Free Time Across the Life Course

In most industrialised countries, citizens enjoy a very large amount of free time towards the end of their lives, when they are retired, but find it very costly to access free time during the middle part of their lives. This is concerning because those who die early are deprived of the reward of free time that retirement holds. Extreme discrepancies between a time-rich old age and a time-scarce middle age are not, however, inevitable: some states incentivise long work hours during middle age in combination with early retirement, whereas others incentivise shorter work hours during middle age and later retirement. This variation raises the thus far unexplored question of how a just society should design policies that affect the costs of access to free time across the life course. I answer this question by using a hypothetical decision-situation where prudent choosers must allocate access to free time across different life stages.


Introduction
Many people in industrialised countries feel pressured for time. For example, 48% of Americans say they don't have enough time and 35% say they're always rushed (Newport, 2015). One explanation for this is that there are big differences in the amount of free time different individuals enjoy (Hamermesh, 2018;Robinson, 2013). For example, single mothers with low income have much less free time than high-earning childless couples (Goodin, 2008: Ch. 3). These interpersonal discrepancies in free time raise interesting questions of social justice to which political philosophers have recently devoted some attention (Rose, 2016). But there is also an intrapersonal reason for why people can lack free time that is often overlooked -that is, a reason that has to do with how free time is distributed across different stages within people's lives. 1 In industrialised countries, citizens typically enjoy a very large amount of free time towards the end of their lives, when they are retired, but it is much more costly for people to access free time during the middle part of their lives (Eriksson and Rooth, 2014;Eurostat, 2003;Nunley et al., 2017;Pedulla, 2016;Weisshaar, 2018). To illustrate, workers are often reluctant to interrupt their careers because re-entering the job market after a lengthy break from work can be very difficult. This intrapersonal discrepancy between a time-scarce middle age and a time-rich old age provides another reason for why so many people in industrialised countries feel pressured for time.
The fact that middle-aged people face high costs in accessing free time is by no means natural or unavoidable. States can choose between different regulatory regimes that expand or reduce citizens' opportunities to access free time at different life stages. At one end of this spectrum are conceivable regimes that incentivise citizens to work as much as possible during middle age, in order to retire as early as possible. Under such regimes people might have an average work week of 70 hours and an average retirement age of 50. We can refer to a regime that tends towards this extreme of possible distributions of free time across the life course as a postponement regime because it incentivises citizens to postpone their enjoyment of free time until they're retired. 2 At the other end of the spectrum are conceivable regimes that incentivise the middle-aged to enjoy as much free time as possible, although this means that they must continue to carry out some work until they're quite old. An example would be a regime with an average work week of 25 hours and an average retirement age of 75. Let us refer to a regime that tends towards this pattern as a frontloading regime because it leads citizens to 'consume' a large part of their lifetime share of free time when they're still relatively young.
To illustrate the distinction between postponement and frontloading tendencies, consider the regulatory regimes of Japan and the Netherlands. Japan's economy encourages workers to postpone the enjoyment of free time until old age. 3 It does this, for example, by rewarding workers who don't interrupt their career with wage increases -a practice that is often referred to as 'seniority-based pay '. 4 Beyond that, Japan's pension system makes it attractive (and in many cases mandatory) to retire early. Japanese workers thus work on average 1680 hours per year and retire on average at age 63 (Organisation for Economic Co-operation (OECD), 2019; OECD Labour Force Statistics, 2019). In the Netherlands, by contrast, workers have a right to decrease their working time and employers must justify the refusal of requests to reduce working time. What is more, discrimination between workers based on work hours is prohibited, so that those who opt for part-time work are guaranteed equal treatment with regard to wages and other benefits. Meanwhile, a high legal age of retirement incentivises Dutch workers to not retire early. 5 As a result, average working time during middle age is much lower in the Netherlands than in Japan, at 1433 hours per year, and workers retire later at the average age of 67. The Netherlands are thus closer to the frontloading end of the spectrum, whereas Japan is closer to postponement (Boulin, 2006;European Commission, 2020). 6 The existence of a variety of free time regimes raises the thus far unexplored question of whether some points on the spectrum between frontloading and postponement are preferable to others, morally speaking. How should a just society structure the costs that individuals face when they want to access free time at middle and old age? My main claim is that justice requires contemporary societies, all of which have regimes that are relatively close to postponement, to move to a point that is closer to frontloading. This means that the distribution of opportunities to access free time at different life stages in the Netherlands is preferable to that of Japan. Beyond that, I claim that even countries like the Netherlands should move further towards frontloading and make access to free time during middle age less costly, although this comes at the cost of further increases in the age of retirement. Even the Netherlands is still too close to the postponement end of the spectrum.
