RETHINKING THE STORY OF CART V UPPER TRIBUNAL & ITS IMPLICATIONS

FOR ADMINISTRATIVE LAW

Dr Joanna Bell*

ABSTRACT

An influential set of narratives are beginning to emerge according to which the Supreme Court
decision in Cart v Upper Tribunal marks a major departure from existing judicial approaches to
administrative law. This paper challenges these narratives, arguing that they overstate the sense of
novelty inherent in Cart as well as overlook a number of important seams of continuity with the
past inherent in the case. This analysis is important, not only because it enhances understanding of
a recent major case. Reflection on Cart and its relationship with the past brings to light three
important, but sometimes overlooked, themes which have long characterised administrative law.
These themes, in turn, both give rise to some highly important and challenging questions about
how legal scholars ‘do’ administrative law and emphasise the importance for those working
within the generalist administrative law tradition of ‘taking legislation seriously.’

Keywords: administrative law, judicial review, public law, doctrine, legislation, legal
reasoning

Seven years have gone by since the Supreme Court put in place the first major linchpin 1 in
defining the legal relationship between the ordinary courts and the newly constituted tribunal
system in the important case of Cart.2 In this time a considerable body of commentary has
accumulated on the case3 and underlying much of it is a strikingly clear theme: the story of

1
* Fellow and College Associate Lecturer in Law at St John’s College, University of Cambridge and Affiliated
Lecturer in Law, Cambridge Law Faculty ([email protected]). With many thanks to Thomas Adams, Professor
Lionel Bentley, Professor David Feldman, Professor Liz Fisher, Voraphol Malsukham, Alistair Mills, Adam
Perry and the participants at the Centre for Public Law Seminar held at the University of Cambridge in February
2018 for their extremely helpful and thoughtful comments on an earlier version of this article. Thank you also to
the anonymous reviewers for their extremely useful suggestions. Any errors and omissions of course remain my
own.

Other important cases have followed. See especially: R (Jones) v First-tier Tribunal [2013] UKSC 19, [2012] 2
AC 48; R (Great Yarmouth Port Co Ltd) v Marine Management Organisation [2013] EWHC 3052
(Admin), [2014] ACD 23; Revenue & Customs Commissioners v Pendragon Plc [2015] UKSC 37,
[2015] 1 WLR 2838.
2
R (Cart) v Upper Tribunal [2011] UKSC 28, [2012] 1 AC 663.
3
In addition to the literature cited elsewhere in this article see for example: Philip Murray, ‘Judicial
Review of the Upper Tribunal: Appeal, Review and the Will of Parliament’ (2011) 70(3) CLJ 487;
Sarah Craig, ‘Judicial Review: How Much is Too Much? A View of Eba, Cart and MR (Pakistan) from
the Asylum and Immigration Perspective’ (2012) 16 Edin LR 223; Tim Buley, ‘R (Cart) v Upper
Tribunal; R (MR (Pakistan)) v Upper Tribunal’ (2012) 26(1) JIANL 68; Janina Boughey & Lisa
Burton Crawford, ‘Reconsidering R (Cart) v Upper Tribunal and the Rationale for Jurisdictional Error’
[2017] PL 592.
Cart is seemingly coming to be seen by many academics as a tale of considerable change. 4
Commentators have, that is, tended to explain the significance of Cart as lying in its potential
to mark a major change of direction in the judicial approach to administrative law. As will be
seen, the precise nature of the change identified has varied. Some commentators, for instance,
have conceptualised the significance of Cart as the embracement of a new principle of
‘proportionate dispute resolution’5 while others have seen Cart’s significance as, much more
grandly, hailing the judicial ‘abandonment of doctrine’6 and the calling in of a new era of
administrative law in which ‘raw pragmatism’7 and discretionary judgment will rule.8
Regardless of this variation, the clear, overarching theme in this commentary is one of major
departure from that which has gone before.9
A core aim of this article is to offer a counter to these emerging narratives. What it seeks to
show is that while there are certainly senses in which the Cart decision entailed novelty – a
Supreme Court case surely always arises within a legal grey area and so an element of
novelty is unescapable – much of the commentary which has accumulated to date has tended
to overstate the extent to which this is so. There are, furthermore, a number of important
seams of continuity with the past running through Cart which have been significantly
overlooked.
This argument will be developed across three parts. Part 1 will offer a brief overview of the
issues and decision in Cart. Part 2 will then explain and critically analyse three distinct senses
in which it has been suggested that the significance of Cart lies in substantial departure from
the past. The first focuses on the outcome of the case, the second of the court’s reasoning
process and the third on the theoretical implications of the decision. In each case the
argument of the article will be the same: the significance of the change entailed by Cart is
overstated by these narratives. Part 3, finally, will then build on the earlier discussion by
explicitly drawing out three themes which it will be argued have characterised administrative
law case law long before Cart and of which the case can be seen as an important
continuation. These are, firstly, the ever-changing nature of administrative machinery and
judicial creativity in responding to such changes. Secondly, the important role played by the
detail and underlying policies of the specific legislation at the background to a case in
shaping judicial approaches to review. 10 Thirdly, the limited utility judges find in general
4
On the importance of stories in law and legal scholarship see Satvinder Juss & Maurice Sunkin,
Landmark Cases in Public Law (Bloomsbury, 2017), especially the foreword and chapters 1 and 4;
Stephen Sedley, Lions Under the Throne: Essays on the History of English Public Law (CUP, 2015),
chapter 1.
5
Mark Elliott & Robert Thomas, ‘Tribunal Justice and Proportionate Dispute Resolution’ (2012) 71(2)
CLJ 297.
6
Christopher Forsyth, ‘Blasphemy Against Basics: Doctrine, Conceptual Reasoning and Certain
Decisions of the UK Supreme Court’ in John Bell, Mark Elliott, Jason Varuhas & Philip Murray (eds),
Public Law Adjudication in Common Law Systems: Process and Substance (Hart 2016), 155.
7
Philip Murray, ‘Process, Substance and the History of Error of Law Review’ in John Bell, Mark
Elliott, Jason Varuhas & Philip Murray (eds), Public Law Adjudication in Common Law Systems:
Process and Substance (Hart 2016), 108.
8
Forsyth (n 6), 155.
9
Interestingly, the same theme has not tended to characterise the written reactions to the case by
barristers to the same degree. See for example Christopher Knight, ‘Book Review: Public Law
Adjudication in Common Law Systems: Process and Substance’ (2016) 21(4) JR 317, 319.
10
A theme drawn out prominently in the writings of Trevor Allan. See especially Trevor Allan, ‘The
Constitutional Foundations of Judicial Review: Conceptual Conundrum or Interpretative Inquiry?’
(2002) CLJ 87; Trevor Allan, ‘Doctrine and Theory in Administrative Law: An Elusive Quests for the
doctrinal tools such as the distinction between issues of construction, application and fact-
finding11 in supplying concrete answers to legal questions concerning the proper approach to
review in a given case12 but which play an important structuring role in the leading textbooks.
Before turning to these arguments, one obvious question which may be in the mind of the
reader ought to be raised and explicitly addressed: why does Cart warrant the lengthy and
detailed focus of a full article? The answer to this question begins to emerge from the above
and has two parts. Firstly, critical reflection on Cart and the stories which are beginning to be
told about it is an important exercise not only because it enhances our understanding of a
major case. A careful exploration of the relationship between Cart and the case law and legal
ideas which came before it is illuminating also because it teaches us something much broader
about administrative law. Thus, as explained, the focus of part 3 of this article is on extracting
three wider themes which it will be emphasised have characterised administrative law case
law long before the Cart case. Each of these themes is deeply important, but also sometimes
overlooked in legal scholarship.
Secondly, the extraction of these lessons about administrative law in turn gives rise to some
challenging and extremely important methodological questions about how we ‘do’
administrative law. The conclusion will raise and consider one such question in particular: if
the themes drawn out in part 3 of this article truly have characterised administrative law for a
long time what does this mean for generalist administrative law? Is the unescapable
conclusion of this analysis, more particularly, not that the commitment to thinking of
administrative law as a set of general legal structures which apply across all areas of
administration is doomed either because there are in reality a set of specific administrative
laws (tribunal law, housing law, social security law, etc) or because administrative law
ultimately collapses into a ‘wilderness of single instances?’ 13 The conclusion of this article is
not, of course, the space for detailed exposition of this very probing and difficult question.
Indeed, one aim of this article is to generate discussion of it rather than to provide a
conclusive answer. The article will, however, end by offering two important reflections.
Firstly, there are a number of considerations which can explain why it is possible both to
embrace the themes drawn out in part 3 and to believe that there continues to be an important
role for generalist administrative law. Secondly, however, one of the methodological
implications which surely emerges from this article is the importance of building within
generalist administrative law a greater commitment to taking legislation seriously. 14
Limits of Jurisdiction’ [2003] PL 429.
11
A taxonomy which, as will be discussed below, plays a major role in many textbooks: see for example
Mark Elliott & Jason Varuhas, Administrative Law (5th ed, OUP 2017), chapter 2. See further Timothy
Endicott, ‘Questions of Law’ (1998) 114 LQR 292.
12
A point which has been made before, perhaps most memorably by Lord Denning in Pearlman v
Keepers and Governors of Harrow School [1979] QB 56 (CA), 70. See also Jones (n 1), especially
Lord Carnwath’s judgment.
13
A phrase famously appearing in Lord Tennyson’s Aylmer’s Field (1793) and used often in legal
scholarship. See especially DM Gordon, ‘The Relation of Facts to Jurisdiction’ (1929) LQR 458, 459;
Christopher Forsyth (eds), The Golden Metwand and the Crooked Cork: Essays in Honour of Sir
William Wade QC (OUP 1998), preface.
14
This form of expression is of course a twist on the well-known title Ronald Dworkin, Taking Rights
Seriously (Harvard University Press 1977). It is not uncommon for twists of this kind to be placed on
this form of expression in legal scholarship. See for example: Denis Galligan Law in Modern Society
(OUP 2007), 4 and Elizabeth Fisher, Bettina Lange, Eloise Scotford & Cinammon Carlarne, ‘Maturity
and Methodology: Starting a Debate about Environemntal Law Scholarship’ (2009) 21(2) JEL 213, 237
which speak about the importance of ‘taking law seriously.’ See also Elizabeth Fisher, Risk Regulation
 1. A BRIEF OVERVIEW OF CART

