We can help , but there ’ s a catch Nonprofit organizations and access to government-funded resources among the poor

Purpose – Today, low-income people seeking resources from the federal government must often work through non-profit organizations. The purpose of this paper is to examine the constraints that the poor must face today to secure resources through non-profit organizations. Design/methodology/approach – This is a conceptual paper. The authors review cases of non-profit organizations providing federally supported resources to the poor across multiple sectors. Findings – The authors find that to accept government contracts serving the poor, nonprofit organizations must often engage in one or several practices: reject clients normally consistent with their mission, select clients based on likely outcomes, ignore problems in clients’ lives relevant to their predicament, or undermine client progress to manage funding requirements. To secure government-supported resources from nonprofits, the poor must often acquiesce to intrusions into one or more of the following: their privacy (disclosing sensitive information), their self-protection (renouncing legal rights), their identity (avowing a particular self-understanding) or their self-mastery (relinquishing authority over daily routines). Originality/value – The authors show that the nonprofits’ dual role as brokers, both liaisons transferring resources and representatives of the state, can complicate their relation to their clients and the predicament of the poor themselves; the authors suggest that two larger trends, toward increasing administrative accountability and demonstrating deservingness, are having both intended and unintended consequences for the ability of low-income individuals to gain access to publicly funded resources.


Introduction
Over the past half century, the relationship between the welfare state and those in need of its assistance has gradually evolved. An important and long-noted part of this evolution is that the federal government has come to rely heavily on nonprofit organizations to distribute goods and services to low-income individuals and families (Smith and Lipsky, 1993). Relegating service provision to nonprofits has been described as part of the general retrenchment of the welfare state, wherein the US federal government has gradually reduced its direct involvement in support for the poor (Salamon, 1995;Allard, 2009). Nonetheless, this relegation has not necessarily resulted in less government involvement in their lives. In fact, the provision of resources has been a vehicle for substantial intrusion into the lives of the poor, and nonprofits have become crucial brokers in the process.
In what follows, we make a case that understanding these dynamics deserves greater theoretical and empirical attention, requiring a research agenda focused on how the conditions nonprofits face in our current welfare state regime affect what low-income individuals must either do or subject themselves to for receipt of government-funded support. A long literature has examined the complications involved in collaborations between government and non-profits in the USA (e.g. Smith and Lipsky, 1993;Salamon, 1995;Brinkerhoff, 2002;Marwell, 2004Marwell, , 2009Bryson et al., 2006). Our focus, however, is narrowly centered on the consequences of the government-nonprofit relation for what low-income clients seeking services and other resources must do or acquiesce to. While several researchers have pointed to the difficulties low-income individuals face when dealing with welfare officers (e.g. Brodkin, 2013;Lipsky, 1980;Soss et al., 2011;Watkins-Hayes, 2013), we suggest that related, and equally important, dynamics occur among nonprofit organizations, which due to their dependence on state resources may function less as an alternative to than an agent of government. As the welfare state has evolved, so has the brokerage role of nonprofits, which must comply with federal requirements to receive the contracts and grants on which many depend, and which inevitably respond to the direct and indirect signals by the state about who deserves support, how, and under what conditions (c.f., Marwell, 2004).
In a conceptual but empirically informed paper, we examine some of the consequences of the brokerage role nonprofits play for the organizations' relations to low-income individuals and families and for individuals and families themselves. Nonprofits have been shown to be powerful and important brokers of resources, at times making up for flaws and/or limits in government policy. We turn attention to an under-studied question: the constraints that result from government funding. Based on a careful review of published case studies across multiple service domains, we argue that to accept government contracts serving the poor, nonprofit organizations today must often engage in one or more of four practices: reject clients normally consistent with their mission, select clients based on likely outcomes, ignore problems in clients' lives relevant to their predicament, or undermine client progress to manage funding requirements. In turn, to secure government-supported resources from nonprofits, the poor must often acquiesce to intrusions into one or more aspects of their lives: their privacy (by disclosing sensitive information), their self-protection (by renouncing legal rights), their identity (by avowing a particular self-understanding), or their self-mastery (by relinquishing authority over daily routines). From the perspective of the poor, nonprofits are both recipients of the state's resources and institutional enforcers of federal imperatives, particularly, we suggest, around accountability and deservingness (Moffitt, 2015;Watkins-Hayes and Kovalsky, 2016). We call for a robust research agenda devoted to uncovering where, how, and to what extent these experiences form part of how the poor must access resources today.
We proceed as follows. First, we discuss the national context under which nonprofits provide government-funded services and other resources for the poor today. Next, we discuss the impact of government contracting on four aspects of nonprofits' approach to their clients. Then, we discuss the four kinds of intrusions that low-income individuals must acquiesce to, showing that while some would seem clearly necessary for the provision of resources, others appear to be arbitrary practices, the unintended consequences of state action, or the downstream result of high-level political actions. We conclude by calling for a research agenda that pays systematic attention to these questions at the level of the state, the organization, and the individual, and to fully understand, across domains, their scope, causes and consequences.

