Overseas but under the law, revisited

In an article “Overseas but under the law”, published on Journal Online in November 2018, I provided a brief overview of the Court of Appeal’s judgment in British Council v Jeffery; Green v SIG Trading Ltd [2018] EWCA Civ 2253 (“Green”). Those conjoined appeals concerned the territorial scope of various provisions of the Employment Rights Act 1996 (“ERA 1996”). It was an important judgment for a number of reasons, as I explained in last year’s article.

One of the reasons for the judgment’s importance was that the Court of Appeal considered, at length, the test applicable in territorial jurisdiction cases – that of sufficiently strong connection – and whether it was a question of fact or a question of law. The outcome of that issue was significant, for it would directly impact the standard of appellate review and thus, the circumstances in which an appellate tribunal or court would be prepared to interfere with the judgment of the employment tribunal. The judgment in Green was considered and applied by the Employment Appeal Tribunal in Ravisy v Simmons & Simmons LLP UKEAT/0085/18/OO, which concerned claims under the Equality Act 2010 (“EA 2010”). That case will be analysed here along with an analysis of the “law or fact” question from the judgment in Green.

Ravisy: the facts and ET judgment

Background

Ravisy was an equity partner in the Paris office of international law firm Simmons & Simmons from mid-2013 until she was required to retire with effect from the end of 2016. She presented two claims: one for direct discrimination, equal pay and victimisation against the firm, and another for direct discrimination only against a Mr Taylor, the National Practice Group Head of the department in which she worked

The respondents denied that the ET had jurisdiction to various degrees: the firm accepted there was international jurisdiction, in the sense that the ET could competently hear the claim against it under the Brussels I Recast Regulation, but denied that the claimant fell within the territorial scope of EA 2010 (which is generally accepted as having the same territorial scope as ERA 1996), such that she could not rely on its provisions, and therefore the tribunal did not have territorial jurisdiction; Taylor denied the ET had jurisdiction in either sense for the claim against him.

In Foreign & Commonwealth Office v Bamieh [2019] EWCA Civ 803, Lord Justice Gross succinctly explained the difference between these separate requirements as follows: “The Brussels [Recast] Regulation is concerned with which court should hear a claim; it has nothing to say on the content of the substantive law applicable to a claim or the extraterritorial application of the [relevant Act]” (at [51]). I am only concerned with the latter in this article.

The territorial scope of the ERA 1996 and EA 2010 is much litigated and well known. The principal authorities, from which the following is derived, are: Lawson v Serco Ltd [2006] UKHL 3; Duncombe (No 2) v Secretary of State for Children, Schools and Families [2011] UKSC 36; and Ravat v Halliburton Manufacturing & Services Ltd [2012] UKSC 1 (the cumulative result of which is analysed at 2017 Juridical Review 279). I will only briefly address the commonly applied tests to the different types of claimants here, by reference to the ET’s summary in Ravisy (at [79]-[80]):

• type (a) cases – where, at the relevant time or during the relevant period, the claimant worked in Great Britain, there will be territorial jurisdiction;
• type (b) cases – where the claimant worked (wholly, and also lived) outside Great Britain, where there must be something exceptional such that the employment has much stronger connections to Great Britain and British employment law than any other system of law; and
• type (c) cases – where the claimant lived and worked for at least part of the time in Great Britain, where the claimant must show that there is a sufficiently strong connection with Britain and British employment law.

The ET’s conclusion

The ET found that the claimant represented a type (b) case. Although she sometimes visited the London office, these visits were not of a frequent or significant nature such as to change the type of case (and thus reduce the burden on the claimant) to that of a type (c) case. Ravisy’s counsel had sought to draw parallels to the factual matrix in Bates van Winkelhof v Clyde & Co LLP [2012] EWCA Civ 1207, in which the claimant (who was also a partner in a law firm) successfully established territorial jurisdiction, albeit as a type (c) case.

