Association of Asian Studies University of Maryland
College Park, MD
October 27, 2007
Part I: Prologue After one argument they began sleeping separately, without quite knowing whose decision it was. They should have talked about things, but they both seemed to avoid doing so. If nothing else, after being in the truck every day, Atsushi simply didn't have the energy. Unlike in his student days, he was always tired. His exhaustion remained, persisting in the roots of his muscles and in the joints of his bones.
Ito Takami, "Thrown Away", p. 271
The 135th Akutagawa winner paints a grim picture of Japan's future. The core of the story is two individuals servicing vending machines for a living. One is Mizuki Emi, a single mother, about to get married; the other a young man, Sato Atsushi, about to get divorced. The older woman is grasping for stability of some sort, doing whatever it takes to provide for her children. The man finished college with an idle dream of become a screenwriter, but never submitted a script or got involved with productions. Instead, it seemed to merely what helped hold a group of college drinking buddies together. While in college he started to banter with one coed, who became his sexual partner; they drifted into marriage. She had a more concrete goal, to become an editor, and after graduation promptly found a job with a magazine. But she gradually fell into a funk, quit her job and sat around home; after her husband raped her during a drinking bout, whatever tenuous ties they had evaporated. The picture is of the downhill spiral of two individuals with no clear goal, no stable friendships, and no prospect of a higher income or proper job. It is a narrow world; no retirees or relatives appear, there are no thoughts of children – except to fear that a slipped condom might lead to pregnancy.
If this is the future of Japan, then it is a grim one indeed. Yet it does reflect part of the social environment that underlies low fertility. The shift in labor markets, combined with a long period of slow growth, gave birth to a "lost generation" that (quite rationally and responsibly!) does not want to reproduce itself. Research shows that marriage is partly a function of husbands' income – young men in temporary or other low income jobs tend to remain single – and that when women do marry such men, they avoid having children.2 Unlike when Ito Takami wrote his story, as of 2007 the economy had been growing year-on-year for 22 quarters; unemployment among this crucial age bracket has fallen, and the number of individuals in "regular" employment is up. So, at least anecdotally, are weddings. Even if this leads to an increase in the birthrate 10 months later, however, it will not lead to a fundamental shift in fertility. Any change is likely to be small, and most of the economic impact would not be felt until 2 decades or more later, once the (slightly) larger cohorts born in the years leading up to 2010 enter the labor force. This need not, indeed probably does not imply a grim future, even if it is a future that will entail novel challenges.
Part II: Introduction
Japan is the first society in history to see its population decline due to a drop in fertility rather than a rise in mortality.3 In addition, its population is aging rapidly, something that again is historically unprecedented. In contrast, the population of the U.S. continues to increase about 1.0% per annum. While we also are facing the retirement of the baby boom generation – those born in 1946 are approaching age 62, when voluntary early retirement becomes feasible, and when involuntary retirement increases – both the rate of change, and the level relative to the population as a whole, are comparatively modest.
This paper will make three points. That of Part III is simple, that Japan is one of many countries experiencing the impact of fertility decline and of population aging (indeed, they are to some extent the flip side of the same coin). Japan is in this more "normal" than the U.S.; it is merely a bit ahead of the pack. Because of that, in fact, the analysis of aging in Japan promises insights into what will happen in Europe, Korea, China and elsewhere (including southern India). This is a topic that ought to interest individuals across a wide swath of Asian studies.
The following Part IV looks at demography through the eyes of Japan's young people. How did Japan get to where it is today? – that reflects decisions of young women (and of their mothers). Fundamentally, it is a story of prosperity and the increasing choices available to women. It reflects for example the shifting structure of labor markets, as young men are less able to afford a wife, much less a full-time housewife, given the lifestyle to which women are accustomed. However much older Japanese may wring their hands and pine for the good old days, their children (or at least their daughters) do not. It is very easy to forget this, given the rhetoric that we sometimes see. I have little insight into the vision of the future young Japanese see; their expectations are furthermore quite diverse. I think that is in part because they are aware of rapid change, the inchoate forms of a future far different from the Japan of their parents. I am sure that, if I were transported 30 years into the future, I would find myself quite out of place, socialized into the pre-bubble Japan of the late 1970s and 1980s.
In Part V, the third substantive section, I will remain true to the title of the paper, and try to look through my economists' glasses, however darkly. Since Japan's experience will be novel, such projections must be based on theory untested by historical experience. Nevertheless, it is inevitable that aggregate growth will remain quite slow, compared to that of the U.S. However, that is not a crisis and in itself not even necessarily a worry, as per worker income will continue to rise. Likewise, Japan's national debt is not in itself a problem, even though the country has been more profligate than Italy, long the fiscal black sheep of the developed world. Likewise, retirement can and will be financed. The underlying issues are widely misperceived, in particular that the analogy of retirement at the national level to that at the household level is fundamentally wrong; where enough savings have been set aside is not relevant. Retirement is fundamentally, unavoidably pay-as-you-go. I will focus on this point, which has a simple consequence: that the central policy challenge is not fertility, or immigration, or even growth; it is rather the division of the (national) economic pie.