My argument proceeds as follows. Section 'Free Time: Too Much, Too Late' explains why we should be concerned about regimes that incentivise people to postpone the enjoyment of much of their free time until old age. Section 'Free Bargaining?' argues that free and fair bargaining between workers and employers can't achieve justice in the distribution of free time across life stages. In addition to providing conditions under which free and equal bargaining is possible, states must provide a regulatory framework that protects certain opportunities to access free time across the life course. In section 'Hypothetical Choice', I argue that the design of this regulatory framework should mimic the outcome of an idealised decision-situation, where prudent individuals behind a partial veil of ignorance choose between different types of protections to access free time at different lifestages. I claim that prudent persons would choose policies that make access to free time cheaper for the middle-aged and more costly for the old-aged, that is, a framework that is much closer to frontloading than those that currently exist. Section 'Why Do We Observe So Much Postponement?' responds to the objection that postponement lifestyles are so common because postponing free time until old age is what most people really want.
Let me make some preliminary notes before I proceed with the main argument of this article. First, it is important to distinguish between two interpretations of the claim that societies should move closer towards the frontloading-end of the spectrum of access to free time across the life course. One possible interpretation of this claim is that justice requires forcing citizens to enjoy more free time earlier in life. Another interpretation is that justice requires giving citizens opportunities to enjoy more free time earlier in life. It is this second interpretation that I have in mind when I use the term frontloading. One might think that expanding opportunities to access free time early in life is wholly uncontroversial and doesn't require justification because -other things equal -it is always better for individuals to have a larger set of options from which to choose. However, as I will explain in section 'Hypothetical Choice', access to free time early in life can't be expanded gratis -it requires making access to free time later in life more costly, for example by raising the legal age of retirement. Therefore, the more modest proposal to incentivise frontloading, rather than forcing people to frontload, also requires justification. 7 Second, I will frequently refer to two different age groups -the middle-aged and the old. For the purpose of my discussion, I define everyone between the age of 25 and 65 as 'middle-aged' and everyone above the age of 65 as 'old'. 8 Finally, there is a variety of policies that achieve the aim of expanding access to free time during middle age. It would be interesting to know which of these policies is most effective in getting us closer to frontloading. However, I will set this problem of policy-choice to one side in order to focus narrowly on the question of whether moves towards frontloading are morally desirable in principle. 9

Free Time: Too Much, Too Late
The abundance of free time that retirement brings at old age promises to compensate people for some of the hardships they incur during the middle part of life, when they are economically productive. But for several reasons, this promise of free time that awaits us after our work life is a cruel joke.
First, there is a significant number of people who do not reach the age of retirement because they die prematurely. 'In 2017, there were 56 million deaths globally' and 27% among them 'were between 50 and 69 years old' (Ritchie, 2018). Most of these victims of premature death did not reach retirement and the few among them who did reach retirement enjoyed very little of it, in comparison with the many years they spent working. In rich countries, the number of people who die prematurely is lower than the global average, but not by much. In Germany, the share of those who are between 50 and 69 years old when they die lies at 18% and in the United States, it lies at 26% (Ritchie, 2018). Many of these individuals who die early work a lot, but premature death deprives them of the free time they were expecting to enjoy during retirement.
What makes the fate of these individuals even more regrettable is that many of them are already disadvantaged during their lifetime. Among those who die early, the least advantaged are overrepresented, so that early death exacerbates a misfortune that many have already been suffering during their lifetime. In other words, the poor and otherwise disadvantaged have shorter lives. An extensive literature documents that those of lower socioeconomic status are much more likely to die early (Antonovsky, 1967;Case and Deaton, 2020;Lewer et al., 2020;Marmot, 2005;Wilkinson and Pickett, 2010). To illustrate, among men in Norway, those who belong to the richest 1% live on average 14 years longer than those who belong to the poorest 1%. Among women in Norway, this difference is 8 years (Kinge et al., 2019). Among all disadvantages suffered by the poor, their shorter life expectancy might be the greatest and the one that most urgently calls for remediation. It is therefore striking that the poor, who are most in need of compensation for their unfortunate situation, are least likely to enjoy the rewards of retirement.
Second, among those who reach the age of retirement, there are many whose physical and mental health has deteriorated, so that they are less capable of enjoying their free time. Just like premature death, ill health at old age doesn't strike people at random but occurs more often among the disadvantaged, for example, because of hazardous working conditions or other kinds of stressors to which people of lower socioeconomic status are more heavily exposed. According to Crimmins et al. (2009), 'People who have less education and who are poorer are more likely to experience earlier disease onset, loss of functioning, and physical impairment'. In other words, the poor age earlier, which makes it more difficult for them to enjoy their free time when they are old.
Third, some of those who are healthy when they retire have relatives or friends who die prematurely. This reduces the value of the retirees' free time because the loved ones they hoped to share that free time with are no longer around. For example, Bronnie Ware tells the story of John, whose wife Margaret passes away shortly after he retires. Towards the end of his life, after having lived through retirement without his wife, John says, 'I worked too damn hard and now I am a lonely, dying man' (Ware, 2012: Ch. 2). John's retirement was much less valuable to him than it would have been if Margaret had not died prematurely. Again, it bears emphasising that the chances of losing a loved one or a relative are much higher for those of low socioeconomic status. This is because those from lower income brackets are more likely to marry someone who also occupies a lowincome bracket, which as we saw increases one's risk of dying prematurely (Fisher, 2017: Ch. 2). As a result, the risk of having a less valuable retirement because of losing one's partner or one's relative is much higher for those who belong to the group of the least advantaged.
