Reconceiving SNAP

Program Origins

In 1939, the Congress authorized the United States Department of Agriculture (USDA) to implement a Food Stamp Program for individuals receiving direct relief payments. Individuals who bought orange “food stamps” at face value (accepted by store owners for any product) were given free blue stamps worth 50% of the value of the orange stamps. (This amounted to a one-third subsidy up to US$0.75 worth of food, about US$12.77 in 2014 dollars.) The blue stamps could only be used to “purchase” “surplus food,” that is, food that the federal government had purchased from farmers (to place a floor on its price). This Food Stamp Program was ended in 1943, after general economic conditions had improved as a result of the World War II mobilization (Roth, 2013).

Through the 1950s and early 1960s, the federal government continued to distribute surplus food, essentially by providing the food to local social welfare organizations such as food banks. Over time, support grew for a revived, food stamp-type program on the grounds that it would be more effective than direct distribution of surplus food. In 1961, President John Kennedy, who, as a senator, had introduced a bill to create a Food Stamp Program, issued an executive order that created a pilot Food Stamp Program and also proposed legislation to create a permanent program. The Food Stamp Act passed in 1964, making the program permanent, but for the first few years, it remained relatively small. Localities were not required to implement the program, and some decided against doing so (Roth, 2013; U.S. Department of Agriculture, n.d.a).

In 1967, Senator Robert F. Kennedy famously visited rural Mississippi (with the national media accompanying him) to document that hunger was a serious problem for poor, even with the new Food Stamp Program. In his words, they saw Black children with “bellies … swollen with hunger” (John F. Kennedy Presidential Library and Museum, n.d.).

One of the reasons that the problem of hunger was greatest in the rural south is that many local welfare agencies systematically excluded many African American families from Aid for Families with Dependent Children (AFDC) and other government benefits. In the ensuing years, various steps were taken to alleviate malnutrition, particularly by increasing access to food stamps. An important step was 1970 legislation (proposed by President Richard Nixon) that (1) removed the requirement that food stamp recipients pay a defined amount of their income in order to receive food stamps (called the “purchase requirement”) for families with incomes below US$30 a month (about US$185 a month in 2014 dollars), (2) reduced the purchase requirement for families that did not qualify for free food stamps, and (3) increased the value of monthly food stamps by about 40% (Roth, 2013).

Rather than directly take on the Democrats in control of these Southern states—as well as the powerful Southern Democrats in Congress—advocates expanded federal food assistance as an end run around their opposition. Today’s program is still shaped by this political expediency.

The End of Malnutrition

As mentioned above, as late as the 1960s, symptoms of malnutrition (and especially child malnutrition)—including emaciation, kwashiorkor, marasmus, stunting, wasting, and even death—were a reality in America. The liberalization of AFDC and food stamps in the late 1960s and early 1970s, however, and their consequent large expansions (plus increased opportunities for African Americans), all but eradicated these conditions. Besides the US$75-plus billion SNAP, for example, annual federal expenditures for school breakfasts and lunches have grown to US$16 billion and for WIC (the Special Supplemental Nutrition Program for Women, Infants, and Children) to US$6 billion.

In 1997, Rebecca Blank, who recently served as Acting and Deputy Secretary of Commerce in the Obama Administration, reported, “Evidence of severe malnutrition-related health problems has almost disappeared in this country” (Blank, 1997, p. 163). In fact, since the 1970s, the physical manifestations of real malnutrition have all but disappeared from the nation’s health data. Between 1973 and 2011 (the latest year for which data are available because the Pediatric Nutrition Surveillance Survey was discontinued), the percentage of children who were underweight declined from 7.3% to 3.5% and the percentage of children who were short in stature declined from 9% to 6.3% (Center for Disease Control, 2012). Many of these children suffered from other illnesses or diseases that caused their being underweight. Furthermore, over the past 35 years, there have been almost zero cases of children suffering from protein-energy malnutrition (PEM), and where cases of PEM have been diagnosed in adults, the vast majority are the result of chronic diseases or drug addictions and not insufficient food. (Center for Disease Control, 2012).