Of course outcomes are not written into tombstones – yet. Policy can, is, and will continue to make a difference. Politics to date has dictated a focus on effects rather than causes. However much Japan's leaders may decry the baby dearth, concrete action instead has been limited to those approaching death. Many things could have been done to increase fertility – the conclusion of the second section will note such steps – but from my perspective, it is too late: none would have a significant impact within my lifetime as an observer of Japan. Why that should be the case is a fascinating political science case study, and the comparative success of healthcare policy for the aging merits widespread study and perhaps emulation. The real political battle has however yet to commence: how will the baby boomers live, and how will that affect the lives of their children and grandchildren? Unlike the U.S., where discussion of the issues has not taken place, there are at least hints of how Japan will proceed. As the closing vignette of Part VI indicates, the details remain sufficiently hazy that, even if economists prove prescient in their analysis, it is not yet something that the young think about.
Part III: Japan as a "normal" country
The decline in fertility in Japan is normal for a high-income society, and more generally for societies with high educational expectations. In agrarian societies, children have value as sources of labor, and above all as a form of insurance. Such societies tend to be at most partially monetized, while land is lumpy and illiquid; children are the most reliable form of savings against disability and old-age. Both Jewish and Christian scriptures rightly note that it is a pity to be childless – and the person most to be pitied is the childless widow. With high mortality, particularly of infants, on average it was necessary to aim for at least four children to guarantee that one son would reach adulthood. As economies become more urban and monetized, however, the value of children diminishes, and improved public health further reduces the target number. Japan fits that pattern; by 1962, the realized number of children had already fallen below 3 for couples married postwar, and for at least the past three decades, the median desired number of children has been 2 and the realized number of children around 2.2.4
While the range varies, no member of the OECD (the Organization of Economic Co-operation and Development, comprised of essentially all the rich countries) has fertility rates much above 2, and with the exception of the poorest 3 countries (Brazil, Mexico and Turkey) none are above replacement level of a TFR (total fertility rate) of 2.1.5 As recently as 1990 Brazil, Mexico and Turkey had TFRs of 2.8, 3.4 and 3.1, respectively; they are now barely above replacement levels, with TFRs of 2.3, 2.2 and 2.2. The decline in Korea is particularly striking, going from 4.5 in 1970 to 2.8 in 1980, 1.6 in 1990 and 1.16 in 2004, the lower even than Japan. All of Mediterranean Europe is at or below 1.4, from Portugal through Greece. (See Figures 2 and 3 on fertility and marriage age for Japan, Italy, Spain, Greece, and Portugal, and Figure 4 on China.)
In the short run, low TFRs do not immediately lead to a flat or declining population; in some countries, high birth rates 20-30 years ago mean that there is a large cohort of women in their prime reproductive years. Even if individually they have comparatively few children, in the aggregate the number of children can still exceed that of prior years. (See the "echo" in Figure 1.) Immigration can also make a difference. The TFR of the US has been at or below replacement since 1972, a sufficient length of time for the "momentum" generated by past higher fertility to have largely worked its course. However due to high levels of immigration, which is biased towards younger individuals, population continues to rise about 1% pa, the fastest in the OECD.
Low fertility in and of itself does not imply an aging population. However, declining fertility rates do. The age structure of a growing population takes the form of a pyramid, with a large base as each cohort of children is larger than the one that preceded it. (See the separate population pyramid chart sequence.) If fertility stabilizes around replacement level, then the number of children – the size of birth cohorts – will level off. For a stable population, the figure will look more like a stack of equal-sized coins than a pyramid, eventually coming to a point as mortality takes its toll at older ages.
However, if birth rates fall sharply, younger cohorts will tend to be smaller than their predecessors. There will be an "echo" when the boom generation enters its reproductive years. The echo of that echo will be far more muted or (in Japan's case) virtually absent if fertility continues to fall. (See Figures 1 and 8.) As past population growth works its way through, the number of mothers stabilizes, the size of birth cohorts will begin to decline. That results in a bulge that moves towards older age. (See the population pyramid figures.) The faster the decline in fertility, the more rapid will be the rate of aging when the first of the "boomers" hit retirement age. Among the developed countries, Japan's decline was both sharp and early. That in China and Korea has been even faster; in the future, therefore, the rate of aging in those two societies will be even faster.
Falling mortality accentuates the impact of aging; Figure 7 contrasts age-specific male and female mortality for Japan in 1930 and 2000. While I do not provide comparative data herein, Japan has slightly greater longevity than in the other OECD members, but the difference is small. To the extent that aging reflects shifts in mortality, there is again little distinctive about Japan. Infectious diseases, including tuberculosis, once claimed large numbers of Japanese. Stroke and other circulatory diseases followed. Now cancer is the leading cause of death. With extended longevity, debilitation is an issue; dementia of various forms is unusual at younger ages, but is endemic at high levels by age 80. Projecting future mortality trends is at best an educated guess, based on the rate of past improvements in age-specific mortality. The current 2006 medium-mortality projections for 2055 show 12.0 million Japanese in their 80s, 4.6 million in their 90s and 634,000 Japanese in their 100s. In 2055, according to these projections, the single largest cohort in the population will actually be those age 80; with the decline in women of child bearing age on top of continued low fertility, there will be 3 times more 80-year-olds than newborns.6 Faster progress in treatments for cancer and other age-related diseases could push those numbers higher; the faster evolution of antibiotic-resistant bacteria could sway numbers the opposite direction. In any case, Japan is already facing an aging population. But in 2055, the date at which current detailed demographic projections stop, this will not make a large difference. Those just entering the labor market will have been born by 2035; for the most part, their mothers have already been born. Those just retiring are currently 12 years old. Higher fertility could make some difference, as could higher (or lower) mortality. The overall size and structure of the population – absent a large shift in immigration – will not shift due to anything short of a revolution at either end of the process of population generation and degeneration.
