Leif Jensen and Danielle Ely
INTRODUCTION
Poverty is a social problem that is both universally recognized and personally experienced by all too many people in rural and urban areas. One estimate suggests that upwards of two-thirds of all Americans will experience poverty by the time they reach age eighty-five (Rank and Hirschl 1999). Poverty seems familiar to us all, and for many it is a lived experience. It touches all of our lives in one way or another. Whether it conjures up mental images of ramshackle housing tucked away in rural hollows or empty storefronts along blighted inner city neighborhoods, people think they know poverty when they see it. And most people have opinions—often strong opinions—about what poverty is, what causes it, and what, if anything, needs to be done to combat it.
Being ubiquitous, complex, and contentious, poverty is precisely the kind of social problem that first needs to be carefully defined before we can hope to pursue a rigorous and unbiased understanding. Like any measurement in the social sciences, a worthwhile measure of poverty needs to be both valid and reliable. Validity means that the measure of poverty truly captures the underlying concept it is intended to measure. “Are we measuring what we think we’re measuring?” is the way it is often phrased. Reliability means that the measure of poverty operates in a consistent way over time and is relatively free of random error. A yardstick made from flexible material is not very reliable.
A standard operational definition of poverty—one that is both valid and reliable—serves two basic and critical purposes, one scientific and one political. First, it allows researchers to assess the prevalence of poverty through the calculation of poverty rates: the percentage of people or families that are in poverty. Knowing how poverty rates vary across sociodemographic groups, across space, and over time is essential for a full accounting of social inequality. That there are certain categories of people (e.g., recent immigrants or high school dropouts) or certain kinds of places (e.g., rural localities or counties in Appalachia) with high poverty rates invites scientific questions about why this is so. Assessing trends in the poverty rates for people and places over time likewise leads to intense study of the factors causing poverty rates to change. At a more micro level, analyzing survey data on individuals or families that include a variable for poverty status allows for complex statistical analyses of the correlates of poverty. If the data are collected for the same individuals at multiple points in time, social scientists can better understand the factors that give rise to the movement of people into and out of poverty over time. In short, a valid and reliable measure of poverty allows social scientists to confidently explore the complex array of factors that cause poverty, as well as its severity, nature, and consequences.
Second, a standard measure of poverty is important for social policy as well. Although the scientific study of the causes and consequences of poverty has scholarly value in and of itself, it also provides policy makers and analysts with solid information about which people and places face the greatest poverty risks and are in greatest need of attention. At the same time, by providing a dependable understanding of the causes of poverty, how the causes differ in strength, and under what circumstances they are more problematic, social science can contribute to more effective public policy. Trends in poverty rates can provide some evidence about whether or not policies are effective. Finally, a valid and reliable measure of poverty allows for policy targeting. That is, it provides an arguably fair and standardized way to assess what people or places are deserving of and qualified for various means-tested government programs.
Despite the clear value of having a standard definition of poverty, in reality, there is a wide variety of ways to define poverty. This chapter describes some of the basic definitions of poverty, with emphasis on the official U.S. definition of poverty and its strengths and weaknesses. Then, using various definitions, a brief descriptive portrait of poverty in the United States is presented that focuses on how patterns of rural and urban difference are sensitive to the definition of poverty.
ABSOLUTE DEFINITIONS OF POVERTY
Poverty can be defined in absolute or relative terms. As the term implies, absolute poverty refers to a situation in which individuals lack the income or resources needed to maintain even the most basic, minimally sufficient, subsistence-level standard of living (Iceland 2013). In practice, the amount of money needed to attain (purchase) this minimum standard constitutes the absolute poverty threshold (or poverty line), and those with incomes below this amount are defined as poor.
The official definition of poverty in the United States is an absolute measure. It was developed in the early 1960s by Mollie Orshansky of the Social Security Administration and, with minor adjustments, has been with us ever since (Ruggles 1990). Its logic is as follows. Recognizing that the most basic of needs is food, the official definition rests on the cost of a minimally adequate diet—sometimes referred to as a “food basket”—as determined by the U.S. Department of Agriculture (USDA). Although the USDA specified alternative food baskets that differed by ampleness, it is noteworthy that the food plan chosen as the basis for the official measure of poverty was the so-called economy food plan, which was designed for temporary or emergency use (Carlson et al. 2007) and, arguably, is nutritionally inadequate in the long run. Mid-1950s survey research on the U.S. population suggested that families spend about one-third of their income on food, so the cost of the food basket was simply multiplied by three to arrive at the poverty threshold. Different food baskets were specified for families of varying size and composition (i.e., different levels of need), so multiple poverty thresholds could be established. A process known as “equivalence scaling” allowed researchers to specify poverty lines that adjusted for family size, number of children, whether the family head was elderly, and the like. The official thresholds are expressed in annual dollars. They represent the annual pretax cash income from all sources needed to achieve the absolute minimum standard of living; families with annual incomes below this level are defined as poor. Families—people living together who are related by blood, marriage, or adoption—are considered to be the prime income-sharing unit for establishing poverty thresholds. People living alone or with unrelated individuals are essentially regarded as one-person families and assigned poverty thresholds accordingly.
