Have you ever submitted a Freedom of Information request for a specific dataset? When you received the dataset, was it in a form that you could easily use and then re-use in the future or was it annoyingly in a PDF format? Then perhaps you’ll welcome the new dataset changes under the Freedom of Information Act (FOIA).
As of 1st September, FOIA requires public authorities to release requested datasets in a form that enables re-use, often a CSV format, where “reasonably practicable”. They also have to publish these re-usable datasets as part of their publication schemes. These changes were introduced by the Protection of Freedoms Act 2012, which made amendments to sections 11 and 19 of FOIA.
“For the first time, the Act now gives users the right to re-use datasets,” writes Steve Wood, Head of Policy Delivery for the Information Commissioner’s Office (ICO). You can access the ICO guidelines on the changes by clicking here.
When public authorities release these datasets however, the datasets might be subject to a licence that sets out the terms for re-use, like the Open Government Licence, if they contain copyright material owned by the public authority. Public authorities can also charge for making datasets available for re-use.
What’s this thing about licences?
There are three types of licences, according to the new section 45 code of practice on datasets by the Ministry of Justice. They are the following:
- The Open Government Licence where you can copy, publish, distribute, adapt and exploit information for commercial purposes. There are not many restrictions on use and re-use, and you can use the information under the licence free of charge.
- The Non-Commercial Government Licence, which is a little more restrictive than the Open Government Licence. You are free to copy, publish, distribute, transmit and adapt the information, but you cannot use it for commercial purposes. As the blog Act Now observes, “it will be interesting to see if public authorities routinely offer this licence (even though it would be against the spirit of the Act and the new code) just to prevent the private sector from profiting from the requested dataset”.
- The Charged Licence, where according to the section 45 code, a public authority can charge you for the re-use of a dataset.
Coming into force on 1st September, The Freedom of Information (Release of Datasets for Re-Use) (Fees) Regulations 2013 confers powers on public authorities to charge for datasets, and that the total fee shall not exceed the sum of “the cost of collection, production, reproduction and dissemination of the relevant copyright work” and “a reasonable return on investment”. The Act Now blog, writing on these dataset changes, notes that “it will be interesting to see how many complaints are made to the Information Commissioner about public authorities over charging”. Interesting indeed.
Does this mean new rights of access?
‘Fraid not. When you request datasets under FOIA, information officers still have to assess whether the dataset engages exemptions and whether the data contained in these sets breaches the Data Protection Act. As many publications on the new changes emphasise:
“The new provisions are about how information is released, rather than what is released… There is no new duty to provide any information in response to a FOIA request that was not previously accessible, and there are no new exemptions from that duty” – page 2 of the ICO guidelines.
The new amendments are welcomed changes to FOIA, since, as Mr Wood suggests, “the more usable the data, the greater the potential to enhance accountability, transparency and economic growth”. It’s in the spirit of the government’s drive for open data, but the changes do not strengthen our right to access (we’ve blogged about the difference between open data and the right to access).
On a practical level the new dataset changes (as well as publishing datasets in general) carry the risk of data-dumping: publishing lots and lots of information without clearly labelling it, thus making it difficult and confusing for the public to access and analyse. The code recommends that public authorities should issue metadata and contextual information about the datasets, but how many public authorities will be able to achieve this for each dataset they publish?
Furthermore, as previously mentioned, public authorities are under a duty to release datasets in an electronic form that allows re-use where “reasonably practicable”. According to the code, “reasonably practicable” will depend on the circumstances e.g. the cost of doing so, the work involved in converting files and whether any specialist equipment or software is required. This raises the issue as to whether public authorities have the technology to enable re-use, which then begs the question, are public authorities ready for these changes?