In the face of this progress, advocates turned to estimates of “food insecurity” as a rallying point for continuing and expanding SNAP (so does the Obama Administration). Every year since 1995, the federal government has conducted a survey called “The Food Security Survey.” In 2013, it found that 14.3% of American households were “food insecure,” but many think that this is an artificial construct, as it is based on answers to 18 different questions that express some uncertainty about having sufficient financial resources to obtain enough food to meet the needs of all household members even once in the past year (Bavier, 2002). In the same survey, only 5.6% of all households actually reported that one or more household members were hungry—even once in the past year—because they could not afford food (Coleman-Jensen, Gregory, & Singh, 2014). Moreover, only 0.9% of households with children reported that one or more children were hungry at least once during the year. A far cry from the 1960s, the formative years for most federal feeding programs (see Table 1). However, one feels about this controversial and, much criticized concept, it is a very different problem than malnutrition and should not be the basis for making policy for a US$75 billion-plus program (Wunderlich & Norwood, 2006).

OPEN IN VIEWER

Table 1.

Food Insecurity/Hunger (Even Once in the Past Year).

Household Type and Poverty Status	Percent Food Insecure (FI)
	All FI	FI With Hunger	FI With Hunger of Children
All households
With and without children	14.3%	5.6%	—
With children under age 18	19.5%	5.9%	0.9%
Poor households
With and without children	42.1%	18.5%	—
With children under age 18	45.6%	—	2.7%
Households < 130% poverty
With and without children	38.9%	16.7%	—
With children under age 18	44.2%	—	2.4%
Households ≥ 185% poverty
With and without children	6.7%	2.3%	—
With children under age 18	7.7%	—	—

Note. Adapted from Coleman-Jensen, Gregory, & Singh, 2014.

Obesity

Today, instead of hunger, the central nutritional problem facing the poor, indeed all Americans, is not too little food but, rather too much—or at least too many calories. Although there are still some pockets of real and substantial hunger in America, they are predominantly among populations with behavioral or emotional problems. In 1998, for example, then-Agriculture Secretary, Dan Glickman, when discussing the problem of childhood obesity, said that “the simple fact is that more people die in the United States of too much food than of too little, and the habits that lead to this epidemic become ingrained at an early age.” (Besharov, 2002, p. 4).

Today, as many as 70% of low-income adults are overweight, about 10% more than the nonpoor. Adolescents from low-income families are twice as likely to be overweight (16% vs. 8%). Racial disparities are even greater. Almost 82% of African American women, for example, are overweight—almost 30% more than White women. Even more serious, about 57% of African American women are obese—two thirds more than White women (Ogden, Carroll, Kit, & Elegal, 2014; see Table 2).

OPEN IN VIEWER

Table 2.

Overweight/Obesity.

Age, Sex, and Race/Ethnicity	Percent Overweight/Obese
	1961–1962	2011–2012
	1963–1965a
Men	50/11	72/34
Women	40/16	69/37
Children	4/—	15/—
Men White	50/11	73/33
Black	44/14	69/37
Hispanic	—	78/41
Women White	38/14	65/34
Black	59/27	82/57
Hispanic	—	76/43
Children aged 6–11 years
Boys White	4/—	27/9
Black	2/—	39/26
Hispanic	—	49/29
Girls White	5/—	33/18
Black	5/—	37/22
Hispanic	—	44/23

Note. Adapted from Ogden, Carroll, Kit, & Flegal, 2014.

^a1961–1962: for adults and 1963–1965: for children.

