Japan’s Population Challenge

In one of those strange coincidences the world’s second and third largest economies are both scheduled to have elections next month – Japan on 11 September, and Germany on the 18th. The coincidences go further, since these two societies are leaders in another, non-economic, sense: they are leaders in the great global ageing revolution. Japan (at 42.64) and Germany (at 42.16) have the highest median ages of any OECD country (Italy comes third at 41.77, while the US is still a sprightly and young 36.27). The two countries also share the problem that they cannot pump up economic growth by introducing more liquidity (money) into the system, or at least the impact rates of doing so have become very low (see this post on the German problem) .

In fact the comparison goes even further since at the end of the day the elections are about pretty much the same issue: how to make the respective economies grow fast enough to be able to pay for the growing pension and health needs.

The Financial Times today has an interview with outgoing Japan economics minister Heizo Takenaka, and Takenaka is quite explicit: changing demographics is Japan?s biggest challenge.

?From now on, the total population of Japan will start falling,? he said. ?That means if we don?t create a system in which the private sector can carry more responsibility, the burden on taxpayers and on the state will become unsustainable.?

The facts indeed are compelling. Japan is in the process of becoming the first major state to experience structurally declining population (correction: as David points out in comments this is misleading for a number of reasons so lets say Japan is the first post-modern state or society with a developed economy to experience structurally declining population. It was not my intention to indirectly pass any judgement on whether Russia is a major state or not). Population was predicted to start falling in 2007, but preliminary data from the first half of the year suggest that the decline may have already begun. There is no end point in sight. The decline, which is also known as the second stage of the demographic transition, is a result of continuing low fertility, and so far there is no clear way back. Most demographic models suggest that fertility in low fertility countries may bounce back a little, but not sufficiently to reach the 2.1 TRF replacement rate (normally the assumption is in the region of 1.9, and estimates vary as to whether this will arrive in 50 years, or in 100). Meantime the UN median estimate seems to suggest that the total population of Japan (currently something over 125 million) will be around the 45 million mark come 2100. Europe and the US look carefully and take note.

This entry was posted in A Fistful Of Euros, Economics and demography and tagged , , by Edward Hugh. Bookmark the permalink.

About Edward Hugh

Edward 'the bonobo is a Catalan economist of British extraction. After being born, brought-up and educated in the United Kingdom, Edward subsequently settled in Barcelona where he has now lived for over 15 years. As a consequence Edward considers himself to be "Catalan by adoption". He has also to some extent been "adopted by Catalonia", since throughout the current economic crisis he has been a constant voice on TV, radio and in the press arguing in favor of the need for some kind of internal devaluation if Spain wants to stay inside the Euro. By inclination he is a macro economist, but his obsession with trying to understand the economic impact of demographic changes has often taken him far from home, off and away from the more tranquil and placid pastures of the dismal science, into the bracken and thicket of demography, anthropology, biology, sociology and systems theory. All of which has lead him to ask himself whether Thomas Wolfe was not in fact right when he asserted that the fact of the matter is "you can never go home again".

33 thoughts on “Japan’s Population Challenge

  1. David Weman is right: Russia is obviously a major state and its population is already declining. And, equally clearly, Japan is the economically most important state to find itself in this predicament.

    In the short and medium terms stable or falling numbers in population are a greater problem for ethnic states (and probably the greater the more strongly the state defines itself as an ethic state). But in the long term the whole world will probably face the same situation so what the leading edge of the wave does is important.

  2. “David Weman is right: Russia is obviously a major state and its population is already declining.”

    I agree, I’m just going home to eat humble pie.

    “What’s a “major” state, and why do you count Russia as minor?”

    Sorry, slip of the keyboard in one sense, blind spot in another. I was trying to rule out small, localised hiccups which someone might pick me up on, and I forgot a whopper.

    Actually I was thinking of Bulgaria at the time of writing ‘major’ and even then I realised it wasn’t a very satisfactory description: I don’t like the phrase myself, but didn’t want to keep repeating ‘leading economy’ or something.

    The point is, in the general curve of the demographic transition, the central and eastern european societies are ‘outliers’ since they are following a slightly different path from what we have seen and are seeing up to now with the OECD bloc, and this is what I want to get at. Unlike the rest, they may well get old before they get rich due to their complicated 20th century history.

    Anyway, linke Dave Schuler says the ‘leading edge’ is what I’m getting at and want to focus on here. We had quite a discussion on Russia here

    http://fistfulofeuros.net/archives/001754.php

    So I think I’m correcting the original.

  3. For those who like to know a bit more about Japan’s population pyramid and age cohorts. Take a look at table B1 and B2 in “Monthly Statistics of Japan”. First take a look at B1 and note the net increase for the total population and live births (curiously Japanese emigration was actually higher than imigration in 2004!).

    The explanation why this is such a problem for Japan can be found in table B2. Look at the age cohorts and compare the year 1998 with 2004. Note what happens around the 30-34 age group. All the following (younger) groups are getting smaller.

    http://www.stat.go.jp/english/data/geppou/index.htm#b

  4. Edward,

    Ignoring Russia and the eastern European countries seems shortsighted, especially if you are really interested in this in terms of forecasting the future development of Europe. They have had pronounced population loss for ten years or more now, and really should be considered as models, at least to some extent.

    First off, we really don’t know what will happen with fertility rates in Japan, or really, anywhere else in the future. They are notoriously unreliable, and predictions are almost always wrong.

    Second, this is really a problem of affluent societies, at least for democracies. That means that if the problem becomes really acute and the economy starts to falter, this is probably self-correcting.

    Third, so what? Aging of the population can be judged somewhat like the weather. We can change it, but only in unpredictable ways which we really don’t understand.

    What precisely should we do about it? What’s the point in posting this stuff if you don’t talk about what you are really interested in? That seems to be: increasing immigration to Europe, admittance of Turkey to the EU, etc.

  5. The US is currently at 2.1 children, which is the replacement rate. Europe is not, and it will be in dire straits unless European women start having more children. Their other option is allow more immigration. But since the only immigrants they get are from Islamic countries, this will lead to Eurabia. Either way, Europe is in deep trouble.

  6. Just once, I’d like someone to be brutally specific when they say “Europe will be in dire straits”. What does “dire straits” actually mean? Does it mean civil war, mass death, and the collapse of civilization? Does it mean getting slightly poorer over time? Does it mean blue hair rinse displacing Salon Selective on store shelves?

