Legislating in Times of Crisis: The Coronavirus Job Retention Scheme

The focus of the common law on the individually negotiated contract of personal service makes it peculiarly inapt for reaching collective compromises with the workforce on matters such as temporary reductions in hours, pay or staffing levels.

Any attempt to seek agreement with individual employees is liable to unravel. Absent a contractual right to send employees home, an employer will commit a repudiatory breach if it ‘furloughs’ them. If some employees object to changes to hours or pay or a union objects on their behalf, the variation is likely to be legally ineffective, ‘a thing writ in water’: see Asquith LJ in Howard v Pickford Truck [1951] 1 KB 417. The judgment of the House of Lords in Rigby v Ferodo [1988] ICR 29 remains a good illustration of this rule applying in the employment context.

Alternatively, those employees who don’t like the changes could resign and claim constructive (and unfair dismissal) based on what is very likely to amount to a repudiatory breach by the employer. There may also be problems with the change being invalid because not supported by consideration, though here the ‘practical benefit’ to workers of continued employment may come to the rescue (see Lee v GEC Plessey [1993] IRLR 383). While in ordinary times these legal obstructions to variation can support a worker-protective common law, they can create difficulties where there is a need to implement rapid contractual changes in situations of emergency.

Without statutory intervention, collective agreements in UK labour law are also ill-suited to supporting modifications that are binding at the individual level. The coverage of collective bargaining is now confined to around a quarter of employees. In addition, the so-called ‘normative’ effect of collective agreements on individual contracts of employment is based upon the intention of the individual contracting parties. This has necessitated special legal interventions during times of crisis such as wartime, to ensure that collective agreements have compulsory normative effect in individual contracts of employment.

Under wartime legislation, the collective agreement was transformed into an instrument of public economic and social policies (O Kahn-Freund, ‘Collective Agreements under Wartime Legislation’ (1943) MLR). It was no longer an instrument for securing private economic advantages. This required exceptional legal intervention, however, through the system of compulsory arbitration introduced under Order 1305. This implemented a strong version of compulsory normative effect in individual contracts of employment.

Changes to work, hours or pay may also activate two unfamiliar and archaic statutory duties. First, in circumstances where remuneration is paid as a function of the amount of work, a reduction in hours resulting in no pay or less half a ‘week’s pay’ will trigger the provisions on lay-off or short-time working now found in Chapter III of Part XI of the Employment Rights Act 1996 (ERA). These complicated provisions date back to the Redundancy Payments Act 1965 and require the payment of statutory redundancy payments to employees who are not dismissed, provided the employee gives the requisite notice. They have been described by HHJ Hand in Dutton v Jones [2013] ICR 559 as the road ‘less travelled’ and by Lord McDonald MC in MacRae v Dawson [1984] IRLR 5 as ‘the despair of all those who have been concerned with the interpretation of industrial legislation.’

Second, under Part III of ERA, containing provisions introduced by the Employment Protection Act 1975 in the context of the three-day week, if an employee is not provided with any work for some days on which s/he would ordinarily be required to work, the employer must pay ‘guarantee payments’ in respect of the ‘workless days’. These must meet the modest levels in s.31 of ERA, currently £29 a day for a maximum of five days in any three-month period – that is, £145. These guaranteed payments are payable even if the employer and union agree to a temporary cessation of work on some days: see Abercrombie v Aga Rangemaster [2013] IRLR 953.

Faced with the legal difficulties of securing contractual variations binding on the workforce, an employer is liable to resort to the common law’s nuclear weapon – dismissal by notice and re-engagement on new terms that are less favourable to the employee. Here, in contrast to its strong protection of existing contractual rights, the common law safeguards a wider perimeter for the employer’s dismissal powers under the contract. It is little exaggeration to say it imposes almost no restrictions on the decision to dismiss on notice – see Johnson v Unisys [2001] UKHL 13 – and none on the reasonableness of the new terms.  Moreover, because it is a question of fact and degree whether fundamental changes to terms and conditions amount to a dismissal (Alcan v Yates [1996] IRLR 327), the employer might as well put the manner beyond doubt by expressly dismissing everyone with the requisite notice. This outcome would be at complete variance with the current policy to avoid or limit collective redundancies.

