J P Whitter (Waterwell Engineers) Ltd v HM Revenue and Customs

This appeal raises an important point of principle concerning the power of the Respondents, the Commissioners for Her Majesty’s Revenue and Customs (“HMRC”), to cancel the registration of a taxpayer for gross payment under the legislation which governs the Construction Industry Scheme (“the CIS”). The relevant legislation was contained, at all times material to the present appeal, in Chapter 3 of Part 3 of the Finance Act 2004 (“FA 2004”) and the Income Tax (Construction Industry Scheme) Regulations 2005 (“the 2005 Regulations”). The question of principle is whether, before exercising the power of cancellation conferred by section 66(1) of FA 2004, the Board of Inland Revenue (now HMRC) are obliged, or at least entitled, to take into account the impact on the taxpayer’s business of the cancellation of its registration for gross payment.

The CIS was introduced in order to counter widespread tax evasion by sub-contractors in the construction industry. Originally enacted in Chapter II of Part I of the Finance Act 1971, under the heading “Sub-Contractors in Construction Industry”, the legislation governing the scheme was re-enacted with substantial modifications in Chapter II of Part III of the Finance (No. 2) Act 1975 (sections 68 to 71, and schedules 12 and 13), and then consolidated in Chapter IV of Part XIII of the Income and Corporation Taxes Act 1988 (“ICTA 1988”) (sections 559 to 567). The legislation was then again re-enacted with modifications by FA 2004, under the heading “Construction Industry Scheme”. In that form, with minor amendments, it remains in force today.

The overall structure and purpose of the legislation has remained the same since the inception of the statutory scheme some 45 years ago. In a passage which has often been cited with approval in later cases, Ferris J described the background to the legislation, and the advantages to a sub-contractor of being registered for gross payment, in Shaw v Vicky Construction Ltd [2002] EWHC 2659 (Ch), [2002] STC 1544, (“Vicky”) at [2] to [5]:

“2.

Vicky is engaged in the construction industry. In the course of its business it does work in that field as a sub-contractor engaged by another company (the contractor).

3.

In the absence of the statutory provision with which this appeal is concerned Vicky would be entitled, like any other sub-contractor, to be paid the contract price in accordance with its contract with the contractor without any deduction in respect of its own tax liability. However it became notorious that many sub-contractors engaged in the construction industry “disappeared” without settling their tax liabilities, with a consequential loss of revenue to the exchequer.

4.

In order to remedy this abuse Parliament has enacted legislation, which goes back to the early 1970s, under which a contractor is obliged, except in the case of a sub-contractor who holds a relevant certificate, to deduct and pay over to the Revenue a proportion of all payments made to the sub-contractor in respect of the labour content of any sub-contract. The amount so deducted and paid over is, in due course, allowed as a credit against the sub-contractor’s liability to the Revenue.

5.

The need to make and pay over such deductions can be an irritation to the contractor obliged to carry out this exercise. It also adversely affects the cash flow of the sub-contractor. Accordingly it is advantageous to a sub-contractor to have a statutory certificate rendering such a deduction unnecessary. The provision of such a certificate tends to make the sub-contractor holding the certificate a more attractive party for the contractor to deal with and, by enabling the sub-contractor to receive the contract price without deduction, improves the sub-contractor’s cash flow.”

The taxpayer and appellant in the present case is a relatively small family-owned and operated company, J P Whitter (Waterwell Engineers) Limited (“the Company”). As its name implies, the Company carries on business as water well engineers, drilling boreholes and wells for water companies, commercial and agricultural businesses and the domestic market. It operates throughout the United Kingdom, and at the time when its registration for gross payment was cancelled in August 2011 it had about 25 employees, including a number of family members on the administration side. The business had grown steadily since it was started by Philip Whitter in 1972, and its later incorporation in the 1980s. In the three years to 2011, the business had a turnover of approximately £4.4 million, making a net profit over the same period of about £180,000. Approximately £1.9 million of that turnover derived from contracts with United Utilities. Other major customers accounted for a further £900,000.

There is no finding of fact as to when the Company was first registered for gross payment, but it appears from the correspondence in the appeal bundle that this happened in about 1984. Thereafter, the Company’s registration was kept under regular review by HMRC. Latterly, at any rate, such reviews were generally carried out by computer on an annual basis. In July 2009 the Company failed a review for the first time, because of late payment of PAYE, and its registration was cancelled. The Company appealed, and in November 2009 its appeal was upheld. In the letter allowing the appeal, dated 12 November 2009, the officer with conduct of the matter in the Hull CIS Team gave a clear warning that in order to continue to benefit from gross payment status both the payment of tax and filing of returns had to be made on time. Under the new guidance, the rules would be applied strictly with no scope to allow for “minor and technical” failures as before. Reference was made to the latest guidance on the subject, which the Company was requested to bear in mind for the future.

Despite this warning, the Company failed its next annual review in June 2010 and its registration was again cancelled. The Company’s accountants, Wilds, appealed by letter dated 8 July 2010. The letter said:

“We note that you wish to withdraw this gross paying certificate due to late payment of PAYE. Whilst my client agrees that they have been tardy occasionally in making their PAYE remittances they have, I am sure you will agree, always been paid.

This business has been grown meticulously by the Whitter family to a position where it can now confidently apply for tenders from United Utilities and indeed have just obtained two very large contracts with them. The withdrawal of the gross paying certificate will quite obviously put those contracts in jeopardy as United Utilities will not deal with companies who do not have a gross paying certificate … The withdrawal of the certificate will therefore severely hinder the company who have managed to survive over a very difficult last eighteen months to two years.”

HMRC then mistakenly took the point that the appeal had been lodged out of time, when in fact it was clearly in time. When this was pointed out, Mrs D Smith of the Hull CIS Team replied on 20 August 2010, apologising for the error and saying:

“On this occasion I am prepared to overlook these failures, your appeal is upheld and the company will retain gross payment status.”

The letter recommended the Company to make all of its PAYE payments electronically, and again warned that if the Company was to continue to benefit from gross payment status both the payment of tax and filing of returns had to be made on time.

Unfortunately, this second warning was no more effective than the first, and on 30 May 2011 there was another annual review which the Company again failed due to late payment of PAYE. On this occasion, before making a decision whether to withdraw the Company’s gross payment status, the Hull CIS Team wrote to the Company on 20 June 2011 identifying the defaults and giving the Company an opportunity to provide an explanation for them, supported by documentary evidence, or to advise whether it had entered into a formal “time to pay” arrangement. The writer said that an informed decision could then be taken as to whether a “reasonable excuse” existed for some or all of the compliance failures.

