Stay informed: Sign up for eNews Subscribe
Read Chapter 1


Welfare Reform and the Enduring Structural Roots of Distrust

A Brief History of Cash Assistance for Low-Income Mothers and Their Children

The 1990s were not the first time the United States attempted to reform the way it delivers cash assistance to low-income families. In fact, just a mere eight years earlier, the Family Support Act of 1988 had made similar, though smaller, changes. Both pieces of legislation echo themes that run throughout the history of the American welfare state. Since the beginning of this history, politicians and the public have been concerned that "handouts" in the form of cash assistance would create a dependent population and undermine the value of work. Additionally, policy makers have resisted rewarding the "wrong" sorts of people-those deemed undeserving or unworthy of help either because they are able-bodied but are not employed or because their moral character is considered lacking in other ways. Those deemed deserving benefit from more generous social programs and those deemed undeserving receive less generous benefits or, sometimes, none at all. Popular resentment of welfare recipients, considered undeserving by many citizens, has played an important role throughout this history. This theme continues in the most recent reforms. In fact, one can interpret the welfare reform of 1996 as an aggressive attempt to promote "deserving" behavior (welfare exit, employment, marriage) and to transform the welfare population from an "irresponsible" and "undeserving" crowd to a responsible and deserving one. After all, the words "Personal Responsibility" appear in the name of the legislation that ushered in welfare reform.

While all forms of government social spending can be considered "welfare," when people refer to welfare they typically mean cash assistance to low-income families. These families usually, but not always, consist of single mothers and their children. At the time the 1996 changes were being debated, the program that issued such assistance was called Aid to Families with Dependent Children (AFDC). The earliest precursors of AFDC, which were instituted even before the United States had a federal system of social welfare policies, were the Widows' Pensions, a set of state-level pensions designed to support Civil War widows, enabling them to continue raising their children rather than losing the children to orphanages. Widows, especially those who had lost their husbands in wartime military service, were considered a group particularly deserving of government assistance. Furthermore, staying home to raise their children was accepted as the appropriate role for women. This sympathy for war widows supplied the needed political will to pass the Widows' Pensions. In doing so, state governments were willing to act as a substitute for an absent breadwinner in the family by replacing his earnings, at least in part.

The concept behind Widows' Pensions became the basis for a national program of cash assistance to low-income families when Franklin Delano Roosevelt passed his main New Deal legislation, the Social Security Act of 1935, in response to the Great Depression. The Social Security Act created a two-tiered national welfare state in which employed Americans receive social welfare benefits, such as disability and retirement payments, through a set of universal, employment-based, contributory social insurance programs, while those out of the labor market whose income falls under a defined eligibility level receive benefits through noncontributory public assistance programs. Roosevelt initially conceived of cash assistance to low-income families and similar public assistance programs as equal in importance to employment-based social insurance programs, but they quickly became less favored and less generous. Even Roosevelt feared that cash relief without work requirements would become a "habit with the country."

The two-tiered welfare state not only resulted in an uneven welfare state in terms of generosity but also divided the population into those eligible for the more generous programs and those shuttled into the less generous ones. Since women, racial minorities, and members of the lower classes have had less stable attachments to the labor force, the division has largely been based on gender, race, and class lines.

These lines became even clearer in the 1960s when civil rights legislation and President Johnson's "War on Poverty" and "Great Society" programs (inspired by President Kennedy's antipoverty efforts cut short by his assassination) gave political voice and equal access to welfare benefits to many African Americans for the first time. Suddenly the welfare rolls swelled and those receiving cash assistance benefits were increasingly African American. In addition, the single-mother recipients of benefits were increasingly never-married women rather than widows. Coupled with the 1960s' racial unrest in many cities, these changes made middle-class Americans less likely to view single mothers as a group deserving of government aid. While African Americans still constituted a minority of those on the welfare rolls, a public perception of welfare recipients as primarily African American probably decreased support for social welfare spending. Welfare served as a convenient focal point for white America's anger at and judgment of African Americans. As Jill Quadagno writes in her book The Color of Welfare: How Racism Undermined the War on Poverty, "No program better exemplifies the racially divisive character of the American welfare state than Aid to Families with Dependent Children (AFDC). Conservatives attack AFDC for discouraging work and family formation and for rewarding laziness. Such comments are really subtly veiled messages about family structures and employment patterns among African Americans. However, often the attacks are neither veiled nor subtle." These racially tinged reactions to welfare largely spelled the end of the Great Society efforts to eradicate poverty.

As time wore on through the feminist movement of the 1970s and the increasing labor market participation of mothers in the 1980s, the idea that women should be home with children, and that those without a husband's paycheck to enable them to do so should be paid by the government instead, lost support. Cash assistance to low-income mothers and their children was no longer seen as a benefit that appropriately replaced an absent male breadwinner's earnings so that women could fulfill their duties as mothers and children's basic needs could be met; instead, it was viewed as one that rewarded recipients who were undeserving on account of their lack of a work ethic and their sexual immorality. Attitudes gradually shifted from a "maternalist" welfare state, or one designed to support women's roles as mothers, to a "universal worker" welfare state, or one that encouraged men and women alike to provide for children not through stay-at-home care but through paid employment. Some argued that the cash assistance program created perverted incentives that discouraged both work and marriage, creating a pathological "dependency."

