Get old or die trying: Longevity justice in social insurance

Abstract

Of all the risks we face in life, ranging from unemployment to old age, early death is among the most tragic and yet most neglected by modern states. Liberal egalitarians might find it easy to dismiss social insurance against early death, but I argue they should not. Early in this paper, I explain why social insurance should include the risk of premature death by replying to four common criticisms. What follows is a case for a novel form of insurance that is more liberal and egalitarian across ages than the status quo. More specifically, I show that granting storable pensions in all life stages – youth, middle, and old age – promotes liberty and reduces lifetime inequality. Writers on age group and longevity justice often overlook the extent to which liberal lifetime equality can supply insightful arguments against inequality across ages. These call for a policy that gives each age group the freedom to draw a storable pension. I call that policy freetirement.

Introduction

Of all the risks we face in life, ranging from unemployment to old age, early death is among the most tragic and yet most neglected by modern states. ¹ The risk of premature death, by which I understand before old age, is far from negligible. From 2015 to 2020 in Europe, for instance, one-fifth of all deaths were between the ages of 15 and 65 (UN, 2019: 17). Should social insurance protect individuals against the risk of dying early? And, if so, how should it be done?

Early in the paper, I reply to four important objections against early-death insurance. What follows is a case for a liberal egalitarian version of that insurance, which spreads pensions more evenly across all age groups: young, elderly, and middle-aged. The importance of this result can hardly be overstated, especially given the role of early death in making other forms of insurance attractive. Consider what Nicolas de Condorcet (1995) wrote in a founding formulation of old-age insurance:

Here then is a necessary cause of inequality, of dependence and even of misery (…) it can be in great part eradicated by guaranteeing people in old age a means of livelihood produced partly by their own savings and partly by the savings of others who make the same outlay, but who die before they need to reap the reward.

Condorcet notes that part of what makes old-age insurance financially attractive lies in the contribution that the prematurely dead never get to receive. This seems all the more so as we shift pensions to older ages, which is the direction we are now heading. There is a widespread agreement among politicians, philosophers, and economists that old-age pensions should shift to later in life because we are living longer on average. However, the problem is that the more we do so, the more we exclude those who sadly die younger from enjoying those goods. The solution to our longer lives cannot be considered longevity-just unless it also protects those who die early. What is not clear, yet, is what protecting people against premature death requires as a matter of social justice.

The standpoint of social justice adopted here is liberal egalitarian. It is liberal in that it respects and promotes people's ability to pursue and revise their conception of the good life, whatever it may be. It is egalitarian (or prioritarian) because it prioritises those with the least means to pursue their ideas of the good life. ² The maximin (or leximin) criterion I use holds that the involuntarily worst off in life should be as well off as possible, even if doing so comes at the cost of increasing inequalities. ³ In line with liberal egalitarians such as Ronald Dworkin (2002b: 94-5), Thomas Nagel (1995: 69), and John Rawls (2001: 174), I assume that it is the prospective quality of our whole lives that primarily matters for how well off we are. This is known as the ‘complete-lives’ view.

Influential writers on age-group justice tend to endorse this view. Still, they worry that it makes us indifferent to discrimination and inequality across ages. ⁴ The worry is that any age-based differential treatment, as bad as it may be, might treat us equally over our entire lives because we all age. Giving more of some good to younger or older people creates inequality between individuals at specific times that might not translate into a disparity between their whole lives. If the older have been young and the young will one day be old, both end up receiving the same in life. Age is special in this way. Unlike sex-gender or race-ethnicity, it cannot but constantly change across life. To avoid the embarrassing result that lifetime equality is too lenient on age-based treatment, writers on age-group justice have turned to other principles, such as sufficiency, prudence, and relational equality. ⁵

But though we all age, we do not all age equally. The thought that complete-life equality implies indifference towards age-based inequality tells us how little attention we have been paying to longevity. Unlike class, gender, ethnicity, and even age, longevity is not among the common worries of the day. Although longevity is not as socially salient as these other features, it has an interesting relationship with equality, especially with entire-lives equality. It is true that the older have been young, but many young people do not live long enough to be older. Sections ‘Youth pensions’, ‘Old-age pensions’, and `Middle-age pensions´ show that applying our normative standpoint to differential longevity does not justify indifference but, instead, a special concern with all life stages.

In this paper, I focus on the distribution of pensions, which I understand as income transfers paid by the state as compensation or protection against some event. ⁶ These regular payments may, but need not, imply that the recipient has retreated from paid employment. Writers on this topic often assume that longevity justice must be achieved through retirement pensions, but the problem with retirement is that, unfortunately, it is still too much associated with leisure. ⁷ Because the allowances I have in mind leave open the choice of what to do with them, it may be better to speak of pensions more generally. These are unconditional on willingness to be employed, and they may or may not lead one to ‘retire’ temporarily. There is no advantage – financial, liberal, or even egalitarian – in assuming that being compensated for the risk of early death requires one to leave the labour market.