Dismissal brings in its wake a host of statutory duties, almost none of which are tailored to emergencies of the present sort, taking us into some uncharted legal waters.

The duties include, first, the provisions on unfair dismissal in s.98 of the Employment Rights Act. Normally entailing consultation and the like before changes to terms and conditions are implemented, it is a matter of considerable uncertainty what reasonableness requires in the present circumstances.

Second, so long as the employer is proposing to dismiss 20 or more employees within a 90-day period in order to effect changes, the duties of collective consultation in s.188 Trade Union and Labour Relations (Consolidation) Act 1992 (TULRCA) are triggered, backed by penal sanctions. The minimum periods of consultation appear long in the current emergency. While the ‘special circumstances’ provision in s.188(7) may well be engaged, given the ‘exceptional’ nature of the pandemic (Clarks of Hove v Bakers Union [1978] ICR 1076), the employer must nonetheless still comply with the duties so far as is reasonably practicable.

Third, where the employer is proposing to reduce the numbers of employees or fundamentally alter their duties so that the nature of their work will change, the dismissals will entail individual redundancy payments. Those payments fit badly with the terms of the Scheme: an employer who re-engages dismissed workers following the Scheme remains legally liable to make the payments.

There are other oddities and uncertainties in the existing law. The provisions on lay-off and short-time working don’t apply where wages aren’t dependent on the amount of work done. It’s open to doubt, on the present state of the law, whether the coronavirus epidemic would amount to frustration of employment contracts; but it would be unwise to discount this argument being revived in such exceptional circumstances so as to put a common law block on the duties arising in the first place. Indeed, if the law is perceived at imposing excessive rigidities on reaching tailored solutions to preserve employment, arguments based on frustration might become more appealing to employers.

Dismissals might alternatively be justified for the rarely-used reason in s.98(2)(d) ERA in those industries where there has been a Government ban on working, with the uncertainty about the duration of the ban tending to make dismissals fair.

Finally, if employees are dismissed or have a gap in service during the pandemic but then return to work afterwards, they may well find their continuity of employment has been lost on the basis that they ceased to have contracts of employment in this period (see s.212 ERA). The effect of this is the loss of important statutory rights that depend upon a qualifying period of continuous employment.

The difficulties of effecting rapid changes to working conditions in times of acute crisis, coupled with the actual or potential common law and statutory liabilities, may cause the employer to give up, make its workforce redundant and go into insolvency. Where this occurs, state guarantees swing into place: see ss. 166-170 ERA (individual redundancy payments); Part XII of ERA (unfair dismissal basic awards, guarantee payments, some wages and protective awards: see s.184). The social and economic consequences of widespread insolvencies are unthinkable at the current time, reflected in the Chancellor’s announcement, and these liabilities risk adding to the state’s already over-burdened finances. While in ordinary circumstances the payments provide a small cushion for dismissed workers, the social goal now should be on rescue packages in advance not ex post facto.

The details of the Scheme to date suggest it is restricted to employees and will only apply to (i) employees who (ii) would otherwise have been made redundant during the crisis. It seems to be restricted, so far as we can tell, to employees who are ‘furloughed’, meaning those who aren’t working. In order to put flesh on its bones, we suggest legislation is required, building on the existing model of guarantee payments in Part III of ERA and the provisions on lay-off and short-time working in ss 147-154. By s.28(1) of ERA, the right to a guarantee payment arises where a worker cannot work by reason of an ‘occurrence affecting the normal working of the employer’s business in relation to the work of the kind which the employee is required to do’, wide enough to catch the pandemic (North v Pavleigh [1977] IRLR 461).

But the existing statutory provisions need modification if they are to give effect to the Scheme. We suggest the following to ensure the Scheme is clear, workable and fair in its operation and meets contemporary working practices, even if this will add to the cost of the Scheme. The overarching aim is to protect jobs and demand. To that end it is critical that the Scheme is transparent and sums are payable quickly, before employers go bust: anecdotal evidence suggests cash-flow is an urgent problem for many employers. We propose the following changes to the Scheme.