The 1980s ushered in the transition from the "War on Poverty" to the "War on Welfare." President Reagan used the image of the "Welfare Queen," always depicted as an African American woman and often as one who picked up her welfare check in her Cadillac and fur coat, to drum up (or tap into) animosity toward welfare spending. In Reagan's view, the biggest problem with welfare was the fraud it bred, and he used racial tensions to add fuel to the public's anger over paying taxes for a program that, in many people's eyes, benefited only the undeserving. In 1981, he passed the Omnibus Reconciliation Act, which began to chip away at welfare benefits. The Family Support Act of 1988 mandated that each state institute a welfare-to-work training program for welfare recipients and allowed states to require recipient participation, created work supports for recipients transitioning to work, and stepped up efforts to establish the paternity of children receiving benefits in order to increase enforcement of child support payments to reimburse the state for welfare expenditures.

But it was the Democratic president Bill Clinton who took welfare reform across the finish line. Polling during his 1992 presidential campaign bid indicated that his line promising to "end welfare as we know it," which was written by a young aide the night before he first used it, was one of the most popular phrases he uttered on the campaign trail. Once elected, he set out to make good on that campaign promise. He brought top social policy experts from Harvard University and Washington, D.C., think tanks to help develop the right mix of programming. Then the Republican-led Congress drafted its own versions of a bill. Twice President Clinton vetoed these versions, arguing that they placed too much burden on recipients without enough support. When the third version came to his desk in August 1996, he signed. Mary Jo Bane and Peter Edelman, two of the experts he had brought to his administration to work on welfare reform, resigned in protest, claiming the new law would drive many additional families into poverty.

Fears of rewarding the undeserving have been with us since the beginning of our welfare state. Over time, single mothers and their children have lost their status as a deserving group and have been seen increasingly as unworthy of our support. Coupled with frustrations with the cash assistance program on all fronts, including among recipients of benefits themselves, this shift led to increasing demands for welfare reform and, ultimately, to the passage of reform in 1996.

What Policy Changes Did Welfare Reform Make?

While the 1996 legislation made many changes to AFDC and other programs (such as Food Stamps), I outline here the key changes it made to the AFDC program, since, as the main cash assistance program to low-income families, AFDC is often what people mean when they refer to "welfare." Before reform, women like Bethany Grant, whom we met in the Introduction, received monthly cash assistance through AFDC, but reform replaced AFDC with a new program called Temporary Assistance to Needy Families (TANF). The word temporary underscores the idea that women like Susan Schiller, whom I interviewed after reform and whom we also met in the Introduction, should see the program as an emergency, short-term source of support between periods of employment or marriage to an employed spouse. Unlike Bethany, who had no limit on the amount of time she could receive cash assistance, Susan faced a lifetime limit of five years of receipt.

Susan had received cash assistance, or "welfare," four times in her life. Three of these were before reform and began when she gave birth to each of her first three children. These three times totaled about five years. Susan then remained off welfare for almost a decade until after reform, in 2001, when she was between jobs, was suffering from depression, and went back on welfare for about four or five months to get herself together before finding another job. Had reform's time limits been in place during each of the times Susan received welfare, she would have reached her time limit before this last time. Though she had not used welfare in close to ten years, was out of work, was severely depressed, was the single mother of three minor children at the time, and received no child support from the fathers of any of her children, she would not have been eligible. For this reason Susan considered returning to welfare a last resort. However, since only her 2001 short stint was after reform, Susan actually had over four years before she would hit her time limit. If Susan's trouble finding a job or the pressures of her son's legal issues led her one day to return to welfare, she might end up reaching that limit. If so, then Susan would never be eligible for cash assistance again in her lifetime, no matter what financial disaster might befall her. Neither the loss of a job, a new baby in the house, a medical emergency, nor any other circumstance would allow her to be eligible.

Unlike Bethany, who was encouraged but not required to find a job, when Susan was receiving cash assistance in 2001 she did face work requirements. The exact nature of these requirements differs across the states, but in Illinois, for a woman like Susan who was not caring for an infant at the time, was not in a domestic violence crisis, did not have a health condition deemed severe enough to warrant an exemption, was not caring for someone else with a severe health condition, and was not over the age of sixty, these requirements began immediately.

Like most Illinois welfare recipients since reform, Susan spent the first thirty days after she filed her application "working for her check," meaning she had to do volunteer work cleaning, carrying boxes, and filing papers in a welfare office from 9 a.m. until noon five days a week to "earn" the $342 she received in cash assistance for the month. She was then sent to a job-training program that sent her out on job interviews. These interviews gave her little control over the kind of job she might get, her hours, or the job's location. Luckily for Susan, she found a job that suited her needs through her own efforts and left the welfare rolls. It was this job that she left when she was trying to gain control over her son, leaving her in her current situation of getting neither welfare benefits nor earnings from work.