The letter identified seven late payments of PAYE, and in due course the First-tier Tribunal found that the following late payments of PAYE had been made:

The Company replied to this letter on 14 July 2011. The reply was written by Ms Sally Whitter, a daughter of Philip Whitter, who was the company secretary and had worked for the Company for over 20 years. She apologised for the delays, which she said were due to “administrative oversights”. She continued:

“Whilst admitting to these oversights, which we will endeavour to prevent in future, we would point out that losing our gross status would prevent us tendering for contract work and thus cause the company to cease trading. I am sure you will agree that removing gross status will cause great hardship which is disproportionate to the level of the oversights discussed above.”

HMRC then wrote to the Company on 3 August 2011, stating that they were unable to accept the explanation for the compliance failures. The writer said:

“This is because you have not provided me with any documentary evidence, as requested, to support your claim. I also note that this is the third failed review and assurances have previously been given about the future compliance.”

As the FTT subsequently found, and is not in dispute, this is the letter which effectively evidenced HMRC’s decision to cancel the Company’s registration.

Wilds then appealed the decision on 22 August 2011, and also asked for it to be reconsidered. In their letter, Wilds said:

“Whilst we accept that our client has made multiple compliance failures these are of a trivial nature with payments being an average of three weeks late … we feel that the punishment is disproportionate to the crime.”

They then said that the Company’s major customer would only deal with businesses which had a gross status, so if that status were removed the customer would be lost, the Company would be unable to continue trading, and several people would have to be made redundant.

On 15 September 2011 HMRC refused the Company’s appeal, saying that no reasonable excuse had been provided for the failures. Wilds replied on 23 September, requesting an internal review. They said they were aware of other cases where gross status had been allowed to continue after a review on the basis of disproportionate hardship. The review was concluded on 12 December 2011, when the original decision was upheld. The reviewing officer clearly thought there was no reasonable excuse for the defaults, as indeed the Company now accepts. On the issue of proportionality, the officer considered that the decision of the High Court in the Hilton case (Barnes v Hilton Main Construction [2005] EWHC 1355 (Ch), [2005] STC 1532) meant that no test of proportionality could be read into the legislation, and neither HMRC nor (on an appeal) the FTT could take such a factor into account.

In due course, the Company prosecuted its appeal to the FTT. The appeal was heard by Judge Cannan and Mr Peter Whitehead on 9 August 2012. In its decision released on 18 October 2012 (“the FTT Decision”), the FTT accepted the evidence of Ms Whitter that United Utilities and other major customers would be likely to withdraw work from the Company if it lost its gross payment status. They found that the effect of losing such customers would be a fall of some 63% in turnover and profits. The number of the Company’s employees would probably shrink to some 5 or 6 from the current 25. The Company would not be able to tender for any utility or large commercial work. Even if the registration was lost and regained 12 months later, the nature of the work was so specialised they would not be able to recruit suitable employees. It would take the Company “10 years or so” to regain its present position: see [2012] UKFTT 639 (TC) at [47].

Having made these findings of fact, the FTT held that HMRC had erred in law in not taking into account the adverse effect on the Company’s business when exercising their discretion whether to cancel the registration. The FTT distinguished Hilton on the basis that it concerned an application for renewal of a certificate prior to its expiry, rather than the cancellation of an existing registration. The FTT therefore allowed the Company’s appeal, on the basis that HMRC had failed to take into account a relevant factor and the decision was accordingly wrong in law. The FTT considered that they had no power to substitute their own view based on the facts found by them, nor could they be satisfied that, even if HMRC had taken into account the effect on the Company’s business, the decision would inevitably have been the same: see the FTT Decision at [73].

HMRC then appealed to the Upper Tribunal (Warren J and Judge Colin Bishopp), which by its decision released on 13 July 2015 (“the UT Decision”) allowed the appeal and reinstated the cancellation notice: see [2015] UKUT 392 (TCC), [2016] STC 204.

The Company now appeals to this court, with permission granted by the Upper Tribunal on 20 August 2015.

The Company’s two grounds of appeal are:

The parties are represented by the same counsel as they were before the Upper Tribunal, namely Mr Thomas Chacko for the Company and Mr James Rivett for HMRC. I express my gratitude to both of them for their clear and concise submissions.

The main relevant legislation in FA 2004 and the 2005 Regulations is conveniently set out in the annex to the UT Decision, reproduced at [2016] STC 227 to 233. It is therefore unnecessary for me to set it out at length again. In the following summary, which is to a large extent based on that contained in the UT Decision at [7] to [18], references to section numbers are to sections of FA 2004.

Sections 57 to 59 introduce the CIS, and describe who are sub-contractors and contractors.

Section 60 defines “contract payment” as meaning any payment made under a construction contract by the contractor to a sub-contractor. A payment is not a “contract payment”, however, if the person to whom the payment is made is registered for gross payment at that time: see subsections (2) and (4). This is important, because section 61(1) obliges the contractor to make a deduction on account of tax when making a contract payment. The amount to be deducted is a sum equal to the relevant percentage (specified by regulations) of so much of the payment as is not shown to represent the direct cost of materials used in carrying out the relevant construction operations. It follows that there is no obligation to deduct where a contractor makes a payment to a sub-contractor who is registered for gross payment.

Section 63 deals with the registration of sub-contractors. There are two types of registration: registration for gross payment and registration for payment under deduction. In either case, certain documentation and information must be provided under section 63(1): nothing turns on this requirement. The section goes on to provide:

“(2)

If the Board are satisfied that the requirements of subsection (2), (3) or (4) of section 64 are met, the Board must register –

(a)

the individual or company, or

(b)

in a case falling within subsection (3) of that section, the individual or company as a partner in the firm in question,

for gross payment.

(3)

In any other case, the Board must register the individual or company for payment under deduction.”

The registration provisions are therefore highly prescriptive. If the relevant requirements of section 64 are satisfied, HMRC must register the applicant for gross payment, but in any other case the applicant must be registered for payment under deduction. It can also be seen that payment under deduction is the default position, which applies unless HMRC are satisfied that the requirements of registration for gross payment are satisfied.

The relevant requirements in the case of a company which is not a partner in a firm are those in section 64(4), which in turn requires the conditions in Part 3 of Schedule 11 to FA 2004 to be satisfied. Part 3 of Schedule 11, headed “Conditions to be satisfied by companies”, comprises paragraphs 9 to 12. By virtue of paragraph 9, the conditions in each of the following three paragraphs must be satisfied.

Paragraph 10 lays down the business test, and paragraph 11 the turnover test. Nothing turns on these conditions, because the Company has at all times fulfilled them.

