Article

Cities Matter: Shifting the Focus of Welfare Reform

Bruce Katz and Katherine Allen

Discussions of welfare reform almost always begin—and sometimes end—by citing the remarkable halving of the national welfare caseload since 1994. Much less frequently noted is that the decline in the caseload has not been uniform nationwide. Most strikingly, the decline has been slowest in the nation’s urban areas. As a result, the welfare rolls have increasingly concentrated in the cities. Today nearly 60 percent of all welfare cases can be found in 89 large urban counties.

When Congress passed Temporary Assistance for Needy Families (TANF) in 1996, there was much talk of how devolving responsibility for welfare would allow states to tailor their reform strategies to match their own particular needs and circumstances. As reform proceeds, the focus must shift from the states to the places where reform is most urgently needed.

Where Is Welfare?

Since 1997, the Brookings Center on Urban and Metropolitan Policy has tracked welfare caseload declines in the nation’s largest cities and urban counties. In our latest study, released last July, we found that while urban welfare caseloads are indeed falling rapidly—faster than most welfare experts predicted at the start of welfare reform—they are declining more slowly than national caseloads. In the 89 urban counties that contain the 100 largest American cities, the aggregate caseload fell 41 percent from 1994 to 1999—more than 10 percentage points short of the 52 percent drop in the national rate. The national caseload has declined by roughly 2.6 million welfare cases since peaking at 5 million in 1994, whereas the aggregate urban county caseload has fallen by almost 1 million cases from a 1994 total of 2.4 million.

As a result of these slower declines, the nation’s welfare cases are becoming more concentrated in urban areas. In 1999, the 89 top urban counties, with 33 percent of total U.S. population, accounted for 58 percent of the nation’s welfare cases, up 10 percentage points since 1994. Ten urban counties now account for roughly one-third of all U.S. welfare cases. Clearly, urban areas are at the epicenter of welfare reform.

State and Regional Variations

What explains this spatial concentration? State policy accounts for some of it. While caseloads have generally fallen at different rates across states, two of the largest urban states, California and New York, were slow to implement reform and consequently experienced relatively slow caseload declines—29 percent in both states between 1994 and 1999. The two states accounted for 27 percent of the national caseload in 1994, but 34 percent in 1999. As a consequence of the slower reductions, five of the ten counties with the largest 1999 caseloads were in California. In 1994, only three California counties had made the “top ten” list.

Other states chose to pursue welfare reform efforts much more aggressively and consequently achieved more dramatic declines, even in the cities. Wisconsin, where state-driven reform efforts predated federal reform, saw its caseload shrink by 90 percent between 1994 and 1999. Even heavily urbanized states like Florida and Illinois reduced their welfare rolls faster than the nation as a whole: by 67 percent and 58 percent, respectively. Major cities in these states also watched their caseload fall rapidly: 83 percent in Milwaukee County; 81 percent in Duval County (Jacksonville); and more than half in Dade County (Miami) and Cook County (Chicago) (52 percent and 53 percent, respectively).

This diversity of experience was quite intentional—Congress envisioned welfare reform as 50 distinct state experiments. But state policy doesn’t explain everything going on with welfare reform. There is also much diversity within states and regions.

For instance, while Texas rolls fell 52 percent between 1994 and 1999, two of the state’s largest urban areas—Houston and El Paso—had very different results. Harris County (Houston) cut its welfare caseload 70 percent; El Paso County, only 36 percent. Within California’s relatively slow average decline of 29 percent lies much variation by county. San Francisco and Orange counties both outpaced the state and declined slightly more than 40 percent, while Kern County (Bakersfield) declined only 12 percent.

Within regions, too, major differences arise. Caseloads in Wayne County (Detroit) declined by 38 percent between 1994 and 1998. Caseloads in suburban Oakland and Macomb counties, in the same metropolitan area as Wayne County, declined faster, 50 percent and 51 percent respectively. And in the Philadelphia area, the suburbs dramatically outpaced the central county in caseload decline. Between 1994 and 1998, Philadelphia County’s caseload declined by 25 percent, while those in suburban Montgomery and Bucks counties fell 47 percent and 51 percent, respectively.

