Interoperability enables information systems to do more with the same data — which can change lives.

An in-depth look at Open Referral with Greg Bloom

We might wish that we could all agree to use one system, but the reality is that organizations have different needs for the same information.

Greg Bloom is the founder of Open Referral, which is promoting open access to resource directory data (i.e. information about the health, human, and social services available to people in need). Previously, Greg managed communications for Bread for the City, the District of Columbia’s preeminent anti-poverty service provider. He is a certified cooperative developer and a dedicated community organizer, with more than a decade of experience in GOTV, class-action labor lawsuits, municipal budget battles, death penalty abolition campaigns, community wireless networks, and even a backyard-chicken legalization movement. He has served as a fellow with Provisions Library and Civic Hall Labs, and a resident of the Elsewhere Museum. He has also published writing in In These Times, Civic Quarterly, Personal Democracy Forum, and Code for America’s Beyond Transparency.

The following is a conversation between Daniel Ramirez-Raftree and Greg Bloom, reflecting on his work with Open Referral. Read it on StoryEngine site as well.

Greg’s story

Could you start by telling me about yourself, your career, and what you learned when you worked at Bread for the City?

Professionally, I’ve worked in fundraising and communications for various nonprofit organizations leading up Bread for the City. Beyond that professional role, avocationally, I’ve been a community organizer in many contexts, something I’ve gravitated to instinctively. I enjoy finding opportunities for people to get together to build relationships and learn from each other.

This is why at Bread for the City, on top of my job as a fundraising/communications person, I played an unofficial role finding issues around which people wanted to connect and creating spaces for them to come together to learn, form relationships, and build power. The project I’m currently working on — Open Referral — actually emerged from a series of side projects like this at Bread for the City.

Can you describe Open Referral and its mission?

Open Referral’s vision is a future filled with accessible information about the resources that are available to people in need — resources like health, food, social services, and legal assistance. We want to see a world in which this information is easy to find and use, so that anyone who is looking for essential resources is able to get reliable, accurate information about them.

To that end, our mission is to make it easier for the organizations that maintain health, human, and social services directories to produce information and cooperate with each other to publish it.

Specifically, we’ve developed data exchange protocols that enable interoperability among the information systems that keep track of what resources exist and how people can access them. Furthermore, we’ve also organized a community of practice. This community is made up of people in various organizations that use our protocols and tools to manage their information more effectively.

Can you describe what it’s like for people encountering out-of-date information about critical resources? In other words, what’s it like living with a system of inaccurate and inaccessible information?

It’s a story that we saw playing out over and over again at Bread for the City. We offered dozens of different services onsite — and on top of that, one of the most important things that Bread for the City did was provide information about other services people needed. Even though people might have come to Bread for the City specifically for food, if they were also wondering where they could find services like childcare or behavioral health counseling, Bread for the City was able to point them in the right direction because it kept track of these other resources using a database that it built for itself.

And Bread for the City had to build the database itself because there was no other way to get that information reliably. Oftentimes, the information is just not out there, even if you know what you’re looking for, which much of the time people don’t.

When they do exist, the directories that are designed to provide this information have serious shortcomings. For example, some are designed to be used exclusively by phone, others are only made available through a poorly-made website, and still others have information that’s simply not that useful. Right now, if you want useful information, you have to collect it yourself, and so people would come to Bread for the City because they knew that we had good information.

It was important that we could provide this information, but it didn’t come cheaply. To maintain the database, social workers would, on top of everything else they had to do, spend hours throughout the week calling different organizations to compile the information. Then they would take more hours out of their Friday afternoons to compile their notes and enter all that information into the database.

Then, when other organizations wanted to collaborate, that created even more work. For example, a food justice coalition reached out to us asking to partner with them to build a website that would connect people to food resources across the city. We agreed to help. Flash forward three months and now there was a new website for food resources that we were updating at the same time that we were updating our own internal database, which meant we had to copy and paste the same information into two different systems. So in trying to provide more uses for the information, we ended up making more work for ourselves.

So this state of affairs is bad for the people who are looking for services, and it’s bad for the people providing services. If somebody needs food assistance, they probably need several kinds of assistance, and if a service provider isn’t able to solve all of a service seeker’s problems with whatever it is they do, then they need better information to be able to effectively help the service seeker, and then that becomes their job to chase down and maintain this information.

