Service Provider Perspectives on the Differences between Place-Based and Scattered-Site Permanent Supportive Housing in Los Angeles County after the Onset of the COVID-19 Pandemic

Abstract

Background.

Permanent supportive housing (PSH) is an evidence-based solution to chronic homelessness. There are two common PSH models: place-based (PB) programs where clients live in one building with services provided onsite and scattered-site (SS) programs, which use community apartments coupled with mobile case management and support. Understanding the relative strengths and weaknesses of PB and SS is important for PSH planning and service delivery. This paper explores homeless service provider perspectives on these two models after the onset of the COVID-19 pandemic.

Methods.

Service providers (N = 37) from across 5 PSH agencies in Los Angeles that provided either PB or SS services during the pandemic participated in focus groups. Discussions were recorded, transcribed, and analyzed using template analysis, grounded theory, and inductive techniques.

Results.

Providers identified four major differences between PB and SS services: (1) challenges in finding placements; (2) managing relationships with landlords/property managers; (3) frequency of contact; and (4) community integration. Advantages of PB included ease of finding units, ease of managing relationships with landlords/property managers, greater ability to serve clients efficiently, more frequent client contact, and more community among residents. SS was seen to provide tenants with more opportunities to grow, live in healthier environments, and develop independence. During the pandemic, finding units for SS clients became more difficult, while differences between PB and SS related to frequency of contact and community integration became more attenuated.

Conclusions.

PB can be advantageous for clients with higher levels of acuity, whereas SS could be more appropriate for clients who are more stable and independent. PB programs are seen to have practical and logistical advantages, but some providers prefer SS services. Clients and providers should be matched to PSH configurations that best match their needs and preferences, and providers should be aware that public health emergencies may impact PB and SS settings differently.

1. Introduction

Each night in 2022, roughly 582,500 people in the United States experienced homelessness [1]. Nearly one-third of these individuals experienced chronic homelessness, meaning they had a disability and experienced homelessness for at least 12 months, or had at least four episodes of homelessness totaling 12 months in the previous three years [1]. Nationwide, the number of individuals experiencing chronic homelessness increased by 15.6% from 2020 to 2022 [1]. Permanent supportive housing (PSH) is an evidence-based solution to chronic homelessness that provides immediate access to subsidized housing coupled with comprehensive support services that may include case management, mental health services, substance use treatment, trauma-related services, primary healthcare, education services, job services, life skills training, support groups, social groups, clothing and food assistance, legal services, transportation, and assistance with benefits acquisition and maintenance [2–5]. Currently, there are two commonly accepted models of PSH. In place-based (PB) programs, nearly all residents in a building are PSH clients, and they receive housing services that are colocated within the building and delivered onsite. In scattered-site (SS) programs, PSH recipients lease private market units in the community using rental subsidies and receive support services at offsite locations or through field visits to their apartments [6, 7]. As communities across the United States continue increasing the number of PSH units they provide to address the homelessness crisis [1], it is critical to consider the benefits and drawbacks of PB and SS models and design service systems accordingly.

To date, research comparing PB and SS has focused on identifying the perspectives and preferences of PSH clients. This work has shown that PB programs may offer clients more supportive services and a greater sense of community than SS [5, 6] and be more effective in improving disability severity, psychological community integration, and recovery outcomes [8, 9]. However, participants often prefer SS placements [10], possibly because they dislike the rules, limited privacy, and what they perceive to be erratic behavior of others in their building in PB settings [8, 11].

Little is known about how PSH service providers view PB and SS. Provider perspectives on the strengths and weaknesses of PSH models are important to understand because whereas clients have expertise on one person’s experience in PSH (their own), providers have worked with many clients, giving them a broader insider perspective on the differences between PB and SS [12]. Insights on how providers understand the differences between PB and SS models and how such differences impact their work can be used to improve service delivery in both types of service settings. Moreover, the homeless service sector suffers from high levels of burnout, staff turnover, and workforce shortages [13–18], all of which have been exacerbated by the COVID-19 pandemic. The pandemic also brought new challenges to service provision and required swift and creative adaption by providers to ensure the safety of clients and continuity of care during a public health emergency. Even as the concerns of COVID-19 have largely subsided, the practices adopted by providers in response to the pandemic have transformed provider-client dynamics and revealed new insights into understanding of PSH and its implementation. Hence, it is critical to take recent provider perspectives on structural program design elements such as PSH service configuration into account when designing homelessness service systems.

