Francis Maude, then Minister for the Cabinet Office, famously said in 2012 that he would "like to make Freedom of Information redundant, by pushing out so much [open] data that people won't have to ask for it".

Months earlier Prime Minister David Cameron had reportedly told a commitee of MPs that FOI was "furring up the arteries" of government and that the current question-and-answer format could be superseded by regular publications.

By 2014 Maude had warmed to his theme ("My view is that we should be proactively making public everything that is appropriate.") and was signalling an intention to roll back some aspects of FOI ("I am a transparency zealot. But actually I think the Freedom of Information Act probably doesn't protect some things that need to be protected vigorously enough.")

Fortunately this position never gained widespread support either among FOI practitioners or within the open data community. As the Information Commissioner said in 2012, "It is not either/or. It is both/and." There is a broad recognition that without rights of access to information, open data could become simply a communication tool for government rather than a driver for transparency.

The Government established an independent Commission on Freedom of Information, but when it reported in 2016 the recommendations were mostly sensible and unthreatening. For the time being the statutory basis for access to information in the UK seems safe, even if there are doubts about how well FOI is observed in practice.

However: could there be a grain of truth in the idea that releasing open data can reduce reliance on the Freedom of Information Act?

It's difficult to get recent or reliable data on the costs of administering FOI across government, and I am in any case generally unsympathetic to complaints that FOI is a "burden" on public authorities. The Freedom of Information Act has been with us for nearly 20 years now, and is a necessary means of holding government to account at a level of granularity we cannot achieve through the ballot box.

Authorities that struggle with FOI tend to have more general problems with information management and public communication – it's rare to find an authority that is good at open data but bad at FOI, and vice versa.

That said, there are obviously resource requirements to administering FOI and we should look for efficiencies where possible.

Really we don't have a good understanding of the effects of open data on the nature or volume of FOI requests. Demand for information can be a vicious or a virtuous circle, depending on your perspective. Releasing open data without enough detail may just generate more questions. A climate of distrust in government may also increase demand for information. It is difficult to disaggregate these effects and identify clear trends.

Cameron and Maude were mainly interested in release of transparency data: information on government spending and performance that could be leveraged to promote scrutiny and reform of public services. Theresa May's government refreshed the agenda in 2017.

While this focus has produced some notable successes that could not have been achieved as effectively through FOI, such as last year's widely publicised release of data on the gender pay gap, the extent to which the UK transparency agenda as a whole has provided datasets that the public actually want is a matter of debate.

Internationally there is some concern that the open government movement may be shifting attention of activists away from FOI, and that demands for proactive publication of information are diverting resources from fulfilment of access requests. However I suspect this dynamic plays out differently by country. In the US, for example, FOI rights are weaker than in the UK but data initiatives are more likely to have a statutory basis.

There is some anecdotal evidence that release of open data can reduce FOI requests or make them unnecessary, if the data is published to meet an identified demand. For example local councils routinely receive FOI requests for business rates data and some such as Carlisle Council have responded by publishing this data as they generate it. DfE's publication of school lists on Gov.uk was partly a response to frequent FOI requests for this type of data.

As suggested in the US by the Sunlight Foundation, there is likely to be some benefit in sharing and analysing FOI logs to identify common information requests and using those metrics to prioritise data for publication.

For public authorities in the UK this approach has several advantages:

  • If a potential requester can discover the data online themselves, and it is sufficient, there is no FOI request to handle.
  • Even if the requester fails to discover the data and submits a FOI request anyway, the request can be handled more simply by pointing the requestor to the published location and applying the section 21 exemption ("information reasonably accessible to the applicant by other means").
  • Adding the dataset to the authority's publication scheme makes it easier to manage demand for updates by applying the section 22 exemption ("information intended for future publication").

This approach can be applied to other types of public information, not just data, of course.

Signposting the availability of open data at the point of contact for FOI requests, and presenting the published data with clear documentation, is likely to be crucial to any reduction in FOI requests. A recent Sunlight Foundation study of local government in 52 mid-sized US cities found that adopting an open data portal alone did not cause a significant reduction in the volume of access requests, but that implementing a "robust" open data portal decreased volume by 33% compared to the control group.

Public authorities may also find there are benefits from disclosing their FOI responses on the web (with the requestors' details anonymised, of course). This can reduce the likelihood of duplicate requests for the same information, in particular when the subject matter is topical and may be covered in the media.