Homeless Services during the COVID-19 Pandemic: Revisiting Salamon ’ s Voluntary Failure Theory

: This paper examines government and nonpro ﬁ t relations in the context of federal homeless policy with speci ﬁ c attention to Continuum of Care (CoC) homeless networks during the COVID-19 pandemic. Applying Salamon ’ s voluntary failure theory, we conduct a qualitative study answering the following question: How, during the COVID-19 pandemic, did government address the voluntary failures of the nonpro ﬁ t sector? Results expand Salamon ’ s work by o ﬀ ering a more robust picture of what happens after federal funds are allocated to support the failures of the nonpro ﬁ t sector. Government funds both help and generate their own limitations, which are further alleviated by nonpro ﬁ t solutions.


Introduction
Homelessness in the U.S. has been a significant challenge facing communities small and large with roughly 582,500 people experiencing homelessness on a single night in 2022. 1 The COVID-19 pandemic posed unique risks and severe illness for people experiencing homelessness due to underlying and unattended medical conditions and a general disconnect from the healthcare system.The vulnerability of people experiencing homelessness was heightened when homeless services are often provided in congregate settings and shelter in place forced key nonprofit services reduced to its minimum (Finnigan 2021;Rodriguez et al. 2022).The National Health Care for the Homeless Council reports that COVID-19 positivity rates averaged between 9 and 12 % for people experiencing homelessness when compared to 7.8 % for the general U.S. population.2Some of individuals chose to stay outdoors in encampments due to the fear of contraction from congregate shelter where the bed capacity was reduced by providers to meet health and safety protocols.At the same time, the COVID-19 pandemic caused significant disruptions to government agencies and nonprofit providers of homeless services.Early results during the pandemic suggested that homeless service entities struggled to secure financial resources, keep their staff and clients safe, and adapt to a new reality of extensive technological use (Jang et al. 2021;Shi et al. 2020;Utah Foundation 2020).In response to the hardships caused by the pandemic, numerous federal policies and programs were implemented to support homeless service provision and prevention, including the Emergency Rental Assistance Program3 (ERAP), which provided rent and utility funding assistance to individuals at risk of experiencing homelessness.Little to date is known about the effects of these federal programs-particularly, the degree to which these efforts addressed inefficiencies within the nonprofit sector.
This paper takes a close look at government and nonprofit relations in the context of federal homeless policy with specific attention to Continuum of Care (CoC) homeless networks during the COVID-19 pandemic.These networks are created at the local level, with jurisdictions often spanning a city, county, or entire region or state.The focus of these CoC networks is to coordinate collaboration across the community, to pool local resources, and to strategize on ways to reduce the incidence of homelessness within their jurisdiction.During the COVID-19 pandemic, the U.S. federal government adopted policies that funneled financial resources to states and local governments.Following Salamon (1987) voluntary failure theory, we examine the complexity of relationships between the government and the nonprofit sector amid an unprecedented surge in demand.We analyze the degree to which these federal policies and interventions addressed the inability of the nonprofit sector to be effective providers of homeless services.Our work seeks to answer the following research question: How, during the COVID-19 pandemic, did government address the voluntary failures of the nonprofit sector?To answer this question, we relied on a qualitative case study approach with interview data of 26 homeless serving network leaders.In the following section, we review the literature on government and nonprofit relations with a focus on Salamon's voluntary failure theory.In the final sections, we discuss our research design, results, and directions for future research.
2 Literature Review