It is important to begin the analysis in this article with a brief but clear overview of the issue
which arose, and the conclusion reached, in Cart. Cart arose against the background of the
most comprehensive reorganisation of the English tribunal system which has ever taken
place.15 An important part of the restructuring project was the creation of a new institution –
the Upper Tribunal (‘UT’)16 - which exercises a series of functions including that of hearing
appeals from an array of specialist First-tier Tribunals (‘FtTs’). 17 Cart concerned a single
important legal question: say an individual, dissatisfied with the conclusion reached by a FtT,
applies for permission to appeal to the UT and is refused, to what extent, if any, should
judicial review of the permission-refusal decision be available?
One of the possible solutions to this question urged by the applicant appealed to none of the
judges at any level.18 According to this solution, judicial review should be available in the
ordinary way and on the full array of legal grounds. The reason why this option was rejected
is not difficult to find: Parliament had elected to entrust the decision as to whether an
application for permission to appeal disclosed an arguable legal error in an FtT’s reasoning to
the UT.19 It had been open to Parliament to provide for an additional right of appeal of
permission-refusals to the Court of Appeal, but Parliament had chosen to explicitly and
specifically exclude this possibility.20 In light of this, and in light of the practical assimilation
of the judicial approach to appeal and review in recent decades, 21 the provision of ordinary
judicial review would amount to a direct contravention of Parliament’s allocation of
functions.
Having accepted, therefore, that some kind of limit had to be placed on judicial review, if it
was to be available at all, the core question the court was required to address was which legal
device should be drawn on for that purpose. The reasoning process of which the Supreme
Court made use to answer this question is highly interesting and will be discussed in detail
below. For now, however, what matters is the conclusion reached by the court. The Supreme
and Administrative Constitutionalism (Hart, 2007), 36 which emphasises the importance of ‘taking
seriously legal reasoning.’
15
Tribunals, Courts and Enforcement Act 2007 (‘TCEA’). For background see especially The Report of
the Committee on Administrative Tribunals and Enquiries (1957) (‘Franks Report’); Tribunals and
Inquiries Act 1958; Sir Andrew Leggatt, Tribunals for Users: One System, One Service (March 2001)
(‘Leggatt Report’); Robert Carnwath, ‘Tribunal Justice – A New Start’ [2009] PL 48.
16
TCEA (n 15), s3.
17
Note the UT’s functions also include exercising a limited ‘judicial review jurisdiction’: ibid, ss15-21.
18
R (Cart) v Upper Tribunal [2009] EWHC 3052 (Admin), [2010] 2 WLR 1012; R (Cart) v Upper
Tribunal [2010] EWCA Civ 859, [2011] QB 120.
19
Note that the TCEA (n 15), s11 itself does provide much guidance as to the test to be applied by the
UT. The tendency, however, has been to interpret it as one of arguability. See especially Upper
Tribunal Immigration and Asylum Chamber, ‘Guidance Note 2011 No 1: Permission to Appeal to
UTIAC’ July 2011 (amended September and July 2014) (available here:
https://www.judiciary.gov.uk/wp-content/uploads/2014/08/guidance-note-no1-permission-to-appeal-to-
utiac.pdf), [12]. See further Civil Procedure Rules 1998 (‘CPRs’), Part 52, rule 6(1)(a) which makes
arguability the applicable test in ‘first appeals’ to the ordinary courts.
20
TCEA (n 15) 2007, s13(8).
21
See especially E v Secretary of State for the Home Department [2004] EWCA Civ 49, [2004] QB
1044.
Court’s unanimous preferred solution was to retain the availability of judicial review but to
invoke, what are known as, the ‘second-appeals criteria’ by way of limitation. The origins of
these criteria will also be discussed below, but very briefly they require that permission to
proceed is to be refused unless the applicant is able to demonstrate either that:
(a) The [case] would raise an important point of principle or practice, or
(b) There is some other compelling reason for the [court] to hear it.22
Following Cart, then, when a disappointed party seeks judicial review of a permission-refusal
decision the High Court judge at the leave stage will grant permission to proceed only if she
is satisfied that the case falls within one of these categories 23 and that the applicant is able to
make out an arguable error of law.24 If permission is granted, an amendment to the Civil
Procedure Rules made following Cart25 provides for two options: either the permission to
appeal application is directly remitted back to the UT for reconsideration or the parties may
request a substantive judicial review hearing.26 If the latter option is favoured, it would seem
that the High Court will decide the issue by applying the ordinary grounds of judicial
review.27

2. CART’S STORY AS A TALE OF CHANGE?

As noted in the introduction, the Cart decision has attracted a considerable body of
commentary and underlying much of this commentary is a strikingly clear theme: Cart’s
story is seemingly coming to be seen by many as a tale of significant change. That is, a series
of narratives are emerging according to which Cart’s significance lies in its marking a turning
point in the judicial approach to administrative law. This part of the article is focused on three
versions of these emerging narratives. The first focuses on the outcome of the case, the
second on the court’s reasoning process and the third on the theoretical implications of the
decision. The aim in each instance is to both identify the sense in which it has been suggested
that Cart entails a change of direction and to show why the case’s relationship with the past is
both more nuanced and characterised by greater continuity than these accounts suggest.

22
Access to Justice Act 1999 (‘AJA’), s55.
23
Note that there has been some ambiguity as to whether the test to be applied at the permission stage is
whether the second-appeals criteria are satisfied or whether they are arguably satisfied. The case law,
howevr, seems to be moving more firmly towards the former option. See especially discussion in
Joanna Bell, ‘The Relationship between Judicial Review & The Upper Tribunal: What Have the Courts
Made of Cart?’ (forthcoming in Public Law); R (Essa) v Upper Tribunal [2012] EWHC 1533 (QB),
[2012] 3 CMLR 26, [2]; R (HS) v Upper Tribunal [2012] EWHC 3126 (Admin), Official Transcript,
[35]; R (AA (Iran)) v Upper Tribunal [2013] EWCA Civ 1523, Official Transcript, [5].
24
CPRs (n 19), rule 7A(7).
25
Ibid, rule 7A. See discussion in Richard Drabble & Christopher Jacobs, ‘New CPR to Implement Cart’
(2013) 18(1) JR 113.
26
Ibid, rule 7A(9).
27
Note there has, again, been some ambiguity in the case law concerning the proper approach to be taken
at the substantive hearing stage. The emerging approach, however, seems to be that the Administrative
Court will proceed simply by applying the grounds of review in the ordinary way. See discussion in
Bell (n 23) and especially R (G) v Upper Tribunal [2016] EWHC 239, [2016] 1 WLR 3417.
 A. NOVELTY IN THE OUTCOME?

In explaining the first sense in which it has been suggested that there is novelty inherent in
the Cart decision, it is helpful to begin by noting that the two commentators who have most
strongly criticised the Cart decision are probably Philip Murray28 and Christopher Forsyth.29
At least part of the complaint of these authors is that the solution preferred by the Supreme
Court entailed a major step away from pre-Cart administrative law. The argument can be
explained as follows: as discussed above, the core question which arose in Cart was which
legal technique the Supreme Court ought to draw on for the purpose of restricting the
operation of judicial review. For both Murray and Forsyth, invoking the second-appeals
criteria entailed reaching far beyond the established toolbox of administrative law. By
contrast, for both of these commentators there were certain long-established orthodox
techniques for which the Supreme Court could and should have reached.
Take, for instance, Murray’s argument.30 For Murray, a better solution to Cart was for the
Supreme Court to reach back to the 1841 decision of Bolton.31 As Murray has explained in a
series of extremely thorough and comprehensive articles, 32 the Court of the King’s Bench in
this case developed a particular approach to the distinction between jurisdictional and non-
jurisdictional error. At the heart of this approach were considerations of timing. The idea can
be simply put: if an administrative decision-maker made an error which related to something
it could and ought to have known before commencing its inquiry, the error was to be
regarded as jurisdictional. If, however, having legitimately commenced an inquiry, a public
authority was alleged to have fallen into error the error would be regarded as non-
jurisdictional. For Murray, this timing-based distinction between different kinds of error
provided a mechanism on which the Supreme Court could have drawn in Cart; the general
availability of judicial review could have been maintained but the Bolton distinction invoked
as a way of restricting its operation.
There is certainly an element of truth in the suggestion that there is novelty in the Supreme
Court’s invocation of the second-appeals criteria in Cart. The key point to note, in particular,
is that the second-appeals criteria were not originally a judicial creation but a Parliamentary
one. Pre-Cart, they had put to use in only two places in English law. Firstly, the Access to
Justice Act 199933 introduced them as the test to be applied at the permission stage when an
applicant seeks to appeal a decision from a lower court to the Court of Appeal. Secondly,
Parliament put the criteria to use again in the legislation governing the newly constituted
tribunals system. Where, in particular, a party seeks permission to appeal a UT decision to the
Court of Appeal her application is to be determined by reference to these criteria. 34 One sense

