Working Retirees? A Liberal Case for Retirement as Free Time

Abstract

Retirement is often viewed as a reward for a working life. While many have reason to want a work-free retirement, not everyone does. Should working retirees have to give up their retirement pension and, consequently, their status as retirees? The answer, I argue, boils down to whether we conceive of retirement as free time (need-free) or as leisure (work-free). In this article, I put forward a liberal case in favour of free time, despite whether our liberalism leans towards perfectionism or neutralism, with social primary goods being a case in point. Applying this case for free time to retirement yields two significant policy implications. First, it demands “free synchronic combination” – that retirees may use their retirement pensions however they see fit, including to work. Second, it also yields “free diachronic combination” – that, within limits, individuals have discretionary control over how to combine retirement and work across time – thus challenging the idea that retirement should be available only in old age and not earlier in life. So far, the literature on free time focused only on narrow temporal units, such as hours and days, but there is much to gain by extending the concept into retirement.

1 Working Retirees?

Retirement is often viewed as a reward for a working life. With so many working lives being frequently exhausting, it is understandable that many workers are eager to receive it. But this does not mean that we are all equally thrilled about retiring. To some, it may trigger fears of a “permanent vacation”. While some dream of a work-free retirement, many others will want to work – to continue old projects, begin a new working life, or keep contributing to their society.¹ Many are too passionate about work to picture a good retirement without it. To ensure that retirement is also a reward to them, a society could adopt one of two policies,

• A. Free Diachronic Combination – Let individuals choose how to combine work and retirement over time. They might work until as late as possible or go back and forth between work and retirement.

• B. Free Synchronic Combination – Let individuals combine retirement with work at the same time, i.e., use retirement pensions to keep working until as late as they want.

These are distinct policies. Policy A leaves us free to arrange our retirement in various ways. For instance, it allows us to work far beyond the retirement age or retire for a couple of years with a full pension and return to work afterwards. As flexible as it might be, this policy is still compatible with reductions in our pension benefits to the extent that we work. In general, it still asks us to decide between working and retiring. By itself, it does not allow working retirees to use their standard retirement pension plan to work in the labour market, with or without a wage. We tend to present retirement benefits as conditional on earlier years of work and on reaching a certain age. But we often forget the third requirement that one should stop working to receive a pension. This requirement is challenged by B, which commits to the view that work should be no threat to our pension benefits and, consequently, to our status as retirees.²

In this article, I make the case for free combination, the get-together of A and B. I show how the seemingly trivial distinction between leisure and free time and their relation to Liberalism bears on this policy. Leisure and free time inspire two different conceptions of retirement, with implications for what retirement schemes should do. If retirement is about leisure, and leisure is the absence of work, there can be no such thing as working retirees (∼ B). Such rules are soft reminiscences of mandatory retirement, with the notion of forced leisure being implicit in the meaning of compulsory retirement as forcing retirees out of the labour market.³ If retirement is instead conceived as free time, as I argue it should be, there is nothing contradictory about working retirees. As free time is not about “work-freeness” but about discretionary control over our time,⁴ it can justify both A and B. So far, the philosophical literature on free time focused only on narrow temporal units (e.g., hours, days).⁵ This article expands the discussion to years of free time, thereby vindicating the importance of retirement. Given how expensive it has become due to population ageing, it has never been more pressing to defend retirement against those who insist that we would be better off without it.⁶

There may be several reasons for free combination. For instance, preventing retirees from working is likely an unnecessary waste of productive capacity. So, we might have efficiency-based reasons to allow for either A or B. To this, we may add egalitarian or prioritarian reasons if, for instance, efficiency gains will raise the lowest wages or pensions. These might be good reasons for retirement as free time, and I will say more about them in the conclusion. In this paper, I focus on the contribution that liberal arguments can bring to this debate, which can hardly be overstated. By liberal, I mean an argument that appeals to assisting individuals with their ability to pursue, maintain, and revise their conceptions of the good life, whatever these may be. We must be able to revise our life plans, especially when previous plans are no longer desirable or feasible for us, as often happens in old age. Yet, I take it that liberals should also be concerned with the ability to retain plans (i.e. maintenance). Retirement can be stressful. Given how much it could change our lives, we should not feed into fears that retirement must be away from work, but instead let retirees continue working while still receiving their full state pension.⁷

