Colm O’Cinneide. Senior Lecturer in Law, UCL

Council Directive 2000/78/EC (the ‘Framework Equality Directive’) requires the UK to introduce comprehensive legislation prohibiting age discrimination in employment and occupation by December 2006. [1] The long-awaited age discrimination regulations, initially published in draft form in July 2005 in the Coming of Age consultation document, were finally approved by Parliament in March 2006 and come into force on 1st October 2006. In place of a long-established and largely ineffective policy of relying upon non-binding codes of practice to encourage employers to avoid age discrimination, the regulations transform age into another unlawful ground of discrimination, albeit one which is subject to some very important exceptions.

Age discrimination legislation inevitably gives rise to a series of complex and genuinely difficult issues, both conceptual and practical in nature. Many of these difficulties stem from real uncertainty about why age discrimination should be prohibited, and how far should this prohibition be extended. [2] However, comparative experience from countries that have well-established prohibitions on age discrimination in place demonstrates that age discrimination cases will probably be relatively common, often hard-fought and may be difficult to resolve. In the Republic of Ireland, which has comprehensive anti-discrimination legislation extending across all the six major equality grounds, 17% of employment discrimination claims referred to the Equality Tribunals in the period 2000-2003 concerned age discrimination. [3] In the US in the 1990s, 170,000 age discrimination claims were filed with the Equal Employment Opportunities Commission (EEOC), and EEOC support for litigants under the Age Discrimination in Employment Act (ADEA) has produced several high-profile damages awards in recent years, including a 1999 settlement with a New York insurance and brokerage firm, which resulted in $28 million being paid out to 13 individuals following the firm’s implementation of a policy that required employee-directors to retire at age 60 or 62. [4]

Age discrimination legislation is therefore a potential minefield for judges, lawyers, litigants and policy-makers. At first glance, it may appear to be a ‘backwater’ equality ground, which attracts less political controversy and dispute than other areas of anti-discrimination law: however, its potential impact is considerable and should not be underestimated.

Prior to the coming into force of the age regulations, some degree of protection against age discrimination was offered by unfair dismissal legislation and other aspects of employment law. [5] For example, contractual terms of employment have been interpreted in some contexts so as to give some protection against dismissal on age grounds, in particular when it comes to establishing the existence of a ‘normal retiring age’ for a particular class of employee. [6] However, this protection was very limited, with no remedy against the application of contractual terms permitting age-based discrimination, or the use of age-based criteria that fall short of obviously ‘unfair’ standards of treatment. [7] Other forms of anti-discrimination legislation, in particular the prohibition on indirect sex discrimination, can offer some degree of limited protection against certain forms of age-based discrimination.

However, the recent House of Lords decision in Rutherford v Secretary of State for Trade and Industry shows the limits that can exist to this protection. In this case, the denial of unfair dismissal rights to those over the age of 65 (now removed by the Age Regulations) was held not to disproportionately impact upon men notwithstanding the greater proportion of men immediately affected by this age limit: the Court of Appeal and Law Lords agreed that the suitable comparator groups were the proportion of men and women employed in the population at large and not those immediately affected, which meant that no clear disproportionate impact could be established. [8]

The Employment Equality (Age) Regulations 2006 now change the picture, prohibiting direct and indirect age discrimination, harassment and victimisation in employment and occupation. Age has become the latest example of what Owen Fiss has called the ‘proliferation of the protectorate’, [9] becoming another ‘controlled’ ground of classification as required by the Framework Equality Directive.

The scope of the Regulations and their provisions on harassment and victimisation are similar to the scope and provisions of other anti-discrimination legislation in the field of employment and occupation. [10] However, the Regulations differ in some very important respects from other forms of anti-discrimination legislation, and it makes sense therefore to focus on these differences for now.

One important difference that immediately leaps out is that the prohibition on age discrimination is confined to employment and occupation, and, unlike the other forms of anti-discrimination legislation, does not extend to discrimination in the fields of goods and services, education etc. Jonathan Swift in a recent article in the ILS has placed great emphasis on this decision to confine the scope of age discrimination legislation to the employment context, suggesting that it indicates how different age discrimination is from other forms of inequalities. [11] However, it should be noted that many jurisdictions (such as Canada and Ireland) do extend their prohibition on age discrimination across the full ambit of social activities covered by other forms of anti-discrimination law. [12] There is also considerable pressure on the UK government to do likewise from Age Concern, Help the Aged and other activist groups, which may ensure that an extension of age discrimination to goods and services will be given serious consideration as part of the Discrimination Law Review. It would be wise not to assume that the Age Regulations represent the end of the process of legislating against age discrimination.

However, just focussing on the current scope of the Age Regulations, the most obvious difference between age and the other equality grounds is that there is considerably more scope for justifying the use of age-based criteria than is the case for most of the other equality grounds (a comparison could be made with disability discrimination, but this is potentially very misleading: see below). The Directive (uniquely) permits direct discrimination on the grounds of age to be objectively justified.

The Regulations therefore make provision in Reg. 3 for a general objective justification defence for both direct and indirect age discrimination. Secondly, the Regulations also make provision for a series of specific exceptions from the scope of the prohibition on age discrimination: the use of certain age-based or age-linked criteria, in particular in imposing mandatory retirement ages, fixing age limits in minimum wage and New Deal programmes, using seniority-based benefit schemes and fixing access requirements for occupational benefits, are exempted from any requirement to show that they are objectively justified.

The government has justified the insertion of these specific exemptions on the grounds that they constitute circumstances where the use of age criteria is clearly objectively justified. They can be challenged for being incompatible with the requirements of the Directive, and the TUC, Age Concern and other organisations have been very critical of the scope of some of these exceptions. Heyday has already launched a judicial review of the retirement age exception. Nevertheless, taken together, the availability of a general objective justification defence combined with these specific exceptions makes age discrimination unique, and as such it deserves close and critical attention.

It is true that as the Ontario Human Commission has noted, age discrimination seems not to give rise to the sense of moral opprobrium that other forms of illegitimate discrimination attract. [13] It is not usually based on bigotry, hatred or the fundamental denial of equal status as human beings: while age stereotyping is a problem, it cannot be said in normal circumstances to take the same virulent forms of other forms of prejudice.^[14] As a consequence, the argument has been made that the moral imperatives to eliminate discriminatory and stereotyping behaviour on the grounds of race and sex discrimination do not apply in the age context, or at least do not apply with anything like the same imperative moral force. However, discrimination on the grounds of age denies individual equality of opportunity whenever it occurs, and can constitute a virulent form of irrational stereotyping. [15] It is also often experienced as demeaning by those exposed to it, even if it may not be as objectionable in the grand scheme of things as certain other forms of prejudice.

The argument is also sometimes made that age discrimination legislation can cut across or conflict with other equality concerns.^[16] There is some evidence that the principal beneficiaries of the American age discrimination legislation have been older, white males: therefore, the argument is made that age discrimination legislation will inevitably results in special protection for well-off older litigants in established posts, while blocking promotion and employment possibilities for younger workers.^[17] However, the suggestion that age discrimination legislation inevitably only benefits specific and relatively privileged social groups is contestable. The US experience is a very unreliable guide, as the ADEA applies just to discrimination against those over forty years of age. In contrast, the Framework Equality Directive requires age discrimination legislation to protect all age groups in employment and occupation, which may break down an assortment of age barriers across the work force, including those that impact upon younger workers. [18] In fact, older or younger women, ethnic minorities and disabled persons may be the prime beneficiaries of age discrimination legislation, as they may be the most exposed to discriminatory treatment, the most vulnerable to its negative effects, and the least able to combat it. In addition, substantial public policy reasons support the imposition of reasonably tough justification requirements: the social exclusion of particular segments of age groups remains a real problem. [19]

Some commentators have suggested that age discrimination may serve useful economic purposes. For example, in the United States, Evan Pontz has argued that the ADEA should not be interpreted as permitting ‘disparate impact analysis’ (otherwise known as indirect discrimination claims) to be applied in the context of age discrimination, due to fears about its potential interference with standard employment practice.^[20] Richard Posner has gone so far as to argue that the range of circumstances in which age discrimination is a useful tool for effective decision-making in employment is so considerable as to justify the repeal or substantial dilution of the US age discrimination legislation. In particular, Posner has suggested that while age is not always a very accurate proxy for individual characteristics, it is economically an efficient way of ascertaining certain types of relevant information about individuals, such as willingness to work for longer hours or productive capacity, than many other methods.^[21]

However, many of these arguments rest on reasonably tenuous foundations. For example, Posner tends to assume that productivity will usually decline with age, while some evidence exists that, except in a particular types of jobs, work performance does not in fact deteriorate with age to an appreciable degree.^[22] The US experience has shown that the use of alternative non-age based methods in employment decision-making is possible without the infliction of crippling economic costs.^[23]

Other ‘sceptical’ arguments that suggest that age discrimination may be hardwired into our economic and social systems, and any unfairness is counterbalanced by the benefits of age discrimination being enjoyed by different groups at different times. ‘Fair innings’ or ‘life cycle’ arguments make the argument that age is not an immutable characteristic, as a person’s age changes over time and all humans pass through the ageing process. As a consequence, it can be legitimate to offer more opportunity in certain circumstances to those at earlier stages of the aging process, on the basis that those further along this aging process previously had their fair share of extra opportunities. In other words, victims of age discrimination may have benefited from beneficial treatment based upon their age at other points of their life-cycle, unlike the case with other forms of inequality: therefore, this trade-off could be seen as reducing the ‘sting’ of age discrimination.

