A City Leader's Starting Guide to Addressing Homelessness

Start here

Homelessness feels unsolvable when you try to solve all of it at once; it becomes manageable when you solve it one group of people at a time. This handbook is a free starting point for city staff and elected officials who are new to the issue or inheriting one that has stalled.

Who this is for: city managers, department staff, council members, mayors, and their regional partners.

What it is: a way to orient: what to consider, what to ask, and where to begin.

What it is not: a full playbook, a legal opinion, or a one-size-fits-all plan. Your community's numbers, housing market, and state laws will shape every answer.

The core idea in four moves

1. Count and name. Know who is homeless in your community, by group.
2. Separate the groups. People enter homelessness for different reasons and need different responses.
3. Build for one group at a time. Create a system that can reliably house that group, reach "functional zero" for it, then move to the next.
4. Hold the line on public spaces. Decide what you will and will not allow, and pair every limit with a real place to go.

How to use this handbook

• Read Parts 1 to 5 as a team before making any decision. They are the shared mental model.
• Use the question lists in Parts 2 to 11 as agenda items for working sessions, not a checklist to finish in a week.
• Use Part 12 to pick a first move you can start within 90 days.

Know your numbers first

You cannot size a solution until you know how many people are homeless, who they are, and how long they stay. Most cities already have more data than they realize, held by different agencies.

What to find out

• Point-in-time count: the annual snapshot your local Continuum of Care (CoC) reports to HUD. It undercounts, but it is the common starting reference.
• Homeless Management Information System (HMIS) data: shows inflow (newly homeless), outflow (housed), and length of time homeless over a full year, which is far more useful than a single night's count.
• A by-name list: a real-time list of every person known to be homeless, with their history and needs, so you work with people and not averages. Communities that have reached functional zero rely on this.
• Local signals: 311 calls, police and fire contacts, ER visits, jail bookings, school district counts of homeless students, and encampment reports.

Why this comes early

The inflow-versus-outflow gap is the single most important number. If more people enter than exit, adding services without changing that math only slows the growth.

Questions to ask

1. Who owns the data today, and will they share it with the city?
2. How many people become homeless each month, and how many exit to housing? Is the gap growing or shrinking?
3. How long does the typical person stay homeless? How long do the longest stayers?
4. Where does the city's data disagree with what staff and residents see on the street?

Homelessness is not one problem, it is three

Most people who become homeless are homeless briefly and do not return; a much smaller group stays for years and consumes most of the system's resources. Treating them all the same wastes money on some and fails others.

The three-group pattern below comes from long-running research on shelter use. Exact shares vary by city, so verify against your own HMIS data.

Why the distinction drives strategy

• A family evicted over a $1,800 rent gap and a person who has lived in a tent for six years are not the same problem, even if both are counted the same night.
• The visible problem (chronic, unsheltered) is the smaller share of people but draws the most public attention. The larger, quieter problem (transitional) is the cheapest to fix and the cheapest to prevent from becoming chronic.
• A single program model, for example one shelter or one case-management approach, will usually serve one group well and the others poorly.

Questions to ask

1. What share of our homeless population falls into each group today?
2. Which group is growing fastest? Which group drives public complaints? Are those the same?
3. Which group is our current system built for, whether or not we meant it to be?

Ending homelessness is different from fixing why it happened

A person is homeless because they lack a home; they became homeless for many reasons. Cities succeed when they solve the first problem reliably, and stall when they make the second a precondition for the first.

Why the order matters

• It is far easier to treat a health condition, hold a job, or stay sober with a door that locks than on the street. Housing is often the platform for the second problem, not the reward for solving it.
• Making people "earn" housing by first fixing every underlying issue is the most common reason systems stall.
• At the same time, housing alone does not solve everything. For people with serious needs, housing without support can fail, which is why Part 6 matches infrastructure to each group.

Questions to ask

1. Where in our system do we require someone to fix their underlying problem before they get housing? Is that requirement working?
2. Who in our community actually owns the "why" (behavioral health, hospitals, courts)? Are they at the table?
3. What does the city control directly, and what does it have to coordinate with the county or state?
4. How do we measure housing outcomes separately from recovery outcomes so neither hides the other?