Overweight and obesity refer to excess amounts of body fat. The commonly used ways to determine whether a person is overweight or obese are based on medical data indicating weight levels (for a given height) that are associated with increased mortality and various health risks. The standard measure is the body mass index (BMI), calculated as weight in kilograms divided by the square of height in meters (or weight in pounds divided by the square of height in inches multiplied by 703). A BMI of 25.0 or more is used to define overweight. In children, overweight is defined as sex- and age-specific BMI above the 95th percentile, based on growth charts from the Centers for Disease Control (CDC). Obesity is defined as a BMI of 30.0 or more. Other methods used to measure overweight and obesity in epidemiologic studies include waist circumference, skin-fold thickness, and waist-to-hip ratio. For example, a man 5 ‘10” would be considered overweight at 175 pounds and obese at 210 pounds. A woman 5’4” would be considered overweight at 145 pounds and obese at 175 pounds.

Being overweight is not simply a matter of aesthetics. The growing girth of Americans is a major health concern. The Harvard School of Public Health has summarized a number of studies of the effects of obesity. Among the findings are that women with a BMI of 35 or higher have a risk of developing type 2 diabetes that is 93 times higher than women with BMI lower than 22, individuals who are overweight have a 32% higher risk of coronary artery disease compared to individuals with normal weight, and those who are obese have an 81% higher risk, and that being overweight and obese increases the risk of stroke by 22% and 64%, respectively (Harvard School of Public Health, n.d.). Obesity, of course, is more serious, causing an estimated 50–100% increase in premature deaths (estimated to be 300,000 deaths per year; Allison, Fontaine, Manson, Stevens, & VanItallie, 1999).

Despite this massive increase in overweight and obesity among the poor, federal feeding programs still operate under their nearly half-century-old objective of increasing food consumption. In 2012, the Department of Agriculture reported that SNAP “raises food expenditures and improves nutritional availability. Participants in SNAP spend more on food than they would in the absence of the program” (U.S. Department of Agriculture, 2012a, p. 1). Few experts are willing to say that federal feeding programs are making the poor fat, although the evidence points in that direction. But no expert thinks they do very much to fight this growing public health problem.

SNAP benefits work as intended, raising caloric consumption by as much as 10% more than if recipients were given cash. It’s like when you buy tickets for a set number of rides before entering an amusement park. The tendency is to buy more than one needs and, rather than return the unused ones for a refund, it is easier to take that one or two more rides before leaving. That’s of course why the parks sell them that way. The only difference is that unused food stamps can’t be turned in for cash. (The fact that people do not want to use all their food stamps for food helps explain why a black market has developed with them.)

A 2008 research synthesis by USDA economists found that some evidence that long-term receipt of SNAP benefits increased obesity in non-elderly adult women by between 4 and 10 percentage points, but they did not find any effects on obesity for other subgroups (Ver Ploeg & Ralson, 2008). However, these were econometric studies that, by their nature, have difficulty in controlling for selection effects or other factors that might affect obesity rates. On the other hand, we do know from more rigorous methods that SNAP benefits can lead to increased consumption.

In the early 1990s, the USDA commissioned two random assignment studies of the Food Stamp Program where some recipients were provided cash instead of food stamps. Peter Rossi summarized the findings of these studies: “The reductions in food expenditures were $0.18–0.28 for each dollar provided in the form of cash, compared with conventional food stamps.… These studies show that providing income in the form of food stamps leads to more food consumption than an equivalent dollar amount given in unearmarked form” (Rossi, 1998, pp. 36–37). This “cashing out” of food stamps did not result in unhealthy diets nor the mismanagement of family finances. Recipients continued to get well above the recommended dietary allowances for most nutrients (Carlson as cited in Fasciano, Hall, & Beebout, 1993).

It is unclear, however, what effect, if any, the 2002 adoption of the Electronic Benefit Transfer system has had on this behavior. Since the implementation of Electronic Benefit Transfer cards to all SNAP recipients in 2002, SNAP recipients have been allowed to rollover unspent benefits at the end of the month to the next month, which may have a dampening effect on overconsumption in a given month. Such savings, however, cannot be used for purchases of other goods, so it is likely that recipients spend the excess in future months. The failure to be clear about SNAP as a form of income support has removed the possibility of using this important program to address America’s dietary and obesity problems.

