Sign-up for updates about our Freedom of Information work and campaigning.
This blog post is a companion post to a shorter blog post explaining the significance of this polling to mySociety’s FOI work.
We know very little about the real picture of Freedom of Information use because there are not comprehensive statistics. Information on users of Freedom of Information is very hard to come by. We have some information through a survey we run on WhatDoTheyKnow, but we know this only covers the minority of requesters who use our service.
Knowing about this picture is important to us for several reasons.
The first reason is one of the big benefits to society of WhatDoTheyKnow is that we make public information easier to discover without explicitly asking for it. If we can know more about how many FOIs are being made in total, we can have a better sense of what proportion of this information we’re publishing (based on some of the maths below, It’s probably somewhere from 5-10%).
The second reason is that conversations about the pros and cons of freedom of information can be dominated by the problems journalists experience in requesting information from the central government. This is a big and important problem, but it shifts the general understanding of the impact freedom of information has had on our society. Through WhatDoTheyKnow, we get a glimpse of a bigger world when citizens are making requests that affect them and their communities – but we don’t see everything, and getting more information about this is vital in informing how we approach our policy and campaigning work.
As part of a “Giving Tuesday”, Opinium gave five survey questions (for a national representative survey panel) to a number of charities, including mySociety. We used one of these questions to find out how many people had made a freedom of information request. The rest of this blog post explains the results of that survey.
The question we asked
Our data comes from an Opinium survey of a representative selection of UK adults that ran between 30th November – 3rd December 2021. Respondents were asked:
The Freedom of Information Act gives you the right to request a large range of information from public authorities (government departments, local authorities, NHS trusts, schools, etc). These are called Freedom of information requests. Have you ever made a Freedom of Information request?
Respondents had the option of responding:
- No I haven’t made a request, and I am not aware of Freedom of Information
- No, I am aware of freedom of information but haven’t used it
- Yes, as part of my job
- Yes, to find out something that might be useful for me personally
- Yes, to find out something that might be useful to my community/society in general
- I’m not sure / NA
mySociety/Opinium polling in 2021 found that 10% of UK adults have used FOI to try and get information they thought would be useful to themselves, their community or society. When including people who made a request as part of their work, this figure goes up to 14%. When looking just at personal use, the figure is 6%. Overall, a majority of people (62%) had either used FOI or were otherwise aware of it.
There was a small gender difference in both awareness and use of FOI, with men having higher awareness than women (68% to 57%), and greater use (16% to 11%). Our polling found that the 18-34 age group were the least aware of Freedom of Information (55%), but were also the age group most likely to have made an FOI request (25%). This is possibly partially explained by a much higher rate of using it as part of employment in younger demographics (12% compared to 6% overall), but the number using it for other reasons is still notably higher (some more discussion of this further down). Looking at respondents by nation/region, there was a less than expected proportion of people who made a request in Wales (6%) and Northern Ireland (2%), but a greater number who made a request in London (28%).
Validating these figures
When I first saw some of these figures, I was a little surprised and wanted to explore some different ways to validate the number.
Digging into it, I found that other polls asking different versions of the question show a similar figure, and back of the envelope calculations based on known statistics suggest the basic ballpark is right – there are millions, rather than hundreds of thousands, of people who have used the Freedom of Information Act.
Part of the reason this figure might be surprising is that our statistical picture of Freedom of Information is so poor, we have very little idea of the scale of it – and what we do know is misleading as to that scale. For instance, a recent Financial Times article, when highlighting the (bad) trend of how central governments are withholding more and more information requested, falls into the trap of assuming that this picture represents all freedom of information requests. But departments and ministries are not the only public bodies that receive Freedom of Information requests. In fact there’s good reason to believe they receive only a small percentage of the overall total.
Most FOI requests in the UK are not covered by official statistics. In 2017, we did a meta-FOI to ask local authorities about the number of FOIs they received. We calculated around 467k were made that year, compared to 46k made in that same year to the central government. From running WhatDoTheyKnow, we know that only 10% of requests made through the site go to the central government departments that are covered in the statistics.
As the number of Freedom of Information requests is much higher than the official statistics show, this helps explain why the number of requesters can be far higher than expected. Not only are there many more public bodies outside central government, but these bodies are closer to people’s day to day lives, and so a broader range of people might want information, and find it through the Freedom of Information Act.
Polling by the UK’s Information Commissioners
The clearest reassurance of the 10%ish figure is that a similar poll found a very similar number. Polling by the Information Commissioner’s Office in 2021 found 49% were aware of “the right to request information held by public organisations”. This is lower than our equivalent question, where 62% had either made use of FOI, or were aware of it and not used it. However in the same 2021 survey, 10% of respondents to the ICO’s survey said they had already made use of “the right to request information held by public organisations”. In the previous year this figure was 12%. This figure is very close to our figure of 10-14% making use of FOI, and it is reassuring to see something in this ball park come from a different survey company.