Fourth, among those who expect to reach old age in good health with their loved ones by their side, there are some who would prefer to carry out some amount of work when they are old in order to be able to work less intensely during the middle part of their lives. In other words, they would like to redistribute some of the free time that becomes available through retirement at old age to the middle part of their lives. 'Frontloading' some of one's retirement in this way can be attractive for several reasons. Some people simply prefer to work less intensively at each point in time although this means that they will have to retire later. Others want to spend a lot of time on non-work activities that must be carried out before reaching old age. For example, the raising of a family can't normally be postponed to old age, especially for women, and those who want to engage intensely in such 'middle-age-exclusive' activities can find it attractive to retire later in order to have more free time during middle age. What's more, there are some activities that can in principle be postponed to old age but that are more rewarding when carried out during middle age. For example, it is often more rewarding to learn to speak a foreign language or to play a musical instrument while we're still relatively young because this allows us to enjoy using those skills during a longer part of our lives. Finally, there are some activities that aren't more rewarding but more enjoyable when carried out during middle age. An example of this may be the climbing of mountains, which becomes more difficult as physical strength declines with old age.
These four considerations highlight a need to scrutinise regimes that incentivise postponement. In particular, the question arises whether within the existing range of regimes in force, frontloading isn't in fact preferable to postponement, morally speaking. Is there a principled way to answer this question -that is, to identify morally better and worse ways of enabling certain patterns of free-time across the life course?

Free Bargaining?
One straightforward way of determining the costs that individuals should face when trying to access free time is to allow workers and employers to freely bargain for working time and retirement arrangements that suit their individual preferences. According to this 'free bargaining approach', decisions about how to structure the costs of access to free time are best left to individual choice within the context of a free labour market. Those workers who find it risky or otherwise unattractive to postpone the enjoyment of most of their free time to old age can bargain for contracts that involve retiring later, in order to work less during middle age. Those who prefer working more intensively during middle age in order to retire sooner can strike agreements with their employers that facilitate this option. Just as we allow individuals to decide how many apples, oranges, and other consumer goods they consume at various stages of their life, so too perhaps we should allow individuals to decide how much free time they 'consume' at various stages of their lives.
Free bargaining is appealing because it doesn't impose controversial time-use patterns that at least some citizens will reject because they don't fit their personal preferences and life plans. As I mentioned in the introduction, I am committed to the idea that governments should ensure that citizens have opportunities to frontload free time if they wish to do so -rather than forcing everyone to work less during middle age. The free bargaining approach fits well with this opportunity-focused interpretation of frontloading. To illustrate why the imposing of time-use patterns is problematic, consider the following scenario: Barcelona: imagine that the administration of Barcelona is convinced that it is best for citizens to work in the morning and to enjoy leisure in the afternoon. Motivated by this conviction, the administration penalises the use beaches in the morning, say by charging a fee for accessing them, and subsidises the use beaches in the afternoon, for example by providing free sun umbrellas and Mojitos.
This kind of policy is objectionable because it can't be justified to those who have good reasons for wanting to relax in the morning and work in the afternoon. Similarly, we might worry that promoting long-term time-use patterns such as postponement or frontloading will necessarily conflict with the reasonable life plans of some citizens. For this reason, it is tempting to conclude that states shouldn't pass legislation that facilitates access to free time at particular life-stages, but instead that it should allow each individual to negotiate agreements that suit their particular preferences.
There are two main concerns with this free bargaining approach: a fairness concern and an efficiency concern. The fairness concern is that many workers have only little bargaining power, so that they can't negotiate for contracts that accommodate their freetime-preferences. Those who own neither wealth nor productive means are in a much weaker bargaining position, and will often have to agree to contracts that are unfair to them. Members of the working class, for instance, normally depend on market income for their subsistence and must eventually agree to some work contract, even if it specifies bad working conditions, while employers can often afford to wait and withhold their offers, which gives them superior bargaining power. Apart from material wealth, workers' bargaining power is also affected by how attractive their skills are to employers. Some individuals have skills that are in high demand in the labour market, while others have skills that are barely sought-after. Those with highly sought-after skills will be able to bargain for favourable working conditions. In particular, they'll be able to specify working conditions that fit their free-time-use preferences to a much higher degree than the less skilled, who will find it hard to convince their employers to design schedules that are compatible with their life plans. Consider the following example: Jacob is 50 years old and lives in a society where the legal retirement age is set at 65 years. He would prefer to retire later, say at age 69, in order to be able to work less now, so that he can pursue important life goals that require more free time than he currently has. His employer doesn't agree to this amendment of his contract. What's more, Jacob's skills aren't sought-after in the labour market, so he can't find another job that better suits his preferences. As a result, he must give up some of his life goals.
Workers like Jacob, whose labour can easily be replaced by other workers (or by machines), will also find it hard to bargain for other ways of accessing free time, such as sabbatical years or long paternity leave.
