Poverty and Education

OVERVIEW
Christy Brady-Smith
Rebecca C. Fauth
Jeanne Brooks-Gunn

CHILDREN AND ADOLESCENTS
Rebecca C. Fauth
Christy Brady-Smith
Jeanne Brooks-Gunn

OVERVIEW

In 1998, more than 13 million children (19 percent of all children) under age eighteen lived in families with incomes below the official poverty threshold. Although children age eighteen and under represent 26 percent of the United States population, they comprise nearly 40 percent of the poverty population. Despite a steady decrease from 1993 (23%) to 1999 (17%) in the rate of children in poverty, the United States still ranks highest in childhood poverty among all industrialized nations.

In the United States, income poverty is defined by the poverty threshold, developed in 1959 and based on expected food expenditures (thrifty food basket) for families of varying sizes. Each year the threshold is adjusted for the Consumer Price Index cost of living. In 1999, the poverty threshold for a single mother raising two children was $13,423. Researchers have criticized the poverty threshold on numerous counts. First, government transfers such as food stamps and housing subsidies as well as tax benefits (e.g., the Earned Income Tax Credit) and tax payments are not included when assessing the poverty threshold. Second, regional and urban differences in the cost of living are not considered when computing the poverty threshold. Despite the criticisms levied against the way poverty is assessed in the United States, the current review highlights research that has used this definition of poverty, while acknowledging its weaknesses.

This article reviews the literature linking family poverty to children's cognitive and educational outcomes such as achievement tests, grade completion, and high school graduation. Timing of poverty has been shown to make a difference vis-à-vis child outcomes; thus, the discussion focuses on three stages of childhood: early childhood (age two to four), middle childhood (age five to twelve), and adolescence (age thirteen to eighteen). For each stage, we examine the effect of income poverty on children's cognitive ability and school achievement. Depth and duration of poverty are also considered, as these features have been linked to numerous domains of child development.

Consequences of Income Poverty on Children's Educational Outcomes

Simple comparisons between children in poor families and children in non-poor families using national datasets indicate that poor children are more likely to do worse on indices of school achievement than non-poor children are. Poor children are twice as likely as non-poor children to have repeated a grade, to have been expelled or suspended from school, or to have dropped out of high school. They are also 1.4 times as likely to be identified as having a learning disability in elementary or high school than their non-poor counterparts.

As with many studies examining the effect of poverty on children, correlational research is deficient in many ways. Simple comparisons between poor and non-poor children do not account for other family or child characteristics–such as single-parent households, maternal education, or child health problems–that may contribute to child development apart from the effect of poverty itself. Selection bias is another potential problem. Some researchers, such as Susan Mayer (1997), argue that poverty status often depends upon decisions made by family members, such as whether to apply for government services or seek employment, resulting in upwardly biased estimates of the relationship between income and child outcomes. Few feasible strategies are available that are capable of completely eliminating the problem of selection bias, though some are more successful than others are.

Recently released national data sources address many of the inherent problems in research on family income and child development. These studies include longitudinal indicators of socioeconomic variables (e.g., occupational status and prestige, years of schooling, and family and neighborhood income), measures of family income, and numerous child development assessments. Utilization of the data from these longitudinal studies has made possible the distinction between family income and its correlates on child development and outcomes. Moreover, longitudinal designs allow for an assessment of how the timing (early vs. late) and duration (transient vs. persistent) of poverty may differentially affect children's outcomes. When possible, this discussion presents finding from studies that used standardized tests of school readiness, achievement, and cognitive ability, and controlled for key family and child characteristics, in its review of the impact of poverty on children's educational outcomes.

Early childhood. During the 1990s, the nation was inundated with reports on the importance of the early years on children's brain development and later cognitive achievement. While some of the reports may have overstated the issue and understated the importance of a child's later years on development, evidence suggests that the early years may be a critical period of development in which family poverty has particularly strong effects on young children. As seen in Table 1, poverty occurring early in a child's life (age two to four) is associated with large effects on indices of child school readiness and cognitive outcomes.

Judith Smith and colleagues (1997), using data from two national datasets, showed that family poverty was significantly associated with lower scores on several measures of child cognitive and school readiness outcomes for children age three to four years, even after controlling for the effect of mother's education, family structure, and child race, birth weight, and gender. Moreover, according to Pamela Klebanov (1998), the effect of poverty was seen among children as young as two years of age.