28
Murray (n 3); Murray (n 7); Philip Murray, ‘Escaping the Wilderness: R v Bolton and Judicial Review
for Error of Law’ (2016) 75(2) CLJ 333.
29
Forsyth (n 6).
30
Note that Forsyth (n 6) recommends an alternative approach. For Forsyth, the Supreme Court should
have reached back to Re Racal Communications Ltd [1981] AC 374 and Lord Diplock’s drawing of a
distinction in that case between administrative decision-makers and inferior courts.
31
R v Bolton (1841) 1 QB 66
32
Murray (n 7); Murray (n 28).
33
AJA (n 22), s55.
34
TCEA (n 15), s14B.
of novelty inherent in Cart’s story, then, is that this is first instance in which the courts
invoked the second-appeals criteria and put them to use in a new legal context.
To thereby characterise Cart’s story as one in which the Supreme Court went astray in
reaching beyond the established toolbox of administrative law to find a solution, however,
would be overly simplistic. The picture is rather more nuanced and two related points in
particular ought to be noted. Firstly, Cart is far from the first example of the courts being
faced with a major change to administrative machinery and looking beyond the immediate
legal context for an appropriate legal solution. Indeed, the history of English administrative
law is rich with examples of courts being faced with such changes and thinking creatively
about the solutions to be deployed in responding to them.
Bolton itself is a very good example of this point. As Murray has cogently explained, the
administrative state at the time Bolton was decided was undergoing a process of major
change.35 In the early 19th century administrative power was largely yielded, and thus most
challenges related to, a singular body – Justices of the Peace. Parliament at this time was,
furthermore, consistently pursuing a policy of relinquishing the control of the King’s Bench
over Justices of the Peace, preferring to provide for oversight by appeal in the quarter
sessions.36 In Bolton, the King’s Bench was looking for an approach to drawing the
distinction between jurisdictional and non-jurisdictional error, which had by this time come
to serve as the ‘gate-keeper’37 of certiorari, which would be sufficiently restrictive so as to
show due respect to Parliament’s policy. As explained above, the technique which the court
preferred was the ‘timing-based’ distinction between jurisdictional and non-jurisdictional
error. Importantly, however, the court did not develop this technique by building on existing
certiorari doctrine. The court, rather, borrowed from a very different legal and remedial
context: private law, where the distinction was already in use as a mode of restricting the
personal liability of administrative decision-makers in trespass claims.38 There is, then, a
close analogy between Bolton and Cart: in both instances the courts responded to a major
change in administrative machinery by borrowing from outside of the immediate legal
context.
Bolton is, furthermore, far from an isolated example of judicial creativity of this kind. 39
Another, for instance, can be found in the judicial rediscovery of the largely forgotten notion
of ‘error on the face of the record’ in Shaw40 in 1952. This judicial manoeuvre served to
resurrect an idea which had long been out of use in the context of certiorari applications 41 in

35
Murray (n 28), 356.
36
Ibid. See especially Summary Jurisdiction Act 1848; Justices Protection Act 1848; Poor Law
Procedure Act 1848; Quarter Sessions Act 1849; Summary Jurisdiction Act 1957; Summary
Jurisdiction Act 1879.
37
Ibid, 388.
38
See especially Brittain v Kinnaird (1819) 1 Brod & Bing 432; Cave v Mountain (1840) 1 Man & G
257.
39
Another example might be the introduction of the idea that affidavit evidence could be admitted in a
certiorari application where ‘jurisdictional error’ was alleged: see Murray (n 7); Murray (n 28). More
recent examples include the courts’ recent embracement of proportionality (Pham v Secretary of State
for the Home Department [2015] UKSC 19; [2015] 1 WLR 1591) and open justice (Kennedy v
Information Commissioner [2014] UKSC 20, [2015] AC 455) as techniques for regulating
administration from outside of the immediate legal context of administrative law.
40
R v Northumberland Compensation Appeal Tribunal, ex parte Shaw [1952] 1 KB 338 (CA).
41
See for example Racecourse Betting Control Board v Secretary for Air [1944] Ch 114.
order to provide an appropriate bolstered approach to review following the massive
expansion of the administrative state during the course of two world wars 42 and following the
election of the first majority labour government in 1945.43
The second important point which ought to be noted is that it is difficult to see how it is
useful to think in terms of a neat dividing line between legal mechanisms which form part of
the settled administrative law toolbox and those which fall outside of it. It rather seems clear
that this, like so many, is an issue of degree and not of kind. Take, for instance, Murray’s
suggestion that the proper solution to Cart may have been to reach back to Bolton. There is
clearly a sense in which doing so could be thought of as an exercise in working within the
existing administrative law toolbox. But there would also seem to be an equally strong sense
in which reaching backwards to Bolton would entail reaching significantly beyond
administrative law. In the first place, there has been a prevalent way of thinking according to
which the Anisminic decision put to rest the idea that there is a distinction in English
administrative law between jurisdictional and non-jurisdictional errors of law. On this view,
the distinction has not been a feature of English law since the late 1960s and certainly not
since the decision in Page44 in 1992. In the second, the timing-based approach to the
distinction developed in Bolton was but one of many attempts to draw a distinction between
jurisdictional and non-jurisdictional errors across the course of the nineteenth and twentieth
centuries.45 Against this background, a more accurate, or perhaps a more cynical way, of
describing Murray’s proposed solution would be to say that it would involve invoking just of
one of a significant number of different ways of formulating a distinction which has not
played a significant role in English law for nearly 50 years.
The key point, then, is that it seems plain that Murray’s suggested approach to resolving Cart
would have entailed drawing on a legal mechanism which could be neither straightforwardly
characterised as falling inside or outside of the existing administrative law toolbox. 46 The
same point, furthermore, might be made about the second-appeals criteria themselves.
Although, as noted, these criteria were developed by Parliament, the Court of Appeal has put
them to regular use in deciding whether to grant permission to appeal, including in relation to
appeals from the Administrative Court, since 1999. The High Court, furthermore, since
200647 had been making use of a mechanism for limiting judicial review of the permission-
42
For discussion see for example Lord Hewart, The New Despotism (London: Ernst Benn Ltd, 1929);
Cecil Carr, ‘Administrative Law’ (1935) 51 LQR 58, David Williams, ‘The Donoughmore Report in
Retrospect’ (1982) 60 Public Administration 273.
43
See discussion in Derek Fraser, The Evolution of the British Welfare State (4th edn, Palgrave
Macmillan 2009), chapter 9. The key innovations of this government include such ground-breaking
legislation as National Insurance Act 1946; National Health Service Act 1946; Coal Industry
Nationalisation Act 1946; Electricity Act 1947; Transport Act 1947; Town and Country Planning Act
1947; National Parks and Access to the Countryside Act 1949.
44
R v Lord President of the Privy Council, ex parte Page [1993] AC 682 (HL).
45
For general discussion see Paul Craig, Administrative Law (8th edn, Sweet & Maxwell 2016), 473-480.
See furthermore especially Gordon (n 13); DM Gordon, ‘Observance of Law as a Condition of
Jurisdiction’ (1931) 47 LQR 386; DM Gordon, ‘Jurisdictional Fact: An Answer’ (1966) 82 LQR 515;
B Gould, ‘Anisminic and Jurisdictional Review’ [1970] PL 358.
46
Note that a similar point could be made about Forsyth’s suggestion that the proper answer to Cart lay
in the invocation of Lord Diplock’s dicta in Racal (see (n 30). As Dyson LJ (as he was) remarked in R
(Sivasubramaniam) v Wandsworth County Council [2003] 1 WLR 475, [42], the courts have not tended
to draw upon this idea in determining the proper approach to review of inferior courts and the pre-
restructure tribunals.
47
R (Sinclair) v Lands Tribunal [2005] EWCA Civ 1305, [2006] 3 All ER 650.
refusal decisions by the Lands Tribunal48 which functioned in a remarkably similar way to the
second-appeals criteria. Dyson LJ (as he was) had even gone as far in 2010 49 as suggesting
the court might embrace the second-appeals criteria as a tool for moderating scrutiny of the
Social Security Appeals Tribunal. The invocation of the second-appeals criteria in Cart, in
other words, did not come as strongly ‘out of the blue’ as some of the literature might be
taken to suggest.

