eJournal Woodring College of Education Western Washington University eJournal

Western Washington University

ISSN 1935-7699
Journal of Educational Controversy

Problematizing Payne and Understanding Poverty: 
An Analysis with Data from the 2000 Census

Jennifer C. Ng, University of Kansas
John L. Rury, University of Kansas

Of all the preposterous assumptions of humanity over humanity, nothing exceeds most of the criticisms made on the habits of the poor by the well-housed, well- warmed, and well-fed.
 - Herman Melville, “Poor Man’s Pudding,” 1854

Poverty is a problem with important educational implications.  Poor children, for example, often struggle academically in school, pose unique disciplinary problems for teachers and administrators, and may require additional resources to be successful (Byrne, 2008; Jackson, 2008).  Given these challenges, school districts have devoted considerable time and money to professional development intended to better prepare educators to work with children from poor backgrounds.  Perhaps the most visible contemporary spokesperson on poverty and education to whom districts have turned is Dr. Ruby Payne, author of A Framework for Understanding Poverty (2005).

Measures of Payne’s influence are remarkable to consider.  Her aforementioned book has sold over one million copies and been translated into other languages such as Spanish since its publication in 2005.  Payne has also launched a speaking career by conducting professional development workshops in 38 American states and internationally.  She trains approximately 40,000 educators a year and reports having worked with 70 to 80 percent of the nation’s districts over the last decade with the assistance of her staff and consultants (Shapira, 2007). 

While Payne’s popularity cannot be disputed, her work has generated great controversy and criticism.  For example, questions have been raised about the methodological validity of her work and subsequent self-proclaimed “expertise” (Baker, Ng & Rury, 2006).  Others have criticized the deficiency-oriented nature of her views on poor people that results not only in blaming the victim for being poor in the first place, but also blaming the victim for not exercising the power to alleviate his/her poor condition (Bohn, 2006; Osei-Kofi, 2005; Gorski, 2006a & 2006b).  Reviews of Payne’s published materials also indicate her inaccurate characterization of existing social science research and reliance upon stereotypes that poor people are disproportionately more immoral, lazy, and promiscuous than middle-class or wealthy individuals (Ng & Rury, 2006).  And lastly, a careful analysis of the 607 “truth claims” she makes in her text reveals that the majority of her assertions actually contradict the findings of empirical work in fields such as education, anthropology, and sociology (Bomer, Dorin, May & Semingson, 2008). 

Clearly, the merits of Payne’s work on poverty are highly disputed.  Her extensive involvement with school districts across the country underscores educators’ interest in the topic, however, and the desire to be equipped with accurate understandings of poverty and its effects.  This is a challenging task since poverty is a historically dynamic, governmental construct with real consequences affecting people across different racial, ethnic, and gender backgrounds; living in a variety of family arrangements; and residing in rural, suburban, and urban communities.  In other words, it is a complex problem that cannot be captured by pat generalizations.

In this article, we examine the statistical parameters of childhood poverty in the United States with critical reference to the views of Ruby Payne.  We begin by utilizing data from the 2000 U.S. Census to describe the conditions of poverty among American children and youth.  In particular, we consider the scope of poverty and its correlation with a number of other circumstances in children’s lives that may bear on their success in school, such as family structure and parental education.  Because poor children are not evenly distributed across American society, we also assess the geographic dimensions of poverty that highlight where they live and the degree to which they are concentrated in certain types of communities.  As it turns out, each of these aspects of poverty is important to understanding its impact on the lives of children at this point in history. 

Our analysis culminates by considering how poverty and other conditions of children’s lives are associated with accomplishment in school, measured by their likelihood of attainment.  We find that children’s poverty status is one of many characteristics statistically associated with educational outcomes, but it is probably not the single most critical factor in predicting their school attainment.  We also follow Payne’s lead and consider the circumstances of children contending with multiple risk factors who might well be described as “the truly disadvantaged.”  Success in school appears to be especially problematic for these children; yet, poverty is again but one of several conditions that appear to affect their education.  We conclude by discussing the implications of this analysis as they relate to the educational dilemmas of working with poor children today, especially in light of Ruby Payne’s limited treatment of the issue and the pressing need for a more accurate understanding.

Data & Methods

Payne makes broad generalizations about the causes of poverty, how it affects children, and how it should be remedied with little or no reference to systematic data on these questions.  In order to better inform the understanding of educators and the general public, we believe it useful to present some statistical information from the 2000 census to illustrate the extent of poverty in the U.S. and how it is associated with different social conditions and life experiences.  To do this, we utilize data drawn from the Integrated Public Use Microdata Samples (IPUMS) database at the University of Minnesota Population Center (n.d.), a 1% sample of children aged five to seventeen, specifically.[1]  We focus on a section of the northern United States—a broad region extending from Wisconsin and Illinois eastward to New Jersey and then New England in the north.  While the child poverty rate in this 14-state region is slightly lower than that of the country as a whole (15.5% vs. 16.5%), this sub-sample includes more information on the particular types of communities in which children live than the data collected for other parts of the country.[2]  We term this a geo-spatial dimension, as it refers to the geographic arrangement of different communities in relation to each other, and as we shall demonstrate, it is a factor that clearly affects children’s lives. 

The IPUMS data also permit consideration of other variables including characteristics of individual children, their parents, and their households.  For example, several of the key educational measures examined include the highest level of schooling a child has completed and his/her current enrollment; the single- or two-parent structure of a child’s family; and the age, educational attainment, recent work history, and disability status of a child’s family household head.  We exclude children living in group quarters arrangements, such as correctional or mental institutions or college campuses, for the purposes of this analysis, however, because information on their parents’ characteristics is incomplete.  Although the IPUMS data provide only a cross-sectional look at this sample, it includes a rich set of indicators for considering the effects of poverty on children and their education.

In the discussion that follows, we utilize these data to answer several questions.  First, what do we know about the social and economic realities of children who live in poverty?  And secondly, how does poverty affect the educational attainment of these children, measured by their reported completion of a certain grade level in school at a given age?  Our results yield descriptive information about poor children generally, as well as an understanding of how various factors in children’s lives are associated with growing up in poverty.  In light of these findings, we conclude with a discussion of implications for interpreting the work of Ruby Payne and others on poverty and education.

Who are the Poor Children?