Voluntary Failure Theory
The relationship and collaboration between government agencies and nonprofit organizations have long held the attention of scholars across disciplines.Early on, Lester Salamon argued that more theoretical attention needed to be paid to this critical relationship, noting the shifting trend of government developing and funding policy instead of being the primary deliverer of services.Specifically, Salamon (1987) theorized that the government mainly funded human services and delegated the responsibility of implementation to other organizations, including nonprofits, resulting in a "third-party government."At the time of Salamon's writing, economic theories of the sector prevailed, explaining that nonprofits existed in communities because of inherent failures of both government and the market.
Against these prevailing theories, Salamon (1987) argued that there is a longstanding partnership between government and nonprofit organizations in human services and that these distinctive sectors agreed to share responsibilities to meet the unsatisfied demand for collective goods.Salamon's theory focuses on the government's intent to serve and ability to respond but also the condition of partnering with the nonprofit sector as the primary deliverer of services by compensating the nonprofit sector's inabilities.Salamon's theory turns our attention to the government instead responding to voluntary failures rather than nonprofit organizations filling gaps of government failures.Specific failures of the nonprofit sector further explained the natural connectedness of government and nonprofit organizations.Salamon suggested that we consider nonprofits as the primary response mechanism to local needs as opposed to the government and that nonprofits also had inherent failures prompting the cooperation of government agencies.Salamon (1987) voluntary failure theory identifies four voluntary failures: (1) philanthropic insufficiency, (2) philanthropic particularism, (3) philanthropic paternalism, and (4) philanthropic amateurism.These failures resulted in government and nonprofit interdependence developed by government supporting the nonprofit sector with the government's tax authority and democratic policy-making processes.
Philanthropic insufficiency suggests that nonprofit organizations lack the financial resources to carry out their charitable work in the community.Nonprofits may be unable to satisfy demand or expand in scale/scope, leaving some groups underserved or not served.Instead, through structured and consistent government support, those same organizations can grow and sustain their services to satisfy existing demand.On the other hand, philanthropic particularism refers to the tendency of nonprofits to focus their work on specific subgroups-leaving other groups without services.Leaving subgroups without services generates challenges of equity, diversity, and inclusion.In these cases, government funding can help these subgroups replicate their models and extend their services to traditionally excluded groups.Philanthropic paternalism reflects cases of influential community stakeholders with financial power having significant influence on nonprofit governance, which results in nonprofit agendas and missions driven by board members with wealth and influence that may not necessarily serve the best interests or needs of the broader community.Lastly, philanthropic amateurism suggests that the sector sometimes suffers from the lack of management capacity and professionalism, particularly less reliance on professional staff and field experts.This failure also reflects the long-standing tradition of relying on volunteers to run charitable programs instead of paid staff.
According to Salamon, the government is better off when public policies and programs remedy these inherent limitations of the nonprofit sector and take advantage of nonprofit organizations' service infrastructure, local knowledge, flexibility, and access to philanthropic resources.At the same time, government support of the voluntary sector helps nonprofit organizations serve more customers in need with comprehensive strategies to respond to a shortage of collective goods.Salamon's theory offers an analytical framework that explains the essential role of nonprofit organizations in providing government-financed services and describes a nonprofit sector that is a critical part of public policies and programs.Research applying Salamon's voluntary failure theory is voluminous, and some recent scholarship has studied the tools that partnership success will require (Salamon and Toepler 2015), the degree to which nonprofit involvement in advocacy is influenced by government funding (Salamon and Geller 2008), and how government and nonprofits make up for each other's weaknesses using resource exchange and contractual relationships (Salamon and Anheier 1998).
Since the pioneering work of Salamon, the literature on government-nonprofit relations has evolved with additional frameworks explaining these sectoral relations (Salamon 2002;Salamon and Toepler 2015;Smith and Lipsky 1993;Young 2000) and with a more robust understanding of the reasons that these sectors collaborate (Gazley 2010), the process that they undertake to make partnerships work well (Ansell andGash 2008, 2018), and the conditions that drive collaborations to positive and effective outcomes (Valero and Jang 2020).Other work using Salamon's perspectives has also explored the funding relationship between government and nonprofits, particularly the effects that government support has on nonprofit growth and density (Kim 2015), governance (Guo 2007), advocacy and the use of specific tactics (Lu 2018), and the effects of government funding in international contexts such as China (Yu, Shen, and Li 2021).Cheng (2019), in his study of park nonprofit's engagement in funding decision, found that nonprofit organizations also impact government funding decisions, and the government and nonprofit relationship is a two-way dynamic.More broadly, some studies have attempted to understand nonprofit density in relation to government services and funding by applying government failure theories and Salamon's interdependence theory and reported that lack of government services may attract nonprofit growth (Jeong and Cui 2020;Lecy and Van Slyke 2013).Ultimately, this study seeks to add to this growing literature on government-nonprofit relations by extending the application of Salamon's voluntary failure theory to the context of homeless policy where extensive collaboration between the sectors exists and is understudied.(Valero and Jang 2016).Community collaborations and nonprofit organizations working to address homelessness have long predated these federal policies and programs aimed at reducing homelessness across the U.S. Thus, this is consistent with Salamon's argument that nonprofits tend to be the primary and first deliverer of community programs.Current federal homeless policy (HEARTH Act of 2009) sets expectations for communities to build local level government and nonprofit partnerships that share responsibilities to eradicate homelessness.Nonprofit organizations are often tapped to lead the network and provide a large share of services for the community.There are 387 CoC homeless networks in existence across the U.S., and they bear responsibility for developing local policy and approaches to homeless programming.Even before the HEARTH Act was established, many communities already had homeless service coalitions, with nonprofits playing significant and key roles in congregating relevant stakeholders and working in partnership with other sector organizations.Current federal policy codified preexisting best practices into law modeled by long-standing community collaborations and HUD suggested informal practices (Jang and Valero 2022).Therefore, since collaborations predated federal policies and interventions in homeless services, community-based collaborations have served a vital role as the primary service mechanism for people experiencing homelessness (Kim, Sullivan, and Lee 2022;Mosley 2021).There is a growing body of literature that studies the interesting dynamics in building and leading CoC networks, including the tactics they use to engage in advocacy for greater resources and support (Mosley 2012), the role that nonprofit organizations play in leading CoC networks and leadership in networks (Jang, Valero, and Jung 2016;Valero and Jang 2016), and the variety of health care services provided by CoCs and their impact in the community (Jang, Valero, and Jeong 2020).

The Application of
Despite their best efforts to reduce the incidence of homelessness, nonprofits and CoC networks experience inefficiencies, which draws the support of philanthropists and of government.Nonprofits and other community-based groups struggle with adequate financial and human resources to support the growing homeless service demand.The CoC network as a collaborative mechanism is a community-driven action that aligns and distributes scarce resources and reduces fragmentation across the medical, public health, and social service sectors through community-wide planning and interactive decision-making (Kim, Sullivan, and Lee 2022).Individuals need diverse services to step out of homelessness, including case management, street outreach, legal services, health care services, mental health and substance abuse services, and others (Jang, Valero, and Jeong 2020;Mosley and Park 2022).However, it is well-established that communities have consistently struggled with pooling the sufficient resources to completely eliminate homelessness because of a number of factors beginning with a lack of affordable housing.
Nonprofit organizations are experiencing growing competition in their respective fields, reduced resources, and increased demand for community services (Ashley and Faulk 2010;Topoalogu, McDonald and Hunt 2018).Nonprofits serving people experiencing homelessness, such as housing entities, primary health care providers, hospitals, and food pantries, share similar experiences.Homeless policy also tends to be a highly politicized policy issue, with philanthropic funders, advocates, and other key stakeholders often playing a heavy role in influencing the policy direction of local interventions.In some cases, homeless collaborations are scarce, with agencies serving people experiencing homelessness working in silos and serving subgroups of this population, such as those with substance abuse needs, medical challenges, women and children, veterans, and/or youth.It has been observed that nonprofits have increasingly grown dependent on volunteers to carry out essential functions, especially in homeless serving spaces such as shelters, food pantries, and in yearly Point-in-Time (PIT) counts of people experiencing homeless.The pandemic exacerbated some of these conditions, stretching nonprofits thin and, in some cases, organizations closing their doors (Amandolare et al. 2020;Shi et al. 2020).These conditions of the nonprofit sector serving people experiencing homelessness, particularly during the COVID-19 pandemic, amount to multidimensional failures that create pathways for governmental interventions and funding supports to strengthen capacity and duplication of services (Jang et al. 2021).