Welfare state
In the history of federal support for the poor, the period beginning in the 1980s has often been described as one of retrenchment and retraction (Salamon, 1989(Salamon, , 1995McMurtry et al., 1991;Wacquant, 2010). Policy makers increasingly believed that cash welfare support and social programs tends to produce "net harm" by incentivizing behavior contrary to responsibility and success, particularly with respect to work, marriage, and self-sufficiency (Murray, 1984, p. 218). Researchers such as Murray (1984) argued that federal welfare support made it more rational for mothers to remain unmarried and off the labor market than to marry and work. Politicians often described welfare recipients as women willing to bear children they could not support and expecting taxpayers to foot the bill (Gilman, 2014, p. 247;Hancock, 2004, p. 51). A series of legislative changes promoted workfare through the 1980s (e.g. Omnibus Budget Reconciliation Act of 1981; Family Support Act of 1988) (Caputo, 2011, pp. 31-37). By the 1990s, political commentators often discussed poverty as partly "self-inflicted" (Somers and Block, 2005, p. 268), and a prevailing narrative suggested that relief policies of the preceding decades had harmed the poor by perversely incentivizing dependency on welfare (Somers and Block, 2005, pp. 265-266).
These political trends culminated in the passage of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996, which replaced the federal Aid to Families with Dependent Children (AFDC) with Temporary Aid to Needy Families (TANF). The latter funded block grants to states for them to offer cash support to families under local state rules while meeting federal expectations to generally reduce the welfare rolls (Somers and Block, 2005, p. 268). TANF introduced, among other things, time limits, work requirements, and caps on the number of children that could be assisted. Over the ensuing four years, due to a number of factors, including robust, job-producing economic growth, the number of Americans on welfare decreased by 6.5 million, or 53 percent (Lichter and Jayakody, 2002, p. 119). By 2014, the share of poor families with children receiving cash welfare had fallen to 23 percent, down from 68 percent before the passage of PRWORA (Tach and Edin, 2017, p. 544).
Direct cash assistance to non-elderly able-bodied families decreased (Moffitt, 2015). Nevertheless, expenditures on other social security programs and welfare entitlement programs, such as Supplemental Security Income, the Earned Income Tax Credit, dependent tax credits and Medicaid, grew steadily from 1970 to 2015, and all forms of assistance toward the elderly increased (Moffitt, 2015). The major entitlement programs -Medicaid, Medicare and Social Securityhave been and continue to be expensive budget items (Centers for Medicare and Medicaid Services, 2018;Congressional Budget Office, 2016). Other programs, including those typically understood as welfare programs, constitute only a small percentage of a typical budgetin 2016, programs providing aid (other than health insurance or Social Security benefits) to individuals or families facing hardship comprised only 9 percent, or $366bn, of the federal budget (Center on Budget and Policy Priorities, 2017). In all, overall funding has increased but changed., as Moffitt (2015, pp. 742-43) put it: "(1) a redistribution away from non-elderly and non-disabled families to families with older adults and to families with recipients of disability programs; (2) a redistribution away from non-elderly, nondisabled single-parent families to married-parent families; and (3) within single-parent and married-parent families, a redistribution of transfers away from the poorest families to those with higher incomesthose with incomes just below and just above the official government poverty threshold."

Nonprofit organizations
Concurrent with these general changes, the federal government has increasingly relied upon and funded nonprofits to deliver social services, a trend that, coupled with welfare reform and rising incomes, has encouraged the expansion of the non-profit sector (Boris et al., 2010;Grønbjerg, 2001;Joassart-Marcelli and Wolch, 2003, p. 71;Smith and Lipsky, 1993;Salamon, 1995;Smith, 2006). As Smith and Lipsky (1993) have argued, the relationship between the government and nonprofit organizations has evolved in two important ways. One, "government contracting with nonprofits has expanded to meet a wider variety of needs," particularly increasing in fields previously restricted to privately funded organizations, such as daycare, homeless shelters, counseling, child protection, and legal aid, among others (Smith and Lipsky, 1993, p. 10). Two, the nature of the government-nonprofit relationship has changed. Governments are more likely to "contract for whole programs," to "create providers where they otherwise do not exist," and to institute extensive terms that "shape the sorts of services offered by private providers" (Smith and Lipsky, 1993, p. 10).

111
Governmentfunded resources among the poor Nonprofit organizations today constitute a critical component of social service delivery systems. Between 2003 and 2013, the USA saw a growth of 2.8 percent in the number of nonprofits registered with the IRS, with the total jumping from 1.38 million to 1.41 million (McKeever, 2015, p. 2). The true scope of the nonprofit sector is greater still, given the indeterminate number of unregistered nonprofit organizations (McKeever, 2015, p. 2). There are more than 30 types of nonprofits, including public charities, private foundations and other tax-exempt organizations (e.g. health maintenance organizations, advocacy groups, labor unions, business leagues, and social and recreational clubs) (National Center for Charitable Statistics, n.d.;McKeever, 2015). Public charities constitute the largest and fastest growing category (McKeever, 2015).
Among public charities, the two largest types are our focus: human service organizations (e.g. foodbanks, homeless shelters, youth services, sports organizations and family/legal services) and educational charities (e.g. booster clubs, parent-teacher associations and organizations and financial aid groups) (McKeever, 2015). These organizations are particularly important given the increase in the frequency with which the nonprofit sector has been called upon to deliver antipoverty services and in-kind income transfers ( Joassart-Marcelli and Wolch, 2003, p. 71;Smith and Lipsky, 1993), with individuals across the political spectrum in the USA looking to nonprofits to overcome the "alleged inefficiencies of both for-profits and government" in supporting the poor ( Joassart-Marcelli and Wolch, 2003, p. 71).
Nonprofit public charities receive their revenue from both public and private sources, including government contracts and grants, private charities, fees for services and goods from the private sector (such as tuition payment, ticket sales, membership fees and hospital patient revenues), charitable donations, investments and other income (McKeever, 2015). The two main mechanisms by which federal government allocates funding to nonprofit organizations are contracts and grants[1], although other forms of government financing, such as tax credits, tax-exempt bonds, tax deductions, vouchers, and fees-for-services, also exist (Smith, 2006). Based on a nationally representative sample of large charitable nonprofits in eight human service fields, Boris et al. (2010) estimate than in 2009 governments at all levels in the USA paid about $100bn to human service nonprofits, entering about 200,000 grants with about 33,000 organizations. Government funding comprised about 65 percent of their total revenue and constituted the primary funding source for 60 percent of human service nonprofit organizations receiving grants and contracts [2].
Funders hold large sway. Under most state laws, a nonprofit organization's governing bodysuch as a board of directors or board of trusteesis ultimately accountable for how these resources are used (Ebrahim, 2016). If the charity does not comply with funders' regulations, funders hold the power to "revoke funds, impose conditionalities, or even tarnish nonprofit reputations" (Ebrahim, 2016, p. 119). Funders may play a role in a number of issues, including defining who is eligible for services, what kind of services are offered, and how they are administered. As Smith (2006, p. 228) has argued, contracts are increasingly "performance oriented, with many contracts tying agency reimbursement to specific performance measures" (also Smith, 2016). Furthermore, funding schemes often tie payment to the eligible client, rather than the organizations, and may reimburse organizations at a vendor rate regardless of the actual costs of providing the service (Smith, 2016). Researchers have contrasted this state of affair to government contracting as late as the 1960s and 1970s, when, under cost-reimbursement contracts, agencies were reimbursed based on the contract terms and budget, rather than outcomes, and funding was tied to agencies, not clients (Smith, 2016). These changes in the contracting environment, as well as retrenchment following the 2008 recession, have increased nonprofits' revenue uncertainty, competition for contracts and clients, reporting requirements, and pressure to follow the norms and policies of government (Smith, 2016;Tschirhart and Bielefeld, 2012;Smith and Lipsky, 1993).
Nonprofits are a crucial component of federal and state governments' delivery of services. And they have been an important source of innovation, and even political maneuvering, in the context of service delivery (Marwell, 2004(Marwell, , 2009. At the same time, nonprofits themselves gain substantially from the billions in public funds spent in the USA on contracts with nonprofits of the kind we have discussed. These funds, however, are not without strings attached, and to the extent nonprofits come to depend on such funds they must respond to government demands that might well affect the organizations' approach to the families seeking their services and resources.
Two trends These demands have evolved over time. With respect to the services nonprofits provide to the poor and to how the poor are, in turn, affected, two particular trends are important: the federal push for administrative accountability and the continued demand for demonstrated deservingness. We discuss each in turn.
One of the most important trends has been the substantial increase in expectations regarding administrative accountability. Although accountability and evaluations have long been an important component of US public services [3], the passage of the Government Performance and Results Act (GPRA) in 1993 marked a "high-water mark of the performance measurement reform movement," when management philosophy that providers of public services should demonstrate results and accountability firmly took hold in the US federal government (Frederickson and Frederickson, 2006, p. 1). The law required federal agencies to specify their goals and objectives in strategic plans, to identify performance measures to evaluate whether these goals are being met, and to annually report their progress during the budgeting process (Leroux and Wright, 2010;Frederickson and Frederickson, 2006). John Mercer, who is said to have drafted early versions of the GPRA, described the law as an effort "to improve the effectiveness of federal programs as measured by their actual results, and to do this by improving the performance of those programs through better management" (US Congress, 2001, p. 3, cited in Frederickson andFrederickson, 2006, p. 2).
Nonprofits, in turn, have been subjected to demands for accountability through the federal funding programs on whose resources they rely (Leroux and Wright, 2010). With the expansion of "performance-based contracting," nonprofits are increasingly required to meet benchmarks identified by the government in order to be reimbursed for their services (Smith, 2016;also. Eikenberry and Kluver, 2004). Funders often require nonprofits to use tools such as logical models to link goals to outputs and track program efficacy, and to report these outcomes as proof of sufficient performance (Ebrahim, 2016, p. 107). A survey of 189 non-profit organizations throughout Indiana (71 percent of whom report receiving funding from Medicaid or federal, state, county or local government) by Carman and Fredericks (2008) showed that 71 percent produced reports for their public and private funders regarding program activities, 70 percent produced reports for funders about financial expenditures, and 55 percent experienced site visits by funders or regulatory agencies. Based on a nationally representative survey, Boris et al. (2010) recently found that 76 percent of human service nonprofits identified governments' reporting requirements (their complexity and time consumption) to be an issue.
In addition to the push for accountability, there has been pressure to demonstrate deservingness, a cultural, rather than administrative, trend that is different in nature but no less important. This trend reflect a gradual but long standing cultural shift in ideas about which poor are and are not deserving of public support (Moffitt, 2015;Watkins-Hayes and Kovalsky, 2016). Beyond reducing the number of people on cash assistance, the welfare reform of the 1990s typified a change in focus from categorical and means-tested programs for the poor toward entitlement programs for which more middle class Americans were eligible (McMurtry et al., 1991, p. 235;Salamon, 1989, p. 14). As Soss et al. (2011, p. 1) have argued, today's social programs are oriented toward "temper[ing] the hardships of poverty 113 Governmentfunded resources among the poor and ensure that they do not become disruptive for broader society," to "manage" poor communities, and to "shepherd" poor people "into the lower reaches of societal institutions" (Soss et al., 2011, p. 1). Today, and possibly to a greater extent than in the past, low-income people seeking institutional assistance must abide moral discourses that stress individual responsibility and personal reformation. These expectations, which are articulated often by politicians and other commentators, may well filter down through the work of government agencies into the behavior of nonprofits.

