Measuring Open Data


#1

Open data may be free to the public but publishing comes at a cost to Governments who are often resource constrained. Governments may have policies (or laws) to publish data openly but how do they know that the policy is driving improvement and achieving the desired outcomes?

What metrics can help governments justify the continued expense of publishing open data?

Some resources to help start the conversation:

What metrics do you use?


#2

I’m not confident any metric will help a government justify publishing data.

Even if you look at established social good works you’ll find that Government will have a party and/or political skew on what they justify as a good investment. The literacy level, for example, is part of a broader discussion on what investment should be made in education. Similarly any number of education or health statistics can be used to debate the need for investment in Government programs.

The metrics we hear time and again in relation to open data are related to economic growth or sustainability. Social benefits in health and education are often reduced to such arguments in terms of workforce capability, innovation and sustainability. As such, I’d suggest the most effective metrics to measure will need to have an intrinsic relationship with economic indicators for whatever jurisdiction is being considered.


#3

Thanks for your thoughts Steven. Have you seen any open data metrics based on economic indicators?

Metrics I’ve seen don’t go to far beyond:

  • number of datasets published
  • quality of publishing
  • count of case studies / success stories

#4

When looking at ‘digital transformation’ within Government the measure is often a cost reduction in processing requests for services. Something which cost $10 per transaction moves to $1 and the $9 difference is measured as a cost saving over thousands of subsequent transactions.

This can be measured for digital services via website log files, and more easily with google analytics from multiple Government sites feeding into a dashboard. The UK’s GDS use this approach with great success.

Translating this to open data might be of benefit. For example, for each dataset you’d store in the metadata the FOI request charge, or some other relevant cost assumption. Having this data would allow both the platform owners and app developers to measure the assumed savings they are providing when people use the data.


Impact of open data on FOI enquiries
#5

How about looking at this from the angle of interest. Government agencies normally focus on a particular sector. We could report measurement such as the reduction in service delivery in the sector as it relates to citizens that used the open data available.

  • Social impact index could be given a metric,
  • Number of direct engagement of the citizen with the available open data,
    -Revenue generated by startups using the available data.