No Answers Only Questions

One person who could rightly claim to know more about global ageing and its possible consequences than anyone else in the business is the German Director of the Manheim Research Institute for the Economics of Ageing Axel B?rsch-Supan. If there’s a conference being organised, he seems to be there. Actually his comments at both these meet-ups are well worth reading in and of themselves (here, and here).

In a sense B?rsch-Supan is almost uniquely qualified to express opinions on the topic since he has both devoted a large part of his professional career to studying the question, and he lives and works in a society which is already reeling under the impact. As he says:

“Today?s Germany has essentially the demographic structure that the United States will reach in a quarter of a century. The dependency ratio (the ratio of persons aged 65 and over to those aged from 20 to 59) is at 28 percent, and it will reach 75 percent in 2075, if we dare project that far. Almost one-fifth of the German population today are aged 65 and over. One quarter are aged 60 and over, which is relevant because the average retirement age in Germany is 59.5 years. Thus, in this sense the United States is not ?entering largely uncharted territory,? …. Rather, they can look to Europe?in particular to Germany and Italy?to see what will happen in the United States.”

I mention B?rsch-Supan because he serves as a good pretext for going over where we are to date with the issue. As he says himself. watching demography change is rather like watching a glacier melt, on a day-to-day basis it’s hard to see that anything is happening, but over time the impact is important.

One of his recent papers has the intriguing title: “Global Ageing: Issues, Answers, More Questions“. It is a good up-to-date review of the ‘state of the art’, and a quick examination of the points he makes probably serves as a good starting point, since I can’t help thinking, in the case of global ageing, it isn’t so much what we know that matters, it’s what we don’t know.

So here we go, a review of what we “know”, what we think we know, and what we don’t know:

What we think we know:

* The expected change in the age structure in virtually all industrialized countries ? but also in many developing countries ? is dramatic and will lead to a substantially higher proportion of older people in the world.

* Global aging will affect labor, product and capital markets in fundamental ways which will change the growth path of GDP and the wealth of nations.

* The ageing process will deeply affect future labor, financial and commodity markets.

On a macroeconomic level:

* labor is becoming relatively scarce in the ageing countries while capital will become relatively more abundant.

On a microeconomic level,

* the age composition of the labor force will change and this will in all probability have an impact on labor productivity.

* consumption and savings patterns are likely to alter as the elderly become a larger proportion of consumers and savers.

* globalisation will mean that international flows of capital, goods, services, and labor will be among the most important mechanisms through which the ageing process will make itself felt.

* aging is global, there are marked international differences in the speed and the extent of the aging processes.

* dependency ratios are important, as are their variations. Not all societies are ageing at an equal pace. Germany?s old-age dependency ratio in 2003, for example, corresponds to that which the United States may expect in about 2023, ie in twenty years time.

* world population is still growing continuously, but the annual growth rate is slowing, from 2.04% 1965 to 1970, to the current 1.2% annually. It is expected that this decrease in the rate of world population growth will continue.

Three questions seem to stand out as fundamental:

* the structural changes occurring on the labor market,
* the age structure of workers,
* their labor productivity and wage structure.

Most developed countries have produced projections of labor supply, but there is no good compilation of them, and even less information on the exact assumptions, including the underlying projections of pension rules. Projections of the impact of global aging on the wealth of nations are therefore usually based on relatively coarse projections of labor supply.

Ageing changes the structure of consumption demand, and as a consequence the sectoral demand for labor. If the Continental European economies keep failing to permit quicker adjustments to changed circumstances, employment will fall quicker than labor supply due to unemployment in those sectors that are shrinking, and a lack of available new jobs in the potentially growing sectors.

Unemployment among less-qualified persons remains constant because the root cause of such unemployment (the large wedge between marginal productivity and total labor compensation) has not been removed

Higher labor force participation makes Japan less vulnerable to ageing than Europe. By the same token, policies affecting employment – such as an earlier labor market entry age via education reform, or a higher labor market exit age due to pension reform – are crucial weapons in our efforts to minimise the economic effects of population ageing.

Germany: despite the increase in participation rates predicted by the most probable scenarios, the size of the labor force as measured in 2000 is nonetheless set to fall by around 8 million in the long term. In other words, the labor market will contract ? in absolute terms by more than twice the current number of unemployed.

It is very unlikely that this decline can be compensated by an equiproportional increase in productivity and/or capital accumulation. Total German GDP will therefore almost surely decline.

The modal age of the German workforce will rise rapidly: in 2000 it is 36; 10 years later the peak age will increase to 46; and a further 10 years on it will have risen to 54 years.

This fundamental change in the age structure of the working population will have profound effects on the microeconomics and the sociology of the labor market. The most important ? and most controversial ? aspect is the potential effect on labour productivity. If labour productivity is age dependent, a shift in the age structure will also bring about a change in aggregate productivity, even if age-specific productivity were to remain constant.

What We Almost Certainly Don’t Know

Despite the fact that we have some idea of the relative magnitude and direction of the anticipated demographic changes, there are plenty of details which we still are unsure of.

