Skip over navigation

Journal Issue: Opportunity in America Volume 16 Number 2 Fall 2006

Children's Health and Social Mobility
Anne Case Christina Paxson

The Relationship between Economic Status and Health in Childhood

Numerous studies have analyzed the relationship between income and children's health. They have examined a variety of health measures, ranging from health status broadly defined to very specific health conditions experienced by children of different ages. A general conclusion is that lower-income children are more likely to be in poor health than are children from higher income groups.

Economic Status and Global Health Status
The National Health Interview Survey (NHIS), a nationally representative annual survey of U.S. families, asks respondents (or, for children, their adult caregivers) whether they are in excellent, very good, good, fair, or poor health. The resulting summary measure of health, called global health status, although crude, is highly correlated with specific types of illnesses and health conditions in childhood. Adults who report poorer global health status are more likely than others to become ill and to die sooner rather than later.1 Using the NHIS surveys conducted from 1997 to 2003, we estimate how parents' reports of their children's health vary with family income.2 Figure 1, based on these estimates, charts the share of children of different ages and with different family income who are reported to be in excellent or very good health.

For all age groups, children from higher-income families are more likely than those in other income groups to be in excellent or very good health. Among children from birth to age three, for example, fewer than 75 percent of those with family incomes less than $10,000 a year were in excellent or very good health, as against more than 90 percent of children with family incomes greater than $100,000. The relationship between health and income is apparent throughout the income range: middle-income children are healthier than lower-income children, and upper-income children are healthier than middle-income children. Moreover, income-related differences in health become more pronounced as the children grow older. Among children from birth to age three, those at the highest income level are 21 percentage points more likely to be in excellent or very good health than those at the lowest income level. This difference increases to 29 percentage points for children aged fifteen to seventeen. Children from poorer families are substantially more likely than their wealthier peers to enter adulthood with health problems.

Other researchers report similar findings. An analysis we conducted with Darren Lubotsky, using three large nationally representative data sets—the Panel Study of Income Dynamics, the National Health and Nutrition Examination Survey, and earlier years of the National Health Interview Survey—documents both that children's health differs by family income and that the gaps widen as children age.3 Paul Newacheck and several colleagues conclude that the health of teens from poorer families is worse than that of teens from wealthier families.4 Elizabeth Goodman reaches a similar conclusion in examining participants in the National Longitudinal Study of Adolescent Health, a nationally representative data set in which adolescents rate their own health.5 Other broad (and arguably more objective) measures of poor health—including days spent in bed because of illness, school days missed because of illness, and hospital episodes—also decline as income rises.6

These income-related differences in health are not attributable to differences in health insurance coverage. In an earlier study, we found that even among children who have private insurance, higher-income children are in better health than lower-income children.7 Nor do these differences exist only in the United States: Janet Currie and Mark Stabile find a nearly identical link between children's global health status and family income in Canada, which has universal health care.8

Finally, racial and ethnic differences in health status do not account for the income-related differences either. Using the same methods as for figure 1, figure 2 shows the share of black non-Hispanic, white non- Hispanic, and Hispanic children of various family incomes in excellent or very good health. African American and Hispanic children have worse global health status, on average, than white children with the same family incomes.9 But within each racial and ethnic group, wealthier children are in better health. All the U.S. studies we have mentioned above also find strong links between family income and children's health after adjusting for differences in health across race and ethnic groups.

Socioeconomic Status and Birth Outcomes
Income-related disparities in childhood health are evident at birth or even before. Much research on this topic focuses on low birth weight, which provides a measure of the quality of both the intrauterine environment and the medical care received during pregnancy. Small newborns are categorized as being “low birth weight” (less than 2,500 grams), “very low birth weight” (less than 1,500 grams), or "extremely low birth weight" (less than 1,000 grams). Low birth weight stems from preterm birth (defined as less than thirty-seven weeks of gestation), prenatal growth retardation, or both. Almost all babies with very low birth weight are born preterm. Although low birth weight is not uncommon, only a small fraction of infants have very low and extremely low birth weights. In 2002, for example, 7.8 percent of infants had low birth weight; 1.5 percent, very low birth weight; and only 0.7 percent, extremely low birth weight.10

