Table of Contents

INTRODUCTION

BACKGROUND

• National Standards
• Adequate and Appropriate Services
• Homeless People’s Legal Right to Survive
• Examples of San Francisco Encampment Response

FAILURES OF HSOC MODEL

• Poor Structure Leads to Low Placement Rates
• Not Enough Beds for Encampment Residents
• Inappropriate Service Offers
• Leading with Cleaning
• Left Without Shelter
• Zero Compliance with Bag & Tag Policy
• Sidewalk Shuffle is Trauma-Inducing and Ineffective

RECOMMENDATIONS

• Recommendations to Fix a Failing Street Response
• Recommendations for Informal Settlement Relocations

CONCLUSION

Appendix: Dissecting the Data

Introduction

“‘Wanda’ is a young woman in her 20’s who is currently pregnant and residing on the streets in the South of Market area.  She is desperately trying to get a place to stay that is safe for her and her unborn child in the middle of the delta surge.  She has been forcibly displaced from her camping spot on more than two occasions by the city and remains on the streets today, as no appropriate placement was offered to her.  These operations have made her less safe, separating her from her support system, and causing undue stress during her pregnancy.  Wanda is just one example of literally thousands of people who have been hurt by HSOC operations.  Wanda matters.  Her future child’s life matters.” 

– Kelley Cutler, Human Rights Organizer with the Coalition on Homelessness

On Monday, August 2, 2021 the San Francisco Department of Emergency Management’s (DEM) executive director, Mary Ellen Carroll, gave a quarterly presentation to the Local Homeless Coordinating Board regarding the impact of the city’s Healthy Streets Operating Center (HSOC) on the unhoused population within the city. The presentation highlights the role of the public health-centric team in using a service-led model to connect people in need with available resources in an efficient and compassionate manner. Through a regimented assessment and resolution process, the HSOC staff evaluates what resources are needed to “resolve” an encampment, to bring adequate shelter resources to meet the need present, and to lead with offers of shelter to every encampment resident, before conducting street cleaning in a humane and respectful manner.  Carroll’s stated approach aims to prioritize upholding public health, where the primary objective is connecting unhoused people with long-term solutions. 

To achieve  this objective, the City would have to adopt equitable distribution of scarce shelter resources based on need, rather than complaint or nuisance. The City would also have to recognize that every individual has a right to self-determination: that different people require different types of resources, and that connecting the right people with the right resources requires time, care and intensive outreach. According to Director Carroll, the service-led model is the key to success in addressing the city’s homelessness population. 

Carroll insists on the resilience of the model even when the  statistics revealed during the presentation that the recent service connection rate was merely 30%. Carroll interjects that this was the responsibility of unhoused San Franciscans, implying that HSOC provided sufficient resources for every client they served and that the low rate of connection is due to unhoused individuals’ negligence. In addition to these “service resistance” individuals, Carroll asserts that advocates who document actions taken by the city, inhibits HSOC’s ability to expand on their efforts. 

On Tuesday, August 3, the morning after Director Carroll’s presentation, an  advocate witnessed a Department of Public Works (DPW) employee, who was working  with the HSOC resolution team, unceremoniously drag a tent about 5 yards along the sidewalk. Inside the tent was an unsuspecting woman, who after being dragged across the ground, continued to be harassed by the entirety of HSOC staff until the advocate helped her move her tent to the next available sidewalk. The next day, that sidewalk too was “resolved,” and she was yet again displaced. 

This incident, as opposed to the idyllic resolution process described in Director Carroll’s presentation, is  a better representation of what advocates from the Coalition on Homelessness have been witnessing over the course of the pandemic. It reflects the perpetual displacement, lack of meaningful efforts to offer adequate and appropriate services, and the unjust treatment that unhoused San Franciscans often describe when asked about their interactions with HSOC. It also reflects what the city’s own documentation demonstrates – that HSOC almost never has an adequate number of beds to offer those who are being forcibly displaced as required by law, that they are failing to make appropriate and lasting placements, and that they are illegally discarding the property of those in encampments. 

This report is spurred by the recently acquired access to publicly released data that reinforces our experience monitoring HSOC operations.  This is an effort by the Coalition on Homelessness to expose the reality of HSOC’s operations, how greatly they differ from public presentations by HSOC, and to outline how HSOC is ineffective and harms public health.

Background

“We don’t do sweeps.”

– Mayor London Breed, speaking to Nathan Heller of The New Yorker in May of 2020

“Man sleeping on bench on Hayes st near Gough. Can someone come ASAP. I’m in the area having lunch”

– Mayor London Breed, in a August 2019 text message to, among others, SFPD Chief Scott, released by public records request in May of 2020

During the past four decades, San Francisco relied on using enforcement in an attempt to curb the presence of unhoused people, including the issuance of hundreds of thousands of citations and making countless arrests.  If individuals could not pay their citations or if they failed to appear in court, a warrant would be issued.  Eventually, after numerous tickets, the individual would typically be taken into jail, with time served “paying off the debt”.  In October, 2015, the Superior Courts halted the practice of issuing warrants, and the police department decreased by almost half the number of citations given out annually.  According to San Francisco Police Department (SFPD) Lt. Lazar, the founder of HSOC, “the police don’t feel it is worth it anymore, so many have stopped issuing homeless citations”.  On October 12, 2016, the DOJ/COPS released its initial assessment of SFPD civil rights practices and provided 272 recommendations for improvement within the department. The areas of review included, Use of Force, Bias, Community Policing, Accountability, and Hiring and Personnel Practices. Many of the recommendations criticized the SFPD response to homelesness.  A homeless advisory body was formed to address these recommendations under then Lt. Lazar, but most of these meetings were cancelled or rescheduled, and rarely was there any review of the DOJ/COPS recommendations within those meetings.  Contrary to the DOJ/COPS recommendations that essentially called for SFPD to back off of homelessness, the city decided to take a markedly different route and doubled down on unhoused people – in particular those in encampments.  On January 16, 2018,  HSOC was started within the Mission Police station by Lt. Lazar, with the very intentional goal of eliminating large tent encampments in that same district.  HSOC consolidated the city’s encampment response by establishing a coordinated multi department effort to respond to 911 calls regarding encampments with 6 or more tents.  These calls were rerouted to 311, HSOC was dispatched with police in the lead, with departments such as Public Works, Public Health and Homelessness and Supportive Housing close behind.  