B. A NEW POSITIVE FEATURE IN THE SUPREME COURT’S REASONING?

Whereas the first sense in which it has been suggested that Cart entails significant departure
from the past focuses on the legal tool on which the Supreme Court drew, the second sense
focuses on the way in which the judges decided which tool to deploy. The suggestion, more
particularly, is that the significance of Cart lies in some new and positive feature of legal
reasoning which has not played a significant role in administrative law adjudication before.
In explaining the idea, it is helpful to begin by painting a picture of how the Supreme Court
set about finding the answer to Cart which is presented in much of the literature. 50 According
to this picture, the court reasoned as follows. Firstly, the court identified a set of
considerations which pointed in favour of providing some level of judicial review. These
considerations, roughly, constituted ‘rule of law’ considerations which place value in
ensuring scrutiny of decision-making by an independent judicial body in order to ensure
compliance with the law.51 Secondly, the court identified a set of considerations which
pointed against providing review of permission-refusal decisions. These considerations
included the virtues of finality and speed which would come with regarding the UT’s
permission-refusal decision as conclusive, making efficient use the expertise and institutional
competence of the UT, and, perhaps most controversially, the benefits that restricting judicial
review would have in terms of not adding to the Administrative Court’s already substantial
caseload. Thirdly, the court then sought to find an appropriate way of striking the balance
between these two sets of practical consideration. The second-appeals criteria were chosen
because they appealed to the Supreme Court as the best way of striking this balance; they
would safeguard against the concern that errors of law would become embedded in the
system52 whilst also ensuring that in most cases the UT’s decision would be final.
This mode of reasoning has been suggested by some to be a new feature of administrative
law. The point has been expressed in different ways, some forms of expression being more
provocative than others. Mark Elliott & Robert Thomas have put the point fairly mildly.
Cart’s significance, the authors argue, lies in the embracement of a new legal principle which
they label ‘proportionate dispute resolution.’53 In the authors’ own words:
48
Created by the Lands Tribunal Act 1949.
49
(Wiles) v Social Security Commissioners [2010] EWCA Civ 258, Official Transcript.
50
See for example Elliott & Thomas (n 5).
51
Note this consideration is probably regarded as a basic component of the rule of law by all theorists.
See for example: Joseph Raz, ‘The Rule of Law and Its Virtue’ in The Authority of Law (2nd edn, OUP
2009); Tom Bingham, The Rule of Law (Penguin 2011); John Laws, ‘The Rule of Law and the
Presumption of Liberty and Justice’ (2017) 22(4) JR 365.
52
Cart (n 4), [92]; In Re Woodling [1984] 1 WLR 348 (HL).
53
Elliott & Thomas (n 5).
 The Supreme Court’s explicit recognition of the need to attain a proportionate balance
between the cost of challenges and their importance is notable.... the explicit
invocation of efficiency and resource arguments in Cart represents an important shift
in judicial reasoning: proportionate dispute resolution has received judicial
recognition.54
Others have expressed the point in stronger language. Philip Murray, for instance, has said of
the case that its significance, or to use Murray’s own language, its ‘fundamental problem’ 55
lies in ‘the Supreme Court’s eschewal of doctrine in favour of raw pragmatism.’56
As with the suggestion that there is novelty in Cart in terms of the outcome of the case, there
is an important element of truth in this idea. The notion of striking a balance between
different sets of consideration certainly did play a striking role in their Lord and Ladyship’s
reasoning. Once again, however, to thereby suggest that Cart’s story is one of straightforward
change would be overly simplistic; the picture is rather more nuanced.
A number of points in particular are worth emphasising in this regard. The first is a
preliminary, but highly important, point: the considerations on which the Supreme Court
drew in Cart were not invoked from thin air. The court in this case did not reason by reaching
beyond the legal sources to draw on ideas from some extra-legal dimension on no other basis
than that the judges thought them to be normatively desirable. The considerations on which
the courts drew were, rather, carefully extracted from the legal materials which were at play
in the litigation including, importantly, the legislative framework at the background to the
case: the Tribunals, Courts and Enforcement Act 2007 (‘TCEA’).57
Thus take, for instance,58 the most controversial consideration on which the Supreme Court
drew: the importance of ensuring the manageability of the High Court’s caseload. This factor
was not simply ‘invented’ by the judges. The consideration was rather carefully extracted as a
core legal policy underlying the TCEA. Thus in the course of rendering judgment, the
Supreme Court placed emphasis on a series of considerations which strongly indicated that
one of Parliament’s aims in creating the TCEA was that of creating an adjudicative
mechanism capable of alleviating the substantial case law pressures which continue to afflict
the Administrative Court. Firstly, Parliament’s restructuring programme was preceded by a
series of legislative changes clearly directed at alleviating these pressures especially in the
context of immigration and asylum cases.59 Secondly, pre-TCEA materials, especially the

54
Ibid, 314-315.
55
Murray (n 7), 108.
56
Ibid.
57
TCEA (n 15).
58
Other examples could be discussed. The ‘rule of law’ concerns the Supreme Court relied on, in
particular, were drawn from two sources. Firstly, they are commonly thought to be embedded in the
common law itself: see especially R v Secretary of State for the Home Department, ex parte Leech
[1994] QB 198 (CA); R v Lord Chancellor, ex parte Witham [1998] QB 575 (CA); R v Secretary of
State for the Home Department, ex parte Simms [2000] 2 AC 115 (HL); R (UNISON) v Lord
Chancellor [2017] UKSC 51, [2017] 3 WLR 409. Secondly, the Supreme Court also clearly understood
the preservation and protection of this value as an inherent policy goal underlying the TCEA (n 15)
itself reflected in various features of the tribunal restructure including the central role that would be
played by sitting High Court and Court of Appeal judges (see Cart (n 2, [29]) and the overarching
functions of the Court of Appeal (see Cart (n 2), [48]; TCEA (n 15), ss13-14).
59
Especially Nationality, Immigration and Asylum Act 2002 (‘NIAA’).
initial governmental White Paper60 and the Leggatt Report61 which greatly influenced the
TCEA, strongly indicated that it was no part of Parliament’s aims to undo the improvements
made by these changes. As Lord Dyson put it, it was therefore clear that one of the policy
goals which to drove the TCEA was the idea that ‘there should not be a return to the... days in
immigration and asylum cases when the courts were overwhelmed with unmeritorious
judicial review claims.’62
If, in other words, it is helpful to think of the Supreme Court’s approach to deciding Cart as
being characterised by ‘pragmatism,’ we ought to think of it as a kind of ‘legal pragmatism.’
Though the Supreme Court sought to navigate an array of competing policy considerations,
each of these considerations was carefully extracted from legal materials, including especially
the TCEA, and not simply invoked from some extra-legal sphere.
With this preliminary point in mind, two other main points ought to be noted. Firstly, there is
nothing new in the idea that, in deciding on the proper approach to judicial review in a given
case, a major focus of the court will be on navigating the details of legislation and attempting
to extract the policy goals underlying it. Such a focus, on the contrary, is a pervasive and
long-standing feature of judicial reasoning in administrative law. The article will shortly
discuss two examples of well-known cases – Puhlhofer63 and Croydon64 - which neatly
illustrate this point. As will be explained, the court in each was faced with a difficult question
considering the approach to review which ought to be taken in assessing the lawfulness of an
administrative decision-maker’s conclusions. In each case, the judges understood their
primary task as being that of navigating the statutory scheme at the background of the case in
order to try to extract the policy goals underlying it so that these goals could be drawn on in
shaping the proper approach to review.
The second important point which ought to be noted builds on the first. Not only is it
common for judges to take the statutory scheme at the background to the case as their
starting-point in determining the level of judicial review which is appropriate. It is,
furthermore, not unknown for the courts, in the course of trying to navigate legislative
schemes in this way, to extract the kinds of consideration on which the Supreme Court drew
in Cart and to make use of them in shaping their conclusion. One example will suffice to
illustrate.65 Take the case of G decided in 2004.66 This case arose against the background of
the Nationality, Immigration and Asylum Act 2002, one of the major pieces of legislation
which preceded the TCEA. The 2002 Act, among other things, sought to replace judicial
review of decisions by the Immigration & Asylum Tribunal (‘IAT’) with a process by which
a High Court judge would decide on the papers whether to overturn or leave intact a
determination by the IAT. A significant focus of the Court of Appeal in deciding the issue of
60
Cart (n 4), [126].
61
Ibid. See also Leggatt Report (n 15).
62
Ibid.
63
R v Hillingdon LBC, ex parte Puhlhofer [1986] AC 484 (HL).
64
R (A) v Croydon LBC [2009] UKSC 8; [2009] 1 WLR 2557.
65
Another overlooked example of this point is Page (n 44), 704 itself. Lord Browne-Wilkinson in this
case, in particular, concluded that a more restrictive form of review should be available of the
university visitor here in part because one of the rationales underlying the existing of this adjudicative
system was its ability to produce ‘speedy, cheap and final answers to internal disputes.’ See further
Thomas v University of Bradford [1987] AC 759 (HL), 825; Patel v University of Bradford [1978] 1
WLR 1488 (Ch), 1499-1500.
66
R (G) v Immigration Appeal Tribunal [2004] EWCA Civ 1731, [2005] 1 WLR 1445.
whether judicial review of the IAT’s decisions ought still to be available was on extracting
the legislative policies underlying the 2002 Act. As Lord Phillips put it, the object of the new
process ‘was to provide a much faster procedure than that which was involved when asylum
applicants sought to challenge a refusal of permission to appeal to the IAT’ 67 and thereby to
ensure a better balance between scrutiny of decision-making and the efficient use of judicial
resources. This consideration, furthermore, played a central role in shaping the outcome of G.
Thus, at least partly on the basis of it, the Court of Appeal introduced a new restriction on
judicial review: after G, the High Court could regard it as ‘a proper exercise of [its] discretion
to decline to entertain an application for judicial review of issues which have been, or could
have been, the subject of statutory review.’68

