Debating ageism and pragmatism

The BSG (2020: 1) issued its statement on COVID-19 on 20 March 2020, in which it ‘urge[d] the Government to reject the formulation and implementation of policy based on the simple application of chronological age’. It argued that risk was distributed across the age spectrum, with many older people being relatively less vulnerable and many younger people being relatively more vulnerable. It also questioned the arbitrariness of drawing a boundary at 70, when the risks of two people aged 69 and 70, respectively, were likely comparable. In conclusion, the BSG (2020: 7) ‘reject[ed] firmly the ageist and stereotypical assumptions that underly public and policy pronouncements that rely solely on the application of chronological age’. It is this final claim in particular – that the stratification of people and development of policy in terms of chronological age is ageist – that has subsequently stimulated debate. It is important to recognise that this statement was prepared rapidly in the context of tumultuous events and was not intended for rigorous critical evaluation. Moreover, it contained various other important insights and recommendations which, had they been heeded, might well have lowered the UK's dire death toll. I use the statement here as an example of broader contradictions within notions of age and ageism that the COVID-19 crisis has opened up.

One of the more notable figures to counter the claim that chronological quarantine is ageist is former BGS president David Oliver. In a BMJ article published on 22 April, Oliver (Reference Oliver2020) accepted the diversity of health statuses within older age groups, but cautioned against the conflation of perceived fitness with invulnerability to the disease. From a medical perspective, he noted that ‘even fit older people show poorer immune responses than their younger selves in the face of infection’. Speaking directly to the issue of using chronological age to dictate COVID-19 policy, he argued that ‘the speed of the covid-19 pandemic doesn't allow us to assess each person over 70 for individual risk – and the government has made a pragmatic decision’. Here, where the BSG sees ageism, Oliver finds epidemiological practicality. This discrepancy echoes far broader ethical–legal debate regarding when discrimination becomes wrong (Hellman, Reference Hellman2008).

For Oliver, it is unfeasible to assess and communicate the risk profiles of individuals, and health policy must therefore take crude blanket approaches to entire populations to protect some members of that population from grave harms. This is common in public health. Many of the aforementioned women aged between 50 and 71 will have negligible personal risk of developing breast cancer, but at a population level it is considered epidemiologically justifiable to impose slight constraints on the lives of many to improve dramatically the lives of a few. In terms of COVID-19, the basic calculation is that imposing heavy restrictions on millions of older people is preferable to the preventable deaths of tens, if not hundreds, of thousands of people in that population. Other notable voices have weighed in on either side of the argument, including Baroness Ros Altmann (Reference Altmann2020) arguing that chronological quarantine is ageist and Professor Julian Savulescu (Savulescu and Cameron, Reference Savulescu and Cameron2020) arguing that it is not, and it continues to play out across traditional and social media with considerable passion.

The issue is not limited to the British context. For instance, Help Age International has criticised the government of Bosnia and Herzegovina for imposing a lockdown on people aged over 65, claiming that this is an example of ageism (Help Age International, 2020). The Italian Society of Anaesthesia, Analgesia, and Intensive Care issued a statement early in the crisis suggesting that intensive care treatment might be age-restricted to maximise the life-saving effectiveness of limited resources, an announcement criticised as ageist by some Italian gerontologists (Cesari and Proietti, Reference Cesari and Proietti2020). In a manner reminiscent of the BSG statement, the Canadian Association on Gerontology criticised ‘the use of arbitrary age cut-offs and the generalized (mis)attribution of COVID-19 risk to aging’ for ‘reinforce[ing] and intensif[ying] negative age stereotypes, prejudice, and discrimination’. Similarly, the president of the Irish Gerontological Society criticised ‘arbitrary age cut off[s]’ and argued that ‘prioritizing younger, healthier patients with a higher chance of recovery, as has been suggested in other countries, is “ageist” in the extreme’ (O'Shea, Reference O'Shea2020). At its heart, this is a debate about the boundaries of ageism that has been animated by COVID-19, raising several critical issues for gerontology.