Nonprofits as resource brokers
The core predicament faced by nonprofits today may be understood by adopting notions from research on brokerage in social networks. Network analysts have defined brokers as actors that serve as intermediaries between two entities (Simmel, 1950;Gould and Fernandez, 1989;Burt, 2005). These intermediaries can play multiple kinds of roles, and may ultimately be in a position to serve themselves, other parties, or the system as a whole. Our focus is those who, at least in theory, benefit a system (Obstfeld, 2005), in this case, the system of resource transfer from the federal government to low-income individuals (Small, 2006). Nonprofits in this system serve as brokers.
Just as brokers can play different kinds of roles, they can also play multiple roles at once. In an important study categorizing kinds of brokerage roles, Gould and Fernandez (1989) identified two kinds of roles relevant to our discussion. What they call a "liaison" are brokers whose role is to link entities that did not have a prior relationship. Nonprofit organizations are clearly playing a version of this role by connecting federal or state governments and the poor. But they also play a different role, which in Gould and Fernandez' (1989) language would be termed a "representative," wherein the broker acts on behalf of one party to establish connections with outsiders. While nonprofits are liaisons with respect to their general standing, they are also, by dint of the contracting process, representatives of the government with respect to funding. The push for administrative accountability and demonstrated deservingness would affect them little as liaisons, but has a major impact on their work to the extent they are representatives, since in varying degrees they must act on behalf of the expectations of their funders. The different expectations behind these two roles, as we will see, often result in conflicts within the nonprofits themselves and with respect to their clients.

Nonprofits and low-income clients
The nonprofits' brokerage position in the current political and funding context has myriad implications. Yet discussing the full extent of administrative consequences of nonprofits' dependence on federal resources lies far beyond scope. Instead, and keeping with our focus on low-income individuals and families, we center narrowly on the possible consequences for how nonprofits approach clients. Based on a wide-ranging examination of published case studies, we have identified at least four possible consequences. Nonprofits may be forced to reject clients they otherwise might not, select clients based on likely performance outcomes, ignore problems in clients' lives not captured by accountability metrics, or undermine their clients' progress in some measures to manage accountability expectations. We discuss each in turn.