    Frankly, most of the world would kill to have the “problems” western Europe is currently facing.

  7. zman, immigrants to EU came to from Latin America too. And from sub Sahara Africa, India, Pakistan, and China.

    DSW

  8. ?From now on, the total population of Japan will start falling,? he said. ?That means if we don?t create a system in which the private sector can carry more responsibility, the burden on taxpayers and on the state will become unsustainable.?

    How trusted (and how honest) are Japanese politicians.
    I can tell you that if a US politician said this (or anything along these lines) I would immediately dismiss it for what it is; yet another attempt to further today’s agenda in a way that has no actual real-world relevance to the problem. (cf Social Security)

    (1) From now on, the total population of Japan will start falling.
    Probably true.

    (2) That means if we don?t create a system in which the private sector can carry more responsibility, the burden on taxpayers and on the state will become unsustainable.
    Explain this one to me in more detail, dude; I don’t quite follow the jump here.

    My point is that, if you want to convince me that Japanese politicians are serious about dealing with the issues of future demography, you’re going to have to do better than a recycling of the good old “Why the Bombings Mean That We Must Support My Politics” argument.
    http://www.adequacy.org/public/stories/2001.9.12.102423.271.html

  9. “Frankly, most of the world would kill to have the “problems” western Europe is currently facing.”

    Exactly, right now this is “still” a 2nd order problem. Europe has the skills, technology and resources to tackle this. However the general public still isn’t aware or ignores these long term problems. At some point hard choices will need to be made so there needs to be a show of political leadership to force unpopular decisions.

  10. Hektor Bim wrote:

    Second, this is really a problem of affluent societies, at least for democracies.

    Hektor, I think the issue goes well beyond the affluent democracies (depending on how you define both “affluent” and “democracy”, I suppose). Mexico which while middle-income is generally not characterized as affluent has a similar demographic issue. China, which is now considered middle-income (amazing, isn’t it?), does, too. Check the projection for China in 2050. I’ve written on the graying of China on my blog.

    Note my selection of words, also. Is the increasing degree of seniorization really a problem? I’m not so sure. It’s going to require some adaption by societies that have created social structures based on the assumption of an ever-increasing mob of young workers following on the heels of the seniors. Adapting is part of life.

  11. The problem of ageing usually presented as a decline in workforce participation, that is, a labour shortage. So it should raise eyebrows that many of the countries leading the charge into old age are simultaneoulsy suffering from elevated unemployment. The stock answer is that their labour markets are insufficiently flexible, an independent problem.

    But the question I want to ask is, what if inflexible labour markets are partially causational in low fertility? When employers pay for non-wage benefits on a per-employee basis, either into state-run social programs or via private third parties, they have a strong incentive to favor fewer employees doing more work per employee than more employees doing less work each.

    For the single-income family, this means that the working parent spends more time outside the home, which works against shared parental responsbility (a vital factor in maintaining fertility in highly developed economies).

    For the two-income family, this means that both parents spend more time outside the home, which works against parenting in general.

    The happy medium might be a “one-and-a-half income family”, where each parent works outside the home some 30 hours a week, and shares parental responsiblity. Having two incomes, the family is placed at less risk in the unemployment of either parent, nor is either parent in the position of strict dependency on the other.

    However, as discussed above, when the employer has per-employee (as opposed to per-labour-unit) costs, part-time employment is scarce, especially skilled part-time employment.

    If market conditions could be changed to not discourage skilled part-time employment, the benefits of increased labour flexiblity would extend beyond having possibly more fertile families. Aside form parents of young children, retirees represent a second pool of talent not employed under a full-time or no-time regime, whether due to an increased preference for leisure or to mild disability; and most commentators have agreed that if retirement can be delayed (or in this case, gradualized) the drop in workforce participation associated with ageing can be mitigated.

    As a related observation, a worker with two half-time jobs may have greater confidence as a consumer than one with a single full-time job, since they already effectively have unemployment insurance, without state intervention.

  12. “However the general public still isn’t aware or ignores these long term problems.”

    I would say that’s not entirely true here in Sweden.

  13. “I would say that’s not entirely true here in Sweden.”

    Well, what about in Germany, they don’t seem exactly thrilled by the reform programme, if that is any indication.

    What I would say David is that the general public isn’t aware of the importance of the problem, just like with climatic problems. Everyone knows its getting warmer, but they don’t see any need to do too much right now, its the future’s problem (some people call this inter-generational solidarity). Well likewise todays pensioners aren’t too thrilled about programme cutbacks.

    I’m convinced that the implications of all this are much more serious than even most specialists realise. But I don’t expect people to believe this just because I say it. I think it needs to be demonstrated, both theoretically and by explanatory models which help understand what is happening, and that, even if it sounds fairly abstract, is what I’m really into.

    “curiously Japanese emigration was actually higher than imigration in 2004!”

    Thanks for the info CapTVK. One of the thinks which will help amplify all this beyond what people are expecting are the negative feedback elements, emigration of young people is one of these. We should not be surprised that this is happening, but it will make things worse.

    Ditto Robert’s point.

    “When employers pay for non-wage benefits on a per-employee basis, either into state-run social programs or via private third parties, they have a strong incentive to favor fewer employees doing more work per employee than more employees doing less work each.”

    Well it’s much more than this, they have strong incentives to take their work elsewhere, ie offshore. The lower the dependency ratio the higher the level of government income needed to support the aged extracted per job. There must be some critical numbers in there somewhere that really work against employment creation, and Germany must be somewhere near those numbers.

    “Aside form parents of young children, retirees represent a second pool of talent not employed under a full-time or no-time regime, whether due to an increased preference for leisure or to mild disability; and most commentators have agreed that if retirement can be delayed (or in this case, gradualized) the drop in workforce participation associated with ageing can be mitigated.”

    I agree with this too, this is one of the structural reforms we need, but remember the productivity issue.

    Think of some famous footballer, lets say Luis Figo. When he is 20 he is still learning, when he is 25 he may be at his height, and when he is 35 he is probably not capable of playing at Champions League level. He has a lifelong earnings curve which reflects this. In the case of a footballer this is relatively uncontroversial, but everyone faces a productive effectiveness curve like this.