What I call a ‘pension’ is, basically, a special income transfer that protects individuals against a particular kind of event, say unemployment, old age, or premature death. A pension can be specified in several ways, three of which are particularly important. First, it is likely paid directly, as any typical subsidy is, but it does not have to be. For instance, it is possible to indirectly raise people's net income by significantly reducing their fiscal liabilities, if not exempting them from taxes altogether. ⁸ Secondly, pensions may or may not be conditional on work or contribution. Thirdly, pensions can be given as a one-off capital grant or as a regular instalment throughout a period. This one is especially connected to early death. As Philippe Van Parijs (1997: 46) once noted, regular instalments, such as Universal Basic Income (UBI), give ‘those living to an older age (…) a larger total basic income than those dying earlier, and the distribution seems therefore bound to violate the leximin criterion’. ⁹

Only a year after Condorcet wrote his proposal, Thomas Paine (2004) published his pamphlet ‘Agrarian Justice’, where he advocated for an estate tax upon adulthood to fund universal old-age support and a pension for the young, a fixed grant paid to all citizens once they enter adulthood. ¹⁰ The proposal of youth pensions seems more friendly to those who die young than UBI, and it is close to what I will advocate in the ‘Youth pensions’ section. Yet, I show that youth pensions are still insufficient. Surprisingly, early-death insurance also requires old-age pensions, in line with what Paine had once proposed. Finally, I also make a case for the necessity of middle-age pensions. Together, these three allowances bring about a freer and more even distribution of pensions across life, which I propose to call freetirement.

Social insurance against premature death?

Insurance against early death might have been neglected because, we think, there is no meaningful way to reimburse those who die young. Often, we cannot know in advance who will be short-lived. We only know who they are once they have died, after which our ability to compensate them for early death seems limited. It is difficult to identify the short-lived ex-ante and to compensate them ex-post (Fleurbaey et al., 2014). Because of this problem, it seems that the best way to compensate those who die young while they are still alive is to shift goods (such as pensions) to the ages when they are still alive; that is, towards everyone at younger ages. Since this pension arrangement is better for those who die early than the status quo, the question is not so much about whether we ‘can’ compensate people against the risk of premature death. Instead, the more controversial question is whether we ‘ought’ to do so. This section examines four reasons to answer this question negatively.

Social or cosmic?

It might be emphasized that the risk of premature death is just a biological fact and that such facts are neither just nor unjust. That, as John Rawls (1971: 87) once put it, ‘what is just and unjust is the way that institutions deal with these [natural] facts’. The ‘cosmic’ fact that some humans are alive for longer on planet Earth is, one might say, beyond social justice.

In part, it is a biological fact that people do not live equally long lives. Some are born with genetic diseases they cannot do anything about. But societies can also exacerbate or attenuate these discrepancies. These are two independent claims. That premature death is purely natural does not detract from the possibility that the prematurely dead can make complaints about social justice. No one intends to say, for instance, that paraplegic persons cannot be victims of social injustice if their inability to walk is rooted in natural causes. Rawls's quote above cannot deny that it is possible to mistreat people based on such facts. It is a relevant case of social justice if the lives of paraplegic persons or short-lived persons contain fewer goods, opportunities, and rights, than others. So it is not the bare fact that some live longer than others that is the source of (dis)advantage here, but the goods and bads that lives of unequal length contain. ¹¹

Equal chances of growing old?

Given this complaint, I will not assume that longevity is itself the source of (dis)advantage. Instead, I take it that the disadvantage lies in the fewer goods that shorter lives typically contain. This is often true ex post: after uncertainty is gone, and we know how long we lived, it will likely be that, all else being equal, the prematurely dead had fewer resources and welfare in life (Ponthière, 2020b). This is not as obvious if we take equality to apply ex ante, which many do. ¹² For example, Rawls describes the worst-off by reference to the expectations of their representative member. When we all have the same odds of living through every age, those who die early are not worse off before their longevity ‘cards are dealt with’, since the risk of premature death and old age is equal for all. For this reason, it is tempting for defences of early-death insurance to reject ex-ante equality. ¹³ Yet, I think this is unnecessary. This article may be seen as an ex-ante case for early-death insurance.

One problem with this line of argument is that, in reality, it is well-documented that disadvantaged socioeconomic classes and ethnicities are statistically more likely to die prematurely (Chetty et al., 2016). There is, one might then say, a predictable socioeconomic and race bias in social security: redistributing from shorter to longer lives involves (regressive) net transfers from the worse to the better off and from people of colour to whites. ¹⁴ So those who are made worse off by socially salient characteristics tend to die younger, with one known exception. The exception is that women can expect to live longer despite being relatively poorer than men. ¹⁵ This means there may be a feminist argument for higher old-age transfers. The link between longevity and other characteristics may well call for targeted policy measures, which are well in line with what I will argue for in this paper.

What about if such correlations disappear and longevity is entirely unpredictable? Would ex-ante egalitarians no longer have to worry about premature death if we all had an equal chance of getting old? Even if we were all equally likely to live equally long lives, ex-ante egalitarians can still defend the possibility of insurance against the risk of early death as an instance of bad (brute) luck. The idea is not that such insurance should be imposed on individuals, but only that it should be available. It could be reasonable for people to take up insurance against early death if the offer was on the table. On a plausible version of ex-ante equality, the fair insurance approach, we should respect, within limits, the insurance decisions symmetrically situated individuals would make about the risks they face. ¹⁶ To make the discussion more challenging, I assume (unless I state otherwise) that we are symmetrically situated concerning the risk of early death and old age.