Judith Smith, Jeanne Brooke-Gunn, and Pamela Klebanov (1997) found that children living in families with incomes less than half the poverty threshold (deep poverty) scored nine to ten points lower on

TABLE 1

cognitive ability than children in near-poor families (150 percent to 200 percent of the poverty threshold) at age three to four. Persistently poor children, those who spend all of their childhood years in poverty, experience more negative cognitive and educational outcomes than their peers who encountered only short-term or transient poverty. Children who lived in persistent poverty scored six to nine points lower on measures of cognitive ability and school readiness than children who were never in poverty. Although transient poverty was associated with worse child outcomes, the effect was not significant–children who had experienced short-term poverty scored only two to three points lower, on average, than children who were never in poverty.

Middle childhood. Measures of child educational outcomes in elementary school typically include school achievement test scores, grade failure, and learning and attention problems. Among school-age children, the effect of family poverty on child achievement parallels the findings for young children. Family poverty was significantly associated with lower reading and mathematics achievement scores for children at age five to eight. Other researchers, such as Thomas Hanson, Sara McLanahan, and Elizabeth Thomson, have found modest associations between family income-to-needs ratio (income divided by poverty threshold for family) and school performance and grade point average.

Children living in deep poverty scored ten to twelve points lower than near-poor children on measures of achievement and cognitive ability at age five to six. This effect was also found for older children–those in deep poverty scored seven to nine points lower on the achievement measures at age seven to eight. The differential effects of persistent and transient poverty were also seen in middle childhood. Five-to eight-year-old children who lived in persistent poverty scored six to ten points lower on measures of cognitive ability and school readiness than children who were never poor. Children who had experienced transient poverty scored three to six points lower, on average, than children who were never in poverty. A study in Canada by Linda Pagani and colleagues also found that persistent poverty was significantly related to academic failure. Children who had experienced poverty throughout their lives were twice as likely as never-poor children to be placed in a non-age-appropriate classroom.

Adolescence. Academic measures of adolescent educational outcomes include high school dropout and graduation rates, school engagement, GPA, and achievement test scores. Overall, the effect of poverty in adolescence on educational outcomes is relatively small. Findings suggest that a 10 percent increase in family income is associated with a .2 percent to 2 percent increase in the number of school years completed. Income poverty in the early years, however, is more strongly associated with high school completion than poverty during later childhood and adolescence, and most pronounced among those in deep poverty.

Pathways through Which Poverty Affects Children's Educational Outcomes

Poverty's effect seems to be the strongest when it occurs early in the child's life, when it is persistent, and when children live well below the poverty threshold. Poverty may influence child development through at least five pathways: (1) child health and nutrition,(2) parent mental health and affective interactions,(3) provision of a stimulating home environment,(4) school and child care quality, and (5) neighbor-hood conditions. Emerging evidence suggests that the influence of family poverty on children's cognitive outcomes may be entirely mediated by these pathways. The latter four pathways are briefly reviewed below.

Parental mental health and affective interactions. Parents who are poor or who have a history of welfare receipt are more likely to have worse emotional and physical health than those who are not poor. Researchers such as Rand Conger and colleagues, using the family stress model, have linked parental depressive symptoms to more conflict with adolescent children, which in turn results in less optimal emotional, social, and cognitive outcomes. Poor parental mental health also is associated with less stimulating home environments and more discordant parent-child interactions.

Provision of a stimulating home environment. Family income directly influences the material resources available to children in their homes. Higher income children benefit from higher levels of cognitively stimulating materials available in their homes compared to low-income children. The provision of a stimulating home environment, in turn, accounts for much of the effect of income on the cognitive development of preschool and elementary school children and may be the most important pathway through which poverty operates.

School and child care quality. Children from poor families are also exposed to lower quality school and child care settings compared to their non-poor counterparts. Findings from two national child care studies, as described by Deborah Phillips et al. (1994), indicate that up to 60 percent of subsidized and low-income child care centers failed to conform to legal child to staff ratios in toddler classrooms and most (70%) received low ratings on scales of appropriate caregiving and the provision of appropriate activities. Lower quality child care is associated with lower math and language ability, negative peer interactions, and more behavior problems.