Government Intervention During the COVID-19 Pandemic
Federal funding is a foundational resource for CoC networks and a significant benefit for establishing a community-based CoC network and keeping it running for the longer term.CoC participating agencies have developed programs and services based on federal policy guidelines to respond to the ever-increasing service demand since the McKinney Vento Act of 1987.Currently, the Continuum of Care Program utilizes three major federal funding channels for homeless service networks: yearly allocation from HUD to CoC Program funding, Emergency Solutions Grants (ESG), and Community Development Block Grants (CDBG).The CoC Program funding, which is allocated by Congress yearly, is distributed by HUD to CoC networks through a competitive and formula process based on community needs, performance, and other key indicators and serves as the primary federal funding for CoC networks.CoC Program funding supports permanent supportive housing, rapid rehousing, and other housing and shelter service activities.The Emergency Solutions Grant (ESG) is distributed using a formula-based process.While an ESG recipient does not have to be a member of a CoC network, they must consult with the CoC network in the recipient's jurisdiction on how the ESG funds will be used.ESG funding may be used in various activities such as shelter operations and services to offer individualized housing for medical concerns of people experiencing homelessness and people at risk of homelessness.Lastly, Community Development Block Grant (CDBG) funding can be used for rental assistance and other COVID-19 related responses without a cap for income-qualified households.
While strict rules and regulations govern these federal funding programs for homeless services, the COVID-19 pandemic generated new challenges and needs that led the federal government to suspend or waive particular requirements that generated much needed flexibilities for CoC networks. 6dditionally, and perhaps more importantly, the federal government responded to the nation-wide needs associated with the pandemic with the passage of key legislation.The Coronavirus Aid, Relief, and Economic Security Act (CARES) alone provided more than $12 billion for critical housing, homelessness, and community development programs in 2020.ESG and CDBG funding experienced an overwhelming increase in available funding due to the CARES Act allotment for homelessness.For example, ESG funding increased by $3.7 billion and CDBG increased by $1.7 billion between 2019 and 2020 (Gerken and Boshart 2020).The CARES Act also created the Coronavirus Relief Fund (CRF), which provides state and local government discretion to use funds to mitigate COVID-19 challenges by assisting public health programs to comply with federal health guidelines and providing economic support to those in need of emergency rental assistance and assistance in utility payments.Additionally, the FEMA Public Assistance Program supports public health compliance for unsheltered persons and outreach programs and services for non-congregate medical shelters.The American Rescue Plan (ARP) Act of 2021 provided $21.55 billion to help households with financial assistance, housing stability services, and to help with costs associated with other affordable rental housing and eviction prevention services. 7ltimately, these government funding programs were created to respond to the housing instability and to slow the spread of COVID-19 among people experiencing homelessness and demanded collaborative action between government agencies at all levels and nonprofit groups.The passage of the CARES Act, FEMA Public Assistance Program and ARP Act were essential in generating trillions of dollars to support communities by addressing housing and medical needs and implementing emergency protective measures and prevention of homelessness during the pandemic.Communities have received a historical scope of federal funding to prevent and reduce homelessness.What remains largely unknown are the effects of these programs on homeless serving organizations and particularly the degree and ways by which government addressed the voluntary failures of the nonprofit sector.
To examine government and nonprofit relations in homeless policy and the impact of government funding on homeless CoC networks' ability to overcome voluntary failure, we conduct an exploratory case study.The case study approach gives us an opportunity to answer our research question and use theory as a lens to explore the cases (Ospina, Esteve, and Lee 2018).To do so, we conducted semi-structured interviews with 30 CoC network leaders from 26 CoCs 8 during the summer of 2021(June-August). 9The CoCs that participated in the interviews were in COVID-19 hotspots and represented areas of the US with high incidences of homelessness. 10We sent e-mail requests for interviews to 80 CoC network leaders that we identified as having met these criteria and a total of 26 CoC networks responded favorably (response rate 32.5 %).A virtual interview was scheduled within one to two weeks after receiving a response from CoC leader.As shown in Table 1, various CoC sizes were included in the final sample, from rural to urban areas and major metropolitan regions.CoC leaders included executive directors, housing directors, board presidents, and other executive-level positions.Executive-level positions were selected because they had the most comprehensive knowledge of the CoC funding streams and budgets, volunteer and staff shortages, programs, and service capacity during the pandemic.Each interviewee signed a release form allowing us to record and transcribe the interviews.Each interview lasted, on average, 1 h.
During the interview, each interviewee answered a series of open-ended questions about federal funding, how it was used, and challenges faced by the CoC during the pandemic. 11The semi-structured nature of the discussion allowed participants the flexibility to discuss other issues and topics not covered by the interview guide, given the novelty of the virus and the limited knowledge base on the effects of the pandemic on homeless service systems (Svedin and Valero 2020).This data allowed us to understand the challenges and disruptions caused by the COVID-19 pandemic and how government funding addressed those challenges.Once the interviews were completed, each recording and transcript were stored and analyzed using content analysis and axial coding in NVIVO software.We conducted content analysis in repeated reading workshops led by the lead 8 We have a total of 30 interviewees participated in 26 interviews since some of CoCs had more than 1 interviewees participated in.20 % of interviewees were male.9 IRB 20-608.researchers to review transcriptions using coding and then organized the codes into overarching themes (Onwuegbuzie, Leech, and Collins 2012).Consistent with other qualitative research analyses, we also conducted axial coding, which involved selecting coding categories (Strauss and Corbin 1998).
For this study, the coding categories were organized into four conditions of voluntary failures, which Lester Salamon originally proposed.Salamon's theory helped us analyze the hardships CoCs experienced during the pandemic and the limitations of government interventions.While there was no consistent vocabulary to identify the categories and themes across all interview transcripts, we used an iterative process to define and refine the categories and themes during our series of workshops.We relied on multi-coders to improve the trustworthiness of data and to mitigate the implication of subjectivity. 12oC networks provide local solutions to homelessness in a geographic area.Because they are local solutions, each CoC network is unique in its location, size, structure, service area, and budget.Table 1 details the location, category, governance, the homeless population in the CoC coverage area, and how much the CoC received from HUD as CoC program funding in FY 2020.The geographic areas are described by rural, suburban, urban, and major city designations as defined by the U.S. Census. 13CoCs also vary in terms of their governance structure as having a "lead" or "network administrative organization" (NAO) according to how the interviewee described their network's governance style.A "lead" form of governance is considered more centralized and is depicted by a single organization and member of the network leads the whole collaborative unit of member entities.CoCs with network administrative organizations have stand-alone nonprofit organizations that represent the CoC and are charged with coordinating the CoC's management and community-wide activities (Jang and Valero 2022).

