THE Riot (Damages) Act 1886 rarely looms large in the legal imagination. It is brief and apparently straightforward. Section 2(1) of the Act obliges police authorities to compensate owners whose property has been damaged by “persons riotously and tumultuously assembled together” in their area. There is, notably, no requirement to show that the police were at fault. Two appellate cases on the Act arose from disturbances at the Yarl's Wood detention centre in 2002 (see D.J. Feldman [2010] C.L.J. 433). The London (and nationwide) riots of 2011 have occasioned further consideration.
In Mitsui Sumitomo Insurance Co (Europe) Ltd. v Mayor's Office for Policing and Crime [2014] EWCA Civ 682; [2015] Q.B. 180, compensation under the Riot (Damages) Act 1886 was held (unsurprisingly) to extend to financial losses consequential on damage to property. The second, more controversial issue was the proper construction of “riotously and tumultuously assembled”. “Riotously” bears the same meaning as the offence of riot (Public Order Act 1986, ss. 1, 10(1)). What though does “tumultuously” mean, and why did Parliament include this additional condition for liability?
In the leading case, JW Dwyer Ltd. v Metropolitan Police District Receiver [1967] 2 Q.B. 970, Lyell J. held that “tumultuously” connotes an “excited, emotionally aroused assembly” of considerable size. The “specific reason” for this additional criterion was, His Lordship stated, to confine liability to cases where the police manifestly should have been aware of the threat to the claimant's property and therefore could have acted to protect it. The right to compensation exists because of the failure “to protect the public who were menaced by a threat which was, or ought to have been, obvious to the forces of law and order”. Lyell J.'s judgment was deemed “admirable” and applied by the Court of Appeal in DH Edmonds Ltd. v East Sussex Police Authority (The Times, 15 July 1988). In one of the Yarl's Wood cases, Bedfordshire Police Authority v Constable [2009] EWCA Civ 64; [2009] 2 All E.R. (Comm.) 200, Longmore L.J. referred to the “notional responsibility” of the police for preserving law and order as the 1886 Act's rationale. Its genesis had been traced back as far the Statute of Winchester 1285 but, “happily for non-historians”, that rationale was confirmed by modern authorities – notably Dwyer. Longmore L.J. stated that the purpose of the 1886 Act had been to transfer liability from the public at large (as under its predecessor, the Riot Act 1714) to the police force, precisely because the police are notionally in breach of their duty to keep the peace when rioting breaks out.
At first instance, in Mitsui Sumitomo Insurance [2013] EWHC 2734 (Comm); [2014] 1 All E.R. 422, Flaux J. summarised these authorities. Liability under the Act
only arises where it can be said that the notional responsibility of the police to protect the public against rioters is engaged, in other words where the “riotous and tumultuous assembly” manifests itself in such a way that the police ought notionally to have been aware of it and to have prevented it, even though the obligation is a strict one and not dependent on establishing fault. In other words, it is irrelevant whether the police could or should have in fact stopped the assembly, what matters is whether it was sufficiently manifest, i.e. “riotous and tumultuous” that, albeit notionally, the police ought to have prevented it.
But the Court of Appeal disapproved this reasoning, and Lyell J.'s in Dwyer. The test of a “tumultuous” assembly was not “whether the police should, even notionally, have prevented the damage”. It was wrong to undertake “a counterfactual inquiry into whether the (or a hypothetical) police force could have prevented the damage”. That would come close to asking whether the police had been at fault, yet fault is clearly irrelevant to their liability.
It is suggested that the Court of Appeal may have mischaracterised the reasoning of Lyell and Flaux JJ., which is compatible with strict liability. The Riot (Damages) Act 1886 is best rationalised as an incentive for the police to prevent and quell riots. Prudently, the Act does not require fault. As the Court of Appeal rightly stated, deciding whether the police had been negligent in the way they dealt with a riot would be extremely difficult for a court. (Common law negligence liability is highly restricted: e.g. Michael v Chief Constable of South Wales [2015] UKSC 2; [2015] 2 W.L.R. 343.) But the hornets' nest of fault need not be poked to reinforce the police duty to keep the peace. Under the 1886 Act, they must pay for all property damaged in rioting (irrespective of fault). Thereby, the police enjoy a direct financial incentive to take all cost-justified measures to protect property. Any destruction prevented will redound automatically to their own financial benefit (through liabilities averted). (See further J. Morgan (2014) 77 M.L.R. 434.) The incentive will be sharpened the greater the proportion of riot damage for which the police must pay. Thus the confirmation in Mitsui Sumitomo Insurance that consequential loss is covered by the Act is welcome.