However, such ‘life-cycle’ arguments attribute certain group characteristics to all individuals within particular age groups, irrespective of their actual circumstances or life history. Some older persons may have been able to maximise their life experiences: others have not been able to so do, often due to lack of political power or social capital, and deserve not to be judged on the grounds of their age. [24] Jolls argues that age discrimination legislation can be seen as an effective tool for ‘tying the hands’ of employers in general, preventing them from reneging upon their side of the ‘life-cycle’ bargain. [25] ‘Fair innings’ arguments in employment are also often founded upon the ‘Lazear contract model’, which relies on the premise that older workers are being paid more than their productivity would justify: [26] however, there is limited evidence to support this assumption, and some to contradict it. [27] As with other forms of anti-discrimination legislation, time-honoured assumptions tend to shrivel, once exposed to the rigour of justification demands. To lock the ‘life cycle’ into place by diluting age discrimination legislation may be to protect unnecessary and often unfair employment practices, where better alternatives exist.

However, there may be a narrow but distinct range of situations where an arguable case for using age as a differentiating factor can be made out. Firstly, there may be limited circumstances where the use of age to differentiate between individuals may serve some useful social or economic objectives, while not involving harmful stereotyping. This can arise when the use of the age-linked criteria may yield accurate information about particular individuals. There also may be some circumstances where no reasonable alternative to the use of age as a proxy for individual characteristics is available, and the use of age in these circumstances can provide objectively relevant information about the individuals in question. [28] An employer may have too many employees to make individual assessment a practical proposition, or concerns may arise about whether individual testing will satisfy health and safety concerns.^[29]

Age may also have to be used as a proxy for need in identifying appropriate recipients of particular forms of employment-related social insurance, special support or positive action. Making differentiations between different age groups may also be necessary to adjust to changing socio-economic conditions and to ensure what Robin Allen has described as ‘cross-generational equity’.

Many of these exceptions involve direct discrimination the grounds of age, or age-linked criteria. Nevertheless, they are reconcilable in the abstract with the general prohibition on age discrimination: they usually either involve scenarios where age and age-linked characteristics are not being used to stereotype, or are being used for the purpose of defining groups in need of positive action or special treatment, or are linked to achieving cross-generational equity and therefore can be seen as advancing wider equality concerns.

However, these exceptions are also all capable of encompassing a broad range of age-based distinctions, and of being used as a cloak to camouflage the perpetuation of unnecessary and damaging age-based differentiation. The use of the objective justification test familiar from other anti-discrimination contexts will be necessary to ensure that these exceptions are only applied where their use is necessary and justified. The justifications for prohibiting age discrimination may be different in certain respects from those that underpin other forms of anti-discrimination law, but how age discrimination legislation may be applied and develop in practice may not be so different after all.

However, before going on to examine the circumstances when age discrimination will be deemed to be justified, an initial point needs to be made. Comparative experience from Ireland and the US indicates that much of the legal angst that will be generated by age discrimination litigation will not always surround issues of justification, even though these issues loom large in academic and political debates: instead, problems often arise as to when inferences should arise that that ‘less favourable treatment’ has occurred on the grounds of age.

Determining when ‘less favourable treatment’ has occurred in the age context will involve applying the usual comparator approach familiar from other equality contexts. However, the fluid and changing ways in which a person’s age and the differences in position and expectations that often exist even between persons of very similar ages may make the application of the comparator test difficult in this context. For example, if a 60-year old worker’s contract is not renewed and she alleges that this treatment was based on her age, citing the fact that a 57-year old co-worker on the same grade had his contract renewed at the same time, should an inference arise? What happens if the employer cites the continuing employment of the 57-year old as an example of their willingness to continue to employ older workers? Does the three year difference constitute a significant age distinction at that stage of that particular career, or is the age difference of no real significance? The same problems could obviously arise in respect of indirect discrimination claims.

Courts and tribunals may often be able to identify with ease the appropriate actual or hypothetical comparators from the specific circumstances of a particular case. However, where an obvious comparator does not exist, then difficulties may arise and creative solutions might be required. [30] Otherwise, there is a danger that age discrimination litigation will often degenerate into a debate about comparators that could prove even more intractable than similar debates in other equality cases, as has tended to occur in the US. [31]

Complexities can also arise in deciding whether ‘less favourable’ treatment has taken place when differentiating criteria are used which are not linked directly to an individual’s age, but instead refer to characteristics that are in reality closely linked to age. ‘Seniority’, ‘maturity’ and ‘experience’ are all examples of this type of age-linked characteristic, as they are often only capable of being acquired by those who have spent a number of years in an activity. If an employer selects employees for redundancy on the basis of ‘experience’, this may come close to making age the determining factor.

Bob Hepple has suggested that that courts and tribunals will simply apply the usual ‘but for’ test utilised by the UK courts, [32] and ask whether a claimant’s age was a substantial factor in the decision to subject the claimant to less favourable treatment, or played a role in the making of that decision. [33] The US Supreme Court adopts a similar approach in its age discrimination case-law, [34] as does the Dutch Equal Treatment Commission in its age discrimination case-law. [35] Hepple suggests that this test would mean that the use of age-linked factors which are ‘analytically separate’ from age, such as seniority, experience, qualifications and pay levels would not be deemed to constitute direct age discrimination (even if they may give rise to issues of indirect age discrimination). However, factors that are essentially ‘age proxies’, that is, which are directly linked and act as a proxy for a person’s age, will be deemed to constitute direct discrimination, as age will actually be a ‘material factor’ in the decision-making process. [36]

It may be important to keep age-based (i.e. where the treatment would not have occurred ‘but for’ a person’s age) and age-linked (such as seniority) factors distinct. The former may involve direct discrimination claims; the latter indirect discrimination. While both types of discrimination can be justified under the Regulations, how the test may be applied may vary: it may be likely, for example, that indirect discrimination based on age-linked factors will be easier to justify than direct discrimination on the grounds of age.

Many different types of criteria used in employment-linked decisions may have some sort of disparate impact upon one or more age groups. Experience, ‘know-how’, educational qualifications, decision-making capabilities, and emotional maturity are all neutral criteria that might put persons of a particular age at a disadvantage. [37] However, commonsense application of the objective justification test should be capable of distinguishing stereotyping use of age-linked criteria from legitimate use in making rational economic decisions. [38]

The issue of when age discrimination will be objectively justified nevertheless remains the central legal issue that distinguishes age discrimination legislation. The wording of the objective justification defence used in Reg. 3 for both direct and indirect discrimination has been criticised by Age Concern and others as introducing a weak and uncertain test of objective justification, and for not replicating the apparently stronger wording used in the Directive itself. However, the government’s explanatory notes that accompanied the text of the draft Regulations stated that the ‘legitimate aim’ test used in the Regulations was intended to correspond with the well-established ‘real need’ test in EC sex equality law, [39] and the ‘proportionate means’ stage of the same test corresponded with the ‘appropriate and necessary means’ test set out in Article 2.2(b)(i). [40]

Therefore, the objective justification test in the Regulations should be interpreted in line with how the test is applied across the other areas of anti-discrimination law: this means that employers attempting to rely on the defence will have to show a) the existence of a legitimate aim, and that b) proportionate means where used to achieve this aim.

