Building a Safety Net for the 21st Century

Table of Contents:

1- Introduction from Chief Organizing Officer

2- Strategic Framework

3- History of Open Referral

4- Progress to date

4- Technology:

5- Pilot projects:

5- An emerging ecosystem of interoperable tools

8- Resource Data-Driven Advocacy for Systems Change

9- Feedback

9- On HSDS:

10- On Ohana:

10- On Open Referral’s Leadership:

12- Appendices and Thanks

Letter from the Chief Organizing Officer

Dear friends and collaborators,

Together, we’re taking a new approach to a very old problem: as our world grows more complex, it becomes ever harder to keep track of who does what for whom. The resulting chaos — proliferating resource directories, managed in ad hoc silos — makes it hard for people to know where to go to find help. And for decision-makers, it’s hard to know what’s working and what needs remain unmet. [p2] 

Technology now makes it possible to solve this old problem in new ways: data about who does what can flow across boundaries — institutional, technological, etc — which previously contained it. In Internet ages, such technology is already generations old. In the world of non-profits and governments, however, it’s still brand new. This poses a tremendous opportunity. Yet we know that technology is only part of the solution to complex social problems; effective solutions consist of people and ideas. Culture.

In 2014, we created Open Referral to bring all these pieces together. In conversation among a diverse group of smart people, we identified the need and opportunity to develop a common language — so that information systems could ‘speak’ to each other — for the sharing of community resource directory data. In collaboration with the Knight Foundation-funded Ohana Project and Code for America, Open Referral mobilized a network of stakeholders who seek to apply this approach in their communities. [p3]

In 2015, we took coordinated action. We published Version 1.0 of the Human Services Data Specification in March. [p4] Our initial pilots in DC and San Francisco have been joined by projects in Chicago Il, Boston, MA, Ontario, Canada, Madrid, Spain and elsewhere. Software vendors — such as Purple Binder and iCarol — are enabling their systems to ‘speak’ this language. [p6] Interoperable, open source tools are emerging across our network. [p5] People are using resource data to advocate for change in their communities. [p8] Along the way, we’re collecting feedback about what should come next.

Moving forward we must learn from these experiments, and demonstrate the viability of open and interoperable models. We know HSDS must evolve to become both more robust and easier to use. [p9] We’ve learned that stakeholders need ‘many-to-many solutions’ for maintaining accurate resource data. [p10] To realize Open Referral’s full potential, we must build capacity for community organizing, infrastructure engineering, and sustainable institutional development from local to global levels. [p10]

We’ve made great progress already, yet our journey is a long one. Many thanks to all who have brought us this far. Together, we’re building a safety net that’s fit for the 21st century — one in which everyone can find reliable information about the helpful resources available in their community.

Onwards,

Greg Bloom

Chief Organizing Officer

Strategic Framework

The Problem: A landscape of siloed directories

It’s hard to see the safety net. Which agencies provide what services to whom? Where and how can people access them? These details are always in flux. Nonprofit and government agencies are often under-resourced and overwhelmed, and it may not be a priority for them to push information out to attract more customers.

So there are many ‘referral services’ — such as call centers, resource directories, and web applications — that collect directory information about health, human, and social services. However, these directories are all locked in fragmented and redundant silos. As a result of this costly and ineffective status quo:

• People in need have difficulty discovering and accessing services that can help them live better lives.
• Service providers struggle to connect clients with other services that can help meet complex needs.
• Decision-makers are unable to gauge the effectiveness of programs at improving community health.
• Innovators are stymied by lack of access to data that could power valuable tools for any of the above.  

The Opportunity: Enabling systems to ‘talk’ to each other

Many attempts to build centralized solutions have come and gone, and new apps emerge all the time. However, if the many different kinds of ‘community resource databases’ could all recognize a common ‘language,’ then resource directory records could be published once and accessed in many ways (such as call centers, web engines like Google and Yelp, mobile apps, or other channels).

In collaboration with Code for America, Google.org, the Alliance of Information and Referral Systems and others, Open Referral has just made this possible. Our data format can establish interoperability between conventional systems, new systems, and the Web itself.^[1] Now we have to make it easy.