The article proceeds as follows. I begin by distinguishing between leisure and free time (§ 2). Afterwards, I offer a liberal argument in favour of free time over leisure, regardless of whether our liberalism leans towards perfectionism or neutralism, with primary goods being a case in point (§ 3). Applying this case for free time to retirement yields two significant policy implications. One is free synchronic combination, the view that retirees may use their retirement pensions however they see fit, including to work (§ 4). The other is free diachronic combination, which asks individuals to choose, within limits, when to retire and when to work, even before old age (§ 5). In Sect. 6, I conclude that, unlike with leisure, this liberal case for retirement as free time vindicates free combination.

2 Leisure and Free Time

The word leisure can have two meanings. It can be understood as the portion of time we do not devote to work. The definition of leisure as work-freeness is widespread, especially among economists.⁸ However, in some languages, as well as in the etymology of leisure, we are told that leisure can also be seen as “time at one’s disposal”.⁹ This idea of “time for what we will” aligns with what the literature in political philosophy now refers to as free time.¹⁰

Though it is common to use free time and leisure interchangeably,¹¹ free time can be better understood as need-free time – time over which we have autonomous control, after having satisfied the demands of necessity.¹² It is the time left after we discount the time necessary spent in various dimensions; typically, paid labour, household labour, personal care.¹³ These are unlikely the only dimensions. For instance, necessary time away from paid labour, of the sort that retirees might be subject to, also constrains their free time. For many retirees, leisure can be a demand of necessity. Our time is not only unfree when it is necessary for us to work. It is also unfree when we must steer away from it.

For this reason, it would be a mistake to see free time as “work-free”. Doing so fails to capture two situations. First, some limitations to our freedom are not work-related. For instance, sleeping and eating are usually work-free. But they are typically not need-free, insofar as we need to do them to survive. Second, we can use our free time to work.¹⁴ For instance, the time of working retirees is not work-free, but it is well described as time free from necessity. An interesting difference between need-free time and work-free time (leisure) is that, unlike the former, the latter always carries an “opportunity cost”. While work always takes away from your time, need does not do so in the same direct sense.¹⁵ In line with the literature, I take it that it is intuitive to use the labels leisure and free time to refer to work-free time and need-free time, respectively.¹⁶

Understanding leisure in terms of work-freeness raises the tricky question of what work is.¹⁷ While work is often understood as paid employment, the idea that work must be paid is problematic for a number of reasons. Many throughout history have not been paid for their work, especially women. Our reaction to such cases is that they should have been paid for it, not that they were not working in virtue of not having been paid. It is not easy to come up with a plausible definition of work. So one potential advantage of free time is that it requires no such definition. What is ultimately relevant for free time is not how we end up using it. This is sometimes an advantage, but not always. Consider, for example, what Walvin (1995) wrote about free time under slavery,

“The central problem in assessing slaves’ free time is the fact that much of that free time was granted to allow them to work, in their gardens and provision grounds. (…) This scarcely seems like leisure. (…) [A] great deal of slave time freed from the planters’ scrutiny was, nonetheless, destined for work. The freedom to work (in order to eat) sits uneasily with most accepted definitions of leisure.” (p.5).

“In societies where masters expected a great deal of the slaves’ freetime to be put to fruitful use, to the benefit of both master and slaves, it is hard to see how slaves’ free time can readily be equated with leisure. Only in the sense that it was time at the slaves’ own disposal does it come close to leisure time.” (p.11).

Free time might be open to abuse because, unlike leisure, it includes the freedom to work. Many unions are often not enthusiastic about working retirees for a similar reason. They suspect the reason why retirees work is that their pensions cannot sustain their living standards.¹⁸ In these cases, however, a more sympathetic reading through the lenses of free time is possible. We could insist that time is not free from need if one still needs to work after a working day, or working life. Having the real freedom to enjoy leisure is also necessary, but not sufficient, for free time.¹⁹

In this section, I distinguished between leisure and free time. Free time requires both the freedom to work and leisure, while leisure concerns itself only with the latter. If so, it might be obvious that liberals will welcome free time, since it seems to enlarge individual choice by simply adding work-related options to leisure. Although there are cases of leisure without free time – such as forced leisure – their compulsoriness detracts from their appeal. There is indeed a close connection between free time and Liberalism. As we shall see, their connection may even be more robust than it already appears.