Neighborhood conditions. The neighborhoods in which poor families reside are another pathway through which income poverty may negatively affect children's educational outcomes. Financial strain limits the housing and neighborhood choices available to low-income families, constraining these families to live in neighborhoods characterized by high levels of crime and unemployment, low levels of resources, and a lack of collective efficacy among the residents. Neighborhood residence, in turn, is associated with child and adolescent school outcomes above and beyond the effect of family poverty.

Conclusion

The challenge currently facing researchers is how to accurately estimate the effect of income poverty net of other factors, such as measured and unmeasured family and individual characteristics. There are several statistical options available that may provide upper and lower bounds of effects. Regressions, such as those reviewed above, that control for a host of demographic and parental characteristics are one option that provide an upper bound of effects. Including more specific controls, such as maternal verbal ability or IQ scores (which most of the findings in this article did not do), would provide a more accurate estimated effect of poverty on child outcomes. Another option is to control for assets or permanent income. Sibling analyses, which control for many unmeasured family characteristics, also provide a more accurate effect of poverty. Experimental studies, in which participants were randomly assigned to receive income supplements or the usual welfare program, provide some of the most powerful evidence thus far that income influences child achievement. Parents randomly assigned to programs that increased earnings had elementary school-aged children who scored consistently higher in the area of school achievement, compared to those in the control program.

Although the complex pathways through which family poverty affects child educational outcomes have yet to be fully understood, it is clear that childhood poverty compromises the educational prospects of children and adolescents. Many poor children begin life at a disadvantage, due to family income, low maternal education, single parents, young parents, or a combination of these factors. The level of disadvantage may become exacerbated through the lack of cognitively stimulating or safe home environments, conflicted parent-child interactions, poor school and child care environments, and poor neighborhood conditions. As knowledge of the pathways through which income poverty affects child cognitive and educational outcomes expands, this nation will come closer to understanding how best to address the problem of childhood poverty.

Although the percentage of children ages 18 and under living in poverty in the United States dropped from 18.9 percent in 1998 to 16.9 percent in 1999, children in the United States are more likely to be poor than any other age group. Moreover, U.S. child poverty rates are well above those in other industrialized nations, such as Spain, Canada, and Australia. Put simply, 12.1 million children in the United States are living in families with earnings at or below the conservative income poverty line ($17,601 or less for a family of four, according to the U.S. Census Bureau's 2001 guidelines), indicating they are likely to be lacking the finances to fund basic needs such as food, shelter, and clothing.

The rise in both divorce rates and out-of-wedlock births in the last quarter of the twentieth century is partially responsible for the staggering number of poor children in the United States. One-third of U.S. births in the early 1990s were to unmarried mothers. Single-parent families are more likely to be poor due to both their dependence on a single wage earner for income and the probability that the household heads are younger and less educated than their dual-parent counterparts. Given this information, it is not surprising that 53 percent of poor families are headed solely by a female adult.

Changes in employment patterns, declines in the manufacturing sector, relocation of jobs into the suburbs, and wage polarization may also explain the high national poverty rates. The Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996 also has the potential to affect U.S. poverty rates. With PRWORA's prescribed time limits, work requirements, sanctions, and categorical restrictions for cash assistance, many families may find themselves unable to receive needed benefits. Moreover, the current welfare legislation does not offer a safety net for parents facing difficulties transitioning into full-time work.

Effects of Income on Child Outcomes

Findings are modest but significant regarding the direct role of poverty in influencing children's cognitive and school-related outcomes throughout early, middle, and late childhood. Children reared in conditions of poverty face more adverse developmental outcomes than their non-poor counterparts. Much of the research prior to the mid-1990s on the impact of poverty status on children suffered from shortcomings such as flawed measures or unrepresentative samples. Additionally, the differences between income and socioeconomic status (SES) were often blurred, elevating the risk of using inappropriate indicators of SES such as occupational prestige or level of maternal education (fairly stable variables) as proxies for income (which is often quite volatile). Large, longitudinal studies such as the National Longitudinal Survey of Youth (NLSY), the Infant Health and Development Program (IHDP), and the Panel Study of Income Dynamics (PSID) have remedied many of these methodological problems and include adequate child development assessment instruments, measures of family income, and socioeconomic status indicators. A common current practice of researchers is to calculate an income-to-needs ratio, which takes into account family size relative to the current poverty line; a ratio of one or below denotes living in poverty. Utilization of this poverty measure compared with a two-pronged (poor versus not poor) approach allows for a more thorough analysis of how different depths of poverty affect children's outcomes.