The concept of HSOC as a police-led, complaint-driven coordination of city departments and resources designed to lower tent counts and break up large encampments appeared in all the early materials including the Controller’s report which states that it aims to “[e]nsure [that there are] no tents within the geographic boundaries of the Mission District.” Specifically, HSOC’s metric of using tents as a measurement to evaluate the “success” invites detrimental policies that dehumanizes unhoused individuals. Indeed, while the San Francisco’s Office of the Controller emphasizes HSOC’s mission to “[l]ead with compassion and respect” and to coordinate response “to unsheltered persons experiencing homelessness,” the main action items for various working groups detailed in the report prioritize reducing the number of tents within an area, rather than meaningfully decreasing homelessness. 

In early 2019, HSOC moved out of the police department and into the Department of Emergency Management (DEM), the department that oversees the 911 system, emergency dispatch, and (more recently) the city’s COVID response.  Along with that move, they replaced former HSOC lead in SFPD with Jeff Kositsky, who previously led the city’s homeless department (who himself has since been replaced following his recent resignation). Over the course of the COVID-19 pandemic, Kositsky’s HSOC has made use of the city’s rapidly changing shelter system to greatly reduce the number of tents on San Francisco’s streets.

National Standards

Despite HSOC’s purported claims on prioritizing public health, the HSOC process to control the unhoused population is through the police department.  Specifically, once HSOC receives a 311 phone complaint regarding homelessness, police officers are included in the response and are sent to the area to remove the individuals. This approach fails to address the underlying problem of the housing crisis, and is an ineffective use of government resources. 

In fact, communities and governments across the country are re-examining a police response to homelessness. The federal government now penalizes municipalities in their McKinney Act funding applications for failing to address criminalization, including the elimination of enforcing misdemeanors for status crimes such as lodging.  The U.S. Interagency Council on Homelessness (USIACH) has also put out guidelines for addressing encampments which include recommendations to work collaboratively over a couple weeks with communities to relocate individuals into permanent housing.  If permanent housing is unavailable, the USIACH suggests moving folks into temporary shelter until permanent housing is available.  The 2009 Congressional HEARTH Act delegated the United States Interagency Council on Homelessness (USICH) with the task of “develop[ing] alternatives to laws and policies that prohibit sleeping, eating, sitting, resting, or lying in public spaces when there are no suitable alternatives, result in the destruction of property belonging to people experiencing homelessness without due process, or are selectively enforced against people experiencing homelessness.”  As we outline later in this report, the national standards established by these policies shows how HSOC’s process to address homelessness is subpar. 

The transition away from policing in homelessness  exemplifies a growing recognition in the failure of criminalization to combat homelessness and of the need for a different strategy to address the growing crisis. Despite this, many who are forced to stay on the streets of San Francisco are consistently denied fundamental human rights such as the right to accessible water, sanitation and hygiene (WASH) facilities. Meanwhile,  extensive resources are spent on enforcement-led responses to complaints about the presence of homeless people that at best push them from block to block. This approach creates high economic costs for the City to funnel resources into this growing problem, with little progress to show for it. 

During the COVID-19 pandemic, there were guidelines established by the Centers for Disease Control and Prevention (CDC). The COVID-19 virus and now the Delta variant puts even more risk on municipalities’ practice of displacing and criminalizing unhoused community members.  Congregate settings are currently not considered by many medical professionals as a viable option due to individual medical risks and public health risks of COVID-19 Delta variant.  In addition, the CDC recognized that displacement of encampments and removal of tents created several health hazards, including an inability to shelter in place when no adequate alternatives are available and a difficulty in contract tracing if outbreaks occur.  According to the CDC’s Interim Guidance on Unsheltered Homelessness and Coronavirus Disease 2019 (COVID-19) for Homeless Service Providers and Local Officials: 

If individual housing options are not available, allow people who are living unsheltered or in encampments to remain where they are. Clearing encampments can cause people to disperse throughout the community and break connections with service providers. This increases the potential for infectious disease spread.

– the CDC’s Interim Guidance on Unsheltered Homelessness and Coronavirus Disease 2019 (COVID-19) for Homeless Service Providers and Local Officials

Adequate and Appropriate Services

Unhoused community members are diverse in terms of need.  Most are simply poor, cannot afford rent, and do not have accumulated wealth to fall back on when times get tough.  Approximately a third have mental health and/or substance use issues, more than a quarter are working, many suffer from chronic health conditions, many experienced adverse childhood trauma, while others suffer from Post Traumatic Stress Disorder.  In order to ensure that individuals are connected with appropriate services, trust must be developed, and a variety of services secured.  Critical to a successful relocation process is the embracing of self determination on the part of the unhoused neighbor, and detailed information on what options are available.  There may be barriers, such as PTSD which for some people prevents them from being able to sleep in congregate settings, or domestic violence that prevents them from staying at a particular site where a former abuser is present.  Many options can be inappropriate for a myriad of reasons, and follow up work must be done to secure more appropriate options.  This is a process that can take time.  If there are not appropriate options, the individual may have to stay put, and in those cases garbage, clear passage and hygiene needs can be addressed until appropriate placements are available.  If the person is displaced, it will often be impossible to connect with that individual once appropriate placements are available. HSOC operations as we will later outline, include the offering of congregate shelter for a portion of the encampment residents.  Rarely do they include the offering of a wide variety of services appropriate to meet an individual’s need. 

Homeless People’s Legal Right to Survive

“We’re getting cops called on us all the time. We’re in other people’s space. They don’t understand that we don’t want to be in their way, we’re just looking for a space to exist. We’re just looking for a place to be. We try and find these spots that are out of the way, try and find areas that people don’t frequent.”