First, a caveat. To speak of ‘gerontology’ as a discipline is admittedly problematic given its widely noted pluralism, begging the longstanding question: ‘What constitutes gerontology's disciplinarity?’ (Katz, Reference Katz1996: 105). Throughout the 20th and 21st centuries, commentators have suggested that gerontology denotes little beyond a juxtaposition of more traditional disciplines that engage with notions of age, ageing and/or agedness. However, at the very least, there is a substantive international institutionalisation of gerontology in various groups, programmes and publications. Moreover, many eminent scholars self-identify as gerontologists, indicating that a gerontology exists at least in the imaginations of some devotees (Kastenbaum, Reference Kastenbaum1992). Ballenger (Reference Ballenger2006: 56) refers to the ‘gerontological persuasion’, encompassing a diverse array of actors who have, since the early 20th century, coalesced around a belief that a better later life can be extricated from unnecessary pathologies, be they cognitive, molecular, socio-cultural or politico-economic. This gerontological persuasion is an unsettled but nonetheless meaningful coalescence, and this paper is itself a manifestation of that unsettledness.

Populations and persons

The first gerontological quandary that is catalysed by the chronological quarantine debate is that of the incompatibility of population level and person-level imaginings of biosocial phenomena, of which age is a potent example. Generally speaking, public health does not deal directly with persons as we experience them in our daily lives. Instead, it deals with the epidemiologic subject, a hypothetical being who embodies select average characteristics of the wide range of individuals who make up a given population (Bunge, Reference Bunge1999). The 80-year-old epidemiologic subject who contracts COVID-19 has a 9 per cent risk of death (Ferguson et al., Reference Ferguson, Laydon, Nedjati-Gilani, Imai, Ainslie, Baguelin, Bhatia, Boonyasiri, Cucunuba, Cuomo-Dannenburg, Dighe, Dorigatti, Fu, Gaythorpe, Green, Hamlet, Hinsley, Okell, van Elsland, Thompson, Verity, Volz, Wang, Wang, Walker, Walters, Winskill, Whittaker, Donnelly, Riley and Ghani2020; Webb, Reference Webb2020), which may differ substantially from the particular vulnerabilities of the specific 80-year-olds who are known to us personally. In a similar manner, my father is a 70-year-old highly active manual worker. At the time of writing, he is confined to his house due to his age-associated vulnerability, and yet he is likely fitter than most of his middle-aged compatriots. His situation seemingly manifests a certain tyranny of averages that is bound up with uses of chronological age to understand populations rather than persons. Of course, the tyranny of averages is a problem in quantitative social science generally (though qualitative research is not immune), but it is especially important in gerontology because the discipline's core subject matter is typically conceived of numerically. It is easy to refute the categorisation of populations in reference to average characteristics because in many cases those averages are almost farcical. As Sabat notes:

Here, the ownership of 1.7 automobiles is a methodological artefact, as is a 9 per cent risk of COVID-19-related death. By representing many persons loosely and the epidemiologic subject precisely, such artefacts often represent no one person well. Thus, in some instances, the social phenomena that we observe at a population level may actually not exist at all at a person level, opening up something of an epistemic paradox for the macroscopic social sciences (Bunge, Reference Bunge1999).

The BSG statement weaponises this paradox and the tyranny of averages to critique chronological quarantine. It does so by appealing to one of gerontology's key contributions to social scientific knowledge – aged heterogeneity. Aged heterogeneity denotes the recognition that a range of inequalities (social, economic, biological, psychological, etc.) generally increase exponentially over time, so that older populations are marked by substantial diversity across numerous measures (Dannefer, Reference Dannefer1987; Nelson and Dannefer, Reference Nelson and Dannefer1992). Aged heterogeneity fundamentally undermines conceptualisations of later life in relation to averages and select characteristics, because it emphasises the great extent of deviation. It is hence a powerful humanist expression of the person in opposition to the epidemiologic subject, a reminder that, when it comes to the observation of age associations, abstracted artefacts can often be far removed from the persons whom they are intended to represent. In this tradition, the BSG statement highlights ‘the diversity of older people’ (BSG, 2020: 6).

