1 of 8

Living data: sharing patterns to accelerate housing

Samuel J. Klein

Knowledge Futures Group | Underlay Project

​

​

meta.sj@gmail.com

underlay.org

John Hack, 1977

​

​

2 of 8

Underlay.org

HUD requires that continuums of care conduct a regular homeless count of people who are without shelter, or in emergency or transitional shelters or in safe havens, on a single night.

National housing data

The Underlay organizes collaborative data for

the public good.

​

We aim to make knowledge accessible, connectable, and inspectable, by storing data �with context —

sharing patterns to make sense of the world

Homeless Population

Point in time count: 575k ('19)

School children: 1.4M (NLIHC '17)

HomelessWorldCup.com

Jail population: 250-330k ('16) �

Regions with highest per capita ct:

DC: 0.9% CA: 0.4%� NY: 0.5% OR: 0.4%

​

​

A lower bound

so it can be discovered by people working on different

facets of the same problem.

3 of 8

An living layer of housing data

The need for Living Data is a barrier to acceleration that we can fix without policy changes. This data must be connected, contextualized, and updatable — alive. �

Good data motivates investment at scale,

convinces policy-makers, and

defangs the specter of risk in trying something new.

4 of 8

Good data will answer three core questions:

  • What is the current challenge?
    • Who is affected
    • The cost of not acting (short + long-term)�
  • How can we address it?
    • Patterns that have worked elsewhere (and which fit your current context)
    • Programs + resources available near you (patterns of activity)�
  • Are we making progress, and what should we do next?
    • How current programs are working (metrics, self-reports, external audits)
    • Urgent needs to meet in your community (and how to help)

5 of 8

Patterns of housing failure

What is the current state of � homelessness?

    • Who is affected �- Homeless counts on the street, in jails, in schools�- Demographics individual, family, community�- Regional overviews city, state, national�
    • The cost of not acting (short + long-term)�- To individuals, to cities�- Compounding costs work, health, addiction�- Misdirected costs police, prison

Water Quality

Shelter Capacity

Health Services

Homeless Population

Point in time count: 550-575k ('17)

School children: 1.4M (NLIHC '17)

Jail population: 250-330k ('16) � prisonpolicy.org �

Regions with highest per capita ct:

DC: 0.9% CA: 0.4%� NY: 0.5% OR: 0.4%

​

​

Standard reference, but a lower bound

6 of 8

Patterns of success

How can we end homelessness?

    • What patterns have worked elsewhere? Which fit your current context?�
    • What programs + resources are available�near you?

​

Resources for Veterans

Universal Basic Housing

Modular Construction

City Commitments

Annual homeless count per state

​

Getting to Zero: ⚉ , ⚇ �COVID-19 Pledge: ▨�Mayors Challenge: △ �UBH Coalition ◆

​

​

Overlay of many national datasets

7 of 8

Measuring progress

Where are we making progress?� What should we do next?

    • How are current programs working? (metrics, self-reports, external audits)�
    • What are the pressing gaps to bridge in your community?

​

Youth Shelters

Resources for Veterans

State Commitments

Permanent Housing Trends

Permanent housing placement 42% �Permanent housing retention 96%�Successful exit 36% �Recidivism (24mo) 19%�Median stay 29 days

​

​

Decadal change

​

Annual statistics

8 of 8

Register + share data you use

housing.pubpub.org.� see also: oshi-la.com �

​

Add to the pattern language.

We can travel farther, together.

What you know may complete someone else’s pattern —�

​

Thank you.�

Geeta Dayal

​

​