3 Liberalism and Free Time

Before applying the distinction between leisure and free time to retirement, I should say something about its connection with Liberalism. Crucial for my argument is the suggestion that Liberalism is generally much more aligned with free time than it is with leisure. In this section, I show that both perfectionist liberals (2.1) and neutralist liberals (2.2) have reasons to protect need-free time, while there is not necessarily such a converging case in favour of work-free time.

3.1 The Perfection of Free Time

Some liberals are said to be ‘perfectionists’ in the sense that they rely on a thick theory of the good: their political principles are built upon metaphysical and ethical doctrines concerning the nature of value and the good life.²⁰ Perfectionist liberals can defend either free time or leisure, depending on which is best included in that doctrine.

One example of a perfectionist account of leisure comes from Aristotle, who saw leisure as the cultivation of mind, spirit, and character – “the goal of all human behaviour, the end toward which all action is directed”.²¹ Aristotle thought leisure was of great importance, but had quite a specific idea about how people should use it. But if Aristotle thought this, what would he say about free time? It seems that free time preserves all the relevant benefits of leisure in terms of the cultivations Aristotle was concerned with. To this, free time adds the condition that people can work, which appears more suitable to advance the full development of a free mind, spirit, and character with which Aristotle was concerned. On a plausible interpretation, Aristotle conceived leisure as time free of necessity, which resembles what I mean by “free time” (Kazez 2023, 529). Interestingly, Aristotle’s view of leisure was also not too different from the perfectionist way Marx (2005, 631) describes free time. Consider,

“The saving of labour time [is] equal to an increase of free time, i.e., time for the full development of the individual (…) It goes without saying, by the way, that direct labour time itself cannot remain in the abstract antithesis to free time in which it appears from the perspective of bourgeois economy… Free time – which is both idle time and time for higher activity – has naturally transformed its possessor into a different subject, and he then enters into the direct production process as this different subject.”

Along perfectionist lines, Marx considers free time to be “time for higher activity” and “for the full development of the individual”. Perfectionist liberals should agree that need-free time is more conducive to such development than work-free time is. Liberals can also agree that, ideally, free time should not be opposed to labour time. Indeed, the development afforded by free time may require labour time. The disagreement with Marx might be about whether labour time in a “bourgeois economy” can foster that development.

Of course, none of the authors above is liberal in the sense specified in this paper. But the discussion nevertheless applies to perfectionist liberalism, according to which political morality is characterized by a commitment to helping individuals lead autonomous lives and make valuable choices.²² If the liberal state is to help individuals live up to the pre-established ideal of what a valuable life is, and work has a place in that ideal, as it should, then the liberal state should be more engaged in giving its citizens free time rather than leisure.

3.2 The Neutrality of Free Time

Not all liberals are perfectionists. Many are neutralists, as they do not take a position on what a good life necessarily involves. They rely on a thin theory of the good, which identifies goods one needs for whatever rational plan in life one may have (Rawls 1971, 380). What is essential here is the capacity and means to formulate and act upon a conception of the good, whatever it may be.

To suggest that leisure will violate liberal neutrality and that free time is compatible with even the most neutral theories,²³ I employ the list of social primary goods. These are the goods any free person should need and want in life, whatever their conception of the good. What makes this case study insightful is that the most famous proponent of this list, John Rawls, suggested counting leisure as a primary good as an amendment to his first proposal.²⁴ The change was met with criticism,²⁵ which perhaps Rawls could have avoided if he had added free time instead. I do not defend in this article that free time should be a primary good, but I must still explain why leisure violates the neutrality test for a primary good.²⁶

Consider the widespread intuition that the involuntarily unemployed are victims of bad luck. How can involuntary unemployment be an instance of bad luck if leisure is a primary good – a good that we are all presumed to want, no matter our final aims? The involuntary enjoyment of other primary goods, like income, wealth, and self-respect, is typically seen as good luck. With all other primary goods, having more of them through no fault of our own is considered good luck. But we do not have the same intuition regarding leisure. Even if the involuntarily unemployed do not fare worse in other respects (say, economically), we still have reasons to regard them as in one way worse off. What explains this difference between leisure and other goods is, I suspect, that work-free time is not a primary good. Leisure time does not pass the neutrality test for anyone who plans to work. In contrast, free time offers a straightforward explanation of why the involuntarily unemployed are victims of bad luck: forced work and compulsory leisure are bad because they both constitute necessary time.