Any employer attempting to satisfy the ‘objective justification’ test will therefore have to first show the existence of a legitimate aim. Neither the Directive nor the Regulations fix what aims will be deemed legitimate. [41] Article 6(1) of the Directive lists certain examples of potential legitimate aims, namely ‘legitimate employment policy, labour market and vocational training objectives’. These examples are not exhaustive. Both UK consultation papers suggest that there are a wide variety of other potential legitimate aims, as long as they correspond to a real need on the part of the employer, including economic needs. [42]

Criticism by Eurolink and other age equality NGOs of the text of the Directive centred on the very broad and vague wording of the examples of legitimate objectives listed in Article 6(1), such as ‘legitimate employment policy’, labour and training objectives. Their concern was that including such open-ended and broad examples in the text might encourage a loose approach to the objective justification test on the part of governments and national courts. [43] The draft Regulations contained a similar and even perhaps more expansive list, which however did not appear in the final text.

It could be argued that giving employers a wide ambit in defining their aims makes sense, given the amount of age-linked factors that exist and the frequency with which their use may be entirely legitimate: looking for experience, for example, should clearly be a legitimate aim in most circumstances. However, applying the standard approach in assessing the legitimacy of an aim in sex discrimination and human rights cases, the aim cited to justify the use of an age distinction cannot be discriminatory in itself. [44]

Jonathan Swift has expressed strong concern that if this rule is applied, then employers will not be able to argue that it is legitimate for them to attempt to attract or market directly to a youthful audience or assert the legitimacy of similar aims, as these aims could be seen as age-linked. [45] However, again, it is important to maintain the distinction between age-based and age-linked factors. When an employer is trying to sell goods to younger people, then the aim in question is to sell goods to an audience that will buy them: if this audience happens to be predominantly young people, then that aim is still nevertheless legitimate, as the aim in question is in itself no way discriminatory. Attempting to attract female customers will similarly not constitute sex discrimination. There is nothing in the legitimate aims stage of the age objective justification test which has proved problematic in other jurisdictions. Swift’s concerns may be unfounded.

In addition to showing the existence of a legitimate aim for their use of an age-based distinction, employers will have to satisfy the test of proportionality by showing the use of age was ‘appropriate and necessary’. Age limits that are not linked to a clear justification may be very vulnerable, as will age limits which could be replaced with less restrictive methods of achieving the legitimate aim in question, such as the use of job specifications based on clear, objective job descriptions.

Some criticism has been directed at the list of examples of justified uses of age distinctions set out in Article 6 of the Directive, and in particular the suggestion that minimum age requirements can be justified to ensure minimum conditions of seniority, return on training and required experience. The fear has been expressed by Age Concern and others that the potentially permissive nature of these examples could mislead employers about how easy it is to demonstrate the existence of objective justification, and give the false impression of a ‘presumption’ that the practices cited will be lawful. This concern is legitimate: however, the practices contained in the list are examples, not exemptions. Employers will, if challenged, still have to show that in their circumstances the practice pursues a legitimate aim, and is an appropriate and necessary (or proportionate) means of achieving that aim (the test of objective justification).

As will probably be the case with the genuine occupational requirement (GOR) defence that the Regulations also permit, the use of age as a ‘proxy’ or stand-in for other characteristics (such as when a person’s age is used as a marker or proxy for their state o health) should only be usually permitted where individual assessment of each affected person’s specific circumstances and characteristics is impracticable. A person’s health maturity, ability to learn, experience, skill, willingness to work may often be ascertained by vetting procedures, individual assessments and good job specifications. In a Dutch case, a number of referees successfully challenged the age limits of 47 and 49 used by the Royal Dutch Football Association (KNVB) on the basis that an individual assessment of each referee’s capability was entirely possible, and it was a breach of the proportionality principle to set a fixed age limit. [46] Even where age can be statistically linked to trends, as with increased ill health, the individual deviations from the statistical trend are so great that the use of age can constitute a very inadequate proxy for any given individual.^[47]

Age limits may be necessary in particular industries to ensure a ‘turnover’ of workers and to encourage recruits into a profession: the Dutch Supreme Court has upheld the imposition of a compulsory retirement age upon airline pilots for this reason. [48] However, the use of age limits that are intended to simply shift the age profile of the company or which unreasonably narrow the age spread of new recruitees may face great difficulties in showing objective justification. Sargeant has suggested that this might conflict with some of the government’s use of ‘age diversity’ rhetoric. [49]

How the objective justification test will be applied in practice may very much depend on what view is taken by courts and tribunals of the importance and status of age discrimination as an equality ground. This general objective justification test could be applied by adopting the standard proportionality requirement developed and applied in EC and UK indirect sex discrimination case-law, whereby an employer would have to show the existence of a clear legitimate aim and that the use of age distinctions was reasonably necessary and clearly proportionate to achieve this aim. Alternatively, a looser ‘rational’ standard could be applied, similar to the unfair dismissal standard and controversially adopted in the context of the justification defence for ‘less favourable treatment’ under the Disability Discrimination Act 1995. [50]

It is in applying this test that the question of how ‘serious’ is age discrimination arises again: how much leeway should be given to employers to use age distinctions and to argue that these distinctions are in fact objectively justified? Jonathan Swift has suggested that there is a lack of consensus or clarity as to how this test will be applied in practice, which stems from a fundamental lack of social consensus about how seriously age discrimination should be treated. Given the lack of this consensus, he suggests that the looser standard used in the disability discrimination context might be more appropriate. [51]

Courts in various countries have indeed often chosen not to apply a strict proportionality analysis in human rights cases challenging the use of age distinctions, often classifying age discrimination as a less problematic form of inequality than other forms of discrimination. [52] Courts and commentators have frequently suggested older people as a group have in general neither suffered historical discrimination nor a lack of political power, and therefore do not constitute a disadvantaged group.^[53] For example, the Canadian Supreme Court in applying the Canadian Charter’s s. 15 guarantee of equal treatment in McKinney v University of Guelph [54] applied a ‘rational’ standard of analysis, in contrast to the ‘strict’ standard required under race and sex discrimination cases, and concluding that the University of Guelph had a ‘reasonable basis’ for their decision that mandatory retirement impaired the right to equality as little as possible given their legitimate objective of ensuring staff renewal.

Similarly, the House of Lords in Reynolds rejected an Article 14 ECHR challenge to the lower rates of jobseekers’ allowance and income support paid to those under 25, on the basis that the age differentiation could be justified due to those under 25 having in general lower living costs than older persons, as many were still living with their parents. [55] While this generalisation did not necessarily apply to many of those unemployed and under 25, the Law Lords were willing to accept that the Secretary of State’s arguments were sufficient to cross the lower justification threshold that exists for age.

In his leading judgment, Lord Hoffmann drew a distinction between forms of discrimination that potentially infringed upon fundamental dignity and equal respect for the individual, which would require very strong justification to survive a HRA challenge, and less objectionable forms of discrimination, where the ‘general public interest’ could be taken into account. He gave discrimination against older people as an example of a type of ‘borderline’ form of discrimination, which might belong to either category. [56] However, he considered the use of age limits in paying less income support to younger workers as clearly justifiable in the public interest, and presumably therefore falling into the second category of discrimination.

There are good arguments for being cautious in applying a strict justification standard in human rights case-law, in particular given how often age is used as a differentiating tool in social welfare payments and other forms of social regulation, as where age limits on voting, purchasing alcohol and drawing down a state pension exist. However, it should not be presumed that a loose justification test will be adopted in applying the Age Regulations. The US, Dutch, Canadian and Irish courts have all applied a stricter standard than that advocated by Swift in applying age discrimination legislation, as distinct from constitutional and human rights standards.

For example, the US Supreme Court in Western Airlines v Criswell adopted a variant of the ‘necessity’ proportionality test in interpreting the US age discrimination legislation. [57] Similarly, in applying the age discrimination provisions of the Ontario Human Rights Code, the Canadian Supreme Court in Ontario Human Rights Commission v. Etobicoke [58] held that the employer had to show that the age requirement in question was reasonably necessary, although the Court recognised that the employer need not show that this standard was justified in respect of every employee affected where this was ‘impractical’. [59]

Two questions have generally been asked by the Australian and Canadian courts in age cases: a) are the characteristics that are cited to justify the act of discrimination legitimate and justifiable grounds for distinguishing between two people, and b) is age an effective proxy for the relevant characteristics or a necessary differentiating tool for determining whether an individual possesses those characteristics?^[60] Much will depend on whether the European Court of Justice interprets the Directive in a similar manner, and in particular whether it carries over its strict approach to the proportionality test over from the sex discrimination context.