C. NOVELTY IN CART’S THEORETICAL IMPLICATIONS?

Moving, finally, to the third sense in which it has been suggested that Cart entailed a
significant change of direction from the past. Whereas the first sense focused on the outcome
of the decision and the second on the Supreme Court’s reasoning process, this third sense
focuses on the theoretical implications of the decision. The argument, more particularly, is
that the significance of the decision lies in the Supreme Court’s rejection of a major tenet of
‘orthodox’69 administrative law theory: the central structuring role played by the idea of
‘jurisdiction.’
The argument is best understood as having two main components. Firstly, it has, to use
Philips Murray’s language, long been ‘orthodox theory’ 70 that the courts’ ‘starting point’ 71 in
deciding issues within administrative law is the House of Lords’ decision in Anisminic.72
According to conventional theory, this decision eradicated the distinction between
jurisdictional and non-jurisdictional errors and established that in any case where an
administrative decision-maker falls into error when construing a statute, 73 or into legal error
more broadly,74 the consequence will be that it steps outside of its assigned jurisdiction. This
basic idea, for authors such as Murray and Forsyth, has, certainly since the 1992 decision of
Page,75 served as the cornerstone of administrative law; it is the foundation on which the
courts have built legal doctrine.76 Thus, as Forsyth puts it, post-Anisminic ‘jurisdiction’ has

67
Ibid, [8].
68
Ibid, [26].
69
Murray (n 7), 108.
70
Ibid.
71
Forsyth (n 6), 152.
72
Anisminic v Foreign Compensation Commission [1969] 2 AC 147 (HL).
73
See especially Lord Diplock’s judgment in O’Reilly v Mackman [1983] 2 AC 237 (HL).
74
See especially Lord Irvine’s judgment in Boddington v British Transport Police [1999] 2 AC 143
(HL). See also discussion in Mark Elliott, ‘Boddington: Rediscovering the Constitutional Logic of
Administrative Law’ [1998] PL 92.
75
Page (44).
76
See William Wade & Christopher Forsyth, Administrative Law (11th edn, OUP 2017), 26.
come to serve as the ‘organising concept’ 77 of the entirety of administrative law. Or, as one
leading textbook puts it, ‘all of [modern] judicial review is about jurisdiction.’78
Secondly, against this theoretical backdrop, the significance of Cart lies in the Supreme
Court’s rejection, both explicitly and implicitly, of these basic and long-standing theoretical
foundations. Explicitly, there are passages in the judgments in Cart where the judges declined
to frame the issue in the case by drawing on the language of ‘jurisdiction.’ Lord Dyson, for
instance, expressed his agreement with De Smith’s suggestion that ‘the distinction between
jurisdictional and non-jurisdictional error is ultimately based on foundations of sand’ 79 and
therefore declined to make use of this language in expressing his judgment. Implicitly, and
perhaps more fundamentally, the outcome of the Cart case is extremely difficult to square
with these basic tenets of ‘orthodox theory.’80 Imagine, for instance that in adjudicating on a
legal issue a FtT misconstrues a statute and so falls into legal error. The UT, however, fails to
identify the error and so refuses permission to appeal. Post-Cart, it is not enough for
permission for judicial review to be granted that the applicant is able to make out an arguable
case of error of law.81 She must also convince the High Court that either the case raises ‘an
important point or principle or policy’ or that there is ‘some other compelling reason’ for the
matter to be reconsidered.82 This gives rise to a very striking possibility: there may be cases in
which the FtT has fallen into legal error in relation to which judicial review will not be
available because the requirements of the second-appeals criteria are not made out. This sits
very uneasily with the ‘orthodox’ 83 way of thinking according to which all legal errors lead to
an excess of jurisdiction and according to which it the central goal of judicial review to
ensure that administrative decision-makers remain within the boundaries of the legal powers
assigned to them.
Collectively these points lead some commentators to suggest that the significance of Cart lies
on a far grander plain than the Supreme Court’s invocation of the second-appeals criteria for
the first time, or the embracement of a principle of ‘proportionate dispute resolution.’ For
Forsyth, for instance, Cart may signify the Supreme Court’s ‘abandonment of jurisdiction as
the organising principle of administrative law.’ 84 Or, as Elliott & Thomas put it, Cart may
mark administrative law’s transition into a new ‘post-jurisdictional world.’85
As with the previous two senses in which it has been suggested that Cart entails novelty,
there is certainly an important grain of truth in this characterisation of the case’s significance.
There are, in particular, numerous examples of oft-cited judicial dicta from a handful of very
well-known judgements – the classic examples 86 probably being the judgements of Lord

77
Christopher Forsyth, ‘The Rock and the Sand: Jurisdiction and Remedial Discretion’ (2013) 18(4) JR
360, [28].
78
Elliott & Varuhas (n 11), chapter 2.
79
Cart (n 2), [111] citing Lord Woolf, Jeffrey Jowell & Andrew Le Sueur (eds), De Smith’s Judical
Review (6th edn, Sweet & Maxwell 2007), para 4-046.
80
Murray (n 7), 108.
81
CPRs (n 19),
82
Ibid, Part 54, rule 7A(7).
83
Murray (n 7), 108.
84
Forsyth (n 6), 155.
85
Elliott & Thomas (n 5), 154.
86
For further examples see Lord Denning’s judgment in Pearlman (n 12) and Lord Neuberger’s
judgment in Ahmed v HM Treasury [2010] UKSC 5, [2012] 2 AC 534.
Diplock87 in Racal88 and O’Reilly,89 Lord Browne-Wilkinson’s in Page90 and Lord Irvine in
Boddington91 - which make use of the idea that Anisminic and the eradication of the
distinction between jurisdictional and non-jurisdictional errors forms the starting-point of
administrative law. The Supreme Court’s judgment in Cart, by explicitly and implicitly
declining to make use of this idea, therefore sits in significant tension with these cases.
As with the previous senses in which it has been suggested that Cart’s story is one of
significant change, however, the picture is rather more nuanced than the narrative set out in
this part suggests. One consideration in particular is worthy of detailed discussion. 92 The key
point can be simply put: it is far from clear that the courts have ever fully and in a meaningful
way embraced the idea that a broad reading of Anisminic (as interpreted in Page), according
to which all misconstructions of a statute or legal errors more broadly result in an excess of
jurisdiction, serves as their ‘starting-point’ 93 in administrative law. The use of these ideas as a
starting-point is, rather, more commonly a feature of textbook analysis than it is of judicial
reasoning itself. There has, as a result, sometimes been a disparity between judicial
approaches to resolving certain issues within administrative law and the leading textbook
accounts of the case law; the use of the Anisminic decision as a ‘starting-point’ has often
been something crafted onto judicial decision-making by textbook analysis retrospectively,
rather than being a feature of judicial reasoning which actually played a meaningful role in
the courts’ search for an answer to the legal question at hand.
This very abstract point is best illustrated by considering a concrete example. 94 In introducing
the example it is helpful to begin with an important point recently emphasised by Sarah
Nason:95 many of the judicial review cases which come to the attention of the High Court
concern a complaint of a particular kind - an applicant comes to court in order to argue that a
public authority, tasked with applying a statutory or other legal provision to a set of facts, has
gone in some way awry. A core question which will arise in such cases is the level of rigour