Then you also have all these people who want to try and solve the problem by building better tools, but everyone who tries ends up having to start from scratch. Then, all of these attempts at solving the problem end up competing with each other over dominance or funding.

It’s just a big mess.

You mention in one of your talks that attempts to create a centralized database for information have failed. Can you describe that problem and explain why interoperability is a solution?

Let’s bring it back to our experience at Bread for the City. It maintains its own internal database, then we have a website for the DC Food Finder, and then there’s a database of homelessness services that another organization maintains — and as we’re getting together to work out a better process, inevitably the first thing someone suggests is creating a centralized system that everybody could use. As it turns out, people had tried to build such a system roughly every five years starting about 15 or 20 years ago. So when that idea comes up, the people who had been in the room for 10 or 15 years are like, “oh boy, here we go.”

This is the way it is. When thinking about this problem, the first thing that comes to your head is to build a centralized system, but that’s not actually how people or organizations like to behave. We might wish that we could all agree to use one system, but the reality is that organizations have different needs for the same information, or they need slightly different versions of the same information. Some organizations need information about services to be really simply described and understandable for people with low literacy levels. Others need information about services to be really finely detailed and to have very precise technical terminologies so that they can be sure they’re making the right referrals to the right places. One information system is not going to serve all these needs effectively.

On top of that, the reality is that there’s also political struggle. If there’s going to be one centralized information system, who will own that information system? That becomes a complicated political question that has potentially unhealthy implications for a community.

As we learned this history about attempting to create a centralized system, I was also learning about modern data sharing technology. We now have the ability to use the same data in many different systems simultaneously. In fact, the technology is generations old in internet years. It’s like 15 or 20 years old.

With this understanding in mind, we envisioned that rather than trying the old failed model of building one system and trying to get everybody to use it by hook or by crook, we would instead get all of the existing systems and all the emerging systems to be able to talk to each other. We sought to find cooperative solutions that would enable people to use different systems to meet their own particular needs while still making the same information accessible to many users.

It’s a little bit of a paradigm shift, even though the goal is still the same. We want everybody to have access to the same information, but we don’t live in a centrally planned society where there’s a single governing mechanism. That’s just not how we live. So we need tools that reflect the reality of how we actually live.

There’s a comic you use in one of your talks to frame the debate around creating a universal standard. Can you talk to me about that comic?

The comic is this XKCD cartoon called Universal Standards. It has three panels. The first is captioned, “Situation: there’s 14 competing standards.” In the second panel, this guy says “this is ridiculous! we need one standard that meets everybody’s use cases.” Then in the third panel you see, “Situation: there’s 15 competing standards.”

I’ve found that when talking with a group of technical people about problems that could be solved through standards, the likelihood that someone will bring up this comic approaches 1. And techies — men, especially — tend to bring it up as a reason for why you shouldn’t bother trying to use standards at all. They think this cleverly shows the whole concept is naive. But that’s only because this thinking stops with the recognition that standards pose a collective action problem: IF everybody cooperated, everybody would be better off — but, people have these incentives to compete, so everyone tends to suffer. And it’s true, collective action problems are hard; but the techbros tend to assume that they’re totally unsolvable, or only solvable if someone wins outright and achieves cooperation through domination. And if collective action problems are inherently unsolvable, then attempts to solve them might actually make things worse.

In Open Referral, we read that comic differently. There aren’t many words so it’s not too much to ask to read slowly: it’s saying that there are 14 competing standards, and then someone comes along and is like “I’m going to solve the problem for everyone” — in other words, his idea is to beat out all of those other standards. This obviously yields… 15 competing standards. But what if the new standard was designed to enable cooperation among other standards? That’s a different story. That potentially leads to a collective action solution. There might be 14 competing standards, but if you develop a standard for the purpose of translating across standards — not to solve the problem for everyone, but to enable different solutions to work in harmony with each other — then you don’t have to create a standard that everybody uses, you just have to be willing and able to cooperate with any standard that anybody uses.

That kind of thing is hard work. It’s not sexy. It entails a whole lot of process, a whole lot of relationship building, a whole lot of translation. Trust and accountability. And that’s work that many people, especially technical men, just don’t want to do; they want to build apps. Fortune and glory. Well for the rest of us who are really thinking about the whole complexity of the world, this collective action problem might actually be the most important problem for us to solve.