In its first decision on the interpretation of the Framework Equality Directive and on age discrimination, Mangold v Helm, [61] the ECJ applied the proportionality test in a rigorous manner. German labour law provided that fixed-term contracts could only be concluded with employees if it was of less than two years duration, and could only be renewed three times within the two year time-limit: otherwise, objective justification was required. However, to encourage greater employment of older workers, legislation was introduced permitting employers to conclude fixed-term contracts with workers over the age of 58 without these restrictions, and with workers over the age of 52 until December 2006. This was challenged as constituting a lowering of protection for older workers contrary to the commitment of the German government to take effective measures to implement the age discrimination provisions of the Directive by December 2006, and not in the interim to take action which would conflict with the goals and purpose of the Directive.

The ECJ agreed that the less favourable treatment afforded to older workers did constitute discriminatory treatment contrary to the German government’s commitment not to take steps that conflicted with the Directive’s requirements prior to its coming into force, and the general principle of equality recognised by EC law. The ECJ also rejected the German government’s argument that the special measures were justified by the goal of ensuring greater employment of older workers.

The ECJ therefore was not satisfied that the German government had discharged its obligation to show sufficient objective justification for the use of age distinctions, despite the greater leeway conventionally given to national governments in applying proportionality analysis in the context of national economic and social policy. This judgment indicates that the proportionality test may be applied with some rigour in age discrimination claims, although it remains to be seen how consistent the ECJ will prove in practice.

Similarly, in its consultations on the age regulations, the UK government appeared to assume that the proportionality test would and should be given this strict interpretation. In the consultation on the draft age regulations, it was stated that ‘the test of objective justification will not be an easy one to satisfy. The principle remains that different treatment on grounds of age will be unlawful: treating people differently on grounds of age will be possible but only exceptionally and only for good reasons’. [62] Concern still remains that UK courts and employment tribunals may veer between strict and looser applications of the test, and it may take time before a consistent standard is established. However, it is by no means certain that Swift’s argument supporting the application of a looser standard will in fact be adopted, or that it is necessary and desirable: the tendency is for national courts to adopt a tougher line, and the ECJ appears so far to be doing likewise, as evidenced in the Mangold decision. This may reflect the reality that sufficient social consensus and normative weight underlies the idea that age discrimination legislation is a necessary and useful anti-stereotyping measure, even if it may not necessarily be identical in structure, content and underlying normative justification to other forms of equality legislation.

Aside from the general objective justification defence, the Regulations also set out a number of specific areas where the use of age-based requirements by employers is automatically deemed to be justified. To summarise, the key ones are:

If an age-related criterion, provision or practice falls within the scope of one of these specific exceptions, then it will be deemed to be lawful: an employer will not have to demonstrate the existence of an objective justification. The government’s intention in establishing these specific exceptions was to provide greater certainty for employers by delineating areas where the use of age criteria would be automatically treated as justified and save employers the time and expense of showing the existence of objective justification. However, it should be remembered that these specific exceptions must comply with the provisions of the Directive and be themselves objectively justifiable as general exceptions.

Article 7 of the Directive allows states to maintain positive action measures to compensate or prevent disadvantage, meaning that the proportionate use of age distinctions to compensate for disadvantage will be deemed to be objectively justified, if such measures are ‘…introduced or adopted by a member state’. Article 6(1)(a) also recognises that the promotion of vocational integration and the protection of vulnerable groups is a legitimate aim, which chimes with the overall ethos and orientation of the Directive.

The Regulations have however taken a narrow approach to positive action, which parallels the highly restrictive approach adopted for other equality grounds.

29.—(1) Nothing in Part 2 or 3 shall render unlawful any act done in or in connection with—

(a) affording persons of a particular age or age group access to facilities for training which would help fit them for particular work; or

(b) encouraging persons of a particular age or age group to take advantage of opportunities for doing particular work;

where it reasonably appears to the person doing the act that it prevents or compensates for disadvantages linked to age suffered by persons of that age or age group doing that work or likely to take up that work.

Therefore, anything done in connection with giving persons of a particular age access to vocational training or encouraging persons of a particular age to use employment opportunities will be lawful, if reasonably expected to prevent or compensate for disadvantages suffered by such persons. Similar exceptions exist for trade organisations.

Satisfying the requirements necessary to come within this positive action exception will be easier for employers than jumping the hurdle of the objective justification test, as the legitimate aim of such measures is already approved by the regulation and in place of the requirement to show objective justification, it is only necessary to show that it reasonably appears that the measure pursues the aim of preventing or compensating for such disadvantages. [63] However, this exception will only apply to discrimination in two narrowly circumscribed situations: giving access to vocational training facilities and encouraging persons to take advantage of employment opportunities. This resembles the limited positive action provisions in the sex and race discrimination legislation, and any wider forms of positive action can only be justified through the general test of objective justification. [64]

As these exceptions are narrow in scope, being essentially confined to training and encouragement, their usefulness will be limited. The government consultation on the draft regulations suggested that the positive action defence would be less significant than it could be for other equality grounds, due to the existence of the general objective justification defence. [65] This is a valid point, but the scope of the defence may prove uncomfortably restrictive.

The position is more complex when employers implement practices which involve differentiation on the grounds of age and lie outside the scope of the positive action exception, but are designed to benefit older workers, especially those nearing retirement. This type of practice can include pre-retirement courses, extra payments, extra holidays and adjustments to work loads. Much will depend on the nature of the benefits in question, and whether they are designed proportionally: ‘preparing for retirement’ should not be allowed to become a legitimate aim that could justify any preferential treatment for older workers. However, it would be also problematic if an excessive concentration upon achieving formal equality were to prevent employers introducing special provisions for older workers moving towards retirement.

While the Canadian courts have tended to treat these policies as justified, [66] the Dutch Equal Treatment Commission have taken the opposite view: in Case 2004/150 of 15 November 2004, the Commission held that an employer’s policy of gradually reducing working hours and granting extra holiday time for employees of 57 years age or older was not objectively justified. [67] This decision was based upon a strict application of the objective justification test, but it can be queried: should the provision for time-off for older workers to prepare for retirement be recognised as a legitimate positive act? Is the need to maintain age-neutral policies really so great as to require that such pre-retirement policies be jettisoned?

As already discussed above, the use of age limits as tools of government policy is common: they are often used as a cut-off point to determine beneficiaries of particular policies, or to identify vulnerable groups who are to be excluded from a particular activity or subject to special requirements. The use of age limits in this way may deprive some individuals of benefits or support on the grounds of age alone: however, given the reality of resource limitations, public programmes may have to be able to be targeted at specific elements of the population to achieve maximum impact, where there is no other reasonable method of targeting the appropriate group. Therefore the use of age distinctions in employment policy could be found to be objectively justified, unless the proportionality of the measure vis-à-vis the social harm it is designed to address cannot survive close scrutiny, as was the case in Mangold. [68]

For example, some ‘New Deal’ programmes have been restricted to the under-26s and over-50s, and upper age limits have been placed upon apprenticeships, student loans, or access to Jobcentre Plus and New Deal advice and training. The use of age limits in these circumstances could be replaced by more nuanced ways of identifying who might be the most appropriate beneficiaries of this special support: however, these alternative methods might be deemed to be excessively time-consuming and resource intensive.

Similar arguments could apply when age limits are used to protect disadvantaged groups, as with the work restrictions imposed on younger persons by the Young Persons’ Directive 94/33/EC: in the absence of any way of assessing each and every person’s competency and maturity, age limits may be objectively necessary as they set a definite cut-off point that can be effectively enforced across an industry or across the country.

Regulation 27 exempts anything done under statutory authority, so any age limits provided for in legislation will remain in force and can only be challenged via a reference to the ECJ. [69] Another exception exists in Regulation 31 for the age disadvantages faced by younger workers in several national minimum wage schemes, where they are often entitled to a much lower rate, or else completely excluded. [70] The exception allows employers to pay a lower minimum wage rate to those in young age bands (those aged 16-17 or those aged 18-21) than that paid to other workers, provided that no age differentiation is maintained within these bands.

The UK government has argued that this differential treatment can be justified as necessary to promote the interests of younger workers, who might otherwise face exclusion from the labour market due to their relative cost when compared to their lack of experience and training needs. Hepple suggests that the UK government has partially based its position on the advice of the UK Low Pay Commission, which would go some way towards showing justification. [71] In addition, challenges to similar age restrictions in Canada have been dismissed by the Canadian Supreme Court, [72] and by the House of Lords in the HRA decision of Reynolds (see above). However, the less favourable treatment of younger workers in this context remains controversial, based as it is on the assumptions discussed in Reynolds that younger workers have parental resources to fall back upon, and that employers need special incentives to hire younger workers. [73] A challenge to this exception may now be likely in the wake of Mangold.