The aim here is not to argue that free time is a primary good, but to show that it reinforces our access to other already acknowledged goods. The most notable among them are basic liberties. According to Rawls (1982, 15 − 6), these are necessary to safeguard peoples’ agential capacities to form, pursue and revise their life plans. The paradigmatic examples would be freedom of conscience, free speech, free association, and political liberties. While leisure and free time might not have a significant impact on the first two, they do promote associational freedoms and political liberties (Rose 2016b). Sometimes, free time and leisure coincide, as when we spend time with family or take time off to vote. Other times, we may need free time to pursue activities better described as work, such as forming organisations with others or running for public office.

Next in the list of primary goods, there is occupational choice. This item may be where the difference between leisure and free time is the clearest. While both promote the ability to engage in non-work-related activities, free time enables any occupation and career path. In large part, a significant benefit of free time over leisure is the much wider set of occupational options it provides. After occupational choice, there is freedom of movement. Here, too, both free time and leisure help people move around more freely (Eurostat 2021). How much one adds to the other will depend on how many of us would use our freedom of movement to work in other countries.

Other notable items on the list are income and wealth. Economic resources are often necessary for us to have freedom to enjoy leisure. That is why the time of enslaved people, whose income was too small to enjoy leisure, was not free in a real sense. The more economic resources we have, the more control we have over our own time.²⁷ Also here, free time has an advantage over leisure. Given how many cash-generating activities it curtails, leisure will make it much more complicated for workers to increase their levels of income and wealth in life.

Finally, there are the social bases of self-respect, which purport to make us feel that our plans in life are of value and worth carrying out (Rawls 1971, 386). Free time promotes our sense of self-respect more than leisure; again, it provides us with the means to carry out any plan in life, regardless of whether it belongs to work or leisure. Free time offers a more robust basis of self-respect because, unlike leisure, it signals that more of our goals are of potential value and worth carrying out, not just those involving abstention from work.

To sum up, free time is more closely aligned with Liberalism than leisure is, regardless of whether our liberal principles lean more towards perfectionism or neutralism. Free time gives us much more of what liberals value than the freedom to leisure does. The claim may seem obvious when stated so simply. Much less obvious are the insightful implications arising from its application to retirement. I will now examine two respects in which retirement as free time differs from retirement as leisure. Taking these policy features together, it becomes clear that, unlike leisure, free time supports a retirement scheme that embodies free combination (A + B).

4 On Synchronic Combination

Entitlements to pension benefits tend to be conditional on career length and on reaching a certain age. But we often forget a third requirement: the idea that one should stop working to be entitled to them. In this section, I argue against the requirement that pension benefits should be conditional on renouncement to work.

Suppose I work as a carpenter, and that I can and want to work until I am 80 years old. As we saw at the beginning of the paper, two policies are available to me. One is free diachronic combination, which allows me to go back and forth between work and retirement as freely as I wish. The only condition would be that working comes at the expense of a reduction in pension benefits. As a carpenter, I could stop working for two years (with a full pension) and then start working again (losing my entitlement to benefits while I am employed). A smoother transition is also possible, as one could slowly phase out their career for ten years and get a pension that decreases the higher the percentage of time dedicated to work. There is room for different kinds of transition between work and retirement.²⁸

Since this policy carries no obligation to leave work, one might wonder why it is insufficient. In this section, I argue that it is: taking the liberal case for retirement as free time seriously also requires free synchronic combination (B). The policies above are still conditional on the requirement that retirees renounce work once they retire, thus preventing them from using their retirement benefits to work. Against this restriction, I suggest that a liberal retirement scheme should render retirement benefits unconditional on how they are used, thus allowing our retirement to take up any content.

