American poverty as a structural failing: evidence and arguments.

Hirschl, Thomas A.

Empirical research on American poverty has largely focused on individual characteristics to explain the occurrence and patterns of poverty. The argument in this article is that such an emphasis is misplaced. By focusing upon individual attributes as the cause of poverty, social scientists have largely missed the underlying dynamic of American impoverishment. Poverty researchers have in effect focused on who loses out at the economic game, rather than addressing the fact that the game produces losers in the first place. We provide three lines of evidence to suggest that U.S. poverty is ultimately the result of structural failings at the economic, political, and social levels. These include an analysis into the lack of sufficient jobs in the economy to raise families out of poverty or near poverty; a comparative examination into the high rates of U.S. poverty as a result of the ineffectiveness of the social safety net; and the systemic nature of poverty as indicated by the life course risk of impoverishment experienced by a majority of Americans. We then briefly outline a framework for reinterpreting American poverty. This perspective incorporates the prior research findings that have focused on individual characteristics as important factors in who loses out at the economic game, with the structural nature of American poverty that ensures the existence of economic losers in the first place.

Keywords: American poverty, low wages, U.S. economy, social welfare, human capital, social structure

**********

Few questions have generated as much discussion across time as that of the causes of human impoverishment. The sources and origins of poverty have been debated for centuries. As the historian R. M. Hartwell notes, "The causes of poverty, its relief and cure, have been a matter of serious concern to theologians, statesmen, civil servants, intellectuals, tax-payers and humanitarians since the Middle Ages" (1986: 16). The question of causality has found itself at the heart of most debates surrounding poverty and the poor.

In recent times these debates have often been divided into two ideological camps. On one hand, poverty has been viewed as the result of individual failings. From this perspective, specific attributes of the impoverished individual have brought about their poverty. These include a wide set of characteristics, ranging from the lack of an industrious work ethic or virtuous morality, to low levels of education or competitive labor market skills. On the other hand, poverty has periodically been interpreted as the result of failings at the structural level, such as the inability of the economy to produce enough decent paying jobs.

Within the United States, the dominant perspective has been that of poverty as an individual failing. From Ben Franklin's Poor Richard's Almanac to the recent welfare reform changes, poverty has been conceptualized primarily as a consequence of individual failings and deficiencies. Indeed, social surveys asking about the causes of poverty have consistently found that Americans tend to rank individual reasons (such as laziness, lack of effort, and low ability) as the most important factors related to poverty, while structural reasons such as unemployment or discrimination are viewed as significantly less important (Feagin, 1975; Gilens, 1999; Kluegel and Smith, 1986).

This emphasis on individual attributes as the primary cause of poverty has also been reinforced by social scientists engaged in poverty research (O'Connor, 2001). As the social survey has become the dominant methodological approach during the past 50 years, and with multivariate modeling becoming the principal statistical technique, the research emphasis has increasingly fallen on understanding poverty and welfare dependency in terms of individual attributes. The unit of analysis in these studies is by definition the individual rather than the wider social or economic structures, resulting in statistical models of individual characteristics predicting individual behavior. Consequently, the long standing tension between structural versus individual approaches to explaining poverty has largely been tilted within the empirical poverty research community towards that of the individual. As Alice O'Connor writes,

 That this tension has more often been resolved in favor of the
 individualist interpretation can be seen in several oft-noted
 features in poverty research. One is the virtual absence of class
 as an analytic category, at least as compared with more
 individualized measures of status such as family background and
 human capital. A similar individualizing tendency can be seen in
 the reduction of race and gender to little more than demographic,
 rather than structurally constituted, categories (2001: 9).

The argument in this article is that such an emphasis is misplaced and misdirected. By focusing on individual attributes as the cause of poverty, social scientists have largely missed the underlying dynamic of American impoverishment. Poverty researchers have in effect focused on who loses out at the economic game, rather than addressing the fact that the game produces losers in the first place. An analysis into this underlying dynamic is critical to advancing our state of knowledge regarding American poverty.

Of course, not all social scientists have abandoned the importance of structural considerations with respect to poverty. The work of William Ryan (1971), Michael Katz (1989), Herbert Gans (1995), Douglass Massey (1996) and Joe Feagin (2000) come to mind. However, it should not be a surprise that most of these scholars have taken a theoretical or historical approach, rather than a statistical one. There is a need to articulate the quantitative evidence supporting the argument that U.S. poverty is ultimately the result of structural failings at the economic, political, and social levels.

Current Understanding of American Poverty

The current research emphasis upon understanding American poverty has largely focused on the individual and demographic characteristics of the poor. These characteristics have, in turn, been used to explain why particular individuals and households experience poverty. This approach has revealed the extent to which the risk of poverty varies across particular individual and household attributes.

Repeated cross-sectional national surveys such as the Current Population Survey have indicated that the likelihood of poverty varies sharply with respect to age, race, gender, family structure, and residence. For example, the U.S. Census Bureau (2003a) reports that while the overall U.S. poverty rate in 2002 was 12.1 percent, it was 16.7 percent for children and for those residing in central cities, 24.1 percent for African Americans, and 28.8 percent for persons in female headed households. Other demographic characteristics closely associated with the risk of poverty include giving birth outside of marriage, families with larger numbers of children, and having children at an early age (Maynard, 1997).

In addition, cross-sectional research has shown a close association between human capital characteristics and an individual's risk of poverty--those who are lacking in human capital are much more likely to experience poverty than individuals with greater levels of human capital. Specifically, lower levels of education, less marketable work skills and experience, and having a physical disability that interferes with an individual's ability to participate in the labor market are all highly correlated with an elevated risk of poverty (Blank, 1997; Schiller, 2004). On the other hand, research comparing the attitudes and motivation of the poor versus the non-poor, have found relatively few differences between these two groups (Goodwin, 1973; 1983; Lichter and Crowley, 2002; Rank, 1994; Seccombe, 1999) and little in the way of their being a causal factor leading to poverty (Duncan, 1984; Edwards et al., 2001).