The Plan: local pilots to shift the global market

In pilot projects, lead stakeholders — consisting of government champions, referral providers (2-1-1s, startups, etc), and community anchor institutions —- will collaborate to establish open interoperable resource data infrastructure, tested by specific implementations that yield tangible value to service providers and people in need. Through the pilot, stakeholders will evaluate results and plan for long-term sustainability. The Open Referral team will a) develop tools that make it easy to produce, validate, and publish standardized data, b) organize our community of practice, c) identify best practices and viable business models, d) plan for post-pilot governance, and e) scale the solution.

The impacts: a cascade of benefits from interoperable, open resource data

As various institutions adopt open standards and open platforms, we expect the following outcomes:

• Higher quality resource data can be produced at a lower cost than siloed directories of status quo;

• Innovative tools and applications can proliferate, and become easier to re-deploy and adapt;

• People can more easily find services, and service providers can more readily meet complex needs;

• Researchers, policy-makers and funders can better understand public health and community needs.

• All of this can yield healthier people and more resilient communities.

The History of Open Referral

Open Referral was instigated by the DC Open211 project, and launched in 2014 with co-sponsorship from Code for America (CfA) and the Ohana project. It is now an independent community of practice.

DC Open211

Between 2009 and 2012, the DC Open211 project emerged from collaboration among Bread for the City, DC’s Department of Human Services, and other local institutions that produce or wish to use community resource data. These stakeholders crafted a vision in which various organizations could share the same data for use in various interfaces. In 2013, in partnership with Code for DC, the DC Open211 project experimented with merging various directories and loading them into CfA’s new Ohana API, while exploring long-term paths to integration with local 2-1-1 systems.

Code for America and Ohana

In CfA’s fellowship program, teams of technologists work with local governments to address specific challenges. In 2013, CfA fellows worked with the San Mateo County Human Services Agency to enhance the value of its official (printed) resource directory. The team developed an ‘Application Programming Interface’ (API) that enables San Mateo’s community resource data to be accessed by external applications. The Knight Foundation awarded the 2013 Health Challenge to this team to redevelop its API (named Ohana, Hawaiian for ‘family’) into an open source resource directory platform.

Google.org and Schema.org

In the summer of 2013, a Google.org team proposed a "civic services schema" for inclusion in the Schema.org project. The World Wide Web Consortium (W3C) approved this schema, instantly making it a de facto standard that is recognized by search engines and web platforms.

Open Eligibility

One of the major barriers in this field is the challenge of describing types of services in a coherent and precise way. In 2012, Aunt Bertha published Open Eligibility, an open source taxonomy of types of human services and situations. Open Eligibility is in use by various systems, including Ohana.

Publication of Towards a Community Data Commons

In 2013, Greg Bloom (instigator and facilitator of the DC Open211 project) published the essay ‘Towards a Community Data Commons’ in Code for America’s Beyond Transparency. The essay ignited a broader conversation about interoperability and openness in the field of ‘information and referral.’

Formation of Open Referral

In 2014, with co-sponsorship from Code for America, the Ohana team joined DC Open211 to initiate Open Referral, along with representatives of the Alliance of Information and Referral Systems, Google, various providers of service referral products, and experts on interoperability of human service informatics. Open Referral publicly launched in March of 2014. Its first workshop was in July of 2014.

Moving forward

The initial 2014 cycle was a success: the Ohana team finished up its work in January of 2015 and the Human Services Data Specification (v1.0) was published in March. Open Referral is now an independent community of practice, supporting the implementation of HSDS and evaluating its performance in anticipation of iterative cycles towards v2.0.

Progress to date

Technology: designing the building blocks for open data ecosystems

Open Referral is stewarding a set of tools that can help communities build open ecosystems around interoperable resource directory data. With a ‘universal data exchange format,’ and open source API software that works with it, communities can now experiment with new ways of keeping track of who does what, and new means of helping people meet their needs.

Published Data Specification

The Human Service Data Specification (HSDS) is a format for publishing machine-readable data about health, human, and social services — including their organizational provision and local accessibility. HSDS is designed to enable interoperability among heterogeneous information systems — including both 211/AIRS-compliant call centers and W3C/Schema.org-compliant web platforms (such as Google, Yelp, Facebook, etc). HSDS V1.0 was drafted by Sophia Parafina and published in March of 2015, after nine iterations of public comment and testing.

