Careful with That Long-Term Planning:

Robert Samuelson writes, in the Washington Post:

Europe as we know it is slowly going out of business. . . . It's hard to be a great power if your population is shriveling. Europe's birthrates have dropped well below the replacement rate of 2.1 children for each woman of childbearing age. For Western Europe as a whole, the rate is 1.5. It's 1.4 in Germany and 1.3 in Italy. In a century — if these rates continue — there won't be many Germans in Germany or Italians in Italy. Even assuming some increase in birthrates and continued immigration, Western Europe's population grows dramatically grayer, projects the U.S. Census Bureau. Now about one-sixth of the population is 65 and older. By 2030 that would be one-fourth, and by 2050 almost one-third. . . .

There's much to Mr. Samuelson's article, which may well be generally correct. But I think it's something of a mistake to make demographic predictions for what happens "in a century" "if these rates continue," or even in 2050. I doubt that a century ago we could have anticipated the demographics of Europe in 2005; I doubt that 45 years ago we could have anticipated the demographics of the U.S. in 2005 [UPDATE: I particularly had in mind the magnitude of Hispanic and Asian immigration, especially given that Asian immigration was largely opened up by a political decision in the mid-1960s]; I doubt that today we can anticipate the demographics of Europe in 2105 or even 2050. I doubt, for instance, that someone 100 or even 45 years ago could have guessed Italy, seat of the Papacy, would have such a low birth rate.

Too much depends on shifts in culture, immigration, economics, and to some extent medicine. Some things one can plan on with somewhat more confidence: People who are 30 today will be 75 in 2050, and barring a major war, plague, massive emigration, or massive immigration of older people one can make a good guess about how many of these 75-year-olds there'll be in 2050. (Immigration, I suspect, is the main variable, but one can have a decent idea of how many will immigrate in the next 10 years, and past that we're talking about immigration of 40-to-75-year-olds, which I suspect is rarer than immigration of younger people.)

But birth rates and rates of immigration of young people are, I think, much harder to estimate. And it seems to me a mistake to just assume that things will stay more or less the same over that long a timespan. They rarely do.

I've enabled comments.

Greg Erickson (mail):
Dear Professor Volokh:

While I very much agree with your warning that it may be "a mistake to make demographic predictions for what happens in a century," (or to rely heavily on any long term predictions, for that matter) I think the problems associated with declining birthrates in Europe actually have been a matter of concern for 40-50 years. (I believe a little research will confirm that the problem - and its implications - has been widely recognized for the last 20-30 years.)

The magnitude of this problem, however, is clearly not understood. Our debate (or lack thereof) regarding Social Security reform should begin with a look at its European counterparts, for despte their many differences both are "pay as you go" systems. Both WILL's just a question of when and how bad will it be.

(I won't even begin to discuss the underlying geo-political and ideological problems that the decline of the West might portend.)
6.15.2005 2:28pm
Chris Lansdown (mail) (www):
"But I think it's something of a mistake to make demographic predictions for what happens "in a century" "if these rates continue," or even in 2050."

Actually, the value of such predictions lie in trying to avoid them. If one is speeding off of a cliff, it's true that most people will vary their speed and direction so as not to drive off of it, but in most cases it's the knowledge of what will happen if they don't alter their course that prompts the alteration.

The other value in such predictions is that a group's power is, while mostly a function of their current power, also a function of their likely power. That's why lame duck presidents have that nickname, and why they have a harder time negotiating if their likely successor will be less tough — what direction Europe's future is headed does to some degree determine how influential Europe can be now, because it tells whether time is on the side of the current Europeans.
6.15.2005 2:29pm
gr (mail) (www):
Does this mean that all that talk about social security having a projected deficit is bunk, because we can't know? I think we still get some benefit of talking about current rates, and at least working out scenarios and guessing their probabilities.
6.15.2005 2:32pm
Bill Harshaw (mail) (www):
I thought the disclaimer: "if these rates continue" adequately covered Mr. Samuelson's derriere. As a pundit, he's careful to avoid any estimate of the likelihood of these rates not continuing, which is your point. Maybe you want to blog on Judge Posner's "Catastrophe"?
6.15.2005 2:36pm
Abe Delnore (mail):
This is an old issue. The case I know best is France's.

