British
and Irish Legal Information Institute
Freely Available British and Irish Public Legal Information
[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
Scottish Court of Session Decisions
You are here: BAILII >> Databases >> Scottish Court of Session Decisions >> NHS HEALTH BOARD AGAINST THE COMMISSIONERS FOR HER MAJESTY'S REVENUE AND CUSTOMS [2020] ScotCS CSIH_14 (09 April 2020)
URL: http://www.bailii.org/scot/cases/ScotCS/2020/2020_CSIH_14.html
Cite as: 2020 GWD 13-199, [2020] CSIH 14, 2020 SC 351, [2020] STI 1062, 2020 SLT 369, [2020] BVC 6, [2020] STC 1112, [2020] ScotCS CSIH_14
[New search] [Printable PDF version] [Help]
Page 1 ⇓
FIRST DIVISION, INNER HOUSE, COURT OF SESSION
Lord President
Lord Drummond Young
Lord Glennie
[2020] CSIH 14
XA5/19
OPINION OF THE COURT
delivered by LORD DRUMMOND YOUNG
in an appeal to the Court of Session by
NHS LOTHIAN HEALTH BOARD
Appellant
against
A decision of the Upper Tribunal dated 2 July 2018
against
THE COMMISSIONERS FOR HER MAJESTY’S REVENUE AND CUSTOMS
Respondent
for repayment of input value added tax
Appellant: Edwards; Clyde & Co (Scotland) LLP
Respondent: DM Thomson QC, Roxburgh; Office of the Advocate General
9 April 2020
Introduction
[1] The decision of the House of Lords in Fleming (trading as Bodycraft) and Condé Nast Ltd
2008 have given rise to a significant number of claims for overpaid value added tax of a
Page 2 ⇓
2
historical nature. The present case involves such a claim. As in many such cases, the primary
issue is not the existence of the taxpayer’s claim to recover overpaid value added tax but the
quantification of that claim, and in particular whether the claim can be quantified with
sufficient accuracy to permit an order for repayment of tax to be made.
[2] The taxpayer is a National Health Service trust responsible for the provision of
health services within the Lothian area, and is the successor to health boards that provided
similar services within the same area. The taxpayer and its predecessors operated
44 scientific laboratories during the period between 1974 and 1997. Most of the work
performed in the laboratories was carried out for the clinical purposes of the taxpayer and
its predecessors. To that extent the work of the laboratories consisted of non-business
activities for value added tax purposes. Nevertheless, the laboratories carried out work for
persons outside the NHS, including local authorities and pharmaceutical companies. For
value added tax purposes such work amounted to business activities. The Upper Tribunal
records that such work included non-patient tests, work for the National External Quality
Assessment Scheme (NEQAS), drug trials and food and water testing. It is agreed that the
input tax paid by the taxpayer and its predecessors between 1974 and 1997 in respect of
laboratory expenditure for such business activities was not reclaimed by the taxpayer or by
any government body acting on its behalf.
[3] The taxpayer now seeks to recover such input tax. It has made a global Fleming claim
for the whole period from 1 April 1974 to 30 April 1997 which covers, inter alia, the activities
of its laboratories. The claim in respect of the laboratories has been reduced from its original
level, and is now for £929,874.69 of what is alleged to be recoverable input tax. That claim
was rejected by HMRC, and the taxpayer appealed to the First-tier Tribunal. The First-tier
Tribunal heard from a number of witnesses who gave evidence on behalf of both the
Page 3 ⇓
3
taxpayer and HMRC. These included four scientists who had been employed in the
taxpayer’s laboratories between 1974 and 1997 and two accountants who gave evidence
relating to the taxpayer’s financial management. The First-tier Tribunal rejected the
taxpayer’s appeal on 26 June 2017. That was followed by an appeal to the Upper Tribunal,
which rejected the appeal on 2 July 2018. The taxpayer has now appealed to the Court of
Session against that decision.
[4] The principal argument presented to the Upper Tribunal was that the First-tier
Tribunal had erroneously adopted an approach based on partial exemption instead of
apportionment between business and non-business activities. The Upper Tribunal rejected
that argument on an analysis of the First-tier Tribunal’s reasoning. We should record at this
stage that on this matter we are in full agreement with the reasoning of the Upper Tribunal.
In the argument before the court, however, the taxpayer’s principal criticism was directed at
the First-tier Tribunal’s decision to refuse the taxpayer’s claim in its entirety on the basis that
the evidence available was inadequate to permit proper calculation of the claim; the
evidence produced, it was contended, was the best that was available, and a reasonable
methodology was suggested for calculating the amount of the historical overpayments. In
support of that argument the taxpayer relied on the EU law principle of effectiveness.
[5] In our opinion the argument based on the adequacy of the available evidence is well
founded, for reasons that we will discuss. In order to place that argument in context,
however, it is necessary for us first to discuss the scheme of value added tax, and in
particular the right to deduct input tax in cases where the taxpayer carries out both business
and non-business activities; secondly, to describe the history of the right of a taxpayer to
recover overpaid value added tax as a matter of both EU and UK law, and in so doing to set
out the present state of that right; thirdly, to set out the historical background to the
Page 4 ⇓
4
activities of the taxpayer and its predecessors; and fourthly, to describe briefly the EU law
principles of effectiveness, certainty and proportionality. Thereafter we will consider the
evidence led before the First-tier Tribunal and the findings of that Tribunal as to the
business and non-business activities of the taxpayer and its predecessors during the period
from 1974 onwards. On the basis of that evidence, we will consider the approach that a
Tribunal should take to evidence available in historical claims for recovery of overpaid
value added tax. In so doing, we have regard to the difficulty of obtaining detailed evidence
after the lapse of many years, and to the fact that the inability of the taxpayer and other
comparable bodies to adduce evidence has in large part been produced by decisions made
by government within the UK, through failures to implement properly the right under EU
law to recover overpaid value added tax and through decisions by the government
departments responsible for the NHS that NHS boards and trusts should not reclaim input
tax on their non-business activities.
[6] An appeal from the Upper Tribunal to the court is only competent on a point of law:
Tribunals, Courts and Enforcement Act 2007, section 13. The primary question that arises in
the appeal to the court is the approach that should be taken by the Tribunal to evidence in
historical claims for recovery of overpaid value added tax. That in our opinion can be said to
raise a matter of general law. Moreover, it raises a question of the application of the law to
the facts as found by the First-tier Tribunal, which is also a point of law. Indeed, on its facts
the present case raises the question of how the legal rules that permit the recovery of
overpaid value added tax after the time when input tax would normally have been
deducted apply to historical claims for such recovery. That raises general legal issues, which
we discuss subsequently. The secondary argument before the court, which was the primary
argument before the Upper Tribunal, relates to whether the First-tier Tribunal made a
Page 5 ⇓
5
fundamental error in its approach to the case, by asking the wrong question or by taking
account of manifestly irrelevant considerations. That also raises a point of law, as the Upper
Tribunal held (at paragraph 17).
The scheme of value added tax
[7] Value added tax has its origins in EU legislation; the tax is now governed by the
Principal VAT Directive (Directive 2006/112/EC), although the present claim is governed by
predecessors of that Directive. In the scheme of VAT a fundamental distinction is drawn
between economic (or business) activities, that is to say supplies of goods and services made
in a market for monetary consideration, and non-economic (or non-business) activities
which are not made for monetary consideration and are therefore outside any market; the
distinction is found in article 9(1) of the Principal VAT Directive and, so far as the United
Kingdom is concerned, in section 4 of the Value Added Tax Act 1994. EU Directives use the
expression “economic” or “non-economic”, whereas the United Kingdom value added tax
legislation uses the expressions “business” and “non-business”, but these have the same
meaning for present purposes. Most of the activities of the taxpayer in the present case are
classified as non-economic or non-business activities. In respect of those activities, the
taxpayer is not obliged to charge value added tax on supplies, and has no right to recover
value added tax that has been incurred for the purpose of making those supplies.
[8] Nevertheless certain of the taxpayer’s activities involve the making of business
supplies. Those are subject to the ordinary value added tax regime. Output tax is chargeable
on the consideration received for such supplies, and input tax that is attributable to the
making of such supplies is recoverable. Prior to 1994 the value added tax liability of NHS
boards was handled by the Scottish Office, and during that period (and until the tax year
Page 6 ⇓
6
1995/1996) it was the general practice of public bodies not to make claims to recover input
tax on business activities. Consequently at that time the general practice in relation to
business supplies by NHS boards and other public bodies was that value added tax was
ignored. That practice changed subsequently. Nevertheless, a major factor in the failure of
the taxpayer’s predecessors to recover input tax during the period down to 1994 was the
manner in which government had organized the activities of health boards within the NHS
and decisions made by government in relation to value added tax on such activities.