In its original formulation, lower poverty thresholds were assigned to farm families who would presumably produce some of their own food, but this distinction has since been dropped. Apart from that, the only notable change to the original thresholds has been to adjust them for inflation, using the consumer price index (CPI). In this sense, the definition remains constant over time, which is another distinguishing feature of absolute definitions (Iceland 2013). In 2016, a family of four with two adults and two children was defined as poor if their annual pretax income was less than $24,339.
Budget for a Family at the Poverty Level
See if you can sketch out a detailed family budget for this family of four with earnings of $2,028 per month, given the prices of necessities. Imagine where this family of four is living, and under what circumstances (e.g., employment status, housing, transportation, and so forth). Itemize all of the various expense categories, estimate the monthly cost of each, and sum these expenses for a monthly total.
Can your family “make it” at that income level? What assumptions did you need to make? If you are doing this exercise in a class, how do the budgets differ? What agreements or disagreements are there over necessities and their cost? How do the imagined circumstances of the family differ, and what does this say about what makes it easier or harder to make ends meet? Is the official poverty threshold too high, too low, or about right?
You can get the current poverty threshold from the U.S. Census Bureau, or you can adjust the 2016 threshold used here for inflation using the CPI (which you can look up) or an inflation calculator.
The official definition of poverty in the United States has given us a reasonably valid and reliable way to measure poverty, and as such it has served many research and policy functions. However, it has not been without its critics. In the mid-1990s, a landmark National Academy of Sciences (NAS) study provided an in-depth critical appraisal of the official definition of poverty in the United States (Citro and Michael 1995). The study criticized the official measure for failing to count noncash in-kind income (e.g., food stamps or subsidized housing) toward a family’s total income, for having too restrictive a definition of the income-sharing “family” unit (e.g., not counting unrelated foster children as part of the family), for completely ignoring cost-of-living differences across space, for neglecting the disproportionate increases in the cost of nonfood necessities, for equivalence scaling that was somewhat crude and ad hoc, and for other reasons. Accordingly, the NAS study proposed an alternative poverty measure that would correct as much as possible for these shortcomings. That study took place in the mid-1990s, and it is symptomatic of the inertia of official government indicators that now, a generation later, the official definition of poverty has not changed. Fortunately, in recent years the U.S. Census Bureau has begun to collect the data needed to calculate a new poverty measure based on the NAS critique, the so-called supplemental poverty measure (SPM) (Short 2015). Comparisons of rural and urban poverty rates using the official measure and the SPM are examined later in the chapter.
RELATIVE POVERTY AND OTHER DEFINITIONS
Absolute poverty exists when income is insufficient to achieve a minimally adequate standard of living. If an individual or family has an annual income just a few dollars above their poverty threshold, they are not defined as poor in absolute terms. However, as the previous exercise might have indicated, living at or just above the poverty line is often associated with having a relatively deprived standard of living. Beyond having just enough income to barely eke out a living, is having an income that is much less than average meaningful? Those who define poverty in relative terms think that it is, arguing that this relative deprivation is associated with various forms of social exclusion from mainstream society and institutions (Iceland 2013). Specific definitions of relative poverty abound, but a typical relative definition is having an annual household income that is less than one-half of the median annual household income. In 2015, the median (fiftieth percentile) U.S. household income was $55,775 (Posey 2016); the relative poverty threshold is computed as half of that, or $27,888. Absolute and relative poverty rates are compared in more detail later in the chapter.
The United States officially defines poverty in absolute terms, whereas in Europe relative poverty is emphasized. It is instructive to think about what is needed to bring about improvement in absolute versus relative poverty. In general, reductions in absolute poverty require a lifting up of the real (i.e., inflation-adjusted) incomes of those below the poverty line such that proportionately fewer people or families are below the line. By contrast, reductions in relative poverty require that the incomes of those at the bottom of the income distribution move closer to those in the middle. In other words, the latter requires that income be distributed more equitably. Why the United States clings to an absolute definition of poverty and European governments focus on relative poverty is a fair question to consider.