Longitudinal studies examining the dynamics of poverty have addressed the length of time and particular factors related to a spell of poverty. This body of work indicates that most spells of poverty are of modest length. The typical pattern is that households are impoverished for one, two, or three years, and then manage to get above the poverty line (Bane and Ellwood, 1986; Blank, 1997; Duncan, 1984; Walker, 1994). They may stay there for a period of time, only to experience an additional fall into poverty at some later point. For example, Stevens (1994; 1999) calculated that of all persons who had managed to get themselves above the poverty line, over half would return to poverty within five years. Since their economic distance above the poverty threshold is often modest, a detrimental economic or social event can push a household back below the poverty line.

Longitudinal research has also focused on the nature of such events and individual changes that result in a spell of poverty (Devine and Wright, 1993; Duncan, 1984; Walker, 1994). The most important of these have been the loss of employment and earnings, along with changes in family structure. For example, using the Panel Study of Income Dynamics (PSID) data, Duncan et al. (1995) found that two thirds of all entries into poverty were associated with either a reduction in work (48 percent) or the loss of work (18 percent). Divorce and separation were associated with triggering approximately 10 percent of all spells of poverty. Blank (1997) found that employment and family structure changes were also influential in ending spells of poverty. Two thirds of those below the poverty line escaped impoverishment as a result of increases in the individual earnings of family members or increases from other sources of income, while the remaining third had their spells of poverty end as a result of changes in family structure (such as marriage or a child leaving the household). In addition, research has shown that illness and incapacitation are also important factors contributing to falls into poverty (Schiller, 2004).

A substantial body of work has also examined the dynamics of welfare use and dependency. This research has shown that individuals utilizing public assistance programs and who experience longer spells of welfare are often at a distinct disadvantage vis-a-vis the labor market (Bane and Ellwood, 1994; Boisjoly et al., 1998; Harris, 1996; Moffitt, 1992; Pavetti, 1992; Rank, 1988; Sandefur and Cook, 1998). Consequently, those with work disabilities, low education, greater numbers of children, and/or living in inner-city areas are more likely to extensively utilize the welfare system. The results from these studies largely mirror the findings that have been gathered regarding the length and duration of poverty spells.

The above body of work has provided an important understanding into who are the economic losers in American society. Yet at the same time it has failed to address the question of why there are economic losers in the first place? The premise of this article is that in order answer this question, it is essential to analyze specific failings at the structural level.

The Structural Nature of Poverty

Three lines of evidence are detailed in order to illustrate the structural nature of poverty--1) the inability of the U.S. labor market to provide enough decent paying jobs for all families to avoid poverty or near poverty; 2) the ineffectiveness of American social policy to reduce levels of poverty through governmental social safety net programs; and 3) the fact that the majority of the population will experience poverty during their adult lifetimes, indicative of the systemic nature of U.S. poverty. Each of these lines of evidence are intended to empirically illustrate that American poverty is by and large the result of structural failures and processes.

The Inability of the Labor Market to Support All Families

Several of the pioneering large scale studies of poverty conducted at the end of the 19'th and beginning of the 20'th century focused heavily on the importance of labor market failings to explain poverty. The work of Charles Booth (1892-1897), Seebohm Rowntree (1901), Hull House (1895), Robert Hunter (1904), and W. E. B. DuBois (1899) all emphasized the importance of inadequate wages, lack of jobs, and unstable working conditions as a primary cause of poverty. For example, Rowntree (1901) estimated that approximately 57 percent of individuals in poverty were there as a direct result of labor market failures (low wages, unemployment, irregularity of work).

Yet by the 1960's the emphasis had shifted from a critique of the economic structure as a primary cause of poverty, to an analysis of individual deficiencies (e.g., the lack of human capital) as the underlying reason for poverty. As Timothy Bartik (2001) notes, U.S. antipoverty policy has focused heavily on labor supply policies (e.g. increasing individual's human capital or incentives to work through welfare reform) rather than labor demand policies (increasing the number and quality of jobs). As mentioned earlier, social scientific research has reinforced this policy approach by focusing on individual deficiencies to explain individual poverty.

Yet it can be demonstrated that irrespective of the specific characteristics that Americans possess, there simply are not enough decent paying jobs to support all of those (and their families) who are looking for work. During the past 25 years the American economy has increasingly produced larger numbers of low paying jobs, jobs that are part-time, and jobs that are lacking in benefits (Seccombe, 2000). For example, the Census Bureau estimated that the median hourly earning of workers who were paid hourly wages in 2000 was $9.91, while at the same time approximately three million Americans were working part-time as a result of the lack of sufficient full-time work being available (U.S. Census Bureau, 2001). In addition, 43.6 million Americans were lacking in health insurance, largely because their employer did not provide such benefits (U.S. Census Bureau, 2003b).

Studies analyzing the percentage of the U.S. workforce falling into the low wage sector have shown that a much higher percentage of American workers fall into this category when compared with their counterparts in other developed countries. For example, Smeeding, Rainwater, and Burtless (2000) found that 25 percent of all U.S. full-time workers could be classified as working in low wage work (defined as earning less than 65 percent of the national median earnings on full-time jobs). This was by far the highest percentage of the countries analyzed, with the overall average falling at 12.9 percent.

One of the reasons for this has been the fact that the minimum wage has remained at low levels and has not been indexed to inflation. Changes in the minimum wage must come through Congressional legislation. This often leads to years going by before Congress acts to adjust the minimum wage upward, causing it to lag further behind the cost of living.

Beyond the low wages, part-time work, and lack of benefits, there is also a mismatch between the actual number of available jobs and the number of those who need them. Economists frequently discuss what is known as a natural unemployment rate-that in order for a free market economy to effectively function, a certain percentage of laborers need to be out of work. For example, full employment would impede the ability of employers to attract and hire workers, particularly within the low wage sector. Consequently, a certain degree of unemployment appears systematic within a capitalist economy, irrespective of the individual characteristics possessed by those participating in that economy.