The next section asks why fertility should have fallen so far and whether a near-term reversal is likely. The section after that notes the potential for higher female and old-age labor force participation to offset the economic impact of aging, but stresses the overall dilemmas that aging will impose. The postlude returns again to an excerpt from a recent Akutagawa-winner novelette.
Part IV: The Behavior of the Young: the Origins of Aging
Japanese speak of a baby boom in the late 1940s; that after all is the normal reaction to a society after a disaster, natural or man-made, that effects a demographic disaster. Reproductive decisions reflect a target family size. In Japan's case families were presumably smaller than desired, helped by extended marital delay (and within marriage, marital abstinence) enforced by physical absence and accentuated by mortality from the targeted U.S. firebombing of civilian population centers. The data in fact suggest more of a blip than a boom: standard sources provide no data for 1944-46, but the birthrate at its 1947 peak was only 10% above the level of the early 1940s, and below that of the early 1920s. A moving average highlights that: the wartime series leaves off near where the postwar series begins. (See Figure 1; the "ordinary" nature of late-1940s fertility would be more apparent if the graph was extended to the early 1930s.) Japan in fact did not experience a baby boom; it experienced a fertility decline, with aging as one consequence.
The origins of that fertility decline, as the earlier comparisons suggest, lie primarily in economic considerations and not in culture. There are cultural components to reproduction; in Japan, very little childbearing takes place outside of legally recognized marriages. That varies across cultures; in parts of Europe and in the US, for example, single motherhood is not unusual. In Sweden, less than half of couples are formally married. By and large, however, there is little apparent correlation between such patterns and the level of fertility. There are however interactions with the nature of labor markets and other "local" conditions; work, which takes up a goodly portion of waking hours, is certainly a component of culture. But the substantial shifts in labor markets over time suggest that in this component culture is quite flexible, adapting to and helping individuals operate within labor markets. Instead of culture, then, reproductive choices are made on the basis of the costs and benefits of children. Over time the benefits have decreased, and the costs have increased, as a natural consequence of increased levels of income.
A large literature on economic demography indicates standard factors affecting desired fertility. By the mid-1950s, the benefits of having children had declined; they were not needed on the farm or in family businesses, and thus served much more as a consumption item. From that perspective, as noted above, there is cultural consistency in viewing 2 children as an ideal (though about 1 in 5 women have 3 children); that children are desirable is a social norms. However, Japan also has a norm that childbearing should be within the context of marriage, which was reinforced by the financial difficulties a single mother faced. The marriage decision is thus an intermediary, and that has changed markedly over time. In addition, tastes (and opportunities) for women to work evolved over time. That affected the marriage decision, but also the choice of whether to actually have children once married.
Up until 1975, the average age at first marriage was stable at around age 25. While not all women married then, the oft-cited comparison of single women to "Christmas Cakes," which lose value after the 25th, reflected actual behavior. In 1950 almost have of women married before age 25, and 85% married before age 30. Thereafter the average marriage age increased inexorably. (See Figures 5 and 18.) In the labor market as a whole, the age-wage profile flattened over time; the entry of the comparatively large cohorts born in the late 1940s helped drive that. Those men began entering the labor market after middle school and especially after high school, depressing starting wages. But by 1975 the "baby boom" chorts were all in the labor force. With higher starting wages – and lower rates of increase – it was only somewhat later in their career that men had sufficient income to set up a separate household.7 By the mid-1980s young workers were becoming relatively scarce; job opportunities for women expanded, while the "new middle class" had higher educational aspirations for their children. From 1986, the proportion of women continuing on to post-secondary education rose, particularly as women began attending 4-year colleges in greater numbers. In round numbers, today 40% of women go on to a 4-year college and a further 10% attend 2-year junior and technical colleges. (See Figure 17.)
Women's participation in the labor market has responded to these opportunities and one suspects for recent college graduates, parental expectations.8 Figures 12 and 13 trace male labor force participation over time. Except for an increase in the rate of retirement, and decreases at the youngest ages corresponding to changes in schooling, there is little change. That is not the case for women, as seen in Figures 14 and 15. In 1965 participation exhibits a sharp "M"-form, as over 70% of women entered the labor market after schooling, but then dropped out with marriage and children, only to re-enter 15 years later, corresponding to the entry of their (2nd) child into middle school. The most recent data show participation peaking in the age 25-29 bracket at 75%, while 65% of women in their 30s remained in the labor market. The annual data of Figure 15 illustrate that more clearly, with a monotonic increase in the 20-24 year bracket starting with the recovery of the economy from the first oil crisis. Figure 16 reorganizes the data on a cohort basis; the very different behavior of the youngest cohort stands out. What of course no one knows is whether the higher labor force participation of this most recent cohort will continue, and eliminate the "M". If so, it hints at increased delays in marrying and as a consequence an even greater divergence between realized fertility and the cultural norm of two children.
