What Is Not Measured Does Not Exist: Ageism in the Production of Knowledge
By ageism we do not only mean cultural prejudice, but the kind also that manifests itself in language, in treatment, in the interactions of everyday life. Or in algorithms, as we saw recently. Ageism is a much deeper form of discrimination, with a greater dimension than one might think a priori, and it also appears in the form of institutional biases in policies, data, and, in general, information systems. It is not something that happens in one specific country, in isolation; rather, it seems that no country or institution is truly “safe” from repeating prejudices. This is pointed out, for example, by the World Health Organization’s Global Report on Ageism, which makes explicit that ageism includes institutional practices that exclude or underrepresent older people in public decisions and in the production of data.
But what happens if we do not appear in the data? Or, more specifically, how does statistical invisibility affect older people? Well… what is not named does not exist, and therefore in a much more serious way than we might think: exclusion from statistics and records leads to a systematic underestimation of their real needs. If what happens to older people is not measured properly, it is much more difficult for it to become part of the political agenda, or even to be recognized as a problem or a limitation. And then we will not be able to solve it.
This underestimation of their needs, of their reality, happens for various reasons. For example, many posts ago I already spoke about the unjustifiable exclusion of people over the age of 50 from surveys that analyze sexually transmitted infections, especially at a time when these are increasing — yes, also among older people. There may be numerous reasons that explain their non-inclusion: moral ones? assumptions such as the absence of a sex life among older people? Let us think of other examples, such as the Statistics on Homeless People, which does not systematically or analytically disaggregate by age brackets among those over 65. I suppose — I do not know — that it assumes there is hardly any homelessness among older people, perhaps because they have access to housing arrangements that do not exist for other ages, such as nursing homes, but the fact is that we cannot know this if we do not have data.
If we look more closely, it is not only a matter of disaggregation by single years of age, year by year, but rather that the approach lacks an age perspective. For example, the focus is on administrative status or other issues — undoubtedly of great interest — but attention is not paid to life trajectories, which would be a very interesting matter for better understanding not only the situation of the oldest people, but also life processes themselves. In other words, the invisibilization of old age prevents us from completing our understanding of other stages of life. Homelessness in old age disappears statistically, even though it exists in non-visible forms.
The form of measurement is also less refined when it comes to older people, and there is an enormous underestimation that prevents us from knowing unexpected realities. We know, for example, that older people work in a very low percentage and that not many of them start businesses, but we do not have refined data on this. There is an apriorism that prevents us from knowing it: it is assumed that they do not do it, it is not asked, and so… it is concluded that it does not exist. The assumption that something does not exist ultimately causes it not to exist: if it is not in the data, it is as if it did not exist. I do not know how significant these shortcomings are, mind you, but it would be good to know whether they are.
Therefore, an enormous homogeneity is usually assumed in what is probably the most heterogeneous group that exists and the one with the longest age span. What I have commented on other times: it is not the same to be 66 as to be 95, even though we place both people within the same group. Considering that, thanks to data, we can design more or less appropriate policies in relation to real needs — since it is data that makes it possible to structure public action — the risk is enormous.
We have already talked about ageism associated with AI, but it is worth remembering that if older people appear systematically invisibilized in the data or underrepresented in the datasets that AI will later use to automate systems or respond to problems… it will undoubtedly overlook them. Their problems, their needs, their existence! will not even be detected. We are thus designing a new form of society by erasing part of it in its very definition. An AI is trained for a future that does not include older people. And be careful, because those who are removed from the equation are, to a large extent, the very people who are designing that future: today’s young people will be tomorrow’s old people.
Another issue to point out, with a terrible impact, is invisibilization in scientific research. We not only lack knowledge about matters related to the behavior or social dynamics that affect or are led by older people — beyond the “stereotypical” ones, such as loneliness, of course — but we also lack knowledge about issues with an impact on health and well-being. Older people are undoubtedly represented in studies on Alzheimer’s, diabetes… but there is something that the scientific literature has called “epistemic neglect” toward old age.
This term would refer to a bias in the production of knowledge, like a kind of chosen “blindness” that means not enough is investigated and what does exist is not properly interpreted. Although it appears under different terms — it is not as accepted or as widely used as the term ageism — its basis would be closely related to what Miranda Fricker defined as “epistemic injustice,” which would be a form of injustice that occurs when social prejudices unfairly discredit a person as a source of knowledge. And so we lack frameworks for understanding their experiences.
Ageism also manifests itself in the exclusion of older people from clinical trials. Obviously, in the absence of evidence, worse medical decision-making is to be expected. We saw this, for example, during the earliest stages of COVID; the manifestations of the disease were different among people of very advanced age, which sometimes prevented a faster diagnosis.
This invisibilization also tells us which topics do or do not receive funding. Thus, we find ourselves with a newly conquered territory of the life cycle about which we know hardly anything.
Once again, the above forms part of a broader conceptual corpus: the older person appears as an economic burden — watch how tiresome certain sectors become when they want to see the welfare state “burn” — but not as a subject of rights — one with the right to a pension and care after a life of sharing and building, by the way — nor as a social subject with the capacity to decide and act upon their own life. We lack adequate research protocols, which has a great deal to do with the very conception we have of old age: as something that does not matter, or that has lesser importance, that is not worthy of study.
If we do not produce knowledge about old age, if we do not have adequate data, how will we be able to respond adequately to the needs of a more long-lived society?