A proponent of the free bargaining approach might object that this concern about inequalities in wealth and skills doesn't tell against free bargaining. Rather, it tells against unjust inequalities that skew the results of free bargaining. Imagine a society where material resources are distributed fairly, so that workers and employers can negotiate contracts on a more equal footing. In this society, everyone would have fairer opportunities to bargain for contracts that facilitate their preferred way of accessing free time at different points in life. It is much less obvious why free bargaining isn't an appropriate solution to the problem of free time allocation across life stages in such an idealised scenario.
This only gets us so far, however, because there is the remaining difficulty that skills, unlike material resources, can't easily be redistributed between individuals (Dworkin, 2002a: 83-85). Still, it might be possible to address the problem that many workers lack marketable skills by using progressive taxation to redistribute wealth from the talented to the untalented. Redistributing money enhances the bargaining power of the less skilled because it allows low-skilled workers to accumulate savings which enables them, in turn, to quit a job that doesn't suit their free time preferences (Parijs and van Vanderborght, 2019: Ch. 1).
Redistributing money goes a long way in improving free-time-opportunities of lowskilled workers but it's not enough. A society without a regulatory framework for work contract negotiations would have to ensure that all workers, without exception, have enough bargaining power to get contracts that are fair to them. But it would be very costly to ensure that everyone's work contracts are fair without defining a set of rules and regulations that apply to everyone. A state that relies solely on free bargaining to distribute access to free time must find ways of (a) identifying which individuals lack bargaining power and (b) enhancing these persons' bargaining power, as it were 'one by one' without issuing general laws and regulations. This is likely to be quite expensive. Alternatively, states can provide an institutional framework that sets parameters and boundaries to the negotiation of work contracts. Such a general framework helps address the issue of fairness because it can specify certain free time protections that all workers in a society must enjoy.
Apart from the fairness concern, there is also an efficiency concern about free bargaining. Free bargaining involves large transaction costs. To understand this concern, imagine a society without a regulatory framework for work contracts. In this society, all factors that affect workers' access to free time are negotiated from scratch each time a worker and an employer enter a contract. The negotiating parties would have to find agreement on a vast number of issues, starting from worker's basic rights and entitlements to the definition of a retirement scheme, and so on. In a pure 'free bargaining society', there would be no pre-existing health and safety regulation and no rights to sick leave, to paid leave or to refuse excessive overtime work, because all of these issues would be settled through individual negotiation. In a modern complex economy, where workers constantly enter and quit contracts, these negotiations would generate very large transaction costs that would undermine the economic viability of many businesses. Bargaining between workers and employers is much more efficient when it can use as a point of departure a base of relatively uncontroversial pre-existing terms and conditions, specified by the state.
In summary, free bargaining can't achieve justice in access to free time and states must therefore provide some institutional framework to facilitate negotiations in the labour market. This in turn raises the question of how this framework should be designed. After all, states must specify what exactly counts as excessive overtime work, how many days of vacation workers should minimally enjoy, and so on. By the same token, all industrialised countries may wish to specify a legal retirement age that encourages some ways of accessing public pension benefits and discourages others. 10 Some countries, such as, for example, Turkey, Luxembourg and Slovenia, incentivise citizens to retire early by defining a low legal retirement age that requires workers to quit the job market roughly at the age of 60 (OECD, 2017). In these countries, citizens normally enjoy a lot of free time when they're old. Countries like Iceland and Norway incentivise their citizens to retire much later, at the age of 67 (OECD, 2017: 93). 11 At the same time, Iceland and Norway are among the countries with the lowest average annual hours worked per worker in the world (OECD, 2019). This variety of regimes raises the question of whether some ways of distributing access to free across individuals' lives are preferable to others. Is there a principled way of identifying morally desirable 'free-time-guarantees' that protect opportunities to access free time at different life stages?

Hypothetical Choice
As I mentioned in the 'Introduction', a principled approach that determines how much frontloading is desirable is necessary because opportunities to access free time during middle age can't be expanded gratis. If it were possible to provide everyone with the option to enjoy more free time during middle age without imposing any costs, then this would be obviously desirable because it is normally better for individuals to have a larger set of options from which to choose. Policies that expand our opportunities during one stage of life without foreclosing opportunities at other stages of life don't require special justification.
However, policies that protect access to free time during middle age are costly in the sense that they affect the opportunity range we enjoy at old age. Consider the example of a legal entitlement to sabbatical leave. A society that protects a right to sabbatical leave might be economically less productive than a society that doesn't protect this right. This might be because employers must regularly find substitutes for those who are on sabbatical leave, which in turn generates costs related to advertising jobs, hiring new employees, transferring knowledge to these new employees, and so on. What's more, employees that spend a prolonged time outside the labour market might miss out on important developments in their professional field, so that they must catch up on knowledge and skills when they resume their jobs. We can expect that employers would pass these costs onto employees. So we must ask ourselves whether workers would, nevertheless, prefer to live in a society that protects a right to sabbatical leave. Given that expanding individuals' access to free time during middle age is costly, how might frontloading policies such as the right to sabbatical leave be justifiable? Which point on the frontloading-postponement spectrum would a just society occupy?