I do not provide details on changing marriage patterns here; they are consistent with this story, with the coda that an increasing number of women are not marrying until it is too late to have children. Instead I conclude this section by examining shifts in fertility. As Figures 1 and 8 illustrate, fertililty in Japan has fallen over time; what these data highlight is that the timing of fertility shifted as well, from a peak at age 25 to a peak at age 29. While fertility at all ages remains far below the levels of 1950, in 2001 fertility was high at older ages than it was 30 years earlier, in 1970. As births shifted later in women's lives, realized fertility fell. Through the cohorts born in 1955 (whose children were thus born in the early 1980s), women still managed to achieve their target family of 2 children.9 Thereafter, women were unable to make their target before the effects of aging made that impossible. The data suggest that fertility is stabilizing, as the behavior of the 1980 cohort, the most recent to have entered peak childbearing years, is almost identical to that of their elder 1975 cohort. Figure 10, in fact, suggests a very slight increase for the 1985 cohort. However, there is nothing to suggest that fertility will increase more than marginally; absent a sudden shift, the total fertility rate will remain far below replacement levels.
From a policy perspective, it is clear that women face a dilemma, needing to choose between work and family, and that they are opting for work. Fertility is far higher for women who have mothers or mothers-in-law able (and willing) to take on childcare. It is also higher among professionals, who have a greater ability to exit work and return to a high-paying job. But for many women daycare remains inadequate, with hours too short to permit commuting or overtime. The Japanese media in 2006-7 carried many stories about new daycare services, and firms with on-premises facilities. Nevertheless, elementary schools still expect there to be a stay-at-home parent, and in general there is no sense of crisis among policymakers – unlike in the case of eldercare.10
Economic decisions over education and labor force participation tend to evolve slowly; policy in this area in Japan is moving no faster. Over the next 2 decades, young workers will become increasingly scarce, and it is not unlikely that firms (with the support of local governments) will respond on an incremental basis to attract and retain women as workers. This may stem the increase in late marriages, and increase fertility within marriages. A return of today's teenage women to the patterns of their mothers, who were born in 1965, would still leave a realized total fertility rate of 1.6. That is far higher than the current rate, but far lower than that needed to stabilize Japan's population. Even if fertility rises more sharply over the next decade, a child born in 2017 is likely to go on to a 4-year college, and will only just be entering the labor force in 2040. Changes in the behavior of the young will not remove the tensions Japan faces in how to deal with its rising number of retirees.
Part V: Splitting the Pie: The Policy Dilemmas of Aging
The pressures of aging are highlighted by two simple numbers: currently there are 3 people of working age (15-64) for every person age 65 and above. However, in 2023 the ratio will hit 2:1 and by 2055 it will have fallen by over half, to 1.26. (See Figure 19.) Supporting retirees during the coming decades will be one of the major challenges facing the Japanese economy, if not the major issue. Much of the discussion of this issue, including by economists, misses the fundamental nature of the problem. It is not that people may not be saving enough for retirement, or that the "trust fund" of accumulated social security receipts will run out of money. At a national level, it is simply incorrect to think of this as an issue of saving too much or too little, because very few resources can be transferred across time; most consumption consists of services, not goods.11 Retirees' consumption will be out of the current output of the economy. At an economy-wide level, retirement is fundamentally a pay-as-you-go proposition. The issue then is how to see that they get their (fair) slice of the pie.
To clarify this, let us consider what must happen for retirees to be able to consume in a purely private system, in which households must provide for all of their retirement out of their own savings, without any supplement through government pensions and health care coverage. During their working lives, households accumulate savings in the form of bank deposits, life insurance policies, stocks and pension funds, and owner-occupied housing.12 When they retire they gradually draw down these assets. To turn those assets into consumption, however, the purchasers must on net be productive workers, whose saving offsets the dis-saving of retirees. Workers as a group cannot consume all that they produce, they have to restrain their impulse to consume – that is, save – and thereby free up goods and services for retirees to consume.
There is no reason, however, to believe that markets will equilibrate savings by workers with the needs of retirees. Assume for example that retirees consume only half the level of goods and services as the average person of working age. With an eventual ratio of only 5 Japanese of working age for every 4 retirees, they must on average save 40% to provide output sufficient to support 2 persons.13 Because not all output in an economy goes to consumption, the requisite level would be only about 60% of this, or 24%. Nevertheless, since household savings averaged only about 10% of GDP in the 1980s, and fell to 2% in 2005, it seems highly unlikely that private markets would operate in a manner that would allow today's older workers to retire with a modicum of comfort.
Forced savings will therefore be needed. The Japanese government will need to tax workers to get them to lower their consumption, and in turn transfer those resources to older workers in the form of pensions and health care. It is of course already doing that, with social security taxes running at about 14% of GDP. In the future, therefore, the government will need to depress the consumption of worker households by another 10% of GDP. Including social security and healthcare taxes, Japan's level of taxation of around 34% of GDP is low in comparison to the OECD average, which is near 50%. It is thus clearly possible for the economy to handle the burden of retirement without condemning the elderly to penury.
Avenues for adjustment
Whether or how the government will help this adjustment process is not clear; there is after all no precedent in other economies that can inform thinking on this issue. Generational warfare via the ballot box and the formal political process in not inconceivable. There are however a variety of avenues of adjustment – and directions of change that, while important, will not fundamentally change this problem.
