Yesterday there was some discussion on Twitter about a request made to Transport for London (TfL) through WhatDoTheyKnow. People were asking if TfL’s response was compliant with FOI.

The request related to the Barclay’s Cycle Hire Scheme running in London. Earlier in the year, the Mayor of London and TfL (of which he is Chair), had boasted of 1 million journeys made on hired bicycles through the Scheme. The requester congratulated them on this and asked for data on each of the million journeys.

A month later, the response from TfL arrived informing the requester that the data would shortly be made available in their online repository, which would require the requester to register. To give him a flavour, the response included data on 100 journeys.

To cut a long story short, the requester was not happy and has asked for an internal review. So was the response from TfL compliant, and if not, how should they have answered?

In short, and from a purely technical point of view, no. The response didn’t provide the information asked for, and nor did it explain which exemptions were being applied.

It sounds to me like a classic case of a senior manager in TfL having difficulty with the practicalities of FOI and suggesting a response in the hope that the requester would go away until they’d been able to work out how to make the data available. On the plus side, TfL clearly were looking at making the data available online (and have now done so), so it’s not that they are just trying to be secretive . So how should they have answered?

The most likely reason that they were not able to provide the information (and this is alluded to in the response)  is that  supplying the data would cost too much. After all, a million records is a lot of data, and there may be serious technical difficulties in transmitting that data. It may cost more than £450 (the cost limit for locating, retrieving, and extracting data under FOI) to employ a consultant to find a way to do that. If this was the case, they should have explained this in their response.

The other alternative that I would have gone for is the exemption at section 22 (future publication). If you’re going to publish information in the future, as TfL indicated they were going to, and don’t feel able to make it available now, that would seem to be a reasonable approach. It is subject to a public interest test, but the fact that the information would be available very soon, and the likely cost to the public purse of providing it sooner would seem fairly persuasive arguments against immediate disclosure.

There’s obviously a separate issue about requiring users to access data via a site which requires registration. But once the information is made available, you could use section 21 (otherwise accessible) to refuse the request and point to the data repository. It’s likely this would satisfy the Information Commissioner unless there was a significant charge or the requester was able to demonstrate some reason why this didn’t satisfy their needs.

I sympathise with TfL – 1 million lines of data could well be difficult to provide via email (one could even speculate that if it had been disclosed via WDTK it might have caused problems for the WDTK website and/or users who tried to download it). But they could certainly have come up with a better answer than the one they gave.


  1. I’d assume the data are all held in a database, so it would probably be very easy to select the data required from the tables and export it as a csv file. It may only have taken a few minutes!

    I haven’t registered to look at the data provided, but I wonder if the requester can actually handle 1 million records – Excel couldn’t cope!

  2. If you do register on the TFL website, the data is presented as a 25MB zipped text file which WDTK can easily handle.

    We ask authorities to get in touch with the WDTK team to discuss the best way of sending the data if attachments are over 50MB.

    Zeno – quite a few people already have started making interesting visualisations with the data – see

    Alex – WhatDoTheyKnow volunteer

  3. I find it curious that Tfl chose not to rely upon any exemptions. As you say, s22 would seem to be relevant here.

    One other issue that strikes me with this request is the suggestion from the requestor that compliance should be straightforward in technical terms. As an FOI officer myself I’ve received a number of requests in the last year that were technical in nature and quite frankly required some ‘up-skilling’ on my part to answer. Of course this is no defence against non-compliance with the FOI Act but nevertheless requestors should keep in mind that it is often FOI officers, rather than technical staff, who have the job of assembling and presenting the requested data.

    John M
  4. Thanks for the comments. As John M says, often we FOI Officers aren’t in a position to judge whether it’s feasible to provide data in these situations – we have to trust the word of whoever is responsible for the data. The arguments in my post are based around the assumption that there were good reasons for not disclosing the data earlier but just not explained properly in the response. I’d hope that TfL wouldn’t just drag their feet for the sake of it (but then I’m an eternal optimist!). But even if it was just a matter of delaying until a proposed publication date, TfL could still have made the argument for s.22 so really odd that they didn’t do that.