French politicians and social observers defined and recognized a natality problem at least as early as the immediate aftermath of the Franco-Prussian War (1870-1871). The immediate issue was that the French birthrate lagged behind Germany's, thus in the context of German unification, Franco-German antagonism, loss of Alsace and Lorraine to Germany, and war fought by conscript armies, this meant France faced a military deficiency vis-a-vis its most likely enemy. Some thought the best way to reverse this situation was to undertake policies that would encourage more births, and indeed a natalist position is identifiable--usually on the political right--for much of the history of the Third Republic. This idea remained present even after the threat of war with Germany faded. Simply, to have a healthy nation you need lots of healthy babies.

I think you would find similar concerns elsewhere in Europe. And you would also find that natalist policies have failed universally--the birth rate continued to drop no matter what, even under governments, such as practically all fascist regimes (Nazi Germany, interwar Italy, Vichy France) that made increasing the birthrate a national priority. (To be sure under these regimes there were eugenic concerns as well, meaning that policies were directed toward creating more of the "right sort of baby" and fewer of the "wrong sort." The latter goal proved much easier to achieve than the former.)

The answer, of course, to the gap between babies and workers will prove to be immigration as many observers have already figured out. Unfortunately many Europeans on the political extremes want to have their cake (prosperity and welfare) and eat it too (retaining "national character" as defined in ethnic and cultural terms).
6.15.2005 2:45pm
Catfish (mail):
These estimates also assumes that immigrants to Germany, Italy, et al will not intermarry or culturally assimilate. Many turn of the 20th century American demographers likewise predicted "racial suicide," due to a declining WASP birth rate, but eastern and southern European immigrants seem to have intermarried or acculturated pretty thoroughly. Estimates of the "decline of the West" made in, say 1910 would have been widely off.
6.15.2005 2:47pm
cb (mail):
The best example that I heard on why it is dangerous to project present trends 100 years into the future is this: if the trends of the last twenty years continue, by the year 2105, one out of every six Americans will be employed as Elvis impersonators.
6.15.2005 3:27pm
Jim H. (mail):
Yes, yes, it is foolish to make 100 year predictions, but I could not help but greedily rub my hands together in the hope of snapping up a good portion of Sienna at bargain rates (esp if the dollar continues its rally until 2050)with some friends and living out a mock-16th cent life (with the conveniences - and $ - of cent21)in my retirment...
6.15.2005 3:35pm
David M. McClory:
An often unheeded danger is to project a straight line far into the future. It is the progenitor of many foolish political positions.

That we are still alive proves most of such projections wrong.
6.15.2005 4:53pm
Victor (www):
I doubt that 45 years ago we could have anticipated the demographics of the U.S. in 2005

Some examples to ponder ... the babyboom demographic problem was picked up on very quickly in this country; the 1960 Social Security Trustee Report predicted just over 33 million beneficiaries by 2000; I think the real number turned out to be more than 38 million, but perhaps with broader eligibility criteria. The 1983 Social Security Trustee report predicted the graying we are experiencing today with remarkable accuracy; current predictions are little changed from what I can see. Interestingly, the 1960 report predicted a decline in fertility to the status quo replacement rates by about this time (again, the high cost scenario); we are just shy of that mark, although we did get there more rapidly than predicted (with fertility dropping substantially below replacement levels for awhile in the 1970s before rebounding).

Your assertion is an empirically verifiable one. Actuarial Study Note #46 from the Social Security Administration was published in the late 1950s, and could possibly be used as a test, if anyone has access to it.

I'll join in your criticism of the "century" remark because the demographic shift itself may cause an unpredicted backlash that could begin to be manifested within that long time frame. But changing a country's demographics is like trying to steer the Titanic. Further, the ship's momentum is dictated by biology in ways that are unlikely to change anytime soon. His comment about the picture in 2050 may not be as "noisy" as you would think. Barring additional evidence, I think there is reason to accept his shorter run assertions with a fair degree of confidence.
6.15.2005 5:04pm
Whenever I read studies/stories like these, I am always reminded about the "population bomb" scare from the late 60's to the mid 70's. The people behind ZPG should be happy that they have succeded in the quest to reduce population growth.