    If we take earnings data as a proxy, then our peak ‘value’ is normally somewhere in the 50-55 range, after that we are, on average, less productive. So what I am saying is that the cover of young people by elderly workers isn’t perfect, especially in the new technology industries.

    “The US is currently at 2.1 children, which is the replacement rate.”

    zman, this point is valid, but your apparent glee is extraordinarily misplaced. Just look at the US current account deficits. What is behind these?

    Well, the latest cutting edge research suggests that these global imbalances that everybody is going on about are driven by what could be called ‘unequal demographic shocks’. Basically different countries are experiencing the democratic transition in different ways, and this is effecting comarative savings and consumption rates, growth rates etc. Essentially the US is short of customers to pay for all the imports. This isn’t funny for anyone, and basically the US will be where Germany is now by roughly the early 2020s, and from then on will have all the same problems. Presumably you aren’t counting on getting your pension till after 2020?

  14. No one here has spelled out why this is a “problem” per se. It seems to just be assumed. It is still not clear to me why an aging, shrinking population is _by definition_ bad. After all, a more sedentary, smaller population is much more likely to use less energy and resources than a younger, larger population.

    And, frankly, there is an easy fix to this. Just have more children. If people are really convinced this is a “problem”, it could be substantially corrected in a decade, merely by doubling or tripling the birth rate.

    Increasing pension liability is not a convincing argument, after all, we could just reduce pensions! I have no idea what you mean by “global imbalances” by the way, Edward.

    Dave, I have no idea why you pick out Mexico in particular. Is the Mexican population decreasing? I don’t think so. You seriously expect me to worry about something that _might_ happen in China in 2050?

    Give it to me straight. Why should I worry about this as opposed to global warming, terrorism, the spread of nuclear weapons, virulent pandemics, the shocking lack of morality in the young, the decreasing quality of Hollywood movies, etc?

    Also, explain why I should worry about forecasts 50 years in the future when essentially every demographic prediction made fifty years ago was wrong.

    Why should I care?

  15. Takenaka’s conclusion that falling population ?means if we don?t create a system in which the private sector can carry more responsibility, the burden on taxpayers and on the state will become unsustainable.? must be taken as a polite way of saying current benefits to the aged are excessive, unless he is simply obtuse.

    At a steady population, a pay-as-you-go benefits system simply requires each worker to fully cover the cost of his (and his non-working brethren’s) benefits. Should this cost truly be unsustainable, Japanese society would then need to lower standards of living for the aged.

    His warning to shift responsibility for benefits from the public sector to the private just seeks to spare government the (obviously extreme) political difficulty of stiffing the elderly down the road. To the extent citizens assume individual responsibility for setting and saving for their post-retirement standards of living, they’ll be forced to accept greater frugality, either (or likely both) in their working or retirement lives.

  16. Ok Hektor if you need one more argument on the micro level here’s a social one: “Social atomization”.

    It doesn’t seem to be mentioned much but it does strike a personal chord for me. It was one of the predictions made in conjunction to dropping fertility levels by Eberstadt back in 1997. Lutz also makes a similar mention of this in his presentations. See one of Edward’s earlier posts about this (“the fertility trap”)

    And seeing as fertility levels in Italy have not improved much since then it seems Eberstadt’s prediction is now a lot more accurate then back in 1997.

    http://www.junkscience.com/news/eberstad.html

    —-
    Finally, it is interesting to ponder how the demographic revolution to come will affect the family as most of us have experienced it. The U.N.’s projections imagine a world in which the only biological relatives for many people will be their ancestors.

    Consider the possibilities for Italy, currently the country with the world’s lowest fertility level. If Italy’s current regimen is extended for two generations, almost three-fifths of the nation’s children will have no siblings, cousins, aunts or uncles; they will have only parents, grandparents, and perhaps great-grandparents. Under those same assumptions, less than 5% of such a future Italy’s children would have both siblings and cousins.

    Social Atomization

    Projecting the fertility rates for the entire European Union forward two generations only slightly alters the Italian scenario. In time, under “low variant” assumptions, families in the less developed regions would move in this direction. Within another generation or two, a family consisting of siblings, cousins, uncles and aunts would be anomalous throughout the entire world. For many if not most people, “family” would be understood as a unit that does not include any biological contemporaries. The nuclear family may have marked a radical departure from previous sorts of family arrangements. But, as we have seen, the nuclear family does not begin to approach the limits of social atomization that may await us in a depopulating world.

    All this represents merely a sketch of a future whose social, political and economic outlines promise to break sharply with anything in recorded experience. Yet as opaque as these changes may appear today, we may yet manage to discern them very carefully. A good number of us could eventually experience them firsthand: In the U.N.’s “low variant” projections, in fact, half of the world’s current inhabitants would still be alive at the time that global depopulation commences.

    So remember that, in your own future your (great-)grandchildren may not even have uncles nor aunts and, naturally, no nephews and nieces either.


  17. Dave, I have no idea why you pick out Mexico in particular. Is the Mexican population decreasing? I don’t think so.”

    I can’t resist, but yes. Dave indeed has a point with Mexico. There too the TFR has been dropping dramatically, from 5.7 in 1976, too 4.4 in 1981 and in the last few years from 2.4 in 2000 to 2.2 in 2004 (that’s actually just below the UN medium variant btw). By 2015, or even faster if this rate keeps up, it will fall below 2. It wouldn’t surprise me if the Mexican population start to fall if current Mexico->USA emmigration levels keep up.

    http://www.inegi.gob.mx/est/contenidos/espanol/rutinas/ept.asp?t=mpob16&c=3193

  18. Hektor, the reason that I picked Mexico is that you had characterized the increasing seniorization as an issue of affluent democracies. Mexico is a democracy and although middle-income generally not characterized as “affluent”. I also happened to know the projected dependency ratios for Mexico off the top of my head.

  19. Sorry, but Mexico’s current fertility is _above_ the US’s, and its population is still growing. I wasn’t talking about projections, because they _aren’t_ reliable and never have been for population growth. You lumped this is as a problem in Mexico. But this is a “problem” for Mexico in the future, not right now. Just because all you have is lousy data, don’t assume the data has to be correct and build crazy models from it.

    Sorry, but so people don’t have siblings and don’t have uncles and aunts. So? I fail to see how this is a world-wide calamity that will bring civilization to its knees. You’re assuming, again, that just stating something means it is automatically bad. Explain to me why governments need to move heaven and earth and fund many research scientists because people don’t have uncles and aunts anymore.