Completed the Ohana Project

In March of 2015, the Ohana Project formally concluded its work (funded by the 2013 Knight Foundation Health Data Challenge). Ohana’s products include the Ohana API, which enables any resource directory to be transformed into an open platform, and the Ohana Web Search, a free open source mobile-friendly front-end application that enables basic searching for services.  

Ohana also features a basic data administration interface which enables individual users to input information for individual records. The admin system also enables organizational representatives to edit their own records, and a ‘Super Admin’ to edit any record. However, Ohana does not yet offer functionality such as robust permissions, source information, version control, taxonomy management, workflow tools for reconciliation and validation, and other critical features identified by stakeholders. Ohana is partially compatible with HSDS 1.0.^[2]

Progress to date (continued)

Fostered an emerging ecosystem of interoperable tools

In addition to Ohana, we’ve seen the emergence of new, interoperable tools that help people find services.

For example, Link-SF is a mobile resource locator initially developed by Zendesk for St Anthony’s in San Francisco, and is now being adapted to ‘speak’ the Open Referral format. It has already been redeployed in other localities such as Queens.

In D.C., Social Impact Lab — in partnership with the DC Public Library — has prototyped a ‘Logic Library’ that enables users to author conditional logic for Q+A interfaces, so that screening tools can easily generate targeted queries of open resource directory data.

Pilot projects:

The real work of Open Referral happens “on the ground.” Our data specification, APIs, and other tools are only valuable inasmuch as they are used by people to solve problems.

So our initiative is centered around pilot projects, which are themselves centered around stakeholders — i.e. people and institutions who need access to resource data on a regular basis. These stakeholders agree to utilize HSDS, for the purpose of exchanging resource directory data between existing and emerging systems, in ways that help them achieve their goals. In return, their insights guide the development of our specification and our tools.

Our first pilot projects — San Francisco and DC — saw mixed results in their first year. Key institutions have made ambitious pledges, and small projects have shown transformative potential. However, these pilots have not yet cohered around fully resourced teams with clearly established processes. In part, this is because the pilots have acquired funding in modest, narrowly-scoped grants, which limited the depth and rhythm of work.

In the meantime, a number of other communities have joined the Open Referral initiative with steps that range from minor to major — despite in most cases being self-funded or volunteer-driven. Most projects are summarized below (yet several others are not yet at a point of public communication).

In 2016, Open Referral’s leadership will focus on a small number of pilots — no more than three, possibly just one — to help bring them to a point of maturity.

Pilot Projects (continued)...

San Francisco

Bay Area strategy brief; stakeholder outline

Alameda:

Lead partners: Open Oakland / Urban Strategies Council (Steve Spiker), Eden I&R (2-1-1 for Alameda County), First 5 Alameda

Project: Upgrade Eden I&R’s resource database system (FileMaker Pro) into modernized Open211 platform (iCarol). Enable First 5 (and other institutions) to publish updates to Eden, and Eden’s output republished to Alameda county’s open data portal.

Status: On hold, pending Eden I&R system upgrade.

Desired outcome: Open platform for county’s health, human, social service directory data.

Blog: openreferral.org/building-alameda-countys-first-public-health-human-services-data-portal/ 

San Francisco:

Leads: Code for San Francisco Open Referral team (April Steed)

Stakeholders: San Francisco Adult Probation Department, Reentry Council of San Francisco

Project: Upgrade 300 page (MS Word) Guide to Getting out and Staying Out to structured database with export feature.

Status: Completing prototype

Desired outcome: [short-term] decrease costs of maintaining data and increase range of uses and the effectiveness thereof. [long-term] collaborative network of gov agencies and referral providers sharing data (possibly through publish/subscribe hub model).

Blog: https://openreferral.org/san-francisco-open-referral-getting-out-and-staying-out/ 

San Mateo

Contacts: Edwin Chan [echan@co.sanmateo.ca.us], San Mateo Human Services Agency; Moncef Belyamani, Ohana Project

Stakeholders: San Mateo Human Services Agency

Project: Initial Ohana deployment.

