United Kingdom Journals

The sick men and women of Europe: (Case C-400/95) Handes-Og Kontorfunktionaerernes Forbund, acting on behalf of Larsson v Dansk Handel & Service, acting on behalf of Fotx Supermarked A/S [1997] IRLR 643

Kay Wheat

Senior Lecturer in Law, Nottingham Trent University
<kay.wheat@ntu.ac.uk>

Copyright © 1998 Kay Wheat.
First Published in Web Journal of Current Legal Issues in association with Blackstone Press Ltd.

Summary

In this case the European Court of Justice ruled that the dismissal of a woman because of absence due to pregnancy-related illness is not necessarily contrary to the Equal Treatment Directive. The earlier ECJ cases of Dekker and Webb had stressed that pregnancy is not an illness, and it may, therefore, have been inferred that pregnancy-related illness is not to be equated with other forms of illness. However, the Larsson decision states that once the protection of the maternity leave period has ended, the woman who is absent from work due to pregnancy-related illness is to be compared with a man absent through illness. Unless she is treated less favourably, a dismissal for this reason will not be contrary to the ETD. This article examines the implications of the decision.

Contents

1. Introduction
2. The relevant legislative provisions
   1. The Equal Treatment Directive 76/207
   2. The Pregnant Workers Directive 92/85
3. The case law prior to Larsson
4. The case advanced by Ms Larsson
5. The case against Ms Larsson
6. The decision of the European Court
7. Does this depart from the earlier ECJ decisions?
8. Does Larsson depart from the principle of equality?
9. Conclusion

1. Introduction

In Larsson the European Court of Justice had to consider the lawfulness of a dismissal for a pregnancy-related illness. Ms Larsson commenced her employment in March 1990. In August 1991 she discovered that she was pregnant. Due to the pregnancy she was off sick from 7th August until 24th August, and again from 4th November 1991 until 15th March 1992. Thereafter she went on the statutory maternity leave allowed by Danish law (24 weeks). On 2nd April 1992 she gave birth to a child. Her maternity leave expired on 18th September 1992. She then took accrued annual leave until 16th October 1992. As a result of pregnancy-related illness, however, she did not return to work. On 10th November 1992 she was dismissed with effect from the end of December 1992, because of her "lengthy period of absence and the fact that it is scarcely likely that you will at any time in the future - on the grounds of health - be again in a position to carry out your work in a satisfactory manner". The case was referred to the European Court for a preliminary ruling as to whether this was contrary to the Equal Treatment Directive, 76/207/EEC. The ECJ found that it was not.

Top | Contents

2. The relevant legislative provisions

(i) The Equal Treatment Directive 76/207

The purpose of the Equal Treatment Directive (ETD) is to put into effect the principle of equal treatment for men and women under Article 119 of the Treaty of Rome. Article 2(1) states that ... "the principle of equal treatment shall mean that there shall be no discrimination whatsoever on grounds of sex either directly or indirectly by reference in particular to marital or family status". Article 5(1) of the Directive states that "application of the principle of equal treatment with regard to working conditions, including the conditions governing dismissal, means that men and women shall be guaranteed the same conditions without discrimination on grounds of sex".

Top | Contents

(ii) The Pregnant Workers Directive 92/85

This was not relevant for the purposes of Ms Larsson's case because at the time she was dismissed it had yet to be adopted. The court, however, considered its provisions. This Directive (PWD) is not based upon Article 119 but upon Article 118a, which is concerned with health and safety and improvements in the working environment. Its main provisions impose a duty upon employers to take steps to minimise health and safety risks to pregnant/breastfeeding women or those who have recently given birth (Articles 4,5 & 6), and establish a right to a minimum of 14 weeks paid maternity leave which has no detrimental effect upon existing contractual rights (Articles 8 & 11); a right to paid time off for ante-natal examinations (Article 9) and a right not to be dismissed during the period from the beginning of pregnancy to the end of maternity leave, save in exceptional circumstances unconnected with childbirth (Article 10).

The PWD is implemented in the UK by Part VII of the Employment Rights Act 1996 which deals with maternity rights during the subsistence of the contract of employment (providing for a minimum of 14 weeks maternity leave), and by section 99 of the Act which provides that it is automatically unfair to dismiss a woman because she is pregnant or any other reason connected with her pregnancy.

Top | Contents

3. The case law prior to Larsson

In the case of Turley v Alders Department Stores Ltd [1980] ICR 66 the EAT took the simplistic and now discredited view that, since only women could become pregnant, the concept of discrimination could not apply to pregnancy because given the uniquely female nature of pregnancy, no comparison could be made with a man. In Hayes v Malleable Working Men's Club and Institute [1985] ICR 703, the EAT rejected this approach and decided that the correct test in deciding whether a pregnant woman had suffered discrimination was to compare her treatment with that of a sick man, so that, in certain circumstances a dismissal may be justified.