The gerontological assertion of aged heterogeneity in response to chronological quarantine is intellectually sound. However, it is also somewhat hypocritical in that it overlooks the considerable affinities between the government's approach and gerontology's implicit – and often explicit – demarcation of older people and later life as distinct categories. The use of chronological age – the number denoting the years since a person's birth – to delineate the subject matter of gerontology (i.e. older people and later life) has been a core contradiction since at least as far back as late 19th century. Gerontology developed in line with social statistics generally as a tabulation of phenomena of interest to the state, e.g. birth, marriage, death (Bookstein and Achenbaum, Reference Bookstein, Achenbaum, Cole, Achenbaum, Jakobi and Kastenbaum1993). Ageing fell into this category because older people were increasingly seen as endangering welfare states and liberal capitalism towards the end of the 19th century, neither producing nor consuming appropriately (Ballenger, Reference Ballenger2006). Chronological age was the most expedient means of quantifying, and thus knowing and articulating, the problem. The 20th-century ‘disciplining’ of gerontology in relation to chronological age is therefore itself part of a modern politics of ageing wherein older people became a social problem warranting professional investigation and intervention (Katz, Reference Katz1996).

Chronological age hence manifests a longstanding intersection of state concerns regarding supposed social problems and gerontological definitions of apt subject matter, as evident in the continued use of 65 to denote older people, manifesting historic European pension ages. The ‘problem’ qualification is key here, because it entails that gerontological attention always risks problematising that on which it focuses. This ‘problem’ is split between the normative and the pathological – the traditional domains of social- and bio-gerontology, respectively. The former construes older people as socially victimised by virtue of their age or agedness, i.e. through mandatory retirement programmes that financially imperil people once they reach a certain ‘old’ age. The latter presents later life as problematic due to the increased presence of molecular and cognitive dysfunction. To an extent, these ‘social problems’ are the ‘research problems’ upon which gerontology has traditionally subsisted.

This problematisation can shift across populations, because the chronologic parameters of gerontological subject matter have varied widely. For instance, writing in the second issue of the Journal of Gerontology, Stieglitz (Reference Stieglitz1946) identified those aged 40 and above as the appropriate population for gerontological and geriatric attention. Today, many national ageing surveys collect data on older populations defined by different chronological ages (e.g. China Health and Retirement Longitudinal Study (CHARLS): 45+; English Longitudinal Study of Ageing (ELSA): 50+; United States Aging Survey (USAS): 60+; Survey of Disability, Ageing and Carers (SDAC): 65+). These numbers define populations that warrant specific types of concern relating to state welfare and social problems. For example, ELSA's (nd) website reads: ‘Data from ELSA participants informs policy across all aspects of ageing including health and social care, retirement and pensions policy, and social and civic participation.’ While the exact number chosen may vary, such engagements of chronological age to specify the subject of gerontological research and welfare governance are longstanding, and perform discursive work by denoting an entity (older people) around which resources are mobilised (Green, Reference Green1993). As Katz (Reference Katz1996: 129) argues, numerical depictions of age ‘acquire an alarmist hue when inserted into the context of discussions about healthcare, pensions, social security, retirement, taxes, and intergenerational relations’.