Status: Deployment is live. Project is in stasis.

Desired outcome: Ohana as sustainable centralized platform for service directory data in SMC

District of Columbia:  

Contacts: Justin Grimes at Code for DC; Keith Porcaro at Social Impact Lab; Matt Bailey, director of innovation at the Office of the Chief Technology Officer; Susie Cambria at Deputy Mayor of Health and Human Services; Laurin Hodge at Mission:Launch; Joe Cureton at Advoc8

Stakeholders: DC Public Library, Bread for the City, DC Office on Aging, DC Department of Human Services, DC Public Defender Service.

Projects: 2-1-1 Improvement project (DC Gov); Logic Library (SIMLab, DC Public Library);

Pilot Projects (continued)...

Bread for the City platform (Salesforce); Reentry Resource Manual (DC Public Defender Service, MissionLaunch)

Desired outcome: network of interoperable referral systems sharing a harmonized core of open data (possibly via publish/subscribe hub model)

Status: Ohana demo deployed. On hold pending gov action, funding for community partners.

Blog: https://openreferral.org/first-resort-building-social-services-resources-at-dc-public-library/

http://codefordc.org/blog/2014/11/18/dc-open211-at-the-rebuilding-reentry-hackathon.html  

Strategy memo: here.

NYC:  

Lead Partners: Civic Hall [Erin Simpson and Elizabeth Stewart; Micah Sifry and Andrew Rasiej]

Contacts: Sarapis (Devin Balkind), BetaNYC (Noel Hidalgo)

Stakeholders: NYC 3-1-1, ACCESS NYC, NY Public Library, Greater New York Hospital Association (HITE Site), NYC VOADs,

Projects: NYC Open211. Also, NYC:Prepared post-Sandy resource directory platform.

Desired outcome: Open civic infrastructure for all of NYC’s human service directory data.

Status: pending funding.

Blog: openreferral.org/preparing-for-the-worst-hoping-for-the-best-data-standards-superstorm-sandy-and-our-resilient-future/ 

Memo: NYC stakeholder memo.

Chicago:

Leads: Purple Binder

Community partner: Dan X O’Neill, Smart Chicago Collaborative

Projects: Purple Binder API -> Ohana Web Search, mRelief integration, Chicago Health Atlas

Stakeholders: No designated community-based stakeholders (direct service providers or representatives of help-seekers)

Desired outcome: demonstrate viability of private open platform for service directory data

Status: Live API with deployments. However, not yet an ‘open API’ business model.

Blog: https://openreferral.org/open-referral-action-purple-binder-platform/ 

Memo: Chicago Open Referral memo

Boston:

Leads: Noah Teshu, Boston Children’s Hospital; Aaron Pikcilingis, independent

Project: Helpsteps conversion to Ohana / Open Referral

Stakeholders: Boston Children’s Hospital, Boston Public Health Commission, Allston Brighton Health Collaborative.

Desired outcome: demonstrate Helpsteps as sustainable open platform for Boston ecosystem.

Status: Prototype approaching completion.

Blog: https://openreferral.org/helping-helpsteps-step-forward/ 

Pilot Projects (continued)...

Miami, Florida:

Community Partners: Code for Miami (Ed Toro, Rebekah Monson, Ernie Hsiung)

Lead Stakeholders: Switchboard of Miami, the Children’s Trust

Projects: Ohana demo [here; project page].

Desired outcome: Evolution of Switchboard (211) into a sustainable open platform supporting an ecosystem of referral services and other resource data applications throughout Miami-Dade

Status: Demo deployed. Funding requested for next phase of development.

Blog: https://openreferral.org/open-referral-all-over-hacks-at-code-across-and-open-data-day/ 

Memo: Miami Stakeholder memo

Ontario, Canada:  

Lead Partners: PoweredByData (Michael Lenczner), Ontario 2-1-1 (Andrew Benson), Kate Lambacher (KCL Software / CIOC)

Projects: Ontario Open211, Open CIOC

Stakeholders: provincial and municipal governments, local non-profit service providers

Desired outcome: by publishing in HSDS format (via iCarol), facilitate greater and more effective use of Ontario’s resource directory data.