Proposal 1: Application to ‘workers’

It should apply to ‘workers’ and not only ‘employees’. The distinction between these groups is increasingly hazy and is progressively disappearing in labour law (see this blog post). The present crisis calls for a Scheme which is wide in scope, fair in application and certain to whom it applies so that employers (and workers) can know now whether they will have their pay reimbursed. The existing concept of ‘employee’ fails these tests, as a wealth of recent appellate cases show. The existing Scheme will exclude, for example, the approximately one million agency workers who already receive less pay than direct employees despite often working alongside them, doing the same job. Owing to the obscurity of the UK’s common law definition of ‘employee’, it may not apply to around a million workers on zero-hours contracts who are especially vulnerable to losing income where work dries up. It is not in the interests of employers or their workers that they should have to worry now whether they meet the elusive and unpredictable criteria for determining who is an ‘employee’ at common law. Extending the scheme to ‘workers’ should greatly alleviate concerns about the scope of the Scheme.

Whether a similar scheme should extend to the self-employed who aren’t ‘limb (b)’ workers but who are genuinely running their own business for clients, as Keir Starmer has argued and as the Government is apparently considering, is beyond the scope of our economic expertise. At present, however, the Scheme leads to a glaring disparity of treatment between two groups who are often barely distinguishable in fact in their working arrangements: employees, already the most privileged category in other respects in labour law, will receive 80% of their salaries, whereas limb (b) workers, typically subordinate wage labour, are restricted to access to Universal Credit at SSP rates.

Proposal 2: short-time working or unpaid leave

Critically, the Scheme should apply both to those who would have been made redundant during the crisis and to those retained but placed on short-time working or unpaid leave and who lose pay as a result. Otherwise, perverse effects or incentives will arise: worker A sent home with nothing to do has her wages reimbursed as to 80% under the Scheme, whereas worker B working two days a week at home on matters essential to her employer’s business and receiving one-fifth of normal remuneration gets no reimbursement. The employer would thus have an incentive to run the business on a skeletal staff working long hours, even if that damaged the business. Similar issues could arise in respect of those engaged on zero hours contracts, who may not be dismissed or made redundant at all, as the Scheme appears to require to trigger reimbursement, but simply given no work. Such a restriction will also undermine collective agreements made to combat the crisis. BALPA and British Airways, for example, recently entered into collective arrangement by which pilots were placed on two weeks’ unpaid leave and encouraged to undertake part-time work; Ryanair cut staff pay by 50% in order to navigate the troubled waters; other airlines have made entered into similar temporary arrangments. So long as the Scheme is restricted to employees who were or would have been laid off, it risks undermining creative collective solutions that are addressed to the particular needs of the business or sector.

The German scheme – the ‘Act on limited crisis-induced improvement of short-time work’ (the Gesetz zur befristeten krisenbedingten Verbesserung der Regelungen für das Kurzarbeitergeld for keen scholars) already applies to short-time working. So does the Swedish scheme, under which the state covers 75% of wage costs where workers who are placed on reduced hours. For the UK, the short-time working provisions in s.147 ERA provides an existing model for adoption. However, those provisions only apply where an employee is provided with less work than he is required to do under his contract (s.147(2), meaning it won’t apply to those, such as workers on zero-hours contracts, who owe no obligation to work (see Mailway v Willsher [1978] ICR 511, paras 11-13).

To that end, the Scheme should apply, we suggest, where employees in fact work less time during the crisis than they ordinarily would have worked. The mechanism for doing this could be based, for example, on whether the workers receive less than their ordinary pay, determined over a reference period,  (compare the provisions on a ‘week’s pay’ in s.224 ERA). The level of reduction in hours/pay at which the Scheme bites is a political question. The short-time working provisions in s.147 apply where the worker receives less than a ‘half a week’s pay’; in Germany, the legislation requires a reduction in income of 10% of the workforce, with each losing at least 10% of income. We acknowledge that the prioritisation of full cessation in the Scheme may reflect a public health objective to encourage social distancing. Any short-time working arrangements must also be accompanied by an agreed plan for homeworking and other forms of distanced work arrangements. By this means, the overarching goal of the Scheme, to ensure the retention of the workforce, is better achieved.