Involuntary accident?

It is sometimes asked: What if early death is not an involuntary accident but a consequence of how we choose to live? While it may be contentious to hold people accountable for the duration of their lives, ¹⁷ we must address the background premise that a person loses a claim to social insurance against early death if her passing was sufficiently within her control. Note that this claim is not a principled objection to any right to social insurance. It simply specifies the conditions under which it should occur (for instance, only when early death is involuntary). Just as the condition that unemployment must be involuntary is not a threat to the existence of unemployment insurance schemes, the same is true of social insurance against premature death.

If there must be an involuntariness condition, it should be weak. For instance, we might make choices that reduce our employment prospects. Yet, we tend to think that these should not disqualify us from involuntary unemployment insurance, provided one demonstrates some will to work. A similar point should be made about people whose lifestyles render them more likely to die earlier. Early-death insurance should not exclude people who engage in a less healthy or riskier lifestyle as long as one demonstrates some will to live (however understood). Social insurance in liberal societies should refrain from excessively intrusive judgments about how one should live. The (higher) risk of early death of people with unhealthy lifestyles may be considered voluntary only in a very few extreme cases.

Benefit suppression

Early-death insurance might impose high costs on many advantaged persons for the sake of small benefits to a few, short-lived individuals. Consider the pessimistic scenario in which youth pensions reduce the young´s efforts to work, causing negative scarring effects on the lives of most people, especially those who live longer to suffer them. ¹⁸ These far-reaching changes in the economy will likely have effects, and it is not easy to predict what they will be. The pessimistic scenario is a possibility that any real-world implementation of insurance against early death must consider.

It will be difficult for the liberal egalitarian view expressed in this article to be traded against larger benefits to the more advantaged. But difficult is not impossible. It all depends on how we answer two further questions: the empirical question of how large the suppressed benefits would be and the normative question about how much priority we attach to benefiting the worse off vis-à-vis better off individuals. Substantively, there is reasonable disagreement about how large a benefit to the better off must be to outweigh a smaller benefit to the worse off, and, as well, about how many better off persons does it take to outweigh the claims of a fewer worse off. ¹⁹ These are difficult questions which I cannot deal with adequately here. But I do not need to, either. I only want to insist that it is plausible to include insurance against early death as part of the protection a just society owes its worse off. This is not to say that protecting the worst off from unfairness always overrides other, much larger, benefits to the better off, depending on how egalitarian or prioritarian we are. To undermine my argument, the objection would have to show that the costs to the more advantaged are so severe as to justify no insurance against early death whatsoever.

Youth pensions

The short-lived often cannot be identified ex-ante, so we usually only know who they are ex-post. Yet, it might still be possible to compensate them after they die. Consider, for instance, what Condorcet says right after the passage I cited earlier:

[O]r, again, on the same principle of compensation, by securing for widows and orphans an income which is the same and costs the same for those families which suffer an early loss and for those which suffer it later.

Condorcet proposes compensating the prematurely dead by assisting their families. ²⁰ If there can be posthumous harms or benefits, societies could provide targeted benefits to the short-lived by protecting those they care about. Though promising, this strategy still fails to compensate the actual victims of misfortune while they are alive. If we can, it is preferable to fulfil our obligations of justice to people when they are still alive. This is possible if the two following claims are true:

Relevance of Youth: it is only at younger ages that we can directly benefit the prematurely dead while they are alive.

Badness of Death: In general, death is bad to the extent that it deprives us of goods that we would have otherwise enjoyed had we remained alive. ²¹

Both claims seem plausible. Together, they imply that the death of the prematurely dead could be made less harmful if goods were shifted from later to earlier in the lifecycle (Fleurbaey et al., 2014). Recently, Gregory Ponthière (2020a, 2020b, 2023) applied this argument to retirement pensions. Ponthière argues for reverse retirement, a policy that redistributes retirement pensions from the old to the young. It would do so by increasing retirement ages and using the proceeds to fund the retirement of young adults, who now begin adulthood as retirees.

In principle, these youth retirement pensions do not need to be compulsory. ²² Indeed, I shall argue that the young should be able to save them for old age if they so wish. I call them ‘youth pensions’ rather than ‘youth’ or ‘reverse’ retirement because, again, I do not want to assume that early death must be compensated by enhancing access to time away from the labour market. The assumption that shorter lives go better the higher their share of leisure penalizes many of those who, faced with early death, would prefer to work more (precisely because their clock is ticking faster). As I mentioned earlier, this problem is not present in youth pensions (or allowances), as these are neutral between work and leisure (think, for instance, about the grants proposed by Paine (2004)).

An egalitarian case for youth pensions

In this egalitarian case for youth pensions, I argue for two claims. One is that the short-lived are among the worst-off (a). The second is that youth pensions are a better way to benefit the short-lived than reverse retirement (b).