SNAP as the Primary U.S. Social Assistance Program

How should we think about the current SNAP’s role in maintaining this progress? It is most accurate to think of SNAP as a form of income assistance that allows recipients to purchase food. Thus, in its Budget and Economic Outlook reports, the Congressional Budget Office treats SNAP as an “income support” program along with TANF, the EITC, the Additional Child Tax Credit, SSI, and unemployment compensation (Congressional Budget Office, 2015). In fact, the Organization of Economic Cooperation and Development (OECD) classifies SNAP as the primary U.S. “social assistance” program and as the equivalent of other countries’ cash welfare programs. (It does not include TANF because of its narrow scope [Immervoll, 2009].)

SNAP acts indirectly to improve the nutrition and health of low-income Americans by enabling them to purchase and consume more food. Moreover, eligibility for SNAP now reaches to those with incomes high enough to afford an adequate diet—but often not the other necessities of contemporary American life. This does not make the program less socially valuable. As described below, especially since the passage of TANF, SNAP is the major safety-net program for those who have exhausted their UI benefits and have insufficient other income or assets.

Many program advocates, however, have chosen to leave this reality ambiguous—because they believe that it is only the prospect of hunger that is the reason for the program’s strong support among the public (Center on Budget and Policy Priorities, 2015). They could be correct, but the result is to stifle efforts to update the program to reflect developments in other means-tested government programs as well as economic and social conditions generally.

SNAP benefits now far outstrip TANF benefits (in average size and number of recipients), making SNAP (and its predecessor Food Stamp Program) the primary element of the U.S. income support system (Rosenbaum & Neuberger, 2005). This developed by historical happenstance—and the fact that 100% of SNAP benefits are paid for by the federal government (while they were shared under AFDC and are, essentially, a 100% state cost under TANF).

First, what started as a small federal nutrition program was expanded in the 1960s and 1970s because of apparent hunger in states that had low-benefit inadequate and racially discriminatory welfare systems. The federal Food Stamp Program essentially worked around this problem by ignoring state welfare agencies—a disconnect that continues 50 years later even as the initial reason disappeared.

Second, because SNAP would fill in between 30% and 45% of the difference, in the 1970s through 1990s, many states kept AFDC payments lower than they might otherwise have set them. For example, in 1991, California cut its cash assistance (AFDC) to reduce state spending on the poor by US$10.8 billion between 1991 and 1996. The state’s budget analysts, however, calculated that this reduction would trigger a US$4 billion rise in food stamp payments, so the net loss to the poor dropped to US$6.8 billion (Besharov & Baehler, 1993).

Third, again because the federal government paid program costs, there was a tendency to encourage low-income families to leave their time-limited TANF programs while continuing on the federal SNAP (and, when applicable, being transferred to federal disability programs). Pamela Loprest and Sheila Zedlewski of the Urban Institute used the National Survey of American Families to examine former recipients of cash welfare benefits who left the program but not for employment. They found that between 1997 and 2002, the percentage of these “welfare leavers” receiving food stamps increased from about 46% to about 55% (Loprest & Zedlewski, 2006).

Fourth, in the wake of the Great Recession, long-term unemployment was at an all-time high. (Even now, 31.5% of the unemployed have been jobless for 6 months or more [Bureau of Labor Statistics, 2015].) After their UI benefits expire, many unemployed turn to SNAP, especially given recent liberalizations. In 2012, Anderson, Kirlin, and Wiseman (2012) examined longitudinal UI and SNAP data in seven states and found evidence of this phenomenon.

Explaining Recent Increases in SNAP Caseloads

The recent sharp growth of the SNAP caseload began long before the Great Recession. It began under Republican President George W. Bush at a time when employment was in reasonably strong shape, although employment had not recovered from its prerecession levels. Between 2000 and 2013, SNAP spending grew from about US$20.6 billion to about US$79.9 billion (in 2014 dollars) and the SNAP caseload increased from 17.2 million individuals to about 47.6 million individuals (U.S. Department of Agriculture, 2015). (In 2014, as the economy improved, those numbers dropped to US$74.1 billion and 46.5 million individuals.) In comparison, during this same period, the number of individuals in poverty increased from 31.5 to 45.3 million. Since the start of the Great Recession in 2008, the number of SNAP recipients has increased by 68.7% between 2008 and 2013, even as the number of individuals in poverty increased by only 16.5% (U.S. Census Bureau, n.d.; U.S. Department of Agriculture, 2015).