During the past 40 years, U.S. monthly unemployment rates have averaged between 4 and 10 percent (U.S. Census Bureau, 2001). These percentages represent individuals who are out of work but are actively looking for employment. In 2001 this translated into nearly 7 million people unemployed at any particular point in time throughout the year, while over 15 million people experienced unemployment at some point during the year (Schiller, 2004). Certainly some of these individuals have voluntarily left their jobs in order to locate another job (known as frictional unemployment), while in other cases the unemployed may include individuals whose family's are not dependent upon their job for its economic survival , such as teenagers looking for summer work. Nevertheless, a good proportion of unemployment is the result of involuntary reasons, such as layoffs and downsizing, directly affecting millions of heads of households.

Bartik (2001; 2002) used several different approaches and assumptions to estimate the number of jobs that would be needed to significantly address the issue of poverty in the United States. Data were analyzed from the 1998 Current Population Survey. His conclusion? Even in the booming economy of the late 1990's, between five and nine million more jobs were needed in order to meet the needs of the poor and disadvantaged.

The structural failing of the labor market to support the pool of labor that currently exists can be further illustrated through an analysis of the Survey of Income and Program Participation (SIPP). The SIPP is a large ongoing longitudinal study that interviews households every four months over the course of three or four years, gathering detailed monthly information regarding individual's employment and income across these periods of time. It allows one to map the patterns of labor force participation for a large nationally representative sample (for an in-depth discussion of the history, methodology, and specific details of the SIPP data set, see Westat, 2001).

An analysis of the SIPP illustrates the mismatch between the number of jobs in the labor market that will enable a family to subsist above the threshold of poverty, versus the number of heads of families in need of such jobs. Table 1 is based upon the jobs and work behavior of family heads across all 12 months of 1999. From this we can estimate the annual number of hours worked, the annual amount of pay received, and whether such earnings were sufficient to raise a family above the poverty line. The analysis is confined to heads of families who are between the ages of 18 and 64.

In Table 1, we examine whether the jobs that family heads were working at during the year were able to get their current families out of poverty. Three poverty thresholds are examined-below 1.00 (the official poverty line); below 1.25 of the poverty line (the official poverty line raised by 25 percent); and below 1.50 of the poverty line (the official poverty line raised by 50 percent). To illustrate, the poverty line for a family of 4 in 1999 was $17,029. Consequently the 1.25 poverty threshold for this family would be $21,286, while the 1.50 poverty threshold would be $25,544. These thresholds provide us with several alternative levels of poverty and near poverty.

Our focus is on the availability of jobs in the labor market to lift various families out of poverty. We examine this question for three different populations of family heads who are in the labor market. The first panel focuses only on those heads of families who are working full-time throughout the year (defined as averaging 35 or more hours per week across the 52 weeks of the year). The second panel includes those working full-time as well as those who are working at least half-time throughout the year (defined as working an average of 20 or more hours per week across 52 weeks). The third panel includes all heads of families in the labor market (defined as any head of family who has either worked at some point during the year or who has been actively looking for work).

For those employed full-time during 1999, 9.4 percent are working in jobs where their annual earnings will not get their families above the poverty line, 15.3 percent are at jobs in which their earnings will not get their families above 1.25 of the poverty line, and 22.0 percent are employed at jobs that will not get their families above 1.50 of the poverty line. We can clearly see that the jobs one parent family heads are working at are much less able to sustain these households above the level of poverty than that for all families. On the other hand, single men and women are more likely to be able to lift themselves out of poverty through their work. Married couples fall in between these two family types (it should be kept in mind that for these couples, we are only focusing on the ability of the family head's job to lift the household above the threshold of poverty, rather than the earnings of both partners).

The middle panel illustrates that if we include family heads who are working either full-time or at least half-time throughout the year, nearly 15 percent were working at jobs in which their income would not raise their families above the poverty line, 21.4 percent were at jobs that would not get their families over 1.25 of the poverty line, while 28 percent fell below 1.50 of the poverty line. Finally, the bottom panel includes all family heads that were in the labor market at some point during the year. Here we can see that the percentages for the three poverty thresholds are 20.3, 26.5, and 32.7.

Consequently, depending on the level of poverty and the size of the pool of labor, the failure of the labor market to raise families out of poverty ranges from 9.4 percent (utilizing the official poverty line for those working full-time) to 32.7 percent (applying 1.50 of the poverty line for all who are in the labor market). To use an analogy that will be developed later, the supply of jobs versus the demand for labor might be thought of as an ongoing game of musical chairs. That is, there is a finite number of jobs available in the labor market that pay enough to support a family above the threshold of poverty (which might be thought of as the chairs in this analogy). On the other hand, the amount of labor, as represented by the number of family heads in the labor market (and hence the players in the game), is greater than the number of adequately paying jobs. As indicated in Table 1, this imbalance ranges from 9.4 percent to 32.7 percent. Consequently, the structure of the labor market basically ensures that some families will lose out at this musical chairs game of finding a decent paying job able to lift a family above the threshold of poverty.

Table 2 illustrates this in a slightly different fashion. Here we estimate the earnings capacity of jobs held by family heads to support various hypothetical family sizes above our three different thresholds of poverty. What is clear from this table is that for the pool of family heads who are working full-time, the jobs that they are employed at are quite able to support a one or two person family above the official poverty line. For example, only 2.4 percent are in full-time jobs in which their earnings would not raise a one person family above the official poverty line, while 4.7 percent of family heads are working at jobs that would not raise a family of two above the poverty line.

However, as we look at the ability of such jobs to get larger sized families above the thresholds of poverty, we can see their increasing failure to do so. Consequently, 15 percent of these jobs will not raise a family of four above 1.00 of the poverty line. At the 1.25 level the figure is 25.1 percent, and at the 1.50 level it is 36 percent. Thus, the current supply of full-time jobs in the labor market would appear able to lift most one or two person families out of poverty, but it becomes much less effective in raising moderate sized families out of poverty. As we include family heads who are working at least half-time (the middle panel of Table 2) or who are in the labor market (the bottom panel of Table 2) the percentages rise significantly.