* the future evolution of fertility. Research on the feedback mechanisms and the causality of demographic changes and the reasons for international differences in demographic processes has not got very far. We do not really understand why Italy has a lower fertility rate than France, and why the German fertility rate has stayed amazingly constant since the mid-1970s while all candidate variables that are commonly employed to explain fertility have changed. The link between the social security system and fertility remains controversial.

* The largest wild card in demographic projections is migration. The German statistical office has predicted a decline in the German population in each projection since the early 1980s. It has never occurred because migration was always larger than predicted. While we have many theoretical models describing the link between economic circumstances and migration, we do not have quantitative models that are able to predict how global aging will effect migration. Whilst we know that migration alone will not solve the problem, we do not know how chnages in other parameters, investment, domestic demand, relative growth etc, will affect migratory flows.

* the huge variability of the projections and the leverage of employment on GDP make clear that it is important to better understand how public policy can influence labor force participation rates. Structural reforms such as education reform and pension reform have potentially huge ?side-effects? on GDP growth through their impact on employment, over and above the often more prominently discussed impacts on social budgets.

* we still know relatively little about the extent to which an increasing quality of labor will compensate for the decreasing quantity of labor. ?Side effects? of structural reforms on productivity may in the long-run dominate any direct but static effects, because they change the growth path of the economy.

* there is a high degree of uncertainty in projections since they require general equilibrium models which are sufficiently realistic to model demography, capital accumulation and employment, something which is very difficult to achieve.

* it is likely that in the future more retirement income will come from asset income which tends to fluctuate more than annuitized pension and labor income. Such a development will tend to increase precautionary savings and to depress consumption, for any given a fixed level of income. To date we have little evidence to quantify these effects.

* pension reform towards a multipillar system with a substantial portion of funded retirement income will revive the retirement motive for saving in France, Germany and Italy. In fact, these systems will look very similar to the current Dutch system. Hence, it is likely that saving rates among the young will increase (to accumulate retirement savings), and saving rates among the elderly will decline sharply (because they will dissolve their retirement savings).

* the impact of pension reform (and other savings positive measures) on the rate of return on capital. Are current low rates of return a passing phenomenon, or will they become the norm. Will there be a ‘global savings glut’.

* uncertainty about the magnitude of international capital flows. L?hrmann (2002) has investigated whether demographic factors have influenced international capital flows in the past. She uses a broad panel of 141 countries that covers the period 1960-1997 to investigate the effects of demographics on international capital flows. She confirms that cross-country capital flows are indeed influenced by demographic variables. While this has been shown in other studies before, she can also show that relative differences in the age structure across countries are the most important determinants of capital flows.

13 thoughts on “No Answers Only Questions

  1. Good post Edward,

    I’ve read some of those articles at an earlier point . It will have some interesting longterm implications for the world regional capital in/outflows.

    P.S
    For the moment I would like to add one extra argument. Due to advances in medical science we’ll also be able to lead healthier and productive lives even at a higher age. That should be factored in as well.

  2. “Due to advances in medical science we’ll also be able to lead healthier and productive lives even at a higher age.”

    This is surely true. But then we might also live longer than we are now. All good news, but we need to think about the consequences. I think the last ten years of life are always going to be expensive for the support services, at whatever age they arrive.

  3. “The dependency ratio (the ratio of persons aged 65 and over to those aged from 20 to 59)”

    Does this mean that persons under 20 years are people of “independent means” (as they used to say in an earlier age ?

    An ageing population means that there are fewer very young people, who have to be fed, clothed, educated etc. ?

    How does this affect the forecasts, compared to purely looking at old people ?

  4. “The modal age of the German workforce will rise rapidly: in 2000 it is 36; 10 years later the peak age will increase to 46; and a further 10 years on it will have risen to 54 years.”

    For the layperson:

    What does “Modal” mean ?

    And why is modal age used in this example, rather than the more commonly used mean or median ?

    Is the modal age more relevant to this discussion than mean or median ?

  5. “Does this mean that persons under 20 years are people of “independent means” (as they used to say in an earlier age ?”

    No, sorry, this is a slip of the keyboard, it should have read *old age* dependency ratio or, if you like “support ratio” (a term which is sometimes used). Obviously looking at the overall picture of young and old dependency is interesting, but poses different issues. Basically from now to 2020, which is when I think a lot of the action is going to happen, there isn’t much to be done that can change things at the fertility end, since we have the children we have.

    What does “Modal” mean ?

    “The statistical mode is the number that occurs most frequently in a set of numbers. …”

    This is sometimes used in these studies as the modal age tells you something about the ‘boom generation’, and this moving through the age ranges has a disproportionately large impact. See the fancy graphs on page 16 (or 18 in the Acrobat reader) of the cited paper.

    “Is the modal age more relevant to this discussion than mean or median ?”

    They are all relevant, it just depends what you are looking at. In the US the modal age is important because of the earning capacity and accumulated wealth of the boom generation. We are not only talking about numbers of people, but also about economic impacts.

  6. “Basically from now to 2020, which is when I think a lot of the action is going to happen, there isn’t much to be done that can change things at the fertility end, since we have the children we have.”