Paragraph 12 lays down the compliance test. In view of its importance in the present case, I will set out the main provisions of the paragraph:

“12(1) The company must, subject to sub-paragraphs (2) and (3), have complied with –

(a)

all obligations imposed on it in the qualifying period [which by virtue of paragraph 14 means the period of 12 months ending with the date of the application] by or under the Tax Acts or the Taxes Management Act 1970 …; and

(b)

all requests made in the qualifying period to supply to the Inland Revenue accounts of, or other information about, its business.

(2)

A company that has failed to comply with such an obligation or request as –

(a)

is referred to in sub-paragraph (1), and

(b)

is of a kind prescribed by regulations made by the Board of Inland Revenue,

is, in such circumstances as may be prescribed by the regulations, to be treated as satisfying the condition in that sub-paragraph as regards that obligation or request.

(3)

A company that has failed to comply with such an obligation or request as is referred to in sub-paragraph (1) is to be treated as satisfying the condition in that sub-paragraph as regards that obligation or request if the Board of Inland Revenue are of the opinion that –

(a)

the company had a reasonable excuse for the failure to comply, and

(b)

if the excuse ceased, it complied with the obligation or request without unreasonable delay after the excuse had ceased.

(4)

[This sub-paragraph requires the company to have paid any national insurance contributions falling due during the qualifying period when they became due.]

(5)

The company must have complied with any obligations imposed on it by the following provisions of the Companies Act 1985 … in so far as those obligations fell to be complied with within the qualifying period –

[The relevant obligations are then listed, beginning with those relating to the contents, laying and delivery of annual accounts, returns of company officers and changes therein, annual returns, etc.]

…

(7)

There must be reason to expect that the company will, in respect of periods after the qualifying period, comply with –

(a)

all such obligations as are referred to in paragraphs 10 and 11 and sub-paragraphs (1) to (6), and

(b)

such requests as are referred to in sub-paragraph (1).

(8)

Subject to sub-paragraphs (2) and (3), a company is not to be taken for the purposes of this paragraph to have complied with any such obligation or request as is referred to in sub-paragraphs (1) to (6) if there has been a contravention of a requirement as to –

(a)

the time at which, or

(b)

the period within which,

the obligation or request was to be complied with.”

It can be seen that the requirements of the compliance test are again highly prescriptive, and in many cases (notably those relating to the due payment of national insurance contributions and compliance with obligations under the Companies Act 1985) do not admit of any exceptions. Moreover, the test is both retrospective (throughout the qualifying period) and prospective: by virtue of sub-paragraph (7), there must be reason to expect that the company will continue to comply with the specified obligations for the future. Furthermore, sub-paragraph (8) equates any delay in compliance with non-compliance. The exceptions in sub-paragraphs (2) and (3) (for what one might call prescribed minor failures to comply, and cases where the company had a reasonable excuse for the failure to comply) are obviously important, but apply only to obligations or requests falling within sub-paragraph (1).

The prescribed “minor failures”, as I have termed them, are set out in regulation 32 of the 2005 Regulations (“Regulation 32”). They concern obligations to submit various returns, to meet certain requests, and to make payments of tax (including under the PAYE regulations) on the due date. The prescribed circumstances all relate to late compliance with the relevant obligation or request, usually with a short period of grace being allowed. There is no discretionary element: either the non-compliance falls within the prescribed tolerance, in which case it has to be disregarded, or it does not, in which case, subject only to the “reasonable excuse” exception, the applicant cannot satisfy the compliance test.

As the Upper Tribunal point out at [14] of the UT Decision, the provisions now contained in Regulation 32 replaced the more general exception which was previously found in the predecessor legislation (most recently section 565(4) of ICTA 1988), where the compliance conditions were treated as satisfied in relation to a particular failure if the Board were of the opinion that “the failure is minor and technical and does not give reason to doubt that the [compliance conditions] will be satisfied”.

Reverting to FA 2004, cancellation of registration for gross payment is provided for by section 66, the key provisions of which state:

“66 (1) The Board of Inland Revenue (now HMRC) may at any time make a determination cancelling a person’s registration for gross payment if it appears to them that –

(a)

if an application to register the person for gross payment were to be made at that time, the Board would refuse so to register him,

(b)

he has made an incorrect return or provided incorrect information (whether as a contractor or as a sub-contractor) under any provision of this Chapter or of regulations made under it, or

(c)

he has failed to comply (whether as a contractor or as a sub-contractor) with any such provision.

(2)

Where the Board make a determination under subsection (1), the person’s registration for gross payment is cancelled with effect from the end of a prescribed period after the making of the determination …

…

(5)

On making a determination under this section cancelling a person’s registration for gross payment, the Board must without delay give the person notice stating the reasons for the cancellation.

(6)

Where a person’s registration for gross payment is cancelled by virtue of a determination under subsection (1) the person must be registered for payment under deduction.

…

(8)

A person whose registration for gross payment is cancelled under this section may not, within the period of one year after the cancellation takes effect …, apply for registration for gross payment.”

It is common ground that, when HMRC cancelled the Company’s registration for gross payment, the circumstances fell within paragraph (a) of section 66(1): in other words, if the Company had applied for registration for gross payment at that time, the application would have been refused. The Company’s defaults did not fall within the tolerances prescribed in Regulation 32, and (as the FTT had found) there was no reasonable excuse for them. The defaults in question were all late payments of PAYE. The first late payment was not excepted under Regulation 32; and those of the subsequent late payments which, in isolation, would have been excepted because they were not very late, were nonetheless not excepted because of the prior late payments. There were no defaults within paragraphs (b) or (c) of section 66(1) or, if there were, they were not relied on by HMRC.

Appeals are dealt with by section 67. A person aggrieved by the cancellation of his registration for gross payment may appeal by notice given within 30 days after the cancellation. The notice must state the person’s reasons for believing that his registration should not have been cancelled. The jurisdiction of the FTT on such an appeal includes jurisdiction “to review any relevant decision take by [HMRC] in the exercise of their functions under section … 66”.

Section 73(1) confers power on HMRC by regulations to make “such other provision for giving effect to this Chapter as they consider necessary or expedient”. We were not referred to any regulations made pursuant to this power apart from the 2005 Regulations.

Section 3(1) of the Human Rights Act 1998 (“HRA 1998”) provides that:

“So far as it is possible to do so, primary legislation and subordinate legislation must be read and given effect in a way which is compatible with the Convention rights.”

By virtue of subsection (2)(a), section 3 “applies to primary legislation and subordinate legislation whenever enacted”.

Section 6 of HRA 1998 states that:

“(1)

It is unlawful for a public authority to act in a way which is incompatible with a Convention right.