The conventional way of showing concern for the life plans of retirees who want to work has been to adopt free diachronic combination (often known as voluntary retirement).²⁹ The idea has not been conceived as retirees being entitled to retirement benefits early in life but, rather, as remaining entitled to benefits even if they work beyond the retirement age. Such a diachronic combination offers the real choice of whether to retire or work until as late as one might want. Still, it does not entail that people can be working retirees. This is not a problem for views of retirement as leisure, since these see no value in having retirees use their pensions to work. Such views do not give us reason to go beyond diachronic combination. But conceptions of retirement as free time do, as they also imply synchronic combination.

Even if a society implements diachronic combination, there is still a sense in which a retiree´s liberty remains unduly constrained. There are several reasons why retirees may want to work without losing their pension. Consider the case of a cake baker – who cannot conceive of doing anything else but baking; the problem is that, with age, it gets harder to bake as efficiently as before. They could hardly have enough if they depended economically on the cakes they sold or were paid for their work at market wages. Without synchronic combination, the baker must choose between (a) not working and receiving a full pension or (b) working at a salary that significantly reduces their living standards. The choice becomes more difficult the stronger their wish to continue working on activities that can hardly sustain their living standards (with unpaid work at the end of the spectrum).

So far, this assumes that the cake baker’s market wage would be low. But what if the job were well-paid? If the job pays below the baker’s pension level, we can easily see why the baker might be interested in using their pension to work. The point is trickier if the job pays better than their pension. In such cases, it is important to distinguish between the freedom of retirees to work at a given job and the right of retirees to earn the same income from that job as non-retirees. Policies like synchronic combination only require the former, not the latter. They are compatible with retirees, perhaps irrationally, choosing to work at a (lower) pension level rather than a higher market wage. To be sure, nothing of what I say implies that working retirees have a right to the same market wages as non-retired workers. Employers must pay working retirees as, otherwise, retirees would acquire an unfair competitive advantage. Meanwhile, there may be reasons to tax working retirees more heavily and use fiscal revenues for worthwhile aims (e.g., minimum wages and pensions). Still, we must be careful here because the more we do so, the more being a working retiree represents a loss to our option sets.

These worries raise the question of how to balance, on the one hand, an inequality that can emerge between working retirees and non-retired workers and, on the other hand, the idea that a liberal retirement system should not penalize retirees for conceptions of the good life that involve working. For this reason, I doubt pension levels should decrease simply because a retiree works. The liberal case for retirement as free time requires more neutrality between work and leisure, which is undermined if retirees see their pensions shrink simply because they work. It is more neutral to say that the pensions of working retirees remain untouched as long as they are retired. This does not give them the right to earn the same wages as non-retirees on top of their pensions. Working retirees may be taxed at higher rates, depending on the income and wealth they earn from work rather than on the bare fact that they work.

So far, we assumed that diachronic combination was in place. But to see the value of synchronic combination more clearly, suppose that retirement is mandatory. A strong version takes mandatory retirement as a general obligation by employment law to disengage with the labour market. Much more common are weaker versions, where retirement is “mandatory” by being included in long-term employment contracts established by employees and their employers, mandating that the former leave that company after a certain number of years.³⁰ The weaker version does not seek to remove workers from the labour market, but only from a specific job.

While strong versions will cause the loss of countless options, weaker ones leave our options virtually intact. Weak retirement is much less objectionable than its “stronger” counterpart because of its negligible impact on retirees’ free time. Mandatory work-free retirement is always regrettable because it can never be an inherent feature of the good life: it assumes a life in which individuals will have to disengage from the labour market.³¹ In contrast, compulsory free time is immune to these criticisms because it forbids constraints on the content of our retirement, allowing us to use this time as we see fit, including to work. In this view, retirement would still be valuable even if it were involuntary and did not allow for diachronic combination (A). It could still be valuable because synchronic combination (B), by itself, protects retiree’s freedom over their own time.