Findings: Voluntary Failures, Government
Responses and Nonprofit Solutions The qualitative analysis suggests that nonprofits indeed suffered voluntary failures during the pandemic and these failures were responded to by government programs and actions in some degree.We also learn from our analysis that government not only addressed the needs of the nonprofit sector but also generated its own set of limitations as experienced by nonprofits.Nonprofits then tap into their set of unique advantages to develop solutions to some of those government limitations.Thus, government-nonprofit relations during the COVID-19 pandemic in the context of homeless services is very much circular and cyclical instead of linear or static (as depicted in Figure 1).The review of interview transcripts was based on thematic coding of Salamon's voluntary failure theory and government responses to these failure situations.The initial interview transcript analysis informed us of emerging themes of nonprofit failures experienced during the pandemic.The interviewees also mentioned limitations of government responses they experienced and shared how CoCs continued services with nonprofit solutions offered by community foundations and other nonprofit organizations that ultimately helped CoCs' continued services.Accordingly, we coded the qualitative data based on: (1) voluntary failures (   Homeless services have been historically underfunded.Almost all interviewees discussed the lack of funding, which leads to a lack of resources, for the homeless community.Multiple interviewees discussed how they had been resource-starved for years before the pandemic.One interviewee described, "Homelessness is a place that is resource starved all the time."The historic underfunding became even more apparent during the pandemic when homeless services supplied by local government or other providers were either closed entirely or dramatically reduced.CoC leaders faced opening new shelters due to social distancing requirements, locating additional permanent supportive housing for people experiencing homelessness at risk of acquiring COVID-19 in a shelter, providing meals, and medical assistance and completing all these tasks with significantly reduced staff and volunteers.In the following quote, the interviewee described that service to people experiencing homelessness is under-resourced and the loss of staff and volunteers during the pandemic made the resource problem much more significant.
We always have had a scarcity.We have been under resourced in capacity for shelter for housing, and staffing.But nowhere did it become so apparent than during the pandemic because any staff that had any health issues or people who live at home with aging parents left the service and all of sudden you lost all of your volunteers.
Without resources, CoCs could not fund shelters or operate permanent supportive housing units to serve their homeless population adequately.One interviewee stated that they are not established enough to receive additional funding to purchase more housing options in the years leading up to the pandemic."We have more people that are experiencing homelessness than resources."Many interviewees stated the housing issue has not been improved because there were no affordable units to purchase in their area."We could not increase, permanent housing placement very much.We have so little permanent housing that is safe and decent."The major system gap has always been the lack of resources to house people.And one mentioned this is more of a system level issue that has to be changed by government action."We really need help from the government to make the housing market more accessible to our population and to also keep folks from becoming homeless because they cannot afford to stay in their homes."When CoCs could not house people experiencing homelessness due to a lack of available housing units, they still received an onslaught of service demands, from health care to food needs.CoCs faced two different types of resource insufficiencies Homeless Services during the COVID-19 Pandemic relating to service demands.First, CoCs did not have enough paid staff or volunteers.For example, a CoC lost its primary food pantry because it closed during the pandemic.Other CoCs could not find paid staff because potential candidates did not want to work near others without keeping safe distancing to avoid contracting the virus.Other staffing issues were reportedly caused by people being unwilling to work in the homeless system with low salaries.This quote shows how the fear factor caused serious shortage of staff during the pandemic and the resource failure that intensified during this time.
There was no staff in the beginning because of the fear.Now, I think staff had worked past in a fear and stress, but they are burned out and are leaving the field this time of pandemic.We have more programs, so we need more staff.
Second, their community partners were not able to provide agreed services.Several nonprofit community partners stopped providing essential services such as food, shower, laundry, mail services, etc.-leaving the CoC networks taking over all of service needs without the availability of alternative providers.The following quotes describe the scope of resource challenges that were not just limited to the shelters and housing providers but were felt by other providers who had to close their doors without much warning.As a result, CoC networks were forced to scramble and identify ways of addressing those gaps left unfilled by previous providers.
Many people just disappeared.Frankly, it was a little disheartening and frustrating because some agencies that just closed and didn't tell anybody they were closing.We found out because clients went to that agency and came back and said, "Oh, their lights are off and their doors are locked." All our soup kitchens closed, and now there was not a place for people to go and to get food and to charge their phones.We have some smaller nonprofits that serve [people experiencing] homeless[ness] with mail services, laundry, and showers and all that shut down.The big food pantry also shut down.So, we really scrambled to get those services in place.