Why have the caseloads and expenditures increased so much? Although a struggling economy and an increase in poverty certainly contribute to the increase in the enrollment of SNAP, statutory changes and local discretion that result in expanded eligibility and loosened criteria for determining eligibility have also been contributors. Here are some of the key changes in SNAP. (The figures below for number of households involved have not been corrected for likely duplication.)

Nullified assets tests. To meet SNAP asset requirements, a household must have less than US$2,000 in assets (US$3,000 for households with a disabled individual) and no more than one vehicle that must be worth less than US$4,650. (Houses, retirement accounts, and personal property are not counted as assets.) There are two exceptions to these rules. For vehicles, the Agricultural Appropriations Act of 2000 allows states to use the vehicle asset test of their TANF programs instead of the SNAP vehicle asset test (Super & Dean, 2001). As of November 2012, 34 states and DC exclude the value of all vehicles, and another 15 states exclude the value of one vehicle (U.S. Department of Agriculture, 2012c). For the more general asset test, under the categorical eligibility provisions issued by USDA regulations in 2000 (described below), states may use the asset tests in their TANF programs in place of the SNAP asset test. Thirty-six states exercise this option and do not have an asset test for SNAP recipients (Falk & Aussenberg, 2012).

Work Disincentives

One of the most distressing trends of recent years has been the decline in U.S. labor force participation.

Less job seeking

As of January 2015, the U.S. labor force participation rate was only about 72.7% (compared to its high of 77.4% in 1997; OECD, 2014). About 6 million working age, Americans (2.5%) did not have a job and were not looking for one (even as they said they wanted one). That takes them out of the “labor force,” and, hence, not officially “unemployed.” (“Unemployment” is defined by the federal government as being without a job and also looking for one.) Federal Reserve Chairman Janet Yellen has called this “shadow unemployment” (Graver, 2014). (See Figure 2, showing participation in the labor force of working-age adults.)

OPEN IN VIEWER

Figure 2.

U.S. employment indicators. (Organization for Economic Cooperation and Development, 2014).

Another 49.2 million Americans of working age were not actively looking for a job and, when surveyed, said they did not want a job. (They answered “no” to “Have you looked for a job in the past 4 weeks?” and “Do you currently want a job?”). They explained that they were disabled or ill, enrolled in school, retired, or taking care of the house or others.

Some of the declines in employment among working-age Americans reflect underlying demographic trends. At one end of the working age spectrum, higher percentages of young people are choosing postsecondary education (colleges, community colleges, and specialized job training programs) instead of immediate employment, and, at the other end, an aging baby boom generation is predictably accelerating its exit from the labor market. (At the same time, there are more elderly who are working full-time than in the past, presumably because of the asset losses they experienced during the Great Recession. In fact, between 2000 and 2013, the entire increase in the number of individuals in full-time employment has been because of the increase among elderly, those aged 65 and older, who are working full time [OECD, 2014].)

A Council of Economic Advisors report estimates the impact of these demographic factors to be only about 51% of the decline in the labor force participation among all workers (not just working age; Council of Economic Advisors, 2014), but, given data limitations, the true number is probably lower.

No one knows how many of the 49.2 million Americans not in the labor force thought they could not get a job or were well-enough off from other sources income (perhaps supplemented with safety-net benefits). But given the dearth of good-enough-paying jobs, a sense of palpable discouragement pervades the nation. Many of the long-term jobless feel left behind by what appear to be permanent changes in the economy and have all but given up. It is difficult to exaggerate the impact of repeated failed job searches. In a 2014 nationwide poll by Express Employment Professionals, an employment staffing company, 47% of the unemployed agreed with the statement: “I’ve completely given up looking for a job” (Express Employment Professionals, 2014, p. 2).