Great uncertainty surrounds the question of retirement ages and the extent to which the Directive’s age discrimination provisions require existing national practices to be altered. This debate is ongoing across the globe, and is emerging as one of the crucial issues of law and policy in the UK. This section can only give a brief account of the key legal issues, and how they relate to the age discrimination rules.

It is important to distinguish between pensionable ages, state-imposed retirement ages, and contractual or employer-imposed retirement ages. Pensionable age is the age set by a member state at which individuals become entitled to a state pension (as distinct from the age at which individuals retire from work). Article 3(3) specifically excludes state social security systems from the scope of the Directive, and therefore age-based state pension rules are exempt from its scope. Article 6(2) also permits the use of age criteria for admission or entitlement to occupational social security systems including retirement benefits: thus, member states can choose to exempt the use of pensionable ages in occupational pension arrangements (see below).

However, the key issue for many member states concerns employer-set retirement ages, that is, whether employers are entitled to set retirement ages by contract, collective bargaining or unilaterally. [74] Any requirement now imposed by an employer that an employee must retire at a specified age amounts to less favourable treatment on grounds of age and will therefore be contrary to the Directive’s provisions unless objectively justifiable under Article 6, or permitted by national legislation which in turn can satisfy the proportionality requirements.

Previously in the UK, employers could have been able to set retirement ages because legislation deprived employees over that age of protection from unfair dismissal. [75] As denying employment rights to those above a particular age would appear to be now contrary to the Directive, the UK has altered its approach.

In the consultation paper Age Matters, the UK government considered two options: the introduction of a national default retirement age of 70, with the effect that employers would only have to provide objective justification for any retirement ages below the national default retirement age, or the abandonment of any attempt to retain a national default retirement age, with all employers having to individually provide objective justification for the introduction of any retirement age. After a mixed response to this proposal, Coming of Age opted for the ‘national default retirement age’ model, but with the age set at sixty-five; this will have the effect of permitting employers to dismiss employees when they reach sixty-five without having to demonstrate objective justification. [76]

30(2) Nothing in Part 2 or 3 shall render unlawful the dismissal of a person to whom this regulation applies at or over the age of 65 where the reason for the dismissal is retirement.

To supplement this, the current upper age limit on unfair dismissal or redundancy claims will be removed, meaning that employees kept on after 65 will have full employment rights (see below): however, keeping employees on past the national default retirement age will not prejudice employers' rights to retire employees at or after the retirement age. Section 98 of the Employment Rights Act 1996 is amended to add another fair reason for dismissal, ‘retirement of the employee’.

As compensation for the retention of mandatory retirement ages, employers will be subject to a duty to consider requests from employees to stay on beyond the fixed retirement age, who is obliged to give reasonable consideration to requests to work on, similar to their obligations in respect of part-time workers. [77] This however remains a mere duty to ‘consider’, and employers do not have to give reasons. Employers are also required to inform employees of their right to request a continuation of their employment contract at least six months before its termination. A failure to follow this procedure, or a premature termination of a contract, can be remedied through unfair dismissal. [78]

If employers wish to impose a mandatory retirement age before the sixty-five-year limit, they will have to be able to show that this lower retirement age satisfies the general test of objective justification. The particular arguments in favour of the use of a mandatory age that are presented by a specific employer will have to measured against the proportionality standard, and the outcome will be a matter of case by case analysis.

The retention of the ability of employers to dismiss employees at 65 has generated considerable controversy. Heyday, backed by Age Concern, have announced that they will seek a judicial review of the Regulations, arguing that the scope of the current exemption is disproportionate in that it fails to incorporate an objective justification test which employers should be required to satisfy before they can avail of mandatory retirement policies. [79]

The justification offered by government is that significant numbers of employers use a set retirement age as a necessary part of their workforce planning. [80] Also, the government has suggested that at present, if employers only had the option of individually justified retirement ages, this could risk the possibility that employers would reduce or remove work-related benefits they offer to employees to offset any increase in costs. [81] The government has also announced that it will review whether it is still appropriate to have a default retirement age in 2011 (five years after the Age Regulations come into force).

Coming of Age states that the decision to keep or abolish the default retirement age will focus mainly on two factors: whether, in the light of the evidence, the default retirement age remains appropriate and necessary to facilitate workforce planning and to avoid adverse effects on pensions and other employment benefits; and the influence of any other social policy objectives. [82] The government also has indicated that the six month notice time limit should help to encourage ‘planned retirement’, and perhaps provide time for employees to persuade employers to reconsider. Presumably, there is also the hope that it may generate a culture change.

Nevertheless, the decision to retain a form of mandatory retirement remains controversial. Arguments do exist that as a practice, it may not be capable of being objectively justified in many contexts. It may be possible to justify certain exceptions in the context of specific forms of employment: [83] however, it may be contestable whether it can be justified across the labour force as a whole. [84]

The US experience indicates that prohibiting mandatory retirement need not generate substantial negative effects, if only because the number of people who choose to postpone their retirement is likely to be limited, and those who stay beyond the normal retirement date for their employment will tend to do so for only a few years. Employers have become used to accommodate people leaving earlier than expected, due to increased workforce mobility and voluntary early retirement: the accommodating of a small number who choose to remain in employment longer may therefore not pose a considerable difficulty. Meadows suggests that ‘the effect on our national productive potential of any changes in mandatory retirement arrangements is likely to be very small’. [85] The relevant evidence is less than certain (as with so many aspects of the mandatory retirement issue). [86] However, the retention of the ability of employers to have mandatory retirement ages may yet prove to be yet another example of government reluctance to take an imaginative leap in the realm of anti-discrimination and equality law, and to recognise the reality of changing social expectations.

The denial or restriction of employment rights to workers above a particular age also raises serious questions under the Directive, as this will constitute direct age discrimination that again requires objective justification. The UK government has come to a similar conclusion, removing in the Regulations the previous statutory provisions that remove the right to protection from unfair dismissal upon reaching 65, or the normal retirement age for the post in question. [87] The Regulations also removes the lower age limit of 18 and the upper limit of 65 for entitlement to redundancy payments. [88]

Complex issues arose in the course of the implementation process about redundancy payments. Member states have historically provided for special benefits for older employees in redundancy decisions, as does the UK at present. [89] The UK government initially considered removing the current weighting of redundancy payments in favour of older workers, but subsequently decided that it would be proportionate to retain higher payments for those nearing retirement, as their future earning potential would in general be less than other workers. [90] This decision could be seen as preventing ‘levelling-down’, whereby the weighing in favour of older workers may have been simply disallowed. However, the Government’s case could again be critiqued, especially as many older workers may be in a much better financial position than younger worker, and therefore the favourable treatment of older workers might not be capable of being objectively justified. [91]

Length of service is also used to calculate redundancy payments, with minimum length of service requirements to qualify for eligibility: these elements could amount to indirect discrimination. However, the UK government has retained the use of length of service in computing these payments, on the basis that it is objectively justified as ensuring that compensation rewards past loyalty and commitment of the employee. [92] This raises the issue of seniority, and when its use will be justifiable as a method of distinguishing between employees.

Distinctions based on seniority (length of time of service), such as ‘last in first out’ redundancy selection and pay scales which vary according to length of service, are common. However, they may be indirectly discriminatory, as they will often disadvantage younger workers, unless they can be objectively justified. Therefore, variations on the ground of seniority may have to be objectively justified by reference to the specific experience gained by actual work in the job in question.

There are good reasons for maintaining the ability to make distinctions upon the grounds of seniority. [93] Rewarding loyalty and encouraging experienced employees to stay with an employer may be justifiable if they are required for good business reasons. Article 6(1)(b) specifically refers to the acceptability of justification of minimum conditions of professional experience or seniority in service for access to employment or employment-related advantages.

Other states have been content to assume that seniority distinctions are inherently justifiable, irrespective of the exact legitimate aim sought: others have introduced ‘seniority’ exceptions into their new age discrimination legislation. For example, the Irish Equality Act 2004 now allows an employer to fix differential rates of severance payment based on seniority, and there are also specific exceptions for differences based on seniority in relation to remuneration or to conditions of employment. However, in the Irish case of McGarr v Dept of Finance, [94] the Equality Officer held that the express seniority exception in the Irish legislation must be strictly interpreted and could not be extended to permitting seniority requirements for promotion and for special payments which were not objectively justified or specifically exempted by legislation. Similarly, in a recent Dutch case, a pension regulation which remunerated on the basis of a length of service requirement was held to constitute unjustified indirect discrimination. [95] These initial cases clearly demonstrate that seniority-based distinctions may be more vulnerable to challenge than is usually presumed, especially if they reflect an unquestioned assumption that longer service should be matched by greater rewards.