Government Responses and Limitation
In some CoCs, immediate resources came from state and local governments, while federal funding took several months to receive.During the pandemic, local governments worked to connect CoCs to housing resources by securing motels to create non-congregate shelter spaces and setting up hand washing stations and porta potties at unsheltered encampments.And these initial funding sources from state and local governments helped CoCs build digital communication to create remote work systems.Some of interviewees indicated that state and local funding and resources were received quickly, efficiently, and with fewer restrictions than federal funds.
I have a great relationship with our county health department, and they were able to advise us early on.They were very responsive to our needs and our requests and partnered so closely with us to get the isolation quarantine shelter up and running.They also found ways to prioritize vaccinating our population, as well as staff working in direct service with people experiencing homelessness, despite the fact that maybe it didn't exactly fit the state guidelines.
Several of our agencies got money, straight through the governor's office, that our governor made available COVID funding for nonprofits.And it was a very loose application process and actually very quick as well.
After the CARES Act passed in late March of 2020, CoCs reported that they experienced an unprecedented influx of government funds.These public funds and assistance were crucial for shelters, emergency shelters, eviction prevention, and permanent supportive housing programs during the pandemic.Many CoCs we interviewed reported that these financial resources alleviated resource insufficiency, especially the ESG (Emergency Solutions Grant) funding through the CARES Act.One interviewee discussed the impact ESG had on their CoC network as follows.
We suddenly had access to a lot of funding through ESG.We were able to start a lot of programs through this funding.We started a street outreach program, which we had been missing for many years.It was an incredible transformation in our system to have this service available.
[We] also started a rapid rehousing program that was very, very helpful.
Although government support provided necessary funds and resources to CoCs, it was not a perfect solution to resource insufficiency.Time delays in receiving the funds were frequently reported during the interviews.An interviewee reported not receiving their federal funds at the time of their interview (August 2021), even though they should have had them months prior.Here is a quote as an example, "HUD and other federal agencies were putting out money quickly.But we didn't get it.In fact, some of the HUD money was distributed to the surrounding municipalities in May of 2020, and we still do not have it."Another interviewee reported that their CoC did not receive funding when the first round of CARES Act funding was released because they are considered a "rural" community.An interviewee discussed this unequal distribution of HUD funding initially."I do not know what formula they used to determine distribution, but initially they gave most of it to big cities and the rural communities were left with small amounts."This situation indicates an unequal distribution of resources and the hardship experienced in rural communities.

Nonprofit Solutions that Helped CoCs Compensate Government Limitations
In their striving effort to serve the most vulnerable during the early period of the pandemic, CoCs partnered with other nonprofit organizations within their network and community to overcome many government funding limitations and some of their struggles with resource insufficiency failures.Many interviewees praised their local community foundations for providing critical supplies, such as PPE and porta-potties, and unrestricted funds for emergencies, either through zero-interest loans or grants.Community foundations and financial institutions provided bridging funds to nonprofits partnered with CoCs until government money was received so nonprofits could continue to provide homeless services in the community.For example, according to an interviewee, the United Way fundraised specifically for COVID-19 relief and provided some of the raised funds to the local CoC network.While some nonprofits provided funds, others provided services and helped further the conversation on what needed to be done to assist the homeless community.An interviewee said this about Catholic Charities' contribution as follows.
It was our Catholic Charities that started this actual street outreach to hotel rooms, combined with rapid rehousing.They were a key part of services and part of a lot of the conversations as well about how we could use some hotel rooms for quarantine space when necessary.
Nonprofit organizations like Catholic Charities carry strong social capital, networks, and general capacity to step in and support CoC networks in their time of need.On top of funding and resources from the government and nonprofit organizations, CoCs also made new nonprofit partners into the network due to the pandemic.Interviewees reported new partnerships with nonprofit organizations that provided volunteers and staff to help with operations.One example is here; "We actually saw a number of new partners, which was exciting because it helped fill volunteer and staffing gaps."One interviewee shared that the CoC has made a new partnership with health care groups during the pandemic.She said, "One great thing that came out of COVID was some really good collaboration with folks at UNC Health Care, which is a huge health care system in our community."