A criticism of both absolute and relative poverty measures is that by comparing income to a poverty threshold, they neglect how people spend that income. Some people may have incomes well above the poverty line but not be spending that income in ways that contribute to physical or social well-being. Others may have incomes below poverty but be living an ascetic, healthy, and fulfilling life and may not feel poor at all. Income-based measures of poverty capture neither actual lifestyles nor subjective self-assessments of poverty status. Another way to measure poverty uses a consumption-based poverty measure, which takes advantage of information gathered in surveys on detailed household expenditures—what they are actually spending—that can then be summed and compared to an absolute poverty threshold (Meyer and Sullivan 2012). An advantage of this approach is that in U.S. household surveys there tends to be less underreporting of expenditures than of income. A related approach commonly applied to the measurement of poverty in the developing world is to examine not the household expenditures on items but the deprivations a household endures. For their multidimensional poverty index, Sabina Alkire and colleagues (2015) use ten deprivations commonly captured in national-level socioeconomic and health surveys within the three domains of health (e.g., undernourishment), education (e.g., no one in the household with at least primary schooling), and standard of living (e.g., no electricity). Households reporting over a set number of deprivations are then defined as poor. A disadvantage of this technique is that the list of deprivations is specified in advance by researchers, thereby disregarding differences across cultures and countries in what constitutes a deprivation. The next approach to poverty measurement contends with this problem.
A promising consumption-based technique is the consensual or democratic approach to poverty measurement. As described by Shailen Nandy and colleagues (Nandy and Pomati 2015; Nandy and Main 2015), the consensual approach involves a two-step process. First, a representative survey of a population is conducted to assess popular attitudes about what consumption items (e.g., three meals a day) or even activities (e.g., celebrations on special occasions) are considered necessary by a majority of respondents. Second, another survey is conducted in which respondents identify whether they have each of these perceived necessities and, if not, whether this is because the household could not afford them. Necessities that cannot be afforded are regarded as deprivations. Researchers then define as poor those households that exceed some number of these deprivations. The consensual approach is flexible and has been adapted to study poverty among children in the United Kingdom (Nandy and Main 2015) and adults in developing countries (Nandy and Pomati 2015). The method, however, is data intensive. Unfortunately, we know of no applications of this approach in the United States.
The consensual approach just described takes advantage of subjective assessments among populations regarding the necessities of life. This brings to mind the idea of subjective poverty. Two approaches to subjective poverty deserve brief mention. One is to ask survey questions such as “living where you do now and meeting the expenses you consider necessary, what would be the very smallest income you and your family would need to make ends meet?” (Danziger et al. 1984 as cited in Ruggles 1990). Answers to this question along with respondent characteristics can then be used to specify subjective poverty thresholds, and families with incomes below these thresholds are defined as poor. A second and admittedly crude approach to subjective poverty is simply to ask survey respondents whether they consider themselves to be poor. The sensitivities and ambiguities of asking such a question explain why researchers rarely use this measure.
Many social science concepts familiar to almost everyone, such as poverty, are deceptively difficult to define and measure. A valid and reliable measure of poverty is much needed to achieve a sufficient understanding of the depth, breadth, causes, and consequences of poverty, as well as rigorous and fair policy implementation and evaluation. A range of measurement approaches including absolute, relative, consumption-based, and subjective poverty have been reviewed here. These empirical measures differ in their official status, popularity, and the underlying conceptualization of poverty they capture. Some of these measures are used in the analysis that follows, but students of poverty need to be fully aware of all of them.
A PORTRAIT OF RURAL POVERTY TODAY USING ALTERNATIVE DEFINITIONS
INTRODUCTION AND OVERVIEW
The purpose of this section is to present a statistical portrait of the prevalence and correlates of poverty in both rural and urban America using several of the measures of poverty already defined: the official measure of poverty (and variants of it to capture “deep” and “near” poverty); the supplemental poverty measure, which corrects for some of the shortcomings in the official measure; a simple indicator of relative poverty; and a quasi-subjective measure of poverty. The fundamental question is, “Who are the rural poor?” In addressing this question, it is important to pay attention to the implications of how the picture differs depending on the poverty measure used. The analysis is supplemented with conclusions drawn from the empirical literature on rural poverty in the United States.