Before reform, Bethany was guaranteed access to welfare benefits. The Social Security Act of 1935 made cash assistance a federal entitlement for low-income families. This meant that as long as a family met the eligibility criteria, it could not be denied benefits even in times of budgetary stress on the government. Susan enjoyed no such guarantee. Now the federal government gives each state a block grant to cover its cash assistance program for low-income families. If there is high demand for welfare benefits-for example, during a period of economic recession-the block grant could run out of money. If so, the state is allowed to say that it will no longer pay welfare benefits because welfare reform has taken away the federal entitlement to those benefits. Many critics of welfare reform point to this ending of entitlement as the most audacious change reform made because it undoes what Roosevelt did when he established a right to benefits under the Social Security Act of 1935.

To discourage nonmarital childbearing, reform also allowed states to decide not to give welfare benefits to new babies if their mothers were already receiving cash assistance. Illinois initially decided to take this option to impose what is called a "family cap," but it then reversed its decision and began to phase out the family cap provision in August of 2003. If Susan had been a welfare recipient when the family cap was in place and had gotten pregnant, she would not have received any cash assistance benefits for her new baby.

Undocumented immigrants have never been eligible for cash assistance benefits. The 1996 law extended limitations on immigrants' eligibility by allowing states to bar documented immigrants from cash assistance benefits. Both Bethany and Susan were U.S. citizens; several of the other women interviewed after reform were not, but since their children were U.S. born, the children could receive benefits. Reform also allowed states to deny eligibility for benefits to those with a drug felony record.

Mothers are now required to establish officially the paternity of their children before they can receive cash assistance so that states can attempt to reimburse their welfare programs with child support payments from absent fathers. When Susan applied for cash assistance in 2001, she was given an appointment with a separate child support office in order to establish paternity. She named the father of her children, who were still minors at the time, and the court awarded child support, even though he never paid. Some fear that this provision puts women and children in harm's way by reinitiating their contact with potentially abusive or hostile men. Provisions have been made for exemptions for women fearing violence, though critics remain concerned that exemptions are too difficult to get. While child support enforcement serves the financial goal of paying back the state for welfare outlays, proponents also argue that it encourages the reintroduction and involvement of fathers with their children. Welfare reform legislation also encourages states to promote marriage in a variety of ways, including "relationship-training" programs whose goal is to teach parents constructive ways to interact with each other. President George W. Bush was particularly supportive of the marriage promotion aspects of welfare reform.

The post-reform changes in welfare rules are enforced through a system of "sanctions" that remove benefits from recipients who fail to comply with rules by not attending required appointments, failing to file paperwork, not meeting job search and work activity requirements, or committing other similar infractions. Reform also enhanced work supports for low-income parents. It increased money available for child care and transportation subsidies. The Clinton administration fought for additional work supports as well. It succeeded in decoupling eligibility for Medicaid from the receipt of welfare so that families with income low enough could maintain Medicaid coverage even if they were not current or recent cash assistance recipients. Reform also was accompanied by a substantial increase in the Earned Income Tax Credit (EITC), which rewards low-wage and even working-class families for being employed with a tax credit that can be sizable.

What Effects Did Welfare Reform Have on Low-Income Families?

The literature on the effects of the landmark 1996 welfare reform legislation has focused largely on assessing whether reform met its goals of reducing the size of the welfare rolls and increasing the employment rate, earnings, and marriage rate of welfare recipients. In general, this research has reported that in the early years following reform welfare caseloads plummeted, poverty rates declined, and recipients entered the labor force at rates exceeding expectations. Investigators, however, caution that it is difficult to disentangle fully the separate effects of policy and economic changes, especially since reform was implemented during a period of great economic prosperity. Also, despite welfare reform's goals of decreasing the nonmarital fertility rate and increasing marriage, analysts have found reform to have virtually no effect on marriage and childbearing patterns. Even so, early studies celebrated what seemed to some like enormous positive effects of reform or, to those more skeptical, at least much less disastrous effects than originally predicted.

These findings at first seem to throw into question the validity of my assertion in the Introduction that the struggles of the women I interviewed after welfare reform do not look very different from those I interviewed before reform. Adding to the confusion, I am not the only one to make this assertion. Other investigators similarly find that reform did not change the qualitative nature of life for low-income mothers. In her book Flat Broke with Children: Women in the Age of Welfare Reform, Sharon Hays asserts that reform did not help the women she interviewed because of its contradictory messages. She argues that it simultaneously urged mothers to follow "the Work Plan," which promotes single-minded independence, and "the Family Plan," which encourages marriage and commitment to others. Hays concludes that following both at once is nearly impossible and that her respondents were consequently unable to succeed or benefit from reform's initial promise.