Functional zero, one group at a time

You do not need to end all homelessness at once; you need to build a system that reliably ends it for one defined group, then repeat. "Functional zero" means that, for a given group, the number of people homeless at any moment is no greater than what your system can house quickly, so homelessness for that group becomes rare, brief, and one-time. It is not literally zero people, and it is not a claim that no one will ever lose housing.

The repeating cycle

1. Define the group. Pick a subpopulation that can be counted by name, such as chronically homeless adults, veterans, youth, or families with children.
2. Set the baseline. How many people are in the group today? How many enter and exit each month?
3. Size the capacity. How many housing placements per month can your system sustain? That number, not the total count, is your real constraint.
4. Close the gap. Add housing, services, and process improvements until exits reliably exceed entries.
5. Declare functional zero, then hold it. Keep the by-name list and the response running so the group stays at zero.
6. Move to the next group. Reuse the data, partners, and habits you built.

Why this works for a city with limited resources

• It turns an open-ended problem into a bounded goal with a finish line.
• It produces visible wins that build political and public trust for the harder phases.
• It forces the city to build the machinery (data, coordination, housing pipeline) that every later group will also need.
• It lets you pick a baseline your community can actually handle, instead of promising a citywide fix.

Choosing which group goes first

There is no single right order. Consider:

• Size: can your system realistically cover this group with current or near-term resources?
• Cost and harm of inaction: who is costing the most in crisis services, or at greatest risk of dying outside?
• Available funding: are there dedicated federal, state, or philanthropic dollars for this group?
• Public and political urgency: what will earn the trust needed to continue?
• Early-win potential: a smaller group like veterans or families has often been first, because dedicated resources already exist.

Questions to ask

1. Which group will we commit to first, and why that one?
2. What is our monthly housing placement capacity today, honestly?
3. What would "done" look like for this group, in numbers, and who verifies it?
4. What will we do to keep it at zero once we get there?

What each group needs you to build

Each group needs a different set of tools, and most cities have built one or two of them and expected them to serve everyone. Use this table to see where your gaps are.

What to take from the table

• Transitional is mostly a speed and money problem. Fast, small, flexible help ends most of these cases.
• Episodic is a stabilization problem. It needs more than a one-time fix, but usually not lifelong support.
• Chronic is a housing-plus-services problem. It needs permanent housing paired with services that continue as long as the person needs them.
• Everyone is limited by affordable housing supply. If the market has no units, no program will succeed for long.

Questions to ask

1. Which rows in this table do we have, and at what capacity? Which are missing entirely?
2. Who runs each piece today, and how does a person move from one to the next?
3. How long does a person wait between asking for help and getting it? For transitional cases, that wait is the difference between a bad month and chronic homelessness.
4. Where does our service provider capacity (not funding) limit how fast we can grow?
5. Do we have housing that is actually available, or only housing vouchers with no units to use them on?

Laws and ordinances that help or hinder

Your city's own codes often work against its homelessness goals without anyone intending it. Audit them before spending money on new programs. This section is a prompt for that audit, not legal advice; involve your city attorney and check state law.

Know the legal landscape on public space

In City of Grants Pass v. Johnson (2024), the U.S. Supreme Court held that enforcing generally applicable camping bans does not violate the Eighth Amendment's ban on cruel and unusual punishment. That expanded what cities may legally enforce, but it did not answer whether enforcement works, what it costs, or what state law and state constitutions permit. Some states have since passed their own rules, either requiring or restricting enforcement. Verify your current state and local position with counsel.

Questions to ask

1. Which of our own ordinances, if enforced as written, make it harder to house people or open shelter and housing?
2. Do we enforce camping or public-space rules, and if so, is there a real place for someone to go when asked to move?
3. What does enforcement cost per person (staff time, court time, jail time), and what does it accomplish?
4. Which barriers are in city code, and which are in state law we need to lobby to change?
5. Who within city government can say "yes" to a shelter or supportive housing site, and how long does that take?

