Key Takeaways from the CodeCamp Strategy Week

Jun 302014
  • A group of experts discussed ways to make the power of open data matter to citizens during the CodeCamp Bellagio event in Italy.

From May 19 through 30, 2014, ICFJ brought together experts from around the world to the Rockefeller Foundation's Bellagio Center in Italy to find ways to use data to advance quality journalism. Ben Colmery helped organize this key summit.

His third report:

Having “Code for” and open data leaders from five continents at our CodeCamp meant we were able to draw upon truly global evidence and intelligence on how to strengthen citizens through government and media-focused data initiatives. Here are key points I got out of our strategy session (each of which warrants its own post).

Start With Low-Cost Experiments, Then Scale

The Lean Startup movement may seem like a fad rife with buzz phrases and empty promises, but those who do it well show amazing results. This is particularly true in open data, and can be powerful in changing the minds of government and media skeptics afraid to make big investments.

For instance, Code for Africa’s GotToVote launched in Kenya, and cost just $500 for a short code and two days of staff time to build its “minimum viable product.” The experiment proved so successful that it has been replicated in Zimbabwe and Malawi and has helped hundreds of thousands of people register to vote and find their polling stations. This may sound trivial, but it’s actually a major problem in many parts of the world.

Just as important, if the tool doesn’t succeed, it is easy to kill off, because a large, time-consuming investment hasn’t been made that is kept alive just so some kind of value can be extracted.

Start With Non-Threatening Projects That Help People Actually Do Something

This is particularly true with government, where highly-politicized topics can scare public officials off before there’s even a single line of code.

Haidee Bell (at the time) of NESTA - the organization behind Code for Europe - pointed to a great example of this with “Edinburgh Outdoors,” an app that enables people to explore parks and green areas in Edinburgh, Scotland. It was a safe topic that was actually useful to citizens, while also paving the way for replications in other government projects and later successes in their “Make It Local Scotland” initiative.

Along these lines, Justin Arenstein emphasized that to get real adoption, it is vital that projects solve real problems that people experience – for instance, access to basic health services. Through Code for Africa, he steers media away from scandal to using media as a vehicle for helping people solve their own immediate, day-to-day challenges, which are often non-threatening, themselves.

Don’t Displace Indigenous Movements

The wave of excitement around open data and open government has led to a flurry of initiatives by multilateral and international organizations to put pressure on governments to open up their data vaults and to seed programs that spur development and adoption of civic tech and data. While it is of great benefit to developing countries that more resources are now available, there can be consequences. Often, having a bunch of experts parachute in has the effect of colonizing countries with entities created essentially to spread the name and likeness of the organization, displacing local movements and weakening their nodes. Models that have worked elsewhere, often in a significantly different regional context (think global north vs. global south), are sometimes imposed without deep local understanding and buy-in, and end up giving open data a bad name through their inability to deliver meaningful, sustainable results.

This idea of nodes and the lines that connect them was a recurring theme at Bellagio. For strong open data movements to take hold and thrive, it is essential that they be locally grown, by strengthening the nodes that already exist, and the lines that connect them to other local and global nodes. Nodes can be media, activist groups, innovation hubs, tech companies, universities, government agencies, the list goes on.

What I have long valued about our Knight Fellowships is that we have led through this exact approach – strengthening local nodes and lines between them. You don’t see ICFJ Knight chapters sprouting up around the world, but rather Knight Fellows who build local cultures of news innovation that connect globally through our program to spread that learning and innovation.

Don’t Create Things That Last More Than Ten Years

Sustainability has become such a core focus of funders and implementers. While it is certainly important that all of these investments go into outputs that continue to have impact, we must be careful not to have the wrong kind of sustainability. “Code for” initiatives disrupt entrenched, bureaucratic institutions like government and media to flip their switch from “perpetuating the status quo to stay alive” to “becoming innovative to thrive and better serve.” However, creating new institutions as a way to sustain that impact runs the risk of ossifying. They can become resource-hungry beasts, themselves, that focus on mere survival rather than serving the public. Therefore, the emphasis should be on sustaining the impact itself, and not necessarily on the new institutions that generate that impact. Otherwise, they may become a new barrier that must be overcome.

Sustainability and the Licensing Problem

Again, we all want sustainability. And open data offers many opportunities for new revenue models, which are crucial to consider in a world where donor funding – often a key driver in civic innovation – frequently shifts or dries up. But this presents a real dilemma. For instance, what happens when a company wants to pay you for all your hard work extracting and curating data into a nice API package, but doesn’t want anyone else to have access to it, so they can build their own products?