Premise (a) holds regardless of what the goods in question are. People's share of goods in life typically shrinks the earlier they die, be it in welfare or resources. When we look at our share of resources over a complete life – as in terms of income or wealth – we see that our shares are not just lowered when we have little per year we live but also by the fact that we live only a few years. This judgment becomes more robust when we drop the assumption that each of us faces an equal risk of early death and we, then, acknowledge that those poorer in resources are more likely to die younger.

Nothing changes if we add leisure to the picture. The addition of leisure alongside income and wealth often implies that people are worse off the more they work. ²³ And it is true that longer-lived individuals indeed tend to work more in life than shorter-lived people. But contrary to the common rule, this does not mean that they have less leisure as a result. Interestingly, the case of differential longevity is an exception. If the timespan is the same for two people, say a day of 24 hours, then ‘less work means more leisure’. But when timespans are not the same for all, the fact that longer-lived persons work more does not mean they have less leisure time. It is still the short-lived who had, in fact, a lower share of lifetime leisure. In the case of differential longevity, exceptionally, adding leisure to the list of resources will not justify benefits in proportion to working time, as the short-lived still expect to have less leisure time in life.

What if it is not leisure but work that is a resource? ²⁴ There are various goods in work besides money (Gheaus and Herzog, 2016), and the short-lived miss out on them too. This point does not challenge my argument, since the short-lived also expect to work less in life. They will be worse off regardless of their preferences for leisure and work. This point, in turn, supports premise (b). Reverse retirement does not improve the position of the short-lived for whom work is good. Such persons have no reason to anticipate leisure by taking up reverse retirement. This departure from Ponthière´s proposal is significant and necessary. Youth pensions are more liberal in that they welcome our withdrawal from the labour market but can equally be used to support our presence in it. They show equal concern for short-lived persons who want to work early in life by allowing them to not take specific jobs out of financial necessity but to give them the means to pursue those occupations they genuinely want, especially those with high entry costs.

A second respect in which youth pensions may part ways with reverse retirement is that pensions must be a storable claim one can save for later. Storability means that, once one has a right to it, one can take up that transfer whenever one wants. What short-lived people are being given is, of course, the chance to make a choice under uncertainty. Whether persons take up their youth pensions earlier or later ultimately depends on their own values and attitudes towards risk, especially that of dying early vis-à-vis living longer. We can say that a prohibition of storability will not improve the position of those who would like to use benefits at an age later than the one specified by reverse retirement.

What if I postpone my pension to old age just because I had not known myself to be short-lived? If neither the social planner nor individuals know the length of their life in advance, youth pensions give people choices they might regret, because youth pensions do not oblige us to enjoy them early. Does this undermine their egalitarian credentials? Ex-ante egalitarians will not find it unjust that, when people are free to insure against various risks, some will make choices they had regretted if they had complete information about their situation. Ex-ante equality assumes away this information and equalizes us only before uncertainty. The worry that the short-lived have fewer pensions than the long-lived only occurs after (and because) they know about their realized longevity. The problem of regret only arises if what we care about as egalitarians is ex-post inequality.

From an ex-post perspective, we need to ask what is enough as a matter of social justice: that we have access to or take up pensions. If the former, the problem of regret does not lead to injustice. The short-lived persons who postponed their pension had as much access to pensions as those who did not, even ex-post. There is no ex-post inequality because the short-lived who did not take up their youth pensions still had access to them. The problem of regret only leads to injustice if justice in pensions requires that we enjoy roughly the same amount of pension time ex-post, a view I am sceptical that we would find attractive. It is better to remain neutral and let people match pensions with their conception of the good life. Especially if doing this does not lead to relevant inequality of access to pensions between people of unequally long lives, ex-post and ex-ante.

A liberal case for youth pensions

Youth pensions are no less egalitarian than reverse retirement (b). Meanwhile, they show concern for a wider variety of conceptions of the good life than reverse retirement, such as of young people, some of them potentially short-lived, who prefer not to withdraw from the labour market but prefer to save their pension for later in life. Much of what makes youth pensions more liberal is their commitment to not assuming that the best life for the short-lived (and for all of us) is one where they enjoy the most leisure early in life.

From a liberal standpoint, insurance against early death in the form of youth pensions is better than both the status quo and the proposal to ‘reverse’ retirement. A liberal society should promote people's ability to pursue their conception of the good life, whatever it may be. Youth pensions are liberal in that they maximize the choice of using one's initial endowment of resources as one sees fit. Once such storable rights are given to us, we may use them to surf in Nazaré, invest them in our dream occupation, or store them as our old-age pension. Unlike reverse retirement, it does not assume that the short-lived should be compensated through withdrawal from the labour market at a young age, which may not align with how many of them want their lives to go.