…

(3)

In this section “public authority” includes –

(a)

a court or tribunal …”

Article 1 of the First Protocol to the European Convention on Human Rights and Fundamental Freedoms 1950 (“A1P1”), incorporated into domestic law by section 1 of HRA 1998 and set out in Schedule 1 thereto, provides as follows:

“Protection of property

Every natural or legal person is entitled to the peaceful enjoyment of his possessions. No one shall be deprived of his possessions except in the public interest and subject to the conditions provided for by law and by the general principles of international law.

The preceding provisions shall not, however, in any way impair the right of a State to enforce such laws as it deems necessary to control the use of property in accordance with the general interest or to secure the payment of taxes or other contributions or penalties.”

It is common ground that registration for gross payment under the CIS constitutes a possession for the purposes of A1P1. It is also common ground that the contractual rights of a sub-contractor to receive from the contractor and be paid the full contract price, without any deduction in respect of tax for which the sub-contractor would be liable, is in principle another such possession: see Vicky at [47], per Ferris J. Nor is it disputed by the Company that the CIS legislative scheme, viewed as a whole, is compliant with A1P1. This question was considered, by reference to the predecessor legislation then in force, in Vicky at [48] to [50]. The question was also considered, at greater length, by Lewison J in Hilton at [15] to [20].

In summary, the reasons why the legislative scheme as a whole is agreed to be compatible with A1P1 are that:

I have already referred to most of the relevant facts in the introductory section of this judgment. The facts were found by the FTT in the FTT Decision at [22] to [48]. One point which I have not yet mentioned is that the reason for the Company’s late payments of PAYE was a faulty payments system, which according to the FTT made it “inevitable” that payments to HMRC falling due either on the 19^th or 22^nd of each month would be made late. That had been the position for many years, and HMRC had never chased payment or expressed any concern that PAYE payments were late. After July 2011, the Company made significant changes to its systems to ensure that PAYE was paid on time, and there had been no further defaults down to the date of the hearing on 9 August 2012: see the FTT Decision at [25], [26] and [42].

By way of background, and without objection from Mr Chacko on behalf of the Company, we were supplied (at our request) with some figures showing the number of sub-contractors registered for gross payment under the CIS in recent years, together with the total population of registered sub-contractors. In broad terms, the figures for the tax years 2011/12 to 2015/16 show that approximately 10% of the total population of registered sub-contractors are registered for gross payment, the remaining 90% being registered for payment under deduction.

The FTT at [51] described the primary issue between the parties as being whether or not the officer of HMRC who made the decision to cancel the Company’s registration for gross payment should have taken into account the financial effect upon the Company of losing its registration. The principal submission for the Company was that by failing to take into account the financial effect on the Company the decision was flawed and unreasonable.

The FTT went on to reject the submission for HMRC that this was not a factor that could or should be taken into account. They pointed out that the decisions in Vicky and Hilton were clearly distinguishable, on the basis that neither of them concerned the cancellation of a certificate or registration, when HMRC have a discretion even where there is a breach of the compliance test for which there is no reasonable excuse. The FTT said at [58]:

“In our view it is the existence of such discretion which gives a peg on which to hang arguments that the effect on the appellant of cancellation is a relevant factor. Indeed reasonableness, including proportionality type arguments, might be expected to be at the very heart of such discretion. Having said that, when HMRC is exercising its discretion it will no doubt have well in mind the mischief that the construction industry scheme is designed to combat.”

To similar effect, the FTT said at [60]:

“Leaving aside issues of proportionality, it seems to us that the general unfettered discretion given to HMRC in considering whether to cancel an existing registration does at least involve taking into account the effect on a business of losing its registration for gross payments.”

The FTT pointed out at [61] that HMRC had in fact already exercised discretion when previously deciding not to cancel the Company’s registration in November 2009 and August 2010, although neither of the letters sent out at that time indicated on what basis the discretion was exercised. The FTT said they would be surprised if the loss of business with United Utilities was not taken into account on those occasions.

At [62], the FTT said it was easy to see why Parliament might choose to distinguish between the position of a taxpayer applying for registration, and a taxpayer who was registered but whose registration is liable to be cancelled. In the former case, HMRC must simply ascertain whether the taxpayer satisfies the stipulated conditions, and if he does so registration is mandatory. In the latter case, if the conditions are not satisfied HMRC have a discretion “because cancellation of registration can clearly have serious implications for an existing business”.

Applying these principles to the facts, the FTT then found (as I have already said) that HMRC’s decision to cancel the Company’s registration was vitiated by their failure to take such considerations into account.

The Upper Tribunal discussed the impact of A1P1 on the power to cancel registration for gross payment under section 66 at [30] to [33] of the UT Decision. After recording that Mr Chacko did not challenge the proposition that the CIS in its new form is compliant with A1P1 so far as concerns the initial registration of a person for gross payment, and agreeing with Mr Chacko that the question in the present case had not been directly answered by the authorities, the Upper Tribunal said at [31]:

“However, we do think that the reasoning which led Ferris J and Lewison J to their conclusions in relation to renewal of certificates [in Vicky and Hilton] would apply equally to the cancellation of a person’s registration for gross payment had section 66 expressly provided for the mandatory cancellation of registration at any time if HMRC formed the view that they were not satisfied that the circumstances described in paragraph (a) of section 66(1) were present; or, to put it another way, had the word “may” in section 66(1) been replaced by the word “must”, so far as concerns paragraph (a). We can see no material difference, for the purpose of A1P1, between the requirement of HMRC under the old provisions to refuse to renew a certificate and its obligation under the hypothetical express provision just discussed to cancel a registration. In neither case would the scheme give rise to a breach of taxpayer’s rights under A1P1. It follows, a fortiori, that there would be no breach if, instead of being under a duty to cancel a registration, HMRC were given a power to do so which was exercisable only subject to constraints. In particular, if the power conferred by section 66(1) expressly stated that the financial consequences for the taxpayer of cancellation of their registration were not to be taken into account by HMRC when deciding whether or not to exercise the power, the CIS would nonetheless be Convention compliant and there would be no breach of A1P1 when HMRC exercised the power without taking those consequences into account.”

I confess that I find the reasoning of the Upper Tribunal in this passage rather strange. They begin by hypothesising a mandatory version of section 66(1), and saying that it would be compliant with A1P1 for the same reasons as the mandatory provisions for initial registration or renewal of registration are compliant. They then say that a limited power conferred upon HMRC to cancel a person’s registration for gross payment would be even more clearly compliant, and this would be so if, in particular, the power stated that the financial consequences for the taxpayer of cancellation were not to be taken into account. With the greatest respect to the Upper Tribunal, I do not find this kind of reasoning from hypothetical examples of what the legislation might say to be helpful, and more importantly I disagree with the tacit assumption that there is less need for a discretionary power of cancellation to be compliant with A1P1 than there would be for an obligation to cancel to be compliant. In either case, there is an interference with a possession within the scope of the article, and the effect of the cancellation is to deprive the taxpayer of that possession. The deprivation therefore needs to be justified, whether it is compulsory or discretionary.