First, economic growth is certainly desirable, but does not change the underlying problem. Whether the Japanese economy grows at a robust 2% per capita, or stagnates at 0.5% growth per capita will affect the level of consumption. In the former case, in 35 years or by 2042 the average level of consumption would double; in the latter case it would rise a modest 19%. In either case, however, in order for retirees to finance their consumption, the share of total output that the average worker consumes must fall, and voluntary changes will surely be insufficient. Politically, raising taxes when incomes are growing will surely prove easier than if incomes are stagnant. It may also be easier to restrain the level of income of retirees relative to workers when overall economic growth means that the absolute level of retiree incomes still rises. The baby boomers however will retire with expectations that they can enjoy a modicum of comfort; after all, much of the prosperity of the younger generation is due to the sacrifices of their parents and grandparents. In addition, all too many will reach retirement age with little in the way of private savings, in part because they believed they could depend on government pensions and health care programs, and in part because they were unlucky in their careers or had poor health (and in part because of poor planning). It is not clear why the baby boomers should be asked to make unusual sacrifices as they age; surely they will resist changes in that direction. To reiterate, growth in and of itself will not allow Japanese society to avoid unpleasant tradeoffs between the interests of workers and those of retirees.
Second, higher fertility will not solve the problem. The Japanese government provides a range of population projections. The most optimist is their high-fertility, high-mortality projection. Under it the economy hits a working age to retirement age ratio of 2:1 in 2025, and a 1.5:1 ratio in 2047. At the opposite end is the low-fertility, low-mortality projection, which hits a 2:1 ratio 4 years earlier, in 2021, and a 1.5:1 ratio in 2037, some 10 years earlier. By 2055 the range varies from 1.45 in the optimist scenario to 1.11 in the pessimistic one, and 1.26 for the central one.14 A half century out, there is a range of -12%/+15% around the central projection, significant for pension planning, but qualitatively similar in terms of macroeconomic issues discussed here. In other words, it is already too late for an increase in fertility to do more than ease the problem, even a half century out. Nevertheless, it would still be beneficial if fertility would increase towards the replacement level, but at present there is no particular indication it will do so, nor has there been a shift in that direction in the Mediterranean economies whose demographic behavior most closely parallels Japan's.
Third, other margins of adjustment can lessen the tradeoffs that Japan will face by increasing the number of individuals in the labor force relative to the working age population. The origins of Japan's population aging lie in gradual shifts in the reproductive behavior of Japanese women, which in turn are connected to shifts in income, education, labor market opportunities, and the relative wages of men and women to each other and as they age over their working lives. Japan is still distinctive for its "M" shaped pattern of labor force participation by women, as women exit in the labor market in their late 20s and early 30s to raise a family. Gradual shifts in the availability of daycare and the willingness of employers to be more flexible may ease that in coming years; there is certainly plenty of discussion in the Japanese media of the desirability of this, and of firms that are providing on-premise daycare facilities. Nevertheless, the government exhibits no sense of urgency in providing facilities open long hours or on holidays. The current trend instead is for women to delay marriage and, within marriage, to delay childbearing, which will accomplish much the same labor market outcome. However, it is probably too late for older women to launch careers who have not already done so, while the number of young women will continue to fall. As Figure 16 illustrates, the most recent cohort to hit the late 20s in age has set a new record for participation at 75%; if they remain in the labor market, this will boost female participation by at most 10%, and overall labor force participation by 5%.
Another margin of is immigration. In some local communities that is already important; in one municipality Brazilian-Japanese account for 16% of the local population. Overall the number of immigrants has doubled in the past decade, to 2% of the labor force. A further increase is likely, indeed policy shifts anticipate that. Even a steady inflow of 1% of the labor force per annum (or up to 600,000), more than anyone currently anticipates, would have a lessor impact that at first glance.15 First, many migrants will have workers; it is not realistic to think that Japan can run a successful "guest worker" program under which workers will come as single individuals and return home after a few years. To get the nearly 4 million additional workers Japan would need to fully offset the economic impact of the decline in the working age population over the next half-dozen years would require that Japan attract 7.5 million immigrants to get 5 million added workers in the labor force. Furthermore, the Japanese language is idiosyncratic, impeding effective literacy among immigrants; even in other countries, the education of immigrants is not worth as much in the labor market as that of the native-born. In terms of output, one immigrant is thus on average worth less than the average Japanese worker. Using an 80% "replacement" ratio, 7.5 immigrants would those be needed to offset the 4 million drop in the size of the labor force. Finally, only some immigrants will return home, and hence will in due course themselves retire; their children will also boost school enrollments, so that not all of the taxes they pay can go to the support of retirees. The long-run impact of immigration is thereby lessened, absent ever-increasing numbers of immigrants.
One modest margin is the income from foreign investment. Already Japan earns 3.5% of GDP in income from overseas holdings of bonds, stocks, and direct investment in local factories and other businesses, almost as much as the overall trade surplus. It is difficult to project the future level of such income, as various forces will work towards the appreciation of the yen over time, which will lower the purchasing power of foreign income. Nevertheless this will over time permit Japan to finance the purchase of goods and tradable services without needing to export domestic production. In other words, in the future Japan can (and will) run a trade deficit, to its benefit. By importing goods rather than producing them, Japan will be able to shift workers out of manufacturing and into the production of non-tradable services such as healthcare. While limited in scope, it is thus possible through international investment and trade flows to turn current savings into future consumption. At a level of perhaps 3% of GDP, that is equivalent to an equivalent boost in the labor force.
