Transport for London (TfL) refused to disclose annual compensation recovery information regarding damage caused by graffiti in response to my FOI request. But TfL had already published a press release about civil proceedings for compensation against a named individual on its own website. TfL also initially appeared to confirm it held policies for recovering compensation for graffiti, but months later revised its position to Neither Confirm Nor Deny (NCND) following criticism by the ICO. In this post, I explore the route to challenging TfL’s Freedom of Information (FOI) refusals through the ICO and the First-Tier Tribunal, as well as how TfL was ultimately allowed to keep the information secret following a private, closed evidence session.

In 2022, I made an FOI request to TfL in the following terms after reading about the significant amount of graffiti damage caused to railway networks:

My initial FOI request

1. Does TfL have any policies for recovering the costs of removing damage caused by graffiti from offenders, either through debt recovery action, civil court action, or by asking a criminal court for a compensation order? If so, please provide me with those policies.

2. How much money did TfL spend to remove damage caused by graffiti in 2020 and 2021? If an exact figure cannot be provided, please provide your best estimation.

3a. How much compensation did TfL ask to be awarded by criminal courts in respect of damage caused by graffiti in 2020 and 2021?

3b. How much compensation was TfL awarded by criminal courts in respect of damage caused by graffiti in 2020 and 2021?

3c. How much of this awarded compensation has been paid to TfL by offenders?

3d. What action is TfL taking to recover any such unpaid compensation that is overdue?

4a. Against how many people did TfL take debt recovery action (not including civil court action) in respect of damage caused by graffiti in 2020 and 2021?

4b. What was the total sum claimed by TfL in respect of all such debt recovery action?

4c. Out of the total sum claimed, what was the total sum paid to TfL in respect of all such debt recovery action?

4d. What action is TfL taking to recover any such unpaid sums?

5a. Against how many people did TfL take civil court action in respect of compensation for damage caused by graffiti in 2020 and 2021?

5b. What was the total sum claimed by TfL in respect of all such civil court action?

5c. What was the total sum awarded to TfL through civil court judgments (including default judgments & admissions) resulting from such civil court action?

5d. Out of the total sum awarded, what was the total sum paid to TfL in respect of such civil court judgments?

5e. What action is TfL taking to enforce any such unpaid sums awarded?

TfL responded in November 2022. TfL initially also published their full response on their website, but subsequently removed it during my appeal (the relevant URL is https://tfl.gov.uk/corporate/transparency/freedom-of-information/foi-request-detail?referenceId=FOI-1927-2223 if you are interested in looking at it through an internet archiving service). The main relevant extract from TfL’s response is as follows (but note its revision further below):

TfL’s initial response to my request

I can confirm that we hold some of the information you require.

In relation to question 2, note that TfL does not hold the information requested as the removal work is covered within the broader costs of cleaning and maintenance overall.

In relation to questions 1 and 3a through to 5e, while the information is held it is exempt from disclosure under sections 31 (law enforcement), 38 (health and safety) and 43 (commercial interests) of the Freedom of Information Act.

Section 31 applies where release of information would be likely to prejudice the prevention of crime. As you will no doubt be aware, the TfL network can often be the target of graffiti and other vandalism, and anything that encourages this is detrimental to our efforts to combat it. We have avoided releasing information on graffiti on the TfL network in the past and we continue to be very cautious about encouraging vandalism by releasing any information that makes TfL assets an attractive target by increasing the perceived “challenge” to graffiti vandals which we know is a key motivation for their activity. We believe that revealing the extent to which perpetrators are prosecuted and punished, and whether this varies year on year, would encourage people to target the network. We believe section 31 is therefore correctly engaged as release of the information would prejudice the prevention of crime.

Section 38 is engaged because in order to commit graffiti offences it is often necessary for perpetrators to trespass onto prohibited, unsafe areas of the TfL network – such as on the tracks and in our depots. Clearly this can be highly dangerous, as illustrated by the deaths of young people on the National Rail network in the past, as covered in this news item: https://www.bbc.co.uk/news/uk-england-44523953. There have been other deaths associated with possible graffiti vandals, as featured in this news item: https://www.theguardian.com/uk-news/2019/mar/21/two-graffiti-artists-dead-east-london-rail-tracks-electrocution

It is clear to us that any increase in graffiti attempts on our network translates into an increased danger to health and safety, not only for those committing the trespass into inherently dangerous areas of the network (as evidenced in the news articles above), but also to the TfL employees and law enforcement personnel who work in these areas and who have to deal with the consequences.

