Japan's disappearing population
Japan invariably looms as the central case study for the economic and societal effects of rapid fertility decline, population decline and ageing. Japan is, measured by median age, the oldest country on earth, excluding the greying millionaires of Monaco and the some-5,000 people on British St. Helena. At the end of 2021, Japan had a median age of 48.4, well ahead of the second major country on the list, Italy, with a median age of 46.8. Japan is about to get older still. According to preliminary estimates, the country’s fertility rate fell further last year, albeit marginally, while the gap between births and deaths remained wide as ever. The number of live births fell by 5.0%, to 770.774, while deaths rose by 9.0%, to 1.57 million. Japan’s rapidly ageing population is the result of a quicker and more sustained post-1945 fertility transition than in other developed economies.
The first chart above shows that Japan in effect has experienced two separate fertility transitions since the end of the Second World War. Tsuya (2015) notes that first transition was driven by a broad-based decline in fertility across all ages, while the fall in fertility to below replacement levels has been driven mainly by tempo effects. Tsuya (2015) emphasises that the initial post-war transition, which took the fertility rate from almost four in 1950 to around 2 in 1960, was driven to a large extent by falling birth rates among women aged 30 and higher. The second chart above confirms this. Indeed, it shows that the number of children born to women at the peak of their fertility, at the age of 25-to-27, fell only slightly from 1950 to 1970, in contrast to a collapse in births to women aged 30 or more. Put differently, the initial and quick fertility transition in Japan was primarily due to women reducing births dramatically in the latter part of their reproductive life. From 1960 to the middle of the 1970s, fertility rates stabilised with the exception of the bizarre drop in 1966, due to superstition that women born in that year—the year of the fire horse—would bring bad luck to their future husbands. Fun fact; 2026 is the year of fire horse again, so keep your eyes peeled for another drop in Japanese births in a few years. By 1990, the chart above shows a dramatic shift in fertility, driven by both quantum and tempo effects. The number of births to women at peak fertility fell by almost a third, and the mean age of first childbirth rose. This development continued through the 1990s and into the 2000s, though note that Japan’s total fertility rate did rebound from a low of 1.27 in 2005 to 1.43 in 2016, but it has since dropped back, and if the early 2022 estimates are correct, it is now at a new low. The UN’s most recent forecasts predicts little change in the decade from 2020 to 2030, except from an increase in births to women in the latter part of their fertility career. I am not sure where they get that prediction from.
The drivers of the decline in Japanese fertility will be familiar reading to the people who have followed my work on demographics, and who are familiar with second demographic transition more generally. Marriage rates have declined dramatically in Japan in the past 50 years, and because the country’s relatively conservative values all but exclude out-of-wedlock child births—a mere 1-2% of live births have taken place in non-married unions since 1960—the relationship between falling marriage rates and lower fertility has been particularly strong. The numbers cited by Tsuya (2005) are remarkable. Among women aged 25-to-29, the proportion unmarried in 1975 stood at 18%. By 2010, this number had soared to 60%. For women aged 30-to-34, just over a third of women remained unmarried by 2010, up from 8% in 1970. For women aged 35-to-39, the proportion of unmarried was 23% in 2023, rising from 5% in 1970. The shift for men is even more dramatic, support the culturally charged meme that young men either offer too little in the way of companionship to women, or that women simply have better thing to do in a modern era than to get married and have children with what they perceive as hapless and subpar men. In 2010, the share of unmarried men aged 25-to-29 stood at 72%, up from 48% in 1975, while it had increased to 47% for men aged 30-to-34, from 14% in 1960. For men aged 35-to-39, the share of unmarried had increased to 36% in 2010, from 6% in 1975. In other words, in the middle of the 1970s, almost all men were married by their late 30s, but by 2010, this number had declined to around 60%. In 1975, the celibacy rate for men was just 2%. By 2010, it had soared to 20%.
The data above are linked, not surprisingly, to a significant shift in women’s role in the economy, especially in terms of education and labour market participation. This is true even considering Japan’s relatively conservative gender roles, akin to the situation in South Korea. In 1955, only around 2% of women advanced to higher level education. By 2010, this number had increased to 45%. This change, in turn, has driven a significant increase in women’s labour force participation rate, from 50-to-55% to 65-to-70% of prime-age working women between 1960 and 2010. Note that this remains significantly below the labour force participation rate for men, at around 90% in 2010, down from just under 100% in 1960.