DATA AND MEASURES
The principal data set used for analysis is the March 2015 Current Population Survey (CPS). Collected by the U.S. Census Bureau, the CPS is the nation’s key source of unemployment statistics. Based on a nationally representative sample of U.S. households, the survey is conducted monthly. The CPS questionnaire contains a core set of questions that are common to each month, but in any given month the questionnaire includes supplemental questions that cover topics such as child support, volunteering, health insurance, and food insecurity. In addition to unemployment statistics then, the purpose of the CPS is to provide reliable information on the social and economic circumstances and trends among the U.S. population.
The March CPS includes the so-called Annual Social and Economic Supplement (ASEC), which contains detailed questions about income sources and amounts. The ASEC provides the basis for annual estimates of the size and composition of the U.S. poverty population. The CPS is designed to be representative of the entire noninstitutionalized civilian population. Not included are institutionalized individuals and those who are homeless and not in a shelter. Students in dormitories are only included if their information was reported in their parent’s home (DeNavas-Walt and Proctor 2015). The ASEC, unlike the basic CPS, includes some individuals in the military, but only if they live in a household with at least one other civilian adult (DeNavas-Walt and Proctor 2015). The ASEC underwent a partial redesign in 2014 and transitioned completely to the new design for the 2015 data release.1
As noted, several measures of poverty were used in the analysis.2 In addition, measures of deep and near poverty were used in the analysis. Deep poverty is defined as occurring when a family’s income is less than half its poverty threshold (i.e., the income-to-needs ratio is less than 0.5). The near poor are those with incomes less than 150 percent of the official thresholds (i.e., the income-to-needs ratio of less than 1.5).
For comparative purposes, a simple measure of relative poverty was also computed. Based on 2014 median household income for family households and nonfamily households as reported in DeNavas-Walt and Proctor (2015), the relative poverty thresholds were set at one-half of these medians. Thus, individuals were defined as being in relative poverty if their total household income was less than $34,213 for those in family households, and $16,024 for those in nonfamily households (people living alone or with unrelated individuals). This relative poverty measure is not adjusted for family size or composition as is the official absolute measure.
Finally, the SPM3 was also used in the analysis. The SPM was created in response to the legitimate criticisms of the official poverty measure. Whereas the official measure defines a family as those living together who are related by blood, marriage, or adoption, the SPM broadens this definition to include any unrelated children cared for by a family (e.g., foster children) as well as those who are in cohabiting relationships (see Short 2015 for more detail). Instead of being based on three times the cost of a minimally adequate diet, the SPM thresholds take account of actual expenditures on necessities (food, clothing, shelter, and utilities). The SPM poverty threshold is set at 1.2 times the mean (average) total expenditures on these necessities for those two-child families that fall between the thirtieth and thirty-sixth percentiles on such spending (based on consumer expenditure data). (The mean is multiplied by 1.2 to give families extra for other necessities.)4 Whereas the official definition of poverty uses all cash income before taxes, the SPM adds in some noncash income but takes away certain expenses. Specifically, the SPM counts as income the value of food stamp (SNAP) benefits and other in-kind (noncash) resources, but it also subtracts from income things like taxes paid, out of pocket medical expenses, employment expenses, and child support. The SPM poverty thresholds differ by prevailing housing costs across some 385 different metropolitan and nonmetropolitan areas (and by whether someone is renting, owns their home outright, or is paying a mortgage) to adjust for cost-of-living differences by place of residence. Finally, whereas the equivalence scaling underlying the official measure is somewhat ad hoc, the SPM uses a formula that adjusts the thresholds consistently depending on the number of adults and children in the family and whether the head of household is a single parent. Although no measure of poverty is perfect, it is clear that the SPM does correct for significant shortcomings in the official measure. The SPM is not a measure included in the original ASEC files. However, it is made available separately and is easily merged with the ASEC data by matching the household sequence number and the unique person identification number, which was done for this analysis.
“Rural” residence is defined as living outside of a metropolitan area. In essence, metropolitan areas consist of counties with a city of 50,000 or more people (or a total urbanized area of 100,000 or more), plus surrounding counties with strong economic ties. All other counties are defined as nonmetropolitan. This analysis used the CPS variable indicating whether the household is in a metropolitan area, in a nonmetropolitan area, or is not identified.5 A separate four-category variable was used to further subdivide metro households into those located in a principal city of a metro area or those outside of principal cities (i.e., central city versus suburban residents). Other measures are specified as they are introduced into the analysis.