It is widely acknowledged that chronological age is a poor, and frequently poorly used, biosocial variable (Katz, Reference Katz, Baars, Dannefer, Phillipson and Walker2006). In gerontology, it has always been used as a crude proxy, loosely representing a complex composite of socio-cultural, politico-economic, psychological and physiological human states. There is a long tradition of critique in response to the flaws of chronological age. From the mid-20th century onward, scholars have argued that chronological age categories contain too much human diversity to be of analytic use. Moreover, some have gone so far as to suggest that the preponderant use of chronological age in social analyses actually impedes gerontological advancement (Heron and Chown, Reference Heron and Chown1967; Atchley and George, Reference Atchley and George1973; Murray, Reference Murray1951; Katz, Reference Katz, Baars, Dannefer, Phillipson and Walker2006). The categorisation of persons into age groups facilitates the blanket attribution of characteristics to all those within the category, because biosocial categorisations typically imply the homogeneity of those contained within them (Billig, Reference Billig1987; Cruz, Reference Cruz2017). In response, a heterogeneous ‘lifecourse’ scholarship has become increasingly popular over recent decades, some of which enact far less direct age categorisations (Alwin, Reference Alwin2012; Corna, Reference Corna2013). While the lifecourse offers a potential route through assumptive categorisations, it too often relies on them. Alwin (Reference Alwin2012) notes that the most common use of ‘lifecourse’ in research is to denote age, representing little beyond a rebadging of age categorisation.

Notwithstanding its flaws and critics, chronological age is a widely available, conventional and accessible proxy that facilitates generalisable social analysis. In most contemporary societies, it is a highly naturalised means of social stratification, having emerged over recent centuries as a powerful tool for organising institutional and personal life (Hacking, Reference Hacking1990; Bytheway, Reference Bytheway2005; Cruz, Reference Cruz2017). It therefore remains at the heart of much contemporary gerontological research (Bytheway, Reference Bytheway2005). What should be recognised here is that this commonplace age-based analysis represents the same logic that underpins chronological quarantine. Chronological age is a highly flawed but immediately practicable means of stratifying populations to make them manageable, be that intellectually, empirically or politically. As such, the government's activities are not so different from those found in gerontology. The BSG statement calls out the government's selection of 70 as an arbitrary boundary, yet operationalisations of 50+, 60+, 65+, etc. are a central feature of much gerontology. Looking at almost any issue of this journal, or other notable publications in the field, will uncover numerous uses of chronological age to denote implicitly meaningfully different categories of person. A chronologic epistemology, typically taken for granted, is hence foundational to gerontology. It is a troublesome yet essential component of the discipline's very existence.

Gerontology, like the government, is embroiled in a kind of critical realism when it comes to age, albeit unconsciously in many instances. Critical realism denotes a pragmatic epistemological compromise wherein researchers acknowledge their value-commitments and the limitations of their methodologies, but nonetheless remain dedicated to the scientific explication of social phenomena. It is fundamentally an attempt to tackle misleading dualities (e.g. quantitative/qualitative), and in this respect it is useful for reconsidering the current emergence of a pragmatism/ageism duality. Crucially for this paper, critical realism recognises that reality is only accessed via value-laden constructs (e.g. chronological age), but that this inherently flawed approach is still worthwhile insofar as it can increase our understanding of that reality to at least some extent (Danermark et al., Reference Danermark, Ekström and Karlsson2019). Thus, valuable social scientific insights are based on imprecise but useful metaphors. It is important that ageing researchers remain alert to this so that corresponding policy can account for the imprecisions of social science. Gerontology has always been a particularly acute value-empiric amalgam because of its deep-rooted political commitments to simultaneously understanding and transforming the circumstances of older people (Putney et al., Reference Putney, Alley and Bengtson2005; Minkler and Holstein, Reference Minkler and Holstein2008).

Chronological quarantine should be a powerful reminder to all ageing researchers that we cannot always develop methodologies sophisticated enough to fully respect the reality of aged heterogeneity, but we should nevertheless always be trying to advance in that direction. As Oliver (Reference Oliver2020) notes, government responses to COVID-19 emerged in a rapidly evolving and volatile scenario in which chronological age facilitated the swift development and communication of an intelligible public health response. Having achieved that, the government should have endeavoured to develop increasingly sophisticated approaches that were more applicable to persons and less predicated on epidemiologic subjects. Gerontology is in the same position. There is no shame in our pragmatic compromises, but there may be some ageism.