    All I ask is that someone please explain why population decline is bad. Just give me bullets here people. People on this thread obviously have a visceral reaction to this idea. Is it tied in with national greatness, vitality of the Volk, what?

  20. “Recently the Netherlands had the first month (since reports are available) that the population actually decreased.”

    Oh, I didn’t actually know this, but this is the point (David) I was trying to avoid by initially talking about major states. The trend is so general that it is hard to keep track of exactly where is where. In the Netherlands they are still a little younger though – median age 39.04 – so I think it will still be (guessing) a bit more than five yearsbefore you get to the Germany, Japan, Italy stage. Another of the smaller EU countries, Finland with median age 40.97 seems next in line.

    I’m also watching Finland closely since ex-oil and food (the volatile elements), inflation has been hovering round the 0% danger line for some time now, so deflation – another of the ‘ageing’ problems’ – could become a real issue there during the next recession.

    @ David

    Waking up this morning I realised the big reason I wasn’t thinking about Russia when I wrote this (as I have said it was a blind spot, we do have them, and not only in the way Nelson did:) ) and this is the issue of ageing. The median age is 38.15, and the male median age is a staggering 34.99 (ie more or less where China is now!!!). The stylised facts concerning the second stage of the demographic transition are sustained below replacement fertility rates *and* increased life expectancy. In this sense what is happening in Russia is a-typical, a product of a very specific history, a huge tragedy, but not directly relevant for understanding Japan, Germany, Italy, Spain, Greece, and what may be about to happen (as Dave is flagging with China) across much of the developing world. Eastern and Central Europe is ‘path specific’ to use some jargon.

    I am much more interested in, say, Thailand, where TFR is now significantly below replacement (1.7/1.8) since Thailand is still a developing country. Big Indian states (and among the most successful ones economically) – Tamil Nadu, Kerala – are also below replacement. If this trend continues – and there is every evidence that it will – it can present all of us with a huge problem in the not too distant future.

    Dave’s points about China are not to be taken lightly. One view of how to ease the problem in the OECD countries is to export pension savings to gain income from third world development, but the China situation is pre-occupying since it is both ‘young’ and ‘old’. It has both the savings rich properties of a ‘young’ adult society, and the savings heavy proporties of an ageing one where there is no substantial accumulated wealth to live off, so Chinese savings are paying part of the US government deficit.

  21. “After all, a more sedentary, smaller population is much more likely to use less energy and resources than a younger, larger population.”

    This is of course an interesting question, and its one I’ve been giving a lot of thought to of late. I would put this another way, what’s wrong with a zero economic growth society? In principle nothing. Many of the earlier economists – even Keynes – assumed that we would get to a point where living standards were so high that we wouldn’t need to keep growing and growing. Reality, however, has pushed this view to one side. We are hooked on growth. Can we get off it? Well declining birth rates and rising life expectancy means we will be pushed off it, whether we like it or not, the real issues are the transitional dynamics from one regime (the growth one) to another, slow steady contraction.

    You see economists have often noted that for 10,000 years prior to the industrial revolution the world changed very slowly, almost imperceptibly, as population grew as a function of slow steady changes in agricultural productivity: this was the constraint, our ability to feed people.

    In many ways the spaceship earth idea is not so bad as the press it receives suggests. The earth supported the population which it was able to feed. Then came the industrial revolution, which is in part a mis-nomer since really it was, has been, and still is an agricultural revolution. The earth turned out to be able to support a lot larger population that people imagined. This was partly, as the Danish economist Esther Boserup pointed out, because the pressing problem of feeding people stimulated the creative activity to make it possible.

    We are now going up to a ceiling population of (give or take) 9 billion people sometime in the last quarter of this century. This will be a peak, and then population will steadily fall. By the time the global population peak is reached, OECD populations will already have been falling for some time. This is a bit like the not unconnected problem of peak oil, where a local peak was reached, I think, in the US in the late 70s. So the constraint for economic activity will by then have ceased to be ability to feed, and will have become (collectively) ability (desire) to (re-)produce people.

    Now, the big issue. Is there a homeostatic, steady state, happy ending to all this? (I am talking in economic terms).

    Well, there might be. First we need to briefly touch on the principal sources of economic growth. Crudely put these are: (1)increasing population of working age, (2) increasing per capita participation in the cash economy (this is developmental, age structure related, and gender related), and (3) increasing the economic value added of each unit of labour input (or productivity, raising the value of the flow of labour and capital services).

    Now, with a birth rate of just 2.1, and no increase in life expectancy, you would have a stable population with a stable age structure. Assuming the society has already moved from agriculture to industry to services, and that the gender revolution has taken place, there would be no growth from (1) or (2), but there would be a slow and steady trickle of growth from (3)(in todays terms about one third of all growth comes from this source).

    Now, the assumption of no increase in longevity is unrealistic. So what about an increase in longevity which just covers the trickle in productivity. The two components could balance in such a way that the net result was zero growth. This would then be a global steady state, of ‘n’ open economies, all with similar age structures, standards of living etc, and a relatively stable equilibrium. (This assumes we can get 2.1 TFR, but there are undoubtedly going to be a variety of ‘techno-fixes’ on offer by the time we get there, not necessarily ‘human all too human’ but still, it is doable).

    But hang on a minute. Right now we are far from that equilibrium, and it is the right now situation that really interests us.

    Yes, but hadn’t we previously thought of capitalism as a society which ‘guaranteed’ timelessly high levels of growth. This was thought to be one of its properties. Now we can see that the high growth regime is probably a transitional one between two much more stable ‘local equilibria’. Most of the juice which has been squeezed out of the lemon in terms of growth has been connected with the transitional – between steady states – population dynamics. This has had a huge upside for the richest rich countries, and there is every reason to imagine that it will now have a significant downside as the planet ‘rebalances’ and we enter the new equilibrium. That Hektor, is what you should be worried about. I don’t know whether you currently have a mortgage, or whether you plan to have a pension, but both of these will be affected.

  22. @ Hektor

    “What precisely should we do about it? What’s the point in posting this stuff if you don’t talk about what you are really interested in? That seems to be: increasing immigration to Europe, admittance of Turkey to the EU, etc.”