The objective of this paper is to contribute to knowledge concerning the differences between PB and SS from the perspectives of those working in PSH settings in Los Angeles County (LAC). LAC has 20% of the nation’s population experiencing chronic homelessness [1], and a large home-lessness response system that devotes over $680 million to support services for PEH, including approximately 17,500 PSH units [19] that utilize both PB- and SS-PSH configurations. [7] Consequently, LAC service providers are particularly well situated to compare and contrast the differences between PB and SS-PSH models.

2. Materials and Methods

This study utilizes data from focus groups conducted as part of a larger study investigating the comparative effectiveness of PB and SS housing in LAC during the COVID-19 pandemic and its aftermath [20]. The research team facilitated focus groups at five LAC PSH agencies from June 2021 to August 2022, with participation from thirty-seven frontline PSH providers (case managers, care coordinators, social workers, and peer providers). Groups did not have any specific inclusion or exclusion criteria except that all participants needed to be individuals who regularly provide direct services to PSH clients. Each participant received a U.S. $50 gift card incentive at the conclusion of focus groups. All participating agencies provided a mix of SS and PB services, so participants were able to speak to the benefits and drawbacks of both service configurations. Two of the focus groups were conducted in person and three were conducted via Zoom, with each lasting approximately 90 minutes. Focus group facilitators engaged providers in the discussion through the use of a semistructured interview guide that asked open-ended questions about ways that services provided to clients in PB and SS programs differ, ways that different types of housing configurations may impact participants’ health, social integration, and quality of life, and how the COVID-19 pandemic presented differential challenges to either PSH model.

All focus groups were recorded and professionally transcribed. They were then analyzed using template analysis [21], organizing transcript data into a priori areas of focus that included differences between services offered in PB and SS, ways that PB/SS differences may impact outcomes, and differences in how the COVID-19 pandemic impacted service delivery in PB or SS settings. Text was transferred into templates, each of which was then analyzed separately by two members of the study team with expertise in qualitative research (all of whom are co-authors in this manuscript), who used inductive methods [22] rooted in grounded theory [23] to identify key themes that emerged from the data in areas of interest. This process involved open coding of text in templates to develop an initial code book that was later refined through group discussion and ongoing coding and cocoding practices. The lead author then initially reviewed all coded material and subsequent thematic findings and discussed any discrepancies that emerged between analysts’ findings with coders until 100% consensus was reached on the definition and interpretation of all study themes and findings [24, 25].

3. Results

Four major themes emerged from staff focus group discussions about the major differences between PB- and SS-PSH: (1) challenges in finding placements; (2) managing relationships with landlords and property managers; (3) frequency of contact; and (4) community integration. Participants reported that the COVID-19 pandemic impacted three of these domains: challenges in finding placements, frequency of contact, and community integration.

3.1. Challenges Finding Placements.

Providers reported that because SS clients can choose placements in the community and landlords have the right to refuse applicants, the process of finding housing is more complicated in SS than it is for PB clients. Focus group participants described how clients often want placements located in specific neighborhoods, either because of their general desirability or because of their proximity to friends and family. Often, it is difficult to find apartments that are affordable for PSH clients in the location where they would like to live.

In addition, providers noted that when clients want to live near people they know, it can be detrimental to their well-being if these individuals have had negative influences on them in the past. “Some clients want to be in certain locations,” elaborated one provider, “(but) we don’t wanna place them in a location where they’re not gonna be successful based on their mental health, past experience, (and) drug use.” In these cases, the provider explained “we have to work around that (client preference) and try to get them to see that maybe a different location would be best for them and they’d be successful there.”

Providers reported that even when housing that clients want is available, convincing landlords to rent to them is a challenge. Landlords and property managers, providers explained, “don’t trust somebody coming off the street,” harbor “stigma against Section 8,” and “have been burned (by PSH placements) in the past, (so) they don’t wanna work with (them) or they make it very difficult for a client to get in.” To address these concerns, providers need to spend time assuring prospective landlords and property managers who will support their clients, ensure that they pay their rent on time, and monitor them to prevent disruptive behavior.