In American Dream: Three Women, Ten Kids, and a Nation's Drive to End Welfare, New York Times journalist Jason DeParle suggests that welfare recipients are too busy trying to keep their heads above water either to pay attention to welfare policy or to feel any impact of it. As Angie, one of his three main informants, told him when asked about welfare reform, "I don't pay no attention to that crap!" DeParle goes on to write, "With welfare or without it, Angie said, 'you just learn how to survive.'"

While the initial quantitative evaluations of welfare reform showed great promise for the effectiveness of the policy and improved outcomes for low-income families, later work began to question how far-reaching and long-lasting these effects would be. These later quantitative studies are more in line with the results that Hays, DeParle, and I discuss.

For example, one study questioned the standard practice in early studies of focusing on average effects.Results based on what happened to families on average mask variation across families and thus hide the fact that some families may not be helped by reform and may actually be hurt by it. To address this problem, analysts investigating a Connecticut reform program examined effects separately for welfare recipients in the top half of the income distribution (among program participants) and those in the bottom half of the income distribution. They discovered opposite effects of reform measures for the two sets of recipients. While the relatively advantaged recipients experienced earnings gains as a result of reform, the more disadvantaged women experienced no earnings gains and often even had earnings losses. These findings suggest that many current and former welfare recipients may not be sharing in any of reform's potential benefits and may instead even be financially suffering as a result of reform.

Another study showed that while reform might have reduced poverty rates, it did not actually change low-income mothers' standard of living or lower their levels of hardship. Between 1997 (the year reform was implemented) and 2003 (the year before I conducted my second round of interviews), low-income single mothers were no more likely to obtain health insurance coverage or establish their own households than before. They were also no less likely to worry that they would run out of food by the end of the month or to have a child go hungry for an entire day in the previous twelve months. In other words, while officially their rates of poverty might have gone down, their standard of living remained the same. The study authors explain that going to work probably raised the women's incomes but that employment entailed work expenses. The increase in their incomes was spent on transportation, child care, and uniforms, leaving no money to actually make life easier. Since work expenses are not calculated in determining the federal poverty line, increased work hours and income translate into lowered poverty even if employment entails expenses that actually mean lower discretionary income for a family. No wonder my post-reform respondents described the same struggles as the women I interviewed before reform.

Additionally, studies found that as the years wore on, those who left the welfare rolls began to experience less positive outcomes than those who had left shortly after the implementation of welfare reform. This was so in part because the most employable recipients left the rolls early (or chose not to apply for welfare in the first place given the new requirements), leaving behind those less likely to do well. One clear indication that the welfare rolls are increasingly filled with families with more problems is that the number of families on welfare with active involvement with child welfare services has increased significantly since welfare reform.

The increasingly less optimistic findings on reform's effects are also due to the fact that the robust economy of the late 1990s gave way to a recession in the early 2000s. It is too early for a complete analysis of the much more dramatic "Great Recession" that began in 2008, but it undoubtedly has eroded the well-being of families who leave or decline to take up welfare, and it threatens to drive poverty rates to their highest levels since the 1960s.

Today's low-income families also face an environment of budget cuts to a variety of programs from which they might receive support. Since welfare reform ended the federal entitlement to cash assistance, state governments are free to make cuts to their TANF programs. Many in fact are taking money from their TANF budgets to fill other budgetary needs. Mary Jo Bane, who, as noted above, resigned from the Clinton administration in protest of the version of reform that was passed, writes scathingly in a 2012 article of the damage welfare reform did to the safety net that the poor need in recessionary times. She cites that between 2007 and 2011 the number of children in poverty went up by three million, but the number of children receiving cash assistance through TANF went up only by half a million. In 2010, this meant that of the 16.4 million children in poverty only 3.2 million were being served by the cash assistance program meant to address "needy families" in times of economic crisis. She concludes, "I continue to believe that the moral measure of a society is the way it cares for its poor and vulnerable. By that measure, we should all be ashamed."

The later quantitative findings are sobering. Welfare reform coincided with a dramatic decline in the caseload (70 percent between 1994, when the rolls started to decline, and 2008), but how long term and how far-reaching any actual improvements are in low-income families' lives is much less clear. The studies I have discussed provide evidence that many families have not seen positive changes as a result of reform and some have seen negative ones. There are multiple reasons for why that would be so, including the declining economy and a host of barriers low-income mothers face in successfully transitioning to work. This book argues that an additional important, yet mostly overlooked, reason is that welfare reform ignored the sources of distrust in low-income women's lives and hence was not able to overcome distrust's effects.

Why Trust and Distrust Matter

The more mothers like Bethany Grant and Susan Schiller trust others with whom they interact (their "interaction partners"), the more willing they are to take risks like responding to welfare policy's employment incentives, taking a job, leaving their children with a child care provider, marrying their romantic partners, involving their children's fathers in their children's lives, and relying on network members for help. The more they distrust their interaction partners, the more hesitant they are to take such risks.

