wage, that workers command and employers offer. Some workers' skills are valued highly, as reflected in their wages. Other workers' skills are valued such that the market offers comparatively low wages.

Two commonly measured aspects of workers' human capital are educational attainment and lifetime work experience. In this section, we examine the effects of educational attainment on parents' wages and family poverty status, based on the highest education credential that parents have attained. However, lifetime work experience is not measured on the Current Population Survey (CPS). In this analysis, we use parents' age as a proxy for potential on-the-job experience.

Wage rates vary by factors other than human capital. For example, wage rates of African-Americans and Hispanics tend to be lower than those for whites. Women often earn less than men. Numerous studies have examined wage rates by race, ethnicity, and gender, attempting to determine the amount by which wage differentials can be attributed to differences in measurable background factors (e.g. education and work experience) and the degree to which remaining unexplained differences might be due to other, less directly measurable factors, such as discrimination.⁸

Parents' Wages and Educational Attainment

As is commonly known, wage rates tend to be higher for workers with higher levels of educational attainment. Figure 4 shows the distribution of hourly earnings (i.e., hourly wage equivalents⁹) for working parents by the amount of their educational attainment. It shows the hourly earnings by decile (each 10% of the wage rate distribution) for each level of educational attainment. The upward slope of the lines shows that wage rates increase with levels of educational attainment. Further, the widening "spread" of the lines as educational attainment increases shows that the returns to education have greater variability at higher levels of attainment than at lower levels.

Figure 4. Hourly Earnings of Working Parents, by Level of Educational Attainment, 2004

In 2004, the median wage rate equivalent for a parent who had less than a high school diploma was $9.62 per hour—a full-time full-year earnings level that equates to about 132% of the poverty line for a married couple with one child.¹⁰ Based on those earnings alone, such a family would be classified as "economically fragile." With one additional child, the family would be barely above poverty.¹¹ The median hourly wage equivalent of workers with a high school diploma ($12.82) and those with some college but no degree ($14.62) affords a married couple with one earner and one child a standard of living equal to 175% of poverty in the former case, and 200% of poverty in the latter case. It should be noted that these wage rates are the median observed hourly wage rates for working parents, meaning that half of parents at a given education level earn more per hour, and half earn less. The other earnings percentiles show the distribution of hourly earnings at each level of educational attainment.

The figure shows that higher levels of educational attainment do not necessarily translate into higher wages. For example, some workers with less than a high school diploma earn more than workers with a college degree. While 20% of parents with less than a high school diploma earn $15.38 per hour or more, 20% of those with a bachelor's degree earn $13.33 per hour or less. Those wage levels mean that the top 20% of workers with less than a high school diploma can support a spouse and one child at a modest income level, equal to 210% of the poverty level, while the bottom 20% of workers with a bachelor's degree would support such a family at only 182% of poverty, a level considered here to be low-income.

In Figure 5 we show how parents' level of educational attainment and associated wage rates vary in their ability to support a benchmark family of three. In the figure, working parents' wages are translated into full-time full-year earnings equivalencies¹² as a measure of parents' potential annual earnings. As such, we are considering only the effect of wages, and not how much or little a parent works, on the potential economic well-being of his/her family. Parents' potential earnings are compared to a benchmark family's basic needs: the poverty threshold for a three-person family (i.e., $15,205 for a married couple with one child in 2004). The estimates are for all working parents (i.e., primary and secondary wage earners in married-couple families as well as working lone parents). Parents who did not work during the year are excluded from the analysis, as they have no observed wage rate. As noted above, the estimates assume full-time full-year work. For parents working less, or for parents who have larger families to support than the 3-person standard shown here, standards of living relative to poverty would be lower than those depicted; for parents working beyond a full-time full-year equivalent, or having a smaller family (i.e., a lone parent with only one child), standards of living relative to poverty would be higher than those depicted.

Not surprisingly, as workers' educational attainment increases, so do their wage rates and their ability to support a family. Figure 5 shows that while 3 out of 10 working parents with less than a high school diploma would have full-time full-year earnings too low to support a three-person family above the poverty level, only about 1-in-6 parents with a high school diploma, and about 1-in-20 parents with a bachelor's degree, have wages too low to support a benchmark family above poverty. Among working parents with a bachelor's degree, nearly 1-in-4 have wage rates insufficient to provide a modest standard of living, above 200% of poverty, for a benchmark family.

