August 21, 2017
I spent two days last week interviewing officials from across the Mexican government to understand how they are producing and sharing datasets that can be used to combat corruption.
As someone who spent three years in government heading up the open data programme, it was a fascinating to be on the outside looking in.
Earlier this year the Charter published an Open Up Guide that went through the theory of how, with the right conditions in place, greater use of core datasets can lead to more accountability, less corruption and better outcomes for citizens. To start to test whether this happens in practice we’ve teamed up with the Government of Mexico, and civil society groups Transparencia Mexicana and Cívica Digital. You can read more about our approach here.
An important early step in the process of road-testing the Guide was to talk to officials in government who are involved in fighting corruption and those who collect and share data. We wanted to understand what data the government was already publishing, and whether it was in the right format and being used by the public. We have begun this work by seeking to understand the barriers to publishing the 30 anticorruption datasets in the Guide.
Here are the 6 things that I learned:
- By and large, the government officials we spoke to were committed public servants interested in improving the collection, publication and use of data. They talked frankly about the problems they faced and were interested in collaborating with us and our civil society partners in getting better data out there and improving their own data literacy skills.
- It’s not just datasets that need to be to be able to talk to each other — people do as well. We carried out our interviews in small groups of officials doing similar work. It was striking how many of them had never met before and didn’t know that someone else in government was doing similar work.
- There were examples of data being produced and shared, but often it was hard to find where. Files are given long technical names or use acronyms that only policy wonks understand. Accessibility would improve if the government simplified names and use language that ordinary citizens can understand.
- While some data publishers were talking to data users, there was no systematic way of doing this across government. Part of our aim with this project is to get this to happen early on in the process so that producing and sharing data is done with an end user in mind. Greater coordination with access to information requests and maintaining regular engagement with data user groups should help to improve the quality of the data and increase its usefulness.
- A big barrier to publication of data is concerns around privacy, data protection and secrecy laws. While the Charter’s principles recognise the fundamental importance of respecting privacy, this needs to be balanced with the commitment to be “open by default”. Secrecy concerns can sometimes be used as a smokescreen to not publishing public information at all.
- There were some commonly used identifiers across different datasets (e.g. for individuals or companies) but due to the siloed nature of the production of data there was no coordination around this. A shared set of identifiers is crucial to make datasets produced by different bits of government inter-operable. For example, sharing identifiers of sanctioned companies with the register of government contractors could prevent barred companies from obtaining government bids.
There are obviously a number of challenges to deal with. However, I came away from the interviews with a real sense of optimism about the opportunities. Our colleagues from Transparencia Mexicana, Cívica Digital and the government’s Open Data Team will be following up with a blog highlighting their own reflections from the process.