In this way, they must make choices about trust in the face of risk, as did the diamond merchants in James Coleman's classic article on social capital. For the diamond merchants it was an efficient business practice to allow other merchants interested in buying their goods to take stones home to examine them. It saved time compared to having to conduct all such examinations in person. However, this practice risked the possibility that the diamonds would be stolen. The diamond merchants that Coleman describes trusted other merchants not to steal because they all came from the same, tightly knit ethnic and religious community. If their shared values were not enough to ensure ethical behavior, the substantial power of the community to impose sanctions was. The fact that the merchants trusted each other allowed the business of buying and selling diamonds to proceed unfettered by expensive insurance arrangements that would be needed in the absence of trust.

Coleman's study shows us that trust eases cooperation and the achievement of outcomes. Without it, the diamond merchants would not cooperate with each other. By extension, distrust, can thus be a barrier to both cooperation and action and might block business opportunity for the merchants.Francis Fukuyama captures these ideas when he describes trust as a "lubricant."

Such lubrication is particularly valuable when taking action entails risk in the face of uncertainty. Whenever Coleman's diamond merchants were correct in their assessments of others' trustworthiness, the risks they took by allowing others to take home stones paid off by making their work lives more efficient. If, however, they were ever wrong in their assessments and misplaced trust in the untrustworthy, trust would not have paid off. They would begin to practice distrust instead. Trust eases action, distrust blocks action, and trust may pay off-but only if one's partners in the interaction are trustworthy.

Surprisingly, when policy makers discuss the actions of low-income women, they consider a host of factors about the women such as their skills, their access to child care, and their work ethic, but they do not consider the role of trust. Yet trust operates for low-income mothers like Bethany and Susan in the same way that it does for diamond merchants. Unlike the diamond merchants, however, who know that the buyers have an interest in behaving in a trustworthy way (since they would be shunned by the community if they did not), the low-income mothers rarely feel that their interaction partners have an incentive to behave in a trustworthy way. As a result, the mothers are more distrustful than the diamond merchants. When they are right, their distrust protects them. When they are wrong, their distrust may cost them opportunity.

Uncertainty is ever present, and thus almost all action entails risk of some kind. It is the near-constant presence of risk that creates the need for trust in order for action to smoothly take place. One might argue that low-income parents face a particularly risky and uncertain world. This is especially true for low-income parents who live in neighborhoods of concentrated poverty, as many of the women I interviewed do. For one thing, the neighborhoods where they live and raise their children pose certain dangers. In Chicago (and cities as different from Chicago as Stockholm), rates of violence go up as the level of concentrated poverty in the neighborhood goes up. Gang activity and drug dealing may lead to acts of violence that affect bystanders. Children may be pressured to join gangs or to get involved in other potentially unsafe activities. As we saw in the Introduction, local gang members' grip on her son led Susan to quit her job. Another mother, Dolores Rios, pointed to recent shootings on the main street around the corner from her apartment as one of the primary reasons she did not want a job that would interfere with her ability to walk her children home from school.

The mothers I interviewed talked a great deal about their concerns over the safety of their children. Many other aspects of their lives were ridden with risk. When one lives in tight financial circumstances, any false move that results in a loss of income or similar setback is devastating. While the same uncertain behavior in a middle-class family may cost savings, upward mobility, or the ability to pursue opportunities for children, it can cost low-income families shelter, having enough to eat, or the ability to pay for heat. As we saw, Bethany lost her heat in the middle of a Chicago winter and almost became homeless after she risked taking a low-wage job and her caseworker "messed up" her case.

The sheer magnitude of the potential ill effects inhibits low-income mothers' risk taking. The existence of trust thus is particularly crucial for low-income families to take risks. Low-income families are also especially in need of the potential benefits that accrue to risk takers. However, the above discussion suggests that low-income families are less likely to trust those in their environment.

The assertion that trust and distrust are related to social advantage and disadvantage is supported by the research literature. Investigators have shown that those who have low levels of income and education, who live in communities that have high levels of income inequality, who are members of racial minority groups, who are women, or who have experienced trauma in recent years are less likely to trust.

In addition, trust is greater between partners who are on an equal footing. Familiarity breeds trust. Even if one is not personally familiar with an interaction partner, one may trust that partner if he or she is an agent of a trusted institution. But government institutions are less trusted by the disadvantaged-especially if they are deemed unresponsive.

Clearly, then, low-income mothers are less likely to find themselves under the circumstances that promote trust. Most notably, given their disadvantaged position, many of their interactions, especially those in the welfare office and the workplace, are with people with whom they do not enjoy equality. In fact, they are with people who hold substantial power over them. Even if they are familiar with their interaction partners, they appear less likely to trust the institutions that their partners represent, especially if these are associated with the government.