– Tony, unhoused San Franciscan speaking after an HSOC operation on June 23rd, 2021

In addition to the ineffectiveness of temporarily displacing unhoused individuals in public areas, federal law clarifies that the City cannot criminalize homeless people for loitering and sleeping in public property. In Martin vs. Boise, six currently and formerly homeless Boise, Idaho residents, alleged that laws prohibiting them from sleeping outdoors within city limits amounted to cruel and unusual punishment and violated their rights under the Eighth Amendment. This was supported by the Obama administration’s Department of Justice, which submitted a statement of interest, arguing that making it a crime for people who are homeless to sleep in public places, when there is insufficient shelter space in a city, unconstitutionally punishes them for being homeless.  This case received a ruling in favor of the plaintiffs in the 9th Circuit Court of Appeals after the decision was appealed by the City of Boise. 

In Martin vs Boise,  Judge Marsha S. Berzon states: “Turning to the merits, the panel held that the Cruel and Unusual Punishments Clause of the Eighth Amendment precluded the enforcement of a statute prohibiting sleeping outside against homeless individuals with no access to alternative shelter.  The panel held that, as long as there is no option of sleeping indoors, the government cannot criminalize indigent, homeless people for sleeping outdoors, on public property, on the false premise they had a choice in the matter.”

Examples of San Francisco Encampment Response

The process as advocated by the court, is not only more humane, but it is also more effective. In 2012, California Highway Patrol called for the displacement of  a large encampment on King Street.  Spearheaded by Bevan Dufty, the Mayor’s former homeless director, the initiative led to a successful permanent relocation of this settlement.  Duffy had reached out for counsel from community members of the camp, homeless advocates at the Coalition on Homelessness and secured a church where the residents could relocate en masse.  They rented a storage container where belongings could be stored intact and, most importantly, created an exit plan for the church.  After a stay in the church, residents were relocated to housing, with careful considerations for keeping the very human support systems that formed and developed after living together in adversity.  One hundred percent of camp residents were successfully connected with permanent housing.

The King Street resolution is a far cry from HSOC, where currently only 30% of individuals are connected with shelter, and none to permanent housing.  HSOC’s primary function has been to extend the power of the San Francisco Police Department (SFPD) and DPW, coordinating a massive amount of city resources towards the foundational goal of removing tents from the streets of San Francisco.  The focus on the reduction of tents creates two problems: First, it creates problematic policies that exacerbate the housing crisis; second, it leads to questions on how successful the operations are and the validity of their results report. This is an inherently flawed approach. For unhoused community members, this only continues the cycle of being shuffled around by the city and having their belongings trashed or destroyed. These operations make it harder for people to find places that are well lit, or places where they can stay with others whom they know and, by turn, ensure their personal safety. For people with no other options, tents offer a modicum of shelter and privacy, and take the edge off the indignity of living in public spaces. 

While HSOC sometimes offers services before clearing encampments, it often has very little to offer, and at other times relies on the police or public works departments to simply clear an area. At the August LHCB meeting, Director Carroll stated that “Police address enforcement issues” during operations.  Once an area is cleared, HSOC employs strategies like bathroom removal and constructing barriers to prevent people from returning, further limiting safe sleeping areas for individuals who have no choice but to stay on the streets.  They have a “re-encampment prevention” strategy that focuses on ensuring people don’t set up tents in that area again.  This strategy is not about helping individuals who present as unhoused in that area, but focused on preventing people from sleeping in that area via police enforcement.  At the very same meeting, Director Carroll stated that after an operation, “district police stations are notified and asked to patrol the area.”  This action leads to increased criminalization and does not lead to an exit out of homelessness.  

Failures of HSOC Model

“They didn’t offer me any services.  I was busting my ass moving shit all day…I wasn’t offered anything. I never got a time to stop from moving my stuff to say hey, I need a room, I need a shower, I need food.I don’t know what else to do when 3 days in a row we are told to pack up and move.  This is discrimination, harrassment.  And with these people standing here waiting for you to move, its harassment. They don’t want the homeless here.  They want us dead. It’s how the mayor feels about us.”

– Marquis A, recorded speaking to DPW workers during an HSOC operation on 7/15/2021

HSOC pursues its goal of eliminating large encampments through three main types of operation: resolutions, re-encampment prevention, and cleaning. Encampment resolutions are the largest operation, in which the full HSOC team “resolves” an encampment by offering limited services, demanding that residents leave the area and having DPW conduct a full cleaning. Once an area is “resolved,” regardless of whether the people who lived there have moved into a shelter or just around the corner, it is subject to tactics aimed at preventing the re-emergence of the original encampment.

 As the most significant and publicly observable operation, if not the most frequently used, encampment resolutions and their outcomes were the primary focus of Director Carroll’s presentation to the Local Homeless Coordinating Board. In her presentation, Carroll went over in detail the step-by-step timeline of the HSOC team’s activities during these resolutions. However, as several members of the board mentioned after the presentation, this timeline varies considerably from what unhoused San Franciscans and advocates have witnessed at resolutions over the past several months. Below is a comparison of timelines, based on the 28 HSOC operations that Coalition on Homelessness staff and volunteers have observed from during 2021 thus far, between the stated HSOC resolution process and the operations that have been observed in practice. 

