If you hit the century mark in the United Kingdom, you can look forward to two things: one firework of a birthday cake and a card from the queen.
It’s a tradition born of respect, but it hits a snag.
Queen Elizabeth may have sent tens of thousands of premature birthday cards, according to The Economist, but neither she nor her top statisticians know for sure.
That’s because various estimates of people aged older than 90 in England and Wales differ by about 46,000.
It may be hard to believe that a country with birth certificates, death certificates and a decennial census could lose track of a tenth of its nonagenarians. But the experience in the U.K. highlights an important fact: It’s difficult to count people.
Humans are pesky creatures, constantly moving, losing census forms or simply not bothering to fill them out at all. Statisticians rely on projections rather than hard counts to calculate the number and location of people. In the end it’s an extremely well-informed, highly mathematical guess.
On Jan. 27, King County will embark on a census project one demographer described in understated terms as “challenging” — a count of people experiencing homelessness.
Every January, the resulting number splashes across headlines, the executive summaries of consultant reports and political speeches as evidence of the urgency of the homelessness crisis.
Advocacy organizations — Real Change included — give it a prominent position in informational materials and those website popup ads, decrying the now-annual increases in the same breath they ask for the resources to lessen the suffering the number represents.
The number — 10,688 homeless people in shelters, transitional housing and outdoors for King County in 2016 — deceives in its apparent precision, like the number of servings in a package or your Facebook friends.
It obscures both the method and the madness, the how it came to be and why.
Like any statistic, the result of King County’s One Night Count — now Count Us In, due to a change in ownership — is born of and presented in the political sphere.
It is a product of top-down federal policy influenced by the Washington State Department of Commerce and implemented by local government with the support of advocates.
It cannot help but reflect its creators.
That said, it remains the most comprehensive window into the size and contours of King County’s homeless population, offering more than a number, but demographic information used to direct services to those who need them most.
But it also determines who those people are.
The method
For 37 years, the Seattle-King County Coalition on Homelessness (SKCCH) ran the One Night Count. In 2016, more than 1,000 volunteers split into teams and canvassed areas urban and rural between 2 a.m. and 6 a.m., seeking out unsheltered homeless people.
These warm-hearted, otherwise freezing, amateur demographers tallied every individual they spotted, noting general information about their arrangements: on a bus or a bench, standing or in a sleeping bag.
They looked for signs of life in cars and tents tucked away to avoid detection. Each tent, structure or vehicle automatically counted as two people.
The search did not cover every inch of the county, instead covering locations based on experience, research and suggestion that people were sleeping there, said Alison Eisinger, executive director of SKCCH.
In 2016, volunteers did not enter the Jungle, now referred to as the East Duwamish Greenbelt in polite circles. It had been the site of a shooting that resulted in two deaths just days before, meaning the 200 to 400 people evicted later that year may have been missed in the 2016 count unless they were outside the Jungle that morning.
Shelters reported their clientele and surveys went out to providers to collect more detailed information to add to the mix, fleshing out who struggled to stay warm and dry.
All told, the counters found 10,688 people in 2016 with almost as many outside as in: 4,505 people unsheltered, 3,200 in shelter and 2,983 in transitional housing.
This year marks a shift away from tradition. All Home King County, formerly the Committee to End Homelessness, put out a competitive bid for the count, selecting SKCCH and Applied Survey Research (ASR), a California-based firm. The two were to divide duties with skcch managing the One Night Count in January and ASR crunching the numbers. But SKCCH declined, informing its list of volunteers that the One Night Count would not be happening.
A count will still take place, now under the moniker of Count Us In. The concept remains the same, but the strategies have changed. ASR plans to send volunteers into every census tract in the county. Surveys will be filled out over two weeks by the clients, rather than the providers. They’ll collect not just demographic data, but also help create a new way of estimating the number of people living in cars and tents, potentially changing the calculus of the unsheltered count.
Overall, the counts as described sound well-designed, said Adrian Raftery, a sociology and statistics professor at the University of Washington who researches demography.
But he offered a word of caution: The number is almost certainly an undercount. Some people will be missed.
“They’re missed because they’re missed,” Raftery said. “Counters don’t notice them, they’re in out of the way places. The area where they are hasn’t been covered. Or maybe they’re out of the area temporarily on that particular night.”
The madness
Numbers give the perception of cleanliness. They act as a baseline from which policymakers work, a shared set of facts and definitions in the preamble to action.
But government-by-the-numbers becomes government for those who are counted, not those who are missed or those never intended to be counted at all.
The federal Department of Housing and Urban Development (HUD) lays out guidelines each year for communities to conduct its required point-in-time counts, the federal term for the events. This 40-page document describes who will be tallied and how.
Counted: people in shelters, some HUD-funded housing, housing vouchers and those on the street.
Excluded: people in unstable housing, those doubled up, couch surfers or people in rapid rehousing, a program that offers rent assistance for between three and nine months.
