Skip to main content

Living Data: sharing patterns to accelerate housing

Published onDec 03, 2020
Living Data: sharing patterns to accelerate housing

Notes from my presentation @ the Design Impact conference 12/3

Housing is a human right, and homelessness is a collective failure of social design. Solving it is a question of collective action.

I work on the Underlay project, building a distributed public knowledge graph. We help people connect open data, and capture context that can inspire and support its reuse. This helps people working on different facets of the same problem build on one another’s work, and make progress together.

Good data can accelerate a project. A universe of good data can accelerate a nation.

If we are to end homelessness now, without delay, we need to be able to understand and respond at a community level.

We also need to make it easy to take what works in one place, tweak it, and apply it elsewhere. If we can find a way to increase service or capacity, or decrease construction costs and speed by an order of magnitude in one community, we should have a sense of which communities come next.

So what makes data easier to generalize?

  • We need data to capture changes in communities as they evolve, so we can project forward in time,

  • We need context to understand when two communities share enough of a starting point to share solution templates,

  • And we need connection in order to identify collaborators and complementary initiatives – in order to be in touch with the right people to run models that work in our neighborhoods.

Strong contacts in a community helps realize existing programs in a new context.

We need data to describe questions that we need to answer to know what to try next. We need guidance in framing questions that we can evaluate with an initial foray into running a new intervention, to know what version of known solution might work. You need the data to tell you where to look — what is the critical dimension along which to characterize your environment.

This is a combination of knowing what’s been done elsewhere that’s been successful and knowing what questions you could answer quickly to get confirmation of where you think you are.

This is the coordination we need to build an underlay for housing.

Questions we need to answer

Good data will answer three core questions :

  1. What’s the scope of the problem and the cost to your community? How is that changing over time?

  2. What solutions are working? Or have worked in the past?

  3. And how are we doing? Is your community making progress? How effective are local attempts, compared to similar attempts elsewhere or compared to their self-reported goals?

What is the scope of the problem? How is this changing?

The scope of homelessness in our cities, and the opportunity cost of allowing it to persist, is often hidden. When we fail to house people, the resulting cost far exceeds the cost of providing permanent housing in the first place.

To see this, we need repeatable ways to combine many sources to perceive how cost is spread out across the community, and how that cost compounds over time.

For example, “point in time” homeless counts are widely used — people have been quoting 500,000 as an estimate of homelessness in the US today — but that’s a crude lower bound, undercounting the true figure by a factor of 2 or more.

A full accounting of the people and families affected by homelessness should include groups often overlooked, such as those jailed while homeless, a crisis in its own right.

When prison and homelessness data is siloed, imprisoning the homeless looks like a reduction in the homeless count, and a rise in successful crimefighting — both ridiculous distortions of reality. In fact, we are spending massive sums of money to lock people into destructive cycles.

People dramatically underestimate the cost of not acting, and overestimate the nominal cost of doing something. We have accepted a broken equilibrium, and have accepted overpaying for maintaining that broken state rather than moving to a better one.

What is working? How is this changing?

We also need a better understanding of what is working, in which contexts.

Built for Zero and The New Leaf Project are examples of programs that have seen tremendous recent success. Let’s fill out the map of approaches that are working, the contexts in which they work, and where to apply them next.

For example, COVID has spurred both a rise in evictions and large relief bills for emergency housing. Which existing programs have the flexibility and capacity to be part of that work, on a timescale of months?

And if your community has programs that claim to be addressing those problems, how can you see if that’s actually happening?

How are our own communities doing, now and over time?

We need to speed up our feedback loops that track how things are working, to accelerate our impact. Once you've found a configuration for an operating system that works in a wide range of communities, you want others to try it, and confirm that it works for them. Conversely, many costly programs are slow to spend allocated funds, or not working as advertised. We need current data to understand how our cities are doing.

Every day that there is still homelessness in your community, you should be able to ask, what’s missing? and see current gaps, obstacles, and opportunities.

What can we do about this?

Let's address these missing gaps together —

We've started compiling a catalog of projects, designs, and data online, here, on this site. With a form for sharing datasets that you rely on, and a way to get in touch to request data that you need, or suggest other ways to work together.

Please join us.

No comments here
Why not start the discussion?