The Upper Tribunal then considered the question of proportionality. After referring to the judgments of Lord Sumption and Lord Reed JJSC in Bank Mellat v HM Treasury (No. 2) [2013] UKSC 39, [2014] AC 700, they said in [37] that there is nothing in Bank Mellat which casts any doubt on the correctness of Vicky and Hilton. They relied on Lord Reed’s judgment at [75] as underlining the margin of appreciation which a national legislature is to have, and gave the example of national legislation to counter tax avoidance which conferred a discretion on the decision maker. The Upper Tribunal then said:

“If the presence of the discretion is necessary to make the scheme Convention-compliant, then clearly the actual exercise of the discretion in the case of a particular taxpayer must be effected in such a way as to give effect to that taxpayer’s Convention rights. In contrast, if the scheme would be Convention-compliant in the absence of any such discretion, then there is no need [to] impose any constraint on the manner of exercise of the discretion in order to give effect to a person’s Convention rights. Nor is there any reason to take Convention rights into account when it comes to the ascertainment of the scope of the power or the identification of the matters which may, or must, be taken into account or not be taken into account in its exercise. These are matters of purely domestic law.”

Returning to their hypothetical reasoning in [31], the Upper Tribunal then said at [38]:

“In the present case, for reasons which we have given, the CIS would be Convention-compliant even if section 66 imposed a mandatory requirement on HMRC to cancel the registration of a person for gross payment whenever the requirements of paragraph (a) of section 66(1) are satisfied. It follows from our discussion in the preceding paragraph that HMRC must exercise its power under section 66(1) in accordance with the ordinary principles of public law but the exercise of that power is not further constrained by the impact of A1P1. The question whether or not HMRC must take into account (and if so, how) the financial consequences for the Company of the cancellation of its registration for gross payment is a matter of domestic law untrammelled by A1P1.”

The Upper Tribunal therefore concluded that A1P1 has no part to play in regulating the manner of exercise of the power under section 66. Nor does it have any relevance in ascertaining the scope of that power, or in identifying the matters which may, or must, be taken into account or not be taken into account: see [40].

The Upper Tribunal next considered the scope of the section 66 power. After setting out the rival submissions of HMRC (at [45] to [52]) and the Company (at [53] to [61]), they began their discussion at [62] by describing the issue as “ultimately, a short point of construction”. They then said that in their view HMRC’s position was to be preferred, but before addressing the detailed arguments they mentioned two preliminary matters.

The first matter was to emphasise the purpose of the power to cancel a subsisting registration for gross payment. They accepted HMRC’s submission that the CIS is a collection scheme, and not a scheme for the imposition of tax. They said that the strict conditions which a person has to satisfy in order to obtain such registration “reflect the balance which Parliament considered appropriate to ensure the effective collection of tax”, and that a sub-contractor was only to be entitled to receive payment gross if he could “show a good track record”, as Lewison J had said in Hilton at [19]. The Upper Tribunal continued:

“64.

Similarly, it seems to us that a person should retain his registration only if he continues to display an adequate track record. The power to cancel registration is there principally to ensure compliance with the substance of the CIS. In other words, the idea is that persons who retain their registration should comply with the requirements of the regime. There may, however, be circumstances where, notwithstanding non-compliance, a taxpayer should not suffer, immediately, the disadvantages of cancellation of their registration. For instance, the failure might be the late filing of annual accounts at Companies House (see the requirement set out in paragraph 12(5)(a) Schedule 11). Notwithstanding that this would be sufficient to preclude the grant of registration for gross payment, HMRC might form the view that the failure would not be repeated and that there was no need to cancel the registration. Or to take another example, the failure might be the late filing of a contractor’s return in circumstances where there is no reasonable excuse but where HMRC are satisfied, by reason of changes of internal procedures within the taxpayer company, that there will be no repeat of the failure.

65.

In contrast, the financial consequences of a decision to cancel registration for gross payment are irrelevant to any issue of future compliance. Indeed, where the financial consequences are relied on by a taxpayer as a reason for not exercising the power to cancel the registration when otherwise it would be exercised, for HMRC to take those consequences into account to decide not to cancel the registration would have precisely the opposite effect from that which the power is there to achieve, namely that those registered for gross payment should be those who are, or can be expected to be, compliant.”

The second matter was that all sub-contractors registered for gross payment either know, or must be taken to know, of the risk of cancellation of their registration if they fail to satisfy the requirements of the compliance test. It could not be right that a taxpayer, whose business would be destroyed by cancellation of its registration, could repeatedly fail in its compliance obligations and still rely on that inevitable destruction as a ground for challenging a decision to cancel its registration.

The Upper Tribunal said at [67] that, subject to Mr Chacko’s detailed arguments, the two preliminary matters discussed above led them to conclude that the financial consequences of cancellation of a taxpayer’s registration for gross payment were not a relevant factor to be taken into account by HMRC when deciding how to exercise the discretion conferred by section 66(1). The Upper Tribunal then reviewed Mr Chacko’s detailed arguments at [68] to [74], concluding that the arguments gave them no reason to doubt their initial conclusion.

In considering the grounds of appeal, it is in my judgment essential to begin by placing section 66 in its statutory context and asking, as a matter of construction, what Parliament intended to be the scope and purpose of the power of cancellation conferred by the section.

The starting point is that the CIS was enacted to counter not just tax avoidance, but widespread criminal evasion of tax in the construction industry. The remedy chosen by Parliament was a system which required all payments by a contractor to a sub-contractor in the industry to be made under deduction of tax, unless the sub-contractor had applied for, and obtained, registration for gross payment. The default position was therefore payment under deduction of tax, either because the sub-contractor was unregistered, or because it was registered for payment under deduction. Gross payment was the exception, and a privilege which had to be earned by satisfying a wide range of requirements, precisely because gross payment would expose HMRC to the risk of non-payment of tax by the sub-contractor which the scheme was designed to combat.

The conditions which have to be satisfied by a successful applicant for registration for gross payment are stringent. They require a corporate sub-contractor to have a substantial turnover in excess of a specified minimum level, and to demonstrate satisfaction of the compliance test over the previous twelve months, subject to the Regulation 32 exceptions and (where it applies) the “reasonable excuse” exception in Schedule 11 paragraph 12(3). There must also be a reasonable expectation of future compliance. The areas in which compliance is required cover all obligations of substantive tax law and procedure, all requests by HMRC to supply accounts or other information, the prompt payment of national insurance contributions, and key regulatory requirements of company law.