Wage Rates by Parents' Age

Younger workers' wages tend to be lower than those of older workers, all else being equal (e.g., educational attainment). Younger workers have had less potential time in the workplace than older workers, and thus have not yet had the job experience and the increase in wages that accompanies that experience.

Figure 5. Parents' Earnings Capacity by Level of Education, 2004 (Capacity to Support a Family of Three Based on Full-Time Full-Year Work at Parents' Wage Rate)

Figures 6 and 7 show the hourly earnings distribution of parents by their age. The first of the two figures shows hourly earnings in married-couple families with children for the parent with the highest earnings, by that parent's age; the second figure shows hourly earnings for mothers heading families with no spouse present, by the mother's age.

The figures show that older parents tend to command higher earnings than younger parents. The figure also shows that there is considerable wage dispersion among parents at any given age, and the dispersion widens with advancing age. For example, Figure 6 shows that in 2004, the median hourly earnings of the highest earning parent in very young married-couple families (i.e., with parents under age 25) was $11.06. In other words, half of these parents had hourly earnings that were higher than $11.06 and half had hourly earnings below $11.06. One-tenth of parents under age 25 had hourly earnings below $6.11 and one-tenth had hourly earnings above $18.68. On the other hand, among parents age 45 to 49 who were the primary earner in married-couple families, half had hourly earnings of $24.04 or more (median amount), and 10% earned $10.68 per hour or less. However, slightly more than half of parents under the age of 25 earned more per hour than the bottom 10% of parents age 45 to 49.

Figure 6. Hourly Earnings1 Distribution, by Parent's Age,2 Married Couples with Children, 2004

Figure 7, for lone mothers, shows similarly shaped earnings patterns as those for the primary earner in married-couple families (Figure 6). Hourly earnings tend to be lower for younger than for older lone mothers. However, comparing hourly earnings levels of lone mothers to the primary earner in married-couple families, lone mothers' earnings tend to be much lower. For example, the median hourly earnings of lone mothers under age 25 was $7.69, compared to $11.06 for the primary earner of similar age in married-couple families. For parents age 45 to 49, median hourly earnings of lone mothers ($13.94) were only about 58% of those of primary earners of similar age in married-couple families ($24.04). Single mothers' earnings capacities are much lower than those of the primary breadwinner in married-couple families. As noted in an earlier section, among married-couple families with children, the husband is most often the primary earner—about three-quarters of the time.

Figure 7. Hourly Earnings1 Distribution by Mother's Age, Mother-Only Families, 2004

Figure 8 is similar to Figure 5, which translated working parents' observed wages into full-time full-year earnings equivalencies and compared those potential earnings to the poverty line for a family of three (i.e., married couple with one child). Where Figure 5 compared earnings capacities by parents' level of educational attainment, here the comparison is by parent's age. As with the earlier figure, Figure 8 includes all parents (i.e., primary and secondary earners in married couple families, and lone parents) with observed wages. The figure shows, for example, that 35% of parents under the age of 25 have wage rates that are too low to support a family of three above poverty, assuming full-time full-year work. Among parents age 30 to 34, about 13% have wages that would fail to support a family of three above poverty, and only about half who are working have wage rates that would provide a family of three with an income standard above twice the poverty line.

Figure 8. Earnings Capacity by Parents' Age, 2004 (Capacity to Support a Family of Three Based on Full-Time Full-Year Work at Parents' Wage Rate)

Parents' Age and Poverty Status by Family Type

Families headed by young parents are particularly economically vulnerable. As shown above, wage rates of younger parents tend to be lower than those of older parents, as younger parents have had less time in the labor market to capture the wage premiums that typically come with job experience. Younger parents also typically have younger children, whose needs for care may limit the ability or willingness of a parent to work outside the home. A child's birth, especially among young parents, may interrupt parents' education and delay possible accumulation of work experience, thereby affecting future wage possibilities. Younger parents may lack the job security that comes with time on the job, exposing them to greater risk of being laid off during economic downturns.