Chronic unsheltered homelessness and the public

This is the hardest part of the issue politically, because it is the most visible and the slowest to solve. Leaders need to decide, out loud and in advance, what they will do for the people living outside and what they will do for the neighborhoods living alongside them. Those are two responsibilities, and neither cancels the other.

Decisions to make as a leadership team

Record your team's answer under each one.

1. What is our goal for people living unsheltered? Housing for everyone? Safe alternatives for everyone? Reduced numbers? Managed conditions only? Be honest about which, and how long each takes.
2. What is our rule for public spaces? Where is camping not allowed, where is it tolerated, and where is it sanctioned and supported?
3. What is the offer? If we ask someone to leave a location, where do they go, and is that a real option they will accept (not a bed they have already refused for good reasons)?
4. What is our enforcement approach? Citations, arrests, outreach-first, or some sequence? Who decides, and who is accountable?
5. What do we do with people who refuse every offer? This is the hardest case, and a policy that ignores it will fail when it arrives.
6. How will we handle people with severe mental illness or addiction who cannot safely stay outside? What are the state's options for treatment, and where does the city's authority end?

Managing the public impact

Residents and businesses experience homelessness as trash, public drug use, noise, fear, and blocked sidewalks. These concerns are legitimate and deserve a response even while housing is being built.

• Be specific about what you will respond to: public safety, sanitation, blocked rights-of-way, fire hazards, harm to parks and waterways.
• Build a single point of contact so residents are not bounced between departments.
• Provide the basics: public restrooms, trash service, hygiene access, and storage reduce the problems that drive complaints.
• Publish data: how many people were contacted, offered shelter, housed, and what changed. Silence breeds distrust.
• Set honest timelines: do not promise that encampments will be gone in a month if housing will take two years.
• Treat neighbors as partners: hold community meetings that begin with listening and end with commitments.

Common failure patterns to avoid

• Clearing encampments with no destination, so people simply move and the problem relocates.
• Opening shelters or housing without service funding, so they fail and public trust falls.
• Switching strategy every budget cycle, so nothing is ever tested long enough to know.
• Treating enforcement and services as opposing camps rather than parts of one plan.

Questions to ask

1. How many people are chronically unsheltered here, by name, and what does each need to get inside?
2. Which locations create the most harm or complaints, and what is the fastest humane way to reduce it?
3. What would make someone say yes to an offer of shelter or housing? Who has asked them, and how?
4. What can we promise the public in 6 months, 12 months, and 3 years, and can we deliver it?
5. How will we measure whether residents feel the difference, not just whether counts fell?

What the city does

Ending homelessness does not require fixing every social problem. The city's work comes down to three jobs, and each one is a set of ordinary, doable actions with a result you can measure.

• Solve ends homelessness for a group of people: housing, help to stay housed, and the paperwork that unlocks both.
• Manage keeps conditions from getting worse while you solve: emergency housing, outreach, clean and open public spaces, and safety and health response.
• Run the system is the data, contracts, and reporting that make the other two work.

It is the same Solve and Manage split you will use to tag spending in Part 11.

Questions to ask

1. Which of these thirteen jobs does our city already do well, and which do we skip?
2. Who owns each one by name, with authority to make decisions?
3. What do we measure for each, and who publishes the numbers?
4. Which job will residents notice first, and which one stops the problem from growing?

Who needs to be at the table, and who pays

No city can solve homelessness alone, because the causes and the tools are spread across governments and sectors. Someone has to be accountable for the whole effort, and it should be a named person with authority, not a committee with none.

Partners to bring in

Funding sources to map

• Federal: HUD Continuum of Care and Emergency Solutions Grants, VA programs, Housing Choice Vouchers, Medicaid.
• State and county: housing trust funds, behavioral health dollars, opioid settlement funds where available.
• Local: general fund, housing bonds or dedicated taxes, land the city already owns.
• Private: hospital community benefit, philanthropy, employer contributions.

The point is not to list every source; it is to know which pots exist, who controls each one, and what they will and will not pay for.

