By Marianne Page
PDF version of this article (with charts) can be found HERE
Here’s a common mantra: The only enduring solutions to poverty are economic growth and the jobs it delivers. Although the mantra is delivered especially frequently in the case of less developed countries, it’s also sometimes advanced as a poverty-reduction recipe for more developed ones like the United States. If the mantra were true, it would mean that we’d be well advised to focus all of our policy efforts on growing the economy and increasing employment opportunities, thus allowing us to treat more focused, poverty-specific policies merely as temporary stopgaps.
The purpose of this article is to evaluate whether a simple pro-jobs policy of this sort would reduce poverty in the United States as much as we’d like. In carrying out this evaluation, a natural starting point is to examine the empirical association between labor market conditions and poverty. After all, if it is established that the relationship between poverty and employment opportunities is not all that strong in the United States, then providing more jobs would not likely be a viable solution to poverty.
I begin by discussing how the jobs-poverty relationship has been weakening in recent decades, due in part to ongoing changes in (a) the types of jobs that our economy is creating and (b) the sectors of the labor market that are positioned to secure these jobs. After laying out these changes, I’ll discuss their implications for crafting antipoverty policy that works.
The Empirical Relationship Between Jobs and Poverty
It is well known that economic downturns increase poverty. Jobs disappear, working hours are cut, and wages fall. This is especially true at the bottom of the income distribution. The very groups that, even in the best of times, are close to the poverty line—blacks, Hispanics, young people, and the less educated— are those that tend to suffer most during recessions . During the Great Recession, for example, the poverty rate of children increased more than the rate of any other age group. This is because children typically live with younger adults, who, as a result of their relative inexperience, tend to be among the first to lose their jobs during mass layoffs.
Unless safety-net programs fully replace lost income, an across-the-board rise in unemployment will mechanically increase the number of people who are poor. Figure 1 shows the close relationship between the economy’s overall health, as measured by unemployment, and the poverty rate. The correlation between changes in unemployment and changes in poverty is 0.65.
But is the strength of this relationship changing over time? Is aggregate job growth becoming a less effective lever on poverty?
Indeed it is. Figure 2, which graphs the change in the poverty rate against the change in the employment rate (for adults aged 25–54), shows that since the 1980s there has been a weakening in the jobs-poverty relationship. Recent labor market expansions, though similar in both magnitude and duration to the 1960s expansions, do not cut poverty as much as we’d come to expect. From 1962 to 1969, employment grew 4.7 percentage points, and poverty fell 9.8 points, more than twice the employment growth. In contrast, during the mid-1980s, despite significant labor market expansion, poverty fell far less.
Why Aren’t Jobs Delivering?
In understanding why the relationship between the employment rate and poverty is weakening, it’s useful to lay out the parameters that are relevant to the strength of this relationship, parameters that pertain to the types of jobs that are available, the capacity of the low-skill labor force to acquire these jobs, and its capacity to exit poverty through means other than work. These parameters are (a) the availability of unconditional benefits (benefits that are not conditioned on employment), (b) the availability of skill-compatible employment opportunities, (c) the extent to which the available jobs provide adequate wages, and (d) the extent to which these jobs come with other employment-conditioned benefits (e.g., Earned Income Tax Credit) that may compensate for low wages. For each of these conditions, I will lay out the relevant changes and their implications for the strength of the employment-poverty relationship. This discussion is summarized in Table 1.
Unconditional benefits: If nonworking families can acquire benefits that are not conditioned on work, then there’s a road out of poverty that does not require jobs or a booming economy. That is, when unconditional benefits are widely available, the macro-level relationship between jobs and poverty will be weakened.
The main development in this regard is the rise and fall of Aid to Families with Dependent Children (AFDC). The growing prevalence of cash welfare benefits in the form of AFDC mitigated the impact of downturns after the 1960s. Just as AFDC reduced poverty during downturns, poverty did not have as far to fall when the downturn ended and the economy turned around.