    I will take your argument on a good will basis Hektor, but frankly I think that as an autonomous adult *I* have the right to decide what I am really interested in.

    “……we really don’t know what will happen with fertility rates in Japan, or really, anywhere else in the future. They are notoriously unreliable, and predictions are almost always wrong.”

    “I wasn’t talking about projections, because they _aren’t_ reliable and never have been for population growth.”

    “Also, explain why I should worry about forecasts 50 years in the future when essentially every demographic prediction made fifty years ago was wrong.”

    “Just because all you have is lousy data”

    “Aging of the population can be judged somewhat like the weather. We can change it, but only in unpredictable ways which we really don’t understand.”

    I think all of this is at best a cannard, and extraordinarliy misleading. It also happens to be very insulting to some hard working and dedicated demographers, and I – on their behalf, I’m en economist not a demographer – resent the insinuation.

    In the first place you assume demography has made no progress in the last 50 years. This is just nonsense. No-one would suggest that geology was left unchanged by the acceptance of plate techtonics, well, for eg, Bongaarts identification of the tempo and quantum effects is every bit as important, and there is plenty plenty more. Your view is profoundly unscientific (hasta la medula, as we would say in Spanish, and of course speaking more Spanish is just one of the impacts all this is having in the US) and verges on being akin to creationist objections to evolutionary theory. Just because we need to be modest in our claims (which I think is a virtue, not a vice) doesn’t mean there is nothing to say.

    Demography is not weather forecasting, and here are some reasons why not:

    Firstly: we already do know quite a lot of about the demographics of our societies 50 years from now (much more than we do about the global climate) since we know more or less the form of over half the population (the over fifties) since they are already born.(Barring, of course the impacts of wars, natural disasters or pandemics).

    You are right that the longer the term in which we want to assess these issues the higher the levels of uncertainty there are, but the issues themselves are by no means on such insecure foundations as you suggest.

    “And, frankly, there is an easy fix to this. Just have more children. If people are really convinced this is a “problem”, it could be substantially corrected in a decade, merely by doubling or tripling the birth rate.”

    This is incredible. The second stage of the demographic transition has been ongoing for the best part of a century now, it is universal, and there are no significant reversals to be found, and you think we can turn things round in a decade. Just who is proposing ‘crazy’ models here?

    Even if there were ways to influence the birth rate as you suggest, this wouldn’t make any real positive economic impact for at least 30 years. Since all the issues that are currently exercising my mind on the practical (as opposed to the abstract theoretical) level – see last comment – are going to happen in the next ten to 15 years, they are, as far as I am concerned, a ‘done deal’.

    Strangely you seem to ossicilate between suggesting that the issue is so complex that governments can do nothing, to suggesting they can ‘correct the problem in a decade’.

    The thing is there is a difference between general climate change modelling and weather forecasting. On the demographic equivalent of climate modelling I think we have a lot firmer grip on what is happening, and vis-a-vis weather forcesating (which isn’t very solid beyond 3 days) demographic impacts are pretty clear for at least a couple of decades going forward.

    “Just once, I’d like someone to be brutally specific when they say “Europe will be in dire straits”. What does “dire straits” actually mean?”

    Well, I didn’t use the term (especially since this is global, and not local) but I think if you read my posts over time you can get an impression of the seriousness with which I think we should be taking this.

    “Increasing pension liability is not a convincing argument, after all, we could just reduce pensions!”

    Well, this is happening, and of course there will be a lot more, but one of the worries is that if the reforms don’t follow fast enough the pace of the changing demography, the whole pension system might go bust (think Italy, but there will be more), so I don’t think anyone should be taking any of this lightly.

    “Give it to me straight. Why should I worry about this as opposed to global warming, terrorism, the spread of nuclear weapons, virulent pandemics”

    I don’t buy the ‘as opposed to’ argument. The other topics you raise are all important, you don’t need to choose, just worry more.

    “I have no idea what you mean by “global imbalances” by the way, Edward.”

    Well this may be very much to the point, since it seems to imply you are not entirely up to date on where the argument is. You need to read the speeches of eg Alan Greenspan, or the reports of the IMF or the OECD, or I would have thought serious press like the New York Times, or economics blogs.

    The global economy is suffering from serious imbalances. Previously it was thought that some of these could be remedied by market oriented reforms in some economies (these are not invalidated by what I am, saying, but they don’t fix the problem). Now it is becoming clearer that ‘asymetric fertility’ impacts, and the consequences for age structure (yes, that is another plate-techtonics like advance in economic theory, we now do know that age structure matters) may well have an important explanatory role to play. If you want an example of recent work, I am recommending this paper:

    http://www.sensiblepolicy.com/download/esrimckibbin3.pdf

    It is fairly technical, but if you read the intro, the conclusions, and some of the blurb in the middle, you should be able to pick up the gist of the argument. There is plenty, plenty more.

    “People on this thread obviously have a visceral reaction to this idea”

    Forgive me Hektor, but my impression is that the one with the ‘visceral’ reaction here is you. I’m a pretty cerebral type myself :).

  23. @ mkl

    “must be taken as a polite way of saying current benefits to the aged are excessive, unless he is simply obtuse.”

    I think you got the message. Excessive is of course a relative term, relative to ability to pay. The reductions, obviously, may well cause hardship, but really Japan now has little alternative.

  24. I think you got the message. Excessive is of course a relative term, relative to ability to pay. The reductions, obviously, may well cause hardship, but really Japan now has little alternative.

    Japan, having undergone modernisation faster than the West, does have traditional family structure to fall back on, to a point. A significant minority of Japanese elders do live in trigenerational households.

    From an rational-choice point of view, though, elder care does make more sense as a state social service than, say, child care. Unlike childhood, the length of elderhood is quite variable, and the medical expenses one will incur as one approaches the end of life are also quite variable. So while the couple who decides to have a child can know what they’re getting into, the family that decides to support an elder is taking a much larger risk, hence the role of an insurance program.

  25. First of all, Edward, I don’t appreciate being lumped in with evolution deniers. Since I happen to be a scientist with a Ph.D. in Physics, I know a fair amount about what is and isn’t science. I also don’t appreciate being lectured on science by a non-scientist (economist). So why don’t you get off your high horse, and stop cautioning me that papers are “technical”. This has nothing to do with creationism, and frankly I have no idea why you are constructing such a strawman.