HSOC publicly stated operating procedure	Observed operating procedure
Steps 1 – 3: At 7 a.m., outreach workers gather names and offer “sheltering alternatives”	At 7 a.m., or sometimes as late as 8, outreach workers notify encampment residents of the coming resolution, and gather. At this time, there is no specific offer of any shelter resources, but often a general inquiry as to whether individuals would accept a congregate shelter placement. Residents are also told to pack their belongings and leave the area.
Step 4-5: Outreach workers identify clients needing additional support and inform a clinician, office staff look up clients in system	Immediately following outreach workers’ initial engagement, DPW workers begin their cleaning operation on the streets and sidewalks being resolved. At this time, they are mostly picking up loose debris throughout the encampment, but they also often pressure residents to pack their belongings more quickly and search for “abandoned” belongings to throw away. None of these belongings are “bag and tagged” in accordance with DPW policy. As they are typically not able to provide specific shelter offers until after 9:30 a.m., outreach workers are mostly taskless during this time, often taking the opportunity to make a trip to buy coffee.
Step 6: Client Transportation begins	After about 9:30 a.m., outreach workers gain access to the day’s shelter bed allocations, and begin to make offers to residents. By this time, over two hours after the initial engagement, it’s common for most residents to have already left the area, or else be occupied by packing and moving their belongings to their next destination. For those who are still in the area, there is no guarantee that there will be any appropriate resources available.
Step 7: Client given time they need to leave	Often, when there are not enough or appropriate resources, those who waited hours for a place to stay are told that there may be sufficient resources for them the following day, and that outreach workers will return to the area. This despite the fact that the area is being resolved, and residents have been told they are not allowed to stay there any longer. The workers pressuring people to leave at this time are often DPW or SFPD, who are not privy to which residents are working with outreach workers to access shelter. There is no clear process for assessing disability needs or providing reasonable accommodations, and rarely a clear time by which people must move.
Step 8: Public works begins cleaning and “will bag and tag property as appropriate”	By the time shelter offers are being made, most encampment residents have already been forced to move their belongings out of the area. At this time, DPW begins a more thorough cleaning of the area, including more disposing of unaccompanied property, none of which is bag and tagged.
Step 9: Police ensures everyone’s safety and addresses enforcement issues	People who have not packed and moved quickly enough for DPW workers to conduct their street cleaning are repeatedly harassed by DPW workers. Sometimes, this escalates to HSOC police threatening residents with citations if they don’t move quickly enough. Police and other HSOC staff also often threaten residents with citation and/or arrest if they return to the encampment after the resolution.
Step 10: Operation ends between 11 a.m. and noon	Advocates have been told by several members of the HSOC team that DPW workers’ morning shift ends at 11 a.m., causing workers to rush for the resolution to be finished by then. The entire resolution, a process which at one time spanned several weeks in order to effectively match people with services, usually takes no longer than 4 hours under the current model. Most of that time consists of DPW street cleaning, and very little time dedicated solely to intensive outreach without the stressor of DPW and police activity.

Poor Structure Leads to Low Placement Rates

Among the many discrepancies between Director Carroll’s account of the HSOC resolution timeline and the one observed by advocates, lies in how HSOC approaches connecting people in need with the city’s available shelter resources. Based on what advocates have witnessed, HSOC uses a poorly structured resolution process that prioritizes street cleaning over service connection, providing little time for outreach workers to work with residents and creating an atmosphere of stress and panic that makes it difficult for residents to advocate for their needs. What both advocates and the director agree on is that HSOC’s rate of service connection is less than satisfactory. In the three periods of 2021 presented on by Director Carroll, HSOC never surpassed a “service acceptance” rate higher than 35% (with the higher rates of the previous year largely attributable to the availability of SIP hotel beds).

Director Carroll implies while reporting high rates of “service declines” that these low success rates can be attributed to “service resistance.”  When asked about the low rates of successful relocations of encampment residents she stated “there are many reasons people decline services —addiction and mental health are high on that list,” indicating services are not appropriate for these populations. “Service resistance” is the commonly used term that implies people on the streets have been offered alternatives, but have remained where they are by choice. According to Joe Wilson, director of Hospitality House, a San Francisco organization that operates along with other services, a shelter and two drop-in centers: “When individuals do not ‘accept’ services, likely the offer itself is not meeting their individual needs.  The concept that people would rather be homeless or are homeless by choice fails to acknowledge both the severe misery homelessness presents and the systemic inequities such as racism, homophobia and the massive income rent disparities that lead to homelessness. It is our job to adapt the services to meeting the individual’s needs — not the other way around.”   According to a leading scholar on homelessness, Deborah K. Padgett of New York University’s Silver School of Social Work, “there is no evidence to support this notion that homeless persons are ‘service resistant. People on the street often reject the option of crowded, unsafe shelters—not housing in general.”  In her study, she identified barriers to bureaucratic barriers to housing, and a lack of safe, pet-friendly shelter.   

On the other hand, those of us who have been observing HSOC’s resolutions would argue that the greatest factors contributing to the low success rate are the structure of the resolutions, which as previously stated prioritize cleaning over meaningful service connections, and the lack of availability of adequate and appropriate resources. During the over two dozen sweeps witnessed by Coalition on Homelessness staff and advocates this year, rarely, if ever, did HSOC outreach workers have access to enough shelter beds for every resident at an encampment. On the contrary, advocates have been told by outreach workers and other HSOC staff repeatedly that they typically only had access to 10 or fewer beds each day, despite the presence of many more people than that at most resolutions. 

Placement Rates

End of the Period	Operating Days in the Period	Tents from HSOC Count*	Reduction in Tents	Percent Reduction in Tents	Client Placements	Placements /Day	Acceptance Rate	Placements /Tents Reduced**
4/22/2020		1108
6/9/2020		1400
7/28/2020	34	1003	-397	-28.36%	677	19.91	83%	1.71
10/7/2020	51	703	-300	-29.91%	441	8.65	60%	1.47
12/14/2020	43	565	-138	-19.63%	265	6.16	31%	1.92
2/15/2021	40	501	-64	-11.33%	197	4.93	35%	3.08
4/26/2021	53	383	-118	-23.55%	277	5.23	30%	2.35
6/21/2021	38	387	4	1.04%	179	4.71	29%	0.00
Summary	259		-1013	-72.36%	2036	7.86	45%	2.23

“Accepting help is alright when the help comes that’s really helping and not just trying to shove me out of the way. Any type of services I’ve had to debate whether or not I’d take it or not because the people who are coming to see us, they’re asking us to leave, they’re coming with the police, they’re coming with garbage trucks, telling us we got to go, tents aren’t going to be allowed up, and then they said but we’ll send you to tent city.  Ok, I can’t sleep on the streets in a tent but you’re going to put me somewhere where I can be in the street, in a tent.”

– Paul R, unhoused San Franciscan interviewed by Coalition on Homlessness volunteer Christin Evans

In the past few months, release and analysis of several public records requests (Appendix) have revealed this routine under-availability of shelter resources to be a pattern. Across a 37-day period from January to February of this year, HSOC had enough shelter beds to offer a bed to everyone at that day’s operations only twice. On average, they had access to only 52% of the beds they would need to do so.