Reject clients consistent with their mission
One consequence of a nonprofit's dependence on government contracting is that it may be forced to reject clients its staff might want to help, or who might fall under its general mission but whose conditions place them outside the strict limits imposed by the government contract. Consider a case at the Abbott House, an emergency shelter for the homeless studied by Marvasti (2002). Abbot House is a charity organization whose principal mission is to provide a supportive environment and "free services and goods to 'needy' clients," in particular those who, "for whatever reason, find themselves hungry and homeless" (Marvasti, 2002, p. 620). The organization's director and house manager control fundraising efforts, manage the lunchtime soup kitchen, and enforce shelter rules to maintain order in the facility; its social worker is responsible for conducting interviews to assess client needs, determine the length of their stays, and make referrals to other social service organizations (Marvasti, 2002, p. 620). In this context, staff such as social workers and even volunteers determine "which clients will get what, when, and how often" (Marvasti, 2002, p. 620).
Consider the case of Paula, a woman in her late twenties who inquired at the Abbott House for rental assistance from Ann, the social worker. Since Paula was living with a friend at the time of her request, Ann rejected the application, explaining that Paula did not qualify for emergency services because she was not technically homeless. The "only rent money [Abbot House has] available is for people living in shelters," Ann told her. After additional pleas from Paula, Ann explained: "Well, unfortunately, a lot of times agencies' hands are tied because of limited funding or the rules that they have.
[…] I mean the rules on our rent program are prescribed by the federal government. If I don't follow the rules, they may decide next year they're not gonna give us the money. So, there are consequences if I don't follow the rules" (Marvasti, 2002, pp. 647-648). Though Paula's problems are in theory entirely consistent the Abbot House's mission, she was "not literally without a place to stay," according to the caseworker and thus could not be served (Marvasti, 2002, p. 647).
Similar dynamics are present in different domains. Based on survey, interview, ethnographic, and document analysis methods, Trudeau (2008) examined how nonprofits providing services to immigrants in Minneapolis-St Paul were affected by government funding. His fieldwork made clear that state rules and restrictions govern "how, to whom and under what conditions services can be provided" (p. 2812). An agency director explained: "We have little capacity to help clients who lose a job one or more years after arrival, or people who are stuck in inadequate employment. Our funders require us to focus on new arrivals and people on welfare. We had one former client who we placed into a decent position, but his employer never came through with healthcare benefits despite several promises to this effect. The client asked for help finding a better position, but we could only offer some advice and referrals. We could not provide 'hands on' help due to eligibility requirements" (Trudeau, 2008(Trudeau, , p. 2812. Technicalities of this kind, which in looser environments might be easier to accommodate, are difficult for organizations forced to act as representatives, to enforce strictly held rules under penalty of lost funding.

Select clients based on performance outcomes
A different way regulations may filter low-income individuals and families into and out of services is through contracting that narrowly focuses on outcome measures. Some clients may be selected in favor of others because performance outcomes are likely to be more successful or easier to measure. Some evidence has been uncovered in substance abuse facilities. Consider the work of Shen (2003), who examined nonprofit substance abuse treatment facilities in Maine specifically to assess whether performance-based contracting encouraged facilities to select less severe clients. The author analyzed admission and discharge data provided by the Maine Office of Substance Abuse for 1991-1995. Clients in the data are categorized by the severity of their condition, and those classified as "most severe" can be the most difficult to treat successfully. Shen (2003) found that after introduction of performance-based contracting the percentage of clients in that category was decreased by 7 percent across facilities, suggesting that organizations were dropping or avoiding difficult clients: "[…] in response to incentives introduced in 115 Governmentfunded resources among the poor Maine's performance-based contracting system, nonprofit providers may engage in activities to attract less severe clients because these clients were easier to treat in order to improve their performance" (Shen, 2003, p. 548). Performance-based contracting introduce a different dynamic. It is not merely that organizations may decline clients based on the latter's failure to meet very specific needs or demands; the nonprofits respond to incentives, and a focus on concrete outcome measures encourages directors to focus on those clients most likely to show improvement. The situation can be especially difficult for those facing conditions that may not improve much with treatment but are likely to worsen without it.
Ignore client conditions not captured by outcome measures A different set of dynamics might come into play once clients are admitted for service. Many researchers have documented that low-income families tend to face multiple different problemssuch as unemployment, health issues, criminal justice involvement, housing instability and the likeat once, that different forms of disadvantage can be interconnected. As a result, people may have difficulty overcoming one set of issues (e.g. unemployment) because of another (e.g. substance abuse). But a nonprofit's strong reliance on government contracts might force it to ignore problems or issues not captured by the particular outcome measures required by the contract.
An instructive case may be found in the context of job readiness. Consider the research of Johnson Dias and Elesh (2012, p. 149) on welfare-to-work programs, whose services include job-readiness classes, job counseling, educational classes, on-the-job training and job placement. The authors conducted observational, interview, and discussion group research at two welfare-to-work agenciesone for-profit and one nonprofitin a large northeastern city that would seem to be poised for a holistic approach to treating clients. In fact, the funding contracts "require[d] these organizations to utilize a case management approach" to achieve their goals, which involves the use of "individual assessment" and the development of "direct and personalized relationships" ( Johnson Dias and Elesh, 2012, p. 151). This approach may seem to be poised precisely to address the multiple problems underlying a client's lack of employment. Nonetheless, the contracts also specified requirements about client eligibility and the percentage of clients who must find jobs. And if addressing clients' employment needs did not "specifically coincide with efforts to meet a narrowly defined job-placement goal," clients found that barriers to their employment, such as health problems or lack of a high school diploma, might be ignored by agency managers and staff ( Johnson Dias and Elesh, 2012, pp. 151-152). They might even, according to authors, be referred for jobs for which they lack the necessary skills.
Undermine client progress to manage funding requirements Rather than ignoring problems, a nonprofit may feel forced to even slow down progress in the interest, ironically, of ensuring the client continues to have access to government funded services. A case from the health care context, and in particular elder care, is instructive. While restrictions in this domain have emerged for various reasons, a particular important one has been the need to remain in compliance with regulatory policies affecting reimbursement (Lidz et al., 1992, p. 89). When Medicare and Medicaid were rolled out in 1965, they provided "seemingly unlimited federal and matching federal funds for facilities able to provide hospital-like services in a hospital-like atmosphere" (Lidz et al., 1992, p. 34). While the nursing home industry received "a financial shot in the arm" for embedding itself in the medical model, it also found itself mired in "a maze of new regulations" (Lidz et al., 1992, p. 34). For example, Medicare allocated funding specifically for individuals over 65 years who required up to 100 days of convalescence in extended care facilities, and it also specified that facilities eligible for reimbursement were limited to those formally affiliated or maintaining a written transfer agreement for patients and their records with a general hospital (Lidz et al., 1992, p. 32). Lidz et al. (1992) have found that reimbursement-related regulations can have counterproductive and largely unintended consequences for the relation between elder care facilities and their clients. The authors conducted ethnographic fieldwork at a large geriatric facility owned by a nonprofit health system. The facility began as a small hospital but expanded to encompass two skilled, one intermediate, and one mixed skilled and intermediate patient units, along with a three-unit independent living area (Lidz et al., 1992, p. 40). The authors observed and collected data on nearly 100 individuals, including patients and residents, which gave them insights into how many of the regulations in the facility were in place. Some of the findings were surprising.
The authors reported a case wherein a staff member admitted that she would sometimes stop patients from completing physically beneficial activities during physical therapy because completion of these activities would affect reimbursement. Once the authors had turned off their tape recorder, this staff member "freely admitted" that employees will "sometimes stop a patient from walking 100 feet, regardless of whether the patient can walk that distance, if the length of that walk itself would necessitate termination of the patient's physical therapy reimbursement while other needs remain to be met" (Lidz et al., 1992, p. 103). There were other versions of this issue. The authors witnessed staff, when considering if discharge was "possible" for an elderly patient whose goal was to return home, take rehabilitation progress and remaining reimbursement into account and ultimately determine that the patient's goal was "unrealistic" (1992, p. 101). Indeed, the desire to comply with fiscal policies, often viewed by staff as external constraints, can influence patient care in ways that may or may not align with clients' preferences and perspectives on progress.
In sum, nonprofits may often find themselves engaging in practices seemingly contrary to their missions to respond to increasingly precise requirements by government funders. The push for administrative accountability has no doubt reduced wasteful spending by those nonprofits whose operations lacked the capacity to fulfill their goals or mandates efficiently. Nonetheless, the process has also altered many nonprofits' relationship to their low-income clients encouraging practices that in many ways cannot be described as consistent with the nonprofit mission. To be clear, the list of consequences we have discussed is not meant to be exhaustive. Nor do we suggest that all nonprofits under government contracts are forced to make one of these four decisions. Instead, we believe that the list makes clear that the service environment in which some nonprofits and the poor are forced to operate can be, in diverse respects, highly constrained. Nonprofits can be valuable brokers of government resources, a liaison for low-income individuals seeking support. But as representatives they can also be forced to make decisions in potential conflict with their missions. And when nonprofits find themselves in such predicaments, much of the burden for its consequences will fall upon the poor.