Overall, providers reported that taken together, the process of trying to satisfy client preferences, safeguard client well-being, and find landlords willing to accept PSH clients is difficult and taxing. As one provider summed up,

“When it’s scattered-site, I (the client) tell you (the provider) I wanna live in Santa Monica or Venice Beach (expensive beach communities) and then you have to work with me (to find a placement there) even if the only opening is in Downtown or Pomona (areas that are far from the beach). Then after that, once I identify a community, you have to work with me to convince the landlord to take me in. Then you also have to guarantee that I’m gonna behave well.”

Many of these challenges were exacerbated during the COVID-19 pandemic, particularly for SS clients. Focus group participants reported that landlords were extremely reluctant to accept PSH clients or others with Section 8 housing vouchers during the pandemic because they were already struggling financially since many tenants did not pay rent during the COVID-19 eviction moratorium.

3.2. Managing Relationships with Landlords and Property Managers.

In comparison to SS sites, providers felt they had closer and more collaborative working relationships with landlords and property managers in PB sites. Since they are frequently colocated with landlord and property management services, providers can have frequent meetings or informal check-ins with them, allowing them to learn about clients’ disruptive behaviors or eviction risks early and intervene before they escalate. Having landlords and property managers working onsite also increases the likelihood that they will get to know the residents personally, creating the groundwork for relationships where they can become part of clients’ support systems. In addition, participants reported that property managers in PB sites have a greater incentive to be more forgiving of residents’ offenses to avoid having vacant units. As one provider shared:

“I think it’s easier to keep our clients housed because I mean, in my experience, I’m project-based, so property management already has a contract with HACLA (Housing Authority of the City of Los Angeles). They don’t want to have an empty unit. It takes a while to house a client once we find a match. It could be up to three months, so they don’t wanna have any free unit, so they’re more lenient when it comes to their infractions.”

Conversely, providers reported that SS landlords and property managers rarely interact with residents and “tend to see it as a business for them to get paid.” Consequently, they are less forgiving when residents fall behind on rent or paperwork since they can easily find other tenants who are not receiving receiving public assistance to fill their units. Moreover, because they do not know their residents as well, SS landlords and property managers have less empathy and lack patience needed to develop good relationships with some PSH clients. Landlords and property managers who are less accustomed to working with residents who have behavioral health conditions often find PSH residents “abrasive,” and providers need to intervene to assuage tensions between landlords/property managers and their clients more frequently in SS settings.

This is particularly challenging, providers explained, because they need to manage relationships with a number of landlords/property managers across buildings where SS clients are placed. “You have 20 different clients,” explained one provider, “that means you have 20 different property managements, probably 20 different phone calls or meetings that you have to tend to. Maintaining that relationship with however many property managements you have to deal with (can be difficult).”

3.3. Frequency of Contact.

Providers reported that having all clients concentrated in one place in PB programs has many advantages. This is especially the case for PSH residents who may have a hard time following through on appointments with housing providers and other service agencies. “The more you have to get people to follow up with things like that,” explained one provider, “the more likely they’re gonna fall through the cracks and become discouraged.” What makes PB programs “remarkable,” the provider continued, is “having support staff there so they (clients) can actually go down (stairs), knock on a door, walk into the lobby, and get help. That is the whole difference for a lotta people.” As another provider explained, being onsite is beneficial for clients since “whenever they need me, they go to me (with) whatever needs they have at the moment.” Beyond making it easier for clients to reach providers, focus group participants reported that in PB programs they are better able to provide the daily support that some clients need. “I have to educate them (clients) on their substance use, daily reminders of either taking their medication or paying their rent or maybe keeping in contact with their mental health provider,” explained one provider. “I think that’s very beneficial for them, just having us onsite because they really need that support.”

Moreover, because PB providers’ offices are in the buildings where residents live, they “have more contact, more communication with clients, more engagement” than they would in SS configurations. More frequent interaction, they reported, helps them develop stronger, more trusting relationships with clients. “Because they’re working in the building that you (the client) are living in (and) are there more for more hours of the day, you feel much more comfortable with them,” explained one provider. As another elaborated, “(in) place-based (programs) I feel like I have a much better relationship with my clients because I see them.” Knowing clients well and seeing them frequently helps providers not only with rapport-building but also with awareness of clients’ well-being. As one provider explained,

“I have somebody (as a client) who does recycling (collecting cans for money), so I see him every single day, going in and out with a bag. I even know the schedule. Now, if that doesn’t happen for a couple of days, I get worried and then I need to know what happened. Now, if this participant lives all the way in San Pedro (over 20 miles away), I will not know the pattern (and not know that the client may be having problems).”