Status: Prototypes under development.

Blog: https://openreferral.org/new-partnerships-up-north-open-referral-in-canada/ 

Memo: Ontario Open Referral Memo; CIOC Open Data Presentation

Madrid, Spain

Leads: Medialab-Prado (Adolfo Bravo), Amanda Sorribes

Projects: SociaSalud - visualization and search of Madrid’s open directory data assets

Stakeholders:  Fundación San Martin de Porres, Madrid City Council

Desired outcome: open, interoperable data on all human services (especially for elders, migrants, people with disabilities, people experiencing homelessness)

Status: Demo deployed. Pending additional resources.

Blog: https://openreferral.org/huertas-de-datos-open-referral-in-madrid 

Memo: Open Referral Madrid

Advocacy for Systems Change

Resource data isn’t just useful to help people find where to go to get help. It’s also critical for assessing the effectiveness of our safety net as a whole. Towards this end, the community champion of the DC Open211 project — Bread for the City — has already achieved an illuminating data-driven success.

Advocacy (continued)...

Bread for the City told this story on their blog — and it was also reported at the Huffington Post. By enabling smarter use of resource directory data, Bread’s social workers were able to successfully advocate for systemic change to their local safety net. Specifically, food pantries in DC have long required prospective clients to take an extra trip to a third-party organization in order to get a signed document referring them back to the pantry. In other words, Bread for the City had to write permission slips for thousands of DC residents, vouching that they deserved food assistance. Bread had long felt this to be a wasteful and degrading policy — and now they could generate hard data about it. Bread for the City traced almost all of these referrals back to the same 14 food pantries. By presenting hard numbers about the costs in money and time that these policies inflicted upon low-income DC residents, Bread for the City successfully advocated for this policy to be changed. The result is tangible improvements in food accessibility for thousands of DC residents, a victory that Bread for the City describes as ‘racial justice.’

Feedback on our work so far

(See ‘Progress to Date’ for background.) Over the course of the year, we’ve gathered both structured and unstructured feedback from pilot stakeholders on the effectiveness and challenges of HSDS, Ohana, and our leadership. See raw feedback on HSDS and Ohana here.

On HSDS: technically adept (for a first version)

What's good: We’ve received positive feedback from enterprise IT (especially regarding its approach to normalizing the relationship between services, sites, and organizational entities). Several vendors have used it as a model for structuring new products, in addition to data exchange.

What needs improvement: Feedback indicates that HSDS falls short of one of our initial principles of simplicity. The specification is daunting for relatively non-technical users, and this is a serious drawback. For example, managing HSDS-compliant resource data in a spreadsheet would be prohibitively difficult unless additional tools are developed to ease the process.

What we can change: In future iterations, we might consider a dual approach to formatting flat<>structured data, given precedents set by the Open Contracting Data Standard and the Cove tool developed in part through its implementation. I’ve also proposed adopting a ‘Share Alike’ license that will encourage adaptation, while preserving the open source nature of new contributions — thereby balancing the values of innovation and interoperability across an emerging ecosystem. [Read more about this licensing proposal here.] Additional issues regarding HSDS have been logged in the HSDS Github repo. Future development of HSDS will require the re-establishment of technical leadership in some form.

On Ohana: valuable components of an open ecosystem

What's good: We’ve received positive feedback that Ohana is easy enough to redeploy, with clean code and clear documentation. Several service locator sites have been deployed with pre-existing resource data presented in new ways, alongside APIs for developers to build other apps.

What needs improvement: We have received common feedback that the current incarnation of Ohana doesn’t address the challenges of directory data maintenance. Ohana’s admin interface is rudimentary and unwieldy for users. More importantly, Ohana doesn’t assist users in understanding and improving data quality (such as easily identifying out-of-date and incomplete records). It does not collect feedback from external apps about data quality; furthermore, users have requested workflow tools to enable more reliable verification of accuracy of records. Users have also identified the need for more flexible categorical management. [See requested features for Ohana here.]

