فقر کارگران در جنوب کالیفرنیا: به سوی یک اقدام عملیاتی

As a result of welfare-to-work, discussions on working poverty have sprouted among urban scholars and policy makers. Yet, there is little consensus on how to define working poverty, limiting both knowledge of such populations and ability to inform policy. Using 1998–2000 CPS data for southern California, we compare twelve definitions of work and poverty in an attempt to provide a better understanding of who the working poor are and suggest an operational measure of working poverty that is both empirically simple and realistic. Results indicate that working poor families represent a diverse group of people, which varies significantly depending upon the definitions used. However, as expected, we find a dominance of families with young children and fewer adult workers, most of whom are Latino, work in service or laborer occupations, have lower educational attainment and very low unionization rates, and earn hourly wages close to the legal minimum.

The last few years have witnessed a blooming of research on “working poverty”—the condition of people who are working and earning an income, but fail to earn enough to keep themselves and their family out of poverty (Acs et al., 2000; Boushey et al., 2001; Elwood, 1999; Iceland, 2000; Kim, 1998; LAANE, 2000; Wertheimer, 1999). This expansion is no doubt related to the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA) that promoted the entry of thousands of welfare recipients into the labor force. These “work first” reforms led a number of scholars and policy advocates to question whether former welfare recipients were improving their economic conditions or merely entering working poverty. The growth of studies on working poverty is also linked to the rising evidence that the benefits of the 1990 economic expansion have not been equally shared among wage earners, and until recently, had not led to declines in overall poverty. This was especially true in the first half of the decade when workers at the bottom 20% of the wage scale earned approximately 10 times less than those in the top 20%, and actually saw their real wages continuously decline by an average of 0.9% a year (Mishel et al., 2001). Existing studies however lack a common definition of who the working poor are. Because different estimation methodologies are employed, results often seem contradictory and comparisons between geographical areas and time periods are difficult. The resulting confusion has hampered the development of policies such as raising the minimum wage, enforcing a living-wage, expanding the earned-income tax credit, or improving child care, transportation, health and housing services to working families. The purpose of this paper is to systematically compare alternative definitions of working poverty in an attempt to adopt an operational definition that is both relatively simple to implement empirically and accurate in capturing the economic deprivation of the poor. While this is unlikely to end the debate regarding the definition of working poverty, it provides researchers with the tools to understand differences in results of existing studies as well as suggests an operational measure for future research. In the process, it also generates current estimates of the number and characteristics of working poor in southern California, the site of our analysis. In the next section we discuss the data used in this paper and the context of our analysis. In the following section, we review existing definitions of working poverty and compare estimation methodologies relying on alternative definitions of work and poverty. We then turn to the empirical comparison of working poor populations and identification of key predictors of working poverty based on 12 different resulting definitions. We conclude with suggestions for future research.

Working poverty has become a growing concern in a nation where inequality remains high despite lasting economic growth in the 1990s, and welfare reform has required poor people to find employment. However, although most experts agree that the official poverty measure fails to capture many people in need, there is little agreement on how to measure working poverty systematically. To address this question, we investigated definitions of work and poverty, and compared twelve alternative measures of working poverty. Not only did this exercise shed some light over methodological issues, it also provided a detailed picture of working poverty in the southern California region. Our results indicate that the type of measure employed matters. For instance, in southern California, estimates of people in full-time working poor families range between 500 thousand and almost 4.5 million depending on which measure is used. Moreover, as expected, the characteristics of working poor families vary and become more diverse as the poverty threshold rises. Overall, full-time working poor families are more likely to include children (especially young children), have a foreign-born and/or Latino main income-earner who works about 2100 h per year, earns an hourly-wage close to the legal minimum, has a high school education or less, and works in a service or laborer occupation. However, as our analysis suggests, many other types of families and workers live in poverty, especially when we use measures of poverty and work that more closely reflects real life costs of living and work patterns. Deciding which measure to use implies a judgment regarding the meaning of poverty and work. Both quantitative and qualitative evidence suggest that the economic hardships faced by lower middle-class families are real and often left unanswered by public policy. Because it fails to identify this group as poor—over 3 million people in southern California alone—the federal poverty line is clearly inappropriate both as a measurement and as a foundation for social policy. However, given fiscal budget constraints, it may be neither politically feasible nor efficient to target anti-poverty policies to families with income as high as those identified by the most comprehensive measures. For instance, it is difficult to sustain the argument that a third of working families live in poverty or that almost half of the population is poor. As often, a middle ground seems appropriate. The NAS experimental measure, HUD’s relative income approach, and the 150% adjusted Census threshold all provide similar results. Although the first measure is clearly more sophisticated and more appropriate for national or inter-regional studies, the last one is easier to use and makes practical sense in a regional study like this one. The relative definition of work based the number of potential workers per family also seems more realistic than the absolute threshold of 1885 h which only makes sense in the case of traditional families with one full-time worker. While these preliminary results for southern California suggest that a relative measure of work combined with an adjusted official Census poverty threshold may be both a simple and realistic way to measure working poverty, further research is necessary to test the validity of such measures at the national level. For instance, the relative definition of work needs to be streamlined by comparing work patterns by family size in different regions. Similarly, the 150% adjustment may not be as closely related the NAS measure in other parts of the country where costs of living are lower and the adjusted measure may overestimate the number of people in economic hardship. Adopting these measures has important implications for anti-poverty policy. Raising the number of people considered to be poor would require an adjustment of existing policies and probably an expansion of resources allocated to anti-poverty efforts. Thus, it is critical to adopt a realistic and politically sound measure. Further research is necessary to estimate the policy and political implications of broadening the definition of poverty and to identify the most effective policies to address the problems of the working poor. The analysis above suggests at least three factors that deserve special attention in the design of policies to reduce working poverty: the socio-economic barriers faced by immigrants and Latinos (the severity of which depends on the rate of upward mobility), the costs of supporting children, and the disproportionately low wages offered by large segments of the regional economy. While tax-based policies such as the Earned Income Tax Credit and child care expenses deductions play an important role in alleviating the burden of the working poor, such programs must be expanded to effectively lift working families out of poverty. Moreover, even with some expansion, these programs are unlikely to successfully reduce working poverty if wages remain at their current low level in many jobs, affordable childcare solutions are lacking, and access to good jobs is limited by geographical and social constraints. To lift these barriers, we need widespread enforcement of minimum wage regulations and living wage ordinances, development of subsidized child care for low-income families, and increased mobility of the poor through affordable housing, public transportation, and job placement and training programs. Because working poverty, even if concentrated in specific areas, affects the entire region’s potential for economic growth, its political and social stability, and its environmental sustainability, policies must be developed at the regional level. Acknowledgements This research is made possible by a grant from the National Science Foundation, Program in Geography and Regional Science (Award Number 0112475). The author thank Udaya Wagle for research assistance, and Jennifer Wolch, Enrico Marcelli, and anonymous reviewers for comments and suggestions.