Proposal 3: those sent home because of sickness or to prevent spread of the virus

The Scheme should apply not only to those who would have been made redundant but also to those sent home because of sickness or to prevent the spread of the virus, and who suffer a reduction in pay as a result – regardless of whether they were or would have been made redundant. Once more, the threshold where the Scheme bites is to be determined. At present, and absent a contractual right to full or substantial sick pay, employees who can’t work for this reason may only be entitled to SSP. As the Resolution Foundation has explained, the sectors most hit by the coronavirus are dominated by low-paid employees, without an entitlement to contractual sick pay and who often cannot work at home. While the Government’s Scheme already provides for the reimbursement of SSP for employers with fewer than 250 employees, these employees are just as deserving as those who would have been made redundant during the crisis.

Proposal 4: caring responsibilities

Thought should equally be given to extending the Scheme to pay the wages of those who are forced to stay at home because of caring responsibilities. This is especially acute when schools close, which may make it impossible for some employees to work. As a first step, employees who have to stay at home should be protected against dismissal; as a second, the reimbursement of wages under the Scheme should apply to them.

Proposal 5: intersection with existing statutory duties

In the current exceptional crisis, the intersection of the Scheme with existing statutory duties needs to be identified and rationalised. At present, first, if employees had been dismissed and made redundant but are now reinstated owing to the Scheme, their potential entitlements to redundancy payments and protective awards remain. Second, if the employees had been laid off but weren’t dismissed, the employer remains liable to make redundancy payments under ss 147-154 of ERA. Third, the employer is potentially liable for guarantee payments in respect of ‘workless days’ in the interim period under Part III of ERA. It is undesirable and unfair on employers that these liabilities should persist where the Scheme bites.

Proposal 6: impact on pensions

Finally, the impact of the Scheme on pensions needs to be considered, even if this takes us into complicated territory. Most affected employees will be in their employer’s defined contribution pension arrangement.  The employee’s contract of employment will set out what both the employer and employee must contribute as a percentage of ‘pay’ to that arrangement. ‘Pay’ may not be defined in a way that will align seamlessly with what the employee is receiving while ‘furloughed’. If affected employees in a  defined benefit pension scheme are still accruing benefits then, unless the scheme rules are fortuitously tailored to address the current situation, the employee and employer may well still be liable under the scheme rules to pay normal contributions, even though pay is in fact reduced.  The Scheme needs an override to adjust the pension scheme rules to cater for this situation, and for other complications which may arise (such as what is to happen if an employee retires or, worse, dies while furloughed: it would be iniquitous if the calculation of his / her pension or death benefits were detrimentally affected by his lower  pay in in this emergency period). The complications of pensions, and the size of the liabilities, mean that it is hoped the Government have already engaged the lawyers at the Department for Work & Pensions and the Government Actuary’s Department in order to address these important issues.

The Scheme, adjusted as we have suggested, should provide a powerful incentive to employers not to dismiss employees but to retain them in employment. But emergency legislation could go further, and improve the mechanisms by which changes can be affected to terms and conditions. The goal here should be to encourage collective solutions to the crisis, exemplified by the arrangements reached by BALPA and British Airways. The social models adopted in mainland Europe are once more ahead of the game. In Germany, under the new legislation adopted in the wake of the corona crisis, any arrangement on short-time working has to be agreed by the works council, where this exists; in Sweden, any furlough must be approved by a recognised trade union or, where none exists, by 70% of the workers.

Some such scheme should be established in the UK. Subject to the approval of the relevant Minister, a collective or sectoral agreement could provide a means of rapidly effecting changes to terms and conditions – regardless of individual objection. Where no union is recognised, a system could be adopted along the lines of the Swedish model. This would provide both flexibility and legal certainty. We cannot implement a system of enterprise-based works councils rapidly enough to mimic the German solution. However, it might be possible to use off-the-peg solutions, such as the concept of a ‘pre-existing agreement’ under the Information and Consultation of Employees Regulations 2004. Where a workplace agreement meets certain defined criteria (e.g. in writing and approved by the employees) it could be used to implement and enforce short-time working or other urgent changes. Such arrangements must also be subject to a ‘sunset clause’ requiring regular review and renewal.

