Anchorage changed how it counts its homeless population during the pandemic. The number doubled.

For a long time, Anchorage’s estimate for how many people were homeless in the municipality was based on a crude tool: a citywide canvass in the depths of winter to figure out who was sleeping outside and in emergency shelters. Called the point-in-time count, it is a requirement to receive federal funds to help local governments with housing and homelessness.

According to the tally from this past January, there were 1,760 people homeless in Anchorage.

But officials and experts say the real number is roughly double that, around 3,000 people. And they have the data to back it up.

During the pandemic, when the city drastically restructured how it handles housing, social services and emergency shelter, a number of entities started using new tools for tracking individuals cycling through different stages of homelessness. The data is more complex and multifaceted than the point-in-time count — and it’s also more accurate, according to those familiar with the issue.

“It is a blessing in disguise,” said Owen Hutchinson, director of external relations for the Anchorage Coalition to End Homelessness, of the new estimates.

On the one hand, Hutchinson said, they show the city’s challenges with homelessness are more extensive than earlier demonstrated. But on the other, the data is guiding smarter, more refined policy.

It’s also helping city leaders make a sharper argument to state and federal officials that because of policy oversights, the municipality is not receiving adequate support to deal with the issue.


“The people of Anchorage, the leadership of Anchorage, are dealing with a state and national emergency without support,” said Assembly Chair Christopher Constant, one of a handful of members who, along with Mayor Dave Bronson, supported a resolution last week requesting policy and funding changes to better deal with homelessness.

[Short on time and options, top Anchorage homeless official warns Sullivan Arena could again be only choice for winter shelter]

Under the point-in-time estimates, the city already had a disproportionately high rate of homelessness relative to the state as a whole, as well as to cities and jurisdictions outside of Alaska. The municipality counted as many homeless people as cities including Baltimore and Indianapolis, which have total populations two to three times the size of Anchorage.

With the new data, gathered under the Alaska Homeless Management Information System, that disproportionality is even starker.

A single point in time

The point-in-time count the city uses to come up with an estimate for how many people are homeless has merits and shortcomings.

It’s been used for so long, in so many different communities, that it has utility as way of making apples-to-apples comparisons. It’s also helpful for understanding how homeless numbers have changed over time.

“It’s really meant for long-term trends in shelter usage and unsheltered homelessness,” Hutchinson said.

Every year, on one single day at the end of January, workers fan out across the city to count people staying in tents and camps along the city’s greenbelt, and collect figures from shelter operators about how many people are staying there.

But many working on homelessness have long complained the methodology is poorly suited to Anchorage for a number of reasons. The biggest is that people avoid sleeping outside during the depths of winter, and may opt for other options like doubling up in homes, piling into cheap hotel rooms or sheltering in abandoned buildings not fit for human habitation.

“The primary shortcoming is that the time of year that the point-in-time is done is the last week in January, and so the last week in January is the least hospitable time of year in Anchorage,” said Constant.

Campers who choose to stay outdoors all winter long are not necessarily easy to find, particularly at a time of year with lots of snow in the woods and little daylight.

[A death on Fireweed Lane]

A few years ago the city started conducting its own second count in the summer, when more people tend to forgo indoor options and camp outdoors, but that data is not submitted to federal funding entities like the Department of Housing and Urban Development, which allocates funds every year for homeless assistance programs.

“There are many reasons the count is conducted at the end of January,” wrote Vanessa Krueger, a public affairs officer for HUD’s Northwest Region, in an email. “Counting and interviewing people sleeping in unsheltered locations during the winter months can provide a more precise count of people who are unable or unwilling to access emergency shelter or other crisis response assistance. A count on one of the coldest nights of the year can be very effective in raising public awareness of the challenges faced by homeless people without shelter.”

HUD understands the limits of the point-in-time methodology, Krueger said, but still views it as “a critical factor in helping communities understand their homeless population.”

“Because it is easier to count people in shelter than on the street,” she said, “conducting the count on a night when the shelters are most full will lead to the most accurate count.”

A tool for better data

Anchorage is unlikely to be the only jurisdiction with a point-in-time estimate that’s an undercount.


So in recent years, many local governments across the country have started relying instead on Homeless Management Information Systems. In key ways, it is a superior method to the point-in-time count, with centralized databases that collect more detailed, granular data from clients and social service providers and that are analyzed in near-real time.

Though not every service provider or shelter in Anchorage reports data to the Alaska Homeless Management Information System, or AKHMIS, most do. And that data captures not just who is sleeping outside or on an emergency shelter cot on a given night, but individuals in various states of housing insecurity, such as staying in transitional housing, couch-surfing, squatting, doubling or tripling up with family and other arrangements.

Hutchinson said the new way of collecting data picks up many more people without stable housing who do not count in the point-in-time estimates reported to the federal government, including, for example, people in small group shelters or halfway houses.

“Until someone has their name on a lease, they are considered homeless,” Hutchinson said.

The new system reports homeless counts that are averaged across a given month. In March of 2023, 3,034 individuals in Anchorage were considered homeless, up slightly from an average of 2,923 people experiencing homelessness at some point in the last year’s worth of data, and lower than the high of 3,304 counted in May of 2022.