Conversely, in SS programs, providers reported that they often do not know if clients are struggling due to infrequent contact. “They could be sitting in their apartment for two weeks and you wouldn’t notice,” explained one provider, while others told stories of clients who had substance use relapses, experienced drug overdoses, or abandoned their apartments without their providers knowing until after the fact.

Providers also reported that in PB programs, it is much easier for them to conduct wellness and welfare checks for their clients. In PB buildings, providers said that they simply “go knock on the (client’s) door” to check in on clients at least once a week, but “that doesn’t happen regularly at scattered-site. They’ll do the minimum we’re required to do-one visit per month or at least a phone check-in.” Providers explained that, in PB programs, they can check on their clients’ welfare in minutes, while, in SS programs, they reported the same task could take nearly an entire day.

Despite these issues, focus group participants also pointed out several advantages of SS compared to PB configurations. Most notably, providers reported that an unintended consequence of the convenience, accessibility, and frequent contact is that PB programs foster programs foster client overdependence on their providers. “What we’re supposed to do (is) help them (clients) reintegrate back into the world,” explained one provider, “(but) sometimes because they know you (providers) are there, they won’t do things on their own. They’ll just wait for you to do everything sometimes. They won’t just take the lead–they’ll just wait.” In several focus groups, providers described PB clients as being overly demanding, “tugging and pulling” at them for attention instead of solving problems on their own. As one provider elaborated,

“I think clients adapt based on what’s accessible to them. So in the case of PB, I notice that because clients are there and they have case managers onsite, they may frequently use case managers to do things for them because in their mind it’s “you (provider) are here to support me.””

In contrast, providers reported that in SS programs “they become a little more independent based on adaptability, because they have to, because we’re not there all the time.” For some providers, the independence that SS configurations foster is beneficial not only because it makes their job easier, but because it helps clients grow and make meaningful progress. “It’s rewarding to see people’s process and (how) people grow from where they start (in SS),” explained, one provider, while another believed that “SS is best for them because (in SS) I think they are actually really independent.”

Another service delivery-related advantage of SS programs, according to focus group participants, is the variety it brings to their job. As one provider who used to deliver PB services explained, “(I want to) move around. (When I) was coming to one building and sitting there for nine hours, I was going crazy. I was like “I need to be able to come and go as I please.” I feel like scattered-site allows me to do that.”

Providers reported that during COVID-19, some of these differences between PB and SS were attenuated because of shifts in the mode and frequency of contact necessitated by the pandemic. When serving PB clients during the public health emergency, providers reported that “it was more on the phone, just because there was a stop in home visits due to the uncertainty” of how to best prevent virus transmission when seeing clients face-to-face. In this respect, the pandemic made client contact in PB settings more similar to how it normally operated in SS programs. Conversely, the pandemic led providers to contact their SS clients more often, giving them more frequent contact similar to that they normally had with PB clients. “Scattered-site, pre-pandemic, the requirement was at least once a month engagement in person,” explained one provider. “During the pandemic it was required for us to get engaged weekly in order to ensure that clients (were) safe and make sure everyone’s fine, that they knew what they needed to do (regarding COVID transmission and care).”

3.4. Community Integration.

In PB sites, providers reported that clients’ physical proximity to each other, combined with classes, groups, and recreational activities that programs run onsite, help foster a sense of community among residents. As one provider summarized:

“I feel like in my PB (program) what we have is a little community. All the clients meet. We have tenant meetings monthly and they discuss issues. They tend to get along, they tend to help each other out. If one of my clients sees that (someone) needs food or something, another tenant steps in and would feed him for the day, or they take turns just keeping an eye on him. They keep an eye on each other. They help out each other. . .they don’t feel as alone. They become family.”

However, providers reported that close and frequent contact with other residents can also create challenges. Since PSH residents all have histories of serious mental illness, substance use, and/or trauma, PSH programs are, according to one provider, “a concentration of people who have a lot of issues and dysfunction.” Thus, rather than supporting recovery, living in an environment where many residents may face serious emotional and functional challenges can make it difficult for residents to stabilize in PB-PSH.