There are at least two potential issues here. One, it can be hypocritical to be an open data champion and then build your own wall around it. Two, by selling your API to one company, it is important to consider whether you are limiting the social impact of the data by withholding it from other companies or organizations, including those that may not be able to pay as much.

This may not be as much of an issue in the government world, but in the private media world, it definitely comes up.

The closest we came to consensus is that this is important to think through as early in the process as possible, and be transparent about it. It can be a real dilemma for anyone with a true social mission.

Sustainability and the Consultancy Problem

Without pointing any fingers, there has been a noticeable problem in the open data movement of international consultants charging exorbitant rates to build platforms and services for governments and the like that simply don’t work. Even worse, they crowd out local firms and groups that could charge real market rates, thereby limiting the indirect effect of strengthening the indigenous base of innovation. This also runs in direct opposition to the idea of a lean approach in which a minimum viable product is launched, and iterated repeatedly to maximize its usefulness to everyday users, which can be exponentially cheaper in the long run.

This consultancy problem can also apply to “Code fors.” Initiatives like these can easily become “backdoor” funding mechanisms to launch products that fellows turn into businesses after the fellowship. Is this fundamentally wrong? Not necessarily, but it can lead to the distortion in the previous paragraph and the licensing dilemma. Something to consider as a funder or an implementer.

Should the Open Data Movements Open Up Their Vaults, Too?

Countless NGOs, multilateral organizations, funders, media, universities, and so on, are sitting on top of data still hidden behind a wall. Many are collecting data on health, the environment, corruption, etc. Many more run projects that, as projects, are data in the form of baseline measurements of all kinds, usage of tech, project outcomes and metrics, and lessons learned. So much of this is buried in organizational servers and private reports to funders, where it can’t be turned into larger-scale intelligence. If this data were open, perhaps we could learn faster what works and what doesn’t, and allocate resources more effectively and efficiently.

Just as important, there is the matter of moral consistency. With all of this pressure and expectation for governments to open up their data vaults, isn’t it reasonable to expect everyone else to do the same? Are organizations worried people will see the emperor is not wearing any clothes?

Opening the vaults certainly poses a real dilemma for organizations. Even effective organizations could find their funding threatened if projects reported as successes look like failures from small numbers or are otherwise misconstrued because of a lack of sufficient metadata. The risk is real. Still, is it time for civic innovation initiatives to open the vaults?

“Code fors” of the World Unite!

I elaborated on this in my second post [LINK], but it should at least be mentioned here. If groups are solving similar problems using similar models, they should work together in some way, or risk working against each other. This is absolutely true of “Code fors,” whether they embed technologists in government, media or civil society organizations (CSO). Joint initiatives may be problematic given natural tensions between government and media/CSOs. However, they are all stakeholders in the same ecosystem of serving the public, so it is essential to get people in a room together, break down barriers and build off each other’s strengths. Again, connecting nodes, especially those unexpected, is a key to innovation.

It’s Time to Start a “Code By” Movement

What if you work for a government, or a media outlet, and already have a team of developers? The reality is that, just because an organization has a team doesn’t mean it also has a culture of innovation able to utilize data in a way that creates user-centric applications that truly serves and engages them. Gaining buy-in at high levels can be very difficult for anyone, and even more so if people aren’t trained as change agents. That is why programs that bring in people from outside, like “Code for,” can be an effective model. Yet, well-resourced organizations that cover larger territories may not always have the same access to “Code for” programs, particularly given that these programs are still fairly young and haven’t touched all corners of the globe.

We determined that a solution that could help is to create a “Code by” program. In other words, something internally-driven that links change agents within a government or media outlet to each other, and to those in other governments or media, around a cohesive idea and framework.

Haidee Bell pointed to LocalGovDigital, a network of digital professionals in local government in the United Kingdom whose aim is just that – innovate government from within.

As we weren’t aware of many other large-scale examples of this “Code by” approach, we felt this may offer a new opportunity for adding to the ecosystem that changes how these types of organizations use open data.

Don’t Invent a New Wheel When an Old One Will Do - Reduce, Reuse, Recycle

In truth, this warrants its own post. I’m including it here because it was a topic of intense discussion during our strategy session, and affirmed our decision that the hack session should be focused on reuse.

A lot of time and money is spent building tools. Many of these tools solve similar problems, but don’t draw lessons learned or code from each other that could save them resources or headaches from duplicating their failures. Much of the civic hacking world, while not that large, is still cordoned off into country-level or regional pockets, exacerbating this lack of tool and knowledge exchange. We began to surface strategic and technological ideas on which I will elaborate as I blog about the hack session that followed.

Author’s note: This is the third in a series of posts that will highlight key thinking and outcomes of the CodeCamp at the Bellagio Center.