But there is a threshold condition for such no-fault incentives to work. The defendant must be able to prevent the harm from occurring. There is no point in trying to deter incapable persons. This explains why liability was transferred in 1886 from the inhabitants of a locality (Riot Act 1714) to police authorities: if it was ever realistic (three centuries ago) to expect the public at large to put down rioting, that function has long since passed to the organised police force. Officers have training, equipment, and legal powers that citizens lack. But even so, the police cannot prevent the unforeseeable, nor disperse the imperceptible. The very reason for a statutory duty to compensate riot damage specifically, and not crime generally, is that rioting is by its nature public, manifest, and obvious. The parliamentary draftsman's addition of “tumultuously”, according to Lyell J.'s reasoning in Dwyer, served to refine that point. Only riots that actually are obvious trigger the liability. The point is perhaps better made negatively: rioting that is “secret or furtive” (per Lyell J.) should not occasion liability because the police are unlikely to be aware of it. It is not that fault is absent in such a case. Rather that, when there was no realistic opportunity of preventing the damage, there can be no point in imposing incentives on the police to do so.
The Court of Appeal noted the “incentive” rationale. For example, citing Lord Mansfield C.J.'s observation that by placing liability on a locality, the Riot Act 1714 served as an “inducement to the inhabitants to be active in suppressing such riots”: Ratcliffe v Eden (1776) 2 Cowp. 485, 488. But the court preferred to emphasise another dictum of Lord Mansfield's, namely that the inhabitants stood as “sureties” for the rioters: a rationale said to prioritise the compensation of victims over the responsibility of defendants. However, as Rix L.J. observed in Yarl's Wood Immigration Ltd. v Bedfordshire Police Authority [2009] EWCA Civ 1110; [2010] Q.B. 698, at [54], a mixed justification is possible. The 1886 Act may promote both compensation of the victims of rioting and incentives for the police to prevent it. The goals are not mutually exclusive.
Mitsui Sumitomo Insurance has questionably deprecated the deterrence rationale for the 1886 Act. For, without it, there is little reason for the additional condition “tumultuously”. The explanation offered in Dwyer for the “tumultuous” requirement, and generally accepted since, has gone. But it can scarcely have been included for reasons of euphony (“tumultuously” is, admittedly, a pleasing word). Parliamentary Counsel has austerely observed that in order to avoid ambiguity, “legislation speaks in a monotone and its language is compressed”. Statutes have to be unpoetic. Nevertheless, the Court of Appeal seemed to view “tumultuously” as merely a fine metaphor for an ugly situation: “In short, the focus of the inquiry is whether property has been damaged or destroyed as a result of mob violence”, a question of fact and degree.
What, then, is the Act's true rationale? It is anomalous viewed purely as compensation (the general Criminal Injuries Compensation Scheme (CICS) covers personal injuries; why would the 1886 Act extend to property damage, and a small subset at that?). It is odder still, from the compensation perspective, for the financial burden to fall on police authorities rather than the national taxpayer's broader shoulders (again, compare the CICS). The 1886 Act should rather be conceived as strict accountability for failures in riot policing. A recent report has broadly endorsed that principle (although HM Treasury often indemnifies police authorities in practice, which seriously undermines any deterrent effect): Independent Review of the Riot (Damages) Act (Home Office, 2013). Reform seems likely to follow, in the guise of revision rather than outright repeal: Draft Riot Compensation Bill (Cm. 9036, 2015). Therefore, the underlying rationale for riot compensation remains an important, if neglected, question. The Supreme Court has given permission to appeal in Mitsui Sumitomo Insurance (although the specific linguistic controversy would be resolved by the recent Draft Bill, which uses the definition in the Public Order Act in place of the existing “outdated” and unclear provisions: Reform of the Riot (Damages) Act 1886: Summary of Consultation Responses and Conclusions (Home Office, 2015), para. 2.12).
