Eberstadt's presentation slides are available in Adobe Acrobat PDF format

James Kurth introduced Dr. Eberstadt, a Henry Wendt scholar in political economy at the American Enterprise Institute. His work there examines demographics, foreign aid, health disparities and economic development. He has also written extensively on North Korea, obviously extremely relevant in our time, but also a great deal in the past decade or two on countries of the former Soviet Union. He earned his A.B. and Ph.D. from Harvard University and a master’s in science from the LSE. He spent a decade as a visiting fellow at Harvard’s Center for Population and Development Studies. He has also served as a consultant to the World Bank, the U.S. State Department, and USAID. His books include Korea’s Future in the Great Powers (U.Wash., 20__) and Prosperous Paupers and Other Population Problems (Transaction, 2002). His recent articles include an article in Foreign Affairs on the future of AIDS, “The Population Implosion” in Foreign Policy and “Russia: Too Sick to Matter?” in Policy Review.

I’m going to be talking about four unexpected demographic surprises we’ve just encountered on the way to the end of history. Life is full of surprises, no less so for specialists in population studies than for other endeavors, because not to put too fine a point on it, the predictive powers of demography and population science are somewhat limited. To oversimplify just a little bit, there are two different types of questions for demography and population studies, the trivial and the unanswerable. We’re going to look at I hope some of the latter ones over the next few minutes.

The general worldview of demography and population studies is captured by the notion of the demographic transition, which is a concept that the great demographer Frank Milstein introduced several generations ago. The idea was, broadly speaking, that with advances in the diffusion of knowledge and improvements in income, there would first be a great, broad, general decline in mortality and some while after that there would be an equilibrating decline in fertility, through voluntary, deliberate reduction in child-bearing patterns.

This may broadly define the human condition now and in the generations to come, but there are a lot of footnotes here, and we’ll be looking at some of those footnotes.

I. Fertility Decline

Let’s start by looking at fertility decline. I’m not insisting that you look at any of these numbers or names, I present this table simply to show you how very many countries can be classified in a certain manner, namely as subreplacement fertility countries today. Which is to say that these are countries, according to the estimates and projections of the U.S. Census Bureau, whose current childbearing patterns, if continued indefinitely, would lead to an eventual stabilization and indefinite population decline thereafter, barring offsetting immigration.

I think we have about 100 different spots on the world that now qualify as subreplacement fertility venues. Subreplacement fertility is a relatively new phenomenon in the world, it’s a 20th-century phenomenon, when not attended by war, pestilence, famine and disaster. Subreplacement fertility was first seen in Europe in the interwar period and every single country in Western Europe now reports subreplacement fertility. There is a tiny exception of the Faeroe Islands, I believe, where just over 2.1 percent births per woman per lifetime, but Western Europe is subreplacement.

Eastern Europe is all subreplacement. Almost all of the former Soviet Union, certainly all the European regions are subreplacement. But most of the world’s subreplacement population is not located in Western or Eastern Europe or the former USSR, the United States, North America--most of it is elsewhere, in what we would think of as the third world, in Asia, almost all of East Asia except tiny Mongolia is now subreplacement; China, Hong Kong, Japan, South Korea, the latest figures for South Korea being 1.2 births/woman/lifetime; Thailand, Sri Lanka, I believe the latest figures for Vietnam show that that country is also subreplacement fertility. In addition, a large part of Latin America and the Caribbean is replacing subreplacement fertility, Brazil being the most populous of the countries in Latin America and the Caribbean to have estimated subreplacement fertility, but Chile, Cuba, other places as well.

Perhaps most surprisingly in the near East and North Africa, there are locales like Turkey and Lebanon and Tunisia that are now believed to be subreplacement fertility, but my favorite is Iran, which is now a subreplacement fertility society as best we can determine. We believe that much of the Arab Islamic expanse is resistant to modern ideas, but if modern ideas include reductions in fertility, almost all of North Africa has been touched by the trend, although not yet to the point of subreplacement fertility. What this table should show is the tremendous drop in fertility between the 1970s and the 1960s in the crescent from Morocco, Algeria, Tunisia, Libya, Mauritania, a tremendous change little recognized is also affecting those countries.