The use of seniority to encourage and reward loyalty was referred to in Age Matters, which proposed to ‘make specific provision for employers to be able to justify seniority conditions by reference to the aims…[of] encouraging and rewarding loyalty.’ [96] The consultation revealed a considerable degree of consensus among respondents that loyalty-linked employee benefits were desirable and should be maintained. [97] To avoid the danger of employers having to justify each use of seniority criteria on a case-by-case basis, a general exemption for service rewards was built into the Regulations, along with two other more specific exceptions.

Employment Equality (Age) Regulations 2006 - Exception for Provision of Certain Benefits Based on Length of Service

32.—(1) Subject to paragraph (2), nothing in Part 2 or 3 shall render it unlawful for a person (“A”), in relation to the award of any benefit by him, to put a worker (“B”) at a disadvantage when compared with another worker (“C”), if and to the extent that the disadvantage suffered by B is because B’s length of service is less than that of C.

(2) Where B’s length of service exceeds 5 years, it must reasonably appear to A that the way in which he uses the criterion of length of service, in relation to the award in respect of which B is put at a disadvantage, fulfils a business need of his undertaking (for example, by encouraging the loyalty or motivation, or rewarding the experience, of some or all of his workers).

(3) In calculating a worker’s length of service for these purposes, A shall calculate—

(a) the length of time the worker has been working for him doing work which he reasonably considers to be at or above a particular level (assessed by reference to the demands made on the worker, for example, in terms of effort, skills and decision making); or

and on each occasion on which he decides to use the criterion of length of service in relation to the award of a benefit to workers, it is for him to decide which of these definitions to use to calculate their lengths of service.

The use of length of service criterion for awarding or increasing benefits during the first five years of service is deemed to be clearly justified, and a complete and automatic exemption will apply: the employer need show nothing more. In contrast, making use of length of service requirements which are longer than five years may still be justified, but will not be automatically so: Reg. 33 (3) sets out conditions to be fulfilled for this general exemption to apply:

a) awarding or increasing the benefit is meant to reflect the higher level of experience of the employee, or to reward loyalty, or to increase or maintain motivation of the employee; and

b) the employer has concluded that there will be a business benefit resulting from the achievement of these aims; and

c) the employer applies the length of service criterion similarly to staff in similar situations.

These conditions appear to be easier to satisfy than the full objective justification test. The employer does not have to show the existence of objective justification, but just to show that the use of the length of service criterion was done for a legitimate aim, applied consistently and was deemed necessary to achieve a ‘business benefit’. However, these criteria still require the employer to justify the aim and effect of the benefit.

Both of these service exemptions are intended to distinguish between the use of legitimate ‘loyalty reward’ schemes, which may often be appropriate and necessary means to achieve the legitimate aim of encouraging and rewarding loyalty, and the widespread use of seniority systems, which will often not be justifiable. However, some concern has been expressed that the general length of service exemption will legitimise the widespread use of length of service criteria, especially as the employer does not have to show that the use of the length of service condition is objectively justified. Concern has also been expressed that the five year exemption of any length of service requirement may provide employers with too much leeway: five years is a considerable period of time in the contemporary workplace, and this time limit seems to be potentially disproportionate.

It should be noted that length of service requirements may fall foul of the prohibition on indirect sex and race discrimination in certain circumstances. [98] In addition, it will be interesting to see if the ECJ applies the same approach to seniority-based criteria as it did in the recent sex discrimination case of Cadman v Health and Safety Executive. [99] In Cadman, the ECJ has found that an employer could in general use length of service as a reason for pay differences between men and women doing work of equal value, as seniority and experience will generally improve the performance of a particular job and therefore the use of these criteria would be usually deemed to be automatically objectively justified. Therefore, there was no need in usual circumstances to show that the use of seniority criteria was actually required for the performance of a particular job. However, if an applicant can raise serious doubts that the use of length of service criteria is appropriate and necessary in the circumstances of a specific post or type of work, then the employer may have to show that its use can be objectively justified in those particular circumstances. This approach could be readily transplanted to the context of age, and the ECJ may in determining age discrimination challenges to the use of seniority ‘carry over’ Cadman and apply it in the age context.

Article 3(3) of the Directive provides that it “does not apply to payments of any kind made by state schemes or similar, including state social security or social protection schemes”. As previously discussed, this may exempt all state social security schemes from the scope of the Directive. On the other hand, employment-related occupational pensions and insurance are within the scope of the Directive. However, Article 6(2) of the Directive provides that:

‘Member States may provide that ‘the fixing for occupational social security schemes of ages for admission or entitlement to retirement or invalidity benefits, including the fixing under those schemes of different ages for employees or groups or categories of employees, and the use, in the context of such schemes, of age criteria in actuarial calculations, does not constitute discrimination on the grounds of age, provided this does not result in discrimination on the grounds of sex.’

In other words, member states can choose to exempt the use by occupational schemes of age distinctions governing admission or entitlement to retirement or invalidity benefits and the use of age criteria in actuarial calculations under such schemes. This was done to allow occupational insurers and pension-providers to continue to use age-based criteria in offering entry terms, regulating costs across different age groups and in assessing premiums. Member states must actively invoke the exception. Most have done so.

The UK Regulations make it unlawful for trustees or managers of an occupational pension scheme, when carrying out their functions, to discriminate on grounds of age. However, certain age-related rules or practices in occupational pension schemes are exempted, and these are defined in a complex set of provisions in Schedule 2, parts 2 and 3 of the Regulations. These excpetions include the setting a minimum level of pensionable pay for admission; having a normal pension age in a pension scheme (i.e. the age at which a person normally becomes entitled to receive retirement benefits) and the use of age criteria in actuarial calculations. [100]

The provisions in the Age Regulations are more restrictive than those set out in the original draft regulations. This unexpected tightening of the pension provisions was probably driven by concern that the width of the exceptions in the original draft regulations went further than was permitted by Article 6(2), especially those that related to the use of age distinctions in paying out benefits, as distinct from their use in fixing who is entitled to particular occupational benefits. [101] Any exceptions still in the Regulations that lie outside the scope of Article 6(2) will have to be shown to be objectively justified under Article 6.1.

These new provisions inserted at the last moment into the final text of the Regulations have caused a considerable degree of uncertainty and consternation in the pensions industry. There have been differences of opinion between the government and some pension advisers as to the meaning of certain of these provisions, and in particular about whether employees who become entitled to a pension could in fact choose to work on while also collecting their pension. [102] As a consequence, the government has delayed implementation of these provisions until 1^st December 2006 (the absolute final deadline allowed under the Directive) and is considering revising the scope of the exceptions.

It should also be noted that the use of age distinctions in occupational schemes can still be challenged on the basis of sex discrimination, even if a member state has taken advantage of the exception. Given that the use of age-based criteria for admission and entitlement to benefits and of age-based actuarial criteria may give rise to serious issues of indirect gender discrimination, Article 6(2) provides at best a limited shelter from the scope of anti-discrimination law.

Malcolm Sargeant has suggested that ‘It appears that, perhaps partly to limit employer opposition and partly to encourage its diversity agenda, the Government has produced Regulations that effectively legitimise age discrimination in employment.’ [103] There are certainly aspects of the Regulations that can be criticised on this basis. However, the Government deserves some credit for giving some serious consideration to the issue of age discrimination, and assumed the existence of a reasonably strict objective justification requirement: the Regulations appear more thought-through than virtually every other piece of equivalent legislation across Europe. Nevertheless, it remains to be whether elements of the Regulations will require adjustment to meet the standards of the ECJ.

It is possible to worry too much about age discrimination. Experience from the USA, Canada, Australia and Ireland tends to confirm that the introduction of age discrimination legislation rarely generates substantial chaos and legal uncertainty, especially if employers take seriously the possibility that a strict standard of objective justification will be applied. However, it should be emphasised as a final point that medical, social and economic trends have considerably altered or called into doubt many of the presumptions that could arguably once have justified wholescale age discrimination. Critical and ongoing scrutiny of age inequalities, affecting young and old, is on the increase. Jonathan Swift may be correct to say that there is no clear consensus as yet on the extent to which age discrimination is unacceptable. However, the Mangold decision, experience from elsewhere and the underlying logic underpinning the prohibition on age discrimination tends to suggest that a tough standard of objective justification may be desirable, and may also prove to be inevitable.