Early-death insurance is also more liberal than the status quo, which, in terms of longevity risks, only consists of social insurance against living long. Early-death insurance is neutral in a way that old-age insurance is not. While persons could always use insurance against early death to buy protection against the risk of a long life, as by saving theirs for old age, the opposite cannot be done with old-age insurance. One cannot anticipate old-age pensions, which are only paid to us when it is no longer possible, or necessary, to insure against early death. Behind a veil of ignorance in which we do not know our risk preferences, choosing youth pensions as insurance against early death would not commit us to a specific conception of the good life. Unlike old-age insurance and reverse retirement, youth pensions give fairly situated individuals (ex-ante) much more freedom to insure against the chances of a short or a long life according to their attitudes to risk.

To illustrate, consider the capital grant proposal by Paine (2004) and Ackerman and Alstott (1999) to pay a fixed sum to all citizens upon adulthood. It contrasts with the standard UBI proposal that purports to pay citizens in regular instalments across life. One-off grants are better for the worst-off short-lived but also increase our freedom to pursue any conception of the good life. There is one conception of the good life that UBI frustrates but that Paine’s proposal allows: investing more in our youth than in our old age. Yet, there is nothing that UBI allows that one-off grants forbid. A right to storable youth pensions opens up more possible conceptions of the good life than old-age insurance, without committing us to endorse any single one of them. And just as individuals can choose to receive their pension later, they might also prefer to receive it in regular instalments over some time. It is up to them to decide what best fits their plans.

In the end, youth pensions might also be used to purchase additional insurance for other aspects of life, for instance, against sickness or disability. But to be sure, they cannot serve as an argument for letting people go of their state-funded health insurance on behalf of higher youth pensions, even if this aligns with their own attitudes to risk. This liberal form of social insurance against early death allows persons to use their pensions, however, and whenever they see fit, as a complement, and not a replacement, of other already established forms of social insurance. This should now be clearer as we turn to old-age pensions.

Old-age pensions

However necessary they may be, I argue that youth pensions do not exhaust the demands of early-death insurance. Ponthière, too, agrees that reverse retirement is not enough: social insurance should not strive to abolish existing old-age insurance schemes but instead add ‘some dose of insurance against premature death’ to them (2020b: 25). Old-age insurance can easily accommodate the concerns with benefit suppression and incentives discussed earlier, by providing for a better structure of workers’ wages and performance across life (e.g. Halliday and Parr, 2022). ²⁵ Along similar lines, Ponthière concedes that the young short-lived do not stand much to gain from postponing old-age retirement beyond the ages at which productivity becomes quite low and the disutility of labour very high (Ponthière, 2020b: 4).

The challenge with this line of argument lies in showing that old-age insurance does not deteriorate, but indeed improves, the position of the worse-off short-lived. As we have seen in the ‘Youth pensions’ section, it seems that the more resources we transfer to the elderly, the fewer resources will be available for insurance against early death, or that the more resources we transfer to protect the young short-lived, the fewer resources we will guarantee as protection for those who grow old (Gosseries, 2022). The youth pensions I have described do not ensure that people will save resources for later in life rather than spend them all early on. This apparent dilemma between insurance against early-death and old-age insurance – one between shifting goods to the young (or younger selves) or towards old age (or older selves) – seems to arise regardless of how pension schemes are funded.

One ‘solution’ to this dilemma is to give up on old-age pensions. After all, ‘complete-lives’ equality suggests that the longer-lived are well-off compared to the alternative. Here is a funny example. When asked, ‘Michael, how is it to get old?’ The famous actor Michael Caine replied, ‘Well, considering the alternative, fantastic’. ²⁶ Such statements put pressure on the egalitarian credentials of old-age pensions. Of course, liberal egalitarians have good reason to be uncomfortable with the prospect of a crowd of elderly destitutes charged with not having saved enough for old age. ²⁷ As I mentioned earlier, the way writers on age-group justice have tried to avoid such conclusions has been by supplementing their commitment to complete-lives equality with some other principle, such as sufficiency, relational equality, or prudence. ²⁸ Those seem plausible to me, provided meeting those obligations does not deteriorate the position of the lifetime worse off. However, it has not yet been shown that old-age pensions do not worsen the condition of the least advantaged in life.

In what follows, I offer two reasons to believe that old-age pensions can, surprisingly, improve the position of the short-lived. If I am right, complete-lives equality is all we need to explain why youth pensions are not enough and should be complemented with old-age pensions, even if one is committed to giving the most priority to insurance against a short life.

Old-age hell

Old-age pensions induce an inequality between longevity groups when they give more to the long-lived than to the short-lived. Yet, this inequality may be justified by the fact that, to some extent, it improves the position of the worst-off short-lived. We can attenuate the conflict between the interests of short- and long-lived persons by noting that, under some conditions, old-age pensions can render all longevity groups better off than they would otherwise be.

People are made better off by living longer, not only insofar as they enjoy more goods but also to the extent that they avoid bads. While the longer-lived enjoy more goods, they also face more hardship. Health and productivity tend to deteriorate after a certain age, and many problems arise with ageing. ²⁹ Perhaps one could insist that such bads make the lives of the short-lived go better, that the short-lived are deprived of less if the longer-lived go through some kind of old-age hell. ³⁰ Consider, for instance:

The Lucky Die First. ³¹ Imagine a society where we all live great lives until 60 years of age. But soon after, those who live past 60 begin being treated horribly. While their life is not short, it quickly becomes solitary, poor, nasty, and brutish.