The rules on continuity of employment require modification, to cater for workers who are temporarily dismissed or given no work during the crisis. Such breaks in service can lead to them losing continuity, so that the clock is re-set at zero when and if they return to work. Express statutory provision should be made for continuity to be preserved where breaks are due to the crisis, including for absences resulting from caring responsibilities. The breaks may be a ‘temporary cessation of work’ within the meaning of s.212(3) ERA, but we suggest it would be preferable to have a more explicit statutory ‘bridge’. The same should apply, for example, if an employee ceases to have a contract in relevant weeks because they are self-isolating (even if they are not strictly absence because of sickness – see s.212(3)(a).

The existing rules give workers the right not to be subjected to a detriment for leaving the workplace in case of ‘danger’ which the worker ‘reasonably believed to be serious and imminent’ and for taking appropriate protective steps in such circumstances (s.44(1)(d) and (e) ERA). Parallel provisions in s.100 make it automatically unfair to dismiss an employee for such reasons. There is a degree of uncertainty as to whether these provisions would operate to protect employees who leave their workplace because of reasonable fears about the risk of infection from coronavirus, even if the concepts of ‘danger’ and ‘serious and imminent’ have been interpreted widely: see Harvest Press v McCaffrey [1999] IRLR 778. Express statutory wording should put the matter beyond doubt.

More radically still, legislation should provide a stick as well as a carrot to prevent dismissals taking place. In light of the existence of the Scheme, an employer who made redundant employees eligible for reimbursements ought to face a high threshold in justifying such dismissals. It may be that the test of unfairness in s.98 ERA is sufficiently flexible to meet this; alternatively, there could be a statutory presumption that the dismissal was unfair if the reason relied upon was a lack of work during the crisis (perhaps with uncapped liability). At the same time, the legislation should block attempts to resurrect the frustration doctrine.

More radical steps should be taken for large-scale redundancies, whereby employers above a certain size threshold or contemplating a certain number of redundancies would not be permitted to implement collective redundancies or wage cuts without prior approval from the relevant Minister. Under s. 193 TULRCA an employer is already required to notify the Secretary of State where the employer is proposing to dismiss 100 or more workers within a period of 90 days or less. This is enforced by a penal provision in s 194 TULRCA which specifies an offence of ‘failure to notify’. We would go further than this provision in envisaging (i) a lower threshold before the restrictions are triggered; (ii) a requirement of Ministerial approval, without which any purported contractual changes would be of no legal effect; (iii) more serious penal measures in the event of a ‘failure to notify’ and seek approval from the relevant Minister.

It is critical that these provisions apply to a wide range of ‘workers’ and not only employees, thus bringing UK law into line with the ILO Recommendation 189 on employment relationship. Such approval would not be granted unless the employer could demonstrate that it had consulted with a representative trade union or employee representatives, as already foreshadowed by s.188 TULRCA. But, unlike the way s.188 operates at present, consultation and consent would be a pre-condition to dismissals occurring at all rather than a means of providing compensation after the event.

It is worthwhile recalling the restrictions on labour supply in the wartime Essential Work Order (The Essential Work (General Provisions) (No 2) Order, SR & O 1942/1594). As Deakin and Wilkinson explain (The Law of the Labour Market, p 99), this Order emerged against the backdrop of industrial practices of lay-off without pay during the interwar years: ‘the wartime Essential Work Order helped limit this abuse…by requiring the employer to find work for the employee and restricting the power of both sides to terminate the contract of employment at will.’ This was linked to a policy of promoting collective bargaining over Guaranteed Week Agreements, supported by the compulsory arbitration provisions in Order 1305.

Our current crisis is configured differently, of course. We are not yet in the position of requisitioning and directing labour to meet essential needs (though we may yet reach such a situation). Under the current Scheme, there should be restrictions on one party’s power to terminate the contract – the employer’s. And the current policy is not to find work for the employee: it is to ensure the maintenance of the employment relationship either through ‘furloughing’ or short-time working until the crisis has passed.