Seeking services
Many low-income families seeking services and resources from nonprofits avoid rejection, are selected into programs consistent with their needs, and receive the full slate of resources relevant to their condition. Yet the dual forces of administrative accountability and demonstrated deservingness may still structure their relation to non-profits. To gain access to resources, low-income families must at times acquiesce to practices that, in other contexts, might easily be considered extraordinary intrusions into one's life. Specifically, we argue that people may easily be expected to acquiesce to intrusions into their privacy (by disclosing sensitive information), their self-protection (by renouncing legal rights), their identity (by avowing a particular self-understanding), or their self-mastery (by relinquishing authority over daily routines). We discuss each in turn.

117
Governmentfunded resources among the poor Privacy When non-profit organizations receive government funding they are often required to gather and report increasingly detailed personal information about clients (Carman and Fredericks, 2008;Kissane, 2012;Williams, 1996) [4]. Consider Parents Community, a nonprofit organization offering a number of services to single homeless parents and their children, including transitional housing, early childhood education, case management, counseling, career and education contacts, classes, mentorships, technical assistance and other community resources (Binder, 2007).
The nonprofit's "most significant funder is the federal Department of Housing and Urban Development (HUD), through its 'project-based' Section 8 program, which provides vouchers to specific program sites, rather than directly to tenants, as its more well-known tenant-based vouchers do" (Binder, 2007, p. 552). Among the earliest federal subsidized transitional housing programs in the nation, and with an annual budget of $3.7m, "Parents Community receives mortgage assistance on the building, proper, but much more significantly, it receives approximately $700,000 per yearabout a fifth of its budgetin the form of rental subsidies to each of the tenants in the building's units" (Binder, 2007, p. 552). As is the case when receiving assistance from many housing organizations, income-earning tenants must contribute 30 percent of their gross adjusted income to rent, and the Department of HUD covers the difference between that and the market rate. Tenants without an income must pay at least $25, and HUD reimburses Parents Community for the remainder (Binder, 2007, p. 553).
The Parents Community's Housing department, which is one of its three divisions, is responsible for assigning clients to apartment units, managing the housing subsidies, and "enforcing HUD rules on rent remittance, guest stays, income verification, and facility management" (Binder, 2007, p. 561). In the process, it requires clients to divulge an array of personal information in order to secure and maintain access to transitional housing and other services offered at Parents Community. The Housing director explains: "Going through the process of applying for a program like this can be really intimidating. [Applicants] have to provide all kinds of documentation. They have to fill out an application […] It's not an easy process […] It is HUD's stuff. It is highly regulated. It is very wordy […] It is hard for many applicants to understand what we are really asking for" (Binder, 2007, p. 560).
The Housing department's role in the application process, which is "governed exclusively by federal guidelines," includes running background checks on clients' credit ratings, criminal records, and prior housing trajectories (Binder, 2007, p. 560). Some of the required information would seem to be necessary for the process. For example, credit ratings are routinely used in housing contexts to predict the reliability of meeting lease obligations. Nevertheless, as Binder (2007, p. 560) writes, "[i]f applicants have been evicted from any property, have a felony in the past seven years or a misdemeanor in the past five years, or have been reported to a collection agency for failing to pay subsidized housing rent, their applications are immediately terminated [5]." It is unclear why prior evictions should play a role, particularly since these can happen for any reason. For example, Desmond and Valdez (2013) have shown that landlords may begin eviction proceedings for a large number of reasons, including issues such as calling the police to report domestic violence, since landlords may receive nuisance citations for the presence of too many calls to their properties. In such cases, the disclosure of a past eviction only serves to constrain support for legitimately needy families.
The personal disclosures required can be especially difficult for clients or participants when they concern private health information. Consider the Discovery Center, located in the basement of Parents Community, which offers early childhood education (Binder, 2007). This department primarily depends on federal funding, including about $500,000 a year of federally funded yet state-administered child care payments and approximately $100,000 of annual funds from the US Department of Agriculture for lunch and snacks, among other pockets of federal money. The Discovery Center is regulated by the Department of Social Services, which imposes rules upon parents and staff regarding, among other things, children's health, which can be delicate.
For example, parents are required to sign contracts guaranteeing that their children will obtain annual medical exams, a practice that by itself would seem to be a good idea. The records of these medical exams, however, are then reviewed by staff and submitted to the state, giving access to personal medical information to staff at an institution whose primary purpose is not health but education. Binder (2007, p. 556) explains the predicament, noting that while staff members "say they try to accommodate clients' needs against these federal regulations when and where they are able," the Discovery Center's director "makes it clear that '[she doesn't] have a choice' when it comes to basic regulations that keep her department funded." As we have been discussing, the conflict between the role of brokers as representatives and brokers as liaisons becomes clear. Since the government provides funds, the hands of an even strongly motivated nonprofit can be tied. As the director explained: " 'We are regulated by the Department of Social Services, bottom line. I mean, we do not have much freedom as to what our rations should be in the classroom or what qualifications of teachers we should be bringing on board, or what rules parents and staff need to follow regarding children's health […] When it comes to regulations, they need to be met […], and I will not take a chance [of not meeting regulations]. I will not take that risk' (interview with Anna, the director of the Discovery Center, 2005)" (Binder, 2007, p. 555).