And I think Open Referral is succeeding because of this work developing relationships, practicing cooperation, and shifting paradigms.

Permanent link to this comic: https://xkcd.com/927/

How do you get people to start using your standard?

I bring a community organizer’s method to it, meaning that if I’m talking to someone and it sounds like they are struggling with something related to this problem — the problem of resource directory data living in silos, proving unsustainable, creating unhealthy competitive dynamics — then I’ll keep talking with them and keep asking them questions, to see if we’re able to identify that there’s a shared interest. My goal is to get them to see that they have a shared interest with many other people, and that other people are cooperating, and that it makes sense for them to cooperate, in tangible ways, even if they’re skeptical about it at first.

They might hesitate, since nobody’s really excited about taking on a whole new branch of work that is often only indirectly related to their primary objectives; it takes work to adopt a standard, one way or another you’re agreeing to do things at least a bit differently.

So we keep talking, and I see if we can identify additional reasons that cooperation might be worthwhile. I try to bring them together with other people who share their interests, to have them see what happens when they talk to and learn from each other. Once we’ve identified that, in principle, this is actually something that could benefit them and their whole community, we work to identify the most actionable step in the direction of that better world that we want to live in together. We focus on the first ‘minimally viable’ version of that cooperation, the thing that we could actually do now that would be valuable now, while demonstrating what it is that we’re trying to do in the longer term. Even if this is just in a small way, it helps to generate more buy-in from more stakeholders, which in turn, generates more resources from funders. And this is how we set in motion the more and more complex undertaking of our cooperative value proposition.

In the end, it’s really a process of talking to people, writing down what they say, and making sure that what we write helps people understand that they do have shared interests and can act together in the direction of those shared interests.

Sometimes the barriers are really technical, and these can be simple to fix. For example, some organizations need a website for resource directory data. Before us, they’d go through a whole from-scratch design process, it might cost a quarter to a half million dollars or more, and then they’d have to figure out how to maintain the site and the data. Now there are websites out there that use our format, are already built, and are open source — they can be redeployed freely. By using these, organizations could have a website at a tenth of the cost of what it would take to build one from scratch. With that problem solved, they just have to think about the supply of data. And then we can help expand the range of options for solving that problem.

Other times it’s actually much more complex. It could be the case that a funder is asking an organization to give them spreadsheets that are color coded and hand-maintained and matched to their data and so on and so on. It could take hundreds of hours of staff time to get this done, and our challenge is figuring out how to do that automatically. This is Excel files, not a whole website, but it’s actually a more complex problem and there are various ways it could be solved.

I don’t go into any conversation acting as if I have the perfect solution to all problems and that people should just sign on the dotted line and buy our product. We have to actually figure out what’s going on and in what specific way this method is actually going to help a client.

Can you talk to me about the difference between approaching this problem with a solution that’s a commodity versus a solution that’s a public good?

I understand a commodity to be something that a person or organization collects or makes and then turns around and sells. It’s an object that first needs to be produced, and then becomes property that’s transferred to someone for money.

This model doesn’t actually make that much sense in a digital world, because once data is aggregated, anybody can use it, and they can use it over and over again without it being depleted — unlike commodities like corn or pork bellies — so what’s the logic in someone paying for the data? Someone uses it, but it’s still there; if anything, its value increases the more it’s used. So the commodity concept doesn’t really make sense here.

And yet in the current world, especially before Open Referral, that’s how human service directory data is generally treated, like a commodity. It costs organizations money and time to aggregate the information. They have to make thousands of phone calls, conduct research, and enter data to produce a service directory. So when other organizations come and ask for information, the organization that produced it responds, “it costs us money to make it, what are you going to pay us for it?” Nevertheless, that conversation doesn’t tend to go well; people just go off and create their own database. So we end up with a proliferation of redundant, wasteful, and ineffective databases; the commodity doesn’t really get exchanged, instead everyone loses out.

The shift that we’re trying to help this field undertake is from a commodity-based approach to a public-good-based approach. It gets complicated because though it is public information, the government isn’t collecting and maintaining it, like it would with a park or other public goods. Maybe the government should produce it, and Open Referral is exploring that in a few ways. But unless and until that transformation of government’s role happens, organizations have to do it themselves if they want this information, and they currently have this mentality that once they’ve done the work of collecting the data, somebody else should pay to access it if they want to use it in some other way.