Mandatory retirement as free time can be an inherent feature of the good life because it sets no constraints on the content of our retirement: one can use it to lay in bed, travel around the world, or invest in a new occupation. Contrast this with retirement as leisure, which fails to protect retirees’ freedom to work. The prohibition of retirement plans that involve working is pro tanto impermissible on liberal grounds. The loss of options that it brings about depends, of course, on how loosely one defines work. The broader the definition, the more activities retirement as leisure may be precluding and, consequently, the less desirable it is from a liberal standpoint. Conversely, if our understanding of work is narrow – say, for instance, it only includes paid jobs in the labour market – empirical claims about retirement being mostly about leisure will grow stronger. But as leisure includes a wider range of activities, there will also be fewer cases of work to, in part, render the distinction between free time and leisure salient.

In this section, I argued that the liberal case for retirement as free time requires synchronic combination. What about diachronic combination?

5 On Diachronic Combination

Understanding retirement as free time also bears on diachronic combination as a policy. It demands that, within limits, we extend people’s control over how to combine retirement with work across life. Here, too, free time delivers a position distinct from leisure and one that yields a much more liberal retirement scheme.

There is no better illustration of this than what is known as the Fisherman Story.³² Once, an American saw a fisherman in a small boat and asked how long it took to catch the several fish inside the boat. ‘Only a little while’, said the fisherman. Curious, the American asks, ‘why did you not stay longer and catch more fish?’. The angler replied that he spends time with his family and friends once he has enough fish. The American suggests the fisherman could use this extra time to catch and sell even more fish. The fisherman, curious as to why he should do that, asks what he should do with the extra money. The American replies, ‘Well, eventually you retire and use it to spend time with your family and friends’.

The Angler and the American differ in their conceptions of how to distribute free time across life. One wants to preserve daily free time at the expense of a less frequent but larger portion. The other hoards their free time by working hard during the first two-thirds of their life. Each arrangement allows us to do different kinds of things. As liberals, we cannot assert that one is necessarily best. Indeed, it is not hard to see that there is a pro tanto liberal case for neutrality between them. The liberal case for control over our time includes discretionary control over how we answer this question for ourselves (diachronic combination). Each may be entitled to an equal amount of retirement and be left to decide how to divide it according to their conception of the good life.

There are liberal grounds for thinking that, once eligible for pension benefits, we could store them for later and spread them across life however we see fit. More ambitiously, there can also be a case against key constraints on diachronic combination, such as age and career length requirements. So far, arguments against allowing for only old-age retirement often appeal to egalitarian justice between workers who die young and those who live long.³³ The importance of a liberal argument in this debate has been underestimated. This is especially unfortunate, given that such a liberal argument will hold even if all lives are of equal length.

Retirement schemes that prevent most workers from retiring before old age are likely to fail the neutrality test, as they intend to favour conceptions of the good in which years of free time are only valuable in old age. When governments distribute goods, it is liberal to do so early in life and permit them to be saved for later (storability).³⁴ That way, those wishing to retire at older ages have no liberal complaints since they can save their share for later. The main beneficiaries will be those whose plans involve retiring younger, who would now have much more control over their free time than if it were otherwise only given to them in old age.

As liberals, we should not act on the view that individuals must only value years of free time in old age. This is not, however, to oppose valuing free time in old age more generally. While there is a general liberal case for trade-offs, limits to those trade-offs can also pass the neutrality test. One possibility is to appeal to the ability that, as free persons, we must have to revise our ends (e.g., Rawls 1971, 131). Our conceptions of the good and our capacities to act on them may change with time. Especially if being old impacts what is good for a person and what goods one can feasibly pursue.³⁵ Given what we know about old age, we may expect to become unable to work, or else, that our productivity decreases such that state assistance will be necessary. On a realistic view of what ageing entails, it is sensible to guarantee protection for when these risks materialise. The problem with people like Angler is that they fail to preserve the ability to revise their life plans in the face of old age or disability, something which should remain available to free persons. It is compatible with neutrality that, in helping people live the best life for them, we not only consider their plans when they are young but also take seriously the claims of their older selves to revise those plans.