The analysis of the ET and the outcome is not particularly important for the purpose of this article, which is to outline the standard of appellate review in territorial scope cases. Nonetheless, the ET’s conclusion was that it did not have territorial jurisdiction in either claim. The ET disagreed with the suggestion that the present case had such similarities with Bates van Winkelhof so as to establish territorial jurisdiction, on two bases: (1) that it was a different type of case; and (2) there were some significant factors in Bates that had no counterpart in the present case. Given the vast number of factual circumstances taken into account by ETs in the assessment of the sufficiency of connection, this once again shows the dangers in relying on analogies from other cases: the greater the number of factual circumstances taken into account, the greater the scope for distinguishing the respective cases.

Ravisy appealed to the EAT. Before considering the EAT judgment, it is necessary to return to the judgment in Green.

British Council v Jeffery; Green v SIG Trading Ltd [2018] EWCA Civ 2253

Whilst the ET judgment in Ravisy had predated the Court of Appeal’s judgment in Green, the EAT judgment did not. The EAT was thus addressed by both counsel on the standard of appellate review, that matter having split the Court of Appeal in Green.

I noted briefly in my previous article that the Court of Appeal did not envisage that the difference in theoretical reasoning between the majority and minority would have much practical impact. However, in light of the EAT’s discussion on the matter, it is worth returning to Green to explore the standard of appellate review in more detail.

The issue of the appropriate standard of appellate review arose from the House of Lords and UK Supreme Court judgments in Lawson and Ravat. The conclusion of the then President of the EAT, Langstaff P, in Olsen v Gearbulk Services Ltd [2015] IRLR 818 also fell to be considered, since he expressed his conclusion on the standard of appellate review from the judgments in Lawson and Ravat.

In Lawson, Lord Hoffmann commented (at [34]) as follows: “Whether one characterises this [the territorial scope of s 94(1) ERA] as a question of fact depends, as I pointed out in Moyna v Secretary of State for Work and Pensions [2003] 1 WLR 1929, upon whether as a matter of policy one thinks that it is a decision which an appellate body with jurisdiction limited to errors of law should be able to review. I would be reluctant, at least at this stage in the development of a post-s 196 jurisprudence, altogether to exclude a right of appeal. In my opinion, therefore, the question of whether, on given facts, a case falls within the territorial scope of s 94(1) should be treated as a question of law. On the other hand, it is a question of degree on which the decision of the primary fact-finder is entitled to considerable respect” (my emphasis). 

In Ravat, Lord Hope commented (at [29]) that: “The question of law is whether s 94(1) applies to this particular employment. The question of fact is whether the connection between the circumstances of the employment and Great Britain and with British employment law was sufficiently strong to enable it to be said that it would be appropriate for the employee to have a claim for unfair dismissal in Great Britain.”

In Olsen, Langstaff P (at [36]) purported to hold that the question was one of fact and, consequently, that an appeal could only be made on the basis that the ET’s judgment was perverse: “At the outset of the hearing I asked whether the time may now have come to treat the question of ‘sufficiently close connection’ as indeed being a question of fact, such that a decision by a Tribunal, properly directing itself as to the applicable law, with regard to appropriate cases, would disclose no error of law unless it were shown to be perverse. [Counsel for the respondents] submitted that this was so. In the light of Lord Hope’s words, I consider his submission to be well founded.”

Underhill LJ, giving the lead judgment in Green but comprising the minority for the standard of appellate review issue, viewed (at [41]) the appropriate starting point as Lord Hope’s comments in Ravat. Underhill LJ’s conclusion from this was that, because the question of fact will in effect determine the answer to the question of law (as formulated by Lord Hope), the matter is effectively one of fact, which can only be interfered with (on the basis that it then becomes a question of law) if it is outwith the bounds of reasonable judgment: “In the typical case, however, the answer to the former question will depend entirely on the answer to the latter, with the result that in practice the dispositive issue is one of fact, except in a case where the decision made, to use Lord Hoffmann’s phrase in Moyna, 'falls outside the bounds of reasonable judgment', in which case the issue becomes one of law and an appeal will lie.”