Figures 9 and 10 depict family income relative to poverty for married-couple and mother-only families, respectively, by parents' age. For married-couple families, the age of the older parent is shown. Figure 9 shows, for example, that more than one-fifth (21.0%) of married-couples with children in which the parent was under the age of 25 were poor in 2004, based on their immediate family income. Nearly two-fifths (39.6%) had immediate family income below 150% of the poverty line. Among married couples in which the parents were slightly older (age 25 to 29), about one-in-eight (12.0%) were poor, and one-in-four (25.0%) had incomes below 150% of poverty.

Figure 9. Nuclear Family Income Relative to Poverty by Parent's Age, Married-Couple Families with Children, 2004

A similar pattern relating to parents' age and poverty status is evident among single-mother families, although notably, at every age poverty rates are much higher among mother-only families than married-couple families (Figure 10). Among mother-only families, for those with a mother under age 25, more than seven-in-ten (71.3%) were poor, and for those with a mother age 25 to 29, half were poor (50.1%).

Figure 10. Nuclear Family Income Relative to Poverty by Mother's Age, Mother-Only Families with Children, 2004

Parents' Wages by Occupation

Not only do the characteristics of individuals (human capital) affect their wage rates, but so too do the characteristics of their jobs (e.g., occupation, industry, firm size, and working conditions). Adam Smith in Wealth of Nations wrote much about the division of labor and how differing wage rates among jobs reflect both institutional and market conditions. To attract workers to a job, employers set wage rates such that they reflect both the job's educational/skill requirements and its other attributes.¹³

Some occupations typically pay wages that are insufficient to support a benchmark family of three beyond a low-income level, even if the parent works 40 hours per week all year. Table 2 shows some of the occupations that pay parents relatively low wages. Shown are occupations that employed more than 500,000 parents in March 2005 and paid less than $14.62 per hour in that month. (The $14.62 amount is the low-income cut-off shown in Table 1—200% of the 2004 poverty threshold for the benchmark family defined earlier in this report; i.e., a married couple with one child). It should be noted that only parents and their wages are shown in the table—other workers in the same occupations are excluded. The table is sorted by median wage rates of parents in an occupation in that month.¹⁴

In March 2005, the occupation that employed at least 500,000 parents and paid them the lowest median hourly wage was waiters and waitresses, at $7.75 per hour. For parents employed 40 hours per week all 52 weeks of the year, this wage rate would produce an annual income above 100% but below 125% of the poverty threshold for our benchmark family of three. Four other occupations—cashiers, cooks, maids, and child care workers—also paid parents a wage rate that would produce an annual income just above the poverty threshold. Three occupational groups—teacher assistants (employing more than 800,000 parents), nursing, psychiatric, and home health aides (employing 1.4 million parents), and janitors and building cleaners (employing 1.1 million parents)—paid a median hourly wage that would produce annual incomes of between 125% and 150% of poverty for full-year full-time workers. Thirteen occupations paid median hourly wages that would produce annual incomes of between 150% and 200% of poverty.

The median wage information in the table means that, in March 2005, half of the parents in the occupation earned less than the wage value, and half earned more. Thus, though the table shows no occupation employing more than 500,000 parents with an hourly wage that would produce an annual income below the poverty level for the benchmark family of three (at full-year, full-time work), some parents in the occupations shown on this table did earn less than poverty-level wages. This is shown in Figure 11, which illustrates the distribution of wages within these occupations by the proportion of parents in the occupation with hourly wages below 100% of poverty and below 200% of poverty. For example, the median hourly wage for parents who were cooks was $8.20 in March 2005; this wage would produce an annual income above the poverty line for the benchmark family with full-year, full-time work. However, the hourly wages of 37% of all parents who worked as cooks would have produced an annual income below the poverty line, even if the parent worked 40 hours per week all year. More than 9 out of 10 parents who were cooks earned an hourly wage that would have produced an income of less than 200% of poverty for a benchmark family of three, even if the parent worked full-time all year.

Figure 11. Percent of Parents with Hourly Earnings Below 100% and 200% of Poverty (Based on the Poverty Threshold for a Married Couple with One Child), by Selected Occupations, March 2005

Parents' Work During the Year

The previous section focused on hourly wage rates. The second part of the equation that determines annual earnings is the amount of work done during the year—number of hours and weeks actually worked. The hourly poverty wage rates shown earlier in Table 1 reflect the hourly wage needed to achieve a level of annual earnings assuming work of 40 hours per week for all 52 weeks of the year. However, many parents work less than full-time, year-round. Some weeks are spent unemployed, because the parent was either laid-off from a job or is out of work but actively seeking employment. Other weeks may be spent out of the labor force, where the parent is not seeking work. Also, but less frequently, some parents work less than full-time work schedules.