Section 43 is engaged because the consequences of graffiti and other forms of vandalism for TfL can be very costly, both in terms of the need to have assets professionally cleaned but also additional, less obvious costs that might arise such as the following:

Repair and maintenance of security measures – intruders onto the TfL network frequently damage fencing or other parts of the network in the process of gaining access to sites for graffiti. It may also be necessary for TfL to install additional security measures in locations that are frequently targeted, which has both direct and indirect staff costs

Affected carriages have to be taken out of service and replaced, which can lead to difficulties managing the rolling stock and reduces the amount of it available to run services

Some trains have to be taken out of service at short notice and where there are intruders actively on the network, TfL may have to suspend services for safety reasons. This not only leads to significant delays and inconvenience for our customers, but also has direct financial consequences for TfL such as increased passenger claims for delay compensation.

Whilst it can be difficult to fully quantify the financial burden graffiti and other vandalism puts on TfL, and while we have no recent figures, we have previously estimated that it costs millions of pounds per year. Across the wider rail industry, Network Rail have previously estimated an annual outlay of £3.5m which highlights the significant financial burden and disruption felt by train and rail operators because of this criminal activity.

All three of these exemptions are “qualified”, meaning we have to consider whether the greater public interest rests in them applying and the information being withheld, or in releasing the information in any event. TfL recognises that there is an inherent public interest in openness and transparency, and in particular where this relates to the maintenance of public assets and the effective expenditure of public funds. In this case, it may also be of interest in enabling the general public to understand the extent of this problem on our network. On the other hand, there is a very strong public interest in discouraging graffiti and other forms of vandalism on the network, in order to protect the spending of public money and to protect the health and safety of perpetrators and staff. As outlined above, we consider that the publication of this information would be likely to increase graffiti attempts on the network as it is seen as a sign of success by graffiti vandals. We consider that the inherent public interest in openness and scrutiny of our expenditure is to a large extent satisfied by the information that TfL already publishes, such as via our Annual Report and Accounts. We do not consider that the publication of this additional information would add sufficiently to public understanding so as to warrant the risk to TfL’s commercial interests or to health and safety, and therefore believe that the balance of public interest rests with the exemptions applying.

From this, I concluded that TfL did not hold the information in question 2. I also concluded that TfL did hold the information in questions 1 and 3a through to 5e. In other words, I concluded that TfL did hold policies for recovering the costs of removing damage caused by graffiti from offenders, as well as figures about the recovery of compensation from graffiti offenders. However, I concluded that TfL did not disclose this information using the cited exemptions.

I wasn’t satisfied with TfL’s reliance on the exemptions, so I requested an internal review in the following terms:

Q1 and Q3a through Q5e

I do not accept that these exemptions are engaged for the reasons you have provided. Even if they are (which I do not accept), the public interest test clearly favours the disclosure of the requested information. Please see below:

Section 31

– I asked TfL to provide me with its policies and figures for recovering the costs of removing damage caused by graffiti from offenders, either through debt recovery action, civil court action, or by asking a criminal court for a compensation order. TfL states that providing this information would “prejudice the prevention of crime” and it thus, by implication, relies on Section 31(1)(a) for withholding the information. 

– By its nature, the requested information relates to the action taken against persons who have already caused damage to TfL’s assets and how TfL pursues them to mitigate the damage it has suffered by their actions. 

– The requested information does not relate to any information the action TfL (or any other law enforcement body) takes to prevent such damage or offences from occurring in the first place. 

– Furthermore, the requested information relates to prolonged periods of time and a vast number of locations, such that deducing any information about specific locations or periods of time is impossible or extremely unlikely.

– There is no plausible causal link whatsoever between the provision of the requested information and people committing crime. The hypothetical and highly implausible causal link suggested by TfL does not come close to satisfying Section 31 in respect of the requested information. Please see John Connor Press Associates v Information Commissioner (EA/2005/0005, 25 January 2006) at paragraph 15.

– Even if it is determined that such prejudice would occur (which I do not accept), it is clearly trivial at most and as such does not engage Section 31.

– Providing me with the requested information, therefore, does not in any way “prejudice the prevention of crime” by TfL or any other law enforcement authority as TfL claims.

Similarly, the policies requested in Q1 and the figures requested in Q3a through Q5e relate to costs recovery by TfL for the damage caused by persons through graffiti. This does not fall under the way in which potential offenders are “prosecuted and punished”. The recovery of damages/compensation by TfL for damage caused is, by its definition, not punitive in nature: damages/compensation and are intended to put TfL back into the position it would have been in but for the torts committed. They are explicitly not intended to ‘punish’ offenders: indeed, no criminal offence needs to have been committed for TfL to take recovery action in respect of a civil tort. Furthermore, at the point that TfL can ask a court for a criminal compensation order, a separate decision on whether to criminally prosecute would have already been taken based on whether the prosecuting authority believes there is a realistic prospect of conviction for a criminal offence. The requested policies and figures in respect of TfL being compensated for any damage caused do not in any way affect whether or not the test for criminal prosecution is met.

Section 38

You state that providing me with the requested information would “endanger” potential offenders, trespassers on TfL’s network, and TfL staff and law enforcement personnel. This is clearly not the case for similar reasons to those provided above. The requested information relates to costs recovery action that TfL takes after such trespassing has already occurred. In addition, I do not accept that TfL has included any content in the requested information that would in any way directly encourage or make it easier for people to trespass onto the railway. As such, the provision of the requested information does not in any way endanger potential trespassers on the railway network. 

At most, your case is that the disclosure of these policies might just conceivably lead to harm (but I do not accept that even this is the case). This is well below the threshold set in Section 38. Please see Andrew Lownie v the Information Commissioner and The National Archives and The Foreign and Commonwealth Office EA/2017/0087 (12 July 2018) at para 45.

Section 43

You state that you are withholding the requested information under Section 43 because “the consequences of graffiti and other forms of vandalism for TfL can be very costly”. 

While I do not dispute that the consequences of graffiti are costly for TfL, I fail to see how this has any relevance whatsoever to the application of Section 43 to the requested information. If I understand correctly, you appear to be relying on Section 43(2) for similar reasons to your reliance on Section 31. 

Providing me with the requested information is not going to make the consequences of graffiti any costlier for TfL, nor will it in any way prejudice TfL’s commercial interests in the ways you appear to be claiming. You must be able to demonstrate a causal relationship between the disclosure of the information in question and the prejudice you envisage. Similarly to my submissions at Section 31 above, TfL has completely failed to demonstrate a plausible causal link between the provision of the requested information and the prejudice of its commercial interests.  Even if it is determined that any prejudice would occur (which I do not accept), it is clearly trivial at most and as such does not engage Section 43.

Public interest test & relevant disclosures

There is a clear public interest in the requested information being disclosed: costs that are not recovered from offenders are ultimately paid by the taxpayer and TfL passengers through increased fares. In addition to this, I note that:

– By way of the open justice system, information about individual civil cases brought by TfL is already publicly accessible by way of Civil Procedure Rule 5.4C.

– Criminal cases relating to graffiti are similarly open to the public and/or journalists, including details of any compensation orders sought by and awarded to TfL

– TfL has recently received multiple government bailouts and has increased its fares. The public have an interest in knowing how and to what extent TfL is exercising its legal rights to reduce the burden caused by actionable and recoverable graffiti damage on the taxpayer and fare-paying TfL passengers.

Relevant disclosures that are already in the public domain include:

– A press release by TfL entitled “London Underground successfully sues graffiti vandal”: https://tfl.gov.uk/info-for/media/press-releases/2005/february/london-underground-successfully-sues-graffiti-vandal

– Report of the London Assembly Graffiti Investigative Committee, which states that: “The GLA or TfL should appoint an officer who would be would responsible for pursuing offenders through the courts on behalf of all public transport operators. This would then  enable the operators to increase investment in preventative measures.” https://www.london.gov.uk/media/77802/download

“We consider that the publication of this information would be likely to increase graffiti attempts on the network as it is seen as a sign of success by graffiti vandals.” By this logic, all public reporting of criminal offences would be stopped as it may be seen as a ‘sign of success’ by criminals and increase criminal activity. This is clearly not a valid argument, particularly in the context of open, transparent society, public bodies and justice system. A reasonable person is obviously not going to look at the requested information and take that as a ‘sign of success’ or an invitation to start committing offences against TfL, particularly as they render themselves liable to a variety of criminal offences and penalties if they do so (not covered by the requested information).

I further note that, internationally speaking, [some] railway companies proactively publish figures for the amounts they have recovered from graffiti offenders as compensation for any damage caused. This reassures the public that their money is not being unnecessarily spent on damage that can be recovered directly from the offenders responsible.

All this further goes to show that the exemptions you have cited cannot and should not be relied upon for withholding the requested information.

In its response to my request for an internal review, TfL upheld its original position, largely relying on similar arguments. They also claimed publishing the information would allow individuals to create an enforcement database and predict TfL’s enforcement patterns through the mosaic effect. They also claimed that the information I requested could be used by graffiti offenders to gain notoriety and “kudos”.

I believe that the annual, aggregated information I requested about TfL as a whole is highly unlikely to cause TfL the prejudice it claims to envision. Therefore, I subsequently complained to the ICO. Many months later, I received an email from TfL containing the following passage:

Please note that for question 1, namely “Does TfL have any policies for recovering the costs of removing damage caused by graffiti from offenders, either through debt recovery action, civil court action, or by asking a criminal court for a compensation order? If so, please provide me with those policies.” the basis of our position for this question is that we neither confirm nor deny whether we hold the information requested, but in any case the exemptions outlined to you previously still apply for the reasons we have set out.  

I am sorry that this was not made clear at the outset but trust this clarifies our position.

To me, this seemed absurd. After all, I had concluded from TfL’s original response (which had also been published on TfL’s website) that they held the policies. Did you conclude the same?

How could TfL realistically backtrack and revise its position to Neither Confirm Nor Deny (NCND) later on? It made no sense whatsoever.

Shortly afterwards, the ICO issued a decision notice upholding TfL’s position. However, it did not in any way address the fact that TfL had apparently switched to NCND after confirming it held the policies. The full decision notice can be found here: https://ico.org.uk/media/action-weve-taken/decision-notices/2023/4026983/ic-210387-g5p2.pdf

Appeal to the First-Tier Tribunal

I remained of the view that the annual, aggregated information I requested about TfL as a whole is highly unlikely to cause TfL the prejudice it claims to envision (questions 3a through to 5e). In respect of question 1 regarding the policies, I took the view that TfL should at least confirm or deny whether it held the information (but I no longer sought full disclosure of the policies). In my view, this was particularly the case as they had apparently already confirmed they held the information in their original response. I therefore appealed the ICO’s decision notice to the First-Tier Tribunal (formally, the appeal is against the Information Commissioner himself).

To appeal, I completed the GRC1 Appeal Form, available here: https://www.gov.uk/government/publications/appeal-to-the-general-regulatory-chamber-grc1 Thankfully, this can be emailed directly to the First Tier Tribunal – a very welcome improvement compared to the County Court, which only accepts paper claim forms (unless you’re able to use a specific online claim service). I attached a separate Grounds of Appeal document to my appeal form. I have annexed this and my other tribunal documents to the end of this article. In summary, my grounds of appeal were as follows:

Grounds of appeal summary

– I. The ICO was wrong to find that TfL is entitled to neither confirm nor deny (NCND) the existence of the policies requested in Question One, particularly as TfL previously confirmed it held them;

– II. The ICO was wrong to determine that Questions Three, Four and Five relate to the ‘prevention or detection of crime’ (I later dropped this ground);

– III. The ICO was wrong to find a plausible causal relationship between the requested information and the envisioned prejudice;

– IV. The ICO was wrong to find the requested information would likely cause the envisioned prejudice;

– V. The ICO was wrong to determine that the public interest in disclosure was outweighed by the public interest in maintaining the exemption.

While TfL was aware of the ongoing appeal, it did not initially apply to be joined to it. Therefore, only the ICO initially sought to defend my appeal through a written Response.

The first hearing took place through Cloud Video Platform (CVP), an online conference platform managed by HMCTS (similar to e.g. Microsoft Teams). The ICO didn’t turn up, so I was able to present my submissions to the Tribunal judge (a KC) and the two Tribunal members (laypeople who sit on the Tribunal decision-making panel for the case alongside the judge).

At the hearing, the Tribunal panel were also able to see that TfL had still published its original response to my FOI request on its website, where it confirmed it held the information I requested in Part 1 (the policies).

The Tribunal judge was impressed with my submissions, referring to them as “competent and comprehensive” (thanks!) and even asking if I was a lawyer (not qualified just yet!). The Tribunal then decided it was appropriate to join TfL to the proceedings, allowing them a chance to respond at another hearing.

After TfL were joined to the proceedings, they chose to instruct a barrister from 11KBW. We exchanged a number of documents containing our submissions on the case.

It then emerged that TfL had initially approached Part 1 of my request by splitting it into two ‘components’. Their initial response supposedly approached Part 1 by answering the 1st component, which it says was simply for information as to whether TfL had the policies. TfL’s response was that it did hold this information (that is, it held information on whether it had the requested policies) but would not disclose it. Additionally, TfL entirely failed to explain its initial approach when TfL first revised their position, causing further confusion during a significant part of the appeal process.

I found this to be a convoluted and unhelpful approach to part 1 of my request, which was simply structured according to S.1 FOIA. Furthermore, TfL’s response was, in my view, very likely to cause any reasonable reader to conclude that it did hold the policies, but wasn’t disclosing them.

It is unsurprising, therefore, that it turned out the ICO had criticised TfL for its impractical approach to responding to Part 1 during the ICO’s investigation into my complaint. The ICO held that TfL was obliged to inform me whether the policies within scope of Part 1 were held, pursuant to S.1 FOIA.

TfL claimed this was the reason it later revised its position to NCND, but TfL claimed that the substance of TfL’s answer was the same: it would not say whether it held any policies. TfL then decided to remove its initial response from its website because of this ambiguity.

TfL also provided a number of closed submission documents and a closed witness statement to the Tribunal, which I was not able to see. This is because they could reveal the information I requested. Nevertheless, it does put an individual appellant at a disadvantage, because they are not able to respond directly to any arguments made within the closed submissions.

The second hearing

The second hearing was mainly intended to allow TfL to give evidence and turned out to be a lot messier than the first. TfL’s solicitor and barrister attended the Cloud Video Platform hearing, as well as some other TfL workers. TfL also called a witness, their Director of Security, Policing & Enforcement, who also joined the hearing. However, many people experienced technical issues joining the hearing, meaning it only started around an hour after it was scheduled.

I was then able to ask the witness a few questions, but ultimately I had to cut this short due to a lack of time. A large part (around 1 hour) of the hearing was closed, with only TfL allowed to be present to elaborate on the closed evidence. I was not allowed to listen in on this, so I don’t have the details of what was said. However, I was provided with a ‘gist’ of the closed part of the hearing, setting out broadly what topics were discussed. This included, inter alia, ‘Operation Silverback’, which TfL claims supports its view that prejudice was likely to be caused.

Because of further technical difficulties, there was only limited further time after the closed evidence part of the hearing. Therefore, both TfL and I were allowed to make brief closing submissions and were then given a further opportunity to make written submissions after the hearing.

Ultimately, the Tribunal unanimously held that TfL was allowed to withhold the information I requested, because of the closed evidence provided by TfL’s witness. While this is a disappointing outcome for transparency and openness, it is satisfactory that the Tribunal unanimously made its decision, meaning it is likely there was strong closed evidence (the details of which I don’t know). The full judgment can be found here: https://caselaw.nationalarchives.gov.uk/ukftt/grc/2024/778

Nevertheless, I am satisfied that TfL has been effectively held to account for its position through my robust appeal, which they ended up taking seriously. The appeal has also turned out to be important for clarifying TfL’s initial response to part 1 of my request regarding the policies.

Of course, readers will draw their own conclusions from the substantial efforts TfL has gone towards to keep the information a secret. In my view, it’s likely that TfL is failing to make full use of the legal mechanisms available to it to recover compensation from graffiti offenders. Internationally speaking, some railway companies publish annual information about the amount of compensation they recover from graffiti offenders. It is very telling that TfL refuses to do the same: ultimately the repair bill is likely being footed by taxpayers and fare-paying passengers. I would encourage TfL to be transparent about its compensation recovery figures, reassuring taxpayers and fare-paying passengers that they are not unnecessarily bearing the burden of those damaging the railway.

In the meantime, I am making progress with obtaining part of the information I requested through alternative mechanisms. I will update this post once I complete this process.

TfL had no further comment to make and stated they were content with the contents of this publication.

Annex – Appeal documents

Some of these documents have been edited to remove unnecessary personal data etc. I have not published documents drafted by TfL or the ICO.

  1. Grounds of Appeal
  2. Appellant’s Reply
  3. Appellant’s Response to TfL’s submissions
  4. Appellant’s Written Submissions (following the hearing)