Since registration for gross payment is a privilege earned by demonstrating a good track record in all these ways, provision clearly has to be made for cancellation of the privilege, and reversion to the default position, if circumstances no longer warrant its continuation. Section 66(1) empowers, but does not oblige, HMRC to make a determination cancelling a person’s registration for gross payment with effect from the end of a prescribed period if it appears to them that any of the three specified conditions in paragraphs (a), (b), and (c) is satisfied. As one would expect, each of those conditions relates to matters within the purview of the scheme. Thus, the registration may be cancelled if it appears to HMRC that a fresh application by the person to be registered for gross payment would fail, or that he has made an incorrect return or provided incorrect information under any provision of the CIS, or that he has failed to comply (whether as a contractor or sub-contractor) with any such provisions. By virtue of subsections (3) and (4), HMRC may cancel the registration with immediate effect if they have reasonable grounds to suspect that the person (a) became registered for gross payment on the basis of false information, or (b) has fraudulently made an incorrect return or provided incorrect information under any relevant provision, or (c) has knowingly failed to comply with any such provision, whether as a contractor or a sub-contractor.

A number of significant protections for the taxpayer are built into section 66. First, and most obviously, HMRC are not obliged to make a determination under the section. Secondly, they must give the person concerned a notice stating the reasons for the cancellation. Thirdly, a right of appeal is conferred by section 67, and the jurisdiction of the FTT on an appeal includes jurisdiction to review the decision taken by HMRC. Fourthly, the taxpayer may re-apply for registration for gross payment after one year has elapsed from the cancellation. Fifthly, where (as in the present case) the ground for cancellation is that the taxpayer would fail such an application at the time when the determination to cancel is made, the protections for the taxpayer built into the compliance test are indirectly imported.

As a matter of first impression, I cannot find any indication in this tightly constructed statutory scheme that Parliament intended HMRC to have the power, and still less a duty, to take into account matters extraneous to the CIS regime, when deciding whether or not to exercise the power of cancellation in section 66(1). By “matters extraneous to the CIS regime” I mean in particular, in the present context, matters which do not relate, directly or indirectly, to the requirements for registration for gross payment, and to the objective of securing compliance with those requirements. My preliminary view, therefore, is that consideration of the financial impact on the taxpayer of cancellation would fall well outside the intended scope of the power.

The position might arguably be different if no real content could be given to the discretion which the power affords to HMRC unless wider considerations could be taken into account. This was indeed the main burden of Mr Chacko’s argument in support of the second ground of appeal. He submits that where gross payment registration is cancelled for non-compliance with the conditions relevant to the grant of registration under section 66(1)(a), as in the present case, HMRC must be satisfied that those conditions have been breached before their discretion to cancel arises. The risks of future failure, the culpability of the sub-contractor and the seriousness of the breach are all explicitly dealt with in the legislation, and have to be taken into account by HMRC when deciding whether or not they have power to cancel on this ground. It cannot be right, says Mr Chacko, that HMRC are then confined to consideration of those same factors when considering whether or not to exercise their discretion.

Mr Chacko reinforces this argument by reminding us that Regulation 32 replaced the previous exception for “minor and technical” failures: see [29] above. Under the previous law, HMRC were required to decide how serious compliance failures were (i.e. whether they were “minor and technical”, and whether they suggested a risk of future non-compliance) under section 565(4) of ICTA 1988 in order to decide whether or not they had the power to cancel gross payment status under section 561(8)(b), the predecessor of section 66(1)(a). They would then have had to carry out exactly the same exercise when deciding whether or not to exercise that power. It would be surprising, submits Mr Chacko, if Parliament had intended to create a system whereby HMRC are authorised to define a penumbra of tolerable compliance failures within which gross payment registration cannot be cancelled, but had then given HMRC power to define a further penumbra in order to decide whether or not to use their power of cancellation. He submits that there would then be nothing to weigh in the balance against the requirements of Regulation 32, given that the sub-contractor would necessarily have no reasonable excuses for his compliance failures. By contrast, if Parliament had intended to authorise HMRC to take account of the impact on the registered person of the cancellation of his status, there would then be a clear additional factor to be considered by HMRC once their discretion arose, and one which could be weighed against the significance of any compliance failures which went beyond Regulation 32.

These submissions were attractively presented, but I am unable to accept them. In my view there is no need to posit an intention that HMRC should be able to take wider considerations of this nature into account in order to explain the discretion conferred on them by section 66(1). I have already pointed out the highly prescriptive nature, both of the requirements of the compliance test, and of the exceptions to it contained in Regulation 32. It seems to me entirely appropriate, and a substantial protection for the registered person, that HMRC should then be given a discretion whether or not to exercise the power of cancellation, even in cases where the condition in section 66(1)(a) is satisfied. The Upper Tribunal gave two examples, in [64] of the UT Decision, quoted above, of cases where HMRC might properly exercise such discretion in the taxpayer’s favour, without travelling outside what I would regard as the proper scope of the power. It needs to be remembered, in this connection, that the “reasonable excuse” exception does not apply to all the requirements of the compliance test, and in the absence of any discretion even a single minor failure to pay national insurance contributions on the due date, or a minor failure to comply with one of the Companies Act requirements, would be fatal, even if there were a reasonable excuse for the non-compliance. Similarly, the rigid structure of Regulation 32 itself leaves no scope for the exercise of any discretion, even if the relevant test was failed by a narrow margin, the amount involved was relatively small, and although (when viewed in isolation) there was no reasonable excuse for the non-compliance, there was nevertheless good reason to suppose that it would not be repeated. I therefore remain unpersuaded that there is any need to broaden the scope of the discretion conferred by section 66(1) in order to provide it with any worthwhile content.

I am fortified in reaching this conclusion by three further considerations. First, if Parliament had intended to authorise HMRC to take into account the impact of cancellation on the taxpayer’s business, I would have expected this to be expressly stated in such a prescriptive and closely-textured statutory regime. I would also expect some guidance to have been given in the primary legislation about how such a difficult task was to be performed. As the Upper Tribunal said, at [72]:

“Mr Chacko’s submissions lead to difficult questions about what weight to attach to the financial consequences and how that weight is to be reflected in the ultimate decision. If Parliament had intended such consequences to be taken into account, we think that the legislation would have contained, or authorised secondary legislation which contained, some method of ascertaining the principles by which they fall to be taken into account. ”

Secondly, I also find it helpful to refer to the observations of Lewison J in Hilton at [22] to [23]. Although made in the context of a proportionality argument, and section 3 of HRA 1998, the observations are equally applicable to the question of construction which I am now considering. Lewison J said:

“22.