Poor people have been the subject of many debates in American history to determine whether poverty is a matter of defective personal character and rectitude, or the result of unfortunate life circumstances and unforeseen events (Gans, 1995).  While Payne tends to favor the former explanations and ignore evidence of the latter, much social scientific scholarship in the last half century has emphasized circumstantial or contextual explanations of poverty (Jennings, 1999).  This continuing line of academic research has been abetted by the accumulation of quantitative data formulated around a technical definition of the term poverty itself, the federal identification of a level of income adjusted for family size and composition.   In 2006, an income of $20,614 or less for a family of four would place the children of the household in a condition of poverty, by these guidelines (while in 2000 it was $17,603).  Although Payne does not provide a similarly precise definition of poverty, she cites research that does, so it is a useful benchmark in assessing her ideas.  In the following analysis, we identify two levels of income:  one called poverty, which matches the federal definition, and another called low income, which ranges from just above poverty to two times the poverty level ($35, 206 in 2000 or $41,228 in 2006 for a family of four).  This second group can be thought of as representing the lower bounds of the working class, including many individuals and households that struggle to remain above the official poverty line and occasionally do fall below it.  Utilizing both of these household income categories allows us to better assess the relationship of income to a number of other factors in the lives of children who are undeniably living in a condition of poverty. 

Estimates of the number of poor children in the United States vary from year to year, given changing economic circumstances and slight differences in the sampled population.  This figure was as high as 22 percent in the early 1990s, but since 1999 it has fluctuated between 15 and 17 percent (Hernandez, 1997; Mayer, 1997; O’Hare, 2005; Child Trends Data Bank, 2007).  Table 1, located at the end of this paper, provides statistics on the percentage of children living in poor and low income households as well as racial, familial, and educational attainment variables across different geo-spatial environments in the northern United States where residency is specified in the IPUMS data.  While the regional figure for children in poverty is 15.5 percent overall, there is considerable variation between urban and suburban areas on this score.  Another 18 percent of the children in this sample lived in low income households, though this condition appears to be more consistently evident than poverty across different geo-spatial contexts.  Even so, factors of residential location and neighborhood context cannot be overlooked in assessing the causes and effects of poverty, as Payne does in her work.

The figures presented in Table 1 should not be surprising.  The suburbs, communities within metropolitan areas but outside of central cities, exhibited the lowest poverty and low-income rates among all children (7.9 percent and 13.2 percent, respectively), along with relatively high rates of parental education.  Central-city and non-metropolitan areas (largely rural and smaller cities), on the other hand, had higher rates of child poverty and lower levels of parental education.  Nearly 31 percent of central- city children lived in poverty, a rate more than four times their suburban counterparts, and an additional 23.7 percent lived in low-income households.  Indeed, more than half (57 percent) of all children with a known community type living in poverty or low income households resided in central cities even though these areas included fewer than a third of the children in the overall sample.  This is a notably high level of concentration, especially since central cities are more geographically compact than other suburban or non-metropolitan areas.  Non-metropolitan poverty rates are situated between these extremes.  As we elaborate later, a number of sociologists and other social scientists have documented the implications of these circumstances on the quality of life experienced in different settings (Brooks-Gunn, Duncan & Maritato, 1997; Danzinger & Gottschalk, 2005). 

Yet another factor of interest in Table 1 is family structure.  One of the most profound developments in the lives of American children over the past half century has been the decline in two-parent, nuclear families due to rising divorce rates and children born out of wedlock (McLanahan, 1997; Lichter & Qian, 2005).  The extent of these changes can be seen in Table 1, which provides figures on the number of children living in households without an adult male in the role of father, for example.  Again, there is considerable variation in this respect, with the highest percentage of children living in fatherless households residing in central cities (46.1 percent) and the smallest percentage, in the suburbs (18.9 percent).  The proportion of children living in fatherless households outside of metropolitan areas (20.4 percent) is just above the suburban figure, though neither measure is even half that of the urban core.  This, too, is a significant factor to consider. 

A third element of Table 1 concerns race, and our findings from these data are consistent with many readily observable social characteristics of American communities.  African American children comprised 41 percent of those living in central cities and less than 8 percent elsewhere (7.9 percent in suburbs and 1.9 in non-metropolitan areas, specifically).  Altogether, fully 72 percent of all Black children with known metropolitan status in this sample lived in the urban core, a degree of concentration consistent with longstanding patterns of racial residential discrimination in American history, at least outside of the South.  As many observers have noted, the degree of racial segregation within most major metropolitan areas in the United States has been relatively constant over the past several decades (Massey & Denton, 1993; Jargowsky, 1998; Stoll, 2005).  As the information in Table 1 attests, not many Black children live in suburban communities, and thus they cannot enjoy the benefits of growing up in circumstances which include being educated in the region’s more affluent suburban school districts. 

Finally, Table 1 also provides information on educational attainment, namely the proportion of 17-year-olds who have made it to grade 11 or beyond and are still enrolled in high school or have already graduated.  While this measure of attainment is focused on a particular age group, it does reflect the cumulative experience of a child in school.  Defined in this way, school success appears broadly related to geospatial patterns of variation in childhood poverty, racial minority status, family structure, and parental education.  Relatively high rates of child poverty in the urban core can be seen as linked to comparatively low levels of attainment, while the opposite corollary appears to be true in suburban and non-metropolitan communities.  Given the concentration of children from poverty household in central cities, it is clear that schools in these areas contend with a multitude of distinct social and economic challenges (Rothstein, 2004).

Taken as a whole, the patterns identified in Table 1 are consistent with a long line of academic research documenting the uneven prevalence of poverty across geospatial areas and its correlates that emphasize contextual factors and broad processes of societal change.  Social scientists have noted the historical patterns and enduring implications of racial discrimination in metropolitan housing (Massey & Denton, 1993).  For example, as industrial jobs have moved out of segregated core urban areas into the suburbs, African Americans and members of other racial minority groups residing in central cities have been less able to access them (Kantor, 1999; Wilson, 1987; Wilson, 1996; Jargowsky, 1998).  These challenges have grown even greater as the U.S. becomes a postindustrial society with manufacturing jobs moved overseas to capitalize on cheap labor, and domestic demands for either highly skilled knowledge workers or low paying service sector employees remaining (Bills, 2004). 