This consistent lack of shelter beds call into question how widespread the issue of “service resistance” really is. In the statistics presented by Director Carroll, the outcome of each individual at an HSOC operation is categorized as one of three options: Placed, Already Sheltered, or Declining Services. In other words, according to HSOC’s statistics, any individuals who are not transported to shelter after an operation, or were already sheltered, are considered to have declined (or resisted) service. However, it remains in question how this can be true if there were not enough beds available for everyone who is said to have declined them. In other words, it’s not clear how HSOC can consider people for whom there were not enough beds to offer to have refused shelter. In fact, by cross-referencing the service connection data provided by HSOC (from the same 37-day period previously mentioned) with the shelter beds that were available to HSOC during those same days, it becomes clear that people accept the shelter beds available at quite high rates. In those 37 days, despite a 29% overall service acceptance rate, 75% of the beds HSOC had available were filled. 

Shelter Beds Available vs Clients Engaged at HSOC Resolutions, Jan-Feb 2021

Not Enough Beds for Encampment Residents

“DPW took my stuff. All my clothes, my phone, my money, and I can’t replace it so I don’t know where to go now. I don’t know where to go from here.”

“Did they offer you any services?”

“No services, no nothing…I’m mad as fuck.”

Interview between Carlos Wadkins (Human Rights Organizer with the Coalition on Homelessness) and L (Unhoused, disabled San Franciscan) at an HSOC operation on April 13, 2021

Together, these statistics imply that the main problem leading to low service connection rates by HSOC lies in service availability, not service resistance. Addressing this, Director Carroll stated to the LHCB that tent removal operations did not occur, or would be paused, if there were not enough available beds in shelters. However, as previously established, encampment resolutions are close to completion, with most residents having already been displaced, by the time that outreach workers know the exact number and types of beds that they will receive. In the experience of Coalition on Homelessness advocates, when there are not enough resources or the appropriate type of resource, outreach workers are often left to tell people in need that they must wait for the next day in order to be offered a bed. Additionally, there is no section of the outcome data presented by Director Carroll, nor the internal data released through public records request, to account for instances in which operations were paused due to lack of resources. In the event that this had occurred, individuals for whom there were not enough beds would presumably still be categorized under “declining services” for lack of other classifications, lending false evidence to the theory that service resistance is the key factor. 

Inappropriate Service Offers

“The HOT team does not have a ‘housing solution’ for us due to the schizophrenia. Charlie has been homeless and living on the streets since 1976. None of us drinks or does drugs.  I have bipolar, Charlie and Gael have schizophrenia and Tommy has PTSD. We want to stay together, we’ve been on the street together for years and look out for one another.”

– Marlon, speaking about an HSOC operation on 7/29/2021

HSOC Outcomes – Detailed data

Week of Year	First Sweep	Tents Removed	Vehicles Removed	Total “Clients”	“Clients” Served	“Declined Services”	Percent Served	Available Shelter Beds	Percent of available Shelter Beds Filled
2	1/5/2021	53	1	60	25	31	42%	24	104%
3	1/11/2021	106	4	101	32	59	32%	49	65%
4	1/19/2021	84	0	107	34	61	32%	33	103%
5	1/25/2021	80	8	93	22	52	24%	31	71%
6	2/1/2021	82	0	126	29	84	23%	29	100%
7	2/8/2021	86	3	104	33	53	32%	41	80%
8	2/16/2021	39	1	46	12	30	26%	33	36%
9	2/22/2021	63	7	75	23	44	31%	37	62%
	Total:	593	24	712	210	414	Average: 29%	Total: 277	Average: 76%

HSOC Outcomes – Summary

While these numbers can speak to the numerical lack of shelter resources, they cannot speak to how appropriate those resources are for the people HSOC engages with. As we move through sites, one of the questions our outreach team asks is if the individuals were offered any services and if so, which service.  The most common response was that no they were not offered services.  At other times, the individual was offered congregate shelter.  While this number is highly suspicious, obviously people don’t take what does not serve them.  Congregate shelter is advised against by the CDC during the pandemic, and is even more dangerous today than it was at the beginning of the pandemic given high infection and fatality rates of the delta variant. One former HSOC staff person told us, “They offer shelter, but everyone at HSOC knew that staying at MSC-South would be more harmful physically and mentally than staying in the tent, and they did it anyway. There are no ethical guidelines there.”  However, even in “normal” times, congregate shelter is often an inappropropriate placement for many individuals for a variety of reasons including disability accommodations.  Individuals with post-traumatic stress disorder, for example, often are unable to sleep near strangers. Others with severe mental illnesses may shout out during the night, keeping others awake. During the time period that hotel rooms were made available, the acceptance rates of SIP hotels was at about 95%. Hotels offered bathrooms, privacy and dignity. This acceptance rate further demonstrates that when the system is responsive to need, it is able to easily move people off the streets.

While these numbers can speak to the numerical lack of shelter resources, they cannot speak to how appropriate those resources are for the people HSOC engages with. As we move through sites, one of the questions our outreach team asks is if the individuals were offered any services and if so, which service.  The most common response was that no they were not offered services.  At other times, the individual was offered congregate shelter.  While this number is highly suspicious, obviously people don’t take what does not serve them.  Congregate shelter is advised against by the CDC during the pandemic, and is even more dangerous today than it was at the beginning of the pandemic given high infection and fatality rates of the delta variant. One former HSOC staff person told us, “They offer shelter, but everyone at HSOC knew that staying at MSC-South would be more harmful physically and mentally than staying in the tent, and they did it anyway. There are no ethical guidelines there.”  However, even in “normal” times, congregate shelter is often an inappropropriate placement for many individuals for a variety of reasons including disability accommodations.  Individuals with post-traumatic stress disorder, for example, often are unable to sleep near strangers. Others with severe mental illnesses may shout out during the night, keeping others awake. During the time period that hotel rooms were made available, the acceptance rates of SIP hotels was at about 95%. Hotels offered bathrooms, privacy and dignity. This acceptance rate further demonstrates that when the system is responsive to need, it is able to easily move people off the streets.