However, the ECJ in [Case C-177/88,] Dekker v Stichting Vormingscentrum Voor Jong Volwassenen (VJV-Centrum) [[1990] ECR I-3941;] [1992] ICR 325 made it quite clear that, as only women can be adversely treated on the ground of pregnancy, such treatment constitutes direct discrimination, which cannot be justified. The cases of [Case C-421/92,] Habermann-Beltermann v Arbeiterwohlfahrt, Bezirksverband Ndb/OpfeV [[1994] ECR I-1657;] [1994] IRLR 364 and [Case C-32/93,] Webb v EMO Air Cargo (UK) Ltd [[1994] ECR I-3567;] [1994] IRLR 482 confirmed respectively that, a contract of employment cannot be avoided on account of a prohibition under national law of night work during pregnancy, and that a woman recruited for an indefinite period and who is dismissed because pregnancy prevents her from carrying out part of that contract will have suffered direct discrimination.

The specific issue of pregnancy-related illness was considered in [Case C-179/88,] Handels-Og Kontorfunktionaerernes Forbund I Danmark (acting for Hertz) v Dansk Arbejdsgiverforening [[1990] ECR I-3941;] [1992] ICR 332. Ms Hertz's contract of employment began in July 1982. She became pregnant and took maternity leave until late 1983. Between June 1984 and June 1985 she took in total 100 days sick leave. It was not disputed that this illness was consequent upon pregnancy and confinement. She was dismissed in June 1985, as it was the employer's "normal practice to dismiss workers for frequent absence due to sickness".

In an action by Mrs Hertz alleging that the dismissal was in breach of the ETD, two questions were referred to the European Court. The first asked whether Article 5(1) and Article 2(1) precluded a dismissal as a consequence of absence due to illness which is attributable to pregnancy or confinement. If the answer was in the affirmative, the second question asked whether protection against dismissal due to illness caused by pregnancy or confinement would be unlimited in time.

It was argued that it would be unfair to prevent an employer from dismissing a woman on account of frequent periods of sick leave just because these were due to pregnancy or confinement on the grounds that it would entail administrative difficulties for employers and would have negative repercussions for women. The court accepted this view, as such periods of sick leave would lead to the dismissal of a man under the same circumstances. Referring to Article 2(3) of the ETD, the court said that this gave protection during maternity leave, but that the Directive does not envisage the case of an illness attributable to pregnancy or confinement, and "[i]n the case of an illness manifesting itself after the maternity leave, there is no reason to distinguish an illness attributable to pregnancy or confinement from any other illness. Such a pathological condition is therefore covered by the general rules applicable in the event of illness" (para. 16). It was held therefore that the general rules relating to illness dismissals should apply and that the dismissal was not discriminatory.
As the answer to the first question was in the negative there was no need to give an answer to the second.

Top | Contents

4. The case advanced by Ms Larsson

The illness from which Ms Larsson was suffering at the end of her maternity leave was said to be a continuation of a condition which had arisen during her pregnancy/maternity leave. She sought to distinguish this from a case where the illness, though causally connected with pregnancy had arisen de novo following the end of a period of maternity leave. Her argument was that, in her case, the protection of the ETD applied, and that to find otherwise would mean that the protection of the Directive would depend upon the Member State's period of maternity leave.

Top | Contents

5. The case against Ms Larsson

The UK government argued that the ETD includes recognition of a sufficient period of maternity leave to accommodate for the normal consequences of pregnancy. Therefore, if the period is sufficient, even if the illness had arisen as a result of pregnancy and had commenced during maternity leave it should not put her in a better position than a man who is on sick leave, excluding the maternity leave period.

The Netherlands relied upon the PWD which prohibits dismissal from the start of pregnancy through to the end of maternity leave. The argument was that it should be concluded from this that such a period was considered by the Community legislature to be sufficient to protect women against dismissal for absence from work attributable to pregnancy and confinement, without there being any obligation on Member States to provide more extensive protection.
The Commission contended that the PWD adds nothing to the ETD in terms of an extension of anti-discrimination protection, so that it is not discrimination unless dismissal is based wholly or in part upon pregnancy or maternity leave.

Top | Contents

6. The decision of the European Court

The ECJ ruling was that absence following maternity leave was to be regarded in the same way as absence by a man, i.e. if a man would have been dismissed after such a period of sickness absence, then it would not constitute less favourable treatment. It was irrelevant that the woman's illness may have arisen before this or even whether it originated in pregnancy. However during both pregnancy and maternity leave, the woman cannot be dismissed (save for reasons wholly unconnected with her pregnancy and confinement), nor can absence during this time be taken into account upon a subsequent dismissal. This was based upon the decision in Hertz, and the provisions of the PWD, which although not applicable to Ms Larsson, were nevertheless considered by the court.

