Policy to drive improvements in open government data


Governments have often started their open data initiatives with a push to publish large quantities of open data. Then they lament that the data is not being used. They discover that they need to improve the quality of their data publishing, prioritise data releases, stimulate demand, educate, and more.

As I’ve mentioned elsewhere, I’m trying to improve the quality publishing issue in Australia through various volunteer initiatives, but…

What if Governments decided to tackle the problem themselves?

What would a policy designed to make sure value was created from open government data contain?

Measuring Open Data

Whoops. I should have posted in the Policy and Research forum. @Mor or @nealbastek are you able to move it?


Done, the topic has been moved to policy and research.
@jwyg - I think this will find this interesting as well.


From my experience in Romania, the poor quality of open data provided to the public is directly related to the poor quality of data used by the government internally. The main factors are lack of good (or any…) data standards and the incredible lack of communication about various public institutions. In more than one case I’ve seen institutions that create internal data systems without taking into consideration the need to transfer data between each other. This is one point where a lot of work is needed, starting with good data policies.

Second, we want to see open data embedded in any new information system created by the administration. Romania is on a wave of investments in such systems, thanks to access to EU funding. We want a clear policy saying that any new system should have a module that allows for easily exporting to the national open data portal.


@ovoicu thanks for your thoughts and agree 100% - Data standards and procurement requirements are both important to enable sharing.

The Open Data Institute produced a Procurement Guide for the UK. In their 2015 Roadmap they suggest that the government, “Mandate that government procurement contracts require the release of open data”.


when I pushed the isle of man government to localise the UK OGL there were lots of lessons learned, for example:

non-technical civil servants were most comfortable with XLS, DOC and PDF not CSV export

in silos it was very common to have extra XLS tabs with internal calculations and notes to keep everything together and lots of people were just too busy (or reluctant due to unquantified data quality) to review files and filter down to just external open data

arms length statutory bodies were more difficult to engage or required a legal opinion or a business case to amend the website footer to adopt the IM OGL from Her Majesty’s Attorney General (a reasonable but surprisingly not authoritative legal opinion)


Graeme Jones
open data | open government | open banking

(I also happen to be a part-time permanent policy advisor in the Isle of Man Parliament though)


Thanks @g_jones, I have had data stewards rename a file from data.xls to data.csv - job done! (not). Clearly education is a key ingredient in the mix as well as making publishing data a formal part of data stewards role.

I just came across the draft international open data charter and I’m finding it very hard to fault. I do want to check it against the open data maturity model but it seems to cover all the important things and would be a great starting point for any policy development.


On the theoretical side, it is not the government’s business. And it is wrong question to ask.
The openness should be the inherit quality of the government, it does not need practical justification.

Because it is basic right, not business enterprise.
Do we make policies to “make sure” we “create value” from right to speak, or right to vote?


Hi for all, my name is Thiago Ávila, from Brazil, and I am researching about publication of Linked Open Data in public sector.

I want invite all of you to contribute with to the Master thesis research “A proposal of model process for publishing Government Linked Open Data”. This research aims to propose a model process that encourage, guide and contribute to the improvement of the publication of Open Data and Linked Open Data in public sector.

To contribute must answer the following questionnaire:
http://tinyurl.com/lodrecommendations (english version) or
http://tinyurl.com/recomendacoesdac (portuguese version)

The participants of this research must act with publication, consumption (use), research or related activities to Open Data and Linked Open Data. The answers should be based on their experiences (academic and professional) with highlighted topics.

The identification of participants is not required. Feel free to forward this form to others who may contribute to the survey.

     This research is being developed in the Masters in Knowledge Computational Modeling of Computing Institute of Federal University of Alagoas UFAL by Msc student Thiago Tavares José Avila, under guidance of Prof. Dr. Ig Ibert Bittencourt.

All the best,

Thiago Ávila


Hi all, first post on here but been following this interesting conversation from a distance for a little while now. My name is Jack Cornforth and I work with Civicus on its DataShift initiative (www.thedatashift.org).

We’ve just published a paper that will hopefully be of interest to some of you. It explores the idea that governments could host and publish citizen-generated data (CGD) themselves, and whether this could mean that data is applied more widely and in a more sustainable way.


Do take a look at let us know your thoughts.