Worse, the problem seems to be feeding on itself. Many employers have apparently decided that the long-term jobless would not make good employees—because of their attitude, skills, or just plain age (Ghayad, 2014). Barring a major increase in demand for American workers—or some change in their willingness to accept lower-wage employment, many of the long-term jobless may never get back to work.

Those directly affected by this weak labor market have paid a high price in lost earnings and emotional stress and continue to do so. But this high level of nonwork is also an obstacle to the economy’s long-term recovery. On June 6, 2014, a Washington Post lead editorial worried:

Declining labor-force participation may be a new characteristic of the postrecession U.S. economy, and it bodes ill for two reasons: The economy’s capacity for growth depends on robust use of all available factors of production, the minds and hands of U.S. workers very much included; indeed, the surge of women into the workforce was one of the key drivers of economic expansion in the 1980s and 1990s. Declining labor-force participation implies a rising “dependency ratio” of workers to recipients of social assistance.

According to economic theory, a larger supply of potential workers ordinarily leads to more hiring. The additional job seekers increase competition for jobs, thus lowering starting wages, which encourages employers to expand their workforce, which, in turn, raises economic activity (OECD, 2006).

Moreover, as the same Washington Post editorial suggests, the billions more now being spent on means-tested, safety-net benefits are not available for other pressing societal needs. Consider this very rudimentarily calculated example: If the percentage of U.S. households receiving SNAP had remained the same between 2008 and 2013 (about 10.7%), spending on SNAP benefits would have been about US$122 billion lower than the actual amount spent during this 5-year period, adjusting the base to reflect the increase in poverty still leaves a whopping US$93 billion (U.S. Department of Agriculture, 2002–2014). (Of course, some of this money might be spent on less worthy causes or not at all because other programs might not enjoy the same level of political support.)

The Absence of SNAP Work Requirements

As described above, states are financially and politically rewarded when they move people off UI and TANF (programs with at least some activation requirements) and on to SNAP. This incentive was not created deliberately, but, rather, is a historic accident of how and when the programs were established.

The SNAP does have some work requirements. (“With some exceptions, able-bodied adults between 16 and 60 years must register for work, accept suitable employment, and take part in an employment and training program to which they are referred by the SNAP office. Failure to comply with these requirements can result in disqualification from the program.” U.S. Department of Agriculture, 2012b, p. 4) However, as Ron Haskins of the Brookings Institution and others have noted, “These requirements do not seem to be rigorously enforced” (Haskins, 2012, p. 12). The absence of meaningful job search or work-related activity requirements in SNAP can undermine UI and TANF work requirements—because SNAP benefits rise if UI or TANF is terminated or reduced. If the average UI recipient (in a three-person household) loses benefits, monthly SNAP benefits rise from about US$180 to about US$530. If the average TANF recipient (in a three-person household) loses benefits (about US$430 a month), then monthly SNAP benefits rise from about US$400 to about US$530 a month.

It is possible to correct this problem. In April 2014, the University of Maryland and the Secretary’s Innovation Group, with the assistance of the American Public Human Services Association, cosponsored a one-day meeting on how to implement the SNAP pilots authorized in the 2014 Farm Bill in a way that will reduce dependency and increase work levels. Seven state human services secretaries attended the meeting as well as about six senior professional Congressional staff and about five senior administration officials. Through the associated webinar, senior state officials from 13 other states participated (University of Maryland School of Public Policy, 2014).

These pilots, the product of an awkward political compromise, may well point the way to meaningful reform. The problem, however, goes deeper. Although the states administer the SNAP, pay one half of its administrative costs, and essentially decide who will receive benefits, they do not pay for any of the benefit costs. Those are covered entirely by the federal government. Hence, the states have no incentive to reduce SNAP caseloads and in fact, as described above, have an incentive to shift recipients from their state-funded TANF programs to the federally funded SNAP, while keeping the resulting savings and enjoying the political benefit of a reduced cash welfare caseload. In contrast, because states can keep the money that they save in TANF, they are more cautious with spending and focus on limiting the growth of the caseload.