Inequality so ingrained in the structure of American society has compounding effects, as Conley (1999) and Oliver and Shapiro (1997) make evident in their distinction between income and wealth.  While persistent poverty and high rates of single parent families may contribute to what Wilson (1987) describes as concentration effects shaping attendant communal norms and values, others have also recognized the individual agency and collective resilience embodied in kinship networks poor people form themselves to provide the support necessary for survival (Stack, 1997).  The main point—and critical difference—between this extensive body of research literature and Payne’s work is the understanding that most people do not become poor or remain poor because of their individual choices or deficiencies.  These studies dispel the belief that simply changing individuals’ attitudes and behavior can substantively change their social mobility, or that children from poor and low income households have the same objective opportunities for intergenerational advancement as children from more affluent families (O’Hare, 2005).  Without reference to this wealth of understanding about how poverty occurs and why it endures, the research foundation upon which Payne’s Framework for Understanding Poverty rests seems rather impoverished itself. 

The Correlates of Poverty: A Regression Analysis

Descriptive data on patterns of social and economic variation in different geospatial contexts are broadly suggestive, but they cannot pinpoint conditions that contribute to children living in poor households.  A more comprehensive analysis of the relationship between these and other factors is required before anything confident can be said about the causes of poverty. 

Table 2 presents the results of a logistical regression analysis examining various factors associated with childhood poverty.  This is an analytical technique that calculates the likelihood, or odds, of a particular event or condition occurring within a population.  The condition in question, or the dependent variable, in this instance is the likelihood of a child residing in a household at or below the poverty level.  A somewhat larger sample comprised of all five-  to 17-year-olds in the North is used in this analysis, including those for whom metropolitan residency status has not been specified.   This affords a more comprehensive picture of the various elements contributing to a child’s poverty status. 

Since the phenomenon of poverty is quite complex and many factors are potentially involved, this analysis is divided into three parts, represented by models 1, 2 and 3 in Table 2.  The coefficients in Table 2 demonstrate the relationship of a particular characteristic (such as central city residence or family structure) with the likelihood of childhood poverty.  Positive signs on coefficients indicate that a greater likelihood of living in poverty is associated with a given factor than its comparison group, and negative signs are associated with a lower likelihood.  By and large, these coefficients can be interpreted as indicating effect sizes for the various factors examined, permitting comparison across a range of variables.[3]    

The first model in Table 2 simply considers geospatial variables in terms of their association with the odds of a child living in poverty.  Only central city and non-metropolitan residences are considered, since suburban and unspecified children constitute the comparison group.  The signs on both coefficients are positive, indicating that a child living in one of these community types has a higher likelihood of being poor.  Specifically, central-city children are four times (400%) more likely to be poor, and non-metropolitan children are about 40% more likely.  These findings underscore the concentration of poor children in the region’s larger cities, and are consistent with the descriptive analysis offered in the previous section.     

The second model adds race and ethnicity to the analysis, with a variable for being African American and another for having a Hispanic-headed household.  The comparison group is all non-Black and non-Hispanic children.  Logistic regression controls for the effects of all factors in the model, assuming that they are largely independent of one another.  Thus, it allows us to see the impact of these characteristics on the likelihood of a child being in poverty independent of other factors in the model (in this case, the geospatial environment), while also showing the independent effects of the other factors.  As we noted earlier, African American and Hispanic children are heavily concentrated in central-city neighborhoods, especially in the larger cities of the North.  While living in a central-city environment is associated with a greater likelihood of these children being poor, the magnitude of this factor’s effect is reduced substantially once the variables of race and ethnicity are introduced.   This is a consequence of racial and ethnic minority children being concentrated in urban areas.  A child’s being African American or Hispanic increases his or her likelihood of experiencing poverty by a factor of three or more (300%) in this model.   Interestingly, controlling for these factors also raises the likelihood of non-metropolitan residence resulting in poverty, perhaps because so few racial and ethnic minority students live in these areas.   Altogether, this model demonstrates how factors of geospatial location, as well as race and ethnicity, interact in association with the odds of experiencing poverty.  The combined factors in Model 2, however, account for only about 13 percent of the variance in childhood poverty in this sample, indicated in the R/2 statistic. 

Model 3 in Table 2 complicates the analysis even more by adding various household characteristics which have been discussed in sociological status attainment literature focusing on determinants of poverty:  family structure, parental educational attainment, and adult work patterns.  As suggested in the previous section, missing a parent is clearly associated with greater odds of experiencing poverty; however, if the missing parent is the father, then the child’s likelihood of being poor is twice that of any other factor in the model.  Having a parent who dropped out of high school also is associated with childhood poverty.  Since the final two variables are not categorical, they are considered covariates and provide a control for parental age and work patterns to better assess the independent effects of other factors.  Having controlled for these factors, we see that children with older parents are slightly less likely to experience poverty.  Parental work patterns are also notably associated with child poverty, as one might expect:  The more weeks household heads worked in the preceding year, the lower the odds of their children residing in poor households.

Central-city and non-metropolitan residence continued to be important in Model 3, with the odds of children living in those environments experiencing poverty between 30 and 60 percent higher than the comparison group.  Race and ethnicity also remained associated with higher odds of poverty, although the magnitude of their importance is considerably lower in Model 3 than in Model 2.  Altogether, the variables in this model accounted for over 45 percent of the variance in childhood poverty in this sample, a large portion in research of this sort.  The biggest segment of explained variance, though, is associated with the household variables added in the last step, particularly factors related to family structure and parental educational attainment.

If family structure, parental education, and a variety of other factors, including race and ethnicity, employment history, and geospatial environments, are primary characteristics associated with childhood poverty, what does this say about Ruby Payne’s conceptualization of poverty?  It may be true that much of the recent instability in family structure is related to changing American values, and living in a single parent household is strongly associated with a child’s likelihood of experiencing poverty (McLanahan, 1997; Danziger & Gottschalk, 2005).  However, childhood poverty is associated with a number of complex conditions in contemporary society.  Payne discusses very few of these factors explicitly in her writing and gives emphasis to none of them expressly. 