The amount of work that a parent wants to provide is based on tradeoffs between the income derived from work and the value of leisure and home production (e.g. housework, child care) that would have to be foregone in order to work. With children involved, the cost-benefit analysis that goes into the decision to participate in the labor force—and for how many hours in a week, per year—includes costs associated with caring for children. Moreover, wanting to provide work and actually working are not always the same thing. For example, many, if not most, employers offer only fixed work schedules with fixed hours, which may not accommodate some parents' needs or preferences.

Parents' Annual Hours Worked

Figure 12 summarizes annual hours of work for both the mother and father in married-couple families with children in 2004. It shows annual hours by several levels of job attachment—from no work during the year, to being fully engaged in work by working the equivalent of 40 hours per week for all 52 weeks in the year (2,080 hours in the year) or more, and two categories in between: attachment up to three-quarters-time (less than 1,560 hours per year), and near full-year full-time attachment (1,560 to 2,079 hours during the year). The left column of the figure shows married fathers' work attachment, with 80.5% shown as fully engaged in the workforce during the year (2,080 hours or more). A summary of married mothers' work attachment is shown in the second to last row on the chart. It shows that 35% of married mothers were fully engaged in the workforce during the year. It also indicates that about 7-in-10 mothers in married-couple families worked at some time during the year, while 29.2% had no job attachment during the year.

The figure also shows the percent of families with children headed by a married couple by the combination of mothers' and fathers' hours of work during the year. It shows that the most common combination is when both the mother and father worked the equivalent of full-time full-year work during the year. In 2004, 28.2% of all families with children headed by a married couple had both parents fully engaged in the workforce in the year. However, also common was full engagement by married fathers with no work by the mother (23.4% of all married couple families with children).

The last two rows of the figure compare the job attachment of married and single mothers. Single mothers are more likely than married mothers to work, and to work full-time, year-round. Moreover, a comparison of married fathers' job attachment (left column) and married and single mothers' attachment (bottom two rows) shows that despite the historical increase in labor force participation of women, job attachment patterns of married fathers and (both married and single) mothers remain quite different.

Figure 12. Parents' Job Attachment: Married Couples and Lone Mothers, 2004

Weeks Worked and Work Schedules by Family Type

Most families with children (92%) had one parent who worked at some point in 2004. However, the economic status of families depends not so much on whether a parent works, but rather on how much work one or both parents engage in over the year. In the previous section, we examined the total number of hours parents worked during the year. Here, we examine how parents acquire those hours of work—are they working all or just some weeks during the year; when they work, are they working full-time or part-time schedules? To examine this question, we use the U.S. Bureau of Labor Statistics (BLS) definition of full-year employment as working 50 weeks or more during the year, and the definition of full-time work as 35 hours or more during the week. In this section, we find that less than full participation in the workforce—working fewer than 50 weeks, and to a lesser extent, working a part-time schedule—was associated with greater likelihood of poverty and low annual earnings.

Married-Couple Families

Figure 13 portrays the work status of the principal earner in married-couple families in 2004. In the vast majority of cases where at least one parent in a married couple worked at some time during the year (98%), the work status shown is for the parent who earned the most during the year.

Among married-couple families in which the earnings of the highest paid parent were insufficient to lift the family above poverty, almost one-half had a parent working full-time, year-round. Higher up in the income distribution, full-time full-year work was even more common. Among those families who were near-poor (incomes above poverty but below 125% of the poverty line) based on the parents' earnings, 83% had a parent who worked full-time all year, and among families with earnings between 150% and 200% of poverty, the percentage with a full-time full-year worker rose to 90%.

Figure 13. Principal Earner's Work Status in Married-Couple Families with Children, By Earnings Relative to Poverty: 2004

Whether one or both parents are fully engaged in the workforce affects the economic status of families with children headed by a married couple. Figure 14 shows the percent of such families that have one or two full-year, full-time workers within each income group. Here, poverty status is based only on parents' earnings relative to families' needs. As shown on the chart, nearly half (49.6%) of earnings-poor married-couple families with children have no parent fully engaged in the workforce during the year. However, to attain a modest income (i.e., above 200% of poverty), it often takes both parents to be fully engaged in the workforce. Among the more affluent families (earnings equivalent to 300% of poverty and above), about 45% had both parents working full-year at full-time schedules.