Recall that free bargaining appeared to be an attractive solution because it seemed to allow individuals to access free time in accordance with their preferences and life plans. This preference-sensitivity makes the free bargaining approach non-perfectionist: it doesn't rely on controversial assumptions about what allocation of free time over the life course is best for human flourishing. I now want to propose that we adopt an alternative approach to identifying a just free time regime that retains the non-perfectionist character of free bargaining but that avoids its problems. This approach asks what outcome would emerge from an idealised decision-situation, where prudent individuals behind a partial veil of ignorance choose policies that regulate access to free time. It uses the outcome of this decision-situation to identify a desirable free time regime. Different versions of this approach have been developed by political theorists in order to address other questions of social justice, such as, for example, the just distribution of healthcare between citizens in different age groups and the just design of pension systems. My aim, here, is not to elaborate and defend the theoretical details of any one version of this approach but to provide a general outline of it and to show that it is a potentially promising way of identifying a just regime for distributing opportunities for free time over the life course. 12 To understand what the prudent choice approach involves, we must unpack the notion of prudent choice and also understand what it means for prudent choosers to be 'ideally situated'. The goal of a prudent individual is to allocate resources to the different stages of her life in a way that makes her life as good as possible, as judged by her own ambitions and values. In most cases, a prudent person wouldn't make one stage of her life better at the expense of making her lifetime as a whole worse. For example, it would be imprudent to enjoy too much free time early in life, say in our thirties, when we're most able to be economically productive. A person who enjoys too much free time during her thirties might later go through great hardship because she hasn't generated enough savings to sustain herself in old age; this might make her life go worse as a whole.
Prudent choosers must be ideally situated in the sense that they don't have complete knowledge about themselves and their circumstances. There are two reasons for imposing this knowledge constraint. The first reason is that this can help identify what is prudent. To understand this, imagine a chooser who knows her age at the time she makes a choice about how to spread free time across her life. This person might imprudently allocate a lot of free-time opportunities to her current stage of life because she is too short-sighted to fully appreciate the importance of free time at stages of life that lie far in the future. The second reason for why choosers shouldn't have complete information about themselves is that this could distort their decision-making in morally objectionable ways. For example, if a person knows that she is in excellent health and that her genes will allow her to live up to very old age, then she has an incentive to choose a regulatory framework that provides favourable conditions for the long-lived. Basing a regulatory regime on this choice would be unfair to those who die prematurely. In order to avoid biases of this kind, we must devise a 'veil of ignorance' that hides information from the prudent chooser about certain of her morally arbitrary characteristics.
We can specify the knowledge constraint further by imagining that individuals don't have information about their income earning capacity. This is important because if choosers knew, for example, that they will earn a very high wage, then they might devise free time protections that are beneficial only for high-income earners, who can meet their basic needs with fewer hours of work than the less skilled (who instead must work long hours in order to make ends meet). Next, we assume that our prudent choosers know their preferences and life goals, as well as what types of activities and projects they can typically realise during the different stages of their lives. This will help make the regulatory regime they select preference-sensitive. 13 What's more, choosers know about people's average lifespan and the distribution life-lengths across society but they don't know up to what age, they, as individuals, are going to live.
Having observed these ideal choice conditions, we can now turn to the question of how prudent individuals would design a regulatory framework that protects access to free time during middle and old age. 14 To do this, we can survey different strategies of accessing free time that choosers might employ. One possibility is that prudent choosers would opt for a regulatory framework whose sole aim it is to boost their financial income, so that they can always use money to buy more free time when they need it (Rose, 2014). In other words, individuals might not care about devising free time protections because they hope to be able to convert money into free time whenever they need it. This would be an appealing strategy because if successful, it would offer more flexibility than general free time protections that apply equally to everyone. However, it is not always possible to convert money into free time. Consider the example of an employee whose lack of marketable skills forces him to work 60 hours per week. This employee certainly has an interest in choosing policies that boost his income. But imagine furthermore that his job is only available on a 'take it or leave it' basis, so that he can't adjust his working time to say 40 hours per week. This gives him reason to also care about policies that protect access to free time. 15 We should therefore consider other regimes that prudent choosers might select in the ideal choice situation. Consider first an extreme version of a postponement regime -that is, a regime in which people work as much as possible during middle age in order to retire very early. This regime would involve long weekly work hours, short vacations, and weak legal protections for workers' ability to reduce working time if they wish to do so. Some individuals may well prefer this regime, given their life plans. These individuals may want to protect only enough free time during middle age to recover from the burdens they undertake in working and to carry out other essential activities such as eating and sleeping. As I explained in the introduction, among currently existing regimes, Japan approximates extreme postponement more than other countries.