87
See also Lord Diplock, ‘Judicial Review Revisited’ (1974) 33(2) CLJ 233.
88
Racal (n 30).
89
O’Reilly (n 73).
90
Page (n 44).
91
Boddington (n 74).
92
Other important points could be made. There is, for instance, a convincing argument to the effect that
the broader readings which have been conferred on the Anisminic by subsequent case law depart from
the true, and much narrower, ratio of that case. On this point see especially Timothy Endicott,
Administrative Law (3rd edn, OUP 2015), chapter 9 and David Feldman, ‘Anisminic Ltd v Foreign
Compensation Commission [1968]’ in Satvinder Juss & Maurice Sunkin, Landmark Cases in Public
Law (Hart, 2017). In addition to think point it might be added that there is reason to treat with caution
some of the judgments which advanced such broader readings. On the agenda underlying the
judgments of Lord Diplock in the 1970s and 1980s see for example T Arvind & Lindsay Stirton, ‘The
Curious Origins of 99 Judicial Review’ (2017) 133(1) LQR 9 and Jason Varuhas, ‘The Public Interest
Conception of Public Law: Its Procedural Origins and Substantive Implications’ in John Bell, Mark
Elliott, Jason Varuhas & Philip Murray (eds), Public Law Adjudication in Common Law Systems (Hart
2016).
93
Forsyth (n 6), 152.
94
Other examples of this disparity could be given. One often discussed for example is the question of
how the courts determine what the consequences of administrative unlawfulness are. On this compare,
for instance, on the one hand Elliott & Varuhas (n 11), chapter 3 with David Feldman, ‘Error of Law
and Flawed Administrative Act’ (2014) 73(2) CLJ 275; Agricultural, Horticultural and Forestry
Industry Training Board v Aylesbury Mushrooms [1972] 1 WLR 190 (QB); R (Lumba) v Secretary of
State for the Home Department [2011] UKSC 12, [2012] 1 AC 245.
95
Sarah Nason, Reconstructing Judicial Review (Hart 2017), 157.
with which judges should go about reviewing such alleged errors. Now, imagine a student of
administrative law wants to understand how the courts go about finding answers to such
questions and she consults the relevant chapters of leading textbooks in the field. These
leading textbooks will give her slightly divergent advice but there is one thing which unites
them: all suggest that the court’s ‘starting point’96 is with a broad reading of the Anisminic
decision.
Take, for instance, two leading textbooks:97 Timothy Endicott’s Administrative Law98 and
Mark Elliott & Jason Varuhas’ text of the same name. 99 Both of these texts take a broad
reading of Anisminic as their starting-point in analysing these issues.100 According to this
reading, the case ultimately established101 that any error in construing a legal provision will
result in an excess of jurisdiction. Thus, for both textbooks, when an applicant’s case raises a
question of how a legal provision should be interpreted - what its meaning is – the general
principle is that, because all misconstructions lead to an excess of jurisdiction, the court
should review the authority’s interpretation on what might be called a ‘correctness’
standard,102 though both texts recognise that this principle is subject to a list of fairly complex
exceptions.103 Importantly, both texts also draw a distinction between issues of construction
and two other kinds of issue an authority may have to resolve: issues of application 104 and
issues of fact-finding.105 This tripartite distinction plays an important structuring role in these
textbooks which in turn present a particular picture of how the courts must go about finding
answers to the legal questions in cases of this kind. According to this picture, the task of the
judge is to navigate the analytical distinction between three different kinds of issue which
might arise in the case in order to ascertain the proper standard of review which is to be
applied.
What we see in both these textbooks, and other leading textbooks in the field, 106 then, is a
similar idea: in determining the level of judicial scrutiny which is appropriate in such cases,
the courts take a broad reading of Anisminic (as interpreted in Page) as their starting-point.
From this starting-point, the courts will then navigate a series of analytical distinctions
between different kinds of issue or error which follow out from the case in order to ascertain
the appropriate approach to review.
One thing which is extremely striking about analyses of this kind, however, is that the picture
they paint of how judges find answers to legal questions in cases of this kind is not reflective
of the legal reasoning which judges actually offer for their decisions. The courts in deciding
the major cases discussed in these analyses, that is, do not usually take a broad reading of
96
Forsyth (n 6), 152.
97
Other examples which could be considered include Craig (n 45), chapters 16 and 17 and Mark Elliott
& Robert Thomas, Public Law (3rd edn, OUP 2017), chapter 12.
98
Endicott (n 92), chapter 11.
99
Elliott & Varuhas (n 11), chapter 2.
100
Note, however, the difference of tone in which the authors do so. Timothy Endicott, in particular, is
highly critical of these broad readings of Anisminic.
101
Note again that, for Endicott, it is only through an illegitimate broadening of the case’s true ratio in
subsequent decisions that this came to be. See (n 92).
102
Paul Daly, ‘Deference on Questions of Law’ (2011) 74(5) MLR 694.
103
See for example Elliott & Varuhas (n 11), 54-64.
104
Ibid, 65-71.
105
Ibid, 72-78.
106
See (n 97).
Anisminic (as interpreted in Page) as their starting-point and work outwards through a series
of analytical distinctions in order to ascertain the proper approach to judicial review. What
they actually more often do is begin with the legislation at the background of the case.
Judgments in these cases, that is, often read as a close and careful process of navigating the
detail of the statutory framework and an attempt to extract from this framework the
legislative policies underlying it so that these can be used to shape the approach to judicial
review deployed. Close consideration of two well-known cases – Puhlhofer107 and Croydon108
- both of which are an important focus of the textbook accounts just discussed, will suffice to
illustrate this point.109
Take, firstly, Puhlhofer,110 decided in 1986. The applicant in this case applied on behalf of his
family to Hillingdon LBC for assistance under the Housing (Homeless Persons) Act 1977. 111
At the time of application, the family resided in a bed and breakfast facility in which all
members of the family slept in one room and had no access to cooking or laundry facilities.
Hillingdon LBC nonetheless concluded that the family were not homeless within the meaning
of the Act because the family resided in ‘accommodation.’ 112 Puhlhofer challenged the
lawfulness of this conclusion; over-crowdedness and the lack of access to basic facilities,
Puhlhofer argued, inhibited any conclusion that a family were ‘accommodated’ under the
housing legislation.
Three points, in particular, are worth emphasising about how Lord Brightman, with whom all
of their Lordships agreed, set about deciding the case. Firstly, it is noteworthy that his
Lordship made no mention in his judgment of the Anisminic decision at all. Indeed, the case
does not seem to have been cited by counsel, or mentioned by the court, at any level of the
Puhlhofer litigation.113 This is extremely interesting. By the time Puhlhofer was decided the
discourse exploring the proper reading of Anisminic was in full swing;114 Diplock had
published an important article arguing for a broad reading of the case just twelve years
previous and his judgment in O’Reilly115 just had been decided just three. There is every
reason to believe that their Lordships in Puhlhofer would have been aware of this discourse
but, despite this, Lord Brightman seems to have simply not found it helpful, for the purposes
of resolving the legal question arising in Puhlhofer, to take Anisminic as the proper starting-
point.
Secondly, and importantly, what his Lordship did take as the proper starting-point was the
legislation at the background to the case: the Housing (Homeless) Act 1977. The bulk of
Lord Brightman’s judgment is dedicated to an attempt to unpack the logic of the
administrative scheme to which this Act gave rise. His Lordship’s main goals in deciding
107
Puhlhofer (n 63).
108
Croydon (n 64).
109
Other cases which could be discussed include Anisminic (n 72) and Page (n 44) themselves and R v
Monopolies and Mergers Commission, ex parte Yorkshire Transport Ltd [1993] 1 WLR 23 (HL); see
more broadly discussion in Nason (n 95), 166.
110
Puhlhofer (n 63).
111
Housing (Homeless Persons) Act 1977.
112
Ibid, s1(1).
113
R v Hillingdon LBC, ex parte Puhlhofer [1985] 3 All ER 734 (CA); R v Hillingdon LBC, ex parte
Puhlhofer (1985) 17 HLR 278 (HC).
114
See for example DM Gordon, ‘What Did the Anisminic Case Decide?’ (1971) 34 MLR 1; Jack
Beatson, ‘The Scope of Judicial Review for Error of Law’ (1984) 4(1) OLJS 22.
115
O’Reilly (n 73).
Puhlhofer seem to have been twofold: firstly, to uncover the policy goals which had driven
Parliament to pass the 1977 Act and, secondly, to figure out what the implication of these
goals were for the question of how the court should set about reviewing the activities of local
housing authorities. On these issues, Lord Brightman concluded as follows. Firstly, in Lord
Brightman’s view, Parliament’s intention in creating the 1977, as indicated by its long title, 116
was not to create a series of relationships of correlative right and duty 117 between local
authorities and qualifying individuals. Parliament’s intention, rather, was to create a scheme
which would empower local authorities to identify persons living within its area who were in
relative need and to provide them with an appropriate level of assistance. For Lord
Brightman, furthermore, Parliament’s choice of the term ‘accommodation’, coupled with the
failure to define this provision with any more precision, was an important part of this policy:
Parliament’s intention was that local housing authorities would have a good deal of
discretionary judgment in deciding in which category of need to place an applicant. 118
Secondly, this had important implications for the proper judicial approach to reviewing
Hillingdon LBC’s decision: the court, for Lord Brightman, should exercise restraint and
ought, more particular, to interfere only where a local housing authority departed from the
‘ordinary meaning’119 of the language of the Act.120
Thirdly, while it is true that his Lordships did at some points make use of language also
deployed in the sort of analytical framework offered by the leading textbooks - his Lordship,
for instance, at one point refers to the question of whether the family had accommodation as a
‘question of fact to be decided by the local authority’121 – this language was drawn upon in a
very particular, and highly limited, way. The distinction between ‘questions of fact’ and other
sorts of question, more particularly, was not put to any meaningful, ex ante legal work in the
course of Puhlhofer. Lord Brightman that is, did not treat the distinction as a legal
mechanism which helped him to identify the proper standard of review to be applied in
relation to Hillingdon LBC’s assessment. The language ‘question of fact’, rather, was drawn
on as a way of expressing a conclusion once Lord Brightman had identified the proper
judicial approach to be taken in light of the legislative background.
A similar set of points, furthermore, can be made about the Supreme Court’s judgments in the
second example: Croydon.122 Thus, firstly, in the course of rendering judgment in this case
neither Lady Hale nor Lord Hope made reference to either Anisminic itself, nor the case law
such as Page, which has subsequently conferred a broad reading on that decision. Secondly,
the bulk of both judges’ explanations for their decision in this case are, rather, preoccupied
with the same set of legal questions explored by Lord Brightman in Puhlhofer: what were the
policy aims underlying the legislation at the background of the case – the Children Act
1989123 - and what implications did this have for judicial oversight of decisions by local
authorities? As Lady Hale put it, ‘the task in all these cases is to decide what Parliament
116
Puhlhofer (n 63), 517.
117
Wesley Hohfeld, Fundamental Legal Conceptions as Applied in Judicial Reasoning (The Law Book
Exchange 2010).
118
Puhlhofer (n 63), 517.
119
Ibid.
120
This construction of the legislation was no doubt controversial. See for example R (B) v Southwark
[2003] EWHC 1678 (Admin).
121
Puhlhofer (n 63), 517.
122
Croydon (n 64).
123
Children Act 1989.
intended.’124 In relation to this aspect of the case, all of the judges were in agreement: the
purpose of the 1989 Act was to place local authorities under a series of inescapable legal
duties in relation to individuals who sought it and were under the age of 18; as Lord Hope put
it ‘the scheme of the Act shows that it was not parliament’s intention to leave this matter to
the judgment of the local authority.’ As such, in cases of doubt, the courts’ proper role was to
check the correctness of the local authority’s findings. Thirdly, although in the course of
offering judgment Lady Hale, like Lord Brightman in Puhlhofer, drew on language similar
that deployed in the kinds of framework developed by textbooks – her Ladyship drew, in
particular, on the idea that the question of whether a person was a ‘child’ within the meaning
of the 1989 Act was a ‘jurisdictional fact’ 125 – she used this language in a highly limited way,
in particular to express a conclusion reached on the basis of the exercise in construing the
legislation.126
The claim, then, that a broad reading of Anisminic according to which all misconstructions of
a statute and all legal errors more broadly result in an excess of jurisdiction has come to serve
as the ‘starting-point’127 of administrative law and thus that the significance of Cart lies in its
‘abandonment’128 of this tenet of ‘orthodox theory,’ 129 ought to be treated with considerable
caution. It is, in particular, highly questionable whether the courts, as opposed to the leading
textbooks, have every fully and meaningfully embraced this idea. A comparison of the way in
which cases such as Puhlhofer130 and Croydon131 are analysed in the leading administrative
law textbooks with how they were actually reasoned serves to illustrate the point neatly; the
idea that Anisminic (as interpreted in Page) forms the starting-point for administrative law
adjudication has tended to be something crafted onto case law in subsequent analysis rather
than an idea which plays a meaningful and positive role in legal reasoning itself

