What Does Minnesota Know?

In an aging St. Paul neighborhood known as Frogtown, at a storefront social-services agency called Lifetrack Resources, Tina Thompson and Angela Fink are meeting one afternoon to discuss the impoverished clients they are trying to move into the world of work. There is Zainab, an Ethiopian refugee who arrived in the United States with poor English and rusty clerical skills, but who will graduate this spring from an accounting program and almost certainly land a good office job. There is Ginny, pregnant and recently laid off, but eager to plunge back into the job market as soon as she can update her résumé. And then there is Cassandra, who has washed out of several jobs and vocational schools since dropping out of high school at 17. Cassandra turns out to be clinically depressed, and the job counselors want her to complete a course of therapy and antidepressants before she resumes work on her elusive high-school diploma.

This is not the tidy welfare-to-work assembly line that social conservatives envisioned when they overhauled public assistance and created Temporary Assistance for Needy Families (TANF) in 1996. But in Minnesota, the relatively patient and flexible approach works. Minnesota places more welfare recipients in private, unsubsidized employment than the average state. It moves most clients on and off assistance within two years and, unlike most states, actually lifts most client families out of poverty.

With results like that, Minnesota should count as a success in the national welfare experiment that Congress approved in 1996. But not according to the Bush administration. Under ambitious new work requirements proposed by the White House as part of its plan to reauthorize TANF, states such as Minnesota would have to junk this patient, flexible strategy and replace it with something narrower and more mechanistic. The White House proposal requires most welfare parents to participate in government-approved activities for 40 hours per week, with most of that time spent in actual work or workfare. "The range of activities that we offer -- skills training, guided job search, bilingual education -- would be greatly constrained or outright disallowed," says Michael O'Keefe, Minnesota's commissioner of human services. "It's just crazy."

Minnesota isn't alone in this predicament. Oregon, which earns high marks in national evaluations for promoting work and reducing poverty, encourages recipients to choose their jobs carefully, then combines part-time training with part-time work. Illinois, which has seen many clients work their way off assistance and stay off, recognizes a variety of activities in a client's "responsibility plan," including drug treatment and full-time community college.

Officials from Maryland are worried that meeting the administration's work targets would require them to "dismantle effective programs that reduce nonmarital births, improve job retention, encourage completion of secondary education by teenagers and young adults, and reduce substance abuse," according to a recent survey by the National Governors Association. An official from Alabama said: "Our case managers are encouraged to assign clients to a combination of work and educational activities that best meet the client's needs. … We will no longer be able to offer this." Authorities in Washington State say that meeting the administration's work targets would require them to resurrect a workfare program that the state legislature just abolished because it was a waste of money.

Of course, governors always chafe at new rules from Washington, D.C. But this time they're right. Scholars have now conducted a large body of rigorous research on state welfare-to-work strategies adopted since the early 1990s, and it shows that the states that pursued a "mixed strategy" of work and training are the ones with the best records of promoting work and lifting families out of poverty. This will come as no surprise to anyone who has ever set foot in a welfare office. For every client like Ginny -- skilled, experienced, and ready for the workplace -- there's another like Cassandra, who will never hang onto a job or give her children a decent standard of living without a substantial investment in basic education, job skills, or therapeutic services.

The irony is that the 1996 welfare law specifically sought to foster experimentation by the states, with the hope that Washington might learn from the results. But the White House proposal, now embodied in legislation passed by the House of Representatives, seems to ignore the results of these experiments and almost certainly would smother the ones that are most promising.

Perhaps the most rigorous set of evaluations of state programs has been conducted by the Manpower Demonstration Research Corporation (MDRC) of New York. "The evidence indicates that both job-search-first and education-first strategies are effective, but neither is as effective as a strategy that combines the two," MDRC's Gordon Berlin testified before the Senate Finance Committee in March. Under the White House targets, he added, "Instead of focusing on getting people off of welfare, states may become preoccupied with keeping everyone busy while they are on welfare."

White House officials say they are merely trying to build on the success of the 1996 law, which aspired to convert welfare families into working families. "If we are going to simulate the world of work for welfare recipients, we need to simulate the real 40-hour workweek that most Americans have," says Grant Collins, chief of staff in the Office of Family Assistance at the Department of Health and Human Services. "Anything less leads to a different expectation than what is the real standard in the world of work."

The White House also says that governors have misunderstood the complex work requirements and that they will have great flexibility in defining supplemental activities once a welfare client has met his or her core work requirements for the week. "Allowable activities can be determined by the state as long as they meet a TANF purpose," Collins says. "It could be securing stable housing, it could be vocational rehabilitation. The list is endless."

The administration work rules, at first glance, do seem to offer reasonable flexibility. By fiscal year 2007, every state would have to document that 70 percent of its welfare recipients were meeting federal work requirements every month. (As a practical matter, the threshold today is 40 percent or less for most states.) To count toward that 70 percent pool, most recipients would have to prove that they were occupied for 40 hours a week -- 24 hours in direct work or workfare and 16 hours in other activities such as training or night school. The administration plan also makes room for other activities, such as drug counseling or vocational school, but for only three months in any 24-month period. In other words, the great majority of welfare recipients would have to file paperwork in the majority of months proving that they were busy 40 hours per week.