Leading with Cleaning

“I’m close to 40 years of age and i can’t keep up with the demands of the sweeps personnel. i shouldn’t be rushed to move my belonging anywhere without checking ahead of time for services to be dished out. Like why are they still displacing people when COVID is still running rampant in the community. We have a shelter-in-place order in effect so that the transmission of the virus can be limited and through contact tracing, we can overcome this dilemma of fear and fragile feelings.”

– Dawn P, unhoused San Franciscan interviewed by Coalition on Homelessness Intern Marquis A

Another key discrepancy between the two timelines lies in the actions of DPW workers during resolutions. According to Director Carroll’s presentation, DPW workers do not begin their cleaning operation until after all encampment residents have been offered shelter, when transportation has already begun and people have had time to move their belongings. As previously stated, our experience witnessing resolutions has found that DPW workers in fact begin cleaning immediately after the initial engagement by outreach workers, well before residents have begun to receive specific offers of shelter. Furthermore, it is often DPW, rather than outreach workers, that is pressuring encampment residents to leave the area throughout the resolution. This often leads to high tensions between DPW workers and encampment residents, in which case police are present to threaten citation and property confiscation.

Left Without Shelter

Following the presentation regarding HSOC from the Department of Public Works and Department of Homelessness, Supervisor Matt Haney’s doubts of HSOC points to a larger problem in the HSOC policies created to optimize their data. In his speech, Supervisor Haney suggests that HSOC’s mission would encourage officers to discard tents instead of providing services for unhoused individuals, which would leave these unhoused individuals in a more desperate situation. While HSOC frequently boasts its success in reducing the number of tents in San Francisco, this success is not reflected in anywhere near an equal number of exits from homelessness. Tents are not people, and removing tents while doing nothing to change the housing status of the individual does nothing but leave people on the street with even less protection from the harsh conditions under which they live.

Beyond this, HSOC has also significantly changed the way that the City’s resources are allocated. In order to facilitate HSOC’s many operations, many available beds or housing are reserved for use at encampment sweeps. This means that the thousands of unhoused San Franciscans not subject to an operation, who are perhaps in greater need of shelter, are left without a way to access those beds. These resources are often directed to the most politically important areas, such as gentrifying neighborhoods, rather than those most in need. This is often done at the behest of “VIP” individuals and groups, such as San Francisco’s many Community Benefit Districts. 

Zero compliance with Bag & Tag Policy

“I won’t give you a bag.  I’ll call the police if you don’t give me your stuff.”

DPW worker overheard by Coalition on Homelessness Volunteer Flo Kelley on Jan.13, 2021 during an HSOC operation

Director Carroll also claimed that workers “bag and tag” property “where appropriate”, a process where personal belongings are confiscated and stored at the DPW lot, where people can go to get them back. According to DPW policy, abandoned belongings such as tents and other survival gear is to be bagged and tagged, rather than thrown away. 

“Unattended property is not abandoned if it is accompanied by signs of ownership – for example, an unattended tent that is filled with personal belongings, or items that are being stored in an orderly manner (i.e. packed up, wrapped, or covered). In addition, if there is a third party present who states s/he has been designated to watch or secure the items during the owner’s temporary absence, the items are not considered abandoned. By contrast, abandoned items are unaccompanied by objective indications of ownership, for example, an empty or broken tent sitting by itself on a sidewalk with no other belongings.” 

However, in all of the resolutions monitored by the Coalition on Homelessness this year, only one bag and tag was observed after an advocate demanded it be done. Advocates have witnessed, on the other hand, several tents and other personal belongings be thrown away, often into DPW’s trash compacting truck. This experience is reflected in DPW’s own bag and tag logs for the months of January and February, released in response to a public records request. Of the 38 Bag and Tags recorded between January 4, 2021, and February 26, 2021, not a single one was recorded at the same location as the HSOC operations of the corresponding day. In fact, many of the recorded bag and tags were not done by DPW workers at all, but by SFPD officers. In short: across 87 HSOC operations, in which their data claims 593 tents and structures were “removed,” not a single belonging was bagged and tagged. 

Sidewalk Shuffle is Trauma-Inducing and Ineffective

“This morning an officer came by and rudely woke me up asking us ‘Who said we could be here?’ ‘How long had I been here?’ Telling me it was totally unacceptable what my tent looked like and said we had to get packed up and get out and he was just really rude about it. I said, ‘Sir, you don’t have to be rude.’ and he continued to be rude. C’mon, we are all just human beings, you don’t have to talk to us like that. I’m not a piece of shit or piece of trash.’ But that is how he made me feel. He didn’t offer anything but as he left he turned around he said, ‘Ma’am I’ll be back!’”

– Dawn, during a July 19 HSOC operation

Overall, neither HSOC’s own data nor the experience of advocates depict HSOC operations as being centered on public health and permanent solutions for unhoused people. Rather, they depict a DPW-led model that prioritizes displacing people and tents as quickly and efficiently as possible, not leaving sufficient time to notice unhoused residents nor effectively connect residents with adequate and appropriate services. As a result, service connection rates are very low (less than 30% for the January to February period), while the rate of tents cleared remains very high, at 88% for the same period. While this may seem discordant with the presentation given by Mary Ellen Carroll in August, it perfectly aligns with the foundational purpose of HSOC to reduce the number of tents on the streets of San Francisco. In order to achieve higher success in the metrics that actually matter, connecting unhoused people with permanent solutions that keep them from needing to sleep on the street, there must be changes made not only to the model of this city’s street response but to the foundational principles which HSOC was created to pursue. 

Recommendations

Recommendations to Fix a Failing Street Response

“The current state of homelessness in the City and County of San Francisco is truly replusing. I believe that the Mayor and the current board of Supervisors are to blame for this mess. Homelessnes is not a “one solution –  fix all” issue. It touches on many different issues like mental health, substance abuse and treatment, self esteem and self image issues, but would also include evolving discussion around housing justice, financial responsibilities, and community engagement.”

– Marquis A, unhoused San Franciscan and intern with the Coalition on Homelessness

The following are system recommendations to ensure successful approaches to street homelessness. 