Tables 3.1 and 3.2 show the prevalence of poverty using the various definitions of poverty previously described and how these poverty rates differ by residence and race/ethnicity. These tables include all ages of individuals available, ranging from fifteen years of age and top coded at eighty-five years of age. Table 3.1 presents poverty measures by metropolitan status and region of residence. Not surprisingly, the total proportion of the population in poverty varies dramatically when comparing across the various measures. For example, in the total population, 6.7 percent are in deep poverty, 14.9 percent are under the official poverty line, and 24.2 percent are near poor (under 150 percent of the poverty line). The 14.9 percent living in poverty according to the official definition compares to 15.3 percent when using the SPM, suggesting that measured poverty is somewhat worse overall when correcting for deficiencies in the official definition. Finally, according to this crude measure, relative poverty stood at 21.7 percent.
Table 3.1 Poverty Rates (%) by Metropolitan Status and Region
| |
Total |
Nonmetro |
Metro all |
Central cities |
Non-central cities |
| Deep poverty |
6.7% |
7.4% |
6.6% |
8.6% |
5.2% |
| Official poverty |
14.9 |
16.9 |
14.5 |
18.8 |
11.9 |
| Near poverty |
24.2 |
28.0 |
23.5 |
29.4 |
19.7 |
| Relative poverty |
21.7 |
26.5 |
20.9 |
26.0 |
17.5 |
| SPM poverty |
15.3 |
12.8 |
15.7 |
20.4 |
13.4 |
| N = |
199,024 |
37,100 |
159,724 |
49,577 |
77,523 |
| |
Total |
Northeast |
Midwest |
South |
West |
| Deep poverty |
6.7% |
5.6% |
5.8% |
7.7% |
6.8% |
| Official poverty |
14.9 |
12.7 |
13.1 |
16.6 |
15.3 |
| Near poverty |
24.2 |
20.9 |
22.0 |
26.7 |
24.6 |
| Relative poverty |
21.7 |
19.0 |
19.6 |
24.6 |
21.1 |
| SPM poverty |
15.3 |
14.6 |
11.8 |
15.6 |
18.4 |
| Weighted N |
199,024 |
33,252 |
39,362 |
70,439 |
55,971 |
Data: March 2015 Current Population Survey.
Table 3.2 Poverty Rates (%) by Metropolitan Status and Race/Ethnicity
| |
Total |
Nonmetro |
Metro |
Central city |
Non-central city |
| White non-Hispanic |
|
|
|
|
|
| Deep poverty |
4.7% |
5.9% |
4.4% |
5.7% |
3.7% |
| Official poverty |
10.1 |
13.8 |
9.3 |
11.2 |
8.0 |
| Near poverty |
17.4 |
23.9 |
15.9 |
17.9 |
14.1 |
| Relative poverty |
16.3 |
23.3 |
14.6 |
15.8 |
13.0 |
| SPM poverty |
10.7 |
11.1 |
10.6 |
13.1 |
9.4 |
| N = |
118,880 |
28,200 |
89,210 |
18,874 |
46,929 |
| Black non-Hispanic |
|
|
|
|
|
| Deep poverty |
12.3% |
15.8% |
11.8% |
13.3% |
9.0% |
| Official poverty |
26.3 |
33.5 |
25.4 |
28.6 |
20.1 |
| Near poverty |
38.2 |
48.4 |
36.9 |
41.7 |
29.6 |
| Relative poverty |
36.2 |
47.3 |
34.8 |
40.0 |
27.6 |
| SPM poverty |
23.1 |
20.5 |
23.3 |
26.2 |
20.0 |
| N = |
23,088 |
2,905 |
20,067 |
9,807 |
7,541 |
| Asian non-Hispanic |
|
|
|
|
|
| Deep poverty |
5.5% |
4.9% |
5.5% |
6.9% |
4.1% |
| Official poverty |
11.7 |
7.0 |
11.8 |
14.7 |
9.4 |
| Near poverty |
20.1 |
21.4 |
20.1 |
23.7 |
17.1 |
| Relative poverty |
17.2 |
16.0 |
17.2 |
20.1 |
14.9 |
| SPM poverty |
16.8 |
9.0 |
17.0 |
20.1 |
14.9 |
| N = |
11,313 |
628 |
10,654 |
4,622 |
4,944 |
| Other non-Hispanic |
|
|
|
|
|
| Deep poverty |
9.0% |
11.1% |
8.4% |
10.4% |
6.3% |
| Official poverty |
19.9 |
25.3 |
18.3 |
20.8 |
15.7 |
| Near poverty |
30.0 |
38.4 |
27.6 |
29.4 |
24.8 |
| Relative poverty |
25.0 |
31.1 |
23.2 |
26.9 |
20.0 |
| SPM poverty |
16.2 |
15.7 |
16.4 |
19.5 |
14.4 |
| N = |
7,704 |
2,127 |
5,384 |
1,624 |
2,543 |
| Hispanic |
|
|
|
|
|
| Deep poverty |
9.8% |
11.7% |
9.6% |
10.1% |
9.0% |
| Official poverty |
23.8 |
27.5 |
23.5 |
24.9 |
21.7 |
| Near poverty |
38.5 |
43.9 |
38.2 |
40.7 |
34.8 |
| Relative poverty |
31.7 |
35.9 |
31.5 |
34.1 |
28.5 |
| SPM poverty |
25.4 |
21.1 |
25.7 |
28.2 |
24.1 |
| Weighted N |
38,039 |
3,240 |
34,409 |
14,650 |
15,566 |
Data: March 2015 Current Population Survey.