In practice, those targets are less realistic than they appear. Remember that some 80 percent of adult welfare recipients are single mothers, most of them with children under age six, many of them with mental impairments or physical disabilities. It's true that the 40-hour workweek is a venerable American tradition. But a typical working mother in the general population today puts in only about 30 hours a week, and even this has given rise to a generation of soul-searching about overworked parents and overstressed families. It's hard to see why poor, single parents should have to work more hours than the average working mother, especially when they are unlikely to have a reliable car, a dependable neighbor, flexible child care, or a sympathetic boss -- the vital conveniences that typical working parents take for granted.

The administration's targets are unrealistic for states, too, but for a slightly different set of reasons. If you think of welfare families as a population flowing through a system, states such as Minnesota do put 70 percent of their welfare clients to work over a period of one or two years. (Even relatively flexible states such as Minnesota and Illinois promote work first for most of their clients.) But measured at any one point in time, as the administration proposes, the pool of recipients includes lots of families who have just entered the system and can't begin work immediately. They are coping with the very disasters that put them on welfare in the first place: a divorce, an eviction, a medical crisis, a violent husband. The administration would exempt families from work requirements during their first month on assistance, but that is scarcely enough time for most recipients to put their fractured lives together and find a job.

Among the families working with Angela Fink and Tina Thompson on any given day, a large share are interviewing for jobs, hunting for child care, trying to get out of a homeless shelter, or simply waiting for a job-skills class to start. To have 70 percent of the caseload actually at work on any given day, officials such as O'Keefe say they would have to push 100 percent of their clients toward the first available job -- or worse, create workfare jobs to keep them occupied.

Since the governors began complaining, the administration appears to have backpedaled a bit. The White House emphasizes that welfare parents could meet their 16-hour participation requirement with a broad range of family-friendly activities such as coaching sports or volunteering at school. It now appears that the ground for compromise in Congress will be how much latitude to give states in defining this smaller portion of the 40-hour week. Even so, state welfare directors bridle at the prospect of monitoring the private activities of their clients for a certain number of hours every month. "Now you've got to do time sheets on T-ball?" asks Elaine Ryan, director of government affairs for the American Public Human Services Association.

In addition, the administration and House Republicans plan to offer governors something called "State Flex,'' a "superwaiver" from federal program rules. But roughly 40 states already have waivers from federal rules -- that's how many of them got started on their welfare-to-work experiments. Under the administration's plan, they would lose those waivers and have to conform to the new rules or re-apply on terms that satisfy the White House.

What's often forgotten in Washington is that, although Congress created TANF in 1996, most states didn't enact their own legislation until 1997 and many governors didn't actually overhaul the local welfare office until sometime in 1998. As they learned what it would take to actually engage 2.6 million poor adults in the search for work -- something the nation's welfare system had never done before -- it appears that states started moving away from the work-first model, not toward it. Reviewing 17 cities in 13 states for the Urban Institute, researchers Pamela Holcomb and Karin Martinson concluded that, while 11 cities started with a work-first model in 1997, only five still pursued that approach by 2000. The other cities moved toward a blend of work and training or "enhanced services," such as drug treatment and vocational counseling. This shift in strategy reflects one of the startling discoveries of the welfare revolution. Once states actually examined their indigent families and inquired why they weren't working, they found that the welfare caseload is almost indescribably diverse.

Consider Minnesota's Ramsey County, home to the city of St. Paul and the office where Fink and Thompson work. Of roughly 6,000 families on public assistance, perhaps one-quarter fit something like the popular stereotype of a welfare recipient: single mother in her twenties, never married, raised in poverty. Another quarter resemble Ginny: They have substantial education and work experience, but they've had some crisis -- a divorce, a medical emergency, a catastrophic layoff -- that knocked them temporarily out of the working class. Fully one-fifth are immigrants; Minnesota has a large population of Laotian and Ethiopian refugees, many of whom speak little English and have almost no experience with market employment.

Then there are the deeply disadvantaged clients who, many states are discovering, represent perhaps one-quarter of the caseload. The agency where Fink and Thompson work, Lifetrack Resources, recently surveyed 189 of its needier clients, a group that represented about one-fourth of its total caseload. Lifetrack found that 25 percent had no formal schooling whatsoever, 70 percent had no reliable car, 15 percent were homeless, and 43 percent had a physical handicap or a learning disability. "Everyone who walks in our door walks in for a different reason," Thompson says. "We have to figure out what that reason is before we can send them out again."

For all this diversity, job counselors such as Fink and Thompson do favor a work-first strategy for many of their clients. Even an $8-an-hour job can give the family more income than a welfare grant, and many of their clients prefer a steady paycheck to the prospect of going back to school or mental-health counseling. The state's job counselors also advise clients that if they can earn enough money to go off the cash portion of their grant and accept only food stamps and Medicaid, they can stop the ticking of their federal five-year clock and "bank" benefits for any future emergency. Indeed, many advocates in Minnesota think the state already promotes work at the expense of education.

But Fink and Thompson also know that the next client who walks in the door from a street in Frogtown might only founder if simply ordered to go get a job. They could have directed their Ethiopian client, Zainab, straight back into the job market two years ago. And if they had? Says Fink: "She'd be working as a janitor and making $7 an hour for the rest of her life."