Rapid Aging of Population. In Europe, in North America, Australia, elsewhere, in Japan, one of the consequences of subreplacement fertility, only one, is the very rapid aging of populations, a substantial, pronounced aging. And the winner in this aging race will likely be Japan. By 2025, one out of nine people in Japan will likely be over the age of 80. There will be almost as many people over the age of 80 as under the age of 10, 15. It’s going to be a median age of about 50. But Japan, like Europe and like North America, will be a place where people get rich before they get old. In some of these other societies, people are going to get old before they get rich. The most important instance there being China. China’s population, in part due to low mortality, in part due to its awful population control program, but certainly to low fertility, is aging at an almost unprecedented velocity. This is the projection for China’s future aging (slide 4), you’ll see it’s much more rapid than the more developed regions, Europe, the U.S.--it’s about the same as Japan over the recent past. But its an extraordinary pace of aging that can be expected.

China will be getting to old age, having a large proportion of its population at elderly ages, at a much lower income level than Japan or the U.S. is with comparable portions of elderly population. By 2025, China will in effect have over 13 percent of its population 65 or older, I don’t think that I would count as a Sino-pessimist if I say that I doubt China will be here on the income graph by 2025.

With a rapidly aging population, China will have a triple bind. It doesn’t have anything that can be described as a national pension system except for the family. It’s unlikely to put one into place very quickly. The pension system that exists is extraordinarily unsound actuarially, it only covers about 1/6th of the population. So that means old people will have to continue to work if they want to support themselves. But work in China is much more oriented toward intense physical labor than work in Japan or the U.S. is. These types of work don’t favor the elderly or frail, generally speaking. China’s older population, because it’s a poor country, tends to be frailer than the populations of Japan and the U.S. or Western Europe. So there is a humanitarian tragedy in the making that may unfold before our eyes. How this will be handled--the social, economic and international consequences--remains to be seen.

II. Gender Imbalance

But this is not the only demographic disaster waiting for China. Thanks in part to its population control program, there’s been a strange and unnatural increase in the imbalance between baby boys and baby girls. In ordinary human populations, about 103-4-5-6 boys are born for every 100 baby girls. China has broken this natural law, if you’ll call it that, since its 1-child policy went into effect. In the 1982 census, almost 109 baby boys were found for every 100 baby girls; in a micro-census in 1995 it was almost 116. In the 2000 census, it was almost 120 baby boys/100 baby girls. As you can see (slide 9), there are some provinces in China where it was over 130 baby boys/100 baby girls. There is the possibility that this was a statistical artifact, that people might be hiding baby girls to get a crack at having a boy under regimented population control, restricted numbers. That’s why I thought it would be more useful to look at children age 1-4 (slide 10), where presumably there would be less hiding going on. And there, the ratio of young boys to young girls is about 121:100. Again, there are provinces where it’s over 130.

Interestingly enough, the imbalance cannot be explained by old-fashioned attitudes, poverty, or lack of education. If you throw this up on the board, what you find actually is that the higher the female illiteracy rate, the lower the imbalance happens to be in China these days. So better education for women is a predictor for greater imbalance in the sex ratio at birth, not vice versa. . . . If you went through urban and rural, you would find that it is more rural than urban, but you would also find that even in the urban areas we’re talking about 115--and of course it’s the higher parity births, the second and third births, where the imbalance is even more dramatic.

There was a better argument 20 years ago that these ratios reflected in some measure infanticide; because there was much less availability of sonography and prenatal sex determination technology. Now that sonograms and ultrasounds and other sorts of technology are broadly available, the inference one draws here is that this is almost all sex-selective abortion.

Now, given my prejudices and worldview, I initially am looking at this and assuming this was just creepy, China’s population control program that was responsible for this imbalance. However, if one looks at other East Asian venues, my hypothesis goes down in flames. Because there are other places, including South Korea, Taiwan, Hong Kong, and Singapore, where unnatural gender imbalances at birth have emerged. And these are not places where forcible population control policies are in effect. I think what we are seeing instead is a collision or confluence between an extraordinarily strong cultural son preference, subreplacement fertility (which brings each birth more to the point), and the diffusion of ultrasound and other technologies that permit prenatal gender determination and gender-based abortion.

