I still remember the last work meeting I had as a public servant at the Ministry of Development and Social Inclusion in Peru before coming to Harvard Kennedy School in which we spent about three hours discussing whether one ministry was willing to share its data with another one, even though the two ministries were working together on a common project. Months later, the news was not so promising.
What is the deal with data? Data is one of the biggest assets that governments have, but it is still difficult to break the idea that it belongs only to one institution or, even worse, that it belongs to the government itself and not to the citizens.
Understanding data as an asset is a new way of thinking of governance, and this idea has been influencing many cities and countries around the world. DataPortals.org, an organization of leading experts in open data, has mapped more than 500 portals around the world. Although open data portals are much more prevalent in the developed world (the USA and Europe comprise more than 50 percent of these portals), there are already many such initiatives prospering in developing countries.
What’s the point?
There is a lot to be gained by sharing data. It promotes accountability, makes citizens true participants in government, and fosters innovation for better service provision. Exploring primary data opens the door for an infinite number of iterations and integration of information that helps to find opportunities to better address people's needs. In that respect, Open Data 500, an effort of New York University, is mapping how American companies are extensively using data from the government to find new opportunities for businesses. Open data arms citizens with information to discuss government decisions, it gives NGOs the tools for better targeting their interventions, researchers the key sources of analysis, and it gives entrepreneurs inputs to find opportunities for innovation. Above all, it creates a space where ideas can grow and a culture of sharing can flourish.
Not every portal with databases is an open data initiative
Open data sounds like a great idea, but not every portal out there falls into this category. Let's say you are looking for data on education in a certain country and you find a list of indicators, charts, and graphs covering a significant period of time on its Ministry of Education website. Can you call it an open data portal? Well, not yet. There are several lists of the "principles" of open data (Open Government Data, The Open Definition, Open Government Data: The Book, etc.), but I would like to point out three of them.
Raw is better: Open data is not just sharing processed chunks of data that someone decided are useful. What it really implies is the sharing of primary databases that allows any user to play freely with them and to decide the outputs she wants to obtain. Being honest, how many times did you end up discovering totally different — but much more useful — results than what was of interest at the beginning of your research?
Accessibility: Even when some primary data is shared, many times it is accessible at a certain cost or it is not shared in any of the most common languages that a regular computer can read. If it imposes some kind of discrimination in granting access, then it is not open data.
Empowerment: Giving the users the ability to repurpose the data and the outputs of their analysis is a key element of open data (and that is not a condition that goes against citing the source or recognizing the author of the analysis).
A case of success: Kenya Open Data Portal
Under the slogan "Together let's build solutions with Open Data," the Kenya Open Data Portal claims to be one of the first initiatives of open data in a developing country and to have had more than half a million downloads of data since its creation in 2011. Whether or not these numbers are accurate, the truth is the portal shares not only data visualizations of a variety of government topics but also more than 300 primary datasets in nine different digital formats, which ensures machine readability for everyone. Further, there is no restriction to visitors on how to use the data or to declare what is the purpose of its use. In addition, every shared dataset has a discussion board where any user can make suggestions, inform the government about inaccuracies, or ask for additional information.
"It would be useful to organize the data into current administrative boundaries like wards."
(Comment left in Kenya primary schools dataset on Kenya Open Data Portal.)
The launch of the Kenya Open Data Portal was possible due to a combination of political and legal factors. Several attempts to release information publicly in Kenya failed due to fierce opposition from actors that benefited from monopolizing access to the data. However, after strong pressures for transparency from civil society and the private sector, the momentum was created and the launch was possible. Despite the difficulties that the initiative has faced in the three years since its launch (especially the ones related to public entities' reluctance to share information), there are several positive spillovers that the Kenya Open Data Portal has created.
Within Kenya, there are new initiatives that have been created to promote the demand for open data and to change the mindset of institutions and citizens. One of the most remarkable is the Code for Kenya initiative that created apps like The County Safety Crime Visualization, Find My School, Data Story (an app for journalists), Star Health, and Go to Vote (Mutuku and Mahihu, 2014). After the relaunch of Kenya Open Data Portal in 2015, the number of ministries that share data in the portal has risen from four to 31, and the page views have increased from 44 to 58 million. Also, new applications have been added like the budget app, where people can interact easily with budget information and specific information by counties (more info here).
The most important implications of open data
Even while open data portals are spreading throughout the world — and fortunately the movement is also getting some space in developing countries — the true relevance of these initiatives is not only the sharing of data itself but what it is behind them. A sustainable open data initiative implies most of all, a minimum capacity of coordination within the government (between government institutions but also between departments and offices within a particular institution). For those who have worked in government, it rapidly becomes clear that one does not necessarily get unfettered access to datasets even for your own institution. Having an open data portal also implies that the initiative had at least some success in changing the culture of mistrust between public servants and in flipping the sense of power that holding the data creates. This means that open data not only involves technology and programming skills but, most of all, it is about dealing with people. Finally, a sustainable open data portal cannot survive just by uploading datasets every other year; what keeps it alive is the existence of internal users that maintain the need and the pressure for updates.
At the end of the day, coordination failures could be one of the most common issues inside government. Open data is not a solution to this challenge, but it could be the external face of a greater inner change, a change in the public work environment, the creation of a community that is now willing to learn quicker through coordination and cooperation between their members . . . and beyond.