Self-protection
A different kind of expectation involves relinquishing actions an individual might normally engage in for self-protection, including legal self-protection. Consider the aforementioned Housing department at Parents Community. The department rejects about 40 percent of initial applicants. The other 60 percent, if approved by Housing staff in the first round, are sent to the Family Support department, whose staff must also approve them.
After the selected applicants have been approved by both departments, clients are required to not only disclose information but also sign many forms with legal implications. The director of the Housing department explains: "[New residents] have to sign a lot of stuff.
There is a release of information in there. There is an effective maintenance program in there. There is a lead-based paint disclosure in there. I mean things that kind of strike people, or are a little bit startling" (Binder, 2007, p. 561).
At times, low-income families balk at the process, in spite of their need. She recalls: "I had one resident who wanted to refuse to sign some of the paperwork in our lease. I said, 'Well, unfortunately, I count that as a refusal to accept an apartment, and you can't move in unless you sign it, because otherwise we can't do our job for you.' So eventually she signed it" (Binder, 2007, p. 561). The Housing department also updates information about clients at least once a year prior to renewing their leases. The director says, "[Residents] have to report certain things, and they have to sign certain paperwork, and we forward it to HUD. We have to verify everything. [Residents] have to report at least once a year to us […] Once a year they have to come in and do, basically, a renewal of their lease with us" (Binder, 2007, p. 561). She continues: "In between that year period, they would also have to do an interim renewal for updated information. They are required to come in and report it to us. If we find out that they have been making more income than they reported before, then sometimes it's difficult, and we have to go back and back and back, and retroactively charge rent" (Binder, 2007, p. 561). Some clients only meet with Housing staff once or twice a year to update their paperwork and undergo cleaning inspections, but those who repeatedly struggle to pay the rent or have other problems (e.g. noise or cleanliness complaints) experience more frequent interactions.

119
Governmentfunded resources among the poor Low-income individuals seeking services may find themselves complying with requirements and signing forms they normally would not but for the bureaucratically imposed requirements.

Identity
Sometimes to remain in good standing and receive services from certain organizations, individuals must acquiesce to a different sort of intrusion: the expectation that they will avow a particular identity or self-understanding. This practice is at times evident at "strong arm" drug rehabilitation centers, state-subsidized treatment facilities opened by drug courts and other jail and prison diversion programs (Gowan and Whetstone, 2012, p. 70). The courts provide both clients and funds to these facilities, therein transferring their own goals and orientation onto the rehabilitation centers' everyday therapeutic practices (Gowan and Whetstone, 2012, p. 70). Because of their strong ties to drug courts, probation, and parole institutions, and because they mirror these, the centers tend to exhibit "long residential stays, high structure, mutual surveillance, and an intense process of character reform" (Gowan and Whetstone, 2012, p. 70; see also Foucault, 1977). Gowan and Whetstone (2012) conducted fieldwork at Arcadia House, a strong-arm rehabilitation center for drug users in a large Midwestern city contracted by a local drug court. Arcadia housed up to 30 men at a time, two-thirds to three-quarters of whom were black, offering a 60-120 day program that combined "elements of intensive behavioral modification, Alcoholics Anonymous meetings, family therapy, and 'life skills' development" (Gowan and Whetstone, 2012, p. 75). The authors found that clients who arrived at Arcadia to undergo treatment and therapeutic reformation, rather than serve jail or prison time, were "propelled […] into the permanent category of 'addict' " (Gowan and Whetstone, 2012, p. 77). Indeed, Arcadia required that clients, upon entry, "publicly and wholeheartedly" avow the status of addictan agenda that encouraged new clients to admit that their consumption of drugs had incited "a pathological loss of self-control" and "insatiable hungers" (Gowan and Whetstone, 2012, p. 76). The requirement to adopt this self-understanding was universal, even for small-scale dealers opting for treatment to avoid harsh sentencing, many of whom had consumed little more than alcohol or marijuana. They were required to commit to a "drug addict" identity and to affirm that drug use had destroyed whatever sense of stability they had maintained before (Gowan and Whetstone, 2012, pp. 76-77). This requirement was sometimes jarring, as it could conflict with participants' more straightforward understanding of themselves.
Consider the following episode, involving Lamar, a client who had been sent to rehab after getting in trouble for "street-corner dealing"; Damon, another rehab client; and Silas, a staff member. When Silas tried to convince a group of clients to admit how seriously their financial stability had been hurt by drugs, Lamar protested: "Yeah, I have something to say about that. It (marijuana) didn't really affect my finances. I think it only becomes a problem if you start thinking you need it." Damon agreed: "Yeah, me too, I'm not addicted. You might think I have a problem, but I don'tthat's just what you think." Silas then asked the two if they had a drug of choice, and Damon said he preferred marijuana and he was not "some crack addict." Silas responded, "But what's your chemistry?" His voice rose: "[Your chemistry] is wired for dependency. Something happens to us, and we each have our own relationship to some chemical. The basic, bottom line is that your drug use is a medical issue. Now, your behaviors, that's what we need to focus on to fix it" (Gowan and Whetstone, 2012, p. 77).
Other researchers have written about models of this kind. For example, Mijs (2016) has described Safe, a government-funded nonprofit reentry organization in a northeastern US city that provides a variety of services to people who have been released from prison. Safe receives funding from local, state, and federal government and according to the author is the largest hub of government programs in the state. At Safe, in order to access services clients must agree to receive "their service provider's diagnosis of their problems and a prescription for a course of actiona road to reentry" (Mijs, 2016). Staffers tend to explain clients' path to incarceration in terms of societal forces, but in mapping their future road to reentry, they focus on their own choices and encourage them to avoid their old neighborhood and friends, and to instead seek Safe's community of professions and peers (Mijs, 2016, p. 292). Some of these actions can be double edged, and might lead to unintended consequences such as guilt and isolation. Mijs critiques the program for expecting people to represent their present conditions and futures as the product of their choices, not of poverty or any structural conditions. He characterizes this predicament as a "divorcing [of] clients' (structural) pasts from their (agentic) future," which "complicates the development of a self-understanding," isolates clients from their existing social network, and fosters self-blame and dependence on the non-profit organization (p. 305). He proposes that a more fruitful path forward might be to "focus on giving these vulnerable men and women the immediate tools they need to lead dignified lives: housing, mobility, and medical care" and to "aim for a relationship of interdependence between the reentry organization, the client, and his or her community" (Mijs, 2016, p. 306).
The practice of transforming clients' understanding in this way is consistent with the general trend in US policy in which recipients of public services must demonstrate their deservingness by representing an understanding of the self that underplays structural conditions and stresses individual responsibility and personal reformation. Non-profits may or may not see this transformation as essential to the service. And, indeed, values such as individual responsibility would seem worth supporting regardless of circumstances. But low-income individuals may find themselves forced to adopt a self-understanding not consistent with their own (as the dealer forced to "admit" to being an addict) or an account of their lives they may not believe (such as the theory that one's poverty was caused by one's choices, not the constraints in which they were made).