As far as the ETD is concerned, the ECJ considered whether the Directive made it clear that absence during the period of maternity leave could not be brought into the calculation of the total length of absence in deciding whether a dismissal was justified. The conclusion was that this period could not be taken into account in the calculation. If the same rules were applied to men and women in terms of such a calculation it would be direct discrimination because only women take maternity leave (para. 25).

However, the conclusion to be drawn from the application of the ETD is that it does not preclude account being taken of a woman's absence from work between the beginning of her pregnancy and the beginning of her maternity leave. In this case therefore, Mrs Larsson's absence from 4th November 1991 to 15th March 1992 could be taken into account in deciding whether the total length of her absence was such as to justify dismissal. It was found that the PWD, however, would prevent this happening as the 'protected period' under that Directive includes the start of the pregnancy, although adoption of this came too late to assist Ms Larsson.

The argument advanced by Ms Larsson, was that there is a difference between illness caused by pregnancy which arose during pregnancy or maternity leave, and that caused by pregnancy which arose afterwards, because otherwise the protection afforded would depend upon the maternity leave entitlement under domestic law of the Member State concerned. This is an odd argument because, if the 'before or after' distinction is relevant, far from being to the contrary, protection would necessarily depend upon the Member State's period of maternity leave in all cases where the illness arose on or after the minimum period of leave stipulated in the PWD. In the event, her argument was rejected by the court (relying upon the statement in Hertz at para. 17) on the basis that men and women are equally exposed to illness, and the specific time when the illness arose made no difference.

Top | Contents

7. Does this depart from the earlier ECJ decisions?

It has been suggested by Michael Rubenstein (Commentary [1997] IRLR, p 607) that Larsson could be distinguished from Hertz, because in the latter case either the unusual facts (the woman had returned to work for something like six months before the illness manifested itself) or the fact that subsequent cases such as Webb had overtaken Hertz. However, is Larsson really inconsistent with the earlier ECJ cases? It is submitted that it is not.
Dekker had stated clearly that pregnancy is not an illness, and had firmly rejected any assimilation of inability to work on account of pregnancy and confinement to inability to work on account of illness (at para. 14). Webb had unequivocally rejected the view that pregnancy is a pathological condition (at para. 25).

Top | Contents

8. Does Larsson depart from the principle of equality?

Although the decision in Larsson cannot be said to be at odds with these earlier decisions, can it still be argued that it is contrary to the intention of the ETD? It can certainly be argued that as pregnancy-related illness can only be suffered by women, dismissal, or other unfavourable treatment, will always be prohibited by the ETD. However, if this were the case, would it also mean that gender-specific illness would then be brought into the ambit of protection, such as gynaecological problems generally, and, for example, prostate conditions in men? The implication of Larsson and Hertz is that illness relating to the specific biological make-up of the person affects men and women alike and therefore would not be direct discrimination. We would then be left with the argument, based on indirect discrimination, that women more than men suffer from gender-specific conditions, and therefore applying the same requirement or condition in respect of sick leave or sickness-related dismissals, means that a considerably smaller number of women can comply with it. However, the law on pools of comparison may prevent the woman from providing the necessary statistical evidence (if it exists), as she may not be allowed to compare the group of men suffering from gender-specific conditions with the group of women suffering from gender-specific conditions, but only sick men and sick women, regardless of the nature of their condition.⁽¹⁾

Moreover, and crucially, unlike direct discrimination, indirect discrimination is capable of being justified.
Was Dekker over-hasty in rejecting the assimilation of an inability to work on account of pregnancy and confinement with an inability to work based upon illness? Perhaps this is so, but the stress upon the difference is also reflected in the PWD, which, although it refers to the minimum level of maternity pay as being "at least equivalent to that which the worker concerned would receive in the event of a break in her activities on grounds connected with her state of health" (Article 11(3)), nevertheless, in its preamble seeks to emphasise that this is nothing to do with the conflation of pregnancy and illness by stating: "....the concept of an adequate allowance in the case of maternity leave must be regarded as a technical point of reference with a view to fixing the minimum level of protection and should in no circumstances be interpreted as suggesting an analogy between pregnancy and illness" (emphasis added). (That is not to say , however, that the comparison with sickness pay provision may not be relevant in deciding what is an adequate period of paid maternity leave. Under the ETD, it is arguable that paid leave on grounds of pregnancy should be at least comparable to paid sickness leave (see [C-342/93] Gillespie and others v Northern Health and Social Services Board [1996] IRLR 214, and Gillespie and others v Northern Health and Social Services Board (No 2) [1997] IRLR 410.)

It is arguable, however, that this blanket refusal to treat pregnancy as an illness fails to distinguish between a 'normal' pregnancy and one in which exceptional pathological consequences can arise. The PWD, it could be argued, is about 'normal' pregnancy, and that additional protection should be available to women who suffer pregnancy-related illness which is over and above the 'normal'.