Would extreme postponement be selected in the ideal choice situation? For some individuals, postponement can be very rewarding. Imagine someone who is fortunate enough to have most of her important life goals related to, or identical with, the content of her job -for example, an artist who enjoys painting and who can live off the sale of her paintings. This person might think that she hardly needs any free time protections because she is realising her life goals through her work. The most rational thing for her to do would be to spend almost all of her middle age working as a painter, thereby realising her life goals and creating generous savings for old age. To her, it would feel like a waste to have available more than the minimal protections offered by postponement because she doesn't need much free time.
Remember, however, that our artist doesn't know how great her income earning capacity as a painter will be. The 'veil of ignorance' hides from the artist information about whether there will be members of her society who are going to buy her paintings. She must therefore consider the possibility of pursuing painting as a hobby while working on a job to earn money for her subsistence. What's more, she might not be able or willing to delay the pursuit of her passion for painting until the age of retirement. In light of these considerations, it seems plausible that even someone whose work is largely aligned with her life goals would choose more than the minimal protections of free time in middle age that postponement offers because otherwise they would still risk being unable to realise important life goals that require free time and that can only be pursued during middle age. This is even more true, of course, for people whose work is not largely aligned with their life goals.
What about people whose conception of the good involves engaging in a lot of paid work? These people differ from the artist in that they don't value work intrinsically, that is, because of the activities it involves, but instrumentally, because it increases their income and thus their capacity to buy consumer items. Imagine an extreme case of such an income-lover who prefers work over all other activities. Even this person would choose more generous protections of free time during middle age than those provided by postponement because she must be alive to the possibility that her conception of the good will change over the course of her lifetime (Bou-Habib, 2013: 205;Daniels, 1983: 531). We can imagine, for example, that our income lover, who spends most of her waking hours on the job, discovers at some point that she also wants to raise a family. The institutional set-up of a society guided by the strategy of postponement would make it very costly for her to adjust her work schedule to accommodate her newly discovered desire to be a parent. A prudent person would value the opportunity of being able to realise such a change, even though she might end up not making use of the opportunity.
We can contrast extreme postponement with another extreme strategy that consists of enjoying as much free time as possible during middle age -the strategy of extreme frontloading. Two of the considerations I introduced in section 'Free Time: Too Much, Too Late' might speak in favour of maximising the availability of free time during middle age. First, middle-aged individuals normally have a significantly wider opportunity set than the elderly. Physical and mental health decline with old age, thereby limiting the range of activities that the elderly can pursue. Individuals might want to expand access to free time during middle age in order to be able to choose between a wider range of activities. Imagine a person whose favourite activities include diving and raising children. Knowing that these activities won't be available to her at old age, this person might want to pursue them intensely during middle age, while she's still capable to do so. In addition, as I mentioned before, one and the same activity might be less burdensome and therefore more enjoyable for someone of middle age than for someone who is old. Imagine someone who likes hosting barbecue parties and horseback riding. She might still be capable of engaging in these activities when she's old but it'll be much more burdensome for her to do so. This provides another reason to maximise access to free time during middle age.
The second consideration that counts in favour of frontloading has to do with the difficulty that not everyone lives until old age. Individuals who opt for postponement know that if they die before reaching the age of retirement, they will not have had enough time to realise some of their life goals during middle age. Given the fact that the risk of early death can eventuate at any moment, individuals might find it attractive to enjoy as much free time as possible while they're still middle-aged. To illustrate this, consider the following example: Betty and Rufus: In one world, Betty and Rufus both pursue the strategy of postponement. Betty dies shortly before she reaches retirement age and Rufus lives through retirement until very old age. In another world, Betty and Rufus pursue a frontloading strategy. In this world too, Betty dies before she reaches retirement and Rufus lives until very old age.
There is a sense in which the early death suffered by Betty in the first world constitutes a greater disadvantage because she doesn't get to enjoy the later reward for her efforts during middle age. Frontloading the enjoyment of free time partly compensates those who die prematurely for their bad fortune (Fleurbaey et al., 2014). Individuals must also be aware of the risk that early death affects persons who are important to them. Recall the case of John and Margaret, where John regrets having postponed too much of his free time because the early death of his wife Margaret severely impugns his ability to enjoy his free time. There are also instances of bad luck that are less severe than early death but that nevertheless make it attractive to enjoy a lot of free time early in life. For example, the prudent chooser might be concerned about disability or painful illnesses that sometimes come with old age and that make it more difficult to enjoy one's free time.
But there is also an important consideration that counts against frontloading. Workers must use their ability to be economically productive during middle age in order to create enough savings for old age (or in order to contribute to an intergenerational pension scheme), when they are less able or unable to earn income. Prudent persons would at least generate minimally enough savings for subsistence at old age. 16 What's more, they must avoid the risk that time poverty and an obligation to work a lot at old age makes their life as a whole worse than it could have been if they had generated enough savings during middle age.
We must conclude then, that prudent choosers would neither opt for extreme postponement that involves lots of work during middle age and no work during old age, nor would they choose extreme frontloading that involves minimising time spent on work during middle age. Instead, they would choose an intermediate solution that enables middleaged persons to pursue life goals other than paid work before reaching old age.