Real reform probably requires that the states be made financial partners of the federal government. States should have a more direct financial stake in the proper governance of SNAPs, including eligibility determinations. Given that all program funds come from the federal government, a substantial liberalization of eligibility determinations was predictable. State officials have little reason to be cost conscious—as long as program funds seem available. And they have even less reason to take on “street-level bureaucrats” and the vast discretion they enjoy (Lipsky, 1980). As in the case of many of other federal, means-tested programs, states should be required to pay a portion of SNAP’s costs, so that they would have a stake in enforcing eligibility rules. (Properly structured, this would make it possible to give states the flexibility to shift how they spend funds—to spend less on expanding enrollment and more on enhancing services for current recipients, such as spending more time on job training, job seeking, and, yes, nutritional counseling.)

International Comparisons

Starting in the 1970s, many European countries experienced similarly worrisome declines in employment and labor force participation. Across the original 15 members of the European Union (EU-15: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden, and the United Kingdom), between 1970 and 1982, the percentage of the population employed fell from 61% to 57.8% (before beginning a slow increase). For men, the decline was much longer and steeper, from 83.7% in 1970 to 70.5% in 1994 (OECD, 2014). Overall, labor force participation increased during this period, but only because more women were entering the labor force. At the same time, in most countries, new highs were reached in the percentage of the population receiving government benefits from unemployment, disability, and social assistance programs.

In response, a growing number of developed countries introduced policy reforms aimed at “activating” the recipients of safety net benefits who might be able to work, that is, requiring them to perform work-related activities while receiving benefits. (The U.S. welfare reforms of the 1990s were an early part of this movement, but since then, some other developed countries have made more fundamental reforms to their labor activation policies.) Since the 1990s, one country after another has modified its safety net programs, as described in this policy brief. The countries that made the most extensive changes are Australia, Denmark, Germany, the Netherlands, the United Kingdom, and to a lesser extent, Finland, France, Japan, Norway, and Sweden. These countries made both substantive changes (tightening eligibility, limiting the duration of benefit receipt, and mandating job search and other work-first activities) and administrative changes (consolidating programs, decentralizing authority, outsourcing services, and incentivizing systems of financing and reimbursement). The key aspects of the changes in Europe can be summarized under three overarching themes:

Synchronizing benefits across safety-net programs to facilitate seamless benefit receipt over time as well as activation efforts, so that, as individuals were time-limited off UI and disability programs, they were transitioned to cash welfare or subsistence programs;

Few of these changes have been rigorously evaluated. Although no one can say that they have successfully lowered long-term recipiency and increased labor force participation, the evidence from similar policies adopted in the past indicates that if, implemented well, they have the potential to do so (see, e.g., Grubb, 2011 and Tatsiramos & Van Ours, 2011). According to a World Bank report on labor activation programs, “One conclusion from a review of existing evidence is that well-designed policies can have a positive impact on employment outcomes for participants, but that many existing policies have in fact failed to prove effective or cost efficient” (Immervoll, 2012, p. 8). Moreover, the changes seem to enjoy reasonable political acceptance from the left and right. If not initially, over time. And they seem to have maintained the essentials of that nation’s safety net.

Hence, it is worthwhile to review what these countries have done to adjust their safety nets to encourage labor force participation at a time of high joblessness. It is not that their programs should be simply transplanted here. There are surely too many economic, social, and political differences for that to be possible, let along make sense. But just as certainly, the general approaches they adopted are worthy of consideration.

Our focus in the United States should be on rationalizing interactions among our patchwork of safety-net programs—TANF, SNAP, UI, and disability—which too often create a work disincentive for low skilled or difficult to employ citizens. A possible solution is to combine—or at least align—the administration of these programs and to add what the Europeans call “labor activation” (akin to job search requirements) to all recipients of government assistance.