Polkey confusion
The EAT continues to muddy the unfair dismissal law waters, say Elliot Gold and Francis Davey
Until recently, the House of Lords decision in Polkey v AE Dayton Services [1988] ICR 142 made a dismissal unfair if an unfair procedure was used, even if the dismissal would have happened, had a fair procedure been used. The harshness of this rule was improved by permitting an employment tribunal to reduce the compensation given to an employee if the dismissal was still likely to have happened in any event.
Section 98A(2) of the Employment Rights Act 1996 states that a dismissal will be fair where an employer fails to follow 'a procedure' other than one of the statutory procedures but would anyway have taken the decision to dismiss. That would appear to 'reverse' the Polkey principle.
The key question is: what is a procedure? Does it apply only to an employer's formal procedures, or does s 98A(2) make any procedural unfairness irrelevant? Frustratingly, the Employment Appeal Tribunal (EAT) has addressed this question four times, each time coming to the opposite conclusion as its previous decision.
Four seasons
The first decision was in Pudney v Network Rail Infrastructure Ltd [2006] UKEAT/0707/05. HHJ McMullen QC stated that the word 'procedure' in s 98A(2) indicated some formality. It included those procedures that were incorporated in contracts: those set out in policy documents, which might not be contractual; and those that were found in unwritten custom and practice. It did not cover a lapse from general standards of reasonableness, or the ACAS Code.
Next came a contradictory decision in Alexander v Bridgen Enterprises Ltd [2006] IRLR 431. Mr Justice Elias, president of the EAT, stated that there was no limitation on the nature of the procedural breaches caught by the subsection. The term 'a procedure' meant any procedure with which the tribunal considered the employer ought to have complied. If the employer failed to comply with a procedure that ought to have been followed, that would not render the dismissal unfair, if the employer showed that it would have dismissed anyway.
The next decision, Mason v Ward End Primary School [2006] EAT/0433/05, re-affirmed Pudney. HHJ McMullen QC considered that 'a procedure' meant 'a particular procedure in force at a given workplace, the terms of which [were] not carried out'. This included procedures established by custom and practice such as a simple rule of 'last in, first out' in a redundancy situation. However, the ACAS Code of Conduct while containing important and authoritative advice upon the handling of disciplinary matters was not itself 'a procedure'.
The fourth, and most recent, decision in the chain was the EAT judgment in Kelly-Madden v Manor Surgery UKEAT/0105/06/DM, where Elias J re-asserted his view in Alexander, albeit 'diffidently'. He stated that the subsection referred to 'a procedure' without limitation or qualification. Had the intention been to adopt the qualification suggested in Mason, the draughtsman would have so indicated. Otherwise, an employer who had such procedures could take advantage of the section, but not an employer who had not. The statute would operate as a punishment for those employers who had failed to adopt such procedures. To say that there was a breach of the ACAS code was to say that there was a breach of the procedural provisions which the Code recommended; but that was still properly described as a breach of procedures.
Elias J continued by stating that it could not be inferred that Parliament intended that the legislation would apply only to procedural defects found in procedures adopted by the employer, but not reflecting ACAS standards or the standards of the reasonable employer. Otherwise, only certain, very minor, breaches would be unlikely to render a dismissal unfair in any event. The section was broader in its effect. Whenever a tribunal was minded to find that a dismissal was unfair for procedural reasons alone, it was open to the employer to show that compliance would have made no difference. Where the employer could show this, and that the defects did not infringe the statutory dismissal procedures, the dismissal would be fair.
Where are we now?
We are now left in the unsatisfactory position of having two lines of authority that may be followed in the future. If Pudney and Mason prevail, the effect of s 98A(2) would be modest: an employer would still have to attempt to deal fairly with an employee on dismissal, but would not be penalised for failing to follow its own procedures '“ which might be far more generous than those recommended by ACAS '“ where a failure makes no practical difference.
By contrast, if Alexander and Kelly-Madden are ultimately followed, the effect of s 98A(2) would be rather more drastic. An employer would be able to defend all forms of unfairness in dismissal '“ even conduct that fell far below the general standards of fairness '“ provided that the conduct made no difference to the outcome.
Such a position would make it very difficult for human resources departments to persuade management that instituting fair procedures was desirable. It would also seem to place no value on the principle of procedural fairness '“ something that is valued in administrative and judicial contexts for its own sake.
At present, the decisions on s 98(2) have ping-ponged between two judges of the EAT who have failed to come to agreement. This lack of clarity will make it difficult to advice clients and to decide on the best strategy for conducting employment litigation. It is to be hoped that the Court of Appeal will soon provide some definitive guidance on this point.