Finally, we can illustrate this in yet another manner. Using the SIPP data again for 1999, we estimated the annual average hourly wages for heads of families. This analysis indicates that 12.1 percent of family heads were working at jobs which paid an average of less than 6 dollars an hour, 21.2 percent worked at jobs paying less than 8 dollars an hour, 31.7 percent worked at jobs paying less than 10 dollars an hour, and 42.7 percent were earning less than 12 dollars an hour. In order to raise a family of three above the official poverty line in 1999 one would have to be working full-time (defined as averaging 35 hours per week across the 52 weeks of the year) at $7.30 an hour, and for a family of four the figure would be $9.36 an hour.. The fact that nearly one third of family heads are working at jobs paying less than $10.00 an hour, is indicative of the significant risk of poverty that they face.

To summarize, the data presented in this section indicates that a major factor leading to poverty in the United States is a failing of the economic structure to provide viable opportunities for all who are participating in that system. In particular, the labor market simply does not provide enough decent paying jobs for all who need them. As a result, millions of families find themselves struggling below or precariously close to the poverty line.

The Ineffectiveness of the Social Safety Net to Prevent Poverty

A second major structural failure is found at the political level. Contrary to the popular rhetoric of vast amounts of tax dollars being spent on public assistance, the American welfare state, and particularly its social safety net, can be more accurately described in minimalist terms (Esping-Andersen, 1990). Compared to other Western industrialized countries, the United States devotes far fewer resources to programs aimed at assisting the economically vulnerable (Organization for Economic Cooperation and Development, 1999). As Charles Noble writes, "The U .S. welfare state is striking precisely because it is so limited in scope and ambition" (1997: 3).

On the other hand, most European countries provide a wide range of social and insurance programs that largely prevent families from falling into poverty. These include substantial family or children's allowances, designed to transfer cash assistance to families with children. Unemployment assistance is far more generous in these countries than in the United States, often providing support for more than a year following the loss of a job. Furthermore, universal health coverage is routinely provided, along with considerable support for child care.

The result of these social policy differences is that they substantially reduce the extent of poverty in Europe and Canada, while U.S. social policy has had only a small impact upon poverty reduction. As Rebecca Blank notes, "the national choice in the United States to provide relatively less generous transfers to low-income families has meant higher relative poverty rates in the country. While low-income families in the United States work more than in many other countries, they are not able to make up for lower governmental income support relative to their European counterparts" (Blank, 1997: 141-142).

This effect can be clearly seen in Table 3. The data in this table are based upon an analysis of the Luxembourg Income Study (LIS) conducted by Veli-Matti Ritakallio (2001). Initiated in the 1980s, the LIS contains income and demographic information on households in over 25 different nations from 1967 to the present. Variables have been standardized across 70 data sets, allowing researchers to conduct cross-national analyses regarding poverty and the effectiveness of governmental programs in alleviating such poverty (for further detail regarding the LIS, see Luxembourg Income Study, 2000). Poverty in this analysis is defined as being in a household in which its disposable income is less than one half of the median annual income.

Table 3 compares eight European countries and Canada with the United States in terms of their pre-transfer and post-transfer rates of poverty. The pre-transfer rates (column one) indicate what the level of poverty would be in each country in the absence of any governmental income transfers such as welfare payments, unemployment compensation, or social security payments. The post-transfer rates (column two) represent the level of poverty after governmental transfers are included (which is how poverty is officially measured in the United States and many other countries). In-kind benefits such as medical insurance are not included in the analysis. Comparing these two levels of poverty (column three) reveals how effective (or ineffective) governmental policy is in reducing the overall extent of poverty in a country.

Looking first at the rates of pre-transfer poverty, we can see that the United States is on the lower end of the scale. Norway's pre-transfer poverty rate is 27 percent, followed by the United States, Canada, and Germany at 29 percent. The Netherlands pre-transfer rate is 30 percent, Finland stands at 33 percent, Sweden is at 36 percent, the United Kingdom rate is 38 percent, and finally, France possesses the highest level of pre-transfer poverty at 39 percent.

When we examine the post-transfer rates of poverty found in column two, a dramatic reversal takes place in terms of where the United States stands vis-a-vis the comparison countries. The average post-transfer poverty rate for the eight comparison countries in Table 3 is 7 percent, whereas the United States' post-transfer poverty rate stands at 18 percent. As a result of their more active social policies, Canada and the European countries are able to significantly cut their overall rates of poverty. For example, Sweden is able to reduce the number of people that would be poor (in the absence of any governmental help) by 92 percent as a result of social policies. The overall average reduction factor for the eight countries is 79 percent. In contrast, the United States poverty reduction factor is only 38 percent (with much of this being the result of Social Security).

Table 3 clearly illustrates a second major structural failing leading to the high rates of U.S. poverty. It is a failure at the political and policy level. Specifically, social and economic programs directed to the economically vulnerable populations in the United States are minimal in their ability to raise families out of poverty. While America has always been a "reluctant welfare state," the past 25 years have witnessed several critical retrenchments and reductions in the social safety net. These reductions have included scaling back both the amount of benefits being transferred, as well as a tightening of program eligibility (Noble, 1997; Patterson, 2000). In addition, the United States has failed to offer the type of universal coverage for child care, medical insurance, or child allowances that most other developed countries routinely provide. As a result, the overall U.S. poverty rates remain at extremely high levels.

Once again, this failure has virtually nothing to do with the individual. Rather it is emblematic of a failure at the structural level. By focusing on individual characteristics, we lose sight of the fact that governments can and do exert a sizeable impact on reducing the extent of poverty within their jurisdictions. In the analysis presented here, Canada and Europe are able to lift a significant percentage of their economically vulnerable above the threshold of poverty through governmental transfer and assistance policies. In contrast, the United States provides substantially less support through its social safety net, resulting in poverty rates that are currently the highest in the industrialized world.