3. THREE SEAMS OF CONTINUITY WITH THE PAST INHERENT IN CART

The previous part showed how each of the three senses in which it has been suggested that
Cart’s significance lies in departure from the past is unconvincing. The commentary which
advances these reading, it has been argued, tends to overstate the sense of novelty inherent in
the case and overlooks a number of important seams of continuity inherent in the decision. At
this stage of the article it is helpful to return to a question raised in the introduction: why does
any of this matter? Why does, more particularly, Cart and the commentary which has
accumulated on it warrant the lengthy and detailed focus of a full article? The answer to this
question should begin to emerge from the discussion above and it is an answer which has two
parts. The first will be discussed here and the second in the conclusion.

124
Croydon (n 64), [24].
125
Ibid, [32].
126
Ibid.
127
Forsyth (n 6), 152.
128
Ibid, 155.
129
Murray (n 7), 108.
130
Puhlhofer (n 63).
131
Croydon (n 64).
The first part of the answer is that careful reflection on Cart and its relationship with the past
is illuminating and important, not only because it develops our understanding of a major
recent case. There are also important lessons embedded in the analysis undertaken above
about administrative law more broadly. Three themes which have long characterised
administrative law, and of which Cart can be seen as a continuation, are, in particular, worth
drawing out here. Each of this themes draws attention to a deeply important and embedded
feature of English administrative law, though these features are sometimes overlooked in
legal scholarship.
Firstly, major changes in the machinery of administration and judicial creativity in
responding to them are far from novel features of administrative law. The history of English
administrative law is, indeed, littered with examples of the courts being faced with legislation
which seeks to create new administrative programmes and thinking creatively, and sometimes
beyond the immediate legal context, for an appropriate solution to be deployed. A number of
examples of the courts doing this were discussed in the analysis above. Some of these
examples – such as the Court of the King’s Bench’s invocation of the timing-based approach
to drawing the distinction between jurisdictional and non-jurisdictional error in Bolton132 -
were taken from administrative law’s older history. Some – such as the rediscovery in 1952
in Shaw133 of the idea that the court could intervene to correct errors which appeared ‘on the
face of the record’ – from more recent decades. Some again – such as the creation in G134 of a
new limit on judicial review in the context of immigration and asylum claims – from very
modern, post-1977 procedural reform, administrative law. Against the background of
examples such as these, Cart can be seen as but the latest of a long string of examples of the
courts responding creatively when faced with a major change in the administrative state.
Another way of expressing this point is to say that in developing administrative law the
courts are always dealing with something of a ‘moving target.’ The machinery of
administration does not stand still;135 Parliament, as it falls under the control of new
governments136 responding to new problems, is continually pursuing new policy aims and
creating new administrative institutions for their implementation. Cart is a particularly
striking example of this – as noted above, in Cart the Supreme Court was faced for the first
time with the biggest and most comprehensive overhaul of the tribunals system which has
ever taken place – but it is far from the first such example. It is also far from the first instance
of the courts finding that the immediate legal toolbox is inadequate to provide a sufficiently
bolstered or restricted version of judicial review 137 and thus looking further afield for a legal
solution.

132
Bolton (n 31).
133
Shaw (n 41).
134
G (n 66).
135
For discussion of the changing administrative state see especially: Lord Hewart, The New Despotism
(London: Ernst Benn Ltd, 1929); Committee on Ministers’ Powers, ‘Report’ (London HMSO, 1932);
Cecil Carr, Concerning English Administrative Law (Columbia University Press 1941); JDB Mitchell,
‘The Causes and Effects of the Absence of a System of Public Law in the United Kingdom’ [1965] PL
95; Lord Hailsham, Dilemma of Democracy: Diagnosis and Prescription (Harper Collins 1978); David
Williams, ‘The Donoughmore Report in Retrospect’ (1982) 60 Public Administration 273; Derek
Fraser, The Evolution of the British Welfare State (4th edn, Palgrave Macmillan 2009); Sedley, (n 4).
136
See especially Adam Tomkins, ‘What is Parliament For?’ in Nicholas Bamforth & Peter Leyland
(eds), Public Law in a Multi-Layered Constitution (OUP 2003).
137
See discussion in Murray (n 7).
Secondly, as stressed throughout sections B and C of part 2, the specific legislative scheme at
the background to a case plays a very major, though often overlooked, role in legal
reasoning.138 Thus, when faced with a difficult legal question concerning the proper approach
to be taken to review in a particular case, a major focus of the courts has long been that of
thinking carefully about the policy goals which lie at the heart of Parliament’s enactments so
that these aims can then play a role in shaping the approach to review which is ultimately
deployed.
Again, a number of examples of the courts engaging in this process have been discussed in
the course of this article. Thus in discussing Puhlhofer,139 for instance, it was emphasised
how, in deciding how to go about reviewing assessments of whether an applicant was
‘accommodated’ within the meaning of the Housing (Homeless) Persons Act 1977 140 the
starting-point for the House of Lords was that piece of legislation. Lord Brightman, as
explained above, sought to uncover, by thinking carefully about the detail of the statutory
framework, what it was that Parliament was striving to achieve and what, in turn, this meant
for the proper approach to judicial review of local housing authority decisions. In a similar
way, it was stressed that the Supreme Court’s conclusion in Croydon,141 that a court could
reopen the evidence and inquire for itself whether a person was a ‘child’ within the meaning
of the Children Act 1989,142 was based centrally on the judges’ understanding of the
legislative goals underlying that statute. Cart itself, furthermore, is a very clear example of
the important role played by legislation in administrative law adjudication; a major focus of
the Supreme Court in finding a solution to the legal problem which arose in this case, it was
stressed above, was on unpacking the policy aims which underlay the TCEA and on using
these aims to shape the solution reached.143
This focus on legislation, it has been emphasised, is a feature of legal reasoning which is
sometimes overlooked by administrative law scholarship. Thus, as explained above, it is not
uncommon for leading textbooks to weave cases such as Puhlhofer and Croydon, in which
the main focus of the court was on the specific statutory scheme at the background to the
case, into complex and general frameworks built around analytical distinctions. In doing so
however, these accounts paint pictures of how such cases were decided which diverge
significantly from the reasoning which judges actually offer by way of justifying their
conclusions.
The third and final theme is closely related to the second: while general analytical distinctions
such as that between issues of construction, application and fact-finding are commonly
deployed in academic analysis of case law, especially in the leading textbooks, 144 it is highly
138
Note that none of this, of course, is to suggest that judicial review is limited to instances in which
power is conferred by legislation: R v Criminal Injuries Compensation Board, ex parte Lain [1967] 2
QB 864 (QB); Council of Civil Service Unions v Minister for the Civil Service [1985] AC 374 (HL).
The reality, of course, however is that the bulk of administrative law issues arise against the
background of such power. Interestingly, when the court is faced with administrative power derived
from a different source a major focus of adjudication is still on attempting to unpack the legal policies
underlying the existence of the scheme: see for example the reasoning in Page (n 44).
139
Puhlhofer (n 63).
140
Housing (Homelessness) Act 1977, s2.
141
Croydon (n 64). See further R (CJ) v Cardiff CC [2011] EWCA Civ 1590, [2012] 2 All ER 836.
142
Children Act 1989.
143
Many more examples could be added to this list. See for instance (n 109).
144
See for example Elliott & Varuhas (n 11), chapter. 2.
questionable whether judges find significant utility in such tools. Where such language is
drawn on by judges, it was stressed above, it tends to be drawn on by way of expressing a
conclusion reached on some other basis, often following a process of close engagement with
the statutory scheme at the background to a case. These general analytical tools, which play
such a central structuring role in much textbook analysis, in other words, seem to do very
little in the way of ex ante ‘legal work’ in identifying concrete solutions to difficult questions
concerning the proper approach to review, even in the most well-known case law.145