This world is bad for the long-lived and it does not meaningfully benefit the short-lived either. It is doubtful that we have reason to redistribute towards the short-lived in these cases. The motivation for early-death insurance relied on the shorter-lived being worse off in life, but it is not clear that they are so when older persons go through hell in old age. Instead, it is sensible to say that the long-lived are not necessarily better off if they live through the same kind of paradise as the short-lived, but the short-lived do not live through the hell that blights the lives of the long-lived. While it is tempting to consider longevityto be a disadvantage in this case, we do not have to restrict ourselves to cases where ‘the lucky die young’. We need not say that one would be made better off by dying earlier. For our present purposes, it suffices to insist that the short-lived would not be disadvantaged by early death if those who outlive them do not enjoy a sufficiently good old age because they go through some kind of old-age hell. This follows from our earlier idea that the source of longevity (dis)advantage lies in the goods (and bads) that lives of unequal length contain.

Early-death insurance requires, then, a sufficiently high old-age pension that guarantees we can all escape hardship in old age. Without this, early-death insurance may lose its very reason to exist: it becomes doubtful that the short-lived are worse off and owed compensation if a longer life is not sufficiently good because the bads that come with ageing are significant. Short-lived persons will be better off if we retain our reasons to compensate them by ensuring that longer lives are, at least, sufficiently good.

Ageing fears

Old-age pensions can also benefit the short-lived more directly, by freeing them from the fear of ending a long life in poverty. Without old-age pensions, we might end our lives with too little to live with. Since most of us do not know whether we will live long, we all have reason to fear both the risk of early death and the risk of old-age poverty. If the short-lived fear not having enough in old age, the youth pensions they receive might end up being just a formality. ³² It is not something they will likely use to improve their lives. Consider:

Ageing Fears. Suppose we all only have a right to some storable amount of youth pensions. Since longevity is uncertain, one fears using the grant right away. The short-lived among us would feel uncomfortable using their youth pension, as they fear the risk of poverty later in life.

It would be counterproductive if people put their youth pensions aside because they are afraid of using them. To have the real freedom to take up youth pensions, one should not have to risk poverty in old age. The fear of old-age poverty is a reasonable one to have. It would be different if persons feared using their youth pension because they will have less later. There will always be an opportunity cost to any choice about using (or not) our youth pensions early in life. It is not sensible to fear opportunity costs in general, but there is reason to fear old-age poverty in particular. Liberal societies have no reason to shield us from the consequences of our choices. But to guarantee the real freedom to choose an option, taking it must not threaten our ability to live a dignified life. To put the point in terms of social insurance, the real choice to take up early-death insurance in the form of youth pensions requires protection against the risk of ending a long life in poverty.

The point here is not that youth pensions would be worse than the status quo for the shorter-lived. Instead, it is that, without old-age pensions, youth pensions may not be better for the short-lived either. Old-age pensions can benefit the short-lived by undermining the ageing fears that would arise if we only had youth pensions. This way, old-age pensions may contribute to real pension equality between people who fear old age but live unequally long.

What about those who do not fear old-age poverty, who might be willing to spend everything early? Does the argument also apply to those who lack such fears? I would say that it must. To allow people to spend everything early in life because they do not (yet) fear a longer life in poverty amounts to giving people a good reason to fear living longer. Doing so does not contribute - indeed, it undermines - the goal of freeing people from ageing fears. The spirit of social insurance should be to protect and free people from risks and fears, not to add to them. Hence, such insurance schemes should not be in the business of giving people good reasons to fear old-age poverty, even if the person is willing to run those risks.

What about those who know they will be short-lived? For instance, what if a person suffers from cystic fibrosis and knows they will not reach old age? ³³ Must they still save for their old-age pensions? So far, we have been discussing cases of uncertainty about longevity. When the risk of early death is predictable, there may be a call for targeted policy measures (see also the ‘Social insurance against premature death?’ section). In this particular case, it may be permissible to open an exception. People who are fairly certain that they will die early might really have no reason to fear living long and may then have a special claim to ‘borrow’ from their old-age pensions. But we cannot be confident about (not) reaching old age in most cases, meaning that letting go of old-age pensions would still give us reason to fear living long. As much as the state should do to protect persons against the risk of premature death, it should not, in virtue thereof, give the younger reasons to fear growing older.

It is rather surprising that old-age pensions can benefit the short-lived. They are in fact necessary to justify early-death insurance in the first place, and to undermine fears of using youth pensions early in life. It turns out that a liberal conception of ‘complete-lives’ equality can, so far, justify a robust concern with two critical life stages: youth and old age. What about the middle of life?

Middle-age pensions

So far, early-death insurance consists of two pillars: one freeing up pensions early in life and another one ensuring old-age sufficiency. Albeit necessary, they are still insufficient as a matter of longevity justice. Specifically, there is also a liberal egalitarian argument for middle-age pensions, which I understand as a pension paid every couple of years that persons can use as they see fit, including for taking months, weeks or days off. I call them ‘middle-age’ pensions because, while the young may benefit from them to some extent, they mainly support the middle-aged. My argument for middle-age pensions is twofold. The first relies upon the (egalitarian) idea that justice in pensions requires some sensitivity to contribution or work. The second is liberal, and it should lead us to endorse middle-age pensions even if we are agnostic on the need for a link between pension benefits and contribution.