Model 3 in Table 2 makes clear that residing in a fatherless household is the strongest characteristic associated with childhood poverty.  Even after controlling for other factors in the model, children living in families without fathers are almost five times (500%)  more likely to be poor than children living with their fathers.  Payne and like-minded observers might interpret such findings as evidence of individual waywardness or moral corrosion on the part of the parents involved—or, certainly the fathers—who are unable or unwilling to provide appropriate support for their children.  Yet, there are alternate ways of understanding this relationship between family structure and poverty, too, with due consideration to key factors of geospatial context, race and ethnicity, and other household characteristics.

One way to test these explanations is to examine the factors linked to the chances of a child residing in a fatherless household. Table 3 presents another logistic regression assessing factors associated with the odds of a child living in such family circumstances.  This analysis is also divided into several models, with the final one presenting all of the variables under consideration (and accounting for more than 25 percent of the variance in the dependent variable).   Aside from poverty, which we have seen is correlated with family structure, the variables with the largest coefficients are being African American or Hispanic and living in a central city.  Being African American is associated with four times (400%) greater odds of living in a fatherless household than other children, an effect even more salient in statistical terms than poverty.  Living in a central city is associated with nearly 25 percent greater odds, even when the effect of other factors is controlled.  Other factors that also are associated with this condition are ethnicity (Hispanic) and having a disabled parent.  Indeed, the impact of living in a Hispanic household is greater than central-city residence.  Clearly, there are a number of factors that affect the likelihood of a child living in a single parent household besides poverty. 

By and large, these are not behavioral characteristics originating from within individual children or their families that can be changed by schools or other social organizations.  If social scientists are correct in their understanding of contemporary urban poverty described earlier, then the strong association between a child being African American and growing up in a fatherless household can be interpreted as prima-facie evidence about broad structural shifts in the economy and geospatially unequal distribution of resources in metropolitan America, resulting in what Wilson (1987) has described as “the truly disadvantaged.”  Payne says nothing of forces such as these.  Indeed, she goes so far as to argue that race is not related to class, and “one can be examined without the other” (Payne & DeVol, 2005; emphasis in the original).  Payne also admits that her own skin color precludes her from focusing on race because, “The real issue is that I am white, and there’s a huge belief out there that if you’re white, you can’t talk about poverty and race” (Shapira, 2007).  This logic underscores her notion of “expertise” derived from experience alone rather than familiarity that is further informed by careful study of research-based evidence. 

This conclusion is abetted by other factors included in Table 2.  The impact of variables such as central-city and non-metropolitan residence, along with race and ethnicity, in the analysis of poverty points to the salience of historical factors going beyond the basic behavioral explanations provided in Payne’s Framework.  The same can be said of parental educational attainment.  Even after controlling for the impact of household structure, race, and ethnicity, children living in households headed by parents without high school diplomas face more than double the odds (200%) of other children of living in poverty.  Parents in these situations have relatively little human capital to utilize in the job market.  Consequently, even with hard and steady work, they often cannot command the level of income required to provide their children with an adequate standard of living, regardless of their personal values or dispositions.  And other children are at a disadvantage because they live in households with adults who are disabled, reducing their standard of living even further.  In sum, there are a host of factors associated with childhood poverty, few of which Payne even acknowledges before attempting to address how poverty is related to educational success in school.   

Poverty and Schooling: Patterns of Success and Failure 

The point of Ruby Payne’s discussion of poverty, of course, is to recommend policies and programs to schools which she views as critical to aiding children in these circumstances.  Central to her argument is the assertion that poor children are often ill-equipped for success in school because they lack the “emotional resources” developed at home to make thoughtful life decisions based on an awareness of possible consequences.                          

In this section, we conduct a statistical analysis of the relationship of educational attainment to poverty and a number of other characteristics affecting children’s lives.  While we are limited to considering variables about which the census collected information in 2000, we believe the results demonstrate that success in school is related to a number of factors, many of which are quite independent of a child’s poverty status.  

Formal education is a major process of socialization through which children are prepared for adult roles in modern society.  Children who exhibit problematic behaviors or academic difficulties are held back to repeat grades and consequently fall behind their cohort peers, but children who master the expectations associated with this developmental progression are rewarded in most schools by passing from one grade on to the next.  In this way, educationally successful children can generally be identified as those who are enrolled in the appropriate grade for their age.  While this definition of success does not represent academic achievement or school behavior in the fine-grained manner of other studies (Pagani & Tremblay, 1997), it does provide a convenient gauge of different levels of accomplishment and advancement toward graduation.  With a dependent variable constructed in these terms of grade-level completion at a given age, logistic regressions can be used to identify factors associated with educational attainment or overall success in school. 

Since most children start kindergarten at ages five or six, they are typically enrolled in Grade 1 or higher by age seven.  Similarly, the majority of 11-year-olds are in Grade 5 or higher, and the majority of 17-year-olds are in grade 11 or higher (including high school graduation).  In each instance, over 80 percent of all children in our IPUMS data are enrolled at what is considered the appropriate level of educational attainment, suggesting that they have satisfactorily adjusted to the demands of the school system enough to be promoted to the next grade.  The question motivating our present analysis, then, is just what characteristics are associated with a child’s history of success and failure in these terms?

Table 4 presents our results, focusing on the grade and age designates mentioned above.  The independent variables are largely the same as those in Tables 2 and 3 except for the inclusion of a low income household classification, referring once again to families with an income between 100 and 200 percent of the official poverty threshold.  Technically speaking, these children are not living in poverty, even though they may not be far removed from it.  The comparison group for both the poverty status household and low income status household categories are households with incomes more than 200% of the poverty level, or families who could at a minimum be considered members of the middle class.

The results of this analytical approach reveal many of the factors that contribute to school success and failure.  The samples used are restricted to children at the appropriate age for each grade level, and the coefficients in each model reflect the likelihood of a child with a given characteristic being enrolled in the appropriate grade level or higher, controlling for all other factors in the equation.  Thus, we see that seven-year-olds in poverty households are only three-quarters as likely as their middle- class and wealthy peers to be in first grade, and those in low-income households do not fare much better.  At age 11, poor children are slightly less than half (50 percent) as likely as middle-class and wealthy children to be enrolled in grade 5 or better, while low- income children are about 70 percent as likely.  In this instance, poverty appears to be an especially substantial handicap.  The attainment distinction between low-income children and those in poverty households diminished by about half at age 17, although it is still sizeable.  And at each point, the likelihood of age-appropriate attainment or better is lowest for children living in poverty, and it is also relatively low for low-income children.  It appears that income is, indeed, an important correlate of school success, but the effects vary across the ages spanned and affect children both slightly above as well as below the poverty threshold. 