1. Halt the policy and practice of focusing on tents – instead have street teams identify, assess, and place humans in need.    
2. Eliminate barriers to services to existing homeless programs.  This includes  ensuring shelter options are barrier- free, such as ensuring pets, partners and property are allowed, and minimize rules.
3. Invest in permanent solutions to homelessness including housing, living- wage jobs, on-going treatment and medical care over bandaid solutions.
4. Ensure shelters during the surge are safe, such as fully utilizing SIP hotels rooms and pausing the shut down of these hotels. 
5. Have regular cleanings and trash pick-up at set times in areas where unhoused people tend to sleep or park.
6. Halt practices that criminalize individuals for their economic and housing status, such as police responses to homelessness. Fully implement CART —- Compassionate Alternative Response to Homelessness —- by having a deeply trained, well- paid peer- based street team respond to and solve “C” level 911 calls connected to homelessness.
7. Once CART is fully implemented, ensure the Homeless Outreach Team can fully focus on case management and connecting individuals to care, as opposed to responding to complaints.  
8. Halt the enforcement of anti- homeless laws, including the enforcement against individuals residing in areas that have already been swept.  These “re-encampment prevention” activities do not address the needs of unhoused people nor assist them off the streets, and instead rely on failed enforcement measures.  Instead, connect individuals to care.
9. Ensure full transparency and reporting of all street responses, including numbers of people who are connected with care, as is occurring with Street Crisis Response Team and Street Overdose Response Team  

The structure and intent of HSOC fails to provide a sufficient long-term solution to the housing crisis.  A focus on tent removal as a metric to evaluate the severity of the homeless population is not only an inaccurate way of measuring data, but it also creates harmful policies that exacerbate the crisis. We believe HSOC should be abolished because it’s inherently aimed at the wrong goals of removing tents as opposed to addressing homelessness.

Recommendations for Informal Settlement Relocations

Federal guidelines outlining how localities should address encampments also provide a good starting point on how to address the housing crisis, and they encourage municipalities to have clear and transparent communication with campers, timelines and relocation plans.  In the past we have successfully resolved encampments, and not just at King Street. There have been other examples of encampment relocations where elements were effective.  Islais Creek, which was an area that was being redeveloped, was done under the auspices of the Department of Homelessness and Supportive Housing via the Encampment Resolution Team led by Jason Albertson.  This two-week effort entailed deep assessments and lining up resources, resulting in 70% of encampment residents having long-term accommodations.  Earlier in the pandemic, when hotel rooms were available, HSOC collaborated with community-based organizations to place people in hotel rooms.  Of those offered hotel rooms, 95% of them were successfully relocated.  It is not only cruel, but a waste of resources to simply punish and push people who are already in crisis from place to place.  

Comparison of Informal Settlement Relocation Efforts

Recommendations to Ensure Proper Relocation	King Street Example	Current Operations
14-day written and verbal notice, with 7-day reminder	More than 14 days notice was given	No more than 3 days and rarely given
Assessments of need take place during noticing period	Every camp resident received clinical assessment	No real assessments, just asking for name and if they want congregate shelter
Offering of services appropriate to need during noticing time period	New services were garnered for camp residents	Only congregate shelter offered
Permanent housing should be offered, and when not available, temporary accommodations until permanent placements are available	Temporary shelter was offered in a church until permanent housing was available.	No permanent housing currently offered
Cleaning occurs post placement	Cleaning did not occur until after residents moved into shelter	Cleaning occurs at start of operation
Operations conducted in collaboration with culturally competent community-based organizations	Community-based organizations such as Coalition on Homelessness were intentionally pulled in to help	Organizations actively discouraged from participating
All operations done in presence of human rights observer	Coalition on Homelessness served as observers	Human rights observers discouraged from being present
All operations publicly posted	Community including housed and unhoused neighbors informed ahead of time	Operations set up to avoid public scrutiny
Police are not present	Police are not present	Police are present
City entities are appropriate and have experience with unhoused communities and trauma-informed practices. Other entities brought in only on an “as needed” basis (for example, paramedics brought in when there is a medical emergency.)	Only skilled staff present	Untrained personnel present lacking experience with trauma informed operations.
All unclaimed property is bagged and tagged. Individuals offered means to keep and transport claimed property	City provided large containers for property. Containers transported to shelter.	No unclaimed property is bagged and tagged. Owners of claimed property struggle to transport, and often give up in order to enter shelter.
If there are not adequate services for unhoused residents, they should not be moved until such resources are garnered. Meanwhile, garbage, sanitation, clear passage and other mitigations should be addressed until relocation can be achieved via placements.	Operations happened with services already obtained.	Operations consistently happen without adequate and appropriate services.
Homeless people’s legal rights, dignity, agency and self determination should be respected throughout.	Operation was done legally and individuals had agency in housing choice. Outcome: 100% of camp residents ended up in permanent housing.	Legal rights rarely observed, city actively skirting law. Outcome: post offering of hotel rooms, only 30% of encampment residents successfully relocated to temporary shelter.
These fundamental elements should form policy and procedures for the resolution team along with clear ethical guidelines the team is held accountable to.	Operations guided by community input.	No policies and procedures for operations shared with staff. No ethical guidelines.

Conclusion

Conclusion

The current HSOC operations fail to permanently address homelessness, increase economic costs and violate human rights. Only 30% of residents are being relocated, and they are only being relocated to temporary shelter. The remainder are still on the streets.  From the unhoused person’s perspective, they are cruel and trauma-inducing.  From the housed person’s perspective, they are only pushing people into more residential areas.  As a city, we can and should do better by our housed and unhoused neighbors.  Political battles during the epidemic also revealed how profoundly broken the “revolving door” approach the City takes to homelessness is, and the lack of structural investments in long-term solutions has increased systemic inequalities. As UCSF’s Dr. Margot Kushel said in response on how to combat homelessness, “There is no medicine as powerful as housing.”