The one case where the U.S. has effectively reduced the rate of poverty for a particular group has been that of the elderly. Their substantial reduction in the risk of poverty over the past 40 years has been directly attributed to the increasing generosity of the Social Security program, as well as the introduction of Medicare in 1965 and the Supplemental Security Income Program in 1974. During the 1960's and 1970's, Social Security benefits were substantially increased and indexed to the rate of inflation, helping many of the elderly escape from poverty. It is estimated today that without the Social Security program, the poverty rate for the elderly would be close to 50 percent (rather than its current 10 percent). Put another way, Social Security is responsible for getting 80 percent of the elderly above the poverty line who would otherwise be poor in its absence.

The Widespread Life Course Risk of Poverty

A third approach revealing the structural nature of American poverty can be found in a life course analysis of poverty. As discussed earlier, previous work on poverty has examined the cross-sectional and spell dynamic risk. Yet there is another way in which the incidence of poverty can be examined. Such an approach places the risk of poverty within the context of the American life course. By doing so, the systematic nature of American poverty can be revealed.

The work of Rank and Hirschl (1999a; 1999b; 1999c; 2001a; 2001b) has developed this approach. Building upon the longitudinal design of the Panel Study of Income Dynamics, Rank and Hirschl have utilized a technique for constructing a series of life tables estimating the probability that Americans will experience poverty at some point during their adulthood (see Rank and Hirschl, 2001c, for a more detailed description of their methodology and approach, and Hill, 1992, for a further discussion of the PSID).

Table 4 is based upon three separate life tables estimating the age specific and cumulative probabilities of experiencing poverty between the ages of 20 and 75 for the following poverty thresholds--1.00 (the official poverty line); 1.25 (the poverty line raised by 25 percent), and 1.50 (the poverty line raised by 50 percent). Table 4 reports the cumulative percentages of the American population that will encounter poverty at various points of adulthood.

At age 20 (the starting point of the analysis), we can see that 10.6 percent of Americans fell below the poverty line (which is similar to the cross-sectional rate of poverty for 20 year olds), with 15 percent falling below the 1.25 threshold and 19.1 percent falling below the 1.50 threshold. By the age of 35, the percent of Americans experiencing poverty has increased sharply--31.4 percent of Americans have experienced at least one year below the poverty line; 39 percent have experienced at least one year below 1.25 of the poverty line; and 46.9 percent have experienced a year below 1.50 of the poverty line. At age 55 the percentages stand at 45.0, 52.8, and 61.0, and by the age of 75, they have risen to 58.5 percent, 68 percent and 76 percent.

What these numbers indicate is that a clear majority of Americans will at some point experience poverty during their lifetimes. Rather than an isolated event that occurs only among what has been labeled the "underclass," the reality is that the majority of Americans will encounter poverty firsthand during their adulthoods.

Such patterns illuminate the systematic essence of American poverty, which in turn points to the structural nature of poverty. Occasionally we can physically see widespread examples of this. For instance, the economic collapse during the Great Depression of the 1930's. Given the enormity of this collapse, it became clear to many Americans that most of their neighbors were not directly responsible for the dire economic situation they found themselves in. This awareness helped provide much of the impetus and justification behind the New Deal (Patterson, 2000).

Similarly, the existence of the "other" America as noted by Michael Harrington (1962) during the early 1960's, pointed again to the widespread nature of U.S. poverty. The other America was represented by the extremely high rates of poverty found in economically depressed areas such as rural Appalachia and the urban inner city. The War on Poverty during the 1960's was an attempt to address these large scale structural pockets of poverty amidst plenty.

The analysis in this section indicates that poverty may be as widespread and systematic today as in these more visible examples. Yet we have been unable to see this as a result of not looking in the right direction. By focusing on the life span risks, the prevalent nature of American poverty is revealed. At some point during adulthood, the bulk of Americans will face impoverishment. The approach of emphasizing individual failings or attributes as the primary cause of poverty loses much of its explanatory power in the face of such patterns. Rather, given the widespread occurrence of economic vulnerability, a life span analysis points to a third line of evidence indicating that poverty is more appropriately viewed as a structural failing of American society. As C. Wright Mills notes in his analysis of unemployment,

 When, in a city of 100,000, only one man is unemployed, that is his
 personal trouble, and for its relief we properly look to the
 character of the man, his skills, and his immediate opportunities.
 But when in a nation of 50 million employees, 15 million men are
 unemployed, that is an issue, and we may not hope to find its
 solution within the range of opportunities open to any one
 individual. The very structure of opportunities has collapsed. Both
 the correct statement of the problem and the range of possible
 solutions require us to consider the economic and political
 institutions of the society, and not merely the personal situation
 and character of a scatter of individuals (1959: 9).

To summarize, three lines of evidence have been detailed suggesting that American poverty is primarily the result of structural conditions. These include a lack of sufficient paying jobs in the labor market, the ineffectiveness of America's social safety net to pull individuals and families out of poverty, and the fact that a clear majority of Americans will experience poverty at some point during their adulthood years. Other structural failings could have been explored as well (e.g., the inequities in educational quality in the United States, the systematic lack of political power for the economically disenfranchised, or the widespread patterns of racial residential segregation, to name but a few). Nevertheless, the lines of evidence discussed in this section would appear particularly illuminating in revealing the structural nature of American poverty.

Discussion

Given the above arguments and evidence indicating that American poverty is primarily the result of structural failings, how might we reconcile this perspective with the earlier discussed research findings indicating that human capital and individual attributes largely explain who is at risk of experiencing poverty? An approach that bridges the empirical importance of individual attributes with the significance of structural forces has been the concept of structural vulnerability (Rank, 1994; 2000; 2001; 2004). This framework recognizes that human capital and other labor market attributes are associated with who loses out at the economic game (and hence will be more likely to experience poverty), but that structural factors predominately ensure that there will be losers in the first place.