Thus, state policy does not fully explain the concentration of welfare caseloads in urban areas. We expect that a range of other demographic, market, and policy factors also contribute. For example, many of the cities with the highest caseloads—Philadelphia, St. Louis, Baltimore, Richmond—share common elements: a high level of poverty, concentrated poverty within particular inner-city neighborhoods, anemic city economies, extensive decentralization in the metropolitan economy. The Urban Center has commissioned further research to determine how particular factors or sets of factors affect welfare reform in U.S. cities.

What This Means for Welfare Reform

The urban concentration of welfare families raises some big questions for national and state policymakers (particularly as TANF reauthorization approaches), as well as for those local leaders tasked with implementing welfare reform. Decisionmakers need to consider two things: how federal and state rules affect welfare and related programs and how those programs and policies should be implemented to best meet diverse local realities.

Setting the Rules: TANF Reauthorization and Beyond

To understand the local perspective on welfare reauthorization, the Urban Center is interviewing a range of city officials and urban practitioners about what changes they would like to see in federal and state welfare policy. Some of the preliminary results of these interviews should come as no surprise. Urban leaders strongly support maintaining current funding under the TANF block grant. They want to retain the current law’s extensive flexibility to permit the design and implementation of welfare policies that best fit local circumstances and meet the needs of their clients. They worry about how time limits and other uniform rules will affect jurisdictions facing concentrated poverty or high unemployment (particularly in the event of an economic downturn).

Two other federal (and state) issues stand out for urban leaders. First, local officials are focusing on the “next stage of welfare reform”: making work pay for the working poor. Given residential patterns, cities have disproportionately large shares not just of welfare recipients, but of people who have left welfare and of low-income workers generally. While some of these families have just entered the workforce for the first time, others have been struggling for years to make ends meet in dead-end jobs. These households, many with children, have low earnings that keep them near the poverty line. They must struggle to keep up with the rising costs of housing, health care, child care, transportation, and other necessities.

Urban areas thus have an undeniable stake in national policies, such as the earned income tax credit (EITC) or subsidized health insurance, that are designed to make work pay. Federal spending on health care, nutrition assistance, welfare, child care, and associated areas dwarfs the spending on narrow “urban” programs. The EITC in particular has been a broadly effective antipoverty—and pro-work—policy. Annual spending on the EITC should soon surpass that of all programs of the Department of Housing and Urban Development. In 1997, the EITC brought $122 million of additional income into the city of Baltimore, with 28 percent of all city residents receiving the credit and the central city capturing half of all EITC funds flowing into the region. Philadelphia took in $210 million—62 percent of the regional total—with 24 percent of city families receiving the credit.

Yet the urban leaders we interviewed do not believe that making work pay begins and ends with enhancing the earned income tax credit and other federal and state income supports. The hard fact is low-income working families do not have access to mainstream financial relationships and the security (and low cost) that they provide. As Michael Stegman and others have shown, an estimated 10 percent of all American families—including a quarter of all African-American and Hispanic families and of all families with incomes less than $20,000—do not have bank accounts. It is estimated that as many as 75 percent of welfare recipients are unbanked. These families must rely on an alternative set of financial institutions—check cashers, rent-to-own stores, payday lenders, consumer finance companies—that charge high rates and make it hard for low-income families to build savings. The ability of local leaders to regulate these institutions effectively and open up access to mainstream financial institutions is limited. This is an area clearly within the purview of federal and state intervention.

Second, local leaders consistently cite the fragmented administration of governmental programs as a barrier to effective delivery of services to welfare recipients. Many, but not all, states organize their welfare programs along county lines, either delegating responsibility to the counties,or organizing state staff along county lines. Yet county welfare administration can itself pose a fundamental challenge in connecting recipients with jobs. Local welfare systems are particularly fractured in metropolitan areas like Atlanta, where 20 separate counties operate their own welfare systems, all within a single labor market.

The federal government defines the jurisdictional boundaries of other related programs quite differently, but the same problem of fragmentation often emerges. The federal housing voucher program, for example, is administered by thousands of local public housing agencies, most of which serve individual cities, towns, and counties. In the Detroit metropolitan area, 31 public housing agencies administer separate voucher programs. The parochial administration of housing (and other federal transit and workforce programs) hinders low-income families from connecting to economic opportunities in the larger metropolitan marketplace.