    You mean that we *now* know how many children will be born between now and 2020 ?

    Depending on a lot of factors, birth rates can changing fairly quickly. Developments in East Germany after unification are an example.

    We probably know better how many old people there will be in coming years. They have all been born, and death rates are not that likely to change quickly.

    The Modal number alone is not too relevant, even if the rise of the modal age looks dramatic. What we also have to know, is how much the maximum rises above other age cohorts.

  7. Re Germany in the long term
    “In other words, the labor market will contract ? in absolute terms by more than twice the current number of unemployed.”

    Conversely, half the drop in the labor supply could be compensated by reducing unemployment.

    “It is very unlikely that this decline can be compensated by an equiproportional increase in productivity and/or capital accumulation. Total German GDP will therefore almost surely decline.”

    Could you explain how this statement comes about ?
    8 Million is roughly 10% of German population. Even assuming poor economic growth of 1% or so, how can the German economy not grow 10% in the long term ?

  8. “Higher labor force participation makes Japan less vulnerable to ageing than Europe. ”

    Naively, I would have thought it is the other way around.

    If in Japan the working-age population is 100% employed, and you get more dependents, the dependency ratio gets worse and you can do little about it.

    If in Europe there are a large number of unemployed and you have to take care of more dependents, you have a reserve pool that can be tapped.

    Could you explain ?

  9. “You mean that we *now* know how many children will be born between now and 2020 ?”

    No, silly, I mean we now know how many people (maximum) can be 15 in 2020, which is what matters. The rest is only going to *add* to costs in the short term. ie MAKE THING WORSE :).

    “Conversely, half the drop in the labor supply could be compensated by reducing unemployment.”

    This part depends on the labour market reforms being carried through and being effective. Do we have any odds on this in Germany?

    “It is very unlikely that this decline can be compensated by an equiproportional increase in productivity and/or capital accumulation. Total German GDP will therefore almost surely decline.””

    The German economy is on a growth path of a maximum of 1% a year right now. The calculation is an annual one. You get some benefit from increased productivity per head (if you can get that) and then you need to subtract the proportionate element for the reduction in the work force. These are B?rsch-Supan’s calculations. I personally think he is too optimistic.

    “how can the German economy not grow 10% in the long term ?”

    I think we’re saying it isn’t going to grow, it is going to shrink. B-S guesses about 1% a year, I guess more. These calculations will become more complicated if like Japan you hit deflation and we get negative price adjustments to calculate ‘real’ GDP. Don’t worry, in economic conceptual terms we are about to enter the menagerie.

  10. “The German economy is on a growth path of a maximum of 1% a year right now. The calculation is an annual one. You get some benefit from increased productivity per head (if you can get that) and then you need to subtract the proportionate element for the reduction in the work force. These are B?rsch-Supan’s calculations. I personally think he is too optimistic.”

    There’s still a lot of hidden overhead in many German organizations/industries. My take is that, once the babyboomer generation start to retire the overall production will stay at the same level. It won’t rise but it certainly won’t drop much either. Thanks due to extra increases in labour productivity, process&product innovation and some flexibilization on the labourmarket. (under the condition that Germany makes the right decisions in the next 10 years).

    I think there’s still a lot of productivity to be squeezed out of this German lemon! 😉

  11. “Thanks due to extra increases in labour productivity, process&product innovation and some flexibilization on the labourmarket.”

    This would be the big hope, this is obviously what policy has to aim for. This is also why the labour market reforms and flexibility are vital. I suppose I remain sceptical that this can be achieved whilst people are still in denial about the problem. I think strategies to move forward will become much more credible once we move on, and people take this on board. So long as politicians keep diverting the debate onto hedge -funds and the Chinese I remain a doubter.

    Basically Germany would need to do vaious things: put all those extra people to work (ie unemployment down around the 5% mark), get extra productivity (flexibility, positive attitudes to change, less rules and regulations, more start-ups), keep leveraging ousourcing just like they are, and cheer up (ie win the world cup, animal spirits etc). Even if they can’t get positive growth, they could reduce shrinkage to a minimum. The difference wouldn’t be that big, they won’t be the only ones in this situation, not by a long way.

    Now for Italy………..

  12. Renewal

    The science of slowing or regressing ageing is moving quickly. A scientific article appearing in the New England Journal of Medicine about two years ago, illustrated the effect of growth hormone on a group of elderly (60-69) men in the Veteran’s Administration Hospital (US). All experienced increased muscle mass and reduced adipose tissue.

    Another study by an Italian research group, published information showing the effects of a natural hormone (melatonin) extending the lifetime of certain mammals (rats) which experienced a 20% increase from 48 months to 60 months. It is postulated that the same effect would occur with humans, however the experiment would require 100 years.

    This information and much more is available on a non-profit research oriented website called Life Extension Institute.

    Note to German politicians, if you do not want to immigrate any more Muslims, zie mus ein bischen hormon gaben bitte.

  13. “…if you do not want to immigrate any more Muslims…”

    In recent years, most immigrants to Germany have come from Eastern Europe. In Hamburg, for example, Poles now outnumber Turks.