Appendix: Dissecting the Data

    While much of the information presented in this report draws on the experiences and firsthand knowledge of unhoused San Franciscans and advocates, a significant amount of it comes from the city’s internal data and records. This data was made public, and brought to the attention of the Coalition on Homelessness, through an extensive series of public records requests made by Twitter user @dizz_h. These requests have resulted in a rather large and nebulous cache of information about HSOC and its operations, much of which has been made available in this public Google Drive folder. 

    In order to compile this information into a manageable set of data that gives a full picture of HSOC’s operations and their outcomes, this report narrowly analyzes the period of time between January 5, 2021 and February 26, 2021. This analysis primarily takes into account three key areas: HSOC’s operation reports, DPW’s Bag and Tag Logs, and the city’s daily shelter allocation summary. Through comparison and cross-reference of these three, a much clearer picture of HSOC is created than any one source could provide; a picture which largely verifies the experiences reported by those with firsthand experience of HSOC’s day-to-day operations. What exact information came from each source, and how they were used in conjunction with one another in order to reach better analysis, is the subject of this appendix.

HSOC Encampment Reports

HSOC Encampment Report: Week of January 25, 2021

Location	Police District	Neighborhood	Supervisor	Date	Mission	Starting Tents & Structures	Tents & Structures Removed	Tents & Structures Remaining	Starting Vehicles	Vehicles Removed	Vehicles Remaining	Total Clients	Clients Served	Clients Already Housed/ Sheltered	Clients Declining Services	Clients Remaining
Multiple Locations		CATHEDRAL HILL		1/25/2021	Re-Encampment Prevention	15	15	0	0	0	0	17	3	3	11	0
Multiple Locations		TENDERLOIN		1/25/2021	Re-Encampment Prevention	7	3	4	0	0	0	9	2	0	7	7
220 San Bruno		MISSION		1/26/2021	Re-Encampment Prevention	17	17	0	0	0	0	17	3	2	12	0
220 San Bruno		SHOWPLACE SQUARE		1/26/2021	Re-Encampment Prevention	10	6	4	0	0	0	8	6	0	2	2
Division between South Van Ness/ Potrero		MISSION		1/27/2021	Outreach & Cleaning	16	10	6	0	0	0	14	4	5	5	4
220 San Bruno		SHOWPLACE SQUARE		1/28/2021	Re-Encampment Prevention	10	10	0	0	0	0	11	2	3	6	0
30 Onondanga		CAYUGA		1/28/2021	Resolution	3	3	0	0	0	0	1	0	0	1	0
901 Rankin		BAYVIEW		1/29/2021	Resolution	16	16	0	8	8	0	16	2	6	8	0
					Totals	94	80	14	8	8	0	93	22	19	52	13
												Rates:	24%	20%	56%

The weekly encampment reports recorded by HSOC provide a foundational understanding of the day-to-day activities of their teams, as well as some information about their outcomes. More specifically, these reports provide the date and location of every HSOC operation, as well as the number of tents removed and the placement outcome of each “client” engaged. This data, which is very similar to the compiled data presented at the Local Homeless Coordinating Board, and doesn’t provide much insight besides the fact that HSOC has very high success rates in terms of tents cleared, but comparatively low success rates in terms of service connection. These reports become much more informational, however, when contextualized by both the first-hand experience detailed in this report and the other two public records sources. 

DPW Bag and Tags

    While the encampment reports reveal that HSOC’s operations result in a large number of tents and structures being “removed,” they do not detail whether those belongings were disposed of, bagged and tagged according to DPW policy, or taken by “clients” to other locations. Partially helping to clarify this omission, DPW’s monthly Bag and Tag Logs detail every item brought to their lot throughout the month as a part of the bag and tag process. Further complicating the issue, not every DPW bag and tag is conducted as part of an HSOC operation, as many are belongings that have been confiscated by SFPD police officers for various reasons and picked up by DPW workers at the police station. In order to determine how many bag and tags were collected as a part of HSOC operations, the addresses from the logs for January and February were cross-referenced with the HSOC operation locations from the encampment reports. Through this process, it was determined that there was not a single logged bag and tag whose location correlated with that day’s HSOC operations. 

Daily Shelter Allocation Summary

The city’s shelter system, its capacity, and how its beds are distributed are topics that are essential to understanding how HSOC operates, but are not understood widely outside of city insiders and the advocates who work regularly on these issues. To give a brief overview that will help understand this information: As a result of the pandemic, public access points for shelter, including the former 311 shelter waitlist, were done away with as a part of the city’s work to depopulate congregate shelters. Since then, access to all shelter beds, including SIP hotels and Safe Sleep Sites, has been centrally distributed from the COVID Command Center (CCC). Every day (based on the information found on the emails and conversations with frontline HSOC staff, this happens at or after about 9:30 a.m. daily), the CCC sends an email stating how many shelter beds will be made available for that day, and which city teams they will be allocated to. This is not an accounting of all available shelter beds in the system, but a determination based on unknown criteria of how many of those beds can be given out that day. After this email has been sent, outreach workers on the HSOC team can begin offering them to encampment residents. Upon running out of shelter beds, HSOC is occasionally able to borrow allocations from other teams, or tells clients that they will return the next day with additional resources. 

    The public records analyzed in this report include the daily allocation summary emails for every site, every day in the 37-day period discussed throughout. While there are several useful pieces of information in these records —including the time that these emails are sent daily, the total number and types of shelter beds made available every day, and the percentage of available beds that are allocated to HSOC as opposed to other teams —this report primarily focuses on the number of shelter beds allocated to HSOC each day. This number provides much more context to the service connection data provided in HSOC’s encampment reports. For example, on January 25, HSOC reported engaging with 26 total clients across two operations. Of those clients, five were connected with services, three were “already housed/sheltered,” and 18 reportedly declined services. On the same day, the shelter allocation summary shows that HSOC had access to five total beds across all shelter sites, indicating that the 18 remaining encampment residents could hardly be considered to have declined services. As argued in the report, conducting this analysis across the entire 37-day period greatly supported two of the main claims often made by observers and subjects of HSOC operations: that HSOC regularly displaces more people than it has shelter resources for —not to mention the adequacy or appropriateness of those resources —and that the reported rates of declined services are greatly inflated.