When you research and work in an emerging field such as ‘open data’, for good reason, you get many questions and skepticism. When new knowledge and actors emerge, it is a bit like shifting sands. So, when the Latin American community pioneered the first unconference and conference on open data, many argued it was just a fad. I argued the opposite. We are now running the third unconference and conference in the region next September in Chile. Latin American countries in the context of the Red Gealc, are about to formalise a forum and procedures to discuss open data policies in the region. The Open Data Day saw more than fifteen groups across the continent organising events to free, using and discussing public data. A community is now in place, actively working with data to deliver change in several areas.

So if you are still in the skeptical camp you might well say: “Fine, it did not go away. Did it change anything though?” This is a fair question.If you are serious about evaluating public policies, then you might already have encountered a little book “ Theories of the Policy Process ” by  Paul Sabatier. Sabatier notes that to have a proper evaluation of a programme or of a policy field, you may need to wait ten years from the time of its emergence. If you really want to accumulate scientific knowledge, that could take more time. Sadly, people like myself who work in the policy world know about this. We often get evidence of where and how we can to support or oppose an idea. The policy world is imperfect, and, when technology is involved, the buzz can be overwhelming. There is always a demand to perform as if we were running a start-up.

I would like to have a more grounded and realistic conversation about open data and its impact, which could be summarised in this statement: “Fine, it did not go away. What kind of contribution is “open data” making and how it is happening in the field?” In this way our debates and “meta-conversations” about impact become more evidence-based. But we might need to unpack this question. When I write, “open data,” I mean the policies that allow the release of public data in open formats for people and machines to re-use. So far some interesting developments have been found which I would count as “impact.

a) The “living” infrastructure: Government changing role

There is evidence that governments in this region are setting up the basic infrastructure to release public data. This was noted early in the day by public servants, who considered government data as public infrastructure. We are seeing many ways in which governments build this infrastructure, with different emphases according to their national priorities and administrative systems. But, unlike other public infrastructures (e.g. roads) this one is alive. We have evidence showing that, the more data is used (hence becoming critical for certain services or activities), data changes and generally improves its quality. Suddenly when use emerges, there is more attention on how data is collected and assembled. The crucial part is the interaction process, the ebb and flow between users and producers of data. As a result we see evidence of certain government processes that are re-organised to share information that was often ignored, even inside the “owner” organisation. Building this infrastructure often demands the re-deployment and update of several skills in the organization, leading to the emergence of new administrative units (i.e., labs, offices, initiatives) tasked to engage with this policy. A few governments are even setting up new rules to establish certainty about the access, use and re-use of public data. Others are setting up forums to discuss what to release first and how. Thus, we are seeing technical, political and social processes converge.

 b) A new breed of civil society organisations

The map of civil society in Latin America is changing. A new breed of organisations is now emerging in the region. While we don’t yet have a full typology available, they certainly are the “demand side”. Some of them, like others in the global north, use data to check on governments are focusing on transparency and accountability issues. Others are taking advantage of open data policies and building social services in partnership with governments. Such partnerships and arrangements are fairly new. These organisations are part of a new collaborative logic, seldom seen in the region. Open Data is also empowering a new kind of journalism based on data and facts, and, as a result, new skills are needed to harness the potential of open data. Certainly we will need to get much better in terms of defining these skills and characterising these organisations. Part of our research

c) Early impact on policy areas

t is too early to tell whether open data is delivering dramatic change in specific policy areas. We know that the release of information affects the behavior of certain actors in some policy arenas. But it is often the process of releasing the data and building on it, that creates impact in a particular policy area. We might be able to see changes but we still need to develop better rationales or hypotheses of how change could unfold, contributing to a particular policy outcome. Also, we need to develop counterfactuals about what happens in the absence of public data. In short, we need a bit more of experimentation with specific policy domains. Through this year we are running specific interventions to assess this such as the case of atuservicio.uy. Health, Cities and public services are being affected by open data policies as new tools are being developed.

d)  The road ahead.

Things are changing. Occasionally I’ve been approached by colleagues rushing us to set up a baselines to measure some of these changes, or to run experiments based on randomised control trials to check whether the existence or absence of data has any effect at all. I am usually sympathetic to the need for setting up measurements but I often argue that we need to think carefully about this. In this first stage, at least in our research, we argue for identifying key variables and causal mechanisms that eventually could be tested. It is a young field and we need to be rigorous but also realistic about what can be measured and how it would be done. For some activists and advocates I might have stated the patently obvious. But it is often a shared worry about what is working and how. If this field is to grow, then we need to be clear about these things.

At the moment our research is far from predictive. I would love to have an algorithm to calculate the possible scenarios of how open data policies could deliver change in the most unequal region in the world, but I don’t. Still, if I had to tell you three important things to consider for the future, they would be:

1. We need to critically assess the current public infrastructure in place to make sure that is actually open and delivers value.
2. We need to explore the collaboration processes and its key actors to promote data usage and impact on specific policy areas.
3. We need to make sure that the whole agenda moves towards inclusion and appropriation of public value by as many people as possible

While this is frustrating for many, this will demand understanding in the context of countries, polities and policy domains. Open Data can enable change; it may even deliver a few quick wins. But only when data is used towards a greater purpose in a certain context, it  will it become embedded as a public good.