However, with the elimination of AFDC in 1996, the correlation between the employment rate and poverty should have strengthened. While AFDC provided cash benefits to lowincome and primarily single-parent families with children, the new Temporary Assistance to Needy Families (TANF) program imposed strict work requirements and sanctions for non-compliance, making it harder to obtain when jobs are scarce. In short, the countercyclical effect of cash welfare use has been reduced, meaning that the ability to rely on cash welfare during recessions has declined. As a result, relative to the pre-TANF era, we expect poverty to rise more during economic downturns and to fall more during upturns. Because we haven’t observed this pattern, it suggests that other forces must be in play that counteract this expected effect.
Skill compatibility: Why, then, is job growth reducing poverty less than it once did? It’s partly because the economy is not delivering the types of jobs that poor people can fill. As David Autor has shown, most of the job growth since the late 1980s has occurred within either the low-skill or high-skill sectors, with a consequent hollowing out of opportunities in the middle . One reason is that technological advances have led to the automation of (and ultimately to the displacement of) many jobs that involve “routine” tasks. Manufacturing jobs, which used to provide opportunities for workers with moderate levels of education (such as a high school diploma), have sharply declined. The Great Recession has exacerbated this trend, as employment losses have been most severe in middle-skill jobs, both in the white-collar and blue-collar sectors. The higher prevalence of jobs at the bottom should help the poor, but what’s unclear is whether the associated hollowing out in the middle is a countervailing force that increases the competition between the poor and those who had before secured middle-class jobs. All else being equal, this competition may increase unemployment at the bottom of the labor market or lower wages among those who do get jobs .
Wage adequacy: Even if a low-skill job is acquired, it won’t be poverty-reducing unless it delivers enough in the way of wages (or transfers) to push the recipient over the poverty threshold . Over the last 40 years, the wages of low-skill jobs have been stagnant for a number of reasons, including, for example, the declining real value of the minimum wage. Between 1975 and 1995, the 20th percentile of the weekly wage distribution declined from $473 to $386, resulting in fewer jobs that provided an above-poverty wage. Recent studies have shown that a $100 reduction in the real weekly wage among workers in the bottom 20 percent of the income distribution reduces the annual probability of escaping poverty by about 15 percent . The declining payoff to work could also reduce the incentive to work at all, which may in turn lead to a deterioration of skills, further reducing the likelihood of escaping poverty.
Conditional benefits: The Earned Income Tax Credit (EITC) does of course supplement low wages and should thereby raise people out of poverty. The EITC, established in 1975, provides a tax-based earnings subsidy to low-income workers, which increases the income of low-earners and could counteract a decline in the minimum wage or any decline in wages that accompanies a recession. Because the generosity of the EITC expanded significantly during the early 1990s, one might expect that, over time, the relationship between labor market opportunities and poverty would have strengthened at the macro level, rather than weakened. However, although EITC subsidies have a significant effect on the number of families whose total income falls below the poverty threshold, the EITC does not directly affect the official poverty rate, because EITC income is not counted as “money income before taxes.” This measurement artifact helps explain why the official poverty rate changed so little through the mid-2000s despite the EITC’s expansion.
Moreover, as Figure 2 shows, poverty had already become less responsive to economic growth even before the EITC became more generous in 1993. The overall employment growth of 6.2 percentage points during the 1980s was accompanied by a poverty reduction of just 2.4 percentage points, far shy of the 9.8 point reduction in the 1960s.
It follows that the weakening in the aggregate employment poverty relationship is probably driven by (a) the shortage of low-skill jobs relative to the supply of workers competing for such jobs, and (b) the relatively low earning power of the available low-skill jobs. In the following section, I comment on the policy implications of this change in the employment-poverty relationship, with a particular focus on its implications for policies that seek to reduce poverty by increasing employment.
What’s to Be Done?
An antipoverty policy that focuses on jobs and employment will need to be targeted to the current employment regime if it is to have any payoff. A simple policy of “more jobs” has become a less viable poverty solution, but there may be a package of more targeted policies that, taken together, could have substantial poverty-reducing effects.