The most interesting aspects of this analysis, however, are the other factors in each equation.  Living outside of a metropolitan area is a significant handicap for children in the age seven and 11 categories, as is being African American for 11-year- olds (compared to non-Blacks).   For 17 years-olds, having a Hispanic household head or living in a fatherless household are associated with significantly lower odds of being enrolled in grade 11 or higher, as is the circumstance of having a disabled household head.  Living in a household with no mother affected all three age groups, although numerically, a relatively small group of children are identified as living in such situations.  Perhaps the most striking pattern for all children, however, is the significantly lowered odds of age-appropriate attainment for those with a head of household who dropped out of high school.  While the impact of this factor varies across age groups, its effect is robust and significant.  Indeed, it is the most important factor in two out of the three models.  The negative coefficient for this factor in the model of 17-year-olds is the most pronounced factor in the entire table.  It indicates that the odds of a high school student being at the appropriate grade level if his or her head of household dropped out is only 47 percent as high as other children in the same age group.  The effect of all these factors is net of the influence of income levels, which are controlled in this analysis, so these coefficients indicate the odds for children both in and out of poverty.   In other words, these findings suggest that there is a range of factors besides poverty that also negatively impact the odds of success in school, with parental education being perhaps the most consistently important one. 

The patterns of association identified in Table 4 underscore our prior assertion that many conditions account for American children failing to succeed in school.  While childhood poverty is clearly an important factor, Payne’s analysis takes little of this complexity into view.  The findings reported here are consistent with other studies of attainment and achievement utilizing different data sets, and the strong effects of parental education—measured by only one factor in Table 4—are especially noteworthy (Mare, 1995; Gamoran, 2001).  They also are consistent with research showing that poverty alone is not strongly linked to potentially problematic parenting practices (Hanson, McLanahan & Thompson, 1997).  Our analysis shows that living in a household headed by an individual who dropped out of high school is a major disadvantage with respect to a child’s educational attainment and comparable to the measured effects of poverty itself.  Beyond this, the models presented in Table 4 each account for just a small fraction of the statistical variance in school success, as determined by the dependent variable in this analysis.  Clearly, a more complex theory of poverty and its correlates than what Payne offers is necessary to account for why some children fail in school while others succeed.   

Admittedly, analyses like these are academic exercises conducted at such high levels of abstraction that it is difficult to understand how they relate to real children in actual schools.  The results presented in Table 4 do not disconfirm Payne’s basic assertion that poverty is an important element in the lives of many children with related negative implications for their academic success.  And the fact that school success or failure is a complex process influenced by a variety of factors in a child’s life is not a radical insight.  To gain a clearer picture of how childhood poverty and certain other factors affect educational attainment, it is necessary to examine how some of the critical variables identified in the foregoing analyses interact. 

Table 5 offers a glimpse of this by exhibiting the rate of school success attained by a sample of 17-year-olds in poverty who are affected by a number of the other conditions identified in the prior analyses.  This group is drawn from the sample above, and consists of just 1509 poor 17-year-olds, so the margin of error for each of the categories is three percent or less.  Interestingly, there is no statistically significant difference in the school success rates of poor white and Black youth, but other conditions seem to have more telling patterns of association.  For instance, poor teenagers living in fatherless households exhibit slightly lower rates of school success than those with a father present, as one would expect, given the strong association between poverty and fatherlessness noted earlier.  The rate of fatherless households in this sample (70 percent) is quite high.  Similarly, the school failure rate is greater for poor teens living in central cities than those in the suburbs, though the differences are not as significant as one might have expected.  The most striking pattern noted in the table, however, is the difference that a household head’s educational attainment makes: Over one-third of poor teenagers whose household heads dropped out of high school experience school failure themselves, compared to 21.4 percent of teenagers whose household heads graduated.   These patterns of school success and failure highlight the importance of parental education as a factor impacting the lives of children residing in poverty households.  Poverty is shown in this way to be a greater handicap to educational success when it interacts with certain other factors, substantially increasing the likelihood of school failure.  Taking stock of parents’ educational backgrounds may provide critical clues to the challenges that children in poverty face in their struggle for educational success.  Again, this is not addressed in Payne’s analysis.   

Implications and Closing Discussion

Our critique of Payne’s work is not intended to suggest that we disagree with her on all counts.  For example, we share in the belief that poverty is a serious issue affecting many children’s overall quality of life and educational success.  We also agree that certain middle-class orientations are normalized in the values, practices, and expectations of many teachers and the school environments where they work, to the point that such perspectives seem invisible or taken for granted ways of being.  As Lisa Delpit (1995) points out in the title of her book, Other People’s Children, we should be committed to addressing inequities like these in the lives of all children as if they were our very own offspring.  This means continuing to accurately understand the phenomenon of poverty in order to more effectively address it, raising awareness amongst educators about class bias and its effects, as well as providing children who are systematically disenfranchised with explicit access to valuable forms of human, social, and cultural capital.  

Yet, the differences in our perspective and Payne’s should not be minimized.  Whereas Payne explains poverty primarily as the result of individuals’ lacking important emotional resources to change their actions—and thus deflect the ensuing consequences of those actions which result in their impoverishment—we emphasize the systematically inequitable distribution of opportunities, advantages, and material goods across groups of people in society that are correlated with poverty.  These individual and structural perspectives are not absolute or mutually exclusive characterizations of poverty, but adopting the former view yields relatively simple solutions compared to the more comprehensive scope of the latter perspective.  We believe that existing social science research not only warrants but necessitates this more complicated approach to understanding poverty (Brooks-Gunn, Duncan & Maritato, 1997).  While statistical correlations are not necessarily indicative of causal relationships, it is important not to neglect the clear patterns of association revealed by the analysis herein, as well as in the work of others we have cited.

Payne’s self-proclaimed expert status to speak on poverty is a particular challenge for collaboratively advancing the conversation underway between educational practitioners, policy makers, and researchers.  Although expertise may be derived from more than just conducting scholarly research and following defined academic protocols, such professional standards help ensure certain levels of rigor within particular discussions, and also in gathering the basic information required to compare or replicate studies that collectively might benefit the field.  In its current form, Payne has framed an explanation and a conversation about poverty in terms that cannot be engaged by others, but has significant implications for both theory and practice in education. 