The right to deduct input tax
[9] The right to deduct input tax is an important aspect of EU law. The right was enacted
in earlier legislation, but it is now contained in articles 167-172 of the Principal VAT
Directive. So far as material, these provide as follows:
“Article 167
A right of deduction shall arise at the time the deductible tax becomes chargeable.
Article 168
In so far as the goods and services are used for the purposes of the taxed transactions
of a taxable person, the taxable person shall be entitled, in the Member State in which
he carries out these transactions, to deduct the following from the VAT which he is
liable to pay:
(a) the VAT due or paid in that Member State in respect of supplies to him of
goods or services, carried out or to be carried out by another taxable
person;…”.
Thus EU law confers a right to deduct input tax in respect of supplies of goods or services
used for the purposes of taxed services, and the right to deduct arises as soon as the
deductible input tax becomes chargeable.
[10] The way in which the deduction may be made is governed by articles 178-183. So far
as material for present purposes, article 179 provides that the taxable person shall make the
Page 7 ⇓
7
deduction by subtracting from the total amount of VAT due in a given tax period the total
amount of VAT in respect of which, during the same period, the right of deduction has
arisen. Article 180 provides that
“Member States may authorize a taxable person to make a deduction which he has
not made in accordance with Articles 178 and 179”.
Article 182 provides that Member States shall determine the conditions and detailed rules
for applying inter alia article 180.
[11] Under UK legislation, the equivalent right to recover overpaid VAT is currently
found in section 80 of the Value Added Tax Act 1994, a section that has been subject to
extensive amendment. That right is subject to an express limitation period, currently of six
years. Between 1974 and 1989, however, there was no provision in domestic law that gave
any right to recover past overpayments of VAT. Section 24 of the Finance Act 1989 gave a
right to recover overpaid output tax, but this was subject to a limitation period of six years.
The right conferred by that section did not extend to the deduction of input tax that had not
been deducted at the time when it was paid to a supplier. In University of Sussex v Customs
and Excise Commissioners [2001] STC 1495, Neuberger J confirmed that interpretation, holding
that failure to deduct input tax did not give rise to an overpayment of money falling within
section 80. Nevertheless, the taxpayer in that case was held entitled to recover input tax over
a period of 23 years on the basis of statements that had been made by the Commissioners of
Customs and Excise; that right of recovery was not subject to any time limit. Section 80 of
the 1994 Act was amended by the Finance Act 1996 in such a way as to permit taxpayers to
recover overpaid output tax, subject to a limitation period of three years. In respect of input
tax, the right to make a claim for unrecovered input tax was conferred by regulation 29(1A)
of the Value Added Tax Regulations 1995 (SI 1995/2518), inserted by regulation 4(b) of the
Page 8 ⇓
8
Value Added Tax (Amendment) Regulations 1997 (SI 1997/1086). That too was subject to a
limitation period of three years. Both of those time limits were made retrospective.
[12] When those time limits were imposed, no transitional provisions were made in
respect of historical claims. In Case C-00/35, Marks & Spencer PLC v Customs and Excise
Commissioners, [2002] STC 1036, the Court of Justice held that the lack of any transitional
provision was incompatible with EU law, inter alia because it involved a breach of the
principle of effectiveness. Thereafter, in Fleming and Condé Nast Ltd v HMRC, supra, the
House of Lords applied a similar approach to claims for unrecovered input tax; it was
permissible to impose a reasonable time limit, but if that was done retrospectively
transitional provisions had to be made that accorded with the principle of effectiveness, in
such a way as to allow historical claims to proceed. It is that principle that is relied on in the
present case. Following the decision in Fleming Parliament enacted section 121 of the Finance
Act 2008, which in relation to input tax enacted that claims for recovery of such tax incurred
in accounting periods ending before 1 May 1997 could be made without any time limit
provided that the claim was made before 1 April 2009. Those have come to be referred to as
“Fleming claims”.
[13] Numerous such claims have been made. The court was informed that 700 claims
were made by NHS authorities in Great Britain, and that more than 200 of those are still
unresolved. The fundamental problem in most historical claims is the availability of
evidence to support a claim for repayment. Financial records may have been destroyed, in
whole or in part, and the consequent lack of primary documentation may make proof of a
claim difficult or impossible. That is essentially the problem that has arisen in the present
case.
Page 9 ⇓
9
The history of the activities of the taxpayer and its predecessors
[14] With effect from 1 April 1974 healthcare functions in Scotland were transferred to
regional health boards by virtue of the National Health Service (Scotland) Act 1972. The
taxpayer is the successor of the regional health board responsible for health services in
Lothian, which was established on 1 April 1974. At that time all of the health boards in
Scotland had a single VAT registration, although each board managed its own finances.
Claims in respect of value added tax were accordingly made by the Scottish Office on behalf
of the various regional boards. To the extent that value added tax was overpaid, however, it
was the regional board in question that was entitled to repayment.
[15] The registration requirements for value added tax were altered on 1 July 1992
through the creation of a divisional VAT registration for all Scottish health boards in respect
of intra-NHS supplies in Scotland. Thereafter the health boards used individual VAT
registration numbers. In 1994 NHS trusts were set up, and the administration of VAT
returns was devolved to them. In consequence, 43 individual trusts became separately
registered for VAT. That number was reduced with effect from 1 April 1999. On 1 April 2004
the trusts were replaced by health boards. The present taxpayer is a health board established
on that date. It is, however, the statutory successor of a previous trust and the health board
that preceded that trust. Consequently the taxpayer is entitled to make all relevant claims in
respect of the VAT liabilities of its predecessors throughout the period from 1 April 1974. At
this point we should note that since 1983 all NHS authorities have been able to recover
“Contracted Out Services” value added tax directly from HM Treasury; that represents
value added tax incurred on services obtained by them in order to carry on non-economic
activities; the statutory basis for this is now found in section 41(3) of the Value Added Tax
Page 10 ⇓
10
Act 1994. The present appeal does not relate to Contracted Out Services; it is confined to
input tax relating to taxable business activities.
[16] Against that background, the taxpayer has now claimed repayment of a sum
representing the input tax paid by it and its predecessors in respect of the business activities
carried on in its laboratories throughout the period from 1 April 1974 to 30 April 1997. In
view of the material findings in fact, which we discuss subsequently, the issue is not
whether the taxpayer has a claim to repayment of input tax. It is rather whether it is possible
to calculate the amount of that claim with sufficient certainty to permit recovery. VAT was
of course introduced in consequence of the United Kingdom’s joining the European
Communities, and the tax has continued to be governed by EU law and by the principles of
that legal system.
EU law principles: effectiveness, certainty and proportionality
Effectiveness
[17] Effectiveness is a fundamental principle of EU law. The leading case is Case 199/82,
Amministrazione delle Finanze dello Stato v San Giorgio SpA, [1983] 2 CMLR 658, in which the
taxpayer sought repayment of national charges that had been levied contrary to Community
law and the state imposed a requirement that any such claim must include proof that the
charges had not been passed on to other persons. The taxpayer challenged that requirement
of proof as contrary to Community law. The Court of Justice accepted that there would be
no breach of Community law if national legislation prevented the reimbursement of taxes,
charges and duties that had been levied contrary to Community law when it was established
that the person paying such charges had passed them on to other persons (paragraph [13]).
It then continued:
Page 11 ⇓
11
“[14] On the other hand, any requirement of proof which has the effect of making it
virtually impossible or excessively difficult to secure the repayment of charges levied
contrary to Community law would be incompatible with Community law. That is so
particularly in the case of presumptions or rules of evidence intended to place upon
the taxpayer the burden of establishing that the charges unduly paid have not been
passed on to other persons or of special limitations concerning the form of the
evidence to be adduced, such as the exclusion of any kind of evidence other than
documentary evidence …
[15] In a market economy based on freedom of competition, the question whether,
and if so to what extent, a fiscal charge imposed on an importer has actually been
passed on in subsequent transactions involves a degree of uncertainty for which the
person obliged to pay a charge contrary to Community law cannot be systematically
held responsible”.