    I doesn’t surprise me that you aren’t a demographer, so I don’t see why you have to defend them. Most demographers I know are quick to state that most projections of the past were unreliable, and that while there has been some improvement in the past decades, many long-term forecasts are still iffy. I also don’t understand why you are insulted, since you aren’t a demographer. Demographic projections are a hard problem, and they aren’t fully solved even now. For example, there is a lot of disagreement on Russia’s future population, even among demographers, though I do believe some progress is being made.

    The main thing I am reacting against here is the “sky is falling” attitude I see from both your posts and those of the many commentators on this subject. Pension issues and having no aunts and uncles do not seem like world-shattering issues to me. People seem to be discussing these issues as if they are governed by iron laws of mathematics – like your dictum that once fertility falls below a certain level, it never again will increase past it. This is ludicrous for discussion of autonomous agents – people can spontaneously decide to have more children, and various political, social, and cultural trends can profoundly affect these trends. What I get from you is a strongly mechanistic view based on simple models, which frankly is a frequent vice for both economists and physicists and a temptation to be avoided. You can’t just extrapolate predictions to infinity and feel like you have solved the problem.

    “dire straits”: Edward, you still resist listing what precisely you think the bad conditions will be. The best people have come up with are pension problems and social atomization. Neither one of these seem to be world-shattering problems. Why can’t you summarize for me what you think the problems will be? Here, to get the ball rolling, I will list various developments I have heard of and group them as roughly positive or negative:

    Positive:
    decreased population -> less consumption of resources
    decreased activity of population -> less greenhouse gases, less wasteful commuting and creation of exurbs (though more leisure time may lead to more tourism/travel)
    decreased youthful population -> less warfare, less bloodshed
    fewer children -> fewer government resources on non-productive members of society (a significant factor in the growth of the East Asian Tiger economies)

    Negative:
    increased sex-linked infanticide -> more aggressive countries (though this may be self-ameliorating)
    increasing health and pension costs (dependency ratios decreasing)
    social atomization
    increasing conservatism – resistance to structural changes
    elderly hoarding government programs for themselves

    Those are only the most obvious. Do you agree or disagree with this list? Do you have more to add?

    You obviously think there

    HB: “People on this thread obviously have a visceral reaction to this idea”

    E: “Forgive me Hektor, but my impression is that the one with the ‘visceral’ reaction here is you. I’m a pretty cerebral type myself :).”

    No, I don’t actually think you are. I’ve tried to get you to give me real predictions about what you think is going to go wrong, and you so far refuse to do so. You reacted negatively to zman’s trolling by trying to score points off the American economy. It should be beneath you.

    What I have a visceral reaction to is wild-eyed predictions about looming apocalypse made without supporting data or assertions. Why don’t you make some specific predictions about what you think is going to happen? Are you just a bystander?

    Now, your point about moving between two equilibria is interesting, but actually, I don’t worry about the long-term economic health of the US (where I live) that much. We have a slowly growing population, largely from immigration. As long as we can get back to reasonable economic stewardship, we should be all right. I also live in an area of the country, Massachusetts, where there isn’t that much available housing stock, so I don’t worry about my mortgage, especially since all the trends you are talking about are extremely long-range, 50 to 100 years, at which point I will be dead or on anti-agathic drugs anyway. Also, since I am an American, we don’t get pensions here anyway.

    So I looked at the paper you mentioned, which has a fairly simple model for population change, including a common mortality rate – a big warning sign. It also requires all countries have the same tfr in the long term – also worrying. (Just a comment – why do people put all their figures at the end – it always makes it harder to read papers. I hate page-flipping.) It is interesting, though the summary is relatively poorly written – you really have to look at the figures. The basic point seems to be that demographic changes in Japan and other countries interact, causing significant damage to japanese growth and interest rates in the far future – 50 years hence. There isn’t any discussion about how to ameliorate this, though apparently work is ongoing. So, assuming I buy this model, which only seems to have bad effects 20 to 40 years from now.

    Finally, the reason I mention Turkey is that you have previously said that you favor Turkey joining to ameliorate the dependency ratios in Western Europe. That means that you do in fact have policy prescriptions based on what you think is going to happen. So I don’t understand your reticence to mention them, when you clearly have said them in the past.

  26. Very interesting discussion.
    I tend to go with Hektor a bit.
    Tuesday I crossposted an article reacting on Hamiltons (econbrowser) criticism of the Hirsch report on Peak Oil at this other European blog
    My main point of agreement with Hamilton is that he is right, in my opinion, that Peak Oil in itself is not a clearly defined problem.
    Peaking of Oil is a complicating factor in the approach of real problems.

    Something similar goes for the discussion on the demographic transition. Like Peak Oil it influences lots and lots of issues in a complicated manner.

  27. @ Hektor

    OK, thanks for a lengthy answer, and if I’ve taken a bit of time getting back it was because I want to be able to give my response the time and seriousness your comment deserves.

    Frans says:

    “Like Peak Oil it influences lots and lots of issues in a complicated manner.”

    I think if the two of us could just agree on this we could really start going somewhere. The point is we don’t have to agree on how the demographic changes are going to impact, this I think is the open zone of debate.

    This brings me to things like Turkey, immigration, earlier childbirth, more child support, longer working lives, more flexible labour markets etc, etc. All these are things which might help, but there is no clear answer in advance, and maybe we can change our minds as we go. This is what debate is about.

    On Turkey specifically. I am in favour of Turkey EU membership, and partly because they are a ‘young’ society, which I think will be a good balance in an ‘older’ EU. But this is not the only reason. I think having Turkey in the EU will send a message out to the rest of the world about what contemporary European values really are, and I welcome that message. I am aware that others do not, and I think this is a legitimate area of debate. A debate we will have here on an ongoing basis as the negotiations progress, I guess the first frontier will be around the October EU summit. I welcome your input into that discussion, but I think we should have it then, and focus on the specific issue of the impact of demographic changes in developed economies, since I think that impact alone is big enough to warrant attention.

    John Maynard Keynes changed his mind on several times between the ending of WWI and the writing of the general theory, these days I do not consider myself a Keynesian in any significant sense of the word, but I do see myself in the spirit of Keynes. I understand why he felt himself obliged to keep changing his mind, when you face new problems, ones not faced before in the same way, there are no ready made theories and policy tools to hand, you have to fashion them.