Residential differences are, for the most part, suggestive of a rural disadvantage. Whether measured as deep, official, near, or relative poverty, individuals in nonmetro areas have higher poverty rates than their metro counterparts. A decided exception, however, is when poverty is measured via the SPM. That measure suggests that poverty is less common in nonmetro than metro areas. Although this would suggest a rural advantage rather than disadvantage, some caution is warranted. One thing that is driving the apparent improvement in rural poverty when using the SPM is the correction for cost-of-living differences across space. However, that correction is based solely on differences in the cost of housing. At present the SPM does not factor in the cost of necessities such as transportation, which are often more expensive in rural than urban areas. That said, although most measures of poverty give rise to a story of rural disadvantage, according to one new measure designed to be more valid and reliable than the official definition, exactly the opposite story emerges—people in rural areas have lower poverty rates according to the SPM.
Although not the focus of this chapter, when subdividing metro areas into principal cities versus elsewhere, the commonplace disadvantage of central cities in relation to the suburbs is clear, regardless of poverty measure used. However, the results suggest that the gross nonmetro versus metro comparison obscures disadvantages of those in central cities when compared to nonmetro residents. According to most measures, poverty is worse in nonmetro than metro places, but it is the central cities of metro areas that are the worst off (an exception being for relative poverty, which is higher in nonmetro areas than even in central cities).
The second panel of table 3.1 shows poverty rates by region and, again, except for the SPM, indicates a southern disadvantage. When using the SPM, the West has the highest rates of poverty (18.4 percent). The Northeast and Midwest have comparable poverty rates across the varying poverty measures.
Table 3.2 shows the array of poverty rates using the same set of poverty measures broken down by race and ethnicity. Specifically, poverty is examined across five racial/ethnic groups: non-Hispanics who identify as (1) white, (2) black, (3) Asian, and (4) any other race, and (5) Hispanics (who can be of any race). The results underscore persisting racial/ethnic differences in poverty risks with blacks, Hispanics, and “others” (mostly Native Americans) having distinctly higher poverty rates than non-Hispanic Asians and whites. Breaking poverty rates down by race and ethnicity also has the effect of suggesting that poverty prevalence is higher in nonmetro America than even in central cities. That is, for all but the SPM, poverty is greater in nonmetro areas than in principal cities of metro areas among whites, blacks, Hispanics, and others. Black non-Hispanics, with Hispanics and non-Hispanics, have the highest prevalence of poverty for all measures. The results for the SPM are consistent with this finding only with regard to confirming higher poverty among blacks, Hispanics, and others. Otherwise, the SPM again suggests that for all but whites, nonmetro poverty rates are lower than those in metro areas.
AN ASIDE ON SUBJECTIVE POVERTY
In this descriptive portrait of poverty, virtually all of the poverty indicators are income-based. They might define what counts as income differently, or compare that income to different thresholds, but they all have income at their core. As discussed previously, other definitions of poverty are based on consumption or expenditures, and still others are more subjective in nature. One form of subjective poverty is when people define themselves as poor. Perhaps, at the very least, because it seems callous to ask, nationally representative surveys do not attempt to define poverty in this way. One survey that arguably comes close to having a subjective definition of poverty is the General Social Survey (GSS), regularly conducted by the National Opinion Research Center (Smith et al. 2016). The GSS collects data on a variety of opinions and attitudes of the U.S. population, as well as their basic sociodemographic and economic characteristics. Attitudinal items cover issues such as national spending, economics, politics, crime, and intergroup relations, among many others. Although the sample size is modest (roughly 2,000 in any given year or about twice the size of the typical political poll), the GSS is nonetheless considered nationally representative. In this analysis, the 2010, 2012, and 2014 surveys from the GSS are combined.6
The GSS includes the following question (Smith et al. 2016): “If you were asked to use one of four names for your social class, which would you say you belong in: the lower class, the working class, the middle class, or the upper class?” To be sure, this question does not specifically mention “poverty” as being a social class. However, given that it asks respondents to intuitively distinguish the “lower” class from higher class groups, placing oneself in the lower class may serve as a rough proxy for subjective poverty status. Of interest here is how this subjective class status relates to rural versus urban residence. Regarding the latter, the GSS includes the size of the place where the respondent lives.
