Open data + Right to Information = Right to Data

The Open Data movement and the Right to Information movement have many shared aims. The potential for benefit to citizens if these two groups collaborate effectively is significant. In this post, the Web Foundation’s Open Data Research lead, Savita Bailur, shares some thoughts, shaped by our role at the recent International Open Data Conference in Ottawa.

At one of the very first panels of the recent International Open Data Conference, Martin Tisné of Omidyar Network pointed out that “we still haven’t knitted the freedom of information and open data communities together”.

Just a week earlier at  a technology and freedom of information conference in Madrid (organized by Access Info Europe and mySociety) the same issue arose, with participants wondering aloud why these two powerful communities don’t collaborate more.

It was interesting to hear this as “Right to Data” – that open data should belong to us all – is at the core of the Web Foundation Open Data Team’s strategy.  It is a refrain we’ve been repeating in the past year from our response to the Data Revolution Report in December to the Addis Ababa Data Revolution Workshop, At this event, our invited panelist Daudi Were from Ushahidi clearly summed up our position when he tweeted:

We took a deep dive into these issues during Right to Data session at IODC. Attendees included an excellent mix of government officers, journalists, CSOs and researchers from diverse countries such as Tanzania, Jamaica, Indonesia, India and others. We kicked off with four lightning talks from:

  • Silvana Fumega, currently collaborating with the Web Foundation’s Open Data in the Developing Countries (ODDC) research programme Phase 2 (as well as a previous researcher for ODDC Phase 1) on the difference between the RTI and Open Data communities;
  • Jen Bramley from mySociety on the role (and challenges) of technology in bringing RTI and open data together;
  • Bibhu Prasad Nayak from TERI (also a previous ODDC partner) on the intersection between RTI and open data in India;
  • and Eko Prasetyo on Open Data Lab Jakarta’s support and capacity building on open data training for RTI in Indonesia.

Some key takeaways from these talks as well as broader discussions in the session were:

There are significant differences between the RTI and OD (specifically OGD or open government data) communities

Silvana made excellent points that the RTI and OGD communities differ in many ways: in terms of

  • object of the field (largely information in RTI; data for OGD – although of course not always exclusively);
  • difference as to whether copyright licensing is necessary (yes for OGD; not necessarily for RTI);
  • focus (reducing information asymmetry and a focus on access for RTI as opposed to re-use for OGD);
  • a legalistic approach in RTI versus a technical one for OD;
  • and, finally, perhaps most interestingly, RTI tends to adopt a more adversarial approach, whereas OD is more collaborative (how can we leverage open data as a public good) – although this may be contested by both communities.

This summary was one of the most interesting I have seen of the differences between the two communities, but the question is – so … how can they converge? And converge in a way that brings benefit to the proverbial ‘person on the street’. (Although I do recognize as someone who did their PhD thesis on the critique of the concept of community, these are not holistic groups but an artificial creation).

The open data community can learn a lot from the RTI community

The legal recourse expected and demanded around RTI (largely – although again facing its own challenges) seems mostly missing for open data – or perhaps the RTI discourse is just older. A glimpse of the potential in collaboration can be found in Dr Bitange Ndemo’s comment two years after the launch of Kenya’s open data portal, that the “lack of the Freedom of Information Act means that we cannot take any action on government bodies that refuse to release data or refuse to respond”. Open data advocates would also benefit from much closer links with RTI activists who could make open data a much more approachable topic – seen as a means to an end, rather than an end in itself.

On the other hand, the RTI community could be much more demanding on details … drawing on open data practices

The level of granularity that is the core of the open data philosophy – open formats, licensing, data cleaning etc – seems in many cases absent in RTI.  For example, a case was mentioned from Jamaica where an RTI request resulted in disclosure on ministers’ roaming charges. Yet, the ability to do more analysis on this wasn’t possible when at a second stage, an open format was specified in the request. This is not dissimilar to other cases reported, for example in the UK when an applicant complained after a request for information about the 11+ school entry exam was supplied in 184 PDF pages when the requester had asked for an Excel format (and subsequently led to a change in court of appeal ruling on open formats). RTI laws around the world need to factor in formats and standards, where they do not already do so.

Privacy, skills and need for further training straddle both

The group recognized that both RTI and OD needed to work more on privacy, skills and more training for government officers, particularly in under-resourced contexts such as developing country governments  – equally in terms of personal privacy, state interest and data integrity. Jen from mySociety discussed the breaches of privacy often experienced by the UK RTI site  such as sensitive data (children in care; sexual orientation; domestic abuse etc) released by accident by local government in hidden Excel columns (in more detail in this report and in this post).

Unintended consequences: RTI and OD are in collaboration not competition

An important point was made by Bibhu Prasad Nayak from TERI on the question of priorities for an under-resourced government officer. In certain countries – for example, India – penalties are put in place if RTI requests are not responded to within a finite time (usually 30 days). In this case, RTI (reactive disclosure) would be given preference over proactive disclosure. One should not be in competition with the other – how do we ensure both get equal attention?

In conclusion

There’s so much more to write here, but I’ll keep it brief and I’m just excited that RTI/FOI was mentioned at an open data conference. We need to think more of the concept of “open by design” – building openness into government data – whether proactive or reactive, and bring these two communities/networks/movements, whatever we call them closer together. This is happening in pockets – but let’s synthesise it. We will continue to work on this at the Web Foundation and will publish Silvana’s full research in August 2015 – and in the meantime, look forward to the seeing the knitting, Martin!