For the majority on this point in Green, Longmore and Peter Jackson LJJ preferred the view that the question of whether the connection is sufficiently strong is an evaluative judgment made on the underlying facts and that evaluation itself is a question of law. The practical effect of this, because of the deference afforded to a first instance tribunal, is that an appellate court will not interfere with the judgment unless:

• the tribunal failed to take into account factors that it should have (as in the Jeffery appeal, where the Court of Appeal upheld the EAT’s conclusion that the appeal from the ET should be allowed as five relevant factors were not given weight);
• the tribunal took into account factors that it should not have (e.g. if a tribunal were to take into account the substantive law remedies of other jurisdictions involved in the comparative exercise for a type (b) case, contrary to the Court of Appeal’s judgment in Dhunna v Creditsights Ltd [2014] EWCA Civ 1238);
• the tribunal erred in law (e.g. by applying the incorrect standard of test or – as a result of the evaluative judgment being one of law – by erring in the conclusion reached from the evaluative judgment);
• the tribunal was otherwise “wrong” (at [136]).

The outcome in Green means that Langstaff P’s decision in Olsen has been disapproved and should not be followed.

Ravisy: the EAT judgment

In Ravisy, although the EAT was bound by the majority judgment in Green, Mr Justice Kerr set out his preferred approach to the standard of appellate review as obiter.

First, it was difficult to see how the question of sufficient connection could be one of fact, thus agreeing with the Green majority and disagreeing with Underhill LJ. Kerr J summarised the position convincingly as being “a conclusion of law drawn from the primary facts” (at [69]).

Secondly, disagreeing with both the minority and majority in Green, Kerr J was dissatisfied with the standard of “wrong” being introduced as a residual test. The standard of “wrong” has been described as “something more... than personal unease and something less than perversity” (Assicurazioni Generali SpA v Arab Insurance Group [2003] 1 WLR 577 at [197]). It appears to introduce an element of discretion rather than being a question of whether the tribunal got the evaluative judgment, which is a question of law, right or wrong: “If the issue is treated as one of law, as I believe it should be, the tribunal below either gets the law right or gets the law wrong. Deference to its expertise, experience and closeness to the facts cannot extend to upholding a conclusion of law, reached by applying the facts, with which the appellate tribunal disagrees” (at [72]).

Kerr J’s analysis is interesting and convincing. The effect of it (were the EAT not bound by Green) would be to read out the fourth bullet point of the list above taken from Longmore LJ’s judgment in Green. The effect of that reading out (were it possible) would be to create a substantive difference between the majority and minority reasoning in Green on the standard of appellate review. At present, it appears that there is no distinction – as pointed out by Underhill and Peter Jackson LJJ (at [45] and [140] respectively – because the standard of “wrong” introduced by the majority in Green would appear to be of a similar nature to that of Underhill LJ’s minority standard, being outwith the bounds of reasonable judgment.

As to the substantive merits of the appeal, Kerr J disposed of them swiftly on the basis that he agreed with the ET for the reasons given in the ET judgment. The factors that Ravisy sought to place significant weight on – the ownership and control structure of the firm being in London – could not be afforded such weight and, as the ET correctly noted when applying Lawson, would not in any event be decisive.

Conclusion

Kerr J’s analysis provides a more accessible summary of the standard of appellate review from Green – it is difficult not to agree with his comments on the “wrong” standard referred to in Green. However, the current position is that an appellate court can interfere with the first instance tribunal’s decision on the four grounds set out above, and that the question of the territorial applicability of ERA 1996 or EA 2010 to a particular employment is an evaluative question of law based on underlying facts found by the first instance tribunal (the ET). On the basis that it will be arguable as an error of law that the incorrect evaluative judgment was drawn from the facts by the ET, it is likely that unsuccessful parties at first instance will continue appealing ET judgments in this area.