Normatively, things should obviously be different: these services are provided by the government, or funded by tax-deductible donations, so information about them is public information, it should be treated as a public good. But it’s not enough for us to wag our finger and say, “This is public information, you shouldn’t be selling it.” We’re trying to help the organizations understand that it’s not only normatively, morally correct to treat this as a public good — but that it’s also economically strategic. We have to actually show that it’s better for organizations at a fiscal level, as a matter of strategy, for them to consider the information to be free. We need to demonstrate that, when organizations spend all this time and energy collecting this information, it’s really in their best interests to just let others have at it.

At a technical level, what Open Referral has done is simply remove some of the barriers to sharing information, to provide tools that reduce the friction of data exchange. That is our way of getting our foot in the door. But the more complex and important conversations we have are about how to build better business models around this information, moving from property-based scarcity mentality to an infrastructural generative mentality. How to show that information producers will benefit from others using their information, even if some of those users might not pay for it. We have to figure out what those users really will pay for, other than the data itself, and we need to figure out what incentives need to be in place to ensure that the people who are producing the public good don’t lose out. A lot of our work is figuring out how to realign these incentives. And it’s not magical. It takes time and deliberation and a commitment to cooperation. But the win-win scenarios are out there.

How challenging is it creating a business case for why these organizations should switch to the public good model?

Let me give you the simple version of that case. Let’s say you are maintaining a directory of all the resources available in your community. My message to you at the most abstract is that your data is more valuable if more people use it. The benefits of that show up in ways that are both hard and soft. I say soft in the sense that it’s good for you and the community to have more people using the data that you maintain — it makes you look good if more people are using your data. You can require people to give you credit when they use your data for free, which is good for your brand. Also, let’s say 15 organizations are using your data, you can claim you’re helping 15 organizations do their job. This seems like it should be a win even if no money changes hands — if you can communicate about the scale of that impact, leverage those relationships, etc.

That said, we’re really looking for the hard incentives. If an organization is collecting all this information and making it available for free, there are things — services, features, etc — for which other organizations will still want to pay. For example, they’ll want to pay for a certain degree of confidence that the data will never go away, that it will be available for the duration of a contract. That guarantee enables a third party organization to build their operations around data that is already available for free. Or, maybe an organization makes all its data available for free, but a third party only wants to get the specific subsets of data that are relevant to the people they serve — elders, youth, people returning from prison, etc. These users will all need different slices of the whole data pie, and they’ll want those slices served up in different ways. Organizations that collect data can help other organizations use the data for particular purposes — curation, filtering, etc. — and they can charge for that service.

We’re helping this field figure out that they can shift to a public good model by offering services that add value to their data. The data should be free, but they can then recover the costs of maintaining that data by selling their expertise and other values that they are uniquely positioned to add.

Does this approach produce new revenue from donations or grants? Can it reduce costs for organizations?

All of the above. First of all, data collection is a cost center for you, and those costs could come down in various ways. For example, I work with the 211 hotline for Miami. 211 is one of the main incumbent players in this field, it’s usually a nonprofit even though it has an official number (like 911 is for emergencies and 311 is a government complaint line). 211 is the hotline for community resources. It’s responsible for keeping track of all of a community’s resources, a responsibility that, for an average metropolitan area, usually requires the nonprofit to hire two full time employees to make phone calls to keep track of all of the community’s organizations and their services. That’s a substantial investment just in maintenance.

I’m also working with Legal Services of Greater Miami. They keep track of all the legal aid services. Here’s an organization in a specific domain that’s keeping track of dozens of services that 211 also needs to keep track of. However, the legal aid organization is much closer to the information about legal services than 211 is, because they specialize in that, those providers are their peers, they better understand how to maintain that information. They could share that information with 211, and 211 could reduce its cost of maintenance.

Then, once they are sharing information, they can form other kinds of partnerships. Now that Legal Aid is partnering and sharing its information with 211, Legal Aid might also be able to ensure that its own clients have access to other social services using data provided by 211. This can happen much more easily once their systems start to talk to each other.