An analogy can be used to illustrate the basic concept. Imagine a game of musical chairs in which there are ten players but only eight chairs. On one hand, individual success or failure in the game depends on the skill and luck of each player. Those who are less agile or less well placed when the music stops are more likely to lose. These are appropriately cited as the reasons a particular individual has lost the game. On the other hand, given that there are only eight chairs available, two players are bound to lose regardless of their characteristics. Even if all the players were suddenly to double their speed and agility, there would still be two losers. From this broader context, the characteristics of the individual players are no longer important in terms of understanding that the structure of the game ensures that someone must inevitably lose.

We would argue that this analogy applies with respect to poverty. For every ten American households, there are good jobs and opportunities at any point in time to adequately support roughly eight of those ten. The remaining two households will be locked out of such opportunities, often resulting in poverty or near poverty. Individuals experiencing such economic deprivation are likely to have characteristics putting them at a disadvantage in terms of competing in the economy (lower education, fewer skills, single-parent families, illness or incapacitation, minorities residing in inner cities, etc.). These characteristics help to explain why particular individuals and households are at a greater risk of poverty.

However, given the earlier discussed structural failures, a certain percentage of the American population will experience economic vulnerability regardless of what their characteristics are. As in the musical chairs analogy, increasing everyone's human capital will do little to alter the fact that there are only so many decent paying jobs available. In such a case, employers will simply raise the bar in terms of the employee qualifications they are seeking, but nevertheless there will remain a percentage of the population at risk of economic deprivation. Consequently, although a lack of human capital and its accompanying vulnerability leads to an understanding of who the losers of the economic game are likely to be, the structural components of our economic, social, and political systems explain why there are losers in the first place.

The critical mistake that has been made in the past has been that social scientists have frequently equated the question of who loses out at the game, with the question of why the game produces losers in the first place. They are, in fact, distinct and separate questions. While deficiencies in human capital and other marketable characteristics help to explain who in the population is at a heightened risk of encountering poverty, the fact that poverty exists in the first place results not from these characteristics, but from the lack of decent opportunities and supports in society. By focusing solely upon individual characteristics, such as education, we can shuffle people up or down in terms of their being more likely to land a job with good earnings, but we are still going to have somebody lose out if there are not enough decent paying jobs to go around. In short, we are playing a large scale version of a musical chairs game with ten players but only eight chairs.

The recognition of this dynamic represents a fundamental shift in thinking from the old paradigm. It helps to explain why the social policies of the past two decades have largely been ineffective in reducing the rates of poverty. We have focused our attention and resources on either altering the incentives and disincentives for those playing the game, or in a very limited way, upgrading their skills and ability to compete in the game, while at the same time leaving the structure of the game untouched.

When the overall poverty rates in the United States do in fact go up or down, they do so primarily as a result of impacts on the structural level that increase or decrease the number of available chairs. In particular, the performance of the economy has been historically important. Why? Because when the economy is expanding, more opportunities (or chairs) are available for the competing pool of labor and their families. The reverse occurs when the economy slows down and contracts. Consequently, during the 1930's or early 1980's when the economy was doing badly, poverty rates went up, while during periods of economic prosperity such as the 1960's or the middle to later 1990's, the overall rates of poverty declined.

Similarly, changes in various social supports and the social safety net available to families will make a difference in terms of how well such households are able to avoid poverty or near poverty. When such supports were increased through the War on Poverty initiatives in the 1960's, poverty rates declined. Likewise, when Social Security benefits were expanded during the 1960's and 1970's, the elderly's poverty rates declined precipitously. Conversely, when social supports have been weakened and eroded, as in the case of children's programs over the past 25 years, their rates of poverty have gone up.

The recognition of poverty as a result of the way the game is structured also makes it quite clear why the United States has such high rates of poverty when compared to other Western countries. These rates have nothing to do with Americans being less motivated or less skilled than those in other countries, but with the fact that our economy has been producing a plethora of low wage jobs in the face of global competition and that our social policies have done relatively little to support families compared to our European neighbors. From this perspective, one of the keys to addressing poverty is to increase the labor market opportunities and social supports available to American households.

The structural vulnerability perspective thus recognizes the importance of human capital in being able to predict who is more likely to experience economic deprivation, while at the same time emphasizing the importance of structural constraints in guaranteeing that some Americans will be left out of the economic mainstream. In short, the structure of the American economy, accompanied by a weak social safety net and public policies directed to the economically vulnerable, ensure that a certain percentage of the American population will experience impoverishment at any point in time, and that a much larger percentage of the population will experience poverty over the course of a lifetime. The fact that three quarters of Americans will experience poverty or near poverty (at the 1.50 level) during their adulthoods is emblematic of these structural level failings.

As noted at the beginning of this article, social scientists investigating poverty have largely focused upon individual deficiencies and demographic attributes in order to explain the occurrence of poverty in America. As such, they have reinforced the mainstream American ethos of interpreting social problems as primarily the result of individual failings. In addition, a culture of poverty perspective has occasionally been used to explain the occurrence of poverty within specific geographical settings such as inner cities or remote rural areas. This approach also tends to largely place the dynamic of poverty within the framework of individual deficiencies.

We have argued and attempted to demonstrate in this article that such perspectives are misguided. Whereas individual attributes (such as human capital) help to explain who faces a greater risk of experiencing poverty at any point in time, the fact that substantial poverty exists on a national level can only be understood through an analysis of the structural dynamics of American society.