Ageism and discrimination

At the centre of the current pragmatism/ageism debate is the status of discrimination. The BSG statement argues that chronological quarantine discriminates based on age – a sentiment that cannot reasonably be contested – and that such discrimination is ageist, which is a more complicated claim. The statement defines ageism as ‘the stereotyping, prejudice, and discrimination against people on the basis of their age’ (BSG, 2020: 7). Oliver (Reference Oliver2020) agrees that the policy constitutes age discrimination, but crucially diverges from the BSG approach in distinguishing between discrimination and ageism. In this alternative approach, ageism encompasses stereotyping and prejudice on the basis of age, but not necessarily discrimination.

Ageism has long been a paramount gerontological concern, but discrimination has received less attention. Butler (Reference Butler1969) famously characterised ageism as systemic bigotry. He observed that, in mid-20th-century America, old age was culturally reviled, being broadly considered a detestable period of deterioration and decrepitude. This cultural prejudice was manifest in the individual and institutional treatment of older people. For example, Butler noted that older people were more likely to be the victims of muggings than younger people, and that less than 1 per cent of the federal health research budget was dedicated to work on ageing due to its relatively low national importance (Butler, Reference Butler1969: 244). In his seminal paper, Butler conflated age discrimination and ageism in the same manner as the BSG COVID-19 statement, contending:

At its beginnings, then, ageism was entangled with age discrimination. This conflation remains widespread in ageing research today (for recent reviews, see Ayalon et al., Reference Ayalon, Dolberg, Mikulionienė, Perek-Białas, Rapolienė, Stypinska, Willińska and de la Fuente-Núñez2019; São José et al., Reference São José, Amado, Ilinca, Buttigieg and Larsson2019) and is remarkably under-explicated given its centrality to much gerontology (Voss et al., Reference Voss, Bodner, Rothermund, Ayalon and Tesch-Römer2018). Bytheway (Reference Bytheway2005) has characterised the conflation of ageism and discrimination as the ‘narrow’ definition of ageism, in contrast to the ‘broad’ definition that focuses on dominant societal systems of meaning, or ideology, upon which forms of age discrimination are often predicated. However, Higgs and Gilleard (Reference Higgs and Gilleard2020) have argued that depictions of ageism as ideology are overly nebulous, because their proliferation in gerontology is ultimately a political attempt to equate the oppression of older people with that of other disadvantaged populations, likening ageism to sexism and/or racism (note that discussions of ageism frequently cite sexism and racism together as though the two are also unproblematically comparable). Indeed, such a broad approach to ageism, as ideology that begets discriminatory practices, furnishes a conceptual intersection of both the BSG's and Oliver's arguments.

I am sympathetic to those gerontologists (Higgs and Gilleard, Reference Higgs and Gilleard2020) who have questioned the utility of ageism as catchall denouncement of various age-related ills. Indeed, appeals to ageism are particularly surprising in critical gerontology, which has traditionally challenged universalised conceptions of older people and later life (Dohney and Jones, Reference Doheny and Jonesin press). That said, I do not suggest the outright abandonment of any concept of ageism. Instead, I find inspiration in those works that have urged caution regarding the relations between ageism and age discrimination (Bytheway, Reference Bytheway1995; Macnicol, Reference Macnicol2006). With this in mind, I prefer a definition of ageism as the assumption of intra-age-category similarities and inter-age-category dissimilarities beyond date-of-birth. It is these assumptions that can be considered the ‘generative mechanisms’ of ageism from a critical realist perspective (Dohney and Jones, Reference Doheny and Jonesin press). Discriminatory enactments of those similarities and differences may be entirely related to and reliant on ageism, but they are not inherently the same thing. Though perhaps a more coherent definition, its usefulness, especially politically, is open to debate, as evident in the remainder of this paper.