[18] We note two particular features of this statement of law. First, the basic test of
effectiveness is whether requirements of national law render enforcement of a Community
right “virtually impossible or excessively difficult”. The second of these elements is
obviously more significant in practice; what that part of the test entails is a form of
proportionality exercise in which the national court must assess the difficulty of enforcing a
Community right. In doing so, the court must clearly have regard to both the evidence that
is available and the reasons for any apparent deficiencies in that evidence. Secondly, in
paragraph [15] the Court of Justice accepts that in some cases applying the principle of
effectiveness may involve uncertainties of evidence, and it cannot automatically be assumed
that the person who asserts a breach of Community law is responsible for such uncertainty.
This is, we think, of some importance in the present case, where the history of the taxpayer’s
activities indicates that the failure to reclaim input VAT on laboratory activities prior to 2006
resulted from a number of factors, including the failure of the United Kingdom to
implement systematically the EU legislation on the recovery of input tax (see
paragraphs [11] and [12] above) and the practice followed within the NHS, under direction
Page 12 ⇓
12
from the Scottish Office (and its equivalents elsewhere), not to reclaim input VAT on
business activities (see paragraph [8] above).
[19] The principle of effectiveness laid down in San Giorgio has been followed in
numerous subsequent cases. The formulation of the principle has varied over time, and it is
fair to say that some of the recent cases retreat from the more extreme statements of the
principle. Nevertheless, the basic principle as stated in that case remains a fundamental part
of EU law. The principle is now embodied in article 19 of the Treaty on European Union,
which provides that “Member States shall provide remedies sufficient to ensure effective
legal protection in the fields covered by Union law”. It is also found in article 47 of the EU
Charter of Fundamental Rights, which guarantees the right to an effective remedy in all
fields covered by EU law. Indeed, the principle of effectiveness can be regarded as an
application of the well-known dictum ubi jus ibi remedium: wherever there is a right there is a
remedy. As a matter of elementary common sense, it is obvious that if the law accords a
right to any person that person should have a remedy that can make that right effective in
practice.
[20] We were referred to a number of other cases that deal with the principle of
effectiveness, but these tend to turn on their specific facts and we do not do not consider it
necessary to consider them in detail. We should, however, perhaps note two subsequent
cases. First, in Cases C-6/90 and 9/90, Francovich v Italy and Bonifaci v Italy, [1993] 2 CMLR
66, the Court of Justice stated (at paragraphs [32] – [34] ):
“[I]t has been consistently held that the national courts whose task it is to apply the
provisions of Community law in areas within their jurisdiction must ensure that
those rules take full effect and must protect the rights which they confer on
individuals …
The full effectiveness of Community rules would be impaired and the
protection of the rights which they grant would be weakened if individuals were
Page 13 ⇓
13
unable to obtain redress when their rights are infringed by a breach of Community
law for which a Member State can be held responsible.
The possibility of redress from the Member State is particularly indispensable
where, as in this case, the full effectiveness of Community rules is subject to prior
action on the part of the State and, consequently, in the absence of such action,
individuals cannot enforce before the national courts the rights granted to them by
Community law before the national courts”.
Francovich is of course the leading authority on the right to obtain damages from the state for
its failure to implement provisions of EU law. The fundamental principle, however, is
relevant to the present case: to the extent that the state makes it excessively difficult for any
person to exercise rights conferred by Community law, it is the state that must be
responsible for the breach of the effectiveness principle.
Certainty
[21] The second case that we should note is Test Claimants in the Franked Investment Income
the UK Supreme Court, where Lord Sumption, in a partially dissenting opinion, refers (at
paragraph [146]) to:
“... the overriding requirement derived from the Treaty… that national legal systems
should provide a minimum standard of protection for EU law rights. In the case law
of the Court of Justice, the standard of protection required is embodied in two
principles which are restated in almost every decision on the point. First, the
substantive and procedural provisions of national law must be effective to protect
EU law rights (the ‘principle of effectiveness’). Their enforcement in national law
must not be subject to onerous collateral conditions or disproportionate procedural
requirements. They must not render ‘virtually impossible or excessively difficult’ the
exercise of rights conferred by EU law. Secondly, the relevant provisions of national
law must not discriminate between the rules and procedures applying to the
enforcement of EU law rights, and those applying to the enforcement of comparable
national law rights (the ‘principle of equivalence’). There is a third principle which
features less prominently in the case law on this subject but is of considerable
importance because it informs the approach of the Court of Justice to the first two.
This is the principle of legal certainty, which lies at the heart of the EU legal order
and entails (among other things) that those subject to EU law should be able clearly
to ascertain their rights and obligations”.
Page 14 ⇓
14
Counsel for HMRC placed some emphasis on the principle of certainty as described in this
passage. It is undoubtedly correct to state that legal certainty is a significant feature of EU
law which has been regularly restated by the Court of Justice. Indeed, it is perhaps obvious
that in any developed legal system certainty as to the rules of the system is a desirable
feature.
[22] Nevertheless, it is important to note that the principle in question is legal certainty.
What the principle means is that the rules of a legal system should be stated with clarity, so
that those affected by the system can discover their rights, obligations, powers and liabilities
with a reasonable degree of certainty. The principle of certainty does not, however, extend to
the facts of any particular case. It is obvious that the evidence relating to a case may vary
enormously in its extent, in its cogency, in its credibility and reliability, and in its relevance
to the issues that must be decided by the court. The principle of legal certainty is not
relevant to the assessment of the evidence, which is a wholly separate – and very important
– task for the court.
Proportionality
[23] Finally, in relation to EU law, we should note that proportionality is a fundamental
feature of the system. It is mentioned in the foregoing passage from Franked Investment
Income Claimants, in the reference to onerous collateral conditions and disproportionate
procedural requirements. Applying the principle of effectiveness to a particular case will
almost inevitably require an assessment of the proportionality of the procedural or
evidential requirements of national law. Furthermore, the national court, in assessing
whether sufficient and adequate evidence is available to establish a claim, must carry out
what is in essence a proportionality exercise – determining whether the evidence available
Page 15 ⇓
15
is, taken as a whole, sufficient to establish the claim. Proportionality is also relevant to the
standard used, whether that is the normal standard (at least in Scots law) of the balance of
probabilities or a departure from that standard.
The findings in fact relating to the business and non-business activities of the taxpayer
and its predecessors
[24] In relation to their taxable business activities, the taxpayer and its predecessors
divided recovery claims for input VAT into different areas of their activities. HMRC treated
the overall claim as comprising a number of “sectorized” claims, each of which related to a
different area of the taxpayer’s activities. The present appeal relates to the business activities
of the taxpayer’s laboratories, including such activities carried out by the taxpayer’s
predecessors. 44 such laboratories are involved. Most of their work was concerned with
functions of the NHS in Scotland, and to that extent it consisted of non-business activities.
Nevertheless, the laboratories also made taxable business supplies to third parties, such as
local authorities, English NHS bodies and pharmaceutical companies. The input tax that
related to supplies of that nature was recoverable. In so far as it was not recovered at the
time when the supplies were made, it was recoverable on a historical basis, in accordance
with the decision in Fleming and Condé Nast v HMRC, supra.
[25] For the purposes of the appeal to the First-tier Tribunal, and subsequent appeals,
certain facts were agreed between the taxpayer and the Commissioners. These are as
follows:
1) The taxpayer was entitled to make claims to recover historical input tax
attributable to taxable supplies of laboratory services.
2) VAT-inclusive costs were incurred in order to make these supplies.
3) Wages and salaries had been removed from expenses to prevent distortion.
Page 16 ⇓
16
4) Contracted Out Services VAT adjustment had been made.
5) 2006/2007 was the first tax year when a recovery claim for input tax on
laboratory expenses was made by the taxpayer, although laboratory claims
had been made in earlier years by other Boards, which had been accepted by
the Commissioners.
6) The VAT recovery fraction of 14.7% for Lothian Laboratories’ business
activities was derived from the 2006/2007 business activities survey.
7) In the claim period NHS Lothian carried out business activities through its
laboratories.
8) The business activities included NEQAS (National External Quality
Assessment Scheme), clinical trials and research, and public health and
environmental health.
9) No earlier VAT reclaims had been made, save earlier claims that had been
rejected as being out of time.
[26] The First-tier Tribunal heard detailed evidence from four scientists who had worked
in the taxpayer’s laboratories during the relevant period, which extended from 1974 to 1997.