Nonetheless, providers reported that the lack of community and support for clients in SS configurations can be highly problematic for clients. “They get lonely,” explained one provider describing SS clients, “You go to a scattered-site where it’s just you. You have a lot on your mind so you need the extra support more than someone coming out once a week.” Other providers concurred, describing how SS clients tell them that “I wish there was somebody for me to talk to that is right here, or that I could see you (the provider) more often.”

Constraints created by the COVID-19 pandemic limited opportunities for community-building in PB settings, making them operate more like SS programs. Focus group participants explained that to maintain social distancing and avoid COVID-19 transmission, PB programs moved group activities that normally took place in person to online formats, leading to decreases in attendance and engagement. The pandemic also forced PB programs to eliminate open spaces where clients could just drop in and socialize. “Each (PB) building has a community room, and one of the things we do there is foster community spirit,” explained one provider. “COVID closed down the community rooms, and we’ve had difficulty fostering that spirit and having activities for them to do that we would normally have.”

4. Conclusions

From the perspectives of PSH staff who participated in this study, some of the advantages of PB over SS include ease of finding units, ease of managing relationships with landlords and property managers, greater ability to serve clients efficiently, more frequent contact with clients, and a greater sense of community among residents. On the other hand, SS sites provide clients with more opportunities to grow, live in environments where their neighbors do not have serious problems, develop independence, and become less reliant on their service providers over time. Notably, many of the reported benefits of one model (e.g., high levels of support in PB programs and the greater development of independence in SS) corresponded to weaknesses of the other (e.g., greater risk of isolation in SS programs and less development of independence in PB). Thus, rather than showing that either type of configuration is more beneficial than the other, these findings indicate that each model has relative strengths and weaknesses in different areas. Consequently, it is incumbent upon housing service systems to continue developing screening and placement methods that identify not only levels of housing need [26], but also what types of housing would be most beneficial for each individual client [27].

Many of these benefits of PB and SS align with findings from other research showing that from client perspectives, PB is highly beneficial in fostering community and delivering intensive services [5, 8, 9], whereas SS offers more opportunities for autonomy and independence [8, 10, 11]. In discussing how PB is able to facilitate community and community integration with greater ease than SS-PSH, it is important to note the type of community integration that it fosters. Wong and Solomon [28] posit three types of community integration for participants in housing programs: physical, psychological, and social. In PB-PSH, it could be argued that program clients are psychologically integrated in that they feel a sense of belonging and community with other PB housing residents. However, this community is largely confined to program staff and clients. In addition, PB clients experience neither physical nor social community integration with those in their neighborhood. Clients in SS housing, on the other hand, are physically integrated into the community and living in buildings with others who are not program participants, but they are not socially or psychologically integrated into the broader community. Although some view this lack of integration as an opportunity for peace and solitude, the social isolation remains a problematic feature of SS housing [11]. Distinguishing between the different types of community integration that each PSH model offers and their tradeoffs, these findings underscore the importance of matching clients to programs that fit their clinical and support needs, with PB settings often more appropriate for clients with higher levels of acuity, and SS often more appropriate for individuals who are more independent or further along in their recoveries.

Research shows that PB can be more effective than SS in improving disability severity, psychological community integration, and recovery outcomes [9], highlighting the importance of ensuring that clients who struggle most in these domains receive the type of intensive, integrated, and convenient support that PB offers. Similarly, given that clients often prefer SS placements [10] and that they are more conducive to fostering independence, finding ways to get clients who do not need PB services into SS can be beneficial. Challenges in finding placements in SS settings and limitations on the development of community in PB settings during the COVID-19 pandemic also should be noted and considered when weighing the pros and cons of different PSH models during public health emergencies.

To our knowledge, several findings on the benefits and drawbacks of PB and SS from the perspectives of service providers are novel. Difficulties finding placements, managing relationships with landlords and property managers, and delivering services in SS programs are all significant challenges that providers reported they often do not face in PB programs. It may be more difficult for providers to get their clients housed and provide them with the services they need to retain their housing and advance in their recovery if they are serving SS clients spread throughout the community. However, for some providers, the variety involved in delivering SS services and the experience of seeing their clients develop greater independence can make SS preferable to PB. Given that burnout and poor staff retention are major challenges for the homelessness service sector [13–18], it is important for policymaking and programming to consider the challenges and benefits that SS configurations create for many service providers. In addition to matching clients to programs based on their needs and preferences, matching service providers to PB or SS based on their choices and what they value most in their work can help improve job satisfaction, potentially leading to reduced burnout and turnover.