Their distrust in the government probably reflects in part the fact that low-income people have less power and ability to get the state to be responsive to their needs. It probably also reflects the more dire consequences of not getting the government to respond. Middle- and upper-class citizens also experience great frustration with an unresponsive state, but those interactions tend to result in high tax bills, minor unsuccessful battles with bureaucracy, and other such annoyances. When low-income people are unable to successfully interact with government agencies, they may lose the resources needed for their very subsistence, lose custody of their children, or face other extreme outcomes. Again we see that the precarious nature of their lives leads low-income people to become more distrustful and less willing to take the risk of trusting in the face of uncertainty.

In the women's interviews, trust and respect were closely intertwined. When the women were asked why they did not trust various others, perceived disrespect was one of the major answers they gave. The importance of respect cut across almost all of the five social settings studied. Rude caseworkers, rude bosses, unfaithful boyfriends, and gossiping friends and family members were all deemed disrespectful. The child care market was the only setting in which respect did not appear salient for the women.

One might think of respect as an outcome one wants delivered from an interaction partner. If one has reason to suspect that respect will not be delivered, perhaps from past experience, one will not trust that interaction partner. Alternatively, if a low-income mother's growing up as a member of a marginalized group has taught her to see the world as a place that treats her with suspicion and that is best approached skeptically, she may be quick to interpret actions as disrespectful and to distrust that disrespectful interaction partners will deliver on any kind of promises. Thus respect, though not the same as trust, is closely tied to it either as an outcome that one trusts will be delivered or as a trait of interaction partners that suggests whether they can be trusted to deliver other outcomes. Other researchers have similarly argued that for low-income and marginalized groups who are shut out of mainstream material success, respect takes on heightened meaning.

The Key Role of Structure

Many of the women I interviewed encountered interaction partners with an open mind, sometimes even a trusting mind, and learned through direct experience to distrust that partner. It was their direct encounter within a particular structure (such as a welfare bureaucracy, a low-wage workplace, or a child care market) that taught them their distrust. In Tally's Corner, discussed in the Introduction, Elliot Liebow writes,

Each man comes to the job with a long job history characterized by his not being able to support himself and his family. Each man carries this knowledge, born of his experience, with him.... It is the experience of the individual and the group; of their fathers and probably their sons. Convinced of their inadequacies, not only do they not seek out those few better-paying jobs which test their resources, but they actively avoid them, gravitating in a mass to the menial, routine jobs which offer no challenge to the already diminished images they have of themselves.

The distrust exhibited by the women I interviewed both before and after welfare reform similarly stemmed from "the experience of the individual and the group." Sometimes the experience was direct experience with the potential trustee. Sometimes it was experience not with the potential trustee but with people in the same role as the potential trustee (different caseworkers or different boyfriends, for example). Sometimes it was the experience reported by others in the community, such as mothers or friends. Sometimes it was experience reported more distantly in news media or rumor. Sometimes, it was a self-protective approach that mothers drew upon preemptively to combat potential disrespect or mistreatment.-a worldview or strategy based not on first-hand experiences or others' reports about the specific current situation but more generally on collective experience about how things go for those in poverty.

But even in this latter circumstance when the women's distrust was not based on specific experiences, it was very much rooted in structural circumstances. The mothers I interviewed sat at the bottom of the U.S. stratification system, in terms of both their material resources and their prestige. Coming from that position, they pulled distrust out of their "toolkits" to defend themselves against the stigma of their position and the threat of further material deprivation. A distrustful stance served as an armor to shield them from a world in which they had little other protection. And sometimes, perhaps, the power to withhold their trust was their one and only means of wielding power when they controlled so little else.

Many trust theorists argue that trust or distrust comes from calculating expectations about potential outcomes of an interaction. That is, people take into consideration what they know about their interaction partner's interests in the particular social context and what they have observed about the partner's behavior in order to calculate the chances that the partner will reliably come through for them. In other words, they make calculations about whether their interaction partner is trustworthy. These calculations dictate whether a taxi driver decides it is safe to pick up a particular fare or whether a bank gives a particular client a loan.

Much of the distrust that the women I interviewed expressed was indeed based on their collecting information about their interaction partners and, from it, making the calculation that their partners were not trustworthy. This information might be based on direct interactions themselves, similar past experiences, or the experiences of others they knew.

One of the key factors on which people base expectations about trustworthiness is whether they believe their interaction partners share their interests. Because of the women's disadvantaged structural position both in several of the contexts I study and more generally in U.S. society at large, their interests were often at odds with those of their interaction partners. In the five settings studied, low-income women found themselves interacting either with people who held power over them or with people who shared their powerlessness. When interacting with those who held power over them, they rarely felt they shared interests. For example, caseworkers' interest was in moving them off the rolls as quickly as possible while theirs was in maintaining benefits or in moving off only when stable jobs could be found. Similarly, employers' interest was in maintaining a flexible, cheap, and easily replaced low-wage labor force while theirs was in securing stable employment with advancement opportunity.