These described the work that had been carried out for external bodies, and gave varying
evidence as to the amount of time that they had spent on non-NHS work. Evidence was also
led from two accountants who had been employed by the taxpayer regarding its financial
management during the relevant period. As already noted, the taxpayer’s first claim to
recover input tax on laboratory activities was for the year 2006/2007, and the recoverable
amount in that year was ultimately agreed at 14.7% of the total costs of the taxpayer’s
laboratories. The taxpayer sought to extrapolate that percentage backwards to earlier years
in order to claim deduction of input tax for the whole period from 1994 to 1997. That was
rejected by HMRC on the basis that the evidence available was insufficiently reliable.
[27] On the basis of the evidence led, the First-tier Tribunal made a number of findings of
fact that are material for present purposes:
1) The claim was not a new claim (First-tier Tribunal, paragraph 187), and was
therefore not time-barred.
Page 17 ⇓
17
2) The Tribunal accepted as reliable the evidence given by officials from the
various predecessor health boards as to the nature of the business activities
conducted by the taxpayer and its predecessors. These included NEQAS
work, food testing, water testing, non-medical testing of samples, especially
for public health, and research and development. (First-tier Tribunal,
paragraphs 189-190).
3) The nature of the laboratories’ activities did not change during the period of
the taxpayer’s claims, and it was reasonable to infer that the nature of the
inputs would not have changed markedly throughout that period (First-tier
Tribunal, paragraph 189). The claim all along had been for unrecovered input
tax attributable to taxable business supplies made throughout the period 1974
to 1997, and remained essentially the same (ibid).
4) The taxpayer and the Commissioners agreed that the taxable percentage of
the taxpayer’s outputs in the year 2006/07 was 14.7%. That was the figure
attributable to the taxpayer’s laboratory activities, and it was used in
calculating the taxpayer’s VAT liabilities for that year. That agreement,
however, did not extend to other years (First-tier Tribunal, paragraph 195).
[28] Those findings in fact are in our opinion of great importance. In the first place, they
establish that throughout the period of the taxpayer’s claim the taxpayer and its
predecessors carried out business activities which were subject to VAT and in respect of
which input VAT was deductible. That inevitably means that the taxpayer has a valid
Fleming claim for input VAT that was not recovered. The question at issue in the present
proceedings is accordingly not the existence of such a claim but its quantification. We
discuss subsequently the legal significance of this distinction, at paragraphs [39], [55] et seq
and [63]-[65] below. In the second place, on the basis of the evidence led before the First-tier
Tribunal, that Tribunal held that the activities of the taxpayer’s laboratories did not change
during the period of the claims, and that it might reasonably be inferred that the nature of
the inputs (the basis for the recovery of input VAT) would not have changed either, to any
significant degree. That means that, in principle at least, it should be possible to extrapolate
findings of fact relating to the proportion of the taxpayer’s business activities to its non-
business activities throughout the period of the claim. That is a matter of elementary logic.
Page 18 ⇓
18
In the third place, agreement was reached between the taxpayer and HMRC as to the
division of the taxpayer’s activities between business and non-business elements for the tax
year 2006/07. The critical question in the case was, put simplistically, whether that
percentage could be extrapolated backwards to the period of the claim, from 1974 to 1997,
and throughout that period.
The decision of the First-tier Tribunal
[29] As already indicated, the First-tier Tribunal heard extensive evidence relating to the
taxpayer’s claim, including evidence from persons employed in its laboratories throughout
the period of the claim and two accountants with knowledge of the financial affairs of the
taxpayer and its predecessors. The Tribunal accepted the evidence of the officials from the
predecessor Health Boards as to the nature of the business activities conducted, and thought
that it was reasonable to infer that the nature of the inputs would not have changed
markedly during the period of claim. The claim would therefore have been “essentially the
same” throughout the period from 1974 to 1997 (paragraph 189). The Tribunal further found
that the taxpayer and its predecessors made taxable supplies throughout the period of claim
(paragraph 190).
[30] The First-tier Tribunal then described the task before it (paragraph 193):
“[T]he Tribunal has to consider whether the business income of the laboratories can
be calculated. Whilst the evidence of the officials from the various Health Boards was
helpful in determining that taxable supplies were made, that evidence fell short of
facilitating its quantification. While the business income is almost certainly
significant, the Tribunal does not consider that it has been quantified satisfactorily
for the whole period. The Blue Books do not record income except for that arising
from catering. Such other accounting information as is available is insufficient to
show each element of sectorized income of each Health Board. There was for a time
one common VAT return which was submitted for all Health Boards, stating only
aggregate figures. No sales ledgers or copy tax invoices were produced to support
the quantification of taxable supplies. Consequently the amount of output tax paid to
Page 19 ⇓
19
HMRC in respect of the business income received by the laboratories and included in
those common VAT returns has not been established. In the absence of those figures
it is difficult to confirm the reasonableness of any input tax claim relating to that
income”.
On that basis, while the officials’ testimony about business supplies was accepted, that
evidence was “not sufficiently precise” to use as a basis for the quantification of the claim
throughout the relevant period.
[31] The taxable percentage of the taxpayer’s income was agreed at 14.7% in respect of
the tax year 2006/07, but that agreement did not extend to other years. The taxpayer’s
advisers had attempted to extrapolate that figure back to the end of the relevant period (in
1997) and then further back to the start of the relevant period (in 1974). On this aspect of the
claim, the Tribunal stated:
“196. The Tribunal does not consider such an approach reasonable or acceptable.
While the witnesses confirmed that there had been no changes to the general pattern
of activity, there had not been any reference to reliable primary data. The time-scale
involved also undermines the likely accuracy of the process of extrapolation. There is
an interval of ten years between the end of the relevant period and 2006/07, and that
is preceded by a taxable period of about 25 years. The value of the claim (about
£900,000 as now adjusted) is substantial. The ratio of each activity might vary over
an extended period: so too might profit margins. The Tribunal finds that there is no
written agreement concerning the use of 14.70% for any period. It was used in
calculating an agreed amount recoverable in 2006/07. All of this tends to undermine
the validity of 14.70% as a business/non-business fraction used over an extended
period. The Tribunal noted that another Health Board had used a fraction of 35%. No
invoices or VAT returns were available.…
197. The Tribunal would suggest that there is a need to have a verifiable
percentage, calculated by reference to prime records at regular intervals. For
example, it might well be acceptable in a 25 year period to have verifiable figures
every five years, and if there is not significant variation, to use extrapolated figures
for the intervening four Years. The Tribunal observes that in the actual calculation of
the Claim… 14.70% was not used throughout 11% was used, and also 12.15%”.
On that basis the claim was refused. The Tribunal went on to consider the possible relevance
of partial exemption, but for the reasons given by the judge of the Upper Tribunal we are of
opinion that this played no material part in their decision to reject the claim.
Page 20 ⇓
20
The decision of the Upper Tribunal
[32] The argument for the taxpayer before the Upper Tribunal was heavily focused on the
supposed failure of the First-tier Tribunal to distinguish properly between business/non-
business apportionment and partial exemption. It was further contended that, largely
because of that confusion, the First-tier Tribunal had failed to appreciate that the taxpayer’s
calculation of its claim was supported by sufficient evidence and fell within the range of
acceptable methods. Those arguments were rejected by the Upper Tribunal. The Upper
Tribunal’s reasoning was largely concerned with the supposed failure to distinguish
between business/non-business apportionment and partial exemption. It was held that,
while the First-tier Tribunal’s decision made reference to partial exemption (at
paragraphs 200 et seq), its primary reasoning was concerned entirely with the business/non-
business apportionment, and that was not vitiated by the later, essentially incidental,
reference. On that matter we agree entirely with the reasoning and conclusion of the judge
of the Upper Tribunal. The Upper Tribunal gave some consideration to the First-tier
Tribunal’s reasoning set out above, at paragraphs [196]-[197], and held that the extrapolation
exercise put forward by the appellant did not provide a sufficiently precise and satisfactory
basis for a claim for the period between 1974 and 1997. On that basis the First-tier Tribunal
was entitled, on the material before it, to conclude that the appeal failed.