… If the legislation were to incorporate a general test of proportionality that would place a heavy burden on tax inspectors to conduct a prospective review or forecast of the potential effect of refusal of a certificate on individual businesses. Moreover, it is not said that it will always be disproportionate to refuse a certificate if the result would be that the taxpayer would be put out of business. So there would require to be a judgment by the inspector not only whether a refusal would have that effect, but also whether that effect is proportionate to the failures.

23.

There may be social, economic and administrative arguments for and against the imposition of such a burden or there may be other solutions to perceived injustices in the statutory scheme, but they are matters for debate and legislation not for interpretation by a court.”

My third point is related, but of a more general nature. It is the fundamental duty of HMRC to collect taxes imposed, in the public interest, by Parliament. Under their care and management powers, conferred at the relevant time by section 1 of the Taxes Management Act 1970, HMRC have a reasonable degree of latitude as to how in practice they collect tax which has fallen due. They may, for example, enter into “time to pay” arrangements to help businesses that are having difficulties making payments of tax on the due dates, if there is a reasonable prospect that the difficulties will be overcome within a reasonable period. But arrangements of that nature arise in the general context of care and management of the taxes for which HMRC are responsible, and are in my opinion very far removed from the specific power in issue in the present case, which takes its place and has to be construed as part of a closely articulated code designed to combat tax evasion.

For these reasons, and subject to the proportionality issues which I will now address, I consider that on its true construction section 66(1) neither authorises nor requires HMRC to take into account the likely impact on the taxpayer’s business and financial position when deciding whether or not to exercise their power to cancel registration for gross payment.

Although Mr Chacko did not place this at the forefront of his submissions, it is appropriate to begin by considering whether HMRC’s power to cancel must be exercised proportionately as a matter of common law. As Lord Toulson JSC has recently stressed, in R (Ingenious Media Plc) v Revenue and Customs Commissioners [2016] UKSC 54, [2016] 1 WLR 4164, at [28]:

“It is important to emphasise that public bodies are not immune from the ordinary application of the common law … The common law is multi-faceted and remains the bedrock of the English legal system.”

Accordingly, if there was a common law requirement of proportionality in cases of the present type, it would have formed part of the background to the enactment of the CIS legislation, and it would need to be taken into account both in construing the legislation and in its application.

Mr Chacko argued that, because removal of registration would have a significant impact on the Company’s business, the case was analogous to the removal of a licence, where a requirement of proportionality is imposed at common law. Mr Chacko referred us to the decision of the Court of Appeal in R v Barnsley Council, Ex p. Hook [1976] 1 WLR 1052, which concerned the termination of a market trader’s licence in the Barnsley market following an incident where he had been seen urinating in a side street after the market had closed in the evening and the public lavatories were locked. The principal ground of decision was that the rules of natural justice had not been followed in the procedure adopted by the council to revoke Mr Hook’s licence, but two members of the court (Lord Denning MR and Sir John Pennycuick) also relied on the point that the punishment was wholly disproportionate to the offence: see 1057H - 1058B, and 1063B. As Lord Denning put it, at 1057H:

“Now there are old cases which show that the court can interfere by certiorari if a punishment is altogether excessive and out of proportion to the occasion … It is quite wrong that the Barnsley Corporation should inflict upon [Mr Hook] the grave penalty of depriving him of his livelihood. That is a far more serious penalty than anything the magistrates could inflict. He is a man of good character and ought not to be penalised thus. On that ground alone, apart from the others, the decision of the Barnsley Corporation cannot stand.”

The third member of the court, Scarman LJ, while not founding his judgment on any requirement of proportionality (see 1062E), pointed out at 1058G-H that revocation of an existing licence is usually a more serious matter than refusal to grant a licence in the first place.

More recently, in Pham v Secretary of State for the Home Department [2015] UKSC 19, [2015] 1 WLR 1591, Lord Reed JSC, in the course of some helpful observations on the relationship between reasonableness and proportionality as principles of domestic administrative law, referred to the Hook case at [114] as a case “in which a finding of unreasonableness was based on a lack of proportionality between ends and means.”

I am prepared to accept that there are many contexts in which the common law will require proportionality between ends and means to be observed by a public authority in the exercise of its functions, although whether such a requirement exists as an independent ground of review of administrative action, or only as an aspect of review for unreasonableness, remains a controversial question upon which the Supreme Court has yet to pronounce definitively. Even if that assumption be made, however, I do not think that it assists the Company in the present case. The CIS legislation as a whole is clearly proportionate in the balance which it strikes between ends and means, and in the procedural safeguards for the taxpayer which are built into it. In relation to the power of revocation in section 66(1) itself, the existence of a discretion is one of those safeguards, and it seems to me that any common law requirement of proportionality is comfortably satisfied if the matters which HMRC are entitled to take into account are broadly confined to matters relevant under the statutory scheme to the grant of registration for gross payment, but with a wider margin of discretion than the often highly prescriptive terms of the legislation would otherwise permit.

The impact of cancellation of registration on the sub-contractor’s business is in my judgment an extraneous factor, and the mere fact that the financial consequences for the sub-contractor’s business will be severe cannot, without more, make that factor one which it is relevant for HMRC to consider. Gross registration is a privilege which has to be earned by satisfying various conditions, and which is liable to be lost if those conditions are no longer satisfied. The compliance conditions are all matters within the control of the taxpayer, and the consequences of non-compliance are clearly spelt out in the legislation. The taxpayer’s simple remedy is to ensure that it continues to comply with the relevant conditions, and in most cases it will have nobody to blame but itself if its registration for gross payment is cancelled. That is certainly so in the present case, where the Company failed to put in place a system which would ensure timely payment of its PAYE obligations, despite being given two opportunities to do so. Against such a background, it would to my mind be strange if the common law were to subject exercise of the section 66(1) power to a wider requirement of proportionality, requiring a detailed examination of the taxpayer’s present and probable future financial position in the event of cancellation.

For these reasons, I am satisfied that the common law principle of proportionality does not assist the Company.

Mr Chacko’s primary submission is that HMRC were required to exercise their discretion proportionately as a matter of human rights law. He submits that an interference with the Company’s possessions under A1P1 must be proportionate as well as being imposed according to law and justified in the general interest. Mr Chacko referred us to AXA General Insurance Limited v HM Advocate [2011] UKSC 46, [2012] 1 AC 868, where Lord Reed JSC said at [126]:

“In order for an interference with possessions to be compatible with A1P1, it must not only be lawful and in the general interest, but there must also be a reasonable relation of proportionality between the means employed and the aim sought to be realised. This involves an assessment of whether a fair balance has been struck between the demands of the general interest of the community and the requirements of the protection of the individual’s fundamental rights: the individual should not be required to bear an individual and excessive burden: James v United Kingdom, 8 EHRR 123, para 50. In making that assessment at the international level, the Strasbourg Court has allowed national authorities a wide margin of appreciation: see e.g. J A Pye (Oxford) Ltd v United Kingdom (2007) 46 EHRR 1083, para 75.”