For the moment, I'll ignore that the ZPG group failed to mention the down sides of the puposals they made.
6.15.2005 6:05pm
Sam Loftin:
It seems to me than the solution for European countries is to encourage more immigration from the United States. There are many in the U.S. that greatly admire the European way of living, and would probably jump at a chance to move there (assuming that it isn't already too late and they have emigrated to Canada instead.)

The new immigrants from the U.S. would continue to support the social and economic model beloved by Europeans.
6.15.2005 7:07pm
Fred Citek:
Dear Professor Volokh,

I agree that effects of immigration and for that matter emigration are difficult to predict. They depend on the legal framework and attractiveness of the target country as well as the willingness of people to leave their homes and friends.

However, births are another matter. Despite technological advances most women will not be able to conceive after age 40 and most middle age couples are not willing to take on the task of raising children. Unless Europe's dwindling supply of young people decides to become incredibly fecund within the next twenty years (I leave it to your imagination what would trigger that) the age distribution in 2035 is pretty much set.

I would hazard that youth unemployment in Europe has had and will continue to have a negative impact on family formation.

What is likely to happen as Europe depopulates is that the returns for labor will increase but I do not expect that people chose to have children because they (the children) will have better wages.
6.16.2005 9:38am
Perhaps the unnoticed variable in all this is the supposed need for immigrant labor in the future. It seems to me that AI, robotics, and other labor saving devices could reduce the need for human labor to the point where we will all end up being each other's therapists.
6.16.2005 9:40am
Mr. Bingley (www):
I think an important aspect leading to declining birth rates, and I would hazard a guess that this is specific to middle class families with plans for their children to attend a good college is not so much the 'work' involved in raising the kinder but rather the cost. The amount of debt facing a family who has 3 or 4 kids to put through school is staggering, especially mid-to-upper middle class folks for whom no funding is available.
6.17.2005 4:42pm
Taeyoung Jensen:
My guess would be that as far as domestic birth-rates go, a 50 year prediction is not impossible. That's only two generations from the present, after all, and it looks (to me) to have an increasing cultural component to boot--many people nowadays do not merely object to the expense or suffering of childbirth and child-rearing, but believe they have a sort of moral duty *not* to procreate. Contraceptives and other artificial work-arounds make this possible, while decreasing cultural/familial pressure to marry and extend the family and perhaps the increasing isolation of our *concept* of family (widely viewed now as two individuals standing alone, outside the context of a broader society, such as their extended families) make it more acceptable. This active desire *not* to have children may change as my age cohort (20s) grows older, but I am not aware that previous generations larded up the determination not to procreate with the same ideological or moral weight as many of my contemporaries seem to. Or perhaps it is just the people I know. I didn't live through previous generations, after all.

A 50-year prediction may be wrong, of course--there may be some dramatic shift in the next generation, like a backlash as suggested above--but from our present perspective, there's no compelling reason to expect one, and from a policy perspective (i.e. pensions, social security, etc.) it doesn't seem quite sensible to bank on there being such a shift, instead of preparing for the eventuality, Coolidge's wheels notwithstanding. This is so for Europe (and Japan) just as for us.

Immigration as a factor seems much more uncertain, because immigration doesn't deal with slow generational change but attracts from a large pool of already-born people in other countries. There, though, I think much of the impetus for immigration is likely to be economic, and economic factors can shift much more quickly than mere birth-demographics, which have a ~20-yr lag insofar as their effect on the mature population is concerned. To the extent, therefore, that we succeed over the next 50 years in improving the quality of life and economic opportunity in foreign countries (through liberalised trade and the like), and to the extent that globalisation and improved telecommunications infrastructure make jobs more "portable" across borders, this pressure to immigrate will likely decline. Alternately, as foreign conditions grow more miserable, the volume of potential immigration will increase, but I should like to think that's not going to be the case.
6.18.2005 12:54am