The taxpayer’s grounds of appeal
[33] The taxpayer’s first ground of appeal relates to the primary matter argued before the
Upper Tribunal. It is said that the First-tier Tribunal had applied the wrong legal test to the
appeal, in that it approached the appeal as if it were governed by principles of partial
Page 21 ⇓
21
exemption rather than the principles of business/non-business apportionment. The second
and third grounds in effect run together. The second ground is that the First-tier Tribunal
erred in law in its approach to the evidence, and in particular in its approach to the burden
and standard of proof called for in a case of this nature. The third is that the First-tier
Tribunal and the Upper Tribunal erred in law by failing to determine the appeal in
compliance with the principle of effectiveness, which is binding as a general principle of EU
law. In particular, it is said that HMRC should assist the taxpayer and the First-tier Tribunal
in ascertaining a fair and just amount of under-recovered value added tax to be repaid to the
taxpayer. Alternatively, the EU law principle of effectiveness required the First-tier Tribunal,
or the Upper Tribunal on appeal, to find sufficient facts to enable an appropriate
methodology to be determined to secure that a fair and just amount of under-recovered
value added tax may be ascertained for repayment.
[34] The grounds of appeal are followed by a number of questions for the court’s
determination. To some extent these raise collateral issues, which it is unnecessary to
consider. Two questions are significant, however. The first (question No. 2 in the appeal) is
whether the fundamental right to repayment of overpaid VAT, guaranteed by EU law and
the principle of effectiveness, means that even if the taxpayer’s methodology for calculating
the amount of repayment were rejected, HMRC, and the Tribunals, were not permitted to
reject completely the taxpayer’s claim for repayment of input VAT solely on the basis of
difficulties with identifying a satisfactory methodology or difficulties of proof. That is
followed by a supplementary question as to whether the appropriate course was for the
Tribunals to make findings of fact on the basis of the evidence that was available and to rely
on those in determining an appropriate amount of repayment of input VAT. The second
question (question No. 4 in the appeal) is whether, having regard to the fundamental EU
Page 22 ⇓
22
law requirement that the right to recover overpaid VAT should not be rendered excessively
difficult or impossible in practice under domestic law (the principle of effectiveness), HMRC
and the Tribunals should adopt a flexible approach to the burden and standard of proof in
connection with historical claims for repayment. We should indicate at this stage that we
would answer both of those questions in the affirmative, for reasons that are discussed in
detail at a later stage in this opinion.
The calculation of historical claims for recovery of input tax
[35] In historical Fleming claims a problem that is frequently, perhaps normally,
encountered is lack of evidence. The reasons for this are obvious. Financial records for
periods 20 or 30 years previously are likely to have been destroyed or filed in such a way
that they cannot be traced. Notwithstanding the absence of direct records, however, some
evidence will usually be available to support a historical claim. In the present case, for
example, full financial records are available for the year 2006/2007, and evidence was led,
and accepted by the First-tier Tribunal, that the activities of the taxpayer’s laboratories had
not changed significantly throughout the period from 1974 to 1997 and that the taxpayer’s
claim had throughout been for unrecovered input tax attributable to taxable business
supplies made at the taxpayer’s laboratories. In these circumstances the obvious procedure
that will be relied on by a taxpayer is a backward extrapolation from the year for which full
records are available, on the basis that as the nature of the business has not changed
materially the inputs into the business, and the tax paid on those inputs, are unlikely to have
changed materially either.
Page 23 ⇓
23
Effectiveness
[36] In assessing historical Fleming claims, the effectiveness principle is likely to be of
great importance. As we have noted, that principle is based on the general proposition that,
if a legal right exists, a remedy should be available to secure the practical availability of that
right. Any requirements of proof must not make it excessively difficult to enforce the right:
San Giorgio, supra, at paragraph [14]. The court indicated in that case that presumptions or
rules of evidence that place an undue burden on the taxpayer will contravene the
effectiveness principle. The formulation of those examples, however makes clear that they
are only examples (“That is so particularly in the case of presumptions or rules of evidence”:
paragraph [14]). In our opinion the principle is not confined to what may be described as
legal rules relating to matters such as the availability of evidence or the burden of proof. It
also extends to the practical attitude taken by a court or tribunal to the evidence led in a
particular case. That applies in particular to the willingness of the court or tribunal to draw
inferences from such primary evidence as is available; if too strict an approach is taken to
the ability to draw inferences, it may well be “excessively difficult” (or even “practically
impossible”) to establish a claim. In the proof of historical claims the drawing of inferences
is likely to be of particular importance because of the typical restrictions in the primary
evidence.
[37] As we have indicated (paragraph [18]), the application of the effectiveness test
generally involves a form of proportionality exercise, in which a number of elements must
be balanced. In relation to the drawing of inferences from restricted primary facts, regard
must obviously be had to the deficiencies in the facts themselves and to the reasons for
them. The second of these factors is likely to be important in historical Fleming claims,
because an important reason for the taxpayer’s failure to make a claim previously, and to
Page 24 ⇓
24
keep accounting and other records that would have supported such a claim, will typically be
that national legislation prevented claims for the recovery of input tax, notwithstanding the
existence of the right to recover input tax in EU law. That is the case in the United Kingdom,
and in our opinion it is an important factor in the present case. It is also necessary to have
regard to the reasoning that underlies the inference that the taxpayer seeks to draw. In a
historical Fleming claim this will typically be an extrapolation backwards from the period
when clear accounting information exists. That extrapolation must be considered critically to
discover whether it is reasonable in all the circumstances. In this respect, it is to be expected
that there will at least be evidence as to whether the taxpayer’s business activities have
changed significantly over the period in question; if they have not that obviously supports
the drawing of the necessary inference. It is also to be expected that there will be evidence as
to whether the level of taxable inputs in respect of those activities has changed significantly;
once again if there is no significant change that supports the inference. It may also be
relevant to consider the explanation given for the non-availability of any direct evidence for
the period in question. If that is the fault of the taxpayer, that may justify a strict approach to
the primary evidence and a restriction on the extent to which inferences may be drawn from
such evidence as is available.
[38] Nevertheless, we think it important to bear in mind that in historical Fleming claims
the responsibility for failure to make the claims when full evidence was still available lies in
considerable measure with the state – in legislation that did not conform to the EU law right
to recover input tax and, in the case of the United Kingdom, in the convoluted attempts to
impose time limits on historical claims. In the case of NHS Boards and the Trusts that
preceded them, there is the further factor that their VAT affairs were at one time under
government control, and at that time, as a matter of policy, no attempt was made to recover
Page 25 ⇓
25
input VAT in respect of laboratory activities. All of these are factors that suggest that a
reasonably generous approach should be taken to historical claims to recover input tax in
cases such as the present.
[39] Ultimately the exercise involved is one of judgment, weighing up the strength of the
evidence that is available, the reasons for the absence of other evidence, and the coherence of
the reasoning that is said to support an inference as to the amount of input tax that is
recoverable on the basis of the evidence that is available. The underlying question is
whether the taxpayer’s right to recover input tax can be made effective in practice. In this
connection, we would draw attention to the fact that the Tribunal’s task in cases such as the
present is merely one of quantification; it is not the establishment of the right to repayment
of input tax; it is a matter of agreement that such a right exists. Counsel for the taxpayer
placed considerable emphasis on this element, and mentioned that in other areas of the law,
notably reparation for personal injury sustained at work, the courts had been willing to take
a liberal approach towards requirements of evidence and proof. Examples given included
Nimmo v Alexander Cowan & Sons Ltd, 1967 SC (HL) 79, and cases such as Fairchild v
Glenhaven Funeral Services Ltd, [2003] 1 AC 32 (where the logical aspect of causation was
ignored to produce a result that did practical justice). We think that there is a limit to how
far cases of that nature can be followed in other areas of the law. The fundamental point
rather appears to us to be that the effectiveness principle, and indeed so far as domestic
Scots law is concerned the principle ubi jus ibi remedium, require that the court should take a
flexible attitude to the question of remedies, ensuring so far as possible that the parties’
substantive rights are given effect.