This insight is consistent with empirical findings according to which many workers complain about a lack of work-life balance and report always feeling rushed (Hamermesh, 2018: Ch. 3;Sirgy and Lee, 2018;Thornthwaite, 2004). Many would like to reduce their work hours, even if this results in reduced income (Golden and Gebreselassie, 2007). As I explained previously, countries today are all relatively close to the postponement end of the spectrum. Even in the Netherlands, many workers express preferences for a slower transition into retirement and for reduced work hours before retirement. More specifically, 22% of men aged between 51 and 65 in the Netherlands currently work part-time and 57% of men 'would like to reduce their work hours and take some form of gradual retirement, but for some reason are not able to do this' (Kantarci and Van Soest, 2008: 132). Many more women of the same age group already work part-time but among those women who work full-time an overwhelming majority of 90% also want to transition to part-time (Kantarci and Van Soest, 2008: 132). What's more, there are a sizable number of people in the Netherlands who die before they reach old age. In fact, 19.4% of everyone who died in 2017 in the Netherlands was aged between 50 and 69 (Ritchie, 2018). A prudent chooser who knows this fact but doesn't know her life expectancy would count it as a weighty reason to move closer to an intermediate free time distribution over the life course.
A regulatory framework that facilitates an intermediate solution would redistribute opportunities to access free time from the old to the middle-aged, for example, through long vacations, shortened work weeks or through the option for workers to take several sabbatical years over the course of their work life. In order to make these additional opportunities for free time available, the average working time of those of old age must be increased, for example, through raising the age of retirement.

Why Do We Observe So Much Postponement?
Let me now consider an objection to the case I have made for more frontloading. If, as I have argued, more frontloading than is currently implemented across industrialised countries, is the most prudent regime, then one might ask why it is that no country is currently implementing it? Why is it that so many people don't do what would be prudent, namely work less during middle age and more during old age? Doesn't the fact that postponement-lifestyles are so common show that this is what people really prefer?
In fact, there are several reasons for why the prevalence of postponement-lifestyles might not reflect people's actual preferences. First, unjustly low wages make it unavoidable for many that they must work long hours both during middle age and old age. The least advantaged often don't enjoy the luxury of deciding at which stage of life they want to access most free time because their situation forces them to work long hours throughout their whole life. What's more, the least advantaged are often unable to plan how they want to distribute free time across their lives because they work in precarious jobs that they can lose at any moment, so that long-term planning becomes impossible (Cuervo and Chesters, 2019;Kalleberg, 2018: Ch. 1 and 6).
Second, we can speculate that there are many who would decide to frontload more free time if there weren't such powerful incentives to postpone the enjoyment of free time. Third, as I have argued elsewhere, there are many workers who decide not to reduce their average monthly working time because they are trapped in collective action problems that make it very costly for them to reduce their working time (Jauch, 2020). To illustrate, employers often use working time as a proxy for their employees' productivity and commitment and those who work particularly long hours are often awarded benefits such as raises or promotions or are spared from dismissals. This makes it individually rational for each worker to work extra hours in an attempt to out-compete colleagues. In the absence of such competitive pressures, many middle-aged workers would arguably reduce their working time, although this might mean that they must retire later.
Fourth, there are a number of fortunate individuals whose decision to work a lot during middle age isn't the result of poverty, unwelcome incentives, or collective action problems. These individuals decide to postpone the enjoyment of their free time because they are convinced that this strategy is best for them. However, some of them later regret that decision. Recall the case of John, who, reflecting on his previous choices, says, 'I worked too damn hard and now I am a lonely, dying man' (Ware, 2012: Ch. 2). It might be that such retrospective judgements about the value of our life plans are more accurate than judgements that are made earlier in life, when the outcomes of risks like premature death have not yet eventuated. It is psychologically difficult to take the risk of dying prematurely or of losing one's loved ones into consideration when we make decisions about how to plan our lives and this should make us cautious about taking the preferences of 'postponement-lovers' at face value.
Finally, while there are some individuals whose decision to work a lot during middle age isn't the result of poverty, unwelcome incentives, or collective action problems and who don't come to regret this decision, we should not assume that this group of individuals constitutes the majority of persons in society. Thus, while postponement might be widespread, only few people may genuinely endorse this way of allocating free time across the life course. For most people, postponement gives them their their free time when its too late.