In any model of a demographic transition, it is easy to see why the shifts described above go hand-in-hand with falling fertility, mainly through raising the opportunity costs of child-rearing for women, both in terms of the overall number of children, and their timing. Retherford and Ogawa (2005) offers something approximating a classic second demographic list of the causes for falling fertility in their summary of the sustained decline in Japanese fertility since the middle of 1970s:
Educational gains by women
Increases in the proportion of single women who work
Changing values about marriage
Decline in the proportion of marriages that are arranged
Decline in the proportion of newly married couples who cohabit with parents
Increases in premarital sex
Emergence of the “new single concept” (OK to enjoy single life without pressure to get married)
Increasing desire of women for more help from husbands and a more egalitarian marital relationship
Retherford and Ogawa (2005), p. 9
As I have discussed at length in previous posts on demographics, it is largely futile to discuss whether falling fertility is good, bad or whether we should we be indifferent to it. The answer to this question depends on your perspective, and the questions you ask. For Japan as a society, the political economy has, in line with the discourse in most other economies with rapidly falling birth rates, tended to see low fertility as a problem that should be corrected by public policy. Retherford and Ogawa (2005) highlights that Japan has been engaging in pro-natal policies since the beginning of the 1970s, to little effect.
Table 3, “Major Japanese government actions aimed at raising fertility”, Retherford and Ogawa (2005), p. 27.
Japan has had publicly-funded child allowance since the beginning of the 1970s, a policy that was initially designed to help couples have three children, but have since been expanded to cover all order of births. This allowance was means-tested, which means that couples with incomes above a certain threshold would not be receiving the allowance, at least not based on the rules such as they were as of 2004 Retherford and Ogawa (2005). Low fertility didn’t really become a target for public policy until the beginning of the 1990s, however, with the so-called “1.57 shock”, referring to the widespread attention paid to the fact that fertility fell to 1.57 in 1989. From that point, pro-natal policies in Japan took on a well-known format in their attempt to reduce the opportunity cost for women of child-rearing. Put differently, these were policies designed to allow working women to also have children without financial, or ostensibly career, penalties. Later, policies were re-designed to encourage fathers to take more leave, thereby sharing some of the burden with mothers in the early, and most demanding stages, of child-rearing Retherford and Ogawa (2005) and Tsuya (2005). It is difficult to truly tell whether such policies have been successful due to the lack of a counterfactual. It is possible that fertility would have declined further and more aggressively without them. But if we evaluate these policies by their ability to halt the decline of fertility, let alone raise it, they have not worked, in line with the experience in most other developed economies. Fast forward to present day, Prime Minister Fumio Kishida has doubled down on the objective to reverse the country’s ailing birth rate, and rapidly ageing population, by devoting as much as 3.5 trillion yen—$25 billion—to various child care and parental leave support measures. As in South Korea, conservative gender roles loom as a key driver of the decline in fertility itself, and the inability of government to do anything about it. Women of child-rearing age in Japan today are actively distancing themselves from the traditional role as housekeeper, and men find it increasingly difficult to live up to expectations of women, both in terms of their employment and career capabilities as well as their ability and willingness to help out at home. One is tempted to ask exactly what kind of man Japanese women want—the breadwinner or the helper at home—but the reality is that she wants a combination of the two, something which Japan’s still-rigid society is finding it difficult to produce. In the meantime, it is also fair to say that an increasingly large share of Japanese women, as is the case in South Korea, actively choose to forgo the traditional partnership, as an act of rebellion against old, and in their view, out-dated values, and in order to reap the rewards in the labour market, which for their generation, offers real opportunity, unlike for previous generations.
Older and fewer
Whatever sociological and demographic analysis we apply, Japan’s population will continue to age and depopulate. Even if we incorporate a significant reversal of the tempo effect—recuperation of fertility for women who have postponed having their first child—fertility in Japan will get nowhere near replacement levels anytime time soon. Coupled with by now significant population momentum—the fact that each generation of childrearing women is now smaller than the previous—the trends in the two charts below will continue.
Japan’s median age will soon surpass 50, a remarkable number for a major advanced economy, and its population will continue to shrink. According to numbers from the UN, Japan’s population has shrunk by just under 6M since 2005, ex-migration, to just over 124M with the annual decline—the difference between live births and deaths—now standing at 700-to-800k, up from just under 500K in 2015. These numbers, in turn, mean that Japan will retain its role as the great experiment, both in terms of demographics but also overall macroeconomic dynamics given its economy characterised by endemic low domestic demand, low inflation, elevated external surpluses and rapidly rising public debt, facilitated by the central bank. It is, and will remain, a fascinating macro experiment to follow.
Articles referenced above:
Retherford and Ogawa (2005) - Japan's Baby Bust: Causes, Implications, and Policy Responses, EAST-WEST CENTER WORKING PAPERS, POPULATION AND HEALTH SERIES.
Tsuya (2005) - Below-Replacement Fertility in Japan: Patterns, Factors and Policy Implications, in Low and Lower Fertility - Variations across Developed Countries, Rindfuss and Choe (2015).