« Jewish smarts | Gene Expression Front Page | How humans lost a sense (sort of) »
July 24, 2003

AFRAID OF GROWING OLD?

This is aimed at provoking discussion on the problems of an elderly population.

In most Western countries the proportion of old people in the population is rising. The increase is greater in some countries than others, due to differences in migration, birth rates, and life expectancy. It is greatest in Japan, and lowest in the USA, with Western European countries somewhere in between.

Alarming conclusions are drawn from this trend. It is feared that old people will be a serious burden on the working population, and that measures must be taken to encourage a higher birth rate, or higher immigration.

I do not claim that there is no problem, but I do think it is often exaggerated. Some commentators and politicians have reasons of their own for wanting to promote immigration, or a ‘Kinder, Kirche, Kuche’ society.

The growing proportion of old people is due to several causes. One of them is obviously the increase in life expectancy. Between 1900 and 2000 life expectancy at age 60 increased by about 50%. If this were a transition from one ‘steady state’ to another, we would expect there to be roughly a 50% increase in the proportion of over-60s in the population.

In practice, populations are seldom in ‘steady state’. They are either growing or shrinking. If the population is growing, the proportion of old people will be less than its ‘steady state’ level. But if growth slows down or stops, the proportion of old people is likely to rise. Notably, after World War II the birth rate rose sharply, and then fell, producing a bulge of ‘baby boomers’ who are now approaching retirement age. This will undoubtedly increase the proportion of old people over the next 20 years.

Another problem in the long term is the fact that in most western countries the birth rate has fallen below the replacement rate. If this continued indefinitely then populations would eventually shrink to nothing. As I’ve argued previously, predictions of birth and death rates are unreliable, and birth rates may not continue below the replacement rate for very long.

But in the mean time, a low birth rate does have implications for the ‘dependency ratio’. As usually defined by demographers, this is the ratio of people both above and below working age to the working-age population. Both groups of dependents need to be supported. In many countries the dependency ratio is now actually lower than 100 years ago, when there were relatively few old people but a lot of hungry children.

The impact of a low birth rate on the dependency ratio is complicated. If there is a sustained reduction in the birth rate below replacement level, the first effect is obviously to reduce the number of dependent children, without affecting other age groups. It therefore reduces the dependency ratio. This continues for about 20 years (the average period of dependency). Then the working-age population begins to shrink, and the dependency ratio rises, but this is partly offset by a further fall in the number of dependent children, as the child-bearing age-group is also shrinking. The overall dependency ratio does not get back to its original level until about 50 years after the start of the lower birth rate regime. Then it continues rising for about another 10 years, until the number of people reaching retirement age also begins to shrink. After this the dependency ratio falls again, though not quite to its original level.

The main point to note is that the impact of a sustained low birth rate on the overall dependency ratio is very gradual, and smaller than might be expected.
Even with a Total Fertility Rate of only 1.5, the dependency ratio should not rise to more than 10% above its steady state level from this cause alone.

More serious problems arise when several factors are raising the proportion of old people at the same time. At present Germany and Japan in particular are facing a triple whammy, due to (a) very long life expectancy, (b) very low birth rates, and (c) a large bulge in population about to reach retirement age. But even these problems shouldn’t be exaggerated. According to the German statistical office, by 2050 about 31% of the German population will be over 65, 17% will be under 20, and just over 50% will be aged 20-65. A ratio of one worker to one dependent should not be intolerable. And countries as densely populated as Germany, and especially Japan (with over 10 times the density of the USA), could benefit from a gradual reduction of population.

Of course, a lot depends on the level of income expected by old people. At present, German pensioners expect to receive two-thirds or even three-quarters of their previous income. I doubt if this is realistic in the long term.

A lot also depends on the proportion of the population who are actually working. In most countries, relatively small adjustments in the average retirement age, or the number of people doing part-time work, would offset the increase in the dependency ratio.

That’s enough for now, but I can’t resist quoting Frederick the Great (I think), when his troops were trying to retreat: ‘What, do you scoundrels want to live for ever?’

DAVID BURBRIDGE

Posted by David B at 03:40 AM




It was indeed Frederick the Great, at the battle of Kolin in 1757. But the quote is usually rendered 'Dogs, do you want to live forever!?' ('Hunde, wollt ihr ewig leben!?').

Posted by: bbartlog at July 24, 2003 07:02 AM


raising the qualifying age for pensions is strictly speaking not unfair nor really a zero sum game because from what i understand most pension systems were designed for shorter life expectancies.

Posted by: Jason Soon at July 24, 2003 12:29 PM


There already is a job track for semi-retired. A lot of retirees are quite capable of working. My mom is 85 and she's still active and alert, and a friend's father worked six days a week until he was 75. (The Protestant work ethic, which is a wonderful, wonderful thing for other people to have). Many people actually want to keep working, but not necessarily in the job they had before and probably not 40 h/wk. If there ever really is a labor shortage, I bet you'll see jobs being customized for retirees.

Posted by: zizka at July 24, 2003 08:40 PM