As shown in table 3.3, in both rural and urban areas, Americans tend to see themselves as being in the working or middle class. In urban areas, the split between the two is rather even (45.1 percent versus 44.1 percent, respectively), whereas in rural areas, adults are much more likely to see themselves as working class than middle class (52.9 percent versus 34.2 percent). Of much greater interest here, rural adults are far more likely to place themselves in the lower class (12.2 percent) than are urban residents (7.6 percent). To the extent that lower-class identification is a valid proxy for subjective poverty, one can argue that poverty is significantly more prevalent in America’s rural than urban areas.
Table 3.3 Subjective Class Placement
| |
Rural |
Urban |
Total (weighted) |
| Lower class |
12.2% |
7.6% |
531 |
| Working class |
52.9 |
45.1 |
3,000 |
| Middle class |
34.2 |
44.1 |
2,785 |
| Upper class |
0.6 |
3.2 |
188 |
| N = |
|
|
6,505 |
Data: General Social Survey (2010, 2012, and 2014).
CONCLUSION
Poverty is an all too common social problem that, even for those lucky enough to never experience it, affects us all in one way or another. As a commonplace, complex, and controversial problem, it is critically important to define poverty carefully. Measures of poverty that are valid (measure what is intended) and reliable (operate consistently over time) serve two functions. First, they allow for rigorous scientific study, which contributes to the scholarship on the causes, correlates, nature, and consequences of poverty. The second function of poverty measures is political. They allow for policy development and evaluation, and they provide a way to target policies to those most in need and to establish eligibility for support.
Poverty has been defined and measured in a surprisingly wide variety of ways. Many measures are income-based and compare income received during a particular period—usually a year—to some predetermined poverty thresholds. These thresholds can be absolute or relative. Conceptually speaking, absolute poverty means having insufficient income to buy a minimally acceptable standard of living; relative poverty means having an income that is substantially less than average. The official definition of poverty in the United States is an absolute measure developed in the 1960s, and apart from adjusting the official poverty thresholds to account for inflation, it has remained largely unchanged to this day. The official definition of poverty in the United States has come under sharp criticism for ignoring cost-of-living differences across places, for not counting in-kind income, and for other reasons. The supplemental poverty measure corrects for many of these flaws, but it has not replaced the official definition.
In addition to income-based approaches, other definitions of poverty include consumption-based measures, which use expenditures to assess poverty status; and deprivation measures, which examine the actual circumstances of people and families and assess poverty status based on the things they lack—that is, the deprivations they exhibit. The consensual approach takes this a step further by using survey research to first determine the various consumption items regarded as necessary in a population and then using that information to define a set of deprivations. Poverty status is assessed by the number of deprivations people or families endure. By relying on popular opinion to first ascertain local opinion about what families should not be expected to do without, this consensual approach begins to contend with the subjective nature of poverty. What is considered poor in some societies may not be considered poor in others. One way researchers measure subjective poverty is by asking people what they feel is the minimum income level families need to survive, and then define as poor those with incomes less than that amount. Subjective poverty can also be determined most crudely by asking people in surveys to self-identify as poor.
