Notes from Goa Project talk


Some weeks ago I gave a talk at The Goa Project. I ran into my notes from the event and although I don’t have much time to process them into something coherent at this point in time, I figure it doesn’t hurt to put these somewhere more useful than oblivion.

The warning implicit in such notes is of course that they were written out hastily and are lacking in context; the real context exists only in the recordings of the event where these notes were used.

  • When we talk about transprency there’s a slight tendency to lose track of the objective

  • Transparency for transparency’s sake is useless

  • We need it to be effective at informing the public and providing accountability, but that’s not actually enough.

  • History of transparency legislation

    • Swedish constitution for over 200 years (Forskningsfrihedsforeningen and Yttrandefrihedsforeningen) – linking free speech to transparency in a pretty profound way.
    • US law around 1970
    • 3 generations of transparency legislation
      • Request-based vs Index-and-request-based vs Self-service transparency
  • Frequently, asking for transparency implies asking for bureaucracy, because documents cannot be made public unless they have been written.

  • Criticism of bureaucracy

  • Annoying and costly to the agencies

  • Often seen as make-work

  • Not aligned with internal incentives and so there is resistence.

  • It turns out most people in society in practice don’t really care about transparency much unless they’ve been treated poorly.

  • What people do however care about is that government services work efficiently, effectively and cheaply.

  • One of the things that’s becoming increasingly relevant as more people are getting online is the digitization of processes.

  • Many agencies aren’t technically capable.

    • Hospital vs tax authorities vs archealogy institute
    • Different scale of needs for digitization and different scale of operations
  • Increasingly over the last decade, there has been a move towards appointing a Government CTO or technology office to manage the technical infrastructure and capabilities of the government broadly, and various things that many institutions aren’t necessarily naturally capable of. So while a large hospital might have an IT department, a smaller subject-matter focused agency might not.

    • Experts in their field are not necessarily experts in IT, but government agencies shouldn’t be relegated to a second tier just because they haven’t spent part of their business of hiring techies.
    • Most agencies don’t have the expertise
    • What has happened in the past is that a lot of agencies have hired external consultants, but what we’ve seen is that these private sector actors often overcharge, underdeliver, and frequently the agencies in question end up with something that they aren’t fully happy with.
    • So we see cost overruns and predatory contractors.
    • So the role of this kind of government technology office is to assist the agencies in getting appropriate, effective and on-budget technical capabilities and facilitate digitization of processes, harmonize user experience, guarantee security standards, and so on.
    • But in doing this process, a frequent and natural side effect is that the agencies themselves start to become more transparent, in part because their own documentation is digitized and it becomes easy.
  • In practice the people who are running these agencies want to offer better services, be more transparent, and so on, but often lack the technical skill to identify the possibilites and implement them.

  • While this kind of technical office or Government CTO role isn’t a magic bullet for transparency, it helps substantially by aligning the incentives and helping to increase the quality of government services.

  • A proposal went through the Icelandic parliament to do this a few years ago, and there had already been some work underway to establish this kind of role. While Iceland was already reasonably good in digitization of government services, what we’ve seen since then is a significant acceleration of improvement of services.

  • When you ask for transparency, you might be asking for the wrong thing. If you align the incentives and focus on the public utility, there might be ways to improve society and get transparency for free.

  • One final point referring back to Shushant’s talk and the discussion after:

    • Laws, standards and court verdicts must be publicly available at no cost.
    • This is not bureaucracy, because if there is a requirement for people to adhere to something, then they should have a clear positive right to know what that is. There is no space in a democracy for secret laws or non-public requirements.
    • With court cases it is legitimate to protect the identity of the people who are parties to the cases in various cases, but the courts here simply redact the names while making the broad details of the case public, because there is a public good argument for people being able to understand how laws might be interpreted.