A final margin of adjustment is to delay retirement itself. At present older Japanese work in greater numbers than is the case in the US, and in far greater numbers than in Germany. In the US, for example, only 77% of men age 55-59 are in the labor force, while the number remains stable at almost 95% in Japan. While more Americans are delaying full retirement, and participation by older women has risen by 20 percentage points, the movement remains in the opposite direction in Japan, though at present it is restricted to individuals age 60 and above (see Figures 12 and 13). In the short run, therefore, it seems more realistic to expect that the labor force participation rate of older workers in Japan will continue to fall. Policy will eventually encourage that to stabilize, as the age at which full pension benefits will be received is already set to rise gradually over the next 10 years, to age 65 and then age 67.16 Further changes in that direction could encourage older workers to delay full retirement, and perhaps even reverse the trend toward earlier exit from the labor force. Of course this is a tax increase by another name, as it increases taxes paid and lowers benefits received over an individual's lifespan. The potential impact is large. In 2032, for example, a shift from the cutoff for the working age population from 64 to 65 would add 1.6 million individuals to the potential labor force, an increase of 3%, while cutting the number of retirees by the same numerical amount and, because there are fewer retirees than workers, decreasing the number of retirees by 4.5%. Pushing the target working age to 69 would in 2032 shift the numbers by 7.5 million, a 12% rise in the working age population and a 20% cut in that of retirees.17 Such policies could thus make a very large difference in the dependency ratio and the fiscal tensions imposed by aging. Furthermore, they may be more palatable politically than a direct tax increase.
Because retirement needs must inevitably be covered out of current production, at the macroeconomic level it is not possible to save today to finance retirement tomorrow. Furthermore, the private savings of future workers will not free up enough resources to provide for the sustenance of their elders. The government will need to force future workers to cut their level of consumption – force them to save – relative to today's level. That will be politically costly. In addition, aging will slow growth, and tax increases may accentuate that. Slower growth will make the fiscal tradeoffs needed to finance aging more painful, as absent an unprecedented level of productivity growth future workers will see little increase over time in their take-home pay. Productivity increases however will likely remain low, as the flattening of age-wage profile diminishes incentives for education while slower growth will make firms slower to introduce new and better equipment.
In the aggregate, the shift in GDP is substantial, on the order of 10%. Since Japan is currently running large budget deficits and has a high level of debt, the over shift in fiscal position must be larger. Not all of this need be through tax cuts. Increases in female labor force participation, immigration and the retirement age will all help, as will a shift in resources from programs aimed at the young (primarily education) to those aimed at the old (health care). While policy shifts in all of these directions are under discussion as of 2007, the pace of change on immigration and support for female workers is at best slow, and to my knowledge there are as yet no concrete proposals to change the age at which pensions can be received or otherwise encourage a delay in retirement.18 Instead, after 5 years of budget restraint, policy seems focused on tax increases, with a boost in the consumption tax from 5% to 7.5% by 2011.
As social scientists this is a fascinating set of issues. How will the politics develop, as large tax increases appear necessary? Can Japan successfully cap healthcare costs (unlike the US)? Will fertility stabilize and women's labor force participation increase? How about immigration, which is already high in the EU and the US? Finally, what will a society comprised primarily of the old look like? Since some regions of Japan are already much older than the national average, fieldwork can provide insights into this, far in advance of the aging of the population as a whole. Part VI: Postlude
How will this affect the young? They face a future laden with known burdens. In contrast to the prelude, the outlook is not uniformly pessimistic.
The winner of the 136th Akutagawa prize was again by a young author centering on a young character. The 23-year-old author, Aoyama Nanae, starts her novelette "Waiting for my chance" with athe only slightly younger Mita Chizu drifting from high school to living with a boy with whom, other than sex, she has no particular relationship. Estranged from her divorced mother, she ends up living with a distant relative, a 71-year-old woman, working at a series of part-time jobs (for part of the story, two at the same time), jumping into an affair with a boy she meets through her job, and then suffering from his leaving her for a different and more exciting girl met through work. But unlike in the previous year's story, the main character here matures, gradually developing a deeper attachment to the old woman with whom she lives and to society in general. Throughout Chizu reflects a bit of iconoclasm, from stealing small items from acquaintances to her refusal to consider college because she doesn't want to study any more. At the same time, she is pictured throughout as a diligent worker, and at the story's closing this leads to the offer of a permanent job, including an apartment in company housing. Ironically this is just at the point where she seems to be finally at home with the old woman with whom she lives. Still, she values her personal freedom, while at the same time admitting there is something to be gained from the stability of a "regular" job that offers steady employment and perhaps even a career. Her attitudes however remain independent: having been burned before, and having a job she is willing to stick to, she prefers an affair with a married co-worker to seeking a normal boyfriend with all the commitments that might eventually entail.
The lives of the older people are depicted through Chizu's lens, of course, and we get glimpses of that other world. While never detailed, it is clear that her distant aunt went through a lot; her apartment is adorned with pictures of past companions – cats – but there are hints of a more complex history, beyond being a widow. Her aunt's existence is also relatively spartan, a house she owns (but which is small and aging), the occasional meal out with an older male friend, sweets brought back for snacks from outings. It is not clear to what extent Chizu (or her mother, who had been sending money to Chizu) contribute to her budget. What is clear is that she has very little to fall back upon, either in terms of relatives or even a circle of friends; the New Year's envelope she gives to Chizu has only ¥1,000 in it. While she has a social life – she attends a dance group, and sees a male friend Uncle Hosuke – Chizu's one nightmare is finding her aunt dead; there is little sense that she or her aunt know how to deal with the medical system or otherwise handle aging, only a vague unease at her prospects. Her aunt's deep comfort with her surroundings, perhaps best labeled serenity, is a contrast to and steadying influence on Chizu.