Study findings also highlight that programs may not be meeting the standard of providing comprehensive, person-centered treatment services that are essential components of PSH. In focus groups, providers described client housing preferences as burdens rather than priorities, going weeks at a time without contacting clients, and having preferences that were more based on logistical convenience than client well-being. As other research has shown, providers both in LAC [5] and elsewhere [6] frequently have such large caseloads (30–50 clients) that they lack the time and bandwidth needed to provide person-centered care; instead, they need to focus on getting clients housed and doing what is needed to help them avoid eviction. Reducing PSH provider caseloads by hiring more staff and taking steps to increase workforce retention (e.g., increasing salary and providing more training and support) [13–18] could reduce caseloads and improve provider capacity to provide more comprehensive services. The integration of other types of support for both clients and providers (e.g., peers and volunteers) [29, 30] into PSH services could also help programs better meet the array of client needs above and beyond basic housing. For policymakers, study findings also bring many of the challenges of creating PSH placements that work well for both clients and providers into clear relief. Real estate market trends have made it increasingly difficult to create PB buildings [31], and on the rare occasions when it is economically feasible to construct them, local community opposition frequently delays or prevents their development [32]. As a result, communities with large PEH populations need to utilize SS, rather than PB, to meet the immediate needs of PEH. Even though clients often prefer SS configurations [10], this study highlighted how difficult it can be for providers in the current homeless service system to provide patient-centered care for SS clients. Given that the barriers to the construction of PB buildings are unlikely to go away any time soon, it is incumbent upon homeless service systems to ease the burdens facing SS providers and make it possible for them to provide more intensive support comparable to what they are able to give their PB clients. Lowering caseloads, organizing caseloads to minimize provider travel, and facilitating client use of peers, volunteers, and other supports that are near their housing placements can potentially help providers overcome the barriers they currently face when serving SS clients.

The COVID-19 pandemic impacted these differences between PB and SS settings in different ways, making them more pronounced in some respects and less so in others. Providers reported that the pandemic exacerbated the difficulty of finding SS placements, compounding what was already a significant challenge for SS clients that they did not face with PB clients. This underscores the potential benefits of having PB units available to provide housing for clients since SS placements may be even more difficult to access than usual during times of crisis. On the other hand, the pandemic attenuated differences between PB and SS models in domains related to client-provider contact (increasing phone contact instead of face-to-face contact for PB clients and increased frequency of check-ins for SS clients) and community integration (decreasing opportunities for community-building and social activity in PB settings). These findings highlight how many qualities of PB programming that make it more supportive for clients (e.g., frequent in-person client-provider contact, community-building, social activity) can be hindered by public health emergencies, and the need for PSH programs to develop strategies to maintain their supportive elements in times when opportunities for face-to-face interactions are curtailed.

The limitations of this study are as follows. The study took place in one large urban geographic area (LAC), and findings may not be generalizable to other areas. This is especially the case because the homeless service system in LAC is the largest in the United States [1] and covers a large geographic area that spans over 4,000 square miles and includes 88 cities [33]. The challenges of finding SS placements and serving SS clients may not be as salient in areas that are geographically smaller, have less traffic, and have less individuals in PSH. Furthermore, all findings reported are perceptions of a sample of 37 frontline homeless service workers from five agencies, and it is unknown how common or widespread they may be. Future research that elicits perspectives of a larger number of service providers from several communities can help determine how much the findings reported here hold true outside of LAC.

Nonetheless, this research brings to light key differences between PB and SS-PSH from the perspectives of frontline service providers. As the post-COVID landscape of homelessness and PSH services evolves, it will become clearer if observed differences between PB and SS in areas related to finding placements, frequency of contact, and community integration observed are durable or if they shift over time. Future research on how to balance the benefits and drawbacks of SS and PB for both clients and providers can help homeless service systems develop clinically appropriate, effective, and sustainable responses to chronic homelessness moving forward.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request. The data are not publicly available due to privacy and ethical restrictions.