Amateurism Experienced the Early Pandemic
Also, in the early months of the pandemic, CoC networks were overwhelmed with a significant increase in service demand without a strategy plan to scale their services and programs in a short period of time.An interviewee described their level of uncertainty by saying, "We had nothing to prepare us for COVID pandemic, despite many of our homeless service provider management and leadership [team members] have been working in this field anywhere from 10 to 25 years."Other interviewees discussed that their existing emergency and crisis plans, that were mainly to the natural disasters, were not useful to address the pandemic challenges.One interviewee said, "We were not ready.We didn't have a plan.So, we were just flying by the seat of our pants."The following quote shows the early frustration experienced by CoC service networks.
Our systems were just not prepared, even though we had the reminders of the Avian flu in 2008.
We did not have the infrastructure there to support remote work either.The capacity of the infrastructures, those things were just not there.
In addition to uncertainty and the lack of emergency plans, CoCs were overwhelmed with the volume of information exchange with it often contradicting messaging from federal, state, and local governments.Here is an example."Everybody got a different rule.So, it was very confusing."Many interviewees informed how, in the early months of the pandemic, they were trying to keep their heads above water regarding how they managed the information about pandemic response and safety precautions.The following quote captures the frustration of managing shelters when there were directions that were not consistent.
We just didn't know what to do.We didn't know how to plan, because we would make a plan and when we were locked down that was the best plan we had.Then when the governor opened it back up and said you can't shut things down.Then, after a while, it was governor said we can't enforce masks.It became like, "What do we do?"Some CoC leaders reported that a large size of CARES Act funding and other government funding sources caused them a big burden of administration.One interviewee expressed their frustration with those new funding sources, increased demands, and difficulty in recruiting new service providers to make these funding programs work.One interviewee shared as follows: "Dropped a $20 million in our lap [for] creating a rental assistance program.So, we have to coordinate all those referrals … But we wouldn't be enough to tackle [all] the programs.So, we solicited Homeless Services during the COVID-19 Pandemic new providers.And some of these providers are also brand new in this service and had to learn about this new rental assistance services."Several interviewees expressed the lack of administrative capacity to handle the large size of funding from the federal government that either through HUD or FEMA.One interviewee described that they had to decide how to spend money before the money was received."I'm having to make decision much faster and start posting [staffing] positions, even before the money is secured."This rush of decision-making before money was in hand was necessary to have time to build programs, such as hiring new staff and setting up the new operating system that was required preparation for those COVID-19 funding from the government.Here is a quote that shows that CoCs experienced an increased demand of professional management of government funding, but the overall CoC service system was not setup or prepared for the influx of money that had to be managed professionally by the rules of government.
Another challenge that we continued to work through is now that we've received four different rounds of additional funding that we normally would not, have received in a year, which is a great problem to have.But we had no administrative capacity to manage.
All interviewees reported that their work and meetings shifted to virtual platforms, but it was not an easy process for the most of CoCs.Here is a quote that described the challenges of a CoC in its transition to remote work."We all had to pivot to a virtual model but … we didn't even have a telework policy or option for our staff."Uncertainty and inability to properly build capacity to manage service demands and increased funding was complicated for some CoCs because of leadership or the board's unwillingness to adapt to the urgent demand from new environment.

Government Responses to Nonprofit Amateurism
As the pandemic progressed, messaging coordination between the city, county, state, and federal government improved for some of the CoCs.The government addressed much of the uncertainty surrounding the virus through clearer messaging and vaccination assistance.Some interviewees mentioned how helpful the CDC was on providing training, how state and county public health departments responded quickly to CoC needs, or how city governments paid for motels for non-congregate shelters when it was still unclear if non-congregate shelters were an official state or federal guideline.CoC networks also reported receiving flexibility from the federal government in spending and extending time period of expending federal resources.Following quotes are examples.
The county health department did a lot, and they are a really wonderful partner through this … They were part of our planning, and they brought the expertise of the public health response to things like how to do this in a way that prioritizes health.So, they were integral planning partners to what changes we are making as a system and keeping us updated on what the CDC recommendations were because those were changing rapidly and updating us on what we needed to do.
The state agency took a very active role as well, so they were regularly convening people to talk about messaging and getting information out.
We got a lot of support and training from the CDC doctors that were deployed for the testing effort.They came over to the hotel and helped train our folks and they did some PPE training, which was great.
However, not all CoCs reported excellent government responses to their amateurism failures.An interviewee explained that their health department was equally unprepared and offered the CoC no help or guidance.Here is a quote."The Department of Public Health was like a deer in the headlights and trying to figure out how a response for the rest of the public, not just for homelessness.It felt like every time we had a call with them or asked them something, they had no answers for us."Because government responses were not consistent for all CoCs, the CoCs leveraged their existing networks of community nonprofit organizations to provide services in a way that allows for leniency during the pandemic.An interviewee reported receiving support from local nonprofits to manage the coordinated entry process, which helped address their staff and operational capacity issues.Here is an example: "We hired technical assistance providers for other aspects of our CoC, including coordinated entry, and they've done a lot including putting together meetings with our other regional partners."Another interviewee stated that their CoC received service support from another nonprofit organization involved in homeless services."One of the biggest supports we had, came from Tipping Point Community (a poverty fighting nonprofit) in San Francisco.They have provided, very generously, targeted and specific help and data support."

Particularism
Salamon's original conception of particularism failure shows that nonprofit's strength of being focused on mission and serving a subgroup of population can also be a weakness of voluntary section.Focusing on the demand of a subgroup may not serve the broader population needs equally.We apply this original conceptualization of particularism to the case of homeless service to capture the service gaps that can occur within CoC networks.While CoC networks by policy design are supposed to address the continuum of care needs of people experiencing homelessness, they are not immune to creating strategy plans that focus on specific subpopulation groups (i.e.veterans) or to struggle in addressing specific needs of people experiencing homelessness because of the policy direction of their political environment.For example, most of homeless policy work is focused on adopting a Housing First approach which may lead CoCs to focus on housing related services as their main mission and leave public health and other service needs untouched.In our data, there are fewer instances of particularism failures in CoC service networks compared to resource insufficiency and amateurism.We, however, observed some important examples of particularism reported by the interviewees.During the early pandemic, there was confusion between the homelessness service network's response to the pandemic and the public health response to the pandemic.For example, one interviewee pointed out the difficulty to expand service responsibility to include public health services and initiatives during the pandemic situation when their expertise lies in homeless policy and services.
Our book of business is homelessness.So oftentimes people got confused between what is the public health response to the pandemic, and then what is addressing homelessness.
To serve people experiencing homelessness during the pandemic, CoCs needed public health services as well as other supportive services beyond their regular service focus.Some CoC leaders also reported that their networks were not behaving like a comprehensive and continuous service networks but more like fragmented and siloed service providers.An interviewee provided a poignant example of navigating homeless services as follows.
I could send you our system flow map, and no wonder [people experiencing] homeless[ness] cannot navigate our system.If you need a prevention, you have to go to 20 different providers.This situation creates a lot of confusion.
The following quote from one CoC leader exemplifies how a CoC network overcome particularism limitation by quickly integrating the mindset of pandemic public health in their service to people experiencing homelessness."We had to realize this is a response to a public panic, or public health pandemic that is layered on top of the homelessness crisis."To overcome the fragmented service system of CoCs, local governments acted as facilitators to connect service providers and community leaders to develop solutions with resources for services.In the early period of the pandemic, one city started communication for a real action plan immediately with the CoC and other service providers, and this city's leadership helped the CoC to engage in intense stakeholder relationship building with the city and other organizations.The following quote illustrates this dynamic.
It was the first time the mayor's office actually wanted our recommendations.That was powerful when the mayor's office invited the CoC and asked for recommendations.So, I would say that we are in a lucky position right now to be able to leverage relationships with city government.Some of the CoCs had connected with community stakeholders and other service groups to solve problems pre-pandemic.Therefore, they were comfortable calling local leaders or directors of city and county agencies to ask questions to avoid particularism pitfalls during the pandemic.Indeed, the deep connection among cross sector actors helped CoCs take new service roles and responsibilities that were in demand during the pandemic as indicated by an interview.
The collaboration is key.We sit on each other's boards for different pieces in the Human Service Delivery System.And we are always very engaged with each other's workforce and programs and efforts.So, we really know what is out there, what exists where the gaps are.They say we are going to find a way to help you even though we don't have that service.