Self-mastery
A final issue people might be forced to acquiesce to is to relinquish basic autonomy over one's ordinary routines. The practice may seem extraordinary. Yet is has been observed across multiple domains.
One of them elder care. Elder care is a sphere that has become particularly saturated with guidelines; "bureaucratic regulations" have emerged that require facilities to "follow strict rules assessed through inspections," with some states enacting "even stricter" laws than those implemented at the federal level (Ulsperger and Knottnerus, 2007, p. 52). Given the large number of regulations under which nursing homes often find themselves, the elderly must grapple with the thought of relinquishing control over many of their daily activities if they want to reap the benefits of the institutional infrastructure (Foner, 1995, p. 231). Formal rules at times govern "what many would consider routine behavior," Ulsperger and Knottnerus (2007, p. 59) have argued based on an analysis of first-person and biographical accounts. For example, the authors found cases where residents hoped to retain "autonomy and dignity" and "take a bath without help," yet were confronted with laws that "prohibit it on the assumption that the resident will be safer with staff help" (Ulsperger andKnottnerus, 2007, p. 59). Foner (1995) examined a "200-bed nonprofit skilled nursery facility in New York City" (p. 231), hoping to understand how bureaucratic rules affected service provision. She reports on Ms James, one of the nursing aids: "Ms James' attitude toward dressing, bathing, and feeding patients was [that she] […] was determined to get them all done quickly, whether patients liked it or not. Residents, in her view, had no choice but to take prescribed medicines, eat so they would not be forced to go on tube feeding, or 'do a BM' so they would not get impacted. She had no tolerance for patients' resistance which slowed her down.

121
Governmentfunded resources among the poor Besides, she could get in trouble if, for example, their nails were not cut or their weights not done" (Foner, 1995, p. 232;also in Ulsperger and Knottnerus, 2007).
One can also look to the field of service provision to the homeless and unstably housed. Gerstel et al. (1996) track the development of the family homeless shelter system in Westchester County, New York, wherein shelters are subcontracted to provide "supportive housing" services. After the introduction of the Stewart B. McKinney Homeless Assistance Act of 1987, all funded shelters had to qualify for supportive status by offering "a core of essential services, most notably substance abuse treatment, mental health counseling, job training, and education for homeless children," which were assumed necessary for "achieving independent living" (Gerstel et al., 1996, p. 550). Moreover, shelter residents were forced to engage in no less than 20 hours of supervised activity per week (Gerstel et al., 1996, p. 552). These conditions coalesced into a comprehensive series of constraints on residents' everyday activitiesthe "conjunction of programs mandated by the various funding streams and the pressures placed on administrators who enacted these programs in the shelters"which the authors refer to as "therapeutic incarceration" (Gerstel et al., 1996, pp. 552-553). Rules included "prohibitions against in-room visits by outsiders, curfews for adults as well as children, and limitations on the amount of time that residents could spend away from the shelters" (Gerstel et al., 1996, p. 552).
Supporters argue that such rules promote client and staff safety and promote the stability that is often thought to be missing from clients' lives. Nevertheless, they have other consequences. They increase isolation by limiting interactions with friends, relatives, and the surrounding community; they enable systematic supervision over residents' personal lives by, for example, requiring that clients sign in or out every time they enter or leave the shelter and sanctioning searches of personal belongings and routine apartment inspections; and they encourage clients to submit to "therapeutic scrutiny" and "authoritative personal management" by personnel who have access to "potentially discrediting information and, to some extent, the power to limit opportunities for leaving the institution" (Gerstel et al., 1996, pp. 552-553).
Consider the case of Darlene, who faced eviction from her long-term residence when her husband was arrested and his share of the family support was cut off. Darlene chose to enter the shelter system rather than move in with her parents, as the latter would have been only a temporary solution; their apartment was not big enough to accommodate Darlene and her two children. While Darlene believed her case would be handled quickly, the diagnostic process "determined that she needed to officially divest herself of her abusive, criminal husband before she would be 'housing ready' " (Gerstel et al., 1996, p. 556). By the time this "complex process" was complete a subsidized apartment that Darlene was watching had been given to another family, and there were no other units available. Instead, Darlene was assigned to a "service-intensive shelter" in White Plains, where she was given "a full schedule of rehabilitation responsibilities, including psychological counseling and parenting classes." When Gerstel and colleagues interviewed Darlene six months into her shelter stay, she was "restless, discontent, even desperate." She told them she was eager for her husband's release from jail in three more months so that he could " 'rescue her' […] from shelter" (Gerstel et al., 1996, p. 556). Indeed, Darlene had entered the shelter system with intensions that were "diametrically opposed" to those of the shelter: she wanted a housing subsidy to replace the income that disappeared when her husband was incarcerated, while the shelter wanted to prepare Darlene for "gainful employment" and teach her not to rely on "AFDC, her abusive husband's drug income, or the informal economy" (Gerstel et al., 1996, p. 556). In the end, shelter life undermined Darlene's support network, especially her exchange relationship with her parents, while "her angry reaction to her extended shelter stay undermined other benefits she might have derived from the services to which she was assigned" (Gerstel et al., 1996, p. 556). Stuart's (2011Stuart's ( , 2016 fieldwork in Los Angeles's Skid Row is also illuminating. Stuart traces the rise of three "mega-shelters" in Skid Row: the Los Angeles Mission, a 306-bed facility; the Union Rescue Mission (URM), with over 1,000 beds; and the Midnight Mission, which has 360. Some shelters, such as the URM, are bound by "pay per bed" arrangement, meaning they cannot collect on government contracts until after they have rendered services. At one time this arrangement posed a challenge to the URM, as it is not always easy to convince individuals to enter shelters or accept their services (Stuart, 2011, p. 207). Stuart reports that some potential residents were "turned off by stringent religious requirements," while some "fear the racial discrimination, physical abuse, and sexual violence often reported by former patrons, choosing to sleep in the streets instead" (Stuart, 2011, p. 208).
The mega-shelters embraced the therapeutic model of service provision, which addresses "substance abuse, mental illness, poor life skills, and inadequate workforce training"which clients may or may not want help within the name of self-sufficiency (Stuart, 2016, pp. 62-63). In 2002, Los Angeles introduced a program called "Streets or Services" (SOS), which was intended to "funnel [homeless people], and the additional funds they represent, into the hands of service organizations under threat of incarceration" (Stuart, 2011, p. 208). SOS dictated that individuals arrested in Skid Row on misdemeanor charges be offered two options: go to jail or enter one of the mega-shelter's rehabilitation programs (Stuart, 2011, p. 208). If they chose the shelter, once inside they were faced with hosts of rules and with shelter conditions designed to discourage a "comfortable and prolonged residence" (Stuart, 2016, p. 65).
Consider Bernard, a middle-aged man who was detained for sleeping against the wall of one of the rescue missions in Skid Row. The detaining officers eventually uncuffed him and told him he could avoid jail time and eliminate his fines by participating in a program at the adjacent mission, where he would have to perform manual labor, sweep, mop, and prepare food. Bernard entered the program, and in addition to these tasks he was "mandated to participate in daily Bible study and 'work ethic courses' " (Stuart, 2011, pp. 197-198). In return for this labor, Bernard would receive services such as counseling, job training, beds and food (Stuart, 2011(Stuart, , 2016). Yet some individuals feel demeaned and even endangered by shelter conditions. Consider Darryl, who agreed to a 21-day residential rehabilitation program at the URM in return for having charges against him, for sitting on the sidewalk, dropped (Stuart, 2016, p. 4). Once inside, Darryl found that mandatory lights-out and early bedtime amplified his PTSD, and he reported struggling with insomnia and panic attacks (Stuart, 2016, p. 4). Constant check-ins mandated by the shelter made it "virtually impossible [for him] to return to recycling, his usual means of generating income" (Stuart, 2016, p. 4). The URM also "asked him to hand over his monthly General Relief check, insisting that doings so would teach him the merits of saving," which, according to Stuart (2016, p. 4), was the final straw that convinced Darryl to abscond from the shelter.