For example, Payne distinguishes between those who experience situational and generational poverty.  She does not reference any published research to support this categorical distinction, however, only casually specifying that generational poverty is a condition affecting two or more generations of a family (2005, p. 10).  Not only are more clear and measurable definitions needed for researchers to identify and study these groups with available data, but very few existing studies of poverty in fact take place across generations and include large, scientifically drawn samples.[4]  While Payne argues it is particularly challenging to involve the generationally poor in educational endeavors and socially appropriate behavior, it is difficult to discern whether or not substantive differences actually do exist between those who are situationally and generationally poor.  It is even hard to know how many children constitute these groups.

If children who experience persistent poverty pose the biggest challenge to educators, the information in this article can provide a useful addendum—or, in certain respects, a corrective—to Payne’s work.  We have identified many of the ways that poverty is associated with other facets of children’s lives, but there remains the question of the children facing the greatest disadvantages, those most likely to experience enduring poverty.   Although we do not have longitudinal data to measure persistence, it is possible to identify children experiencing perhaps the greatest obstacles: those living in poverty with a single parent (usually the mother) who is not a high school graduate.  Children in these multiple disadvantaging circumstances face the highest odds of school failure, as they are raised by adults struggling to simultaneously provide childcare and maintain jobs without the benefits of a high school diploma.  We believe that this group is the most likely to represent what Payne describes as generational poverty, and studies have amply documented that the prospect of escaping poverty in such circumstances is quite remote regardless of individual effort (Corcoran & Adams, 1997; Mayer, 1997).

Payne discusses generational poverty as a broad and growing dilemma, but how extensive a problem do children in these circumstances represent?  Among 17-year-olds in the North, individuals with all of these characteristics comprise less than a quarter of those in poverty, and just three  percent of the total regional sample, a figure that holds for children of other ages as well.  They are also heavily concentrated in urban areas, with more than two-thirds of them living in central cities.  More than half are African American.  If this is the group of children Payne refers to when she discusses generational poverty, it seems to have important implications for her rationales about why poverty exists and her recommendations about how to combat its influence.  The geo-spatial location of childhood poverty and its implications for education and a variety of other questions is an issue that social scientists have examined in some depth (Brooks-Gunn, Duncan, Leventhal & Aber, 1997).  Unfortunately, Payne provides no indications about the social characteristics of children experiencing generational poverty or where they are most likely to be found, even though this is the group she claims needs the greatest help.  The data we have provided suggest that it is a very particular group, largely concentrated in certain types of communities.  Payne does not address this, and it is among the most glaring problems in her analysis.

The foregoing points to a big question that Payne’s characterization of poverty does not confront: the geospatial dimension of inequality discussed in this article and by other social scientists.  To the extent that expertise is built upon understanding and extending what is already known about a topic, Payne’s Framework would be stronger for acknowledging the varied effects of poverty in different community contexts.  The indisputably high concentrations of poor children in central cities are also those who most likely live in neighborhoods served by schools with less experienced teachers and fewer resources for instructional support services and technology.  The fact that poor children are not randomly distributed in geographical terms points to the social and economic forces that produce and reproduce contemporary patterns of poverty in the U.S.  Given this, Payne’s professional development materials could be made more meaningful by attempting to raise the awareness of participants about how poverty may function differently across different geospatial settings, a key insight for both educational practitioners and policy makers. 

Finally, Payne relays numerous anecdotes of inappropriate and often disturbing behavior patterns exhibited by poor children and their families in her work, as well as examples of the values, language, and related worldviews they embody that appear radically removed from mainstream America.  If it is true that factors like these directly result in long-term poverty and school failure (or, in a best case scenario, the elimination of poverty and ready educational success), as she claims, then more valid and reliable evidence is needed to support her assertion and reasoning.  Our analysis, however, demonstrates statistical associations of varying strengths between children’s educational success and a host of different circumstances impacting their lives.  Poverty itself is a serious issue, no doubt.  Its lone impact may not be as significant as other factors, though, and it often works in conjunction with other disadvantaging variables.  Only such nuanced approaches to exploring poverty and school achievement can help distinguish between stereotypical generalizations versus evidence-based facts, or flawed logic versus systematically established relationships.  

Educators seeking to further their understandings about poverty and its relation to family, classroom interactions, and academic achievement in such a manner might benefit by consulting the work of Shirley Brice Heath (1983), Sue Books (2004), Annette Lareau (2003), and Richard Rothstein (2004), for example.  Their work and the collective scholarship of others cited herein make clear that poverty is a complex phenomenon about which we know much but should continue to explore even more fully.  It is our sincere hope that Payne’s recent popularity will compel all interested individuals and organizations to direct their future efforts toward utilizing the most thorough and trustworthy knowledge of poverty available to date. 


Table 1: Descriptive Statistics, Children ages 5 to 17, Northern U.S., 2000 (IPUMS data)
(Children in non-group-quarters, with known metro status, n=130169)*

Metropolitan Status

Children in Poverty


Children in Low Income Households

Black Children

No Father in Household

Household Head a HS Dropout

17 yr-olds in Junior year +








Urban Core



41  %











Region Total


18  %


27  %







*Metro status “unknown” or “not applicable” similar to regional total figures

Table 2: Logistic Regression, Factors Associated with Poverty,
Children ages 5 to 17, Northern U.S., 2000 (IPUMS data


Model 1

Model 2

Model 3

Odds Ratio

Non-Metropolitan Residence





Central City Residence





Child is African American





Hispanic Household Head





No Father in Household





No Mother in Household





Household Head Not H.S. Graduate





Household Head Disabled





Age of Household Head (covariate)





Weeks Worked in Past Year by Household Head (covariate)





Nagelkerke R Square





Dependent variable: Odds of child living in household at or below poverty level
(Children in non-group-quarters), all variables significant at .000 level

Table 3:  Logistic Regression, Factors Associated with Fatherless Household,
Children ages 5 to 17, Northern U.S., 2000, (IPUMS data