whether justice requires everyone to work less. See for example (Cohen, 1978: Ch. 11) I have taken the helpful distinction between interpersonal, counterfactual, and intrapersonal distributions of free time from Tom Parr, 'Work Hours and Free Time', unpublished manuscript. 2. Note that by a 'postponement regime' I do not mean a regime of postponing retirement, but a regime of postponing the consumption of free time. 3. Recently, the Japanese government signalled that it wants to increase the average age of retirement (Kajimoto, 2019). 4. What's more, pay in Japan is largely 'input-based', which means that workers are rewarded for spending a lot of time at the workplace, even if this doesn't lead to a proportionate increase in productivity (Ono, 2018). 5. Late retirement can also be incentivised in other ways, for example, through sanctioning firms that require older employees to retire early against their will. Some countries, such as Spain and Poland, prohibit compulsory retirement and others, such as Italy and Japan, permit this practice. There is also evidence that increasing pay and improving the quality of jobs make it more likely that workers retire later (Pilipiec et al., 2020). 6. There might be other explanations for these time-use differences between Japan and the Netherlands. For example, social norms that stigmatise leaving one's workplace early are more widespread in Japan than in the Netherlands. I don't deny the existence of such cultural explanations. For my argument it suffices to assume that regulatory regimes have some effect on how much free time workers access at different stages of their lives. What's more, social norms and cultural differences don't exist independently of public policy. Governments can pass legislation that affects social norms. For example, taking time off work to care for children might be less stigmatised in countries where legal entitlements and subsidies to do this exist because such entitlements normalise care work and the associated career-interruptions (Raven, 2009). 7. One might object that my proposal to raise the legal age of retirement forces workers (rather than incentivising them) to retire later because in capitalist societies many workers can't afford to deviate significantly from established legal retirement ages. Whether or not it is true that workers have no choice but to work until they reach the legal age of retirement depends on (1) their level of income and (2) the extent to which deviations from the legal retirement age are penalised. Here I largely set aside issues of access to free time that arise from unjustly low income in order to focus narrowly on the intrapersonal question of how the costs of accessing free time should be distributed within an individual's life. 8. This definition is somewhat arbitrary but this isn't problematic because its purpose is only to a provide a rough schematic distinction between the old and the middle-aged. 9. My normative argument in favour of such moves is particularly relevant in the context of debates about the so-called 'burden of early retirement'. Economists have recently noted that average life expectancy in OECD countries increases steadily over time, and that at the same time, the average age of retirement continuously falls. Together, these two trends exert pressure on countries' pension systems. My argument validates the proposal of some economists to incentivise delayed retirement (Conde-Ruiz and Galasso, 2004). 10. There is also an important gender-dimension to the question of how institutional frameworks should affect the costs individuals face when accessing free time at different life stages. It is conceivable, for example, that some policies that expand access to free time during middle age, such as, for example, paternity leave entitlements, are preferable to others because they have the additional advantage of advancing gender equality. Apart from their effect on gender equality, policies like paternity leave might also be appealing for the simple reason that some people would prefer to access free time through paternity leave rather than through sabbatical leave or other policies. 11. There are also big differences regarding the extent to which states penalise deviations from the legal retirement age. Sometimes workers decide to retire before having reached the legal age of retirement and states often penalise such deviations by reducing workers' pension benefits by a certain percentage for each year of early retirement. The severity of this disincentive for early retirement varies greatly between countries. For example, Germany reduces early retirees' pension benefits by a much larger margin than Austria. In a similar fashion states can penalise or incentivise work at old age, for example, by allowing workers who have passed the legal retirement age to top up their retirement benefits with their market income (Axelrad, 2018;Casey, 1997: 19). 12. For two prominent examples of how prudential reasoning can be employed to settle questions of justice between age groups, see (Daniels, 1983;Dworkin, 2002a: Ch. 8). Approaches that use hypothetical prudential choice to solve (intrapersonal) questions of distributive justice face a number of well-known objections, such as the charge that substantive normative convictions can be smuggled into the theory under the guise of purportedly neutral considerations of prudence. I don't have enough space to address these objections here. For replies to some of these objections see (Dworkin, 2002b). 13. As I noted previously, preference-sensitivity is attractive because it helps avoid perfectionism. Someone might object that there is an easier way to avoid perfectionism: We can hide from choosers their conception of the good. Individuals who don't know their conception of the good will choose a strategy that maximises their possession of primary goods, that is, goods that help advance a great variety of life goals. This way of avoiding perfectionism is problematic because it faces the well-known 'index problem'. We don't know how much value individuals would assign to different primary goods when they compare their relative importance. As a result, approaches that hide from individuals their conception of the good might yield results that are insufficiently determinate to guide public policy. 14. I set aside the question of how much access to free time prudent individuals would protect for young age. 15. As I mentioned previously, the boosting of wages normally increases workers' bargaining power vis-àvis their employers. One might think that this increase in bargaining power is by itself enough to enable workers to turn down, say 60-hour contracts and to force employers to specify 40-hour contracts instead. But the additional bargaining power derived from higher wages isn't always enough to outweigh the employers' interest in specifying contracts that involve very long work hours. Especially in situations where labour supply exceeds demand (i.e. when there is non-voluntary unemployment), employers might still find it preferable to turn down job candidates who 'only' want to work 40 hours. 16. This is only true in societies where individuals don't receive a basic income. In a society where everyone receives a basic income, individuals might decide against generating savings for old age. We can also imagine societies where only the elderly receive a basic income, independently of whether or not they engage in economically productive activities during earlier stages of their life. This would also incentivise people to refrain from generating savings for old age. But assuming such a context would amount to assuming what I intend to show, namely that justice requires expanding access to free time for the middleaged, although the middle-aged must generate savings for old age.