Physical health. Beginning prior to birth, poor children experience more health problems than their non-poor counterparts, even after adjusting for potentially confounding factors. Controlling for mothers' age, education, marital status, and smoking status, women with incomes below the federally established poverty line were found to be 80 percent more likely to bear an infant at low birth weight (2500 grams or less) than women whose incomes remained above the poverty line. Children born at a low birth weight are at risk for negative outcomes well into their childhoods. Compared to full-term children, neurologically intact very-low-birth-weight children (1500 grams or less) present more impairments in arithmetic, motor and spatial skills, language, and memory, and perform worse on measures of achievement. Children with birth weights of less than 1000 grams are at the highest risk. One study revealed that 34 percent of low-birth-weight children were either repeating grades or placed in special education classrooms in school; only 14 percent of normal-birth-weight children experienced the same outcomes. Other research also reports elevated levels of grade repetition as a result of low birth weight. One study of siblings found that low birth weight reduces children's chances of graduating high school by nearly 75 percent when compared with their full-term siblings regardless of family income. Sibling studies control for unmeasured family characteristics and for selection bias–the fact that families with low-birth-weight children may be different in important ways from those who do not have low-birth-weight children–and therefore may provide more nuanced results and greater effects. The detrimental effects of low birth weight remain intact, though diminished, through the adolescent years and possibly into adulthood.

Although the rates of child malnutrition are lower in the United States compared with other countries, poverty is associated with nutrition-related disorders. Often linked with nutritional deficits, poor children in America experience higher rates of growth stunting (low height-for-age) and wasting (low weight-for-height) than their non-poor counterparts (13% and 5% respectively). These height and weight differences between poor and non-poor children are greater when long-term rather than single-year measures of income are used. Stunting and wasting are associated with lower cognitive test scores; short-term memory is especially diminished. Chronic malnutrition, often the cause of stunting, is associated with low cognitive test scores even after controlling for poverty status and its correlates. Children who are born both at low birth weight and short for their age have lower test scores that children who fit in only one of these categories.

Exposure to lead poisoning presents yet another heightened risk for poor children. The prevalence of elevated blood lead levels (20 to 44 μ per deciliter) was four times greater for children from low-income families (16.3%) compared with their peers from high-income backgrounds (4.0%). Rates were especially high (28.4%) for low-income black children. The deleterious effects of lead exposure on children vary depending on the severity of exposure, and also on the developmental stage of the child. At young ages, elevated blood lead levels are associated with stunted growth, decreases in IQ, and various physical impairments. A study conducted by Walter J. Rogan and his colleagues explored the effect of chelation therapy to reduce blood lead levels in poor children ages twelve to thirty-three months. The study found that although the treatment did lower children's blood lead levels, IQ scores were lower and parent-reported behavior problems were higher for treated children than for controls who only received a placebo at a thirty-six-month follow-up. Chelation therapy proved unable to successfully prevent deleterious outcomes for lead-exposed children; therefore, prevention of lead exposure for poor children needs to be a priority.

Health problems and nutritional deficits are important ways through which poverty affects children's cognitive and school-related outcomes. Poor children face an increased probability of being born at a low birth weight and experiencing both nutritional deficits and elevated blood lead levels. Past research suggests that the prevalence of these conditions in addition to other health problems can account for as much as 20 percent of the difference in IQ scores between poor children and their non-poor peers during preschool. Efforts to improve the health and well-being of poor children, including improved access to affordable health care and the provision of nutritional supplements and food stamps, are necessary to promote school-related success and competence.