Paternalism
Evidence of paternalism failures was also present, but at a significantly reduced number compared to insufficiency and amateurism.We defined paternalism to include instances in our data where there was a discussion on the lack of representation on the CoC board, the CoC board being disconnected from the community and service population, or that CoC membership was static.A CEO reported that her CoC restructured the board to be more representing the community from the service focused board to a community-based board including a person who has experienced homelessness too.
We had a service provider base board, meaning that all our board members were also homeless service providers.We restructured within the first 18 months to a community-based board.So now we have not only service providers but of course, businesses, faith community, hospitals, and of course people with lived experience … That way, we make sure that both the community voice, and the provider voice, and the peer voice is always heard.
To address these challenges of establishing a CoC board that truly represents the community it serves, the government has set federal law and guidance on the composition of CoC networks.For example, CoC networks are required to have a community-based board by including voices from the various sectors including hospitals, law enforcement agencies, school districts, housing organizations, and nonprofit providers and this expectation is laid out in the HUD Code of Federal Homeless Services during the COVID-19 Pandemic Regulations (Title: PA-T 578 -CONTINUUM OF CARE PROGRAM regulations). 14To be eligible for federal funding, homeless networks must report their membership composition in their application for HUD CoC program funding.

Conclusions
In the U.S. on a single night in January 2022, 17 out of 10,000 people experienced homelessness and about 30 % of them are family with children.Unfortunately, the homelessness rate has slightly increased by 0.3 % from 2020 to 2022 and this increase reflects a 3 % increase in the number of individuals that are unsheltered.A possible cause for the increase is the pandemic and its economic impact.Unsheltered people are the most marginalized in their access to health care systems and social services (2022 Annual Homeless Assessment Report).As a result, the COVID-19 pandemic disproportionately affected people experiencing homelessness. 15This study aimed to understand government and nonprofit relations by applying Lester Salamon's voluntary failure theory to homeless services during the COVID-19 pandemic.We specifically sought to answer the research question: How, during the COVID-19 pandemic, did government address the voluntary failures of the nonprofit sector?Using data from semi-structured interviews with 26 CoC homeless network leaders, we examine the degree to which these networks experienced Salamon's four types of voluntary failures and attempt to capture government responses to these failures.
Salamon's original work points to the advantages nonprofits bring to the table as first responders to the needs facing communities, but he also describes four dimensions by which nonprofits exhibit failures: resource insufficiency, amateurism, particularism, and paternalism.We used this framework to understand government-nonprofit relations in the context of homeless services and expanded the application of this theory to current reality, notably the COVID-19 pandemic.The findings offer us important insights into the dynamic interdependence developed between nonprofit and government sectors during a challenging time.Our findings specifically highlight that nonprofit homeless service providers experience inherent failures related mostly to resource insufficiency and amateurism.At the same time, government responds to these failures in dynamic ways, including funding through programs such as the CARES Act that allowed communities to utilize a large scope of resources to respond to the increasing demand for services during the early stages of the COVID-19 pandemic.
Figure 1 illustrates the connection across our findings and particularly that the relationship between nonprofits and government is not static or linear, but instead is circular with each sector stepping in to address inefficiencies in each other.In the context of homeless services and during the COVID-19 pandemic, we find that nonprofits and CoC homeless networks have been the first responders of homeless service needs in local communities across the U.S. both before and after the pandemic.These networks and nonprofit groups have engaged in innovative practices that in cases has drastically reduced or eliminated subcategories of homelessness, including veterans and chronic homelessness. 16However, the strategy to make homelessness rare, brief, and one-time is difficult and nonprofit organizations serving in this space did experience significant voluntary failures during the pandemic-mostly of the resource insufficiency and amateurism type.Government responded to these failures and the multidimensional needs of people all over the country, including those experiencing homelessness, through various strategic federal policies and funding programs.Government also served as a key leader and source of information and guidance to communities and CoC networks.Salamon's theory allows us to understand how nonprofit deficiencies are alleviated by government interventions.In extension of Salamon's theory, our work points to continued cycles of government limitations followed by nonprofit advantages that we categorize as nonprofit solutions.For example, the flow of public funding, was experienced as slow, inflexible, and complex by nonprofits serving people experiencing homelessness during the pandemic-signaling to limitations on the part of government interventions.As a result, we learn that nonprofits make up for the shortfalls of government by tapping into solutions unique to the nonprofit sector, including accessing philanthropic resources (individual donations and foundation support), community capacity to reach diverse demands, and established community networks.Thus, our work expands Salamon's voluntary failure by completing a fuller, more comprehensive picture on government-nonprofit relations in the context of homeless services during the COVID-19 pandemic.