Both these surveys might be wrong of course, but polling by the Scottish Information Commission in 2022 found an even higher number. This poll found 36% of a weighted sample of Scottish respondents had at some time “asked for information from a Scottish public body by letter, email or online form”. 18% said they did this annually or more frequently. This is a much higher number than the other survey. There are several possible reasons why.
- There is a genuine difference in awareness and use of rights between Scotland and the rest of the UK.
- The way this question is phrased should also include requests for personal information (subject access requests) as well as freedom of information requests.
- This version of the question does not ask about a right, just if someone did something that might have engaged the Freedom of Information Act. This might catch people who get through the process unaware they may have benefited or made use of information rights.
How should we interpret this? There is no strong reason to believe the use of rights is significantly different in Scotland. The Scottish figure was 63% awareness of freedom of Information, which is higher than the ICO UK-wide, but in the same general area as our UK-wide polling, which did not show a significant difference for Scotland. Similarly, our survey found a statistically significant difference in use of FOI by respondents in London and did not find this for Scotland.
As for subject access requests, we actually know from the (really good) statistics recorded in Scotland that around one-quarter of information requests are subject access requests. So even applying this correction, this question is still suggesting around double the figure from our survey. This is likely to be part of the explanation, but is not all of that.
This leaves the possibility that by not prompting about rights or freedom of information, this is capturing a set of people who are coming into contact with information rights issues without noticing it. It is possible to exercise your freedom of information rights without being aware you are doing so. My first freedom of information request was made this way as a student, asking OFCOM if they held some information from the Broadcasting Complaints Commissions’s archive. If you email your local council wanting to know something, it should be processed under the Freedom of Information Act, even if you were unaware of it. This is likely to include some interactions with authorities that will have existed before FOI (and information may have been made available) but which is now covered formally through the Freedom of Information process. For example, a request for library opening times could be processed as an FOI request, but may well have been answered before the FOI act existed.
If this is an explanation for a higher number in response to the OSIC survey, it might also explain the higher proportion of 18-34 respondents in our poll who had used Freedom of Information for personal reasons. Contact with public bodies for information is more likely now to be by email, and trigger the formal FOI process. There is more to explore here around possible shifting patterns of first contact with FOI.
Back of the envelope calculation based on WhatDoTheyKnow statistics
The relevant polling we have is supportive of our poll not being outrageously high. The other approach is to try a very back of the envelope approach based on known statistics to see if this is a reasonable amount of FOIs to have been made.
Based on previous research by mySociety and the Constitution Unit, we have estimates for the number of FOIs made to local authorities in 2005-2010 and 2017. Filling in the extra years between those dates, extending forward, and doubling the number (roughly 48% of requests made through WhatDoTheyKnow are to local authorities – but we don’t know if this applies more generally or not), this gives roughly 11 million FOI requests all time. On WhatDoTheyKnow there are an average of six requests per user (again, don’t know if this applies more generally) – so applying that ratio gives roughly 2 million requesters all time. A figure of 10% of UK adults would expect roughly 4.3 million requesters all time.
To get the two numbers more into sync some combination of the following could be true:
- More local government requests have been made all time than this assumes.
- Not impossible given this is based on three data points (all of which are incomplete surveys and require some amount of extrapolation).
- A greater proportion of requests being made to non-local government bodies than happens in WhatDoTheyKnow
- No way of knowing this without a complete statistical picture.
- OSIC statistics show a higher 60% statistic in Scotland being made to local government.
- Given there are many more non-local government public authorities in the rest of the UK, it is reasonable to guess it’s closer to the WhatDoTheyKnow statistic of 50%, but could it be lower than that?
- The ratio between requesters and requests is different outside of WhatDoTheyKnow.
- Arguments both ways, WhatDoTheyKnow is missing all the ‘not intentionally using FOI’ one-offs, but also some of the bulk requesters who don’t want the results to be public on WhatDoTheyKnow.
Given so many of these numbers are made-up or trying to generalise from WhatDoTheyKnow to all uses of FOI, there is no real reason not to prefer the figure two separate polls agree on. That said, it is reassuring it is in the right order of magnitude (still talking millions rather than hundreds of thousands of FOI users). This question would be helped by a complete statistical picture of FOI in the UK, and to be honest, that would be so useful, it’d be fine if it proved our current numbers wrong.
Sign-up for updates about our Freedom of Information work and campaigning.
Demographic difference graphs
The following graphs show the demographic split on aspects of the FOI polling. Where the percentage for a category is higher than would be expected statistically if there was no difference between groups, it is highlighted in blue. If it is smaller than would be expected, it is highlighted in red. For non-highlighted categories there is insufficient data to say the category differs from the general average.