4. CONCLUSION: IMPLICATIONS FOR HOW WE ‘DO’ ADMINISTRATIVE LAW

This article has made use of a rather ‘funnel-like’ argumentative structure. It began in part 1
with the issues and decision in Cart. The focus widened out in part 2 to a consideration and
critical analysis of some of the commentary which has accumulated on that decision and in
part 3 the paper broadened out again in order to extract from the earlier analysis three
important, though sometimes overlooked, themes which it is argued have long characterised
administrative law more broadly. Rather than offering a summary of the earlier arguments,
the conclusion of this article will continue with this ‘funnel’ approach and widen out once
again. The aim, in particular, is to explain the second part of the answer to the question of
why there is broader importance in the analysis undertaken in this article. The key point is
that the three themes drawn out in part 3 of this article from that analysis are not only
important of themselves. They also give rise to some important and challenging, though
under-discussed, questions about how we ‘do’ administrative law.
Consider, for instance, the following issue. One of the most striking developments in public
law scholarship over the last century has been the emergence of what may be called
‘generalist administrative law,’ an idea which can be best illustrated through a discussion of
the earliest textbooks on English administrative law. 146 The first two English textbooks on
administrative law were FJ Port’s Administrative Law, published in 1929,147 and John Griffith
& Harry Street’s Principles of Administrative Law, published in 1951.148 With over two
decades of major social and political change to separate them, there is of course a good deal
which is different about these two texts but they do have an important feature in common:
they are structured around an exploration of different kinds of administrative scheme.
Commensurate with the focus of much legal discourse at the time, 149 their primary focus is on
understanding the nature and legal regulation of two varieties of administrative power in
particular: the powers of governmental bodies to create legislation and those of administrative
tribunals to to adjudicate. Now contrast the third and fourth textbooks 150 – Stanley De Smith’s

145
See further Rebecca Williams, ‘When is an Error not an Error? Reform of Jurisdictional Review of
Error of Law and Fact’ [2007] PL 793; Nason (n 95), 166.
146
See discussion in Carol Harlow & Richard Rawlings, ‘Administrative Law in Context: Restoring a
Lost Connection’ [2014] PL 28.
147
FJ Port, Administrative Law (Longmans, Green & Co 1929).
148
John Griffith & Harry Street, Principles of Administrative Law (London: Pitman & Sons 1952).
149
See especially Committee on Ministers’ Powers, ‘Report’ (London HMSO, 1932) (‘Donoughmore
Report’); Franz Becker, ‘The Donoughmore Report and the Franks Report’ (1958) 24 International
Review of Administrative Sciences 31; Hewart (n 42).
150
See further Denis Galligan, ‘Judicial Review and the Textbook Writers’ (1982) 2 OJLS 25.
Judicial Review of Administrative Action first published in 1959151 and William Wade’s
Administrative Law in 1961.152 These texts make use of a markedly different structure to Port
and Griffith & Street. They are arranged around a discussion of general legal techniques,
which are said to apply across all areas of administration for the purpose of adjudicating on
the lawfulness of all uses of administrative power. This swinging of the pendulum from the
study of specific types of administrative power to the study of general legal techniques for
overseeing all areas of administration has had a powerful influence in administrative law.
There is now undoubtedly a widespread and deeply ingrained tendency to think of, research
and teach administrative law as a set of general legal ‘grounds of review’ 153 or ‘principles of
good administration’154 which apply regardless of whether the court is faced with the context
of housing or health administration.
Against this background, the analysis offered in this article gives rise to a challenging and
fundamental question: if the themes drawn out in part 3 are truly long-standing and deeply
embedded features of administrative law adjudication, is this generalist administrative law
project not fundamentally misconceived? Does the important role played the specific
statutory scheme at the background to each case and the limited utility of certain general
analytical tools not mean that there is, in reality not one body of ‘administrative law’ but a set
of ‘administrative laws’ (housing law, social security law, etc) which ought to be studied and
taught as such? Or, more dramatically, do these themes, not mean that, ultimately, all of the
case law studied in this field ultimately collapses into DM Gordon’s dreaded ‘wilderness of
single instances?’155 The broader hope behind this paper is to generate thought and discussion
of this issue, rather than to provide a conclusive answer to it but two main reflections will be
offered here.
Firstly, there are a number of considerations which can help to explain why, at least at first
sight, it is possible both to accept that the themes drawn out in part 3 are long-standing
features of administrative law and to believe that it is important to maintain a body of
research and teaching which looks at the legal regulation of administration generally. One
such important consideration is that the extent of the role played by the specific statutory
scheme may vary depending on the ‘legal ground’ on which the applicant is seeking to rely. 156
The cases discussed in this article are commonly thought of as concerning one legal ground
in particular: ‘jurisdictional error’157 or ‘error of law.’158 As the analysis above makes clear,

151
Stanley De Smith, Judicial Review of Administrative Action (London: Stevens 1959).
152
William Wade, Administrative Law (OUP 1961).
153
Language used in, for instance, Elliott and Thomas (n 97), chapter 12.
154
Language used in, for instance, Dawn Oliver, ‘Is Ultra Vires the Basis of Common Law Judicial
Review?’ [1987] PL 54.
155
See (n 13).
156
Note this point may have important implications for the ongoing ‘ultra vires’ or ‘constitutional
foundations’ debate (see especially Paul Craig, ‘Competing Models of Judicial Review’ [1999] PL 42;
Mark Elliott, The Constitutional Foundations of Judicial Review (Bloomsbury, 2001); Thomas Adams,
‘Ultra Vires Revisited’ [2018] PL 31). In particular, it gives reason to question whether it is coherent to
speak of ‘grounds of review’ in the context of certain aspects of administrative law at all. For further
discussion of this issue see especially Trevor Allan, ‘The Constitutional Foundations of Judicial
Review: Conceptual Conundrum or Interpretative Inquiry?’(2002) CLJ 87; Trevor Allan, ‘Doctrine and
Theory in Administrative Law: An Elusive Quest for the Limits of Jurisdiction’ [2003] PL 429; Nason
(n 95), chapters 6-7.
157
Elliott & Varuhas (n 11), chapter 2.
158
Craig (n 45), chapter 16.
the courts when adjudicating on cases of this kind often view their primary task as that of
unpacking the specific statutory framework at the background of the case. There may be,
however, other ‘grounds of review’ - legitimate expectations or bias perhaps – in relation to
which specific legislation, though important, 159 does not play such a central role and where
general legal structures applying across all areas of administration are more forthcoming.
A second important consideration is that, even in relation to those ‘grounds’ where the
specific statutory framework is truly central, there is surely still an important role for
generalist administrative law scholarship. What flows out from the analysis undertaken in this
paper, more particularly, are a series of important and underexplored questions which can
only be answered by looking across different areas of administration. How do, for instance,
judges go about extracting the policy aims which underlie an administrative scheme and
using these to shape the approach to review deployed in a given case? Does legislation
provide meaningful structure to what the courts do here or are the judges in reality not
exercising a significant amount of normative choice? Has there been any major change in the
techniques which have be used by courts over time?160 Do the courts deploy these techniques
consistently across different areas of administration? Questions such as these surely admit of
no easy answers. The key point, however, is that fragmentation of administrative law
scholarship would prevent their exploration; these are important questions which can only be
successfully explored and addressed by scholars working within the generalist administration
law tradition.
The first point leads in turn to a second: one methodological implication which flows out
from this article is surely the importance of building into the generalist administrative law
tradition a greater commitment to taking legislation seriously. One major theme which runs
throughout this analysis is that judges often take the specific statutory framework at the
background of the case, and the task of unpacking it in order to extract its underlying policy
goals, as their starting-point in adjudication. Another, however, is the tendency on the part of
legal scholarship, and especially in textbook analysis, to ‘gloss over’ this reality and to
instead seek to weave case law into general analytical frameworks. 161 This disparity sits in
considerable tension with David Feldman’s apt warning that:
...theories are aids to understanding or advocacy, not logical truths or divine
revelations. They must be constantly tested and reshaped in the light of reality.
Inadequate theories promote confusion, not understanding.162
If we think of at least part of our task as legal scholars, that is, as being that of promoting
understanding of administrative law adjudication, then it is clear that we need to incorporate
159
The importance of the role played by legislation in legitimate expectations cases is, for instance, well-
illustrated by decisions such as R v Secretary of State for Education and Employment, ex parte Begbie
[2000] 1 WLR 1115 (CA) and R (Badger Trust) v Secretary of State for the Environment, Food &
Rural Affairs [2014] EWCA Civ 1405, [2015] Env LR 12.
160
An importance case in this regard may of course be Padfield v Minister of Agriculture, Fisheries and
Food [1968] AC 997 (HL). See further discussion in Maurice Sunkin, ‘Padfield v Ministry of
Agriculture, Fisheries and Food [1968]’ in Satvinder Juss & Maurice Sunkin, Landmark Cases in
Public Law (Hart, 2017).
161
This is commensurate with Liz Fisher’s point that sometimes in common law systems legislation tends
to be seen as ‘having second class legal status’: see Elizabeth Fisher, Environmental Law: A Very Short
Introduction (OUP 2017), 26.
162
Feldman (n 94), 314. See also further David Feldman, ‘The Nature of Legal Scholarship’ (1989) 52(4)
MLR 498.
into our agenda a greater space for thinking about, and asking questions about, the role
played by legislation in legal reasoning than these issues currently occupy. Specific statutory
frameworks play a major role in adjudication. It is important they play a role in our legal
scholarship too.