A second liberal justification can appeal to the ability to maintain conceptions of the good life (as opposed to revising them). The fact that our abilities change does not always give us a reason to change our lives. The case of the cake baker illustrates this. Though this person could use retirement benefits to revise their ends completely, they want retirement for the opposite reason: to retain their life plans. Back to the distinction between leisure and free time, we can say that leisure recognizes the first (revision-based) justification to provide old-age retirement. Since much of our plans before retiring involve work, leisure makes the most sense if we intend to revise our ends. It makes much less sense as a reason to maintain lifestyles that include working. The concept of free time can, in turn, be used as a ground for both revising and maintaining a given conception of the good life.

For different reasons, leisure and free time may both justify retirement in old age. Indeed, their most divisive case concerns younger persons. I suspect retirement as leisure will be quite restrictive here. Someone who has worked hard for a week surely needs a few days off (i.e., a weekend). Sporadically, there are reasons to give that worker a holiday. But unless that worker cannot work for a long time, there is no reason to give them the right to retire. Claims grounded on leisure will attach to the (in)ability to work. As a result, they will justify shifting retirement to only when individuals can no longer work. Given how this ability tends to evolve, retirement as leisure will most likely locate its years on the extremes of life – old age and childhood – as it is at these ages that freedom to leisure is most needed.

Retirement as free time offers a less restrictive position. It entails limits that protect us against the need to revise or maintain our life plans. However, once these limits have been secured, it aims to minimize restrictions on diachronic combination, to protect people´s discretionary control over their own time. Under free time, it is up to individuals themselves to locate their retirement years as they see fit, provided they abide by these limits. This conception of retirement is much less biased in favour of old age because it recognizes several reasons why people may need years of free time, of which the inability to work is only one. These reasons are more evenly spread across ages than the inability to work is. For instance, the pressure to enter the labour market full-time upon completing education may mean that many young adults are forced to abandon valuable occupations, projects or lifestyles.³⁶ Unlike with leisure, retirement as free time extends diachronic combination (A) to many more age groups than just the elderly.

To conclude, I should mention a possible tension that may arise. On the one hand, I insisted on liberal reasons to shift goods earlier in life to promote our ability to pursue any conception of the good life. On the other hand, I also offered free-time-based reasons to move some of those goods for old age. There is a tension between promoting the ability to pursue whatever life we want – which pulls in the direction of youth – and ensuring that we can revise or maintain that life – which yields a concern with older ages. The more resources we receive at young ages, the more able we are to pursue any life plans, including the possibility of squandering them completely, leaving our older selves destitute.³⁷ Any liberal view needs to find a balance between both sides. Any sensible account must include a concern with the pursuit, revision and maintenance of our plans.³⁸ That liberal account of retirement is well served by distributing years of free time more evenly across ages.

6 Conclusion

The seemingly trivial distinction between leisure and free time makes an essential difference in how we should conceive retirement. Retirement should be free-time-allowing rather than leisure-allowing out of a concern with the ability to pursue, maintain, and revise any plan of life we may have. Free time, understood as control over our (retirement) time, requires enabling both synchronic combination – that retirees may use their retirement pensions however they see fit, including to work – and diachronic combination – that, within limits, individuals have more discretionary control over how to combine work and retirement across time – thus challenging the idea that retirement should be available only in old age and not earlier in life.

Free-time-based views should provide more flexibility than leisure-based ones because there is more discretionary control in free time than in leisure. A life of leisure involves a certain way of living, whereas a life of free time can be of any content whatsoever. Liberals should side with free time whenever there is a side to take. This does not mean, however, that only liberals will agree with the policies I discussed in this paper. One could reach a similar conclusion on behalf of efficiency and equality. Combination-based policies will likely render pension financing less costly, and the revenues may be used to promote more egalitarian distributions. Beyond distributive equality, one could expect relational egalitarians to agree with free combination, a less intrusive and paternalistic policy that avoids segregating retirees from a crucial part of social life. Even in terms of productive justice, this policy can be used to cover holes in particular parts of the labour force. None of the reasons mentioned now is necessarily liberal. Still, all of them can complement the liberal case for a more flexible retirement system. As liberal societies age and the percentage of retirees per worker increases, it has never been more important to be reminded of the place that retirement and free time hold in them, and even more so together.