(For additional reading on the Age Regulations, see in particular Malcolm Sargeant, ‘The Employment Equality (Age) Regulations 2006: A Legitimisation of Age Discrimination in Employment’ (2006) 35 ILJ 209-227; Jonathan Swift, ‘Justifying Age Discrimination’ [2006] 35 ILJ 232; Sandra Fredman, ‘The Age of Equality’, in S. Fredman and S. Spencer, Age as an Equality Issue: Legal and Policy Perspectives (Oxford: Hart, 2003) 21-70; M. Rubinstein, ‘Age Regulations 2006’ (2006) EOR 152, 17.)

[1] Special provision is made for age in the implementation requirements imposed by the Directive. Its general provisions had to be implemented by member states by 2 December 2003, but Article 18 permitted a member state to wait a further three years (to 2 December 2006) before implementing the provisions concerning age discrimination, ‘in order to take account of particular conditions.’ The UK took advantage of this additional time period.

[2] By way of illustration, it took thirty-eight years for the US Supreme Court to confirm in Smith v City of Jackson that the Age Discrimination in Employment Act (ADEA) 1967 permitted recovery for disparate impact discrimination. The Supreme Court based their decision upon the statutory language of the Act and its similarity to the equivalent provisions of the Civil Rights Act, an acknowledgement that the ADEA was broadly equivalent to other forms of anti-discrimination legislation. For a taste of the extensive academic debate that lead up to this decision, see E. H. Pontz, ‘Note - What a Difference ADEA Makes: Why Disparate Impact Theory Should Not Apply to the Age Discrimination in Employment Act’ (1995) 74 N.C. L. Rev. 267; M. Ziegler, ‘Note: Disparate Impact Analysis and the Age Discrimination in Employment Act’ (1984) 68 Minn. L. Rev. 1038; S. J. Kaminshine, ‘The Cost of Older Workers, Disparate Impact, and the Age Discrimination in Employment Act’ (1990) 42 Fla. L. Rev. 229; P. H. Harris, ‘Note, Age Discrimination, Wages and Economics: What Judicial Standard?’ (1990) 13 Harv. J.L. & Pub. Policy 715, 729-30; P. S. Krop, ‘Note, Age Discrimination and the Disparate Impact Doctrine’ (1982) 34 Stan. L. Rev. 837.

[3] This proportion has stayed fairly constant since 2003: see the Annual Reports of the ODEI/The Equality Tribunal, 2000-2003. Initial experience from the Netherlands and Belgium confirms this trend. see Age Concern, Addressing Age Barriers (London: Age Concern, 2004).

[4] See www.eeoc.org (last accessed 5 August 2006). The ADEA combines specific exemptions for particular professions and seniority systems with a general prohibition on ‘arbitrary discrimination’, leaving it to the courts to determine when age discrimination outside the specified exceptions will be deemed not to be based on objective justification. It also only protects those over 40, a significant difference from the requirements of EC law.

[5] See T. Buck and B. Fitzpatrick, ‘Age Discrimination in Employment: legal protection in the United States and in the United Kingdom’ [1987] Anglo-American LR 192

[6] Wall v British Compressed Air Society [2003] EWCA Civ 1762

[7] See Taylor v Secretary of State for Scotland [2000] UKHL 28

[9] See O. Fiss, ‘The Fate of an Idea Whose Time Has Come: Anti-Discrimination Law in the Second Decade after Brown v. Board of Education’ (1974) 41 University of Chicago Law Review 742.

[10] One point to note is that discrimination on ‘the grounds of’ a person’s age is only prohibited when it affects that particular person, unlike the case in the areas of race, gender, religion or belief and sexual orientation discrimination, where discrimination on the grounds of association is also prohibited. As with the similar gap in the disability discrimination legislation, the proposed reference to the European Court of Justice in the case of Corrigan v Attridge Law, Case No. 2303745/2005, Croydon Employment Tribunal, 6^th July 2006 may result in a finding that the Regulations requiring amendment to remedy this.

[11] Jonathan Swift, ‘Justifying Age Discrimination’ [2006] 35 ILJ 232

[12] If interested, see C. O’Cinneide, “Comparative European Perspectives on Age Discrimination Legislation”, in S. Fredman and S. Spencer, Age as an Equality Issue (Oxford: Hart Publishing, 2003), 195-218.

[13] See Discrimination and Age: Human Rights Issues Facing Older Persons in Ontario (Ontario Human Rights Commission, 2000), available at

[14] Note however that older persons can be subject to appalling and degrading treatment in care homes, in residential accommodation and in everyday life, based often on a denial of their basic human entitlements: see J. Watson, Something for Everyone: The Impact of the Human Rights Act and the Need for a Human Rights Commission (London: British Institute for Human Rights, 2001), 46-50.

[15] See S. Fredman, ‘The Age of Equality’, in S. Fredman and S. Spencer, Age as an Equality Issue: Legal and Policy Perspectives (Oxford: Hart, 2003) 21-70, especially 22-35. See also Australian Human Rights Commission, Age Matters: A Report on Age Discrimination (Canberra: Australian Human Rights and Equal Opportunity Commission, May 2000).

[16] Clare McGlynn has argued that evidence in the United States shows that there may be a tension between age discrimination legislation and race and gender discrimination laws, at least in terms of their effects. See C. McGlynn, ‘Age Discrimination and European Union Law’, paper delivered to a workshop organised by the Swedish National Institute for Working Life, Brussels, 6-7 November 2000, p. 11-12.

[17] See Rutherglen, ‘From Race to Age: The Expanding Scope of Employment Discrimination Law’, at 495: ‘Most claims of employment discrimination are now claims of discriminatory discharge. Litigation under the ADEA, which concerns such claims almost exclusively, exemplifies this trend in its most extreme form.’

[18] Thus, for example, the US age discrimination legislation does not apply to firms employing fewer than 20 people or to people under 40, and it only includes direct discrimination. Its creation of a special and privileged protected class by its protection being confined to those forty and above has attracted criticism, as it involves the use of the type of arbitrary age limit the legislation was designed to discourage. and has created difficulties as to when discrimination against a member of this group can be inferred.

[19] See S. Fredman, ‘The Age of Equality’, in S. Fredman and S. Spencer, Age as an Equality Issue: Legal and Policy Perspectives (Oxford: Hart, 2003) 21-70, especially 22-35. See also Australian Human Rights Commission, Age Matters: A Report on Age Discrimination (Canberra: Australian Human Rights and Equal Opportunity Commission, May 2000).

[20] E. Pontz, ‘What a Difference ADEA Makes: Why Disparate Impact Theory Should Not Apply to the Age Discrimination in Employment Act’ (1995) 74 North Carolina L. Rev. 267.

[21] See Posner, Aging and Old Age (University of Chicago Press: Chicago, 1997).

[22] See P. Meadows, Retirement Ages in the UK: A Review of the Literature, Employment Relations Research Series No.18 (London: DTI, 2003), 18. In addition, gains in productivity and performance due to the experience of older workers, their interpersonal skills and relationships, and their embodiment of institutional memory, often offset any adverse effects related to aging. See Meadows, p. 18.

[23] See L. Freedman, ‘Age Discrimination Law: Some Remarks on the American Experience’, in S. Fredman and S. Spencer, Age as an Equality Issue (Oxford: Hart, 2003), 175-194.

[24] . See Australian Human Rights and Equal Opportunity Commission, Age Matters: A Report on Age Discrimination (Canberra: Australian Human Rights and Equal Opportunity Commission, May 2000).

[25] See C. Jolls, ‘Hands-Tying and the Age Discrimination in Employment Act’, (1996) 74 Tex. L. Rev. 1813; Issacharoff and Harris suggest unfair dismissal protection would cover this, but without the rigour of anti-discrimination law standards, employees may find it difficult to challenge employer decisions based on ‘rational’ cost-cutting requirements. See also S. J. Schwab, ‘Life-Cycle Justice: Accommodating Just Cause and Employment at Will’ (1993) 92 Mich. L. Rev. 8; J. H. Verkerke, ‘An Empirical Perspective on Indefinite Term Employment Contracts: Resolving the Just Cause Debate’ (1995) Wis. L. Rev. 837.

[26] E. P. Lazear, ‘Why is there mandatory retirement?’ (1979) 87 (6) Journal of Political Economy 1261-1284. See also R. Hutchens, ‘Delayed Payment Contracts and a Firm's Propensity to Fire Older Workers’ (1986) 4 J. Lab. Econ. 439; J. L. Medoff & K. G. Abraham, ‘Are Those Paid More Really More Productive? The Case of Experience’ (1981) 16 Journal of Human Resources 186, both cited by Issacharoff and Harris.