Figure 14. Nuclear Family Poverty Status Based on Parents' Earnings Alone, by Parents' Job Attachment: Married-Couple Families, 2004

Mother-Only Families

Figure 15 shows the work status for single mothers in 2004. The picture differs from that shown for married couples. Despite large increases in the labor force participation of single mothers from the mid-1990s through 2001 (with some declines since), single mothers still had a relatively high incidence of no work in 2004. Among single mothers who were earnings-poor based on their own earnings, more than four-in-ten (42.5%) did not work in 2004 and thereby had no earnings.

Full-time, full-year work by a single mother is often necessary for her family to escape poverty, but does not guarantee attainment of a modest standard of living, defined as more than twice the poverty line, or even a standard of living beyond near poverty. In 2004, among lone-mother families who were "near poor" (above poverty but below 125% of poverty) based on the mothers' earnings alone, two-thirds (66.9%) had a single mother who worked full-time, year-round. Among single-mother families with earnings between 150% and 200% of poverty, in more than four out of five of these families (83.5%), the mother worked full-time year-round.

Nearly half of all lone mothers (48%) worked full-time, full-year in 2004.¹⁵ Based on their earnings alone, nearly one-in-five (18.8%) were earnings poor, in spite of the mothers' full-time attachment to a job (not shown in figure). Nearly three-in-ten had earnings below near poverty (under 125% of poverty), and the share with earnings below 150% of poverty approached four-out-of-ten families (37.7%) headed by lone mothers who worked full-time, full-year.

Figure 15. Work Status of Lone Mothers

Economic Vulnerabilities and Some Policy Issues

In the previous sections, we examined the relation of wages and job attachment to family earnings adequacy, as indicated by family income poverty thresholds. We showed how the ability to support a family is closely related to parents' education and age. Some parent's earnings capacities based on their observed wage rates are too low to support a couple and a single child above poverty, let alone to attain a level of modest income security above twice the poverty line, even if the parent is able to work full-time, full-year. Families headed by a single parent are at a distinct economic disadvantage compared to married couples, in part due to the unavailability of potential added earnings that come when both parents work, but also due to the absence of a second caretaker who can help watch over children, thereby allowing the other parent to work.

Adam Smith opened The Wealth of Nations by noting that improvement in labor productivity stemmed from the effects of the division of labor. Smith also noted that different occupations resulting from the division of labor command different wages, with market and institutional forces combining to create these wage differentials. Thus, the division of labor provides the means to an affluent society, but also results in an unequal distribution of the rewards from work.

This report has identified several groups of families with children that are likely to be poor or low-income based on the work of their parents:

• Families headed by a single mother. Given that wage rates of many parents' jobs would leave a family in low-income status, it often takes two parents working for a family's earnings to attain a modest income level, above the low-income threshold of 200% of poverty used in this report. Full-time work for many single parents may be difficult to attain, if safe, reliable and affordable child care is difficult to obtain. Obviously, then, single parent families are at high risk for low-income and poverty.
• Young families. The principal earner of families headed by a parent under the age of 25 had median earnings of $11.06 per hour—a wage rate that would classify the three person benchmark family used in this analysis as low-income, but just slightly above the cut-off that would classify the family as economically fragile (i.e., 150% of poverty), based on full-time full-year work. The low earnings of these young parents is a logical consequence of being at the beginning of their adult work careers and not yet having accumulated job experience that raises earnings capacity. In young married-couple families, it may make little economic sense for both parents to work, as child care for young children may be expensive relative to additional earnings that might be obtained. Additionally, many families prefer that a parent stay home to care for and nurture their children, and a parent's labor force withdrawal may be voluntary and temporary. Low income and economic insecurity may be temporary during the early years of family formation, but have implications for family well-being and child development and outcomes, in years beyond.
• Families Headed by a Parent Without A College Degree. Families headed by a parent who lacked a high school degree were likely to be low-income. Median wage rates—even for high school graduates and those with some college, but no college degree—result in the benchmark three-person family being classified in the low-income category, by having earnings below 200% of poverty. Unlike low wage rates caused solely by a lack of job experience, the low-income status for families headed by a parent without a college degree might not be temporary.