Thus far, COVID-19 has certainly inspired some relatively unequivocal examples of ageism in the broad sense, manifested in individually and institutionally enacted negative cultural attitudes. For example, The Guardian newspaper has published separate pieces berating the young (Connolly et al., Reference Connolly, Roy, Holmes, Goñi, Phillips, Ratcliff, Halliday and Hall2020) and the old (Clarke, Reference Clarke2020) for not respecting social distancing rules. Naturally, a more accurate means of group categorisation for this sort of commentary would be those who respect the rules and those who do not, but the appeal to longstanding age stereotypes is good copy. More starkly, the phrase ‘BoomerRemover’ was trending on several social media platforms in mid-March. In these instances, the composite stereotypes and prejudices of ageism are abundantly evident, and the hypothetical Venn diagrams of broad and narrow ageism overlap. However, chronological quarantine has a more complex relationship with the intersections of discrimination and ageism.

To appreciate this relationship, it is important to note that age discrimination is common in many aspects of life, especially regarding younger people, and rarely warrants accusations of ageism. In some forms, age discrimination is so uncontentious that it does not even qualify under the Equality Act 2010. For example, a 17-year-old cannot lawfully buy a bottle of alcohol from a shop. Again, we can see here a certain tyranny of averages. At a person level, some 16-year-olds are seasoned and sensible drinkers who, if allowed, would purchase an alcoholic drink and consume it in moderation without severe repercussions. In contrast, some 20-year-olds have deeply problematic relationships with alcohol, regularly consuming large volumes with deleterious health and social consequences. Policy does not make allowances for these individual cases because it operates conceptually at a population level, governing a populace of persons defined in relation to averages. Lawmakers deem the 18-year-old epidemiologic subject to have attained the appropriate threshold of drinking responsibility.

As noted, for Oliver (Reference Oliver2020), age discrimination and ageism are two different things that are not always united, and chronological quarantine is an example of discrimination that is not ageist. He argues that ‘the Equality Act allows for some “differentiation” based on age, as “a proportionate means of achieving a legitimate aim”’ (Oliver, Reference Oliver2020). He suggests that preventing the deaths of thousands of older people is a legitimate aim, that isolating all older people is proportionate to this aim and that the approach is based on sound scientific evidence. Chronological quarantine is therefore depicted as pragmatic discrimination rather than prejudicial ageism. This distinction between age discrimination and ageism warrants considerable critical reflection. For example, is the age discrimination of rights to purchase alcohol ageist? Probably, yes, in the traditional sense of cultural attitudes, because the decision to refuse these rights to 17-year-olds is partly predicated on widespread age-based assumptions of teenage irresponsibility and inability. It is notable that there is no widespread movement against some forms of institutionalised age discrimination, despite their ageist connotations, with some even garnering gerontological support (Macnicol, Reference Macnicol2006; McNamara and Williamson, Reference McNamara and Williamson2012).

We can approach chronological quarantine in the same way, asking to what extent it is grounded in derogatory age-based assumptions of vulnerability. While such assumptions may feed into the policy, it is primarily informed by robust epidemiological evidence of age-associated risk, as outlined by Age UK's epidemiologist Elizabeth Webb (Reference Webb2020). If we approach scientific evidence as being value-neutral, it could reasonably be argued that chronological quarantine is far less ageist than age-based alcohol restrictions. However, I would suggest that it is a mistake to deem chronological quarantine to be a case of pragmatic discrimination rather than assumption-based ageism based on its grounding in the available scientific evidence. In concurrence with the BSG's statement, the policy is probably somewhat ageist, though perhaps in a manner that is not recognised in current debate. In the spirit of critical realism, I will now conclude by suggesting two important attributes of chronological quarantine that are especially salient to gerontology: (a) the policy is both pragmatic and ageist; and (b) this is largely a reflection of gerontological ageism.

Pragmatic ageism

The BSG's (2020: 1) call for ‘a clear focus on evidence-based practice, using high quality research’ as a means of challenging ageism is ultimately problematic, precisely because the available evidence is based on conventional demographic and epidemiological research that stratifies disease risk by chronological age. It was inevitable that COVID-19 would be categorised by age (and sex and ethnicity and socio-economic status) in research, because that is what researchers ‘know’ to look for, and gerontology is a major purveyor of this type of engagement with society through conventional categories (Bytheway, Reference Bytheway1995). These conventional categories of analysis are generally inert in social research, and their repeated use, based on precedent, reinforces the likelihood of their future use (Cruz, Reference Cruz2017). A collection of assumptions has guided researchers to collect data on the ages of COVID-19 patients, to look for associations between their ages and their outcomes in the resulting dataset, and to generalise and publicise those associations as somehow representing the circumstances of people who did not participate in the research, but who have similar birth dates to those who did.