Model 1

Model 2

Model 3

Odds Ratio

Non-Metropolitan Residence





Central City Residence





Child is African American





Hispanic Household Head





Poverty Status





Household Head Not H.S. Graduate





Household Head Disabled





Age of Household Head (covariate)





Weeks Worked in Past Year by Household Head (covariate)





Nagelkerke R Square




Dependent Variable: Odds of living in household with no father
* not significant at .05 level, all other variables significant at .000 level

Table 4: Logistic Regression, Factors Associated with School Attainment,
Children ages 7, 11 & 17, Northern U.S., 2000 (IPUMS data) 


Age 7 (Grade 1)

Age 11 (Grade 5)

Age 17 (Grade 11)

Odds Ratio (for age 17)

Non-Metropolitan Residence





Central City Residence





Child is African American





Hispanic Household Head





No Father in Household





No Mother in Household





Low Income Household





Poverty Status Household





Household Head Not H.S. Graduate





Household Head Disabled





Age of Household Head (covariate)





Weeks Worked in Past Year by Household Head (covariate)





Nagelkerke R Square





N (Children in non-group-quarters)




Dependent variable: Odds of child attending grade 1 (age 7), 5 (age 11), or 11 (age 17)
*   Significant at .05 level, ** Significant at .005 level

Table 5: Success in School, Youth Living in Poverty,
Under Different Conditions, Northern States,
IPUMS Data, 2000


Percent, Below Grade 11 or Dropped Out

Percent, In Grade 11 or Higher


26.3% (189)

73.8% (531)


25.3% (113)

74.7% (334)




Central City

26.9% (170)

73.1% (463)


22.0%   (70)

78.0% (248)




Fatherless Household

28.1% (300)

71.9% (767)

Father in Household

23.5% (104)

76.5% (338)




Household Head HS Grad

21.4% (197)

78.6% (724)

Household Head HS Dropout

35.2% (207)

64.8% (381)

n= 1509 (Children in non-group-quarters) (n of each cell)


Baker, B., Ng, J. & Rury, J.  (2006, June 7).  Questioning a speaker’s knowledge of poverty.  Education Week, 36.

Bills, D. B.  (2004). The sociology of education and work.  Malden:  Blackwell Publishing.  

Bohn, A. (2006, Winter).  A framework for understanding Ruby Payne.  Rethinking Schools, 13-15.

Bomer, R., Dworin, J. E., May, L., and Semingson, P. (2008).  Miseducating teachers about the poor:  A critical analysis of Ruby Payne’s claims about poverty.  ID Number 14589, Teachers College Record.  Retrieved from http://tcrecord.org.

Books, S.  (2004). Poverty and schooling in the U.S.:  Contexts and consequences. Mahwah:  Lawrence Erlbaum.

Brooks-Gunn, J., Duncan, G., Leventhal, T. and Aber, J. L. (1997). Lessons learned and future directions for research on the neighborhoods in which children live. In J. Brooks-Gunn, G. Duncan, and J. L. Aber (Eds.)
Neighborhood poverty: Context and consequences for children (Volume 1)
(pp. 279-298). New York: Russell Sage.

Brooks-Gunn, J., Duncan, G. J., & Maritato, N. (1997). Poor families, poor outcomes: The well being of children and youth. In G. J. Duncan and J. Brooks-Gunn (Eds.), Consequences of growing up poor (pp. 1–17). New York: Russell Sage.

Byrne, D. P. (2008). Keep the kids in school. Chicago Tribune, August 12, p. 17. 

Child Trends Databank, (2007) http://www.childtrendsdatabank.org/.  

Conley, D.  (1999). Being black, living in the red:  Race, wealth, and social policy in America.  Berkeley, CA:  University of California Press.

Corcoran, M. and Adams, T. (1997). Race, sex and the intergenerational transmission of poverty. In G. J. Duncan and J. Brooks-Gunn (Eds.), Consequences of growing up poor (pp. 461–517). New York: Russell Sage.

Danziger, S. and Gottschalk, P. (2005).  Diverging fortunes: Trends in poverty and inequality.  In R. Farley and J. Haaga (Eds.), The American people: Census 2000 (pp. 49-75). New York: Russell Sage.

Delpit, L. (2006). Other people's children: Cultural conflict in the classroom (updated edition). New York: The New Press.

Gamoran, A. (2001). American schooling and educational inequality: A forecast for the 21st century.  Sociology of Education (extra issue), 135-153. 

Gans, H. J.  (1995). The war against the poor:  The underclass and antipoverty policy.  New York:  Basic Books

Gorski, P. (2006a). The classist underpinnings of Ruby Payne’s Framework.  ID Number 12322, Teachers College Record.  Retrieved from http://www.tcrecord.org.

Gorski, P. (2006b).  Savage unrealities:  Classism and racism abound in Ruby Payne’s Framework.  Rethinking Schools,16-19.

Hanson, T. L., McLanahan, S., and Thomson, E. (1997). Economic resources, parental practices, and children’s well-being.  In G. J. Duncan and J. Brooks-Gunn (Eds.), Consequences of growing up poor (pp. 190–238). New York: Russell Sage.

Heath, S. B. (1983). Ways with words: Language, life and work in communities and classrooms. New York: Cambridge University Press

Hernandez, D. J. (1997). Poverty trends. In G. J. Duncan & J. Brooks-Gunn (Eds.), Consequences of growing up poor (pp.18–34). New York: Russell Sage.

Jackson, S. R. (2008). When education is unequal. Chicago Tribune, August 22, p. 25. 

Jargowsky, P. A. (1998). Poverty and place: Ghettos, barrios, and the American city. New York: Russell Sage.

Jennings, J. (1999). Persistent poverty in the United States:  Review of theories and explanations.  In L. Kushnick and J. Jennings (Eds.), A new introduction to poverty:  The role of race, power, and politics (p. 13-38).  New York:  New York University Press.

Kantor, H. (1999). Race, education, and joblessness in the inner city, 1970-1990.  The Urban Review, 31 (3), 225-242.

Keller, B. (2006, May 3).  Payne’s pursuit.  Education Week, p. 30-32.

Lareau, A.  (2003). Unequal childhoods: Class, race, and family life. Berkeley, CA: ]\        University of California Press. 

Lichter, D. T. and Qian, Z. (2005). Marriage and family in a multiracial society. In R. Farley and J. Haaga (Eds.), The American people: Census 2000 (pp. 169-200). New York: Russell Sage. 