Page 26 ⇓
26
Evidence
[40] In matters governed by EU law it is generally for the national legal system to
determine what evidence is acceptable, although that is always subject to the EU principle of
effectiveness and the further principle of equivalence, under which the rules applicable to a
claim based on EU law should be equivalent to those that apply to a comparable claim based
on wholly domestic law. The treatment of evidence in proceedings was considered by the
Court of Justice in Case C- 621/15, NW v Sanofi Pasteur MSD SNC, [2018] 1 CMLR 16. The
context in that case was EU-based consumer law relating to medical products; it was alleged
that the administration of a vaccine had produced multiple sclerosis. Nevertheless the
principles stated by the Court appear to us to be of general application. The Court noted (at
paragraph 25) that “under the principle of procedural autonomy and subject to the
principles of equivalency and effectiveness, it is for the national legal order of each Member
State to establish the ways in which evidence is to be elicited, what evidence is to be
admissible before the appropriate national court, or the principles governing that court’s
assessment of the probative value of the evidence adduced before it and also the level of
proof required”. The court then (at paragraph 26) restated the principle of effectiveness in
standard terms, to the effect that the detailed procedural rules governing actions for
safeguarding rights that individuals derive directly from EU law must not render
“practically impossible or excessively difficult” the exercise of rights conferred by EU law.
[41] In relation to the assessment of evidence, the Court of Justice stated (at
paragraphs 30-31):
“30 It should also be stated, particularly since, as mentioned by the referring
court, medical research neither confirms nor rules out a link between the
administering of the vaccine and the occurrence of multiple sclerosis, that
evidentiary rules that rule out any recourse to circumstantial methods and provide
that, in order to meet the burden of proof provided for in [the] Directive, the victim is
Page 27 ⇓
27
required to produce certain proof based on medical research of the existence of a
causal link between the defect attributed to the vaccine and the appearance of the
disease, disregards the requirements resulting from that Directive.
31 [S]uch a high evidentiary standard, which amounts to excluding any method
of proof other than certain proof based on medical research, could make it
excessively difficult in many situations or, as in the present case, where it is common
ground that medical research neither confirms nor rules out the existence of such a
causal link, impossible to establish producer liability, thereby undermining the
effectiveness of art. 1 of Directive 85/374 ...”.
In our opinion that case is authority for the view that a national court or tribunal, in giving
effect to EU-derived legal rights, must not apply national rules of evidence that prevent
reliance on any evidence that has, in practical terms, a bearing on the claim that is made.
That appears to us to accord with common sense. In general, any evidence that can
reasonably be considered relevant should be taken into account.
[42] In cases such as the present which involve historical claims for the recovery of
overpaid tax, it appears to us that the same approach should be taken. Evidence should only
be excluded if it cannot reasonably be considered as having a bearing on whether or not the
taxpayer’s claim has been established. Even if a particular piece of evidence appears
relatively weak, it will normally be appropriate to admit the evidence and then to take
account of the overall strength or weakness of the taxpayer’s case at the stage where the
tribunal considers the question of whether the standard of proof on the balance of
probabilities has been met. If a particular item of evidence is weak, it will obviously not
carry great weight at that stage (see below).
The standard of proof
[43] In civil litigation the norm, and indeed the default position, is that proof is required
on a balance of probabilities. Obvious reasons exist for that. At a practical level it treats the
parties as equally as is possible given the fact that a person making a claim must establish
Page 28 ⇓
28
that claim. At a conceptual level it means that courts and tribunals proceed on the basis that
if an argument is more likely than its converse it will be followed. That is elementary
common sense. In some areas, notably criminal prosecution, the law departs from proof on a
balance of probabilities, but that is for important policy reasons. These do not apply to
reclaiming overpaid tax.
[44] Consequently we are of opinion that the standard of proof in a claim for unrecovered
input tax is the ordinary civil standard, proof on a balance of probabilities. The application
of that standard to the tax calculation is perhaps not entirely straightforward. Normally, a
tax calculation is carried out on the basis of detailed financial records, interpreted in the
light of accountancy practice, with arithmetical calculations carried out on the basis of the
information in those records as so interpreted. In those circumstances something
approaching certainty is perhaps achievable in a typical case. We think it likely that the
First-tier Tribunal had that sort of case in mind in the approach that it took to the evidence
available in the present case, and we can sympathize with the Tribunal’s tendency to follow
standard practice. In historical claims, however, much of the evidence is inevitably lacking,
and a typical tax calculation is impossible. The Tribunal must accordingly proceed on the
basis of the evidence that is available, and the critical question for it is whether on the basis
of that evidence it is possible to hold that the taxpayer has proved its calculation of
unrecovered input tax on a balance of probabilities – in other words, whether the taxpayer’s
calculation is, at a general level, more likely than the contention that no unrecovered input
tax, or a much lesser amount of such tax, is due. The words “at a general level” are
important; a historical claim will almost inevitably be based on evidence that is manifestly
incomplete, and the Tribunal must do its best with the evidence that is available. Ultimately
the question is whether, in general terms, the taxpayer’s calculation is more likely to be
Page 29 ⇓
29
correct than the counter argument for HMRC that no or significantly less input tax is due to
be repaid.
[45] So far as deficiencies in the evidence are concerned, it is of course material for the
Tribunal to consider who is responsible. If the fault is that of the taxpayer, for example
through not keeping proper records or destroying records prematurely, it might be
legitimate for the Tribunal to draw the inference that the taxpayer has failed to make out its
calculation of input tax. Where, on the other hand, the absence of evidence relates to a
historical claim which could not have been brought earlier because of deficiencies in the UK
legislation, or government practices relating to claims for input tax by public authorities,
such evidence as is available should be taken into account. On this aspect of the case, the
significance of deficiencies in the evidence is likely to depend on two factors: (i) whether
there is an explanation for those deficiencies that is not the fault of the taxpayer; and (ii)
whether the deficiencies are sufficiently serious to make the taxpayer’s calculation of
unrecovered input tax fail on the balance of probabilities.
[46] Finally, on this aspect of the case, we should note that in historical claims estimates
will frequently be used. That is not unusual in other areas of the law, for example damages
for breach of contract or personal injuries. While these differ from tax claims, we cannot see
any reason that an estimate should not be used in any case where a calculation has to be
carried out on the basis of incomplete information. The critical question in these cases is
whether the premises on which the calculation is based appear sound, having due regard to
the deficiencies in the available evidence and whether on the basis of the evidence that is
available the calculation has been carried out in a manner that is objectively reasonable and
in accordance with proper tax and accounting practice.
Page 30 ⇓
30
Role of the First-tier Tribunal
[47] As we propose to remit this case for reconsideration to the First-tier Tribunal, it is
appropriate that we should comment briefly on the role of that Tribunal in relation to
historical claims based on incomplete evidence. The Court of Justice has recognized that EU
law does not require national courts to raise issues concerning the breach of provisions of
such law on their own motion. In accordance with the practice followed in the domestic law
of most member states, the Tribunal is entitled to adopt a passive role and to confine itself to
the dispute defined by the parties themselves, relying on facts and circumstances put
forward by those parties: see the joint Cases C- 430/93 and 431/93, van Schijndel and van Veen
v Stichting Pensioenfonds voor Fysiotherapeuten [1996] 1 CMLR 801, at paragraphs 19-22. That
accords with the domestic law of the United Kingdom, and it is an approach that the
Tribunal should follow in this case.
[48] The role of the First-tier Tribunal has been further considered by the Upper Tribunal
in recent cases, two of which are relevant for present purposes. First, in Lothian NHS Health
Tyre (at paragraph [19]), held that the onus of proving that an amount of input tax had been
paid and not recovered rested upon the taxpayer and that the standard of proof was the
balance of probabilities. We agree with those propositions. Lord Tyre further held that it
was open to the Tribunal to hold that the claim failed for either of two reasons: either
because it was not satisfied on the balance of probabilities that there was any unrecovered
input tax, or because the Tribunal, although satisfied that there was unrecovered input tax,
was “unable to find, on a balance of probabilities, that any particular – even a minimum –
amount of input tax can be ascertained as having been paid and not recovered”. In the latter
case, it was observed that the Tribunal did not function as a detective with the duty to fix a
Page 31 ⇓
31
figure, even a minimum figure, for input tax paid but not recovered, regardless of the
quality of the evidence. We agree with those observations, but we should observe that the
evidence that was available in that case, although relating to the same Health Board as the
present case, was significantly less than that before the First-tier Tribunal in the present case.
For that reason the decision on the facts is distinguishable.