    Now I see you don’t regard economics as a science, and you are entitled to your view, I do consider it as a branch of knowledge with ambitions to become a science, but I recognise that – as a social science – it has this differing characteristic – that in seeking to understand its object it inevitably changes it. I don’t claim to any expertise in Heisenberg and quantum mechanics, but I suspect that it shares something in common with physics here: in seeking to measure it changes the configuration.

    BTW I unreservedly withdraw the analogy between you and the creationists. I still think that the argument that what people thought 50 years ago about today means we must be equally off target about what may happen 50 years from now doesn’t allow for scientific evolution, but again we will just have to agree to differ. I have a lot more confidence in the consensus opinions of demographers than you probably do. How to measure the consensus, well Wolfgang Lutz et al have evolved a methodology called Conditional Probabilistic Population Forecasting. You can find the abstract and link to one work here:

    http://www.iiasa.ac.at/cgi-bin/pubsrch?IR03052

    This is peer reviewed, published in Nature etc. etc.

    Fuller details of their work, including some simulations they are working on on climate and population trends (which may interest you) can be found here:

    http://www.iiasa.ac.at/Research/POP/proj01/reading.html

    But regardless of whether you are willing to grant these researchers any possibility of advance, I think there is an easier way of looking at this. Einstein recommended keeping things simply so lets for once follow his advice.

    One piece of data we do know with some degree of accuracy is how many children are born each year. We also have more or less accurate figures on morbidity, and a bit fuzzier info on in and out migration. Well using these numbers we can construct a pyramid. I think the simplicity and profundity of Lutz’s insight relates to this. It doesn’t matter what the statistical constructs /TFRs, crude birth rates etc) say. This data is relatively solid. So constructing a simple pyramid is not too hard. But this pyramid then really contains a hell of a lot of condensed information which we can then apply, if only we understand how.

    I could give you an example, the current ‘weak’ US labour market, well its might bet that if we economists could find a way of plugging this data into the standard economic models we might learn a lot about what is happening in the US right now.

    So:

    “You reacted negatively to zman’s trolling by trying to score points off the American economy”.

    This isn’t really fair (although we are all only human) since if you go to Dave Altigs MacroBlog and look at this post and comments:

    http://macroblog.typepad.com/macroblog/2005/08/more_on_labor_m.html

    you’ll see from my comment and Dave’s response that I’ve been discussing this particular issue well before zman came around.

    I am also trying to get across to people in the US that they shouldn’t adopt an I’m alright Jack attitude, basically because you aren’t (alright I mean). The fact that the US is a society built from the ground up (with the honourable exception of the native Americans) gives your society (and Australia, and New Zealand, and Canada) certain distincitive features that the other societies don’t have, this gives your fertility a different ‘look’, but in the longer term you are in the same boat as everyone else.

    Additionally in a globalised world you are affected by what affects the rest. This is really why I mention the global imbalances.

    But back to the pyramid for a moment. Once we have it we can get some fairly precise information going forward. For example, if the cohort between 25 and 35 years of age is twice the cohort between 5 and 15 years, we can know in advance that whatever happens to the TFR for the 5 – 15 group they we be having children on the basis of a 50% female group. This is how population reduction works. So we already know roughly how many children may be born in a given society between 2025 and 2035 with a fair degree of accuracy, allowing for moderate fluctuations in TFRs. We also know how many people (as a minimum baseline) we will have over 50 in 2050, since they are all already born etc etc. All I am trying to suggest is that we do have something to go on, we are not ‘blind guessing’.

    Well that’s the intro. I will now go on to try and address the substantial question

  28. @ Hektor

    Now, the really important point you are raising is that people should be made to justify what they say, and not just engage in far-flung speculation. So what will be the consequences of the panning out of the second part of the demographic transition?

    Well again, forgive me if I hedge, but I do think we need to be prudent. I am happy to admit we don’t know. We can make a number of scenarios. Some of them will be best best best case, and others will be worst worst worst case, and in the end we will probably get either best worst best, or worst, best, worst.

    Lets take something recent which was equally unknowable in advance MadCow disease in the UK. We knew a lot of cows had been infected, but we really had no knowledge of what the epidemiological impact on the human population of nvCJD was going to be (I am talking about the UK here). It was possible to make very differening forcasts. On the basis of some of the more pessimistic of these (which were remember perfectly rational) the NHS adopted all sorts of procedures relating to surgical instruments and tonsil operations, and things like that. In the end (although we still don’t have the full picture and won’t till those who were big adolescent burger eaters at the time reach old age) it seems the worst fears weren’t justified. So should we have taken so many costly precautions? I think, on the basis of what we knew then, we should have.

    Well, I’m going to argue something similar about demographic changes. We don’t know ex-ante what we are in for. But prudence counsels giving a stronger weighting to the downside risks than to the upside ones.

    Well, what could happen.

    I think the first issues is in the third world. Dave’s points about China – where I readily agree the levels of confidence we can place on the projections are fairly low – are very much to the point. Will China get rich before it gets old, or old before it gets rich?

    http://www.csis.org/gai/GrayingKingdom.pdf

    If it is the latter, and given the size of China and its specific weight in the global economy, the consequences will propagate themselves out throught the global economy. What does this mean?

    Well, we know that the property of a young adult society (more or less China now) is that it saves a disproportionately high quantity of its income (I mean really this is Keynes old propensity to consume argument, shifted from class to age). We also know that a old adult society tends to save a disproportionately high quantity of its income (Germany and Japan), so the big danger is that these two processes conflate, and China never passes through the high-credit consumer driven mature adult stage (where the US is now, or Ireland, or, more or less the UK and France). This would then pose an enormous stress on globalisation, since China would have large excess capacity on many levels, and could flood the world market with products over everytype (Airbuses, why not??). Under this pressure it is quite possible that globalisation would break down. Some people would undoubtedly cheer, but again I personally think the worst affected would be the developing countries, who would be able to gain access to the more developed markets, or the necessary inflows of FDI and expertise.