Social and economic behavior that is considered consistent with "personal responsibility"—such as marriage, work, and education—is not sufficient for a family with a child or children to escape low income. Rates of poverty and low income for single-parent families are higher than in married-couple families. However, a non-trivial percentage of poor families with children—28%—were headed by married couples. Wage rates are strongly related to educational status, but in 20% of poor families with children, the principal earner had a college degree.

Moreover, even more affluent families with children may be considered economically vulnerable. Based on the earnings of both parents, 27% of married-couple families with children are low-income; however, based solely on the earnings of the family's principal earner, 39% of such families would be low-income. Thus, the economic well-being of many married-couple families with children is vulnerable to the loss of earnings of the family's second earner. Further, there are reasons that some parents do not or cannot work full-time all year, such as involuntary unemployment, illness, or disability, or the need to take time off for the birth or care of a child or disabled family member.

This analysis has focused on the economic outcomes generated by the labor market. Government policy, of course, also has important implications for the economic well-being and security of families with children, although it will be only briefly mentioned in this report. Government policies that promote full employment while at the same time keeping inflation in check are clearly important to working families. While a strong economy is necessary to promote family economic security, for many families it is not sufficient. Families may still face economic vulnerability, even when the economy is healthy. Table 3 presents a typology of selected existing government programs (focusing on federal or federal-state programs) and how they relate to the economic well-being and vulnerability of families with children. Many of these programs face difficult tradeoffs and competing objectives—such as providing some measure of basic economic security while also promoting personal responsibility. This is especially true of "safety net" programs that provide assistance to mitigate income loss caused by certain events, but result in inherent undesired behavioral effects, such as work disincentives and marriage penalties. Other policies and programs seek to mitigate the undesired outcomes often associated with safety-net programs, by increasing the financial returns associated with work or providing opportunities to reduce the intergenerational transmission of poverty.

• Policies to Mitigate the Risk of Earnings Loss. These programs represent the basic two-tiered "safety net" of social insurance programs and need-tested programs. The social insurance programs base benefits on past earnings, and entitlement to these benefits is earned through work in covered employment. The caseload and expenditures of social insurance programs are far greater than those of the need-tested programs. For families with children, the basic contours of the safety net were established in the Social Security Act of 1935, with the last major additions coming with establishment of Medicaid (1965), Medicare for the disabled, including children (1972), and food stamps, made available nationwide in the 1970s. Most recently, this safety net was affected by the welfare reform law of 1996, which converted Aid to Families with Dependent Children, originally established in the 1965 Social Security Act, into a block grant to states that provides time-limited and work-conditioned cash assistance to poor families with children.
• Policies to Raise the Financial Returns to Work. These are policies to directly raise wages (e.g., minimum wage legislation), provide earnings supplements, or reduce the costs of going to work. Minimum wage legislation—like the Social Security Act—dates back to the Great Depression. However, the 1990s saw a great deal of expansion of policies in this category, including expansions of the Earned Income Tax Credit (EITC), which provides a "work bonus" to families with low to moderate earnings, and increased funding for child care, which helps offset the costs of going to work.
• Policies Aimed at Family Formation. There has been much recent attention paid to this class of policies. Tax policies, such as the deduction for dependents and the (partially refundable) child credit, benefit families based on their number of children. These policies to some extent adjust families' tax burdens based on family size. They may also mitigate, to an extent, the fact that employers consider a worker's assumed productivity when striking wage bargains, rather than the number of dependents a worker must support. The role of cash welfare—cushioning economic hardship when one parent is absent, but thought by some to facilitate single-parenthood—has been controversial. However, collections of child support for both welfare and non-welfare families, through the federal-state child support enforcement program, have increased dramatically since the mid-1990s. Recent attention has been paid to so-called "marriage penalties" in the tax code, and very recently marriage promotion and education programs have been introduced. Meanwhile, family planning grants, authorized in 1970, may have the effect of deferring child-bearing, while adolescent family life (created in 1980) and abstinence education (enacted in 1996) specifically focus on preventing pregnancy among teenagers.
• Policies Aimed at Preventing Intergenerational Transmission of Poverty. As briefly mentioned, poverty among families with children has special consequences, particularly since growing up poor might affect a child's life chances as an adult. Programs such as Head Start, and Title I grants to local school districts for supplementary education and services for disadvantaged children, are intended to help poor children obtain the education needed to escape poverty as an adult.