[49] Secondly, in Revenue and Customs Commissioners v General Motors (UK) Ltd
[2015] UKUT 605 (TCC); [2016] STC 985, the Upper Tribunal considered a submission that the
First-tier Tribunal had misconceived its role as providing a solution rather than determining
whether the claim advanced by the taxpayer had been proved on the balance of
probabilities. The Upper Tribunal (at paragraphs [67]-[68]) accepted that the role of the First-
tier Tribunal was to determine whether it was satisfied, on the basis of the evidence adduced
and to the civil standard of proof on the balance of probabilities, that the taxpayer had
overpaid VAT (in this case on deemed self-supplies of cars between 1987 and 1992). As a
step in its reasoning, the First-tier Tribunal had required to consider and evaluate the
methodology for ascertaining the ratio of the cost of production of such cars to the list price
throughout the whole of the claim period. The First-tier Tribunal were not, however,
confined to choosing whether to accept or reject an expert witness’s model in its entirety.
“68 ... So far as they could properly do so, it was their duty (applying their own
expertise as a specialist tribunal) to ascertain the true amount of VAT (if any) which
[the taxpayer] had overpaid. This result could be achieved either by the FTT
performing the appropriate calculations itself, or by stating the principles by
reference to which they considered the calculation should be made. In performing
this task, the FTT had to act with procedural fairness, and there had to be a proper
evidential foundation both for their findings of fact and for their conclusions. But
their preferred solution did not have to be one for which either side had specifically
contended, either before or in the course of the hearing”.
Page 32 ⇓
32
We agree with the foregoing observations, and we would note that the Tribunal went on to
(at paragraph [71]) approve of the observations of Lord Tyre that we have discussed in the
last paragraph.
Application to the present case
[50] The First-tier Tribunal heard extensive evidence on behalf of both the taxpayer and
HMRC, and on the basis of that evidence it made extensive findings in fact. We have
narrated what we consider to be the critical findings at paragraph [27] above; these must
obviously be read in the light of the agreed facts set out at paragraph [25]. The agreed facts
disclosed that the claim for unrecovered input tax on its laboratory activities made by the
taxpayer for the year 2006/2007 was accepted by HMRC, but that that agreement did not
extend to earlier years. In that year the VAT recovery fraction for the taxpayer’s laboratories’
business activities was agreed at 14.7%; that figure was derived from a survey of the
business activities in that year. It was also agreed that during the claim period the taxpayer
and its predecessors carried out business activities through its laboratories; those are briefly
described in the findings in fact, although a much longer description is contained in the
summary of the evidence in the First-tier Tribunal’s decision.
[51] In relation to the findings in fact, the first important finding was that the evidence
given by officials from the taxpayer and its predecessors as to the business activities
conducted in the laboratories was accepted. Secondly, it was accepted that the nature of the
laboratories’ activities did not change during the period of the taxpayer’s claims. Thirdly,
the First-tier Tribunal held that it was reasonable to infer that the nature of the inputs would
not have changed markedly throughout that period. Fourthly, the claim was for
unrecovered input tax, and would have remained essentially the same throughout the
Page 33 ⇓
33
period of the claim, from 1974 to 1997. Fifthly, the parties had agreed that the taxable
percentage of the taxpayer’s outputs in the year 2006/2007 was 14.7%, that being the figure
attributable to the taxpayer’s laboratory activities.
[52] The fundamental argument for the taxpayer is that in the light of the findings that
the activities carried out in the taxpayer’s laboratories did not change during the period of
the claim to any material extent, and that the nature of the inputs would not have changed
markedly throughout that period, the figure of 14.7% agreed for 2006/2007 should be
extrapolated backwards to earlier years. That was supported by the specific finding by the
First-tier Tribunal that the claim for unrecovered input tax would have remained essentially
the same throughout the period from 1974 to 1997. Counsel for the taxpayer emphasized
that all of the evidence that was now available about the activities of the taxpayer during the
period from 1974 to 1997 had been provided to the First-tier Tribunal; that was not disputed
by HMRC. It was submitted that in the circumstances the Tribunal should have attempted to
carry out the extrapolation exercise, calculating the amount of input tax that is likely to have
been paid by the taxpayer in respect of business activities in its laboratories during the claim
period.
[53] The taxpayer put forward a calculation performed on this basis. It was founded on
such documentary evidence as remained available. This in turn was based on the VAT
incurred in each of the tax years from 1974/75 to 1996/97. Estimates were made of the
amount of income that was attributable to NEQAS, clinical trials and public health and
biochemistry, the three main business areas of the taxpayer’s laboratory activities. On the
basis of those estimates a taxable percentage of those activities was calculated; this varied
from 11% to 14.7% in each of the relevant years. 14.7% was of course the figure that was
accepted by HMRC for the year 2006/07. Those percentages were used to calculate the
Page 34 ⇓
34
estimated amount of input tax for each year. Obviously assumptions are made in this
calculation, including the propositions that the taxpayer’s business activities and the claim
for unrecovered input tax would not have changed markedly over the period in question.
[54] The First-tier Tribunal rejected the taxpayer’s argument. At paragraph 193, which is
quoted at paragraph [30] above, it stated that it had to consider whether the business income
of the laboratories could be calculated, and expressed the view that, while evidence of
officials was helpful in determining that taxable supplies were made, it “fell short of
facilitating its quantification”. The Tribunal therefore concluded that the evidence did not
provide satisfactory quantification for the whole period. Reference is made to the fact that
no sales ledgers and copy tax invoices had been produced to support the quantification of
taxable supplies. Consequently the Tribunal concluded that the amount of output tax paid
to HMRC in respect of the laboratories’ business income had not been established. It
concluded that “In the absence of those figures it is difficult to confirm the reasonableness of
any input tax claim relating to that income”.
[55] In our opinion this imposes too high a test on the taxpayer. The First-tier Tribunal
accepted that the evidence from witnesses helped to determine that taxable supplies were
made, but held that such evidence was not enough to facilitate quantification. The fact that
taxable supplies were made was not in dispute, however. The only issue in dispute was
quantification. It was accepted that full accounting information was not available for the
period in question; the primary reason for this was the historical nature of the claim, which
was in large measure the result of the failure of the United Kingdom to implement properly
the EU law right to reclaim input tax. It was also caused by the decisions made by the
Scottish Office, a government ministry, to centralise the administration of VAT and not to
make any claims in respect of input tax.
Page 35 ⇓
35
[56] Furthermore, it appears from a letter sent by HMRC to the taxpayer’s representatives
on 21 April 2015 that records held by HMRC for the period before 1997 were destroyed; in
2010 all documents older than six years were disposed of. Thus the state has to bear a major
degree of responsibility for the absence of accounting records. In the foregoing
circumstances the failure of the taxpayer to produce sales ledgers or copy tax invoices
cannot be considered significant, as the First-tier Tribunal appears to have thought. The last
two sentences of paragraph 193 suggest that the “reasonableness” of any input tax claim
could not readily be confirmed in the absence of full business records such as sales ledgers
or tax invoices. With respect, this appears a non sequitur. The sales ledgers and tax invoices,
if available, would serve to establish the claim itself by reference to primary evidence. That
is to say, they would establish the actual amount of the claim rather than the reasonableness
of any amount claimed.
[57] In the present case it was plainly impossible to establish the actual amount of the
claim. Consequently the question came to be whether the taxpayer could establish its claim
on the balance of probabilities by reference to secondary evidence and inferences drawn
from such evidence. In addressing that question, the taxpayer relied on such evidence as
was available, making assumptions about the continuity of its business operations over the
period of the claim. The question of reasonableness goes to the nature of the secondary
evidence that was available, the reasonableness of the assumptions made about that
evidence, and the cogency of the inferences drawn from that evidence to establish the final
figures relied on by the taxpayer. The problem with the approach by the First-tier Tribunal
at paragraph 193 is that it relies fundamentally on the absence of primary evidence, and
does not address the main issues in the taxpayer’s argument, namely the inevitable reliance
on secondary evidence and inferences drawn from that evidence.
Page 36 ⇓
36
[58] At paragraphs 196-197, the First-tier Tribunal addressed the calculation of the taxable
percentage; these paragraphs are set out at paragraph [31] above. It declined to extrapolate
the figure of 14.7% for 2006/07 to earlier years. The Tribunal referred specifically to the
evidence that there had been no changes to the general pattern of activity over the period of
the claim (a matter that was held established in the findings in fact), but it rejected the
extrapolation exercise put forward by the taxpayer on essentially five grounds. First, it was
said that there was no reference to “reliable primary data”. On this matter we once again
observe that it was quite clear that adequate primary data, in the form of ledgers and
invoices, was not available, for reasons that we have discussed. The critical question was
whether it was possible to deal with that difficulty by relying on secondary data and
inferences drawn from such data. That question does not appear to us to have been
adequately addressed.
