New code of practice for disclosure of datasets under FOI proposed

The Cabinet Office has published a draft code of practice on datasets (10-page / 80KB PDF) in which concepts relating to the disclosure of datasets under the FOI regime by public authorities are explained in more detail.

Under FOI laws public authorities that are asked to disclose datasets are generally required to provide the information in an "electronic form which is capable of re-use", as far as is reasonably practicable. The Cabinet Office has now proposed that the term requires the organisations to provide the information in a format that is "machine readable, such as Comma-separated Value (CSV) format."

"Datasets are, by their nature, often created in formats that are capable of re-use," it said in the draft code. "Therefore, if a public authority publishes a dataset in a non-re-usable format such as an image file, it should consider also publishing the dataset in its re-usable format before its conversion to an image file, or keeping a re-usable version of the dataset available."

"Where datasets are only held in non-re-usable formats, and it is impractical or too burdensome to convert the dataset into a re-usable format, the public authority is not obliged to convert the dataset before releasing it," the Cabinet Office said.

Public authorities should "as far as possible" comply with the public data principles that Government departments are obliged to adhere to, it said. The principles require, among other things, for public data to be disclosed "using open standards" and in accordance with the "relevant recommendations of the World Wide Web Consortium (W3C)".

The Cabinet Office said that 'metadata' and other information should be provided by public bodies alongside the datasets they disclose in order to comply with the requirements of FOI laws. The Information Commissioner's Office (ICO) has previously described metadata as "information on the properties of electronic documents" that includes details about the "author, dates, editing history, size, file paths, security settings and any email routing history".

"Published datasets should, so long as there are no good reasons for it not to be provided, be accompanied by a sufficient amount of metadata and contextual information about how and why the dataset was compiled or created in order that users may fully comprehend the dataset they are dealing with and as part of compliance with Section 16 (duty to advise and assist) of the Act," the Cabinet Office said.

The information provided alongside the datasets should set out what the terms of re-use are if there are restrictions placed on the information by the application of copyright or database rights subsisting in the information, the department said in the draft code.

"Where the copyright or database right in a dataset is owned wholly or partly by a third party, a public authority can only give permission for re-use of the dataset if it has been authorised to do so by the relevant third party rights holder," the draft code said.

"The ‘re-use duty’ only applies where the public authority is the only owner of the copyright or database right. If the authority does not have the legal authority to make it available for re-use it should make it clear to the applicant when releasing the dataset that this is the case. Even when the public authority is not able to license re-use, it must still provide the dataset in a re-usable format, so far as reasonably practicable, if the applicant has asked for an electronic version," it said.

In a statement, Cabinet Office Minister Francis Maude said: "Enhancing the right to data is a key driver of our transparency agenda, so we have amended the Freedom of Information Act to ensure that public authorities publish datasets for re-use and in a re-useable format."

"To help them meet these new obligations, public authorities must have the best possible guidance. The open online consultation invites feedback on how we can improve the draft code of practice that authorities will use. The prize – as well as ensuring we deliver better value for money in public spending – is to drive real social and economic benefits by making it as easy as possible for businesses and other organisations to exploit datasets held by public authorities," he added.