Historically, the city has not spent much on homeless services. That changed during COVID, when the municipality opened Sullivan Arena as an emergency measure and handled outsourcing operations to subcontractors and footing the bill until federal reimbursements arrived. The situation upended decades of homeless policy in Anchorage, drawing in the municipality as a participant in new ways. And it was during that time that the new data collection system was widely adopted.

While federal officials say HMIS data can be more precise than other tools, it is insufficiently standardized across different communities, according to Krueger.

A funding disparity

As the city has gotten more clear-eyed on the scale of homelessness in Anchorage and across Alaska, it has started to look for money to pursue solutions, including the Assembly resolution backed by Bronson that passed last week asking for outside help.


“The administration fully supports the resolution because we do need more state and federal funding,” said Alexis Johnson, the city’s homeless coordinator.

Aside from several million dollars collected through the local alcohol tax, Anchorage has used very little taxpayer money to fund things like emergency shelter or supportive housing. That has meant relying on nonprofits and charity groups to handle the lion’s share of delivering complex social services to a high-needs group, and occasionally making do with inadequate, repurposed facilities and inconsistent availability when there is a crisis.

“Even right now we have no funding to build or fund a new shelter system,” Johnson said. “Because we’re very limited on funding, we have to close down in summer, and we have to use municipal buildings rather than a purpose-built facility.”

[In April, a record 8 people believed to be homeless died outside in Anchorage]

More resources might be available, city leaders are starting to argue, if federal departments awarded assistance money differently. They are asking the Department of Housing and Urban Development to consider revisiting its funding formula in a way that takes into account a place like Anchorage’s high rate of homelessness relative to its overall population, as well as the high cost of providing services and constructing new housing units.

Even though point-in-time data is not as accurate as what the newer system picks up, it is still helpful for showing how disproportionately little Anchorage receives in federal assistance relative to big cities.

Anchorage counted 1,760 people staying on the street and in overnight shelters in its 2023 point-in-time census, an increase over the 1,494 counted in 2022. Data from 2023 is not available for all 387 jurisdictions across the country that report point-in-time counts to the federal government. But a comparison of 2022 data shows Anchorage has a high rate of homelessness relative to many other jurisdictions.

The number of people homeless in Anchorage, with an overall population just below 290,000, is comparable to places like:

• Indianapolis, which reported 1,761 homeless residents in a population of 882,039, according to the latest Census Bureau figures.

• Baltimore, which reported 1,597 homeless in a population of 576,498.

• Oklahoma City reported 1,339 homeless among a population of 687,725.

• Montana, which reported 1,515 homeless residents for the state as a whole, with a population of 1.12 million.


“We have as many unsheltered people in the point-in-time as much bigger cities,” Constant said.

But federal dollars are allocated based on, among other factors, a jurisdiction’s overall population, not its rate of homelessness within that population.

That’s why larger cities “are still seeing seven to 10 times more than we are in terms of funding,” said Constant.

“We are unfairly funded because of a specific geography and a specific kind of arrangement of homeless services in Anchorage and across our state. The definition of homelessness doesn’t allow us to count couch-surfing and overcrowding, and we have the highest rates in the country,” Hutchinson said. “The funding priorities privilege jurisdictions with high populations … so we will continue to be scored low based on factors that are outside of our control, when the number of people experiencing homelessness is quite high.”

One of the examples of that funding disparity cited in the Assembly’s resolution is Fort Worth, Texas, and its surrounding counties, which reported 1,665 homeless and received nearly $16 million in federal homelessness assistance in 2022. The same year, Baltimore received $26 million for homeless programs and projects.

Anchorage, with its comparable number of homeless individuals within a less populous community, got $4.1 million.


“HUD recognizes the formula could be improved,” Krueger said, adding that the department has been assessing potential changes since 2016.

“It is important to note that while HUD wants to more closely align the need formula to the effort to prevent and end homelessness, we also have to be cautious about creating perverse incentives in our data collection,” Krueger said. “HUD is mindful that if the PIT count became a major driver of a (community’s) funding allocation that would incentivize communities to report numbers that will result in the highest potential funding.”

Johnson, Constant and Hutchinson all pointed to the housing supply shortage as a primary cause for Anchorage and Alaska’s high rates of homelessness. But in the shorter term, all three are part of the push for reformed state and federal policies.

The recently passed resolution asks the governor and Legislature for help expediting applications for Medicaid and some of the benefit programs eligibility unlocks. It also asks the congressional delegation and HUD Secretary Marcia Fudge to reexamine the department’s funding formula to take into account the rate of homelessness rather than just overall population, as well as the inflated costs associated with housing construction in Alaska.

“The idea is: ‘Secretary Fudge, please look at the formula, something’s not right here,’” Constant said of the measure. “It’s just fundamentally unfair.”

• • •

Zachariah Hughes

Zachariah Hughes covers Anchorage government, the military, dog mushing, subsistence issues and general assignments for the Anchorage Daily News. He also helps produce the ADN's weekly politics podcast. Prior to joining the ADN, he worked in Alaska’s public radio network, and got his start in journalism at KNOM in Nome.