Low birth weight is associated with a variety of neurodevelopmental problems, including cerebral palsy, blindness, impaired lung function, and mental retardation. The smallest and most premature children are at much greater risk for these problems, though the rates of major disability among even the most premature infants (born at less than twenty-seven weeks of gestation) are relatively low. Only one-fifth to one-quarter of surviving infants born at less than twenty-seven weeks of gestation experience a major disability, including impaired mental development, cerebral palsy, blindness, or deafness.11 Nevertheless, children born at very low birth weight without a major disability may have more subtle mental and emotional problems, such as attention deficit hyperactivity disorder (ADHD), behavioral problems, and reduced IQ. A recent review of the research concludes that infants who are low birth weight, especially those who are premature, have slightly lower IQs than normal-weight full-term babies.12

Children from low-income families are more likely than other children to have low birth weight. Among poor children, the rate of low birth weight is 10 percent, as against 6 percent among nonpoor children.13 The National Health Interview Survey reveals similar income- related disparities in rates of low birth weight.14 Among children with annual family incomes below $30,000 (measured in 2000 dollars), 9.3 percent were born at low birth weight and 1.5 percent at very low birth weight. Rates for children with family incomes between $30,000 and $60,000 were 6.9 percent and 1.1 percent, respectively. For children whose families earned more than $60,000, 5.6 percent had low birth weight and 0.8 percent had very low birth weight. As with global health status, the disparity is not just between poor and nonpoor children; birth outcomes improve steadily with income.

That poorer children are more likely to be born at low birth weight suggests that socioeconomic differences in health emerge even before birth. Because it is difficult to measure fetal health directly, researchers have instead focused on factors that may affect fetal health, such as socioeconomic differences in prenatal care and the incidence of risky behaviors in pregnancy.15 Much of this research uses maternal education rather than family income as the measure of socioeconomic status, because the former but not the latter appears on birth certificates, which typically provide the data for analysis.

The use of early and regular prenatal care varies widely by maternal education. According to the National Vital Statistics, 68 percent of women without a high school degree began prenatal care in the first trimester of pregnancy, compared with 81 percent of high school graduates and 91 percent of women with at least some college education.16 An important goal of prenatal care is to inform women about proper nutrition during pregnancy, and it appears that this goal is not being met for women with lower socioeconomic status. For example, women with less education are more likely to have folic acid deficiencies (associated with spina bifida and other neural tube defects)—indicating either poorer diets or less use of vitamin supplements during pregnancy.

Analysts observe similar patterns for cigarette smoking during pregnancy, a behavior that has been implicated in preterm birth, intrauterine growth retardation, and subtle but long-lasting effects on cognition and behavior. According to recent statistics, 78 percent of pregnant women without a high school degree refrained from smoking during pregnancy, as against 83 percent of those who were high school graduates and 94 percent of those with at least some college education.17 Although it is difficult to gather reliable information on alcohol and illegal drug use, women with less education also appear more likely to use alcohol and drugs during pregnancy. Self-reported rates of drug use, though, are low. Shahul Ebrahim and Joseph Gfroerer, using data from the National Household Survey on Drug Abuse, report that 2.8 percent of pregnant women surveyed between 1996 and 1998 reported using illicit drugs.18  And during the 1980s and 1990s, although 20 percent of women reported consuming at least some alcohol during pregnancy, only 1.3 percent reported an episode of binge drinking.19 Unless pregnant women greatly underreport binge drinking and illicit drug use, alcohol and drugs cannot account for much of the income-related differences in children's health at birth.

Socioeconomic Status and Health Conditions in Childhood
Children experience a wide variety of health problems, from common ailments such as colds and upset stomachs to rare and more serious conditions such as muscular dystrophy and cerebral palsy. Some problems appear shortly after birth; others develop later. But despite the diversity of these health conditions, lower-income children experience a broader set of specific health problems than do children from higher-income households.