The difficulty in arguing that pregnancy-related illness is discriminatory per se is that, since during the period of maternity leave a woman is protected, and now, since the PWD, illness during the early part of pregnancy prior to maternity leave is also protected, the only circumstances in which the woman is unprotected is after the expiration of her maternity leave. That gives rise to the question as to how long any subsequent period of protection must be. Although it was not necessary for the court in Hertz to decide whether the protection in respect of pregnancy-related illness was unlimited in time, this question would have had to be answered if Larsson had been decided differently. A decision would have had to be made by the ECJ as to how long the period of protection would last.

Would the court have found the protection to extend for an indefinite period of time? It is unlikely that any court would have imposed a duty upon employers to retain a sick employee for an unlimited period. Similarly, if she goes back to work and then develops a further episode of pregnancy-related illness, a decision would have to be made as how long the period of 'healthy' intermission would have to be before the employer could treat her as any other sick employee would be treated. In both cases, we would be left at looking at what is a 'reasonable' period. The only way in which this period of time could be ascertained would be by looking at a variety of factors such as the length of the woman's service, the nature of the employer's business, the urgency of the need to find a replacement and so on. In other words, precisely those factors which would be relevant to a decision about the dismissal of a man who was unavailable through sickness. Michael Rubenstein (op cit) states that: "Instead of dismissals for pregnancy-related illness being regarded as discriminatory per se, subject to remoteness exceptions as in Hertz, the general rule is now that dismissals for pregnancy-related illness are not discriminatory per se....." However, what is to be made of the reference to 'remoteness exceptions'? This is surely only another way of saying that some temporal restrictions must be placed upon the time and duration of the pregnancy-related illness. At present such restrictions can only be imposed by using the analogy of whether it would be reasonable to treat a sick male employee in this way. Additional protection could only be provided by means of further European legislation, for example, to extend the 14 week period and allow for the period of maternity leave to include a number of weeks over a fixed period after the woman has returned to work, subject to the production of a medical certificate confirming that the illness is pregnancy/childbirth-based.

The UK's obligation to implement the PWD is enacted in the Employment Rights Act. Section 99(1)(a) prohibits dismissal for pregnancy or any other reason connected with pregnancy. It might be thought that this latter phrase offers women better protection under UK law. However, the possibility of pregnancy-related illness being an unlawful reason for dismissal is put out of question by later statutory provisions. Section 99(3) states:

"An employee who is dismissed shall also be regarded for the purposes of this Part as unfairly dismissed if -

(a) before the end of her maternity leave period she gave to her employer a certificate from a registered medical practitioner stating that by reason of disease or bodily or mental disablement she would be incapable of work after the end of that period,
(b) her contract of employment was terminated within the period of four weeks beginning immediately after the end of her maternity leave period in circumstances in which she continued to be incapable of work and the certificate remained current, and
(c) the reason (or, if more than one, the principal reason) for the dismissal is that she has given birth to a child or any other reason connected with her having given birth to a child."

The effect of this provision, therefore, is to give the woman a further four weeks protection following maternity leave if she is ill and provides the necessary medical evidence. However, the four week restriction clearly implies that there is no further protection from dismissal due to illness.⁽²⁾

Top | Contents

9. Conclusion

The decision in Larsson has taken the proposition that pregnancy is not an illness to its logical conclusion. It has also made it clear that during both pregnancy and maternity leave, absence due to pregnancy-related illness cannot justify dismissal, and, importantly, that it cannot be taken into account if dismissal takes place after the end of maternity leave. The question is whether this clear ruling, delineating a bright line of protection is marred by the court's refusal to treat pregnancy-related illness as a special case.

Larsson has also made it clear that even if a woman is dismissed for sickness absence after her period of protection has ended, she may still be able to argue that the dismissal is unlawful discrimination if a man absent due to sickness would have been treated more favourably. The question remains open as to whether some additional protection for exceptional consequences of pregnancy is justifiable, and whether the analogy of a sick man will provide it.

(1) ^.In Kidd v DRG Ltd [1985] ICR 405, the EAT held that the definition of the pool is a question of fact for the industrial tribunal to decide in each case, and therefore there are very limited rights of appeal. See also Jones v University of Manchester [1993] ICR 474.
(2).The fact that pregnancy-related illness dismissals would receive no special protection was confirmed in the debate on the Trade Union Reform and Employment Rights Bill, which introduced legislative amendments necessary to comply with the PWD (Hansard (HC) Standing Committee, 12th January 1993. It should also be noted that in the United States, Title VII of the Civil Rights Act 1964 (as amended) protects women from discrimination only to the extent that, if they are affected by pregnancy or related conditions, they must be treated in the same manner as other applicants or employees with similar abilities or limitations.