[27] As Meadows has argued, few older workers remain in jobs that requiring physical strength or rapid processing of new information, where productivity performance does deteriorate with age, while data suggests that in Britain at least, there appears to be little if any difference between the productivity of older and younger workers in most jobs. Meadows, ibid., pp. 27-28. Meadows also notes that Lazear subsequently acknowledged that the restriction of mandatory retirement in the US had not had the effect that he had expected. See Meadows, p. 17.

[28] This could arise when assessing the qualities of each relevant individual is not possible, or would be excessively costly or wasteful of resources in the circumstances.

[29] For example, an airline may want to argue that it cannot test all of its pilots annually beyond a particular age as there may be a risk that eyesight deterioration may not be picked up, and it is safer and more cost-effective to set a fixed retirement age at the point when statistical means indicate that significant sight deterioration begins.

[30] See the excellent and informative discussion of the use of hypothetical comparators in the Irish age equality case-law in M. Reid, ‘Age Discrimination in Employment: Issues Arising in Practice’, lecture delivered at Academy of European Law, Trier, 2 October 2004, available at http://www.era.int/www/gen/f_19096_file_en.pdf (last accessed 17 January 2004).

[31] Questions of when the existence of direct discrimination can be inferred from particular facts have proved difficult and complex in ADEA litigation, where case-law requires a plaintiff to establish a ‘nexus’ that directly links the age of the complainant to the complained of conduct : see Laugesen v. Anaconda Co., 510 F.2d 307, 313 n.4 (6th Cir. 1975); Kelly v. American Standard, Inc., 640 F.2d 974, 980 (9th Cir. 1981) Lovelace v. Sherwin-Williams Co., 681 F.2d 230, 238-41). In O'Connor v. Consolidated Coin Caterers Corp., 116 S. Ct. 1307, 517 US 308 (1996), the US Supreme Court had to clarify that if the claimant was replaced by a person who was also within the age group protected by the US legislation (40 years and older), this could still give rise to an inference of age discrimination, but no automatic presumption would arise.

[32] See e.g. James v Eastleigh BC [1990] 2 AC 751

[33] See Hepple, ‘Age Discrimination in Employment: Implementing the Framework Directive 2000/78/EC’, in Fredman and Spencer, cited above, at 82.

[34] See Hazen Paper v Biggins 507 US 604 (1993)

[35] See I.P. Asscher-Vonk, “Towards One Concept of Objective Justification” in: T. Loenen and P.R. Rodrigues, Non-Discrimination Law – Comparative Perspectives (The Hague: Kluwer Law International, 1999) p. 39-51, at p. 43; see also M. Gijzen, Dutch Baseline Report for the European Network of Legal Experts in the Non-discrimination Field, Transposition of the Racial Equality Directive (Directive 2000/43) and the Framework Employment Directive (Directive 2000/78) into Dutch Law (Brussels: Migration Policy Group, December 2004).

[36] Hepple, ‘Age Discrimination in Employment: Implementing the Framework Directive 2000/78/EC’, at 82.

[37] For an example of this potential complexity, refusing to call ‘overqualified’ applicants for interview may indirectly discriminate against older workers: see the Irish case of Noonan v Accountancy Connections, DEC-E2004-042.

[38] Fredman, 58-59.

[39] Coming of Age, para. 4.1.16.

[40] Coming of Age, ara. 4.1.19.

[41] In Age Matters, the UK government canvassed opinion on whether it should list and definitely define what would constitute a legitimate aim. However, in Coming of Age, it was concluded that an exhaustive list of legitimate aims for direct discrimination would be too restrictive and prescriptive for employers. See Coming of Age, para. 4.1.5.

[42] See Coming of Age, para. 4.1.16: it is also noted there that discrimination will not be justified merely because it may be more expensive not to discriminate, as per the standard approach in anti-discrimination law. Other examples set out in both papers of what may constitute a legitimate aim including health, welfare and safety concerns, including the protection of younger workers; the facilitation of employment planning; the particular training requirements of the post in question, such as the lengthy specialist and expensive training that air traffic controllers must undergo; the encouraging and rewarding of loyalty; and the need for a reasonable period of employment before retirement.’ See Age Matters, para. 3.15; Coming of Age, para. 4.1.17.

[43] See Eurolink evidence to the UK House of Lords Select Committee on the European Union, “EU Proposals to Combat Discrimination”, HL Paper 68, 16 May 2000. See also H. Meenan, ‘Age Equality after the Employment Directive’ (2003) 10 Maastricht Journal of European and Comparative Law 9-38; S. González Ortega, ‘La Discriminación por Razón de la Edad' (2001) 59 Temas Laborales pp. 93-124.

[44] See para. 4.1.18, Coming of Age.

[45] This aim could be cited as part of a justification defence if an employer argued, for example, that their sales staff had to be young in order to give an impression of vibrancy and general hipness.

[46] See Amsterdam Court of Appeal [Hof Amsterdam] 13-01-2000, JAR 2000, 42.

[47] The Canadian courts applied this approach in O’Brien v Ontario Hydro, where a 40-year old man was refused an apprenticeship: the Board of Inquiry ruled that the employer’s argument that age had relevancy when determining whether a person would adjust to particular job conditions was unjustified age discrimination (1981) 2 CHRR D/504.

[48] Dutch Supreme Court, 8 October 2004 - Nr. C03/077HR - 16 pilots v.Martinair Holland NV and the Vereniging van Nederlandse Verkeersvliegers this case concerned a challenge to a policy of compulsory retirement at the age of 56 of pilots employed with Martinair on the grounds of a breach of Article 1 of the Constitution and Article 26 of the International Covenant on Civil and Political Rights (ICCPR). (The case predated the coming into force of the Directive and the Dutch implementing legislation.) The rationale for retirement of pilots at 56 was to guarantee ‘circulation’, based upon the idea that any pilot’s career (starting with a very costly education and ending with early retirement) is structured in terms of the expectation that it will be possible for all pilots to reach the highest seniority level before retirement. Both the cantonal court and the district court ruled that this rationale formed an ‘objective justification’. The Supreme Court affirmed this and rejected the applicants’ claim. See also Dutch Supreme Court, 8 October 2004, Nr. C03/133HR, Applicant v.Koninklijke Luchtvaartmaatschappij NV (Royal Dutch Airlines) and the Association of Dutch Traffic Pilots.

[49] Malcolm Sargeant, ‘The Employment Equality (Age) Regulations 2006: A Legitimisation of Age Discrimination in Employment’ (2006) 35 Ind Law J 209-227.

[50] See Jones v Post Office [2001] EWCA Civ 558; [2001] IRLR 384, and the critique of this standard by J. Davies, ‘A Cuckoo in the Nest? A “Range of Reasonable Responses”, Justification and the Disability Discrimination Act 1995’ (2003) 32(3) ILJ 164. See however the different approach applied to the (subsequently removed) justification defence to reasonable accommodation in Collins v Royal National Theatre Board [2004] EWCA Civ 144.

[51] Jonathan Swift, ‘Justifying Age Discrimination’ [2006] 35 ILJ 232. It should be noted that Swift appears to presume that a clear consensus exists in respect of other areas of anti-discrimination law, which is very debatable.

[52] . See also In the Matter of Article 26 of the Constitution and in the Matter of the Employment Equality Bill, 1996 [1997] 2 IR 321(SC); Massachusetts Board of Retirement et al v Murgia 427 U.S.307 (SC).

[53] In Massachusetts Board of Retirement v. Murgia (1976) 438 US 285, the Supreme Court explained that ‘old age does not define a “discrete and insular’ group ... in need of ‘extraordinary protection from the majoritarian political process.’ Instead, it marks a stage that each of us will reach if we live out our normal span.’ George Rutherglen has similarly argued that the ADEA ‘cannot be justified in terms of opening opportunities to a historically disfavored group.’ See G. Rutherglen, ‘From Race to Age: The Expanding Scope of Employment Discrimination Law’ (1995) 24 J. Legal Studies 491.

[54] [1990] 3 S.C.R. 229. See also Law v Canada [1989] 1 SCR 143

[55] R v Secretary of State for the Home Department, ex p. Reynolds [2005] UKHL 37.

PREVIOUS SPEAKERS:

Colm O’Cinneide
Senior Lecturer in Law, UCL

Back to list of speakers

PREVIOUS SPEAKERS:

Colm O’Cinneide Senior Lecturer in Law, UCL

Back to list of speakers

Colm O’Cinneide
Senior Lecturer in Law, UCL