How to Census the Homeless

Last time, we talked about how Emergency Managers can contact vulnerable populations–for instance, the homeless–with evacuation notices. We discovered that social connections which make communities more resilient as a whole make the homeless in those communities more resilient too.

But as an Emergency Manager, I know that resilience often depends on the effects of long-term, complex social issues. For instance, Haiti sometimes finds it more difficult to recover from earthquakes than it’s sister country, the Dominican Republic, because Haiti is poorer, has non-resilient infrastructure, and social policies which (debatably) prevent economic recovery. Similarly, in America, policy discussions centering around social-political issues like homelessness impact the long-term resilience of a community to disasters.

Policy discussions are important and they rely heavily on statistics. As you know, fair representation in a community means being accurately counted. How can policy makers and advocates make decisions about policy and funding without a careful understanding of the demographics and size of a vulnerable population?  As you might imagine, counting a scattered and mobile population is extremely difficult. This article from the Seattle Times demonstrates how parsing the census data can greatly change the picture; depending on how you ask the question, Seattle has between the third and sixth most homeless in America.

The severity of Seattle’s homelessness crisis is different depending how you measure it — but no matter which way you look at it, we’re in the top 10 worst in the country.
Courtesy of the Seattle Times

 

So how do you count the homeless population in your town? In America, most cities choose to do an annual, night census of the homeless to get a sample.

Counting the unhoused once a year to get a sample.

Most cities have an annual, nightly census where volunteers physically visit known homelessness sites and gather demographic data from those sleeping outside. Sometimes this is accompanied by census data gathered from people using homelessness services. This method of data collection can be problematic for several reasons.

a) A yearly census may have data gaps. While it’s true a yearly census is good for trends over time, it gives only one snapshot of the homelessness situation on this night. Some argue that more frequent censuses would give granularity to the data.

b) Each jurisdiction and agency draw the geography differently. As the Seattle Time’s points out, because of the hodgepodge way each locale accesses funding, each city and county may count people differently. “New York City, for example, counts the homeless people inside its city limits while L.A. counts everyone in L.A. County except the cities of Pasadena, Glendale and Long Beach…. Denver and Boulder group together and count all the homeless people in the six counties around them….” This can make comparing data nationally–or even regionally–a headache.

photo
Pic courtesy of Associated Press

c) It doesn’t count those in shelters (sometimes, depending where you are). This isn’t necessarily a problem depending on what you want the census to tell you. Some cities, like New York are under legal mandate to house their homeless, so counting those strictly on the streets matters to them. But maybe your city wants to know more about how many people might need a certain governmental service. In that case, this statistic might not help you, beware.

d) It doesn’t count the “concealed homeless”–those living on a friend’s couch or in their cars. Like point c, this is a matter of semantics; neither good nor bad, but worth noting.

volunteer
Pic courtesy of San Bernardino Point in Time Count

Is crowdsourcing better? No matter what you do, it’s hard to get a clear estimation of “hidden” individuals. Cities are big and individuals are small. Crowdsourcing has helped gather data on large, complex problems before, maybe–the thinking goes–it can do the same here.

New York City has tried just that with their 311 app. Designed to accumulate citizen complaints for a variety of municipal problems, it was expanded to allow users to help identify the homeless. App-users can take geo-tagged pictures of the homeless and tag them with statements like “NeedsMedicalAid” or “AggressiveBegging”. Unfortunately, some homelessness advocates feel this app has led to shaming and harassment (especially–in my opinion–given that the complaints can be viewed by the public). (View the story here.)

On the other hand, a competing app in New York named “WeShelter” has a gentler image.  App-users “unlock” donations from sponsors which are given to several of New York’s homelessness organizations. Users have an opportunity to share their location with the app by clicking “I’m near a person who is homeless” and can also provide other information like whether or not the person needs non-emergency assistance. This is an interesting combination of location data gathering with activism. And there are many, many more apps with various methodologies out there.

GiveSafe Beacon. Pic courtesy of Smart Cities Council

In Seattle, an app called GiveSafe distributes beacons the size of a quarter to individuals in need. The beacon has a bluetooth that connects to an app on your phone. You can see the individual’s story as you pass them and also donate, if you wish to. The beacon holder can then spend the money at select merchants or non-profits. To keep the beacon active, the holder must keep in touch with a counselor once a month. It was designed hoping that donors would be more willing to donate if they could be assured that their money would be used to help with food (for example) and not vices. (See also StreetChange in Philidelphia & HandUp mostly everywhere)

 

Conclusion

Homelessness census data–like most statistics–are complex and variable. However, policy makers and municipalities rely on this data to distribute funding and reassess their policies. It’s up to us to make our statistics as robust and meaningful as possible in order to support that work. Often, that means gathering different kinds and sources of data to create a mosaic-like picture of the situation, and experimenting with data gathering methodologies.

Further Reading

Advertisements

What do YOU think?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s