Additionally, this might actually be a fundable partnership. Now that there’s cooperation between these two systems on one level, they can readily establish cooperation on other levels. They can more effectively coordinate care for people across these systems. And some of that cooperation can be funded and generate revenue. This is just an example of how the logic of cooperation, once it’s introduced, creates more logic for more cooperation.

Here’s another example. Let’s say a 211 operator gets funding from a philanthropy that also provides funding to hundreds of children and family service providers in the same community. That philanthropy is both funding 211 to refer people to services and funding the services themselves. As a result, the philanthropy can say to all the service providers who receive its funding that their information has to be made available on 211. And that has a lot of value.

With Open Referral, we’re able to make it much easier for synergies like this to happen. Not only can the philanthropy now collect information safely and publish it directly to 211, but 211 can now reduce its costs because it’s getting data about services directly from the funder. And beyond that, 211 can now provide data back to that funder about how they’re using that information. The funder can now get valuable data about everyone who’s calling 211 and being referred to one of its grantees. And this is something 211 can charge for. It can get revenue for the insights it generates about how it’s using data about the funder’s own grantees.

There are business models out there, and those business model happen to be built upon the kinds of things that these organizations should be doing anyways — cooperating with each other.

How does the Tragedy of the Commons fit into this?

When I discussed the idea of information as a public good, I also spent a lot of time talking about how we need to find “business models” — and the reason that’s so important is that the problem with a public good is that everybody benefits from it, but nobody wants to or should have to pay for it. The quote-unquote “Tragedy of the Commons” is a simplistic way of describing that phenomenon, where a resource is vulnerable because many individuals want to use it and none of them want to pay the cost of that use.

Now, the problem with the cliche of the Tragedy of the Commons is that the whole theory comes from this one journal article written in the 1960s by this guy Garrett Hardin — a guy with some pretty extreme views — whose primary conclusion that we should force people to get sterilized since the population is getting too big. He draws this really ugly conclusion from this line of reasoning, though that’s almost always forgotten about when the paper is generally cited as common sense.

Essentially, Hardin says, picture a pasture and everybody wants to bring their cows to eat the grass on the pasture, then eventually people will bring too many cows and the pasture will be destroyed. The argument in the Tragedy of the Commons essay is that the destruction of the pasture is inevitable, and that it is not possible for people to figure out a way to avert that unless there’s a government telling people not to bring their cows to the field or unless the whole field belongs to one person who will know (by dint of their enlightened self-interest) how many cows can graze on that field at any one time. The whole essay lays out this collective problem and outright concludes that this problem is unsolvable — “inexorable.”

This is the tragedy of the Tragedy of the Commons. There’s a whole economic paradigm built on the notion that cooperation is not possible. Well, if you step out of your economics think tank and look at how the world actually works, you find that communities across geography and history actually figure out how to solve these problems all the time. They are problems, they are real, and they are hard, but communities figure out how to share resources. They work out arrangements with each other. They set up rules — maybe each person will only be allowed to bring three cows onto the field at a time. Communities have confronted these problems over and over again throughout history and across different cultures. It’s not easy and it doesn’t always work, but we have known patterns and principles for designing systems through which people can share things. (Elinor Ostrom pioneered this field, and won a Nobel Prize in Economics for her work, Governing the Commons.)

That’s all to say that we launched Open Referral to create a space where we can find the specific forms in which these principles for commons governance should take shape. We need to design the institutional arrangements that ensure that this critical information, this public information, which is currently being treated like a commodity in a failed market, is actually going to be provisioned as a public good. We can decide how that should happen, and we have to work together to make those decisions. It might not actually be the same answer in every community, because communities are different, but by using a common protocol, by using a common language, and by having shared principles, we can find solutions that work.

Can you tell me about a significant moment in the history of Open Referral?

There are two kinds of significant moments.

One type of significant moment is when an organization sees that Open Referral is a good thing and decides that they’re going to take action. Or, similarly, if they’ve already taken action, it’s significant for them to talk about the specific ways in which they’ve benefitted from Open Referral.

I spend much of my time talking to organization after organization, and I eventually find the ones that really get it and want to cooperate. These organizations see our logic, and I help them make that logic real by finding the path they can take in order to benefit. And when they tell stories about the value they’ve gotten out of Open Referral, and when other people hear those stories and understand how organizations are cooperating and benefiting from us — those are important moments.