Does this speak of Japan's future? The young people here are in proper jobs or move into them. Though there is no mention of the bubble and its lingering aftereffects, the sense is that youth will be pulled back into more productive economic roles, and become regular members of society; that is a stark contrast to the "lost generation" of the previous year's winner. It does however reflect common elements. None of the individuals in these two novels have real ambitions; they do not aspire to careers, they are not passionate about a hobby, they even display only the occasional passion for the opposite sex – while showing no aversion to engaging in sex itself. Older individuals barely appear in one book; in the other, older individuals are central, a resource, but at one level living in another world. Dealing with aging is not seen as a problem for the young, who are quite oblivious to the abstract issues of social security.
The following sequence comes towards the end of the book, followed by the parting scenes of Chizu and Ginko-san, and a closing vignette describing her adjustment to her new life as she takes the train that runs past the house of Ginko-san, whom she's not gone back to visit.
The first day of work after New Years I was called over by my boss. The boss, with his white-streaked hair, had decorated the top of his desk with a New Years rice cake, the cheap sort of sold in supermarkets. I still took pains to praise it as cute. After exchanging the requisite pleasantries about our holidays, he paused for a moment, and in a suddenly quiet voice quickly asked, "Would you like to become a regular employee?"
"Uh. Who, me?"
"Yes, you. We've adjusted our workforce a bit, right? Ms. Mita, you seem to be taking your work seriously."
"A regular employee?"
"Indeed, could you think about it? There's also an opening in the company dorm, so if you wanted, you could even move in there."
I told him I'd think about it – what should I do? Had the time finally come to start putting down roots? Even though I'd been working steadily since April, I'd only managed to save 35 man.19 So even though it soon would be a year since I'd come to Tokyo, I was still a long, long way from my goal of saving 100 man. If I became a regular employee, I'd surely earn a bit more. Even though I didn't have any particular reason to save money, maybe because it was the most concrete-seeming goal, well, 100 man, that was the only concrete figure I had.
But when I began to think seriously about where I could I leave that house, I couldn't help but feel bad about Ginko-san. Maybe that's what sympathy is, but after I'd finally become used to living around her, the thought crossed my mind that I couldn't just up and leave.
During lunch break, Ando-san joined me for lunch, and I had a chance to ask her things. Since there wasn't a cafeteria at the employee dorm, she said she'd by things at the local convenience store and eat in the rooftop smoking area. When the weather was nice it was pleasant to go outside, but when it was cold she'd go back in right away.
"The dorm? Well, it's a straight shot on the train to here, that's really great. Mita-san, aren't you coming all the way from Chofu?"
"So, it's a short commute...?"
"Yes. And it's cheap, too, and with fairly nice rooms."
"Cheap, and fairly nice?"
"Yes, why, what's up?"
"So, you mean, you're going to become a regular employee? Of course, in our department two people are quitting on us this month, that's great!"
"I guess it's good to be a regular employee."
"But of course it is?! What do you do about health insurance? You'd be in trouble if you really got sick."
Ando-san had finished eating her pasta, but stopped in the middle of cleaning up her stuff, with a look of disbelief on her face. She turned to look at me, a bit of orange sauce still sitting on her lip. "Ugh. Don't you know? If you don't have insurance, then they really sock it to you when you visit the hospital, and you've got to pay it all yourself."
"Is that the only thing?" I carefully wiped around my mouth with the towelette I'd received at the convenience store.
"Well, I don't really know. Of course, there's probably more than that."
"Uh, can a regular employee save up money?"
"Well, our company isn't doing that great, so you ought not get your hopes up too much. But if you live in the dorm, well, you can manage a bit."
So would even I end up being recognized as a proper OL [office lady]? I'd be paying the resident's tax, and every month there'd be money going for a pension and insurance. Was I becoming an ordinary adult? "An office lady..."
Sitting next to my, Ando-san interjected while puffing her cigarette, "Don't you want to?"
• • •
When the time came around to see "Valentines" emblazoned across department store banners, Ginko-san suddenly announced that she wanted to by chocolates.
"What, you want to give some to Uncle Hosuke?"
"Giving chocolate to Uncle...hmmm, I guess that's good to do."
"Chizu, will you go with me? An old woman like me really wouldn't know what to get."
"Huh, I surely don't know!"
"It safest to rely upon the sensibility of someone who's young."
"Gee, don't old folks know best how to please other old folks?"
Anyway, the next Sunday we headed out to a department store in Shinjuku. Ginko-san had on a light purple two-piece suit and was wearing cream-colored pumps. She'd put her white hair up in a single bun, and came across as a compact, fairly attractive old lady.
I turned my face to the floor when we stopped at Sasatsuka Station. Even with the time that had passed, I still couldn't bring myself to look directly at the station platform. I didn't want to find myself catching eyes with Fujita-kun or Ito-chan. How long had it been since I'd come across them? Would they even remember my face?
When we'd left the station and the train entered the tunnel, I finally could look up. In the window opposite was the reflection of Ginko-san and me. The decked-out Ginko-san sat there with her eyes closed, snoozing. It was quite something to go shopping for chocolate at her age! When the train would take a bounce, she'd raise her head, only to close her eyes again. I asked if she was tired, and she didn't respond.
Would I become like this old woman? Getting myself prettied up even at age 70? Having my own small house, going out to buy chocolate on Valentine's Day? Could I lead such a life?