Mental health. Children's mental health can be assessed via dimensional scales that screen children for various types of less severe subclinical symptoms (e.g., behavior problems) or with scales that more formally determine whether or not a child's behavior falls within a clinical diagnostic category according to the Diagnostic and Statistical Manual of Mental Disorders–Fourth Edition (DSM-IV). Most large national studies contain measures that assess dimensionality/level, such as the Child Behavior Checklist. Parental reports of behavior problems are the most common indicator of children's mental health and are generally classified along two dimensions: externalizing or undercontrolled behaviors including aggression, fighting, and acting out, and internalizing or overcontrolled behaviors such as depression, anxiety, and withdrawal. The reliance on maternal reports of behavior problems is a shortcoming in many of the large national studies. While these studies do control for maternal mental health and other characteristics that may influence how parents rate their children's behavior, higher instances of teacher-rated behavior problems are needed to corroborate mothers' responses.

Studies have established a negative link between income and behavior problems, although modest in size compared to the effects of income on both physical and cognitive outcomes. After controlling for a host of family-level variables including maternal education and single parenthood, the effects of income or poverty on children's emotional and behavioral outcomes are often muted or dissipate altogether. Some researchers, however, have found significant effects of poverty on children's mental health.

Using data from both the IHDP and the NLSY, researchers found that kindergarten-aged children growing up in persistently poor homes had substantially higher internalizing behavior problem scores than their peers who were never poor. Lower self-esteem, which is related to internalizing problems, most notably depression, may also result from persistent poverty, although mothers who are actively involved in their children's lives may lessen this effect. The timing of poverty in children's lives may also be relevant. Children who suffered from persistent poverty early in their lives maintained heightened levels of depression through the age nine in one study. Children who experienced economic hard-ship were also less popular among their peers than those who did not face economic adversity, possibly due to their higher levels of behavior problems.

Research is mixed as to whether current poverty status or persistent poverty contributes to children's externalizing behavior problems, including antisocial behavior. Data from the Charlottesville Longitudinal Study, examining elementary-school-aged children, found that children who endured even one year of family economic hardship possessed higher levels of externalizing behaviors than children who did not experience poverty; boys were more adversely affected than girls. Other research, using an income-to-needs ratio as an index of family poverty, reported a link between poverty and delinquency for young boys studied longitudinally between the ages of 6 and 16 years. Boys experiencing transitory poverty were at greater risk of engaging in extreme delinquent acts than boys who were never poor. Less severe manifestations of externalizing behavior were not influenced by family poverty in this particular study. Thus, families' poverty status may have its greatest impact on externalizing behaviors in their most severe form.

Due to the confounding of race/ethnicity and class and the fact that poverty rates among blacks and Hispanics living in the United States are two to three times higher than that of non-Hispanic whites, it is important for researchers to consider children's racial/ethnic backgrounds when examining links between poverty and child outcomes. One line of research suggests that the effects of persistent poverty on children's behavior problems, both internalizing and externalizing scales, may be stronger for white than for African-American children, possibly explaining some of the discrepancies regarding the relationship between poverty and externalizing problems. Other research found the opposite pattern emerging with internalizing scores. For black children, internalizing behaviors increased with the number of years spent in poverty, whereas the inverse was found for persistently poor white children.

Thus far, the direct impacts of income on children's mental health have been considered. Risk indices, which take into account other factors correlated with family poverty status (i.e., low maternal education, single parenthood, and parental substance abuse and mental health), provide an alternate way of examining the effects of poverty on children. Not surprisingly, the more risk factors children experience, the greater the number of behavioral problems they show at six and seven years of age. Moreover, behavioral maladaptation is more likely to occur for children experiencing numerous risk factors simultaneously rather than one individually. Parental attributes or behaviors such as substance abuse, antisocial behavior, negative life events, and psychiatric episodes tend to be the most devastating risks for children. Thus, it may not be poverty per se that has a negative impact on poor children, but other risks experienced more frequently by poor children than by their nonpoor counterparts.

Parenting behavior and parent mental health. The Family Stress Model has been used by researchers to examine the indirect links between poverty and child development. According to this model, stress generated by economic hardship and income loss results in parenting that is less consistent, less supportive, less involved, and more coercive and harsh. These lower quality parent-child interactions lead to higher rates of behavioral and school-related problems among poor children. Maternal mental health, most notably depression, is also negatively affected by poverty. The higher incidences of maternal depression and distress in poor families most likely drives the use of these less effective parenting styles. Many studies have documented the link between maternal depression and children's internalizing and externalizing behavior problems, indicating that the emotional detachment and disengagement resulting from depression is quite harmful to children. Other indirect influences include the home environment, child care and schools, neighborhoods, health care, and exposure to violence.