Here’s an example. Early on in Open Referral’s history, the market leader for call center software, iCarol, saw that interoperability is where their field would inevitably be going, and they wanted to be part of it, so they decided to adopt Open Referral. This is the market leader for call center operators using software to manage their call centers, their resource directories, and their websites. They’re used by a third to half of all the call centers that do resource referrals, and having them on-board was a huge moment that took a year of relationship building, conversations, identifying opportunities for them to get funding to support the work, and so on and so forth.

Another example is that the New York City government is using Open Referral to publish data about the services that it contracts. It took them a year to decide to use us as a standardized format, and it took them another year to announce that they’d taken this action and to explain why they did it.

Those were important moments because they led to more, similar moments. Eventually we had enough individual organizations saying, “Yes, we benefit from this,” that the Alliance of Information and Referral Systems (which is the trade association for the entire field of call centers) endorsed Open Referral as an emerging standard for the industry. This is an association that represents 1,200 call centers and sets the standards for its field, and they pointed to us last year to say, “These are emerging standards, and we’re going to adopt these protocols for data exchange into our official standards.” That was the breakthrough moment that we’d been building toward for five years.

In addition to the momentous occasion of our transforming from individual projects to a real standard, the most important moments are really when we get people from competing organizations together to have conversations about these topics. These are so important because everybody sees a different part of this problem. Each person might have more or less ability to describe what that problem looks like to them, and each one might have more or less clarity around what a solution might be, but when they get together and are able to compare different aspects of this problem and listen to each other, everyone gets an enriched sense of the problem and possible solutions. So, really the most important moments are when we convene to talk to each other face to face.

By the way, if you read up on the literature of Tragedy of Commons, like Ostrom’s “Governing the Commons,” the key to building a Non-Tragic Commons is precisely that. It’s face-to-face communication about how the community is going to deal with the cows on this pasture. When that happens, when people are able to get together in-person and listen to each other, that’s when you start finding opportunities to escape that collective action logic problem.

It’s as if the assumption is that human beings are these atomized individuals that don’t speak to each other and exclusively do what’s best for them.

That’s right. Well, that is the American assumption. Our founding myth is basically, “Fuck you, I won’t do what you tell me.” But if we want a better world, we actually have to challenge those foundational notions that we are a nation of rugged individuals who are going to triumph purely through enlightened self-interest without having to be responsible to each other.

Can you tell me what the world looks like once Open Referral has created a federation of interoperable databases.

The future that we want is one where anybody who is in any kind of need can easily look around and find resources that meet those needs. We want a future where the information that directs people to these resources is available at everyone’s fingertips. We want people to be able to look on Google or Facebook or Yelp and find an app that’s specifically designed for them to find resources. We want people to be able to go to a guidance counselor or their nurse, and for these service providers to have tools with access to all relevant information. We want it to be easy for people to figure out how to get and use information to access the resources they need.

We also want a future where it’s easy for analysts, researchers, policy-makers, and funders to understand where the resource gaps are. We want them to be able to have insights into questions like how many times a particular service has been requested, whether people were successfully able to access a service, what people are searching for the most, and when are people’s needs not being met. In this way it’s not only important to use the data to directly meet people’s needs, but also to understand the patterns of use. With everyone using the same information, we’ll be able to better understand these patterns and we’ll be better able to match resources to needs.

That’s the dream. From each according to their abilities, to each according to their needs. That’s the world in which I want to live, and I don’t see a way to get there without solving this problem.

Now, let me be really clear, what we are doing is not sufficient. Getting to that world we want will require all kinds of other intermediate steps that Open Referral can’t take itself. Open Referral is not sufficient to bring about a better world — but it may be necessary, and that’s an important distinction.

What does Open Referral still have to do to bring about this vision, within its scope?

Our ultimate goal is a world in which information about the resources available to people in need is easy to find, share, and use, and we have two goals along the way.

We just accomplished our first of these goals, which was to develop a common protocol, language, and method for sharing this information. It took us years to create this. We started by conducting a participatory research process where we compared existing vocabularies and developed a vocabulary that we thought could enable translation. Then we tested that and got buy-in from stakeholders. Eventually we achieved our goal of getting the trade association to give us their stamp of approval as the new standard for exchanging information.