    So China is one whole issue by itself. As I said it is hard to say exactly what is happening in China since we don’t have reliable population data, and you (Hektor) may find this paper outline interesting:

    http://iussp2005.princeton.edu/download.aspx?submissionId=51192

    Then there is Southern Europe and the Asian Tigers. Why are they interesting? Well they are the most recently developed societies economically and they also have had some of the most dramatic declines in TFRs. It is therefore pretty important to establish whether this is an unfortunate coincidence or whether there are real reasons to think that the faster the drop in TFRs to below replacement, the bigger the structural damage to the population, given that this drop also goes with fairly rapid economic development (gender equality at work, consumerism, home ownership) which can have a positive feedback re-inforcing effect on fertility declines.

    I suspect that this latter is what is happening, but we need better documentation, and clearer evidence.

    The pace of drop of TFR in Thailand (which is now 1.7 according to the recent prb report) may be an early warning indicator here. If so many relatively poor societies move to old adult societies without the wealth accumulation this will be a big global tragedy as far as I am concerned. The extent of this tragedy is one area of research.

    Then we could move to the OECD world. What are the possible consequences? Well first you need to ask, with or without reform. Why is this important? Because the reforms will make the difference between whether or not state finances are sustainable.

    Take one example, Italy, where the government debt is currently around 106% of GDP and set to rise towards 110% over the next two years. Without substantial reform Italy will go bust. Italy may reform. All well and good.

    But this is a ‘garden of the forking paths problem’ because if they don’t then the only thing you can say for sure is that there will be a sovereign debt default.

    So even if it isn’t Italy, someone somewhere will default. Not too dramatic still I imagine Hektor, but still distressing, having places becoming like Argentina. And when the first one defaults the pressure will rise on the others.

    Why are defaults more or less inevitable? Well, basically because governments have relied on PAYGO systems where each generation pays for the previous one, so if the generations get smaller, each individual pays more, and that, as Robert was indicating, either encourages either young people or companies (or both) to relocate, so the problem gets worse and worse, in a circular fashion.

    Then there is housing. You see constantly rising house prices rely on a certain underlying demographics, that you have a constant pressure of larger generations pushing against a slightly insufficient housing stock. But if the situation becomes permanently the other way round, that there are always more houses to sell than there are newcomers to buy, well the tonic then will be gradually and continuously declining house prices (in real terms) and in fact we’ve been seeing this in Japan for some time, and we have now been seeing it in Germany for the last few years.

    Well, it is important to remember that housing is the largest source of private accululated wealth, so if the numbers start dropping this also has consequences.

    I think it is important to realise here Hektor that the paper on Japan I suggested was really to give you an indication of the sort of issues which people are talking about, that there will be globally transmitted savings and investment consequences, and these will have general balance of payments consequences etc. But this is simply what I say, an orientative illustration. We still need to calibrate.

    You will notice I have not mentioned anything about 50 years from now, I am talking about what is already discernable in Germany and Japan. So what I am arguing is that if we can already see some of the problems, and identify some relative orders of magnitude, then the problems 50 years from now may well be considerably more important than is presently being contemplated.

    You mention good stewardship, well this is going to be much more difficult, since fiscal policy is going to become much more difficult to apply in an atmosphere of needing to continuously reduce government indebtedness, and monetary policy seems – looking at Germany and Japan – to become less and less effective as interst rates drop towards the zero bound.

    So if you are looking for consequences, this would be another: a blunting of traditional policy instruments.

    Then there is the reform process and democracy. I don’t want to start putting out scare stories, but I am convinced that as ageing makes it more and more difficult to get voters behind things like cuts in pensions and medical services (which are one of the consequences of the pyramid break) since more and more of them are directly affected, we may see the rise of more dangerous demagogic authoritarian movements. Clearly I hope I am wrong, but I don’t think we can entirely rule this out.

    Well, I’m sure this is not an exhaustive list, but these are at least some of the reasons why I think we should be worried about too rapid a transition to a greying elderly society. Coming off growth may not be such a bad thing, but if we do it in a way which provokes a general financial crisis I think we will regret it.

  29. @ Hektor and anyone else with the perseverence to have got this far, and forgive me this is a kind of technical synopsis:

    On the theory side I could give a little more detail. What I am trying to do is bring together the confluence of four different traditions:

    1/. Firstly the idea, currently essentially emanating from Bloom, Canning and Sevilla that age structure matters:

    see this book:

    http://www.rand.org/publications/MR/MR1274/

    or this:

    http://www.globalhealth.harvard.edu/Files/Working%20Paper%20No.%201.pdf

    Really this line of work goes back to Ansley Coale and Edgar Hoover and an important input has come from Jeffrey Williamson.

    2/. Then there is the “asymmetric shocks” work of Ralph Bryant and Warwick J McKibbin plus a host of IMF staff members and people at Brookings. As I have mentioned this one from McKibbin on Japan is a good start:

    http://www.sensiblepolicy.com/download/esrimckibbin3.pdf

    and this whole conference is just full of material:

    http://www.esri.go.jp/en/prj-2004_2005/macro/macro16/syousai-e.html

    My first bonobo menagerie post may explain a little more:

    http://bonoboathome.blogspot.com/2005/08/welcome-to-menagerie-i.html

    3. Then there is the “From Malthus to the Demographic Transition” growth-theory crowd. The best starting point may be Oded Galor in this paper (he tries to develop a unified growth theory):

    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=562085

    or Charles Jones “Ideas & Growth” paper:

    http://elsa.berkeley.edu/%7Echad/papers.html#handbook

    which really goes back to some initial ideas from Michael Kremer:

    http://www.j-bradford-delong.net/Teaching_Folder/Econ_202c/readberk/1-Demography/Kremer_Million.pdf

    The important point is that this growth theory type approach, which has its origins in Solow, is very different to the general equilibrium stuff from Bryant et al.

    4/. Then there is the work of the demographers themselves.

    Bo Malmberg’s work is important here:

    http://www.framtidsstudier.se/aktuellt/2000.6.pdf

    as is Wolfgang Lutz’s, which I have addressed a little here:

    http://bonoboathome.blogspot.com/2005/08/welcome-to-menagerie-ii.html

    and here:

    http://fistfulofeuros.net/archives/001748.php

    What I am trying to do is find a way to bring all of this together. The starting point for my approach is that the demographic transition isn’t simply a shock, there is something more fundamental going on here, and in the light of this, I don’t think Galor’s unified theory is as unified as he would hope. Indeed what Galor identifies as the ‘growth society’ may in fact simply be (as I’ve indicated above) a transition from one steady state to another.