[Aoyama, pp. 146-150]20
1 伊藤たかみ (Ito, Takami). 八月のH上に捨てる (Thrown Away on a Summer Road). Tokyo: Bungei Shunju (文|tHj, 2006. Translation by Michael Smitka. The author did not attach readings for names; I chose what struck me as most natural.
2 For Japan see 岩間暁q (Iwama, Akiko). "女性のA業がo生意~にyぼすe響のWFン_[比r. (A Comparison by Gender of the Impact of Female Employment on Desired Fertility" l口問題研究 (Journal of Population Problems) 62, no. 1-2 (2006): 20-34 and also Suzuki, Toru. "Fertility Decline and Policy Development in Japan." Japanese Journal of Population 4, no. 1 (2006): 1-32. Such behavior is not unique to Japan or to modern times; see Schofield, Roger. "Short-Run and Secular Demographic Response to Fluctuations in the Standard of Living in England, 1540-1834." In Population and Economy: From Hunger to Modern Economic Growth, edited by Tommy Bengtsson and Osamu Saito, 49-71. Oxford; New York: Oxford University Press, 2000.
3 Russia is an example of the latter: a middle income country with a declining population. In Russia's case this is primarily due to a sharp rise in mortality. At least one small social group, the Shakers, did die out due to low fertility – they believed in sexual abstinence, and were unable to maintain in-migration (in the form of new recruits) sufficient to maintain their population.
4 金q龍一, 釜野さおり, 大石亜希q, 佐X井i, rm上正q, Oc[美, 岩V美帆, and X泉理b. (Kaneko et al.) "結婚とoYに関するS国調査:vw調査の結果Tv(第13回o生動向基本調査) (The National Survey on Marriage and Childbearing: An Overview of the Results of the Couples Component of the 13th Survey)."l口問題研究 62, no. 3 (2006): 31-50. This paper draws upon the surveys of marriage and children (1st, 1940; 13th and most recent, 2005).
5 In a given year, the Total Fertility Rate is a synthetic measure of the number of children a woman would have over her reproductive years, given the age-specific fertility rates observed in that year. No single number can capture the complexity of reproductive behavior. The TFR, for example, will normally differ from the fertility rate of a given birth cohort. Likewise, as noted later, the TFR can be falling even while the total number of births continues to rise, because of the age structure of the population. Given the current low rate of female mortality, a TFR of 2.1 is necessary for a population to be stable in the long run.
6 Projections from the National Institute of Population and Social Security Research, www.ipss.go.jp. These data are from Mm1-9.xls (middle fertility middle mortality projection, annual by age by gender, 2006-2055).
7 Consistent with this, the better macroeconomic conditions the past 3 years improved the labor market for men, and according to media reports in Nikkei marriage rates have risen, at least temporary. The statistical sources I have on hand only provide data through 2004, before this occurred. The income-marriage link is widely observed, from medieval England through to various modern societies.
8 I have not data thereon; it would be interesting to know how parental norms have changed.
9 Note however that in and of itself delayed fertility slows population growth.
10 Explaining why this is so is one theme of Schoppa, Len (2006). Race for the Exits: The Unraveling of Japan's System of Social Protection. Ithaca, NY: Cornell University Press.
11 The primary exceptions are housing and consumer durables. In 2005 housing, clothing and household goods averaged 24% of personal consumption, an upper bound since it includes some perishables and services.
12 Children today are only a psychic asset. Not all retirees will have children living nearby, and given increased longevity, the children of many retirees will themselves be retired. See Schoppa (2006) and Jenike, Brenda Robb. "Parent Care and Shifting Family Obligations in Urban Japan." In Demographic Change and the Family in Japan's Aging Society, edited by John W. Traphagan and John Knight, Chapter 9, 177-201. Albany: State University of New York Press, 2003.
13 Consumption is only about 60% of output; other resources flow into corporate investment and government services, though government activities hopefully indirectly benefit consumers.
14 Most sources use only the "central" medium-fertility, medium-mortality projection, including Figure 19 in this paper and the accompanying set of population pyramids.
15 This analysis draws upon that of Robert Feldman of Morgan Stanley Japan, summarized in the Global Economic Forum of January 8, 2004. See the archives at http://www.morganstanley.com/gef.
16 However, Japanese government discussions in fall 2007 suggested that a scheduled increase in co-pays for healthcare would be delayed.
17 These data do not attempt to adjust for likely labor force participation, as individuals with high levels of savings might choose to retire despite financial penalties, while the rate of disability also rises with age. At the same time, participation by those over age 70 might also rise. On net actual benefits would likely be less.
18 Based on a quick scan of background materials made public in fall 2007 by the 経済?政?問会? (Fiscal Policy Advisory Council). The focus is on the estimating the extent of future healthcare obligations, and (given current debt levels) the requisite level of tax increases needed to stabilize Japan's fiscal. The working assumption is that it will be difficult to further trim expenditures as a share of GDP.
19 About $3,500. "Man" is ¥10,000, the largest denomination bill used in Japan.
20 ?山七恵 (Aoyama, Nanae). ひとり日*和 (Waiting for My Chance). Tokyo: Kawano Shobo (河出書房), 2007. Translation is by Michael Smitka; the title is my own choice, and differs from the two other varying titles I have seen in English: "Alone on Fine or Rainy Days" (Fumiko Halloran) and "On My Own" (Japan Literary Prize list at http://www.jlit.net/awards/).