The short-longer-lived

The view that claims to social insurance should depend on work or contribution is prevalent when it comes to pensions, which are usually thought of as a reward for a working career. It is widely held that people should receive higher pensions the more they work or contribute. The reasons for such ‘workist’ or ‘contributivist’ intuitions can vary. Some might hold fairness- or reciprocity-based views, according to which welfare recipients should work in return for their pensions to avoid otherwise imposing unfair costs on others. ³⁴ Alternatively, incentive- or egalitarian-based justifications might point out that contribution sensitivity is an incentive that benefits the lowest pensions. ³⁵ Not only might it push people to work more than they otherwise would, but there is also evidence that people are willing to pay higher taxes if they perceive a link between contributions and benefits. ³⁶ The size of pension systems might have been smaller if there had been no such link.

My aim here is not to argue for any particular reason for this intuition, nor is it to offer a case for the intuition in general. ³⁷ What I aim to show, instead, is that taking this view seriously under differential longevity requires middle-age pensions. There are three options for rewarding a life of contribution or work through pensions. One is to reward people before they contribute or work, which is what youth pensions would do. The other is to give individuals the whole reward after they have worked and contributed, in old age. A final possibility is to reward workers throughout their careers, for which middle-age pensions would be necessary.

The first solution faces an epistemic problem: we do not know how much each will contribute or work before their careers begin. Since youth pensions come before or as soon as we start working, there is not yet information about how much individuals will work and contribute and, therefore, how much they ought to be rewarded. Moreover, rewarding people for what they will still do seems odd. Rewards should operate as prizes for what has been done rather than for what is yet to be done.

The second solution is a much better accommodation of ‘workism’ or ‘contributivism’. If we all lived equally long, old-age pensions could potentially be as sensitive to work or contribution as middle-age pensions. But this is no longer so once we acknowledge the short-longer-lived – people who live short enough to die before old age but long enough to work and contribute for many years. The argument would go as follows. Typically, longer-lived workers contribute or work more than shorter-lived ones (P1). Under ‘workism’ or ‘contributivism’, a person contributing or working for longer should receive more than one who has contributed less (P2). The short-longer lives should, then, receive more than those who die very early. Middle-age pensions seem, then, necessary to give the short-longer-lived their due. Imagine:

Shorty & Middy began working at the same age. They can only receive a pension when young and at 65. Shorty lives to be 35 years old, whereas Middy lives to be 60.

As a short-long-lived person, Middy dies close to old age and, as a result, ends up not receiving their old-age pension despite a whole life of contribution. Neither Middy nor Shorty reach old age, but Middy has also worked and contributed for much longer. Under ‘contributivist’ or ‘workist’ principles, Middy should receive more. Youth pensions do not do this because they give the same to both. Nor do old-age pensions, which benefit neither of them. What can make the difference would be introducing middle-age pensions on top of youth and old-age pensions.

Middle-age pensions ensure that the short-longer-lived will be compensated for their longer working careers. They will do so by ensuring they receive a certain amount in pensions every T amount of time (which can be measured in terms of contribution or work). We should be flexible about this T, as it should be possible for middle-age pensions to come with many different fillings. What we can say is that T should neither be too short nor too long. If it is short, the size of the pension will likely be insignificant because we will be receiving it far too frequently. If T is an extended period, say a decade, the risk of the short-longer-lived missing out on their reward increases. We do not need to commit to a particular T in this essay, since there may be many reasonable specifications within these limits that will justify insurance against a short-longer-life in the form of middle-age pensions.

Should people be free not to use, and instead save up, their middle-age pensions for later? It seems that everything we said about storability when discussing youth pensions applies here, too. There is reason to let people adjust middle-age pensions to their specific life plans, which only they know. Perhaps storability should always be allowed, even in old age. If one can still work and wants to gamble on the chance of receiving higher monthly benefits later, I see no reason to prohibit that on behalf of longevity justice.

Revising conceptions of the good life

A central tenet of Liberalism is that people can form and pursue their conceptions of a good life and be able to revise their pursuits. Revision is not simply about updating one´s beliefs about what choices are worthy but also about having the chance to put those beliefs into practice. The emphasis on revision tames the problem that arises when increasing the ability to pursue certain conceptions at one earlier point comes with severe opportunity costs for the ability to pursue other plans at later points. Older selves must be able to reverse some decisions made by their younger selves. So far, however, youth and old-age pensions do not preserve the freedom to change one's life, as those who spend all their benefits early may have to wait for old age to get another opportunity for a fresh start. ³⁸ Middle-age pensions can then take up the liberal role of ensuring that, to some extent, we all have the chance to reshape our lives, especially the short-longer-lived.