As an example, look at two separate housing programs that bring homeless people indoors. Transitional housing, an up to two-year program with wrap-around services to keep people housed when they leave? Those residents appear in the total of the point-in-time count of homeless people. Rapid rehousing, a relatively new HUD-approved program that puts people into market-rate housing quickly with a public subsidy but fewer services? They are not counted.
These choices have consequences that can create a cognitive dissonance between the reality that people see in front of them and the world depicted in impersonal integers.
Consider the curious case of Utah. The state announced with great fanfare in 2015 that it had solved the crisis of chronic homelessness. It was widely lauded in headlines, and policymakers began to throw around catch phrases like “housing first” with great abandon and little analysis.
But the “Utah Miracle,” as Real Change founder Tim Harris describes it, turned out to be a linguistic parlor trick. Chronic homelessness is a very specific term in HUD parlance, and leaves out the vast majority of people experiencing homelessness. A year after Utah declared victory, 1,041 people slept in its largest homeless shelter, according to The Guardian.
“They’re so willing to live and die by the data,” Harris said. “They count at different times, change the definition of who’s homeless. Couchsurfing, not homeless anymore. Boom, one-third of the homeless declassified.”
Harris has watched the evolution of point-in-time counts since their inception in the mid-1980s, roughly when he began homeless advocacy in Boston, Massachusetts.
The process began as a statistical exercise in shelters in 60 areas used to derive the number of unsheltered homeless. It’s now required at least once every two years for every community in the country.
The national count came into its own at the turn of the century with the introduction of the Homeless Management Information System (HMIS), a database for service providers to track their clients. HUD now requires providers to use HMIS if they want federal funding.
“I think HMIS system developed because they obviously wanted data, to understand who they were dealing with,” Harris said. “Data helps with that, but a big part of that data collection was the priorities that it sets.”
When the federal government sets its priorities, they filter through the system.
A program such as rapid rehousing works in Houston, Texas, because it has high vacancy rates and relatively low cost of housing. Seattle has neither. That didn’t factor into the results of a consultant report prepared by Barbara Poppe, the former head of the United States Interagency Council on Homelessness and accordantly steeped in federal policy priorities. Poppe’s report, which informed Pathways Home, Seattle’s newest plan to combat homelessness, included a heavy emphasis on rapid rehousing, but stopped short of examining the local housing market or otherwise determining if putting formerly homeless and low-income people into market-rate housing in Seattle could be successful.
Policies must fit to the local environment rather than be copied indiscriminately from place to place, Harris said.
“It’s astonishing, the willful blindness,” Harris said.
Coming together
But Count Us In is more than just another box to check in an ongoing list of federal requirements.
It’s a tool of inspiration and organization, an annual chance to take stock, examine why the problem exists and consider how to fix it.
It begins the night of the count, when people go out with neighbors and strangers to comb through their communities for the most vulnerable of their residents.
Eastside cities collaborate on the count, bringing in volunteers regionally rather than trying to coordinate the night individually. It’s similar to how they approach homelessness in general, said Leslie Miller, the human services coordinator in Kirkland.
In 2016, the Eastside documented an 80 percent increase in homelessness and they acted together to get that number down. Kirkland opened a children and family day center, Redmond cares for the young adults and Bellevue opened a 24-hour men’s shelter.
The count is about more than just designing policy, Miller said.
“An important function of the One Night Count was that it was a community function,” Miller said. “City leaders and community coming together to make sure that as many people as possible were counted. That act in and of itself was recognition that folks are not invisible.”
When the count ends, the work does not.
The event kicks off a series of other actions run by organizations throughout the county. The changes to the count – namely that ASR will not produce a final tally until months after the fact – have forced organizations like Real Change to alter traditions like the gong-ringing event, where people come forward to strike a gong once for each person found unsheltered.
The timing of the release, now in the warmer months of spring rather than late January, strikes Harris as a tactic that removes some of the urgency felt by the housed to help people sleeping rough in freezing temperatures.
“Now they’re going to release that number under the conditions of pleasant weather,” Harris said. “You can’t tell me that’s not political, intentional.”
Mark Putnam, executive director of All Home, assures that it is not. King County is just one of the company’s clients, and it takes time to issue the surveys, crunch the numbers and determine the new metric that will be used to estimate residents of cars, vans and recreational vehicles.
The final total will be more accurate and help better target services to help people with their individual needs, Putnam said.
“That’s been an important tradition in our community, and I understand there will be skepticism and cynicism about the why,” Putnam said. “But we remain as committed as ever to reducing these numbers.”
In the meantime, All Home will continue to participate in events like the Washington Low Income Housing Alliance’s advocacy day, held the week after the count.
SKCCH has a host of events as well, including “Beyond the One Night Count,” also referred to as “homelessness advocacy 101” to keep people educated and engaged, Eisinger said. It’s also prime time for advocates to engage with the process in Olympia, where legislators grapple with laws and policies that impact the lives of people experiencing homelessness.
“We’re committed to making this an opportunity to engage people who see the reality firsthand and understand that in our rich country, it is not only unacceptable but outrageous that we have so many without basic shelter and safety, and there are specific actions we can take to fix that,” Eisinger said.