[59] Secondly, it was said that the time-scale involved undermined the likely accuracy of
the process of extrapolation, with intervals of 10 years between 2006/07 and the end of the
claim period and nearly 25 years before that. That is undoubtedly a relevant factor. The
taxpayer sought to address it by relying on the evidence of its witnesses to the effect that the
nature of its laboratory activities did not change during the claim period; that evidence was
accepted by the First-tier Tribunal. The taxpayer further relied on an inference, again
accepted by the Tribunal, that the nature of the inputs would not have changed markedly
throughout the period. It also relied on the fact, again accepted by the Tribunal, that the
claim all along had been for unrecovered input tax attributable to taxable business supplies
made throughout the period, and that the claim remained essentially the same throughout
that period. (The findings in fact are set out at paragraph of [27] above). These are matters
Page 37 ⇓
37
that the First-tier Tribunal required to address in considering the materiality the timescale
involved, but it is not apparent that it did so.
[60] Thirdly, the Tribunal refers at paragraph 196 to the fact that the ratio of each activity
might vary over an extended period, as might profit margins. That is no doubt correct so far
as it goes, at a hypothetical level, but the task of the Tribunal was rather to examine the
evidence from the taxpayer’s witnesses to the effect that laboratory activities and inputs
would not have changed markedly over that period. With all due respect, the fact that in
theory the ratio of the various activities might vary, together with profit margins, is not a
complete answer to the taxpayer’s methodology; it is merely a factor that may be taken into
account in assessing that methodology, and there is no indication that the First-tier Tribunal
performed that task. Furthermore, while the factors referred to might vary, there is
obviously the possibility that the variations might to a substantial extent cancel one another
out.
[61] Fourthly, the Tribunal criticize reliance on the figure of 14.7% as a business/non-
business fraction over an extended period. This is essentially a development of the last
point; it is a relevant factor, but it must be set against the evidence that the taxpayer’s
business activities and inputs were broadly constant over the period in question. Fifthly, the
Tribunal noted that no invoices or VAT returns were available. This of course results
primarily from the historical nature of the claim. The taxpayer’s case accepts the absence of
primary evidence, but seeks to quantify the amount of unrecovered input tax by reference to
secondary evidence and inferences drawn from that evidence.
[62] The First-tier Tribunal continued, at paragraph 197 (set out at paragraph [31] above)
by suggesting that “there is a need to have a verifiable percentage, calculated by reference to
prime records at regular intervals”. It was suggested that verifiable figures might be
Page 38 ⇓
38
provided every five years, and if there were no significant variation it might be acceptable to
use extrapolated figures for the intervening four years. The fundamental point here appears
to be that without primary records such as invoices or ledgers it is impossible to verify a
claim, and if the claim extends over a period such documentation must be available at
regular intervals, possibly every five years. Once again this fails to address the fundamental
problem that, largely as a result of matters for which the United Kingdom government was
responsible, such primary evidence did not exist for the claim period. What was material,
therefore, was whether the taxpayer’s alternative methodology was sufficient to quantify its
claim on a balance of probabilities test. That question is not addressed.
[63] In all of the above, we would stress two particular matters. The first is that the
existence of a claim to recover input tax on business activities is not in issue; all that is in
dispute is the quantification of that claim. That means that some amount of input tax must
be due to the taxpayer. If the First-tier Tribunal were correct, none of that would be
recovered. In a case where it is clear that some repayment of tax is due by HMRC, it should
normally be possible to reach some sort of quantification of the amount due. Indeed, it can
be said that exceptional circumstances would be required to render such quantification
impossible. If necessary the Tribunal should perform the appropriate calculations itself, or at
least state the principles by reference to which it is thought that the calculation should be
made; in this connection we would refer to the decision of the Upper Tribunal in Revenue
and Customs Commissioners v General Motors (UK) Ltd, supra, discussed at paragraph [49]
above. We would also add that in a case where it is accepted that a repayment of input tax is
due but the final amount appears uncertain, it is always possible as a last resort to discount
the resulting figure to reflect that element of uncertainty. We emphasize that this is a last
Page 39 ⇓
39
resort, and that it is necessarily a very approximate procedure, but it is a technique that can
be used to ensure that a claim that undoubtedly exists does not fail entirely.
[64] The second matter that we would emphasize is the principle of effectiveness, which
as we have observed is a fundamental principle of EU law, and has parallels in domestic
law. This is a case where it is accepted that input tax was not claimed during the relevant
period for reasons that are not the fault of the taxpayer. In view of the lapse of time, only
very limited records are available, but the taxpayer has produced all of the evidence that is
now available. It is on the basis of that information, with the extrapolation from the figure
for 2006/07, that it is sought to quantify the claim. The principle of effectiveness appears in
our opinion to demand that if at all possible that claim should be quantified.
[65] The critical question is accordingly whether, in the light of the absence of primary
evidence, the taxpayer has succeeded in proving the quantification of its claim on the
balance of probabilities by using such secondary evidence as exists and drawing inferences
from that evidence. The secondary evidence includes the quantification of the claim that the
taxpayer has produced, and also the testimony of the witnesses who gave evidence. This is
not a test of certainty or near certainty: the critical question is whether the taxpayer’s
calculation of the amount due is more likely to amount to a proper quantification of its claim
than the alternative, which appears to be that no input tax is repayable because of the
impossibility of quantification. That question has not in our opinion been properly
addressed by the First-tier Tribunal.
[66] For the foregoing reasons we are of opinion that the first of the two questions
referred to at paragraph [34] above should be answered in the affirmative. The right to
repayment of overpaid VAT is guaranteed by EU law and the principle of effectiveness. That
means that the Tribunal (and HMRC) should make findings of fact on the basis of the
Page 40 ⇓
40
evidence led and to use those to draw appropriate inferences as to the level of VAT that is
due. In this connection we would refer to the principles put forward in the cases cited at
paragraphs [47]-[49] above, in which the role of the Tribunal is discussed. Ultimately, the
difficulties of proof are not sufficient to reject the possibility of quantification. Even if such
difficulties exist, the First-tier Tribunal must attempt to identify a satisfactory methodology
to permit quantification of the amount of input tax that is due, it being accepted that the
taxpayer is entitled to some repayment of input tax.
[67] We are further of opinion that the second of the questions referred to at
paragraph [34] should be answered in the affirmative. We consider that the Tribunals and
HMRC should adopt a flexible approach to the burden and standard of proof in connection
with historical claims for repayment. The fundamental problem in such cases is that primary
evidence does not exist owing to the lapse of time. The absence of such evidence, at least in
cases such as the present, is not the fault of the taxpayer, and the lack of evidence should not
be held against the taxpayer. In this connection we would refer to our previous discussion of
evidence (at paragraphs [40]-[42]) and the standard of proof (at paragraphs [43]-[46]). It is
also relevant to bear in mind the importance of the overriding principle of effectiveness,
which we have discussed at paragraphs [36]-[39].
[68] Finally, we should note the argument presented by the taxpayer to the effect that the
decision of the First-tier Tribunal was erroneous because it confused business/non-business
apportionment with partial exemption. We agree entirely that those concepts should not be
confused. Nevertheless, in view of our decision on the other aspects of the case, it is
unnecessary to consider this matter further. We would merely state that that we agree with
the decision of the Upper Tribunal to the effect that, although there is reference to partial
Page 41 ⇓
41
exemption in the decision of the First-tier Tribunal, it did not play a material part in the
fundamental reasoning of that Tribunal.
The court’s order
[69] We will accordingly allow the appeal, answering the two questions referred to above
(questions of law Nos. 2 and 4 in the Appeal) in the affirmative. We do not find it necessary
to answer the other questions.
[70] We consider that the appropriate procedure thereafter is to set aside the decisions of
the First-tier and Upper Tribunals and to remit the case to a differently constituted First-tier
Tribunal for reconsideration. Extensive evidence has been led, and it may well be possible
for the Tribunal to make use of that evidence in coming to a decision on the case. We would
not, however, seek to dictate the procedure that should be followed in this respect; details of
procedure are generally best left to the court or tribunal that requires to carry them out. The
fundamental task for the First-tier Tribunal is to reconsider the claim presented by the
taxpayer and the arguments for both parties. In doing so it should obviously have regard to
the considerations that we have set out in the earlier parts of this opinion.