Our open data maturity dashboard aims to show progress in the final output of open data and tell the story of a growing maturity in the overall stewardship of New Zealand’s public data assets.
It began as a commitment in the Open Government Partnership Second National Action Plan to report against the open data action plan. But we quickly realised it could – and should – display more meaningful measures of progress.
The number of datasets released was the only visual measure of progress outside of government, but this on its own does not show the full story of progress. Different measures were needed inside government – ones that show the shift to well managed data leading to sustainable release of quality data – while grappling with competing priorities, legacy systems, legislative barriers, and policy, procedural and cultural changes.
A survey of central and local government agencies and Crown research Agencies (based on the Open Data Institute’s Open Data Maturity Model), highlighted six maturity focus areas that, in time, will lead to the systematic release of quality open data, ensuring good engagement and acting on feedback along the way.
Forming the basis of our dashboard, the six focus areas are:
A prototype informed by the survey and feedback, led to us working with Unfold Data to develop a design, which we hope tells this story well. We’ll repeat the survey every year and update the dashboard to show progress over time.
The annual survey that informs the Dashboard is analysed for key insights, with privacy and risk assessments highlighted this year.
Much, but not all, data warrants privacy and risk assessments before release. While most agencies carry out assessments, our goal is for all agencies to at least consider whether they’re needed.
A priority for us is to make agencies aware of the useful and relevant Privacy Impact Assessment Toolkit, produced by the Privacy Commission.
While some agencies focus on privacy, other potential risks can get overlooked, such as commercial sensitivity, public safety, potential impact on international trade etc. Completing a general risk assessment will help identify potential risks for data releases, and how they might be managed.
The survey indicated most agencies don’t formally raise awareness of open data through staff induction training. We’re exploring the idea of weaving key messages on open data and open government into induction programmes, as a means towards develop more participatory government. By weaving key messages into induction programmes we could bring about more sustainable change in awareness and behaviour.
We need to take the opportunity to negotiate better data outcomes for New Zealanders.
We’re raising awareness around data in the procurement process. We can’t release data for public benefit if we don’t own it. Specifying the ownership of data generated by the service delivered in contracts for services or research, and the intention to release as open data, is vital. This also applies to contract renewal, rather than just rolling over the same terms.
Attention is being given to data in procurement, through Open Data and Procurement training, messages in conference presentations, meetings with agencies, and presentations at events like the government procurement breakfast. We have seen more attention given to data in procurement, and we’d like that to keep growing.
Questions about the mode of publishing data, essentially to get a picture of the use of application programme interfaces (APIs) and where these are outsourced to, were included in the survey for the first time.
This is an area of interest to the developer community, allowing for more dynamic and versatile reuse of the data. It’s also helpful to agencies to see what services are being used by others.
The last part of the dashboard will help us keep track of two key elements of open data: licensing, and machine-readability.
The greatest benefit is achieved if data is released in multiple formats, including a machine-readable format, as this caters for a wide variety of users and uses. But not including a machine-readable format creates a barrier for people wanting to access the data programmatically, or to import into analytical tools.
Excel spreadsheets are the main reason for data not to be considered machine-readable. They’re human-readable but, unless they have CSV format versions (or other formats like JSON) released alongside them, the effort required to reformat for analysis or consumption by software can be significant.
The survey builds a picture of progress and helps agencies to self-assess. We hope it also serves as a reminder to act when agencies identify gaps. Feedback from some agencies confirms that the survey has prompted some reconsideration of priorities.
The information from the survey also helps us prioritise our programme resources to provide support, training, or new guidance resources where needed most.
We hope you find the dashboard informative and useful!