Based on parent reports, nonpoor children are more likely than poor children to have only a handful of relatively minor health conditions, such as hay fever and sinusitis. Poorer children, by contrast, are more likely to have asthma, frequent headaches, heart conditions, kidney disease, epilepsy, digestive problems, mental retardation, and vision and hearing disorders.20 Researchers comparing children in different social classes in the United Kingdom make similar findings.21 Although many of these health conditions are rare, a substantial fraction of children have at least one. Paul Newacheck and Neal Halfon find that 9.6 percent of poor children and 5.7 percent of nonpoor children under age eighteen suffer from a disability, defined as a physical or mental health condition that limits their activities.22 Some mental health and cognitive problems, such as learning disabilities and developmental delays, are also more common among poor than among nonpoor children.23 Evidence on depression is mixed. Research using a nationally representative survey of adolescents finds that poorer adolescents are more likely to experience depressive symptoms.24 A comparative review of studies based on the Children's Depression Inventory, however, finds no link between socioeconomic status and depression in children and adolescents.25

Socioeconomic Status and the Effects of Health Problems on Children
Not only are poor children more likely to have a variety of health problems, they also fare less well than wealthier children who have the same problems. Consider, for example, two children with asthma, one from a low-income family and the other from a high-income family. The low-income child will be more likely to be reported in poor health, to spend more days in bed, and to have more hospital episodes.26 Similar patterns emerge for other serious (although less common) health conditions, such as diabetes and epilepsy. A study of air pollution and children's asthma in California finds that poorer children are not only exposed to more pollution, but also more likely to be hospitalized than nonpoor children who live in similarly polluted areas.27

Poorer children could fare worse than wealthier children with the same health conditions for several reasons. First, there is evidence that poorer children receive less and lower-quality medical care for their problems. Poor children are less likely than nonpoor children to have a usual source of health care.28 Even when poor children have a usual source of care, they are less likely to have continuity of care with a particular primary physician. They are also significantly less likely to be vaccinated for measles and to have received medical attention for specific acute health conditions, including pharyngitis, acute earache, recurring ear infections, and asthma. Second, poor families may be less well equipped to manage their children's health problems. Many such problems, including asthma and diabetes, require a great deal of parental oversight. Parents of children with asthma, for example, must monitor medications and keep their homes free of dust mites and tobacco smoke, which can exacerbate asthma. Parents of children with diabetes must carefully monitor blood glucose levels, administer insulin, and provide an appropriate diet. Evidence for selected childhood health conditions indicates that poorer families are less likely to comply with medical protocols, which could worsen the effects of health problems.29

Do Children's Health Problems Affect Family Income?
Although poor childhood health and low income are linked, it could be that low income does not cause the poor health. It is possible that the relationship runs the other way— that children's health problems lower family incomes. Mothers with sick children may be more likely to stay home rather than work; the stress of having a sick child may lead to a marital break-up that strains family finances; single mothers with sick children may find it more difficult to find new partners to bring income into the household.

Researchers have found mixed support for these hypotheses. For example, Hope Corman, Nancy Reichman, and Kelly Noonan, using a sample of primarily low-income single mothers, find that mothers with children born in poor health are about 10 percentage points less likely than mothers with healthy babies to be working when their children are twelve months old. And when these mothers are employed, they typically work about four fewer hours a week.30 Some researchers find that single mothers with a disabled child work fewer hours than other mothers.31 But others conclude that mothers of children born in poor health are no less likely to work in the three years following the child's birth.32 Research on the effects of children's health on family structure yields somewhat more consistent evidence, at least for the United States. Angela Fertig, using two nationally representative U.S. data sets, concludes that parents of children born in poor health are more likely to divorce.33 That finding, however, does not hold true in Britain. Another study of primarily low-income women who are unmarried when their children are born finds that the mothers of children born in poor health are less likely to be cohabiting or married when their children are a year old.34

Interpreting these findings is complicated because unobserved factors that affect child health could also affect maternal labor supply and family structure. For example, mothers with drug or alcohol problems may be more likely to have children with health conditions and also more likely to divorce. But no matter how these findings are interpreted, several pieces of evidence argue strongly against the theory that reductions in family income caused by a child's poor health can explain the observed link between child health and socioeconomic status. First, as noted, children whose parents have less schooling are more likely to be born into poor health. But except for very young parents, children's health problems cannot lower their parents' educational attainment. Second, our study conducted with Darren Lubotsky shows that the link between child health and family income after the child's birth is the same as that between child health and family income before the child was born.35 A child's poor health cannot possibly lower the family's income before the child's birth, or at least before conception. Thus, we do not believe that the link between low income and poor childhood health is attributable to the fact that children's poor health lowers their families' income.