Using some (though admittedly not all) of these definitions, a contemporary statistical portrait of rural and urban poverty in the United States has been described that shows how patterns of inequality differ depending on the definition used. A key takeaway message is that by most definitions nonmetropolitan residents are decidedly more likely to be poor than their more urban counterparts. The glaring exception to this pattern occurs when poverty is measured by the SPM. This measure suggests that poverty prevalence is lower in nonmetro than metro areas, lower even than in the more prosperous suburbs of metro areas. Although this could be taken to mean that rural residents are not more likely to be poor, two caveats remain. First, the SPM’s cost-of-living adjustment only accounts for the cost of housing and ignores some necessities (e.g., transportation) that may be more expensive in the countryside. Second, there is no evidence for less material deprivation in rural areas despite the presumption rural families can stretch their dollars because it is cheaper to live there (Lichter and Schafft 2016). For all but the SPM, when poverty rates are considered separately for major race and ethnic groups in the United States, nonmetro whites, blacks, Hispanics, and others (mostly Native Americans) have poverty rates higher than their counterparts in metro areas generally, and even in the central cities of metro areas specifically. Finally, a much higher percentage of rural than urban adults define themselves as being in the lower class and are arguably in a state of subjective poverty. Other chapters delve more deeply into the severity, nature, and contours of rural deprivation in the United States.
NOTES
1. The goals of the redesign to the ASEC include improving income reporting, increasing response rates, reducing errors through the use of an automated questionnaire, and updating questions related to retirement income and income from retirement accounts and other assets. More information on the redesign can be found at https://www.census.gov/library/publications/2015/demo/p60-252.html.
3. Further information on the supplemental poverty measure can be found in Short (2015).
4. Because the SPM thresholds are set within a certain percentile range of expenditures, and because they will be continually revised with updated expenditure data, some suggest the SPM is “quasi-relative” (Iceland 2013, 32).
5. To protect the confidentiality of CPS respondents, for a small proportion of all sampled households the Census Bureau will not identify whether the household is in a metro or nonmetro area. Definitions of metropolitan and nonmetropolitan are based on OMB standards. More information on these can be found at https://www.census.gov/population/metro/about/.
6. Further, we use an adjusted weight from the final weighting variable (WTSSALL) to ensure that any analysis accounts for the over- or under-sampling of the adult population.
REFERENCES
Alkire, Sabina, Jose Manuel Roche, Suman Seth, and Andrew Sumner. 2015. “Identifying the Poorest People and Groups: Strategies Using the Global Multidimensional Poverty Index.” Journal of International Development 27:362–87.
Carlson, A., M. Lino, W-Y. Juan, K. Hanson, and P. P. Basiotis. 2007. “Thrifty Food Plan, 2006.” CNPP-19. U.S. Department of Agriculture, Center for Nutrition Policy and Promotion.
Citro, Constance E., and Robert T. Michael. 1995. Measuring Poverty: A New Approach. Washington, DC: National Academies Press.
Danziger, Sheldon, Jacques van der Gaag, Michael K. Taussig, and Eugene Smolensky. 1984. “The Direct Measurement of Welfare Levels: How Much Does It Cost to Make Ends Meet?” Review of Economics and Statistics 66(3):500–505.
DeNavas-Walt, Carmen, and Bernadette D. Proctor. 2015. “Income and Poverty in the United States: 2014.” Current Population Reports. P60–252. Washington, DC: U.S. Census Bureau.
Iceland, John. 2013. Poverty in America: A Handbook, 3rd ed. Berkeley: University of California Press.
Lichter, Daniel T., and Kai A. Schafft. 2016. “People and Places Left Behind: Rural Poverty in the New Century.” In Oxford Handbook of Poverty and Society, 317–40. Oxford, UK: Oxford University Press.
Myer, Bruce D., and James X. Sullivan. 2012. “Identifying the Disadvantaged: Official Poverty, Consumption Poverty and the New Supplemental Poverty Measure.” Journal of Economic Perspectives 26(3):111–36.
Nandy, Shailen, and Gill Main. 2015. “The Consensual Approach to Child Poverty Measurement.” Poverty Brief. Comparative Research Programme on Poverty. Bergen, Norway: University of Bergen.
Nandy, Shailen, and Marco Pomati. 2015. “Applying the Consensual Method of Estimating Poverty in a Low Income African Setting.” Social Indicators Research 124:693–726.
Posey, Kirby G. 2016. “Household Income 2015.” American Community Survey Briefs, U.S. Census Bureau, ACSBR/15-02.
Rank, Mark R., and Thomas A. Hirschl. 1999. “The Likelihood of Poverty Across the American Adult Life Span.” Social Work 44(3):201–216.
Ruggles, Patricia. 1990. Drawing the Line: Alternative Poverty Measures and Their Implications for Public Policy. Washington, DC: The Urban Institute Press.
Smith, Tom W., Peter Marsden, Michael Hout, and Jibum Kim. 2016. “General Social Surveys, 1972–2014.” Machine-readable data file. Sponsored by National Science Foundation. Chicago: NORC at the University of Chicago (producer and distributor).