We are not like Benjamin Button: we are born young and grow older over time (whereas Button aged in the opposite direction). For this reason, young selves are first-comers who may appropriate too much. It is necessary to set up constraints on how much freedom young selves may have if we conceive of persons as beings who preserve and maintain their liberty to revise and alter their ends (e.g. Rawls, 1971: 131). Free persons must guarantee a reasonable degree of freedom to older selves, one that allows them to overcome earlier decisions. Some short-lived persons will not reach middle age, but many others will. Middle-age pensions will then be necessary, especially for those living longer but not enough to enjoy an old-age pension with which they can revise their conception of the good.

There might always be a time when we need to rethink and change our lives. Our capacities to do what we want, and our ideas of what to do, change with age. This affects us all: regardless of how long we live, anyone might need to revise their conception of the good life before old age. Yet, since the longer-lived age more, they are more susceptible to the need to change their pursuits. Those dying younger will not need to revise their views of the good life as much as those who live longer. As a result, the short-longer-lived will require more protection against the need to change their life plans than their shorter-lived peers. In this respect, youth and old-age pensions will not be enough to ensure that we have a real ability to revise our conception of the good life all along our lives. Only middle-age pensions can secure that.

While youth pensions increase our real freedom to pursue any conception of the good life, middle-age and old-age pensions are necessary to ensure that we can revise it eventually. There is a trade-off here, internal to liberalism, between promoting the ability to pursue any life plan by allocating goods at the youngest ages and guaranteeing the ability to revise that plan eventually by saving resources for later. ³⁹ A sensible liberal answer to this trade-off must give weight to both abilities. That will tell us how much weight a liberal scheme should give to middle-aged pensions vis-à-vis youth pensions.

Conclusion

This article argues that, contrary to what may be thought, lifetime distributive equality is not at all indifferent to how we distribute goods across ages. ⁴⁰ There is a liberal egalitarian case for social insurance against early death that justifies a distinct concern with all ages of life: youth, middle, and old age. I suggest calling a system that gives a storable pension to all three life stages ‘freetirement’. Freetirement is a policy that gives each age group the freedom to draw a storable pension. This word bears a close resemblance to the word retirement, which goes back to retire, to withdraw or drawback from somewhere (typically the labour market). ⁴¹ In contrast, ‘free tirer’ is a made-up but phonetically sound name for the idea of drawing something freely. It does not suggest any removal from the labour market, and is well in line with the freer distribution of pensions I am proposing.

A realistic implementation of freetirement would postpone old-age pensions by a certain X number of years. Half would be given as a capital grant or as a paid free time voucher early in life, which one could use or save for later. Workers could receive the other half of X over time, throughout their careers, as a capital grant ⁴² or a paid free time voucher. To determine this X, we must consider what is feasible for older workers, and what is the overall resource constraint of each pension system. Since old age often comes with increasing impairments, some workers might not be able to work much longer than they currently do. It seems impermissible to set X at an age after which asking older people to work longer would put them through significant adversity, akin to an old-age hell. The difference between that age and the standard retirement age will tell us how many years may be freed up for pensions earlier in life.

This point answers economic worries that there might not be budgets for freetirement. The answer depends on the priority we give to each stage of life and to each argument. Old-age sufficiency may take precedence in many cases, if I am right that it is necessary to justify early-death insurance in the first place. So the case for freetirement becomes more robust as old-age poverty is dealt with, and the most basic old-age pensions have been financed. As pension schemes become more generous than that, which they often are, the stronger is our reason to add youth and middle-age pensions to old-age pensions.

For financial reasons, it might be impossible to give everyone a pension at the same time. ⁴³ When unfeasible, we could follow the example of banks, where it is impossible for us all to withdraw our money simultaneously. Just as banks can and must have regulations to prevent this, pension schemes, which are also ‘too big to fail’, must set up various regulations to solve coordination problems between claimants. If a modern state sees the value in having banks knowing that problems such as bankruptcy can occur, they should also find value in freetirement even if we cannot all be pensioners simultaneously.

It is also important to stress that freetirement seeks no guarantee that one will choose to draw all three pensions. We should not expect everyone to prefer buying early-death insurance through youth or middle-age pensions. It is not worrying if some prefer, as they will, to strengthen their old-age pension. What matters is that there is access to pensions early in life, regardless of what one ultimately chooses. It might turn out that no one takes up their youth and middle-age pensions. Everything would then stay as it is now, with one significant difference: individuals would still have the choice to protect themselves against one of the most tragic things that can happen to them. This is already a considerable change in our lives, especially for the short-lived among us.

Many societies today are older than ever before. The widespread reaction to this fact has been to postpone old-age pensions over time, which suggests that it is feasible to ask people to work longer. Freetirement welcomes, if not requires, such policies. However, it does so on the condition that the proceeds be used to spread pensions more evenly and freely across life. If it was once affordable to pay the pensions of the elderly when the demographic pyramid was suitable, youth and middle-age pensions are financially sensible alternatives for ageing societies, where the pyramid reverses, and there are too few young and middle-aged per older person. Freetirement adapts well to our changing reality. Especially in ageing societies, where it is most tempting to forget those who were once visible in their youth but soon disappear while growing older.