What we can change: Additional requests for features have been logged here. Ohana could evolve to meet these needs, or something new could emerge to take its place. This is up for us to decide. This decision will depend upon the interests and resources of existing users, as well as the decisions made by future pilot projects as to the best way to meet their needs. We invite those who have deployed Ohana and/or have experience working with it to join us in this deliberation.

On Open Referral’s leadership: cultivating a community of practice

What's good: A diverse range of people have joined the Open Referral network, and/or influenced our work. This poses an opportunity to learn from government officials, data scientists, civic technologists, non-profit executives, community leaders, and more. We’ve only been able to tell some stories from the range of inspiring work underway — and I’m looking forward to many more stories to come.

We’ve also set a precedent for achieving understanding and compromise around thorny technical issues. For example, approval of HSDS v1.0 hinged upon a design decision that ignited a month-long debate (about data normalization vs denormalization). This underscored the tensions inherent in any effort to construct a useful model out of a complex reality. Our objective was not to ascertain The Truth, so much as hear perspectives from various sides and seek an outcome which would at least be acceptable to involved parties. I believe we rose to meet such technical and diplomatic challenges on a number of occasions.

What needs improvement: While working to build trust — especially in partnership with people who are themselves under various kinds of institutional pressures and constraints — it’s hard to strike the right balance between transparency and discretion. Though we’ve taken care to ensure that our network is able to participate in all deliberations and decisions about itself, much deliberation around partnerships and projects has happened out of view. Perhaps this is

Leadership (continued)...

appropriate, given the realities of institutional politics. However, while striving to sustain this delicate balance, I’ve often allowed long periods of silence in Open Referral’s public channels of communication. In 2016, we should strive to improve the flow of information across our network.

We will also need to re-establish technical leadership capacity which can lead the process of reiteration towards HSDS v2.0, while developing open source tools that make it much easier to produce and circulate of reliable resource data. (It’s worth noting that this need for technical leadership is shared by almost every civic data standardization initiative — such as Open311, Open Contracting Data Standard, BLDS, etc — and there has been some promising discussion of cooperation across initiatives. There is also a great need for education of funders and governments around the significant costs and tremendous benefits of open standardization.)

What we can change: In 2016, we should establish a regular rhythm for open discourse, while limiting demands on partners’ political capital and time. I’d welcome suggestions about how to design more efficient, regular, and valuable forms of open engagement.

Some of these improvements might come through choosing and using more appropriate tools: such as cultivating our own Github repository, starting our own Slack team, even migrating off of Google Groups and into a more sophisticated discussion platform like Discourse.

More strategically, though I’ll always be ready to engage in some kind of dialogue with people who are working to expand access to community resource data in their community, I should also aim in 2016 to narrow the range of my own commitments to a very limited set of engagements — with the lion’s share invested in no more than three projects, and possibly fewer.

On a related note, over the course of this year, I’ve provided much advice and facilitation for institutions who are exploring new solutions to their resource directory needs. Much of this work to date has been pro bono. Now, having demonstrated cost savings and other benefits of open source and/or interoperability, I will seek to provide such assistance through paid consulting services. This will help make my work and Open Referral overall more sustainable — while I also recognize the need to avoid conflicts of interest and to be as transparent as possible about such relationships. I’d welcome suggestions for effective methods for maintaining this balance.

Finally, and most importantly, Open Referral is seeking sponsorship for expanded development capacities. Code for America and the Ohana Project’s initial support were catalytic in getting this project off the ground, yet from the start we knew that these were temporary steps in a long-term process. With the help of terrific advisors, I’m in the process of soliciting support from major foundations, as well as exploring potentially sustainable open source business models. Along the way we’d welcome advisement, leads, and any other assistance in the process of building capacity for this work. Please be in touch.

Appendices

Executive Brief

Pilot projects framework

User Personas

Overview presentation

Assemblies: recordings and notes

Collected feedback on HSDS [Google Docs folder; Github repo]

Bloomberg Data for Good Exchange paper

Many Thanks to Give:

Open Referral has been made possible by the contributions of so many people, it would be unwieldy to list them all. At the risk of skipping essential contributions from wonderful collaborators, here (in no particular order) are some of the many who have provided insight, funding, time, couches, and other resources that have helped us make it this far: