Despite the Ministry of Justice releasing some helpful but limited data relating to judicial reviews, it is hard to know who wins, and who loses, before the Administrative Court. In an attempt to find out the answer, I designed and populated a dataset covering all reported cases handed down by the Administrative Court in 2024. The dataset, when properly analysed, should help us map out success rates before the Court. Further, it might help us answer the question of whether extra-judicial factors play a part in determining case outcomes: do the ‘haves’ hold an advantage over the ‘have nots’? Does the judge hearing a case really matter? In this post, I explain and explore my dataset and my findings, and offer some suggestions as to what needs to be done to verify or strengthen my tentative conclusions.
I am interested in capturing how the Court treats public bodies as compared to private persons. I use these terms, rather than “claimant” and “defendant”, deliberately: most cases before the Administrative Court concern a private party claimant against a public authority defendant. But this does not accurately capture all cases. For example, the Court considers multiple appeals from lower courts, tribunals and other decision-makers: sometimes the appealing party in such cases is the public authority, rather than the original claimant (who now becomes the respondent). So instead of measuring whether claimants or defendants win or lose, I instead seek to determine whether public bodies win or lose. “Private persons” is an inelegant term for ‘the party which is not a public body’ (and does not exclusively include purely private persons – this category includes, for example, associations, campaign groups and companies).
The National Archives report a total of 581 cases handed down by the Administrative Court in 2024. However, not all of these cases proved useful for present purposes. I filtered the dataset to exclude two kinds of cases: firstly, those cases which do not feature a private person against a public body (in some other cases, for example, both claimant and defendant are public authorities). Secondly, the dataset was filtered to exclude any case where there was no identifiable ‘win’ or ‘loss’ in the judgment. For example, the Court published a number of judgments which were effectively case management decisions, or which were solely concerned with the appropriate venue at which the case was to be heard. Where there was no obvious ‘win’ (permission granted, liability found, etc) or ‘loss’ (permission refused, no liability found, etc) to be found in a case, these cases were excluded from the dataset too.
71 cases were therefore removed from the dataset (35 cases did not feature a public/private body dispute, 38 did not involve a win or a loss, and 2 cases were disqualified on both grounds), leading to 510 cases of utility in the modified dataset. Each of these 510 cases was manually read, recorded and coded in the dataset. Data about each case was manually recorded, across a number of variables; the results were then analysed according to different metrics (see below).
It is important to note that whilst these 510 cases provide a relatively representative dataset regarding reported Administrative Court cases, reported cases represent only a fraction of Administrative Court applications. Many, many cases ‘drop out’ of the process at an early stage and only a relatively small proportion of cases will make it to an actual hearing – a significant number of applications will be withdrawn or settled before this point. Further, only a selection of cases heard will be reported in a judgment, particularly if permission is refused. Different judges adopt different practices when it comes to reporting (e.g.) permission refusals and costs decisions.
How Often do Private Persons Win?
With all of these caveats in mind, how often did private parties win in cases against public bodies? The data suggests that private parties enjoy a 33.52% success rate overall, across the dataset. In other words, private parties win around one-third of the time, whereas public bodies win around two-thirds of the time. I did not record success rates according to subject area – trying to do so would involve some very subjective measurements – but I did record success rates in cases invoking rights under the ECHR. In these human rights cases, the success rate was just 20%.
Does the Type of Public Body Involved Make a Difference?
Whilst I was unable to record success rates according to subject area or grounds of review, I was able to record which type of public body was involved in each case and, thus, how often each defendant won or lost. As can be seen, there was an interesting variance between the success rates of various public bodies. If we take the 34% figure as our starting point, representing the average overall win rate, it can be seen that cases against lower courts and tribunals, and cases against prisons or the Parole Board, are especially likely to succeed. By contrast, cases against HMRC, the CPS / DPP, and against Ombuds, are especially likely to fail.
- Prison / Parole Board: 70% of cases won by private persons
- Lower courts: 42% of cases won by private persons
- Central Government: 36% of cases won by private persons
- Local Authority: 34% of cases won by private persons
- Regulatory Bodies: 29% of cases won by private persons
- Requesting State: 26% of cases won by private persons
- Government Agencies: 25% of cases won by private persons
- HMRC: 19% of cases won by private persons
- CPS / DPP: 15% of cases won by private persons
- Ombuds: 14% of cases won by private persons
Some of these results might also tell us something about an associated area of law – for example, requesting states necessarily only defend extradition appeals, HMRC cases will typically involve tax law, and so on. Some of these results are interesting – it may be unsurprising to see that appeals and reviews against ombuds are rarely successful, because such challenges essentially involve claimants trying to take a second bite of the cherry (but the same can also be said for most extradition cases, and cases against lower courts, where the success rate is much higher). It would appear that there is little difference between local and central government authorities, when it comes to success rates. And cases against prisons and the Parole Board appear as a major outlier, as a site where many successful challenges take place. It is worth noting that many of the successful cases in this context were Osborn-style challenges to the denial of a prisoner’s oral hearing, or Sneddon-style challenges to the decision not to move a prisoner to open conditions.
Does the Involvement of KCs Make a Difference?
One interesting question is whether the presence of King’s Counsel makes a difference to success rates – in other words, does employing more experienced (and more expensive) legal representation increase a party’s likelihood of succeeding before the Administrative Court? The answer appears to be, to some extent, “yes”.
Private persons employing a KC succeeded before the Court in 40% of cases. Private persons without a KC succeeded before the Court in 30% of cases. This immediately suggests that KC’s make some difference to case outcomes. But the position becomes starker when we look at cases where only one side employs a KC. In cases where private persons are represented by a KC, and the public body is not, the private persons’ success rate rises to 43%. And in cases where it is the public body which is represented by a KC, and the private persons are not, the success rate drops to just 17% (in other words, public bodies employing a KC in cases where their opponents do not win 83% of the time).
This suggests that KCs make a difference, especially when only one side uses them. But we should be wary; I have shown that there is a correlation between KC presence and success rates, but not causation. Perhaps it is the case that KCs are better at sifting strong cases from weak ones, and thus will only choose to appear in cases where the legal merits are stronger anyway. More work needs to be done, but on this preliminary basis, it appears that KCs do make a difference. This may be good news for legal professionals, who can equate more experience with greater success, but given how expensive it is to engage a KC, this may raise some issues from an access to justice perspective.
Does the Type of Judge Sitting Make a Difference?
A case before the Administrative Court might be heard by one of four different kinds of judge. Firstly, it may be heard by a full-time High Cout Judge (404 cases in the modified dataset). Secondly, it may be heard by a Deputy High Court Judge (72 cases in the modified dataset). Thirdly, it may be heard by a Circuit Judge (54 cases in the dataset). Fourthly, it may be heard by a judge from the Court of Appeal, especially if the Court is sitting as a Divisional Court (48 cases in the modified dataset).
Does it matter which kind of judge hears your case? Perhaps surprisingly, the answer might be “yes”. In the modified dataset, cases heard by a full-time High Court Judge led to a win for private persons just 30% of the time (lower than the average 34% success rate overall). The success rate before Deputy High Court Judges was significantly higher, at 42%. The success rate before Circuit Judges was also higher than average, at 39%.
What might explain this phenomenon? Firstly, less senior judges such as Deputy High Court Judges and Circuit Judges might be given easier cases in practice, which translate into straightforward wins for private parties. Alternatively, it might be the case that part time or less senior judges might still retain a degree of sympathy for underdog parties, and may be less ‘battle worn’ and ‘case hardened’ than their High Court Judge counterparts. Full-time High Court Judges may also be vying for promotion, and may therefore unconsciously ‘fall in line’ with more conservative appellate courts a little more.
Matters are made more complicated by the fact that the success rate for private persons before Court of Appeal judges (on Divisional Courts) sits at 43%. If High Court judges are to be cast as stodgy, unfeeling members of the legal establishment, the same should be true – indeed, it should be more true – of Court of Appeal judges. It may be the case that High Court judges are assigned hard cases, and Court of Appeal judges high-profile cases. But more work needs to be done to determine whether the data presented here is reliable, and if so, what might explain the differences between High Court Judges and other judges.
Does the Specific Judge Sitting Make a Difference?
Finally, a case before the Administrative Court may be heard by one (or more) of a large roster of judges – this raises the question of whether individual judges might have sympathies for certain parties rather than others, or may be more likely than others to find in favour of private persons. We are often told that it is important to “know your judge” – is this really true? When it comes to other courts, including UK courts, “party sympathy” is a salient measure of judicial behaviour. Is the same true on the Administrative Court?
Preliminary evidence suggests that this might be the case. I should stress, at this point, that the results below come from a limited and partial dataset, which means any conclusions drawn, particularly about individual judges, should be taken with extreme caution. Nevertheless, the following “success rates” can be drawn (from judges deciding 10 or more cases in the modified dataset):
- Linden J: 64% of cases won by private persons
- Hill J: 50% of cases won by private persons
- Sheldon J: 45% of cases won by private persons
- Julian Knowles J: 35% of cases won by private persons
- Lang J: 34% of cases won by private persons
- Eyre J: 33% of cases won by private persons
- Holgate J: 27% of cases won by private persons
- Chamberlain J: 20% of cases won by private persons
- Farbey J: 20% of cases won by private persons
- Fordham J: 18% of cases won by private persons
- Kerr J: 18% of cases won by private persons
- Mould J: 10% of cases won by private persons
If this data is representative – and that is a big if – a private person may well be pleased to find that their case is being heard by, say, Linden J rather than, say, Mould J. But the dataset governs just one year of Administrative Court decisions. It is possible, indeed probable, that the data is being affected by numerous messy and unmeasured variables. For example, each judge has their specialist areas and different judges are allocated different kinds of cases. But it remains striking, and provocative, to suggest that different individual judges might take different approaches to private persons and public bodies than others. More research in this area is needed to determine whether these conclusions can be stated with any degree of confidence.
Conclusions and Implications
This post has explored who won and who lost before the Administrative Court in 2024. More work is required to confirm, solidify or rebut the findings I have described above, perhaps looking at different variables, and covering different years. Yet these results tentatively show, firstly, that in cases where private persons are pitted against public bodies, the latter is twice as likely to win. They also suggest that the presence of King’s Counsel makes a difference. The type of judge deciding the case makes a difference. And, most tantalisingly of all, perhaps the specific judge deciding a case makes a difference too.
The author would like to thank Anurag Deb, Gabriel Tan, and the editors of the UKCLA Blog for their helpful comments on an earlier draft.
Lewis Graham, University of Cambridge
(Suggested citation: L. Graham, ‘Who Wins and Who Loses Before the Administrative Court?’, U.K. Const. L. Blog (14th July 2025) (available at https://ukconstitutionallaw.org/))