Conclusion
Important research in recent years has called attention to the centrality of nonprofit organizations to the ability of low-income individuals to acquire resources from the state, documenting their significance as brokers and pointing to their value as liaisons and innovators (Smith and Lipsky 1993;Marwell 2004Marwell , 2009Small 2006Small , 2009. In this paper, we have examined the other side of this processthe constraints involved in this form of brokerage. From the perspective of the federal government, nonprofit organizations are a straightforward way of outsourcing support for the poor; from the perspective of the organizations, the federal government is a source of funds to maintain its operations and enact its institutional mission; from the perspective of the poor, the nonprofit is the point of contact, the place that, as an organizational broker, provides access to information, goods and other sources from the federal government (Allard and Small, 2013;Isett et al., 2011; 123 Governmentfunded resources among the poor Small, 2009;see Obstfeld et al., 2014). This three-way relation, however, is shaped by the larger environment in which the actors operate, and, since the government is the ultimate funder, by the expectations imposed from the top. We have argued that the complicated dynamics that ensue can have detrimental consequences for the relation between nonprofits and their clients and for the decisions of the poor themselves, who are at times forced to acquiesce to substantial intrusions for the sake of government support.
These specific dynamics are not necessarily intrinsic components of the relationship between government funders and contracted nonprofits. Consider a case from Canada. The Manitoba Family Violence Prevention Program (FVPP) arose out of efforts to "establish, maintain, and improve services for persons affected by family violence" and is funded through service-purchase agreements with Manitoba's provincial government (Brown and Troutt, 2004, p. 12). When the government asked the FVPP to adopt these agreements, the FVPP instead insisted on assembling "a group consisting of representatives of all affected organizations as well as the FVPP staff" to consider the request, a process that "produced a well-informed group of participants cooperating to create documents that accurately link funding to the services for which the government is contracting" (Brown and Troutt, 2004, p. 14). The resulting service purchase agreements: are negotiated on a three-year basis, with "all organizations within a particular program set renewing service purchase agreements at the same time"; specify the service expectations of the FVPP as well as the funds that will be furnished; guide spending "in a way that makes the standards possible"; and include "detailed schedules specific to each organization" (Brown and Troutt, 2004, p. 15). In light of this history and these features of the FVPP, organizations funded by the program report "tremendous satisfaction" with it, while client satisfaction is "consistently high" (Brown and Troutt, 2004, pp. 17-18). Indeed, in this context, nonprofits liaison role appears to supersede their roles as representatives.
We suggest not that the practices we have discussed are inherently necessary, that they happen everywhere, that nonprofits have no choice, or that the worst experiences we have recorded here are typical. Instead, we suggest that many of these practices may be traced to well-documented national trends, trends that are either political or administrative in nature, and that a robust research agenda should be devoted to uncovering where, how, and to what extent these experiences form part of how the poor must access resources in the twenty-first century. Such research must clearly operate at multiple levels, including the federal, state, and local government funding sources, the nonprofits providing services, and the individuals and families that are affected. A comparative perspective that, as we have, addresses practices across multiple categories of service provision will be essential to determine the extent to which the practices we have discussed are prevalent. And a continued examination of alternative practices will yield both theoretical and policy contributions. 2. The eight fields were the following: human service multipurpose; housing and shelter; crime and legal-related; community and economic development; youth development; employment; food, agriculture, and nutrition; and public safety and disaster relief. The sample was limited to organizations that reported at least $100,000 to the IRS Form, and was weighted to represent the full population of US social service nonprofit organizations receiving grants and contracts in 2009 (Boris et al., 2010).