Our second goal — our current goal — is to build infrastructure that facilitates the sharing of that information. When I say infrastructure, think roads. If applications are like cars or bicycles or scooters, applications need infrastructure in the way these vehicles need a road. Let me use another metaphor. What we have to do now is build the power lines and the power plants and the utility companies. We’ve already built the outlet. We’ve built the standardized outlet that any application can plug into, but now we need to build the system of power provision that is going to ensure that any application can use the same data and also ensure that the different organizations that produce that data can cooperate with each other.

Our infrastructure may take the form of say a publishing platform for information about resources, and as we start to build this infrastructure, we need to work out a few big questions. For one, we need to figure out the technology for the platform. Second, we need to figure out how to get organizations publishing data from different sources onto the platform, and how to synthesize that information coherently and effectively. And lastly, we have to figure out the business model for the platform.

To answer these questions, rather than trying to go directly at the global scale (which is where we want to get), we’re actually working with communities where they are and at their scale. For example, say we look to a community and see that they have 10 or 12 different organizations maintaining overlapping sets of information about resources in the community, we want to build infrastructure that supports them. Sure, this infrastructure is going to look different in different communities, but if we’re able to use a common standard as we build tools that work in one place, they can be redeployed and adapted to work in other places, and I expect we’ll start to see a common pattern. Eventually that infrastructure could scale to solve the problem on a global level.

Right now Open Referral’s next objective is to support communities through the process of developing infrastructure that works for them, and then we’ll replicate that process and scale up those solutions.

Let me make sure I’m understanding correctly. So you have the standard and you have partners who have adopted the standard, but you don’t have a platform, so how are your partners sharing information?

They’re sharing information one-to-one. Which has been valuable on small scales. And now that we’ve had successful one-to-one exchanges, we need to enable many-to-many cooperation — and that’s where federation comes in.

One-to-one sharing could eventually look just like a centralized system where you have one organization providing information to many organizations because it has that relationship with many organizations. That may be fine and good on some scale and might even be ideal in a city. But that city is near other cities, and they’re part of a region or a state, and they need to be able to cooperate at larger scales. This is why we need to enable that federating capacity where you might have your thing, and I might have my thing, but we’re a part of a shared thing.

Is your biggest challenge going to be designing the platform, designing the incentive structure, or something else?

It’s technical design and interface design, but most importantly it’s institutional design.

Federation can mean different things. For example, before the Constitution there was the Articles of Confederation. Basically, it was a loose agreement among a bunch of states that didn’t have a central government but were working together. In terms of open data, some places might actually work like a confederation in the sense that each will continue to do their own thing while they work in cahoots. Where this could work, Open Referral can facilitate that.

In other places, maybe each organization will want to retain its own autonomy but then also form part of another entity. And that raises other questions. Who should control that entity? How does that entity sustain itself? Will everyone pay in? Will they contribute data or money? We have to figure all this out, and Open Referral is gearing up to play an active role in getting to that more complex institutional capacity and technical capacity for cooperation among multiple stakeholders at scale.

Is there anything else that you’d like to talk about?

Let’s see. We covered the Tragedy of the Commons, we covered infrastructure, we even covered fully-automated gay-space communism…

The challenge here is that this is really abstract stuff. If I sit down with somebody I can talk them through it in a half hour or an hour, but that’s a lot. And in these fields, people are accustomed to getting a clear, simple pitch in minutes — something that tells a story about how, like, Jane was hungry, and she used this app, and now Jane is thriving. It’s a simple story, and it has very little room for nuance and complexity.

Over time that experience of simple simple stories trains us to expect that these complex problems should be communicated on a very simple and heartwarming level, when the reality is messy, abstract, confusing and complicated. On some level, we shouldn’t sugarcoat that. So I am looking for more opportunities to tell complex stories in a digestible way, without treating the audience like children.

Learn more

• Download a photo of Greg
• Greg on Twitter
• Open Referral on Twitter
• Open Referral on GitHub
• Greg’s LinkedIn profile
• gregbloom.org
• openreferral.org
• Original essay that ignited Open Referral
• Essay on “community as platform” in Civic Quarterly

Copyright

This story is copyright Greg Bloom and is published on the StoryEngine website under a Creative Commons Attribution 4.0 International license.