Appendix A. Family Living Arrangements—Effects on Income and Poverty Measurement

This report has principally focused on parents and their children. For this purpose, we focused on parents and their nuclear families as the primary units of analysis. This approach diverges from that more commonly used, in which all related members of a household are considered a family; in this report, we define such families as extended families. Additionally, individuals unrelated to nuclear or extended family members may also reside within a household.

Figure A-1 depicts families with children by their family/household living arrangements. The figure shows, for example, that 91% of married couples with children live as independent nuclear family units, with no other extended family members or unrelated members residing in the household. In contrast, only about 60% of lone-mother families and just 41% of families headed by lone fathers live as independent nuclear family households. Among married-couple families with children, about 8% live in households with extended family members. In comparison, about one-quarter of lone-mother families and nearly an equal share of lone-father families live in households with extended family members. In some of these extended family member households, other nonrelatives may also reside. About 18% of lone-mother families and more than twice that share of lone-father families (41%) reside in households with unrelated household members (the top two tiers of each bar); among lone-mother families, 15% live only with unrelated members, and a small fraction (3%) also include extended family members in the household. Among lone-father families, 36% live only with unrelated members, and 5% also include extended family members in the household.

Figure A-1. Family/Household Living Arrangements Among Families with Children, 2004

In this report, we considered at times only nuclear family members' income and needs when assessing family economic well-being. In other instances, extended family members' income and needs were also included in the analysis. As noted above, some nuclear and related extended families live in households with unrelated members as well. However, the potential effect of unrelated household members' income and needs on nuclear or extended family well-being may vary to the extent that resources are shared among household members. In some cases, unrelated household members, such as unmarried partners, may make significant economic contributions to the nuclear family co-residing in the household. In other cases, the sharing of economic resources among unrelated household members may be minimal. However, there is no clear means of determining the extent to which unrelated household members share income and represent a joint economic unit. In this analysis, we excluded unrelated household members' income and needs when considering family economic well-being.

The effect of family living arrangements on poverty status can be illustrated by a closer examination of the families as economic units under various family configurations and income-sharing arrangements.Figure A-2 illustrates the effects of family living arrangements and the treatment of income and needs on family economic well-being, relative to poverty. The figure is for families headed by lone mothers.

Figure A-2. Poverty Status of Lone-Mother Families by Family/Household Living Arrangement and Poverty Income Definition, 2004

The figure shows that overall, 41.8% of families headed by lone mothers are poor, if just their immediate nuclear family's needs and income are counted for purposes of determining poverty status. However, if all related household members' income and needs are considered, the poverty rate of lone-mother families drops to 35%—the official Census Bureau poverty rate (which considers all related family members as the economic unit for poverty determination). If the economic unit is expanded to include all unrelated household members, the poverty rate for lone-mother families falls to 29.9%. It should be noted, however, that for purposes of income and poverty determination, official Census Bureau statistics do not combine unrelated household members' income or needs.

It was shown earlier in this appendix that about 60% of lone-mother families live as independent nuclear family households, absent any extended related family members or unrelated household members. Among these families, an estimated 37% are poor. Their poverty status is the same under all three poverty definitions: poor based on nuclear family income; poor based on extended family income; and poor based on household income.

Roughly 22% of lone-mother families live with other extended family members only. Among these families, an estimated 51% are poor based on their immediate family income. However, when the income and needs of other related household members are taken into consideration, their poverty rate drops to about 23%—less than half the rate when only immediate nuclear family income was considered, and a rate below that of lone mothers who live as independent nuclear families (37%). In part, the nuclear family poverty rate (51%) of lone mothers in such living arrangements may be so high because they live with extended family members who help provide for their support; this support is not counted in determining the 51% poverty rate but is evidenced by the 23% poverty rate. If these lone mothers were living independently, their poverty rate might not be as high as 51%, but it also might not be as low as the 37% among lone mothers currently observed to be living independently; some would likely work and earn more than they do when living with extended family, or would seek to qualify for government assistance in the absence of extended family income support.