In a few metropolitan areas, efforts are being made to break down administrative barriers to opportunity. Seattle and King County, for example, administer an integrated approach to welfare-to-work. Illinois has consolidated a large array of services under one super-agency and is experimenting with bundling these services for welfare recipients in distressed urban neighborhoods in Chicago and elsewhere. Minneapolis-St. Paul is increasingly administering the housing voucher program on a metropolitan basis.

The federal and state governments could extend these fledgling activities in several ways. They could set the geography of governance of particular programs at the market (or metropolitan) level. Marge Turner and Bruce Katz have recommended, for example, that metropolitan entities compete to administer the housing voucher program. Or governments could reward collaborations across jurisdictions and across systems. At a minimum, federal competitions for scarce resources should give preference to places and institutions that demonstrate market thinking and holistic action.

Implementing Welfare Reform: Understanding Local Conditions

Getting the larger federal and state rules right is only part of the answer to the urban concentration of welfare families. Welfare reform will be most successful, particularly in distressed areas, if local leaders tailor their approaches to the characteristics of the local welfare population and the realities of metropolitan markets.

Although caseload declines have reduced the number of welfare recipients who must be served, those who remain are harder to place into jobs and are less work-ready than the men and women who left the rolls early on. The education level, work experience, and demographics of the caseload all affect how hard or how easy it is to place recipients into jobs. Yet no two welfare populations are alike. Honolulu, for example, where only 18 percent of the caseload are at the lowest skill and literacy levels, may not need to emphasize basic education in its local welfare-to-work program. But the strategy should arguably be different in Dade County, where 51 percent of recipients cannot locate an intersection on a street map, fill out a government benefits application, or total costs from an order. In a world where returns to skill are increasing, education and skills not only go a long way toward making work pay, but are essential if welfare recipients are to get and keep jobs in the first place.

Second, a region’s employment and residential patterns also matter. In general, metropolitan employment in the United States is moving outward from the city to the suburbs while poverty remains concentrated in central city neighborhoods. Yet the extent of concentrated poverty and employment decentralization as well as the reach of regional transit systems varies from region to region. In Chicago or New York City, for instance, enough jobs exist within the city borders to accommodate recipients moving from welfare to work, and mature city transit systems help families get from city neighborhoods to city workplaces without a car.

In places like Atlanta, however, transportation poses a major barrier to work as jobs move farther and farther out into the suburbs and poor families remain in central city neighborhoods. As the Urban Institute has shown, three-quarters of new entry-level jobs in the Atlanta metropolis are located more than 10 miles away from neighborhoods where city welfare recipients live. Yet the Metropolitan Atlanta Rapid Transit Authority does not reach the three counties (Cobb, Gwinnett, and Clayton) where the region’s jobs are growing fastest. The limited reach of public transit is crippling in a region where only 5 percent of welfare recipients in the core counties have access to a licensed vehicle.

With welfare policy, in short, the devil is in the details. The diversity of cities and regions means that local officials are in the best position to make many critical policy judgment calls. Should local welfare agencies aggressively place recipients in central city jobs or focus on linking them to suburban employers? Should a region invest in a targeted public transit strategy, creating bus lines or van pools linking neighborhoods of concentrated poverty and welfare receipt to suburban job centers or instead pursue strategies to subsidize private car ownership options for welfare families? Local officials who understand the complex interplay between recipient characteristics and service needs, regional labor and housing market dynamics, and the regional transportation infrastructure can best interpret the multifaceted picture of their regions.

A New Focus for Welfare

Although welfare reform has achieved some undeniable successes, especially in reducing caseloads, much remains to be done, especially in America’s cities. In 1996 supporters of devolution emphasized that a one-size-fits-all federal welfare policy neglected important differences between the states?Mass- achusetts is not Tennessee, Iowa is not Oregon, Oklahoma is not Maryland. When the new welfare law is reauthorized in 2002, policymakers, not only in Washington, but also at the state and local level, must craft strategies that reflect the enormous differences between urban areas in the country. The next step in welfare reform is to focus on the cities.