Mr Chacko went on to criticise the Upper Tribunal’s conclusion, at [38] of the UT Decision, that the question whether HMRC must take into account the financial consequences for the Company of cancellation “is a matter of domestic law untrammelled by A1P1”. He says that the reasoning which led the Upper Tribunal to this conclusion is flawed, for three main reasons. First, the comparison which they made between a hypothetical system of mandatory cancellation and the actual discretionary system is misleading. As I have already indicated, I consider that there is force in this criticism: see [47] above. Secondly, he argues that the Upper Tribunal were wrong, at [31] of the UT Decision, to see no material difference, for the purpose of A1P1, between the old provisions relating to refusal to renew a certificate, on the one hand, and the obligation of HMRC to cancel a registration under a hypothetical mandatory system, on the other hand. Again, I see force in this point, because it relies on the same flawed comparison between a hypothetical mandatory requirement to cancel and the existing discretionary power to do so.

Thirdly, Mr Chacko says that HMRC’s exercise of discretion is an act of a public authority within section 6 of HRA 1998, and must therefore be carried out in compliance with A1P1. This means, he argues, that HMRC must weigh the importance of taking the step of cancellation against the detriment caused to the person or persons affected. In support of this proposition, he relies on a case involving the right to respect for private life guaranteed by article 8 of the Convention, R (Razgar) v Secretary of State for the Home Department [2004] UKHL 27, [2004] 2 AC 368, where the claimant, an Iraqi of Kurdish origin, who had entered the United Kingdom from Germany and claimed asylum, asserted that if he was removed to Germany as a safe third country he would attempt to commit suicide, and would not receive appropriate treatment in that country. He supported this assertion with expert medical evidence. Despite its very different subject matter, the reason why Mr Chacko relies on this authority is for the clear distinction drawn by Lord Bingham of Cornhill, in particular at [19] to [20], between a lawful immigration policy which, viewed as a whole, is compliant with article 8, and the further question whether the interference with the claimant’s article 8 rights was proportionate to the legitimate public end sought to be achieved by the policy. In relation to this question, Lord Bingham said at [20]:

“The answering of question (5), where that question is reached, must always involve the striking of a fair balance between the rights of the individual and the interest of the community which is inherent in the whole of the Convention. The severity and consequences of the interference will call for careful assessment at this stage … Decisions taken pursuant to the lawful operation of immigration control will be proportionate in all save a small minority of exceptional cases, identifiable only on a case by case basis.”

Mr Chacko also referred us in this connection to the Bank Mellat case, where he submits that both Lord Sumption (at [27]) and Lord Reed (at [68] to [76]) approached the issue of proportionality by considering the detriment imposed on Bank Mellat in particular, in the context of a measure adopted by the Treasury, in the form of subordinate legislation, prohibiting all persons operating in the financial sector in the United Kingdom from entering into any transaction or business relationship with the bank under section 62 of, and Schedule 7 to, the Counter-Terrorism Act 2008. It is worthy of note that paragraph 9(6) of Schedule 7 to the 2008 Act expressly stipulated that the requirements imposed by a direction made under Schedule 7 must be proportionate having regard to various specified matters. This in turn gave rise to the question whether the requirement imposed by paragraph 9(6) is the same as the principle of proportionality as understood in the context of Convention rights. Having posed this question, Lord Reed said at [67]:

“The latter principle is of course relevant to the question whether the decision of the Treasury was incompatible with A1P1 and therefore unlawful by virtue of section 6(1) of the Human Rights Act 1998.”

It was in this context that Lord Reed then embarked on the valuable discussion of the concept of proportionality at [68] to [76], to which the Upper Tribunal referred, and from which they quoted a substantial extract, at [36] of the UT Decision. As I have explained, the Upper Tribunal then went on to conclude that, because the CIS legislation as a whole was Convention-compliant, A1P1 has no part to play in the manner of exercise of the section 66 power: see [48] to [50] above. In reaching this conclusion, the Upper Tribunal drew a distinction between a legislative scheme which, viewed as a whole, would be Convention-compliant in the absence of a discretion such as that contained in section 66, and a scheme where the presence of the discretion is necessary to make the scheme Convention-compliant. In a case of the former type, the Upper Tribunal reasoned, there is no need to impose any requirement of proportionality when the discretion is exercised.

In my respectful opinion, the reasoning of the Upper Tribunal on this critical issue is again vitiated by their focus on other forms which the CIS legalisation might have taken, rather than the form in which Parliament has chosen to enact it. In my judgment, Mr Chacko is right to say that A1P1 has to be considered at the stage of exercise of the discretion conferred by section 66 (1), if only for the simple reason that cancellation of a certificate indubitably involves an interference with the two possessions identified by Ferris J in Vicky. It by no means follows, however, that the proportionality review at this stage always needs to go beyond the proportionality of the CIS regime as a whole. On the contrary, in all save the most exceptional cases it will in my judgment be a complete answer that the discretion as I have construed it forms an integral part of a Convention-compliant statutory regime. And in the circumstances of the present case, I see no more scope for a successful argument based on A1P1, as a ground of challenge to the cancellation of the Company’s registration, than I do for a challenge based on the common law principle of proportionality. In particular, the adverse effect on the Company’s business is in my view an entirely predictable consequence of the Company’s non-compliance, for which it has only itself to blame.

Accordingly, although I would not rule out the possibility of exceptional circumstances justifying a wider proportionality review at the stage of exercise of the power of cancellation, I do not consider that the impact on the Company’s business, as found by the FTT, comes near to satisfying such a test. Given the practical and cash-flow advantages of registration for gross payment, it is always probable that cancellation of the registration will seriously affect the taxpayer’s business. Far from being exceptional, such consequences are likely to be the norm, and taxpayers must be taken to be well aware of the risks to their business which cancellation will bring. In individual cases, of which this may perhaps be one, the result may seem harsh; but a degree of harshness in a regime which is designed to counter tax evasion, and where continued compliance is within the power of the sub-contractor, cannot in my view be characterised as disproportionate. Both deterrence, and ease of compliance, are important factors which help to make the CIS scheme as a whole clearly compliant with A1P1. Furthermore, the observations of Lewison J in Hilton at [22] to [23], quoted at [65] above, will usually be as apposite to the exercise of HMRC’s discretion under section 66 as they are to decisions on initial or renewed registration.

In the result, although there are some aspects of the Upper Tribunal’s reasoning with which I respectfully disagree, I consider that they were right to allow HMRC’s appeal from the decision of the FTT, and that the Company’s appeal to this court must be dismissed.

I agree.

I also agree.