The right-most set of clustered bars in Figure A-2 depict the poverty status of lone mothers living with only their children and at least one unrelated household member. As noted above, about 15% of lone mothers live in such households. When only her immediate nuclear family income and needs are taken into account, about 46% of these lone-mother families would be considered poor. However, if all unrelated household members are considered and their income counted for purposes of determining poverty, the poverty rate among these lone-mother families falls to 14.8%—well below the poverty rate of lone mothers living independently in nuclear families with no extended or unrelated household members. In these households with unrelated members, the extent to which household income is shared among members is uncertain. CRS estimates that about 1 million lone mothers, (approximately 65% of 1.6 million lone-mother families living with unrelated household members) may be in cohabiting relationships with a man.¹⁶ In some cases, unmarried partners may share economic resources and responsibilities much in the same way as married couples do.

Appendix B. Data Tables

Footnotes

1.	For a description of child poverty, see CRS Report RL32682, Children in Poverty: Profile, Trends, and Issues, by [author name scrubbed], [author name scrubbed], and [author name scrubbed].
2.	See Duncan, Greg J., and Jeanne Brooks-Gunn, ed. Consequences of Growing Up Poor. Russell Sage Foundation, 1997.
3.	In 2004, for example, the poverty threshold for a family of three, composed of two parents and one child, was $15,205. Poverty thresholds are lower for smaller families and greater for larger families. For example, in 2004, the poverty threshold for a married couple with two children (family of four) was $19,157, and for a single parent with one child, $13,020.
4.	See Figure 1, above, and Table B-2 in Appendix B.
5.	This is a "static" analysis, and does not consider the effects of government policies on individuals' or firms' behavior that might affect labor supply or wages.
6.	The primary earner in the family is the parent with the greatest earnings among the married couple.
7.	Also, it may be noted that family income poverty thresholds are somewhat higher to support a child than to support an adult in families of equal size.
8.	For a survey of this literature, see Altonji, Joseph and G. and Rebecca M. Blank. "Race and Gender in the Labor Market," in Handbook of Labor Economics, Volume 3C, Ashenfelter, O. C. and Card, D. eds. Elsevier North Holland. 2003. p. 3143-3259.
9.	The hourly wage equivalent is measured here as total reported annual earnings divided by the number of hours worked during the year (i.e., reported number of weeks worked multiplied by reported number of hours worked in a typical week).
10.	$9.62 per hour x 2,080 hours of work per year divided by the poverty line for a married couple with one child ($15,205) equals 132% of poverty.
11.	$9.62 per hour x 2,080 hours of work per year divided by the poverty line for a married couple with two children ($19,157) equals 104% of poverty.
12.	Parents' hourly earnings are converted to annualized amounts assuming the parent is able to work full-time, full-year (i.e., 40 hours per week for 52 weeks, or 2,080 hours annually).
13.	Economists have developed the theory of equalizing or compensating wage differentials to explain the structure of wages. Occupations that require larger investments in education must pay higher wages to compensate workers for paying the up-front costs of these investments. Though the impact of education has dominated recent discussions of wage determination, other attributes of the job might also affect its wage relative to other jobs. For example, jobs that entail physical risk, difficult working conditions, the risk of unemployment, or a higher than average risk of "failure" should pay a "wage premium"—higher wages than they otherwise would pay. On the other hand, occupations that have desirable characteristics, such as pleasant working conditions, higher than average job security, or offer unusual schedules (flexible or part-time) might pay less than otherwise would be paid. See Rosen, Sherwin. "The Theory of Equalizing Differences" in Ashenfelter, Orley C. and Richard Layard (eds.) Handbook of Labor Economics, Volume 1. North Holland Publishing Co., 1986, p. p. 641-692.
14.	Median wage data are from the March 2005 Current Population Survey's (CPS) "outgoing rotation group," comprising about one-fourth of the total March CPS sample. Each month, the outgoing rotation group is asked a series of questions to determine their usual hourly or weekly wage. For those employed by the hour, the information represents the reported hourly wage. For those who report they are not paid by the hour, an hourly wage equivalent was computed by dividing usual weekly earnings by usual hours of employment.
15.	Not shown in figure. See appendix Table B-7.
16.	We identify likely cohabiting couples based on inference. Couples are identified as likely cohabiting if there are only two unrelated adults in the household, of the opposite sex, who are both unmarried and where there are no related subfamily members present.