When the women were interacting with those who were in a more equal structural position (informal child care providers, boyfriends, friends and family), the likelihood that they would share interests was greater and thus the potential for trust was greater. However, their shared powerlessness in relation to the outside world, that is, their mutual relegation to the bottom of what I have called the U.S. stratification system, often meant that their interaction partners were scrambling to survive and hence could not be relied upon to take their interests to heart. The desperation that powerlessness creates means that everyone must be out for him- or herself at times. For example, the workers in Sandra Smith's book Lone Pursuit did not help their unemployed peers find jobs at their workplaces for fear that if these peers did not perform well it would reflect badly on them and they would risk losing jobs they needed to survive. As much as they might wish to help their friends and family members, they did not trust these others to behave acceptably at work.

The Structures That Welfare Reform Did Not Fix

I have described the women I interviewed as disadvantaged in two different kinds of structures. In U.S. society at large, what we might call "macro-level structure," they were at the bottom end of the class system. They had very little income, lived in dilapidated apartments in neighborhoods of concentrated poverty, had few options but to send their children to underfunded schools, and mostly had education levels below those required for jobs that pay a living wage. Welfare reform did not change these facts. The women I interviewed before reform did not differ on these dimensions from the women interviewed after reform. The quantitative studies reviewed above provide evidence that welfare reform did not change these facts for many other low-income women as well.

Welfare reform changed policies about the delivery of cash assistance. It did not ameliorate the forces that create inequality on the macro level. Most notably, it did not change the fact that globalization and increasing technological skill demands have eroded wages for low-skill workers. Income inequality has grown dramatically since the late 1970s. The rewards of economic growth have been concentrated primarily on those at the top of the income distribution.

Low-wage jobs do not provide enough income to lift families out of poverty. When I conducted both my pre- and post-reform interviews, if a woman worked full time and full year in a minimum-wage job, she would still be significantly under the poverty line. This was true not only for a mother of one or more children but even for a single adult. In 2004, 60 percent of families below the poverty line contained at least one worker, and 28 percent contained at least one full-time, full-year worker. Low-wage work is not an automatic ticket out of poverty.

Welfare reform also did not adequately address insufficient labor demand. While many assume that anyone can always get a job "flipping burgers," that is untrue, especially in urban areas of concentrated poverty. Anthropologist Katherine Newman discovered that in central Harlem there were fourteen applicants for every opening at the fast-food restaurants she studied. Even those who find jobs may discover that their employers do not reliably offer enough weekly hours to guarantee stable income or to allow welfare recipients to meet their required number of work activity hours.

In addition to not addressing these macro-level structures, welfare reform brought no change to the structure of the contexts in which women must choose whether or not to trust-often called mezzo- (or middle-) level structures. Here too the women I interviewed found themselves, at both time periods, in a disadvantaged position. Welfare reform established financial incentives for low-income mothers to find jobs and provided for a limited number of subsidized jobs, but it did not create any substantial changes in the way workplaces operated. Since low-wage jobs usually require few hard skills, low-wage workers are easily replaced. Their dispensability gives them little power to demand fair treatment.

Some night-shift employees of Walmart and Walmart's subsidiary Sam's Club, for instance, have been locked in stores overnight, in part to inhibit theft by employees. As a result, workers have been unable to escape even during emergencies. Sam's Club worker Michael Rodriquez learned of this problem when his ankle was crushed by a fellow employee driving an electronic cart. It took over an hour for a manager with a key to arrive at the store and release the pain-wracked Rodriquez. Similar events have involved a worker having a heart attack and expectant fathers delayed from attending their children's births. Academic researchers have further documented the poor conditions and indignities of low-wage work.

Welfare reform did not restructure workplaces. It did more directly affect the organization of welfare offices, but it left key problematic structures in place. Most notably, it added job tasks and responsibilities for caseworkers, but it did not make significant changes in the size of caseloads or training to absorb these enhanced workloads. Nor did it create incentives for caseworkers to treat recipients respectfully or fairly. Evelyn Brodkin's studies of Chicago-area welfare offices show that caseworkers routinely denied or terminated benefits for which recipients were eligible before reform and that these practices continued after reform. In support of Brodkin's findings, another study shows that administrative hearings before termination of benefits are often successful, demonstrating the illegitimacy of some terminations. Several additional studies show the difficulty post-reform welfare recipients have in working with the welfare bureaucracy to get the work supports that would ease transitions off the rolls. Another study finds that post-reform battered clients often do not receive the domestic violence services stipulated by reform measures.

Caseworkers have not faced a penalty for these types of errors either before or after reform. Instead, caseworkers have incentives to shed clients from their caseloads. One can hardly blame overworked and underpaid caseworkers from responding to such incentives. In her study of Boston-area welfare offices, Celeste Watkins-Hayes finds post-reform caseworkers under enormous pressure to meet the demands of welfare reform.