Questions to ask

1. Who is the single accountable lead for our homelessness response, and what authority do they have?
2. Which of our partners are at the table, and who is missing that holds a key piece?
3. What are we paying for now, and what would change if we paid for outcomes (people housed) instead of activity?
4. Which funding is one-time, and what happens when it ends?
5. Is this a city problem, a regional one, or both? Do neighboring cities share responsibility or simply export it?

Fund outcomes, not activity

The most common funding mistake cities make is paying for activity instead of results. Cities do not run any other core function this way: when a street needs repair, the city defines the work, sets the standard and deadline, and asks who can deliver it. With homelessness, many cities hand out money and ask nonprofits how they would like to spend it. A city's job is to solve a problem, not to keep organizations open.

Imagine announcing that Main Street has a mile of potholes and $1 million to fix them, and asking residents what they would do with it. Homelessness funding should work the way road repair does: here is the result we need, here is the standard, here is the deadline, who can do it?

Two funding categories, kept separate

Every dollar should be tagged to one of two jobs. Keeping them apart stops "managing" spending from passing as progress, and stops it from being cut for "not solving" anything.

As "solve" work succeeds for a group, its funding should shrink and be redirected to the next group. "Manage" funding is the cost of the gap, and it should shrink only as the gap does.

Buy it like any other contract

A solicitation should say what the city needs, not ask what a provider would like to do. Check your next solicitation against this list:

Where the road-repair analogy needs care

Paying for outcomes works only if the contract is fair and the city does its part. Build these in from the start:

• The city controls inputs the provider cannot. A provider cannot house people without units, vouchers, permits, or land. Secure those inputs, or contract for them too, before holding anyone to a housing target.
• Guard against cherry-picking. If every placement pays the same, providers will take the easiest cases. Tier prices by need and work the by-name list in priority order.
• Pay for retention, not only placement. Otherwise people are housed and then lost. Track returns to homelessness.
• Solve the cash flow problem. Small providers cannot float costs for months. Upfront start-up funds and prompt milestone payments matter.
• Do not starve the "manage" side. If only "solve" work is funded, street conditions will worsen and public support will erode.
• Give multi-year runway. Outcomes take longer than one budget cycle; one-year grants make delivery impractical.
• Handle the transition with care. Existing providers hold real expertise. Give notice, rebid at renewal, and help strong providers reposition rather than switching overnight.

Align private philanthropy around the same goal

Philanthropy often makes the same mistake: many funders, each making separate grants to organizations that pitch them well, with no shared target. The fix is the same. Track your community's progress:

Questions to ask

1. What do we spend on homelessness each year across all sources, and what does each dollar buy in people housed?
2. How much of our spending is "solve" and how much is "manage"? Can we even tell today?
3. Which current contracts could be rewritten at renewal with a defined target, standard, and deadline?
4. Have the city, the CoC, and local funders ever sat in one room and agreed on a single target?
5. What must the city guarantee (units, flexibility, data access) so that holding providers to outcomes is fair?
6. Are we willing to stop funding activity that advances neither category?

Where to start

In the first 90 days, do not launch programs; get the facts, the people, and one clear target in place. The first visible win comes after that foundation, not before it.

A 90-day starting plan

Mistakes to avoid early

• Announcing a goal before you know your baseline.
• Starting with the hardest group because it is the loudest.
• Funding programs with no plan for the housing they depend on.
• Counting activity (outreach contacts, beds filled) as success instead of people housed and kept housed.
• Leaving out the people the system is meant to serve.

Appendices to add as this handbook grows

• A one-page glossary of terms (CoC, HMIS, coordinated entry, Housing First, permanent supportive housing, rapid rehousing, by-name list)
• A sample data dashboard and the metrics to track
• A sample ordinance audit checklist
• A template for a council briefing and a public FAQ
• Short case studies of communities that reached functional zero for a group, and what they did
• Links to reference organizations and federal resources

A closing note

The goal of this handbook is not to hand a city an answer; it is to help leaders ask the right questions in the right order. The communities that make progress tend to do the same few things: know their numbers, match solutions to groups, build one thing well at a time, and stay at it long enough to see results.

Everything you entered, in one summary

This pulls together your numbers, worksheet answers, and notes. Copy it into a memo or an email for your council or team. It updates as you work.

Clear everything I have entered