Table 1
Inability of the Labor Market to Support Various Family Structures
Above Different Poverty Thresholds

Current Family Status

Poverty All Married Couple One Parent Single
Threshold Families Families Families Families

Heads of Families Working Full-Time
Below 1.00 9.4 9.9 16.9 3.2
Below 1.25 15.3 16.0 26.4 6.6
Below 1.50 22.0 23.0 36.6 10.2
N 9,891 6,053 1,557 2,281

Heads of Families Working Half-Time or More
Below 1.00 14.9 13.5 27.7 9.5
Below 1.25 21.4 19.8 37.0 14.3
Below 1.50 28.0 26.7 46.5 18.3
N 11,312 6,623 1,969 2,720

Heads of Families in the Labor Market
Below 1.00 20.3 17.1 36.8 15.8
Below 1.25 26.5 23.3 44.9 20.4
Below 1.50 32.7 30.0 53.4 24.2
N 12,190 6,972 2,259 2,959

Source: Survey of Income and Program Participation, authors'
computations

Table 2

Inability of the Labor Market to Support Various Family Sizes Above
Different Poverty Thresholds

 Hypothetical Family Size

 1 2 3 4 5 6
Poverty Person Person Person Person Person Person
Threshold Family Family Family Family Family Family

Heads of Families Working Full-Time
Below 1.00 2.4 4.7 7.6 15.0 22.3 29.0
Below 1.25 4.3 8.9 14.0 25.1 35.0 42.4
Below 1.50 7.2 14.5 21.8 36.0 46.3 54.6

Heads of Families Working Half-Time or More
Below 1.00 6.3 10.2 13.8 21.6 28.9 35.3
Below 1.25 9.6 15.2 20.6 31.6 41.0 47.9
Below 1.50 13.2 21.1 28.3 42.0 51.6 59.3

Heads of Families in the Labor Market
Below 1.00 12.0 15.8 19.3 26.9 33.7 39.7
Below 1.25 15.3 20.8 25.9 36.3 45.0 51.5
Below 1.50 18.8 26.4 33.2 46.0 54.9 62.1

Source: Survey of Income and Program Participation, authors'
computations

Table 3

Comparative Analysis of Governmental Effectiveness in Reducing
Poverty Across Selected Countries

 Pre-transfer Post-transfer Reduction
Country Poverty Rates Poverty Rates Factor

Canada (1994) 29 10 66
Finland (1995) 33 4 88
France (1994) 39 8 79
Germany (1994) 29 7 76
Netherlands (1994) 30 7 77
Norway (1995) 27 4 85
Sweden (1995) 36 3 92
United Kingdom (1995) 38 13 66
United States (1994) 29 18 38

Source: Luxembourg Income Study, adapted from Veli-Matti Ritakallio
(2001) computations

Table 4

The Cumulative Percent of Americans Experiencing Poverty Across
Adulthood

 Level of Property

Age Below 1.00 Below 1.25 Below 1.50
 Poverty Line Poverty Line Poverty Line

20 10.6 15.0 19.1
25 21.6 27.8 34.3
30 27.1 34.1 41.3
35 31.4 39.0 46.9
40 35.6 43.6 51.7
45 38.8 46.7 55.0
50 41.8 49.6 57.9
55 45.0 52.8 61.0
60 48.2 56.1 64.2
65 51.4 59.7 67.5
70 55.0 63.6 71.8
75 58.5 68.0 76.0

Source: Panel Study of Income Dynamics, authors' computations

References

Bane, M. J., & Ellwood, D. T. (1986). Slipping into and out of poverty: The dynamics of spells. Journal of Human Resources, 21, 1-23.

Bane, M. J., & Ellwood, D. T. (1994). Welfare realities: From rhetoric to reform. Cambridge, MA: Harvard University Press.

Bartik, T. H. (2001). Jobs for the poor: Can labor demand policies help? New York: Russell Sage Foundation.

Bartik, T. H. (2002). Poverty, jobs, and subsidized employment. Challenge, 45, 100-111.

Blank, R. M. (1997). It takes a nation: A new agenda for fighting poverty. Princeton, NJ: Princeton University Press.

Boisjoly, J., Harris, K. M., & Duncan, G. J. (1998). Trends, events, and duration of initial welfare spells. Social Service Review, 72, 466-492.

Booth, C. (1892-97). Life and labour of the people of London, first series: Poverty. London: Macmillan.

Devine, J. A., & Wright, J. D. (1993). The greatest of evils: Urban poverty and the American underclass. New York: Aldine De Gruyter.

DuBois, W. E. B. (1899). The Philadelphia Negro. Philadelphia: University of Pennsylvania Press.

Duncan, G. J. (1984). Years of poverty, years of plenty: The changing economic fortunes of American workers and families. Ann Arbor, MI: Institute for Social Research.

Duncan, G. J., Gustafsson, B., Hauser, R., Schmaus, G., Jenkins, S., Messinger, H., Muffels, R., Nolan, B., Ray, J., & Voges, W. (1995). Poverty and social-assistance dynamics in the United States, Canada, and Europe. In K. McFate, R. Lawson, & W. J. Wilson (Eds.), Poverty, inequality and the future of social policy: Western states in the new world order (pp. 67-108). New York: Russell Sage Foundation.

Edwards, M. E., Plotnick, R., & Klawitter, M. (2001). Do attitudes and personality characteristics affect socioeconomic outcomes? The case of welfare use by young women. Social Science Quarterly, 82, 827-843.

Esping-Andersen, G. (1990). The three worlds of welfare capitalism. Princeton, NJ: Princeton University Press.

Feagin, J. R. (1975). Subordinating the poor: Welfare and American beliefs. Englewood Cliffs, NJ: Prentice-Hall.

Feagin, J. R. (2000). Racist America: Roots, current realities, and future reparations. New York: Routledge.

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

Gilens, M. (1999). Why Americans hate welfare: Race, media, and the politics of antipoverty policy. Chicago: University of Chicago Press.

Goodwin, L. (1972). Do the poor want to work? A social-psychological study of work orientations. Washington, DC: Brookings Institution.

Goodwin, L. (1983). Causes and cures of welfare: New evidence on the social psychology. Lexington, MA: Lexington Books.

Harrington, M. (1962). The other America: Poverty in the United States. New York: Macmillan.

Harris, K. M. (1996). Life after welfare: Women, work, and repeat dependency. American Sociological Review, 61, 407-426.

Hartwell, R. M. (1986). The long debate on poverty. Paper presented at the Political Economy Seminar Series, Washington University, St. Louis, Missouri.

Hill, M. S. (1992). The Panel Study of Income Dynamics: A user's guide. Newbury Park, CA: Sage Publications.

Hull House (1895). Hull House maps and papers. New York: Thomas Y. Crowell.