Studies have also shown that clear communication between caseworkers and welfare recipients was a problem before reform when recipients had trouble understanding welfare rules. A post-reform study finds that one of the reasons recipients are sanctioned (penalized through benefit cuts for not meeting requirements) is that they do not understand, and hence do not appropriately follow, welfare rules. This finding suggests continued problems in the communication of rules.

Reform similarly fell short in making structural change in the child care market. It created child care subsidies that made it somewhat easier for mothers to afford child care, but it did not address either the supply or the quality of child care services. Instead it has focused almost exclusively on giving low-income mothers subsidies so that they can go out and buy child care of the same quality that existed before reform, when mothers claimed that care of acceptable quality was nearly impossible to find in their neighborhoods. In fact, in the ten-year period between reform's passage and my last interviews, funding for child care doubled but only 4 percent of that funding went into quality improvements in the supply of child care. The other 96 percent went toward helping mothers pay for the same low-quality care that had existed prior to reform. And since the subsidy levels are dramatically below the market rate for many child care providers, it mostly only helps mothers buy the least expensive forms of care. Researchers have found that low-income mothers' lack of confidence in the quality of child care they can access has limited their ability to go to work since reform.

Reform removed some disincentives and created some incentives to marry. As long as their family income is low enough, married couples are now eligible for the main cash assistance program. Prior to reform, they were not eligible. Welfare reform also created a small pool of money for marriage promotion programs, including relationship-training classes that strive to teach low-income people the value of romantic unions and strategies for preserving them. Finally, by removing the permanence of cash assistance, which gave women some small degree of economic independence from men, welfare reform increased the value of marriage if a potential spouse can provide income to the family.

But despite these changes, welfare reform did little to change relationships between men and women (and boys and girls). Multiple researchers have documented that relationships between low-income men and women, including those in the post-reform era, can be marked by distrust and hostility. These characteristics of gender interaction can start as early as adolescence. Several studies show a relationship between women's distrust in men and women's choice not to marry.

Reform also did not change the structural circumstances of low-income men's lives, particularly the inhospitality of the labor market and the ever-present enticement of the drug trade and other illegal activities. It is thus not surprising that welfare reform has had little or no effect on marriage rates. As low-income women's education and employment levels rose over the 1990s, low-income men's actually fell. Women were pushed into the labor market by welfare reform and were also told that marriage was a solution to poverty. Yet as women pursued employment as a means to self-sufficiency, marriage appeared a less promising route given the declining economic prospects of men. Both men and women consider men's employment a precondition for marriage. Unemployment rates are particularly high for low-income African American men, as are incarceration rates. In 1999, for instance, the risk of imprisonment for African American men who left school before getting a high school degree was 60 percent. Having a conviction record creates additional barriers to labor market entry, and incarceration itself removes men from family life. In addition, men's incarceration lowers the probability of marriage.

Women's social networks were a key source of support for them both before and after reform. However, the need to care for network members also drains women of time and resources, making it harder to maintain a foothold in the labor market. Researchers have shown this to be true both before and after reform. Reform was accompanied by rhetoric about transferring responsibility for poor families from the government to their own communities, but as a policy focused on individual low-income parents it did not invest in those communities. Thus the structure of communities themselves and the resources within them were untouched by reform. In addition, by treating welfare recipients as individuals rather than as members of social networks, reform missed an opportunity to capitalize on the potential power of networks to share resources.

What do these policy shortcomings have to do with the trust levels of low-income mothers? The answer is, a great deal. Because welfare reform did not make fundamental structural changes on either a macro or a mezzo level, the social contexts in which low-income women's social interactions occur have remained relatively constant before and after reform. So, for example, before reform, Bethany probably went to caseworkers who had caseloads that were too big to manage effectively. These caseworkers were not sanctioned for being uncommunicative or for inappropriately denying benefits. The caseworkers Susan met after reform were, if anything, more burdened than they had been before reform. Now caseworkers face increased pressures and have greater incentives to get people off the rolls, even if this happens because of poor communication regarding benefits or inappropriate denial of them. In short, for both low-income mothers today and their counterparts in an earlier policy regime the structural setup is one in which caseworkers and recipients have opposite interests and probably feel equally (although differently) oppressed. These conditions produce hostile relationships between caseworkers and recipients. The nature of these relationships creates distrust of caseworkers on the part of recipients (and, undoubtedly, vice versa).

Social interactions in workplaces are similarly shaped by structural conditions. Bethany and many other women before reform held low-pay, low-autonomy jobs that often led them to feel mistreated by-and therefore distrustful of-their supervisors. For some women, this dynamic led to quick turnover. Susan's experience in post-reform workplaces was no different.

Similar stories, detailed in the chapters to come, characterize each of the five settings I discuss. The creation of distrust is related to how the interactions between actors are structured in the specific setting, and none of the settings I examine has changed dramatically since the implementation of welfare reform in the 1990s. Mothers struggling to raise children in poverty today are thus just as likely as similar mothers in an earlier time period to distrust and shy away from potential opportunities in almost every setting in which they find themselves.