Implications of mental health on children's educational outcomes. Poor children's academic performance may be hindered due to the higher rates of behavior problems within this group. A large body of research has established a link between behavior problems and outcomes in the academic or cognitive domain early and late in children's school careers. Difficulty in paying attention, a form of externalizing behavior, is highly associated with underachievement for younger, elementary-school-aged children. Aggression toward peers in fifth grade has been found to predict school dropout. Compared with peers who were never poor, persistently poor children were found to be most at risk for fighting with peers in the sixth grade, even after teacher-rated behavior problems in kindergarten are taken into account. These same children were two times more likely than their never–poor peers to be in a non–age appropriate classroom and to be at risk for experiencing academic failure.

Other studies indicate that both externalizing and internalizing symptoms (including hyperactivity, intrapersonal adjustment, lethargy, and withdrawn behaviors) in kindergarten or first grade may be associated with poor reading and math achievement the following academic year. Early behavior problems have been found to have relatively long-term effects on reading skills, IQ, and class placement during later school years, possibly even into high school. Teacher-reported ratings of internalizing and externalizing symptoms in second and fifth grades predicted grade point averages and achievement test scores during children's freshman and senior years of high school.

Past research has shown that teachers at high socioeconomic level who teach poor children tend to regard their pupils in a more negative light, in terms of behavioral characteristics (maturity/immaturity), academic expectations and performance, even in the early school years. These teachers were also more likely to view the school and classroom contexts less positively than teachers of children at higher socioeconomic levels, possibly creating a self-fulfilling prophecy in which these children begin to act, behaviorally and academically, in line with their teachers' thoughts at very early ages. Some have argued that in order to mitigate the powerful effects of poverty and its correlates on children's school outcomes, teachers must allow students greater autonomy and decision-making, have positive expectations for all of their students, possess good management, disciplinary control, and exemplary organizational skills, and provide a variety of challenging, yet meaningful tasks.

Policy Implications

It is always important to contemplate how government social policy concerning income transfers, employment, child care and schooling, social services, housing, and health care can potentially help or hinder our nation's poor families. Preliminary results from ongoing research examining the effects of mothers transitioning off welfare and obtaining employment suggest that job earnings often are not lifting families out of poverty. Without a significant increase in earnings, it is likely that these working families may not see any marked improvements in children's well-being. Moreover, if these families continue to live in high-poverty neighborhoods with elevated crime rates and inferior public schools, are unable to afford high-quality child care for their children, and do not receive family health benefits, the situation seems dire still. Policy efforts may need to go beyond cash transfers and concentrate instead on the provision of high-quality services such as medical care (e.g., Medicaid), the provision of nutritional information and supplements (e.g., food stamps; Women, Infants, and Children; and the National School Lunch Program), early childhood education (e.g., Head Start and Early Head Start), and housing (e.g., Section 8 vouchers). It is not clear whether programs that target some of the indirect effects of poverty, such as in-home interventions, are effective in improving adults' parenting skills and their ability to facilitate literacy and learning within the home, which are crucial for children's school success. Continued investment in creating new and unique policies and programs that are aimed at either preventing economic deprivation altogether or help to assuage its negative effects are needed.

Conclusion

The effect of poverty on children's physical and mental health begins early and often continues to leave its mark well into the adolescent years. While many researchers have examined the connections between growing up poor, children's health, and their cognitive and school-related outcomes, few independent research efforts have made similar connections between the pernicious effects of low income on youngsters' mental health and the subsequent impact of poor mental health and behavior problems on children's academic performance. The association between income poverty and mental health is not as strong as its effects on physical and cognitive and/or academic outcomes. In fact, when effects are found, they are often mediated by parenting and parental mental health. Children are negatively affected by poverty throughout their lives in many domains, especially when the poverty occurs at a young age and in severe form. Policy efforts to assist poor families in order to bolster children's chances of succeeding in school need to start early and target the entire family and their environment.