Specific imaginings of age, and its importance, are core conventions of contemporary social science, and particularly of gerontology. Recognising this tendency, Bytheway (Reference Bytheway1995: 107) contemplated whether ‘when gerontology becomes scientistic and behaviouristic it also becomes ageist’. He concluded that it did, citing a systemic predilection to engage with populations in terms of assumptions, measurements and prescriptions of age, and to universalise those prescriptions so that knowledge claims can be switched between imagined populations and specific persons, irrespective of the latter's relationship with the former. The gerontological evidence base therefore manifests a type of ageism. Indeed, the existence of a scientific discipline in relation to an imagined population is itself somewhat ‘othering’ of that population (Higgs and Gilleard, Reference Higgs and Gilleard2020). With this in mind, chronological quarantine can be considered ageist because it is based on scientific evidence. It is simultaneously pragmatic, evidence-based and ageist. However, while social science and the claims it generates may be bound up with forms of ageism, that does not necessarily warrant the relaxation of chronological quarantine, just as it does not warrant the end of gerontology. Instead, it requires us to improve our engagements with age. The government's use of age as a means of social stratification manifests a critical realism that is integral to gerontology, albeit often unacknowledged. It is an imperfect means of engaging with a complex social world. In this light, the ethical and empirical purism of gerontological accusations of ageism is suspect.

While ageism has previously been critiqued as a totalising approach to a plethora of age-related ills, the pulling-apart of ageism, prejudice and discrimination in the context of COVID-19 opens up a different set of potentially uncomfortable contradictions within social scientific engagements with age. It reflects traditional ageism critiques back on to the gerontological persuasion, hinting that the accuser and the accused may be embroiled in similar types of social arrangement. Despite the discomfort, the contradictions of ageism are worthwhile. Critical gerontology is at its most useful when it stimulates reflection, debate and ultimately a reimagining of approaches to age, older people and later life (Doheny and Jones, in press). Focusing on the confluences and divergences of ageism, prejudice and discrimination can assist in the pursuit of these ends, whereas insufficient engagement risks undermining progressive gerontological efforts. As a case in point, institutional responses to COVID-19 in relation to older people in the UK, and internationally, seemingly warrant substantive gerontological attention, for which the BSG statement should be a call-to-arms.

Reactive appeals to ageism, wherein discrimination and prejudice are conflated and moralised, risk undermining our capacity to contribute constructive critique. They are criticisms of a system from within that system, and therefore manifest similar assumptions, but they are too often portrayed as criticisms from outside that system. If we cannot escape ageism entirely, for the reasons discussed in this paper, then the trick is perhaps to identify precisely what it is within ageism that we object to and what might be preferable. Ultimately, this is not an argument in support of the government's approach or in opposition to empirical engagements with age. It is an appeal to scrutinise the boundaries and mechanics of ageism, particularly in relation to the study of age, and to reflect on how age and ageism are used intellectually, empirically and politically.

Is chronological quarantine ageist? Yes, slightly, given its reliance on the assumption that age denotes distinct types of people, envisioned in terms of the aged epidemiologic subject. However, a critical realist perspective can grant us a more nuanced appreciation of the role of ageism here. The evidence base manifests a type of deep-rooted ageism and is alienated from the circumstances of real persons, yet it simultaneously draws our attention to life-threatening risk and inspires a practicable means of lowering that risk. The policy thus relies on a widespread and useful epistemic ageism, and one that is principally manifest in conventional biosocial research on ageing. Gerontology is at least as implicated as the government in this ageism, but we should not despair of such affinities. They should ultimately inspire us to reflect on the core assumptions of gerontology, the ways in which we categorise persons and how we might do better.