CoC networks experienced increasing demand for services without the necessary or essential financial resources to address the need in their community.Amateurism was seen in a general lack of capacity to manage the large influx of resources.Particularism and paternalism were felt with lesser frequency when compared to amateurism or resource insufficiency.However, an example of particularism was the deep knowledge of homeless services, but not of public health approaches.Paternalism was observed as some CoCs lacking diverse participation in their governance process.As a result of the needs and challenges experienced across the country as a direct result of the COVID-19 pandemic, the federal government passed several measures that alleviated some of these challenges.For example, federal funding channeled to CoC homeless networks directly or indirectly via state governments helped expand housing services and keep those at risk of homelessness housed.The government also stepped in and provided capacity building support in the form of technical assistance or training to help equip CoC networks with knowledge about the corona virus.Other government measures included facilitating connection and partnerships across the community and improving communication efforts to keep CoCs better informed about the status of the pandemic and resources available.
However, the government response was not perfect.In fact, interviewees from our research indicated that government interventions carried limitations and failures.Thus, in our figure, voluntary failures experienced by CoC networks is followed by government response and later the experience of government response limitation as felt by nonprofit leaders.Government limitations included inflexibility with how and when funds could be used, slow response to the specific needs of homeless subpopulations, and a complex web of procedures that left some communities without the capacity to access those funding sources.In some cases, the voluntary failures already experienced by providers were exacerbated by the limitations found in the government's response efforts.For example, some CoC networks reported delays in received federal funding and that in some cases, rural communities did not receive federal support in early distribution of CARES Act funding.
Our data further illustrates that some providers could overcome the limitations inherent in government response by exercising what Salamon terms as nonprofit advantages or solutions.These refer to special abilities, resources, capacity, and experiences unique to the nonprofit sector.For example, some cases in our study shared how they tapped into private philanthropic resources, which they could do because of their tax-exempt status (501c3).They could solicit grant funding from private sources and offset the government's slow response to provide or reimburse funds.The fact that nonprofit providers were community-based and well connected to their local environment afforded these entities access to a community network of homeless providers and advocates to share knowledge, information, and capacity to respond to the unique challenges of the pandemic.For example, some CoCs developed internal and community response teams to keep all stakeholders informed and involved in the response efforts.These nonprofit solutions, along with federal resources, helped homeless providers to overcome their voluntary failures.
Our work has critical implications for both theory and practice.From a theoretical standpoint, our case study of CoC homeless networks afforded us a unique opportunity to apply Salamon's theory of voluntary failure to the case of nonprofit providers functioning within homeless serving community networks during the pandemic.Our study expands Salamon's work by offering a more robust picture of what happens during the early pandemic and after federal funds are allocated to support the failures of the nonprofit sector.Federal funds certainly help but also generate their own limitations, which are further alleviated by nonprofit solutions.Ultimately, these nonprofit solutions and federal interventions help address the complexity of failures facing nonprofit organizations within the context of homeless services.Thus, as illustrated in Figure 1, our work offers a comprehensive outlook at the process of nonprofits and government working together to address inefficiencies in each other to ultimately help a vulnerable population in our community.From a practical perspective, our work offers real-life evidence of the challenges facing homeless serving organizations across communities with high incidences of homeless and COVID-19 transmission.Lessons can be learned from the stories of these leaders of networks and the unique strategies they tapped into to overcome challenges revealed by the pandemic.Network leaders learned about the importance of developing emergency plans for future pandemic and emergent issues in the community, the value of building partnerships and cementing these to tap into additional resources and supports, and the need to strengthen their capacity to better manage and administer large government resources.Future work should further explore these stories in even more systematic ways (i.e.national surveys) to learn more about the effectiveness of government-nonprofit relations in the context of homeless policy and services.Additionally, future work should explore the degree to which our expanded framework on Salamon's voluntary failure theory holds true in other policy contexts.local government(s) (for example, city or county level government) play in supporting local coordination?4. How the COVID 19 testing has been administered?If confirmed what was the approach to quarantine those confirmed cases among the homeless population?
Later Days of the Pandemic & Into the Future 5. What plans does your community have for continued testing and now the vaccination of the homeless population? 6.Looking into the future, what resources, plans, or programs will be needed in order to stabilize service provision for people experiencing homelessness?

Figure 1 :
Figure 1: Voluntary failure theory in homeless services.
Salamon's Voluntary Failure Theory to Understand Homeless Service Policies and Programs We contend that Salamon's voluntary failure theory is a lens to understand federal funding and its effect on CoC homeless networks during the COVID-19 pandemic.By HUD's definition, homelessness refers to individuals or a family who lack a fixed, regular, and adequate nighttime residence (literally homeless). 4Nonprofit organizations have been essential stakeholders in implementing federal homeless policies since the McKinney Vento Act of 1987 and its reauthorization in the Homeless Emergency Assistance and Rapid Transition to Housing (HEARTH) Act of 2009 5

Table  :
10 Using data from the John Hopkins University Coronavirus Resource Center, we have identified the counties with the most coronavirus cases.https://coronavirus.jhu.edu/us-mapretrieved June 2021.Homeless population data was retrieved from the U.S. Department of Housing & Urban Development's (HUD) continuum of care program.https://www.hudexchange.info/retrievedJune 2021.11 Guiding interview questions are included as an Appendix.Continuum of care homeless service networks that participated in interview.

Table  :
Voluntary failures, government responses to failure and nonprofit solutions.