Mare, R. D. (1995). Changes in educational attainment and school enrollment. In R. Farley, (Ed.), State of the union: America in the 1990s. Volume one: Economic trends (pp. 155-214). New York: Russell Sage. 

Massey, D. S. and Denton, N. A. (1993).  American apartheid:  Segregation and the making of the underclass.  Cambridge, MA:  Harvard University Press.

Mayer, S. (1997). What money can't buy: Family income and children's life chances. Cambridge, MA: Harvard University Press.

McLanahan, S. (1997). Parent absence or poverty: Which matters most?  In G. J. Duncan and J. Brooks-Gunn (Eds.), Consequences of growing up poor (pp. 35-48). New York: Russell Sage.

Ng, J. C. & Rury, J. L.  (2006). Poverty and education:  A critical analysis of the Ruby Payne phenomenon.  ID Number:  12596, Teachers College Record.  Retrieved from http://www.tcrecord.org.

O’Hare, W. P. (2005). Trends in the well being of America’s children. In R. Farley and J. Haaga (Eds.), The American people: Census 2000 (pp. 201-223). New York: Russell Sage. 

Oliver, M. L. and Shapiro, T. M.  (1997). Black wealth/white wealth:  A new perspective on racial inequality.  New York:  Routledge. 

Osei-Kofi, N.  (2005). Pathologizing the poor:  A framework for understanding Ruby Payne’s work.  Equity and Excellence, 38 (4), 367-375.

Pagani, L., Boulerice, B. and Tremblay, R. (1997). The influence of poverty on children’s classroom placement and behavior problems. In G. J. Duncan & J. Brooks-Gunn (Eds.), Consequences of growing up poor (pp. 311–339). New York: Russell Sage.

Pampel, F. C. (2000).  Logistic regression: A primer. Thousand Oaks, CA: Sage Publications.

Payne, R. K.  (2005). A framework for understanding poverty.  Highlands, TX:  aha! Process, Inc.

Payne, R. K. & DeVol, P. (2005). Toward a deeper understanding of issues surrounding poverty:  A response to critiques of A Framework for Understanding Poverty.  aha! Process, Inc.  Retrieved from www.ahaprocess.com

Rothstein, R. (2004). Class and schools: Using social, economic, and educational reform to close the Black-White achievement gap. Washington and New York: Economic Policy Institute

Shapira, I. (2007, April 15). Author’s poverty views disputed yet utilized.  The Washington Post, pp. A01.

Stack, C.  (1997). All our kin:  Strategies for survival in a Black community.  New York:  Basic Books.

Stoll, M. A. (2005). African Americans and the color line. In R. Farley and J. Haaga (Eds.), The American people: Census 2000 (pp. 380-414). New York: Russell Sage.

University of Minnesota Population Center (n.d.) Integrated Public Use Microdata Samples (IPUMS). http://usa.ipums.org/usa/intro.shtml.

Wilson, W. J. (1987). The truly disadvantaged:  The inner city, the underclass, and public policy.  Chicago, IL:  University of Chicago Press.

Wilson, W. J. (1996). When work disappears:  The world of the new urban poor.  New York:  Alfred A. Knopf.


[1] For information on the IPUMS data, see http://usa.ipums.org/usa/intro.shtml.  We are grateful to the Minnesota Population Center for making these data available to us and to other researchers. 

[2] Information on the metropolitan status has been withheld from IPUMS for some households in this sample for purposes of confidentiality.  Because of the size and complexity of metropolitan areas in the region we have selected for study, there are fewer of these cases, some 21 percent, than in other parts of the country.  The region in question is comprised of the following census regions (defined by the Census Bureau): New England, Middle Atlantic and East North Central.  The census region with the lowest number of cases with incomplete information on metropolitan status in 2000 is the Middle Atlantic, with only 12 percent having information withheld.  In comparing the distribution of cases in the Middle Atlantic region and the larger area utilized in this study, we found very little difference between the two.  Because the cases in the incomplete metropolitan status category approximate the regional means for most of the variables considered in this study, we do not believe that their exclusion represents a source of significant bias in the analysis that follows.  For discussion of sampling procedures in the 2000 IPUMS data, see: http://usa.ipums.org/usa/chapter2/chapter2.shtml#2000.  

[3] Logistic regression is used in this instance because the dependent variable is binary, representing a choice of two values.  In this instance, it is a matter of determining whether a household’s income places it below or above the poverty level, using the federal definition of poverty for various household sizes.  Coefficients are logarithmic expressions of the odds, so negative values are less than 1 (even odds) and positive values are greater than 1, in comparison to a specified reference group.   The size of the coefficient indicates greater or lesser odds in either direction, depending on the sign.  In Table 2, for instance, the first factor in the table, Non-Metropolitan Status, has positive coefficients, indicating that it is associated with higher odds of a child being in poverty than its comparison group, which in this instance is suburban and unidentified metro status children.  As most of the factors in Table 2 have positive coefficients, they are associated with higher odds of being in poverty.  The numerical size of the coefficients is an indication of the magnitude of the odds.  The final column in the table lists the odds ratio for the last equation in the analysis.  As indicated earlier, odds greater than one reflect an increased likelihood of an event occurring.   Thus we see that a child living in a non-metropolitan setting in this sample is about 58  percent more likely to be living in a poverty household than a child in the reference group (odds ratio=1.58).  Odds that are multiples of 1 indicate an even greater likelihood of an event occurring.  The odds of a child living in a household with no father of being in poverty (4.928), for instance, is more than four times (400%) that of a child in the comparison group, which is those in two-parent families.  This is a very great order of magnitude, and relatively rare in the social sciences.  Like other forms of regression, this technique holds the effects of other variables in an equation constant when calculating such odds. Thus the pattern of association exhibited between fatherlessness and poverty is independent of other factors in the analysis. For further discussion of logistic regression, see Pampel (2000)

[4] In one of the few analyses to examine this question, Mary Corcoran and Terry Adams have highlighted the significance of race as a factor in the inter-generational transmission of poverty.  Their research also provides considerable support for the geo-spatial arguments of Wilson, Jargowsky and others who argue that analysis of poverty and its effects must be considered in the context of specific community contexts.  See Corcoran and Adams, (1997).