The first, and especially important, part of this package is to promote wage growth within the low-skill sector. This might be done by increasing the minimum wage, further increasing the EITC, or through other interventions in the labor market such as skill-enhancing training programs. The second part of this package is a strong unemployment insurance (UI) system, which plays a critical role in reducing poverty associated with recessions because it provides temporary partial-wage replacement to involuntarily unemployed workers, many of whom have incomes near the poverty line. Indeed, because the rate at which UI replaces earnings varies (negatively) with earnings, UI provides relatively greater protection to low-wage workers. In most states and years, UI benefits can be received for a maximum of 26 weeks, but during the most recent recession Congress enacted emergency extensions that increased benefits in most states to 99 weeks. These UI benefits make it possible for families to maintain their prior levels of food consumption (an important determinant of well-being) in the aftermath of a job loss.
The third and final part of this three-pronged package is the continued use of nutrition assistance (SNAP) and other non-cash safety-net programs. These programs have always been sensitive to the business cycle and have become significantly more responsive to economic cycles in the wake of welfare reform. According to recent studies, SNAP benefits have become especially useful in reducing the adverse income impacts of recessions after welfare reform. When poverty measures include SNAP benefits as income, poverty rates are much lower. For example, the 2009 poverty rate would have been 7.7 percentage points lower if SNAP benefits had counted as income . Although we do not know whether they are as effective as straight-on cash assistance to the poor, we do know that new countercyclical programs, like UI and SNAP, have become critical poverty-mitigation programs in the current economic regime.
This combination of policies would acknowledge, in a real way, the weakening of the employment-poverty relationship. Will the policies themselves affect the strength of that relationship? They very likely will, but sometimes in opposing ways. That is, some of the proposed policies (e.g., more generous wage subsidies) serve to strengthen that relationship, while others work by providing benefits that are not conditional on having a job (e.g., extended unemployment insurance and a preserved SNAP program) and hence will serve to weaken the employment-poverty relationship. However, by keeping the unemployed out of poverty during downturns, both UI and SNAP help to maintain family well-being in the low-skill sector, which may increase employment and reduce poverty in the long run.
PDF version of this article (with charts) can be found HERE
About the Author
Marianne Page is a professor of economics at the University of California at Davis and the deputy director of the Center for Poverty Research. Her research examines the sources of inter-generational mobility and the impact of social programs.
. Danziger, S., Chavez, K., & Cumberworth, E. 2012. “Poverty and the Great Recession.” Stanford, CA: Stanford Center on Poverty and Inequality; Hout, M. & Cumberworth, E. 2012. “The Labor Force and the Great Recession.” Stanford, CA: Stanford Center on Poverty and Inequality.
. Autor, D. 2010. “The Polarization of Job Opportunities in the U.S. Labor Market: Implications for Employment and Earnings.” Center for American Progress Working Paper. http:// http://www.brookings.edu/research/papers/2010/04/ jobs-autor.
. Potential scarring from long-term unemployment adds to the complicated nature of the jobs-poverty relationship. Not only must the poor have the skills necessary for the available jobs, but they might also lose in the labor market to the extent that employers perceive low job readiness among the long-term unemployed.
. Blank, R. M. 1993. “Why Were Poverty Rates So High in the 1980s?” In Papadimitrious, Dimitri B. and Edward N. Wolff (eds.) Poverty and Prosperity in the U.S. in the Late Twentieth Century. London: Macmillan Press.
. Stevens, A. H. 2012. “Poverty Transitions.” In Philip N. Jefferson (ed.). The Oxford Handbook of the Economics of Poverty. Oxford: Oxford University Press.
. Tiehen, L., Jolliffe, D., & Gundersen, C. Alleviating Poverty in the United States: The Critical Role of SNAP Benefits, ERR-132, U.S. Department of Agriculture, Economic Research Service, April 2012.