Hunter, R. (1904). Poverty. New York: Macmillan.

Katz, M. B. (1989). The undeserving poor: From the war on poverty to the war on welfare. New York: Pantheon Books.

Kluegel, J. R., & Smith, E. R. (1986). Beliefs about inequality: Americans' views of what is and what ought to be. New York: Aldine de Gruyter.

Lichter, D. T., & Crowley, M. L. (2002). Poverty in America: Beyond welfare reform. Population Bulletin, 57, 1-36.

Luxembourg Income Study (2000). LIS quick reference guide. Syracuse, NY: Maxwell School of Citizenship and Public Affairs, Syracuse University.

Massey, D. S. (1996). The age of extremes: Concentrated affluence and poverty in the twenty-first century. Demography, 33, 395-412.

Maynard, R. A. (1997). Kids having kids: Economic costs and social consequences of teen pregnancy. Washington, DC: The Urban Institute Press.

Mills, C. W. (1959). The sociological imagination. New York: Oxford University Press.

Moffitt, R. (1992). Incentive effects of the U.S. welfare system: A review. Journal of Economic Literature, 30, 1-61.

Noble, C. (1997). Welfare as we knew it: A political history of the American welfare state. New York: Oxford University Press.

O'Connor, A. (2001). Poverty knowledge: Social science, social policy, and the poor in twentieth-century U.S. history. Princeton: Princeton University Press.

Organization for Economic Cooperation and Development. (1999). Social expenditure database 1980-1996. Paris: Organization for Economic Cooperation and Development.

Patterson, J. T. (2000). America's struggle against poverty in the twentieth century. Cambridge, MA: Harvard University Press.

Pavetti, L. A. (1992). The dynamics of welfare and work: Exploring the process by which young women work their way off welfare. Paper presented at the APPAM Annual Research Conference, October, Denver, Colorado.

Rank M. R. (1988). Racial differences in length of welfare use. Social Forces, 66, 1080-1101.

Rank, M. R. (1994). Living on the edge: The realities of welfare in America. New York: Columbia University Press.

Rank, M. R. (2000). Poverty and economic hardship in families. In D. H. Demo, K. R. Allen, & M. A. Fine (Eds.), Handbook of family diversity (pp. 293-315). New York: Oxford University Press.

Rank, M. R. (2001). The effect of poverty on America's families: Assessing our research knowledge. Journal of Family Issues 22, 882-903.

Rank, M. R. (2004). One nation, underprivileged: Why American poverty affects us all. New York: Oxford University Press.

Rank M. R., & Hirschl, T. A. (1999a). The likelihood of poverty across the American adult life span. Social Work, 44, 201-216.

Rank, M. R., & Hirschl, T. A. (1999b). Estimating the proportion of Americans ever experiencing poverty during their elderly years. Journal of Gerontology: Social Sciences, 54B, S184-S193.

Rank, M. R., & Hirschl, T. A. (1999c). The economic risk of childhood in America: Estimating the probability of poverty across the formative years. Journal of Marriage and the Family, 61, 1058-1067.

Rank, M. R., & Hirschl, T. A. (2001a). The occurrence of poverty across the life cycle: Evidence from the PSID. Journal of Policy Analysis and Management, 20, 737-755.

Rank, M. R., & Hirschl, T. A. (2001b). Rags or riches? Estimating the probabilities of poverty and affluence across the adult American life span. Social Science Quarterly, 82, 651-669.

Rank, M. R., & Hirschl, T. A. (2001c). The measurement of long term risks across the life course. Social Science Quarterly, 82, 680-686..

Ritakallio, V. M. (2001). Trends of poverty and income inequality in cross-national comparison. Luxembourg Income Study Working Paper No. 272, Maxwell School of Citizenship and Public Affairs, Syracuse University, Syracuse, New York.

Rowntree, B. S. (1901). Poverty: A study of town life. London: Macmillan.

Ryan, W. (1971). Blaming the victim. New York: Pantheon Books.

Sandefur, G. D., & Cook, S. T. (1998). Permanent exits from public assistance: The impact of duration, family, and work. Social Forces, 77, 763-786.

Schiller, B. R. (2004). The economics of poverty and discrimination. Upper Saddle River, NJ: Prentice Hall.

Seccombe, K. (1999). So you think I drive a Cadillac? Welfare recipients' perspectives on the system. Needham Heights, MA: Allyn and Bacon.

Seccombe, K. (2000). Families in poverty in the 1990s: Trends, causes, consequences, and lessons learned. Journal of Marriage and the Family, 62, 1094-1113.

Smeeding, T. M., Rainwater, L., & Burtless, G. (2000). United States poverty in a cross-national context. Luxembourg Income Study Working Paper, No. 244, Maxwell School of Citizenship and Public Affairs, Syracuse University, Syracuse, New York.

Stevens, A. H. (1994). The dynamics of poverty spells: Updating Bane and Ellwood. American Economic Review, 84, 34-37.

Stevens, A. H. (1999). Climbing out of poverty, falling back in: Measuring the persistence of poverty over multiple spells. The Journal of Human Resources, 34, 557-588.

U.S. Bureau of the Census (2001). Statistical Abstract of the United States: 2001 (121th edition.) Washington, DC: U.S. Government Printing Office.

U.S. Bureau of the Census (2003a). Poverty in the United States: 2002. Current Population Reports, Series P60-222. Washington, DC: U.S. Government Printing Office.

U.S. Bureau of the Census (2003b). Health Insurance Coverage: 2002. Current Population Reports, Series P60-223. Washington, DC: U.S. Government Printing Office.

Walker, R. (1994). Poverty dynamics: Issues and examples. Aldershot, England: Avebury.

Westat (2001). Survey of Income and Program Participation Users' Guide, Third Edition. Rockville, MD: Westat.

MARK R. RANK

Washington University

George Warren Brown School of Social Work

HONG-SIK YOON

Chonbuk National University

Department of Social Welfare

THOMAS A. HIRSCHL

Cornell University

Department of Rural Sociology