It’s not just East Asia. There have always been very strong son-preference feelings in a number of cultures in society, one of them is Punjab, and if you look (slide 14), here you see that of people who express any preference for the gender of their child in Punjab, the ratio expressing preference for a boy over a girl is some 10:1. You can find this in other places as well, but I thought Punjab was a nice case. And if you look at India’s latest census, you’ll see that for children 0-6 years of age (slide 15), there are 126 young boys for every 100 young girls in Punjab. This isn’t a sex ratio at birth, this is children age 7 or under, and for a variety of reasons we can’t take this as an exact indication of gender imbalance at birth, because people may have migrated, there may have been mortality after birth. But you will see that in a number of states in India there are truly amazing imbalances between boys and girls. These are higher than ever recorded before. With increased affluence, education, and contact with the outside world, the gender imbalance has gone up, not down, in China and in India, and it is starting to do the same in other places in the world. Note (slide 16) the Caucasus for 2001, 115, 118, 119:100; in Armenia, Azerbaijan, Georgia, and in other places that do keep relatively complete records of births and deaths, there are very strange sex ratios at birth being recorded that I don’t know enough about to be able to explain to you. There’s e.g. to wit Cuba, 118:100 in the mid 1990s (slide 17).

The consequences of this imbalance are most likely to be extreme in subreplacement fertility situations where they affect the marriage market a generation hence. And China is poised for having a major, very difficult shock to its marriage market in the decades immediately ahead.

III. Mortality

Another presumption, that with improved income, increased globalization, increased economic development, there would be an increased spread of ideas, knowledge and technology, there would be a gradual decrease in mortality all around the world, the convergence of mortality levels, “day-by-day in every way we’re getting better and better.” And except for wars and other little unpleasantnesses, that was true for France over the long durée between 1800 and the present, as you see in slide 18. And in the post-WWII period, there was a gradual convergence of life expectancy between the more developed and less developed, the rich and the poor areas of the world (slide 19). This was a decrease in inequality about which there was strangely little comment, except for worries that the people in the third world were populating too rapidly--that was a consequence of their health explosion.

The picture is no longer quite so happy, as this graphic (slide 20) indicates. Again, I can show you 40 some countries for which the census bureau is now projecting lower life expectancy in 2010 than it calculated for 1990--a 20-year decline in anticipated life expectancy. Hardly a trivial, minor setback. Very much of it is AIDS--for subsaharan Africa, it’s almost all HIV/AIDS--but you’ll also see that in the former Soviet Union there are expected reversals for quite a number of countries. Much of that is due to increased infant mortality, with the disintegration of the not terribly impressive Soviet-state primary health care system for infants. One of the countries that is not up on the board but should be is the Russian Federation (slides 21-22). I think the Census Bureau is too optimistic about Russia’s trends. Over the last 40 years, Russia has given lie to the generalization that industrialized, literate peacetime societies can’t see health reversals. Russia’s life expectancy today combined for both sexes is about four years lower than it was 40 years ago. And with the reversal in the health as measured in mortality is concentrated in the working age groups. You’ll see it really is true. Increases in death rates for men my age are almost 70 percent over the last 30 years. And of course that has sent Russia to negative natural increase and population decline at a fairly steep pace (slides 23-24). The only question is how fast it will be, I don’t think anyone’s imagining that there’s going to be a turnaround any time soon, in part because of the difficulty of bringing a turnabout.

The causes for this explosion of death in Russia have not been TB or HIV/AIDS, those are horsemen yet to ride through the great steppes. It’s been cardiovascular death and injury, with alcohol being a co-factor (slides 25-27). There was a study some Russian public health people a while ago who said that in the autopsies and in coroner studies that were done, two thirds of the Russian men examined were drunk at the time of death.

Now, here’s the problem. People who are now in their 30s or 40s or 50s in Russia have a lifetime of insults on their health and their bodies and their immuno-systems to contend with. Turning that around is very difficult. If you take a look at Japan (slide 28), each younger generation, each younger birth cohort had better survival chances that the one before it, so that people who were born at the turn of the century had the worst death rates at any given age of life, and people who were 10 or 20 or 30 years younger than them had geometrically better survival chances. This is what Russia looks like (slide 29). The people who’ve got the worst death rates are the youngest of any given age cohort now. Which is to say that people in Russia today are less healthy, to judge by mortality, than their parents were at the same age. And that means that it’s extraordinarily difficult to improve the health of the society as a whole. If one simply got to the point where Russian men’s health was as bad as their fathers’, which would be quite an achievement right now, Russia would have a life expectancy for men of 63, which is a little lower than life expectancy for men in India today. And that would be quite difficult to achieve. And for this reason, Russia’s life expectancy for men is anticipated to have fallen behind that of the third world, and if you’re optimistic like the UN Population Division people, you imagine that Russian male life expectancy may catch up with third-world levels in 20 years (slide 30). But I think there’s a lot of optimism in that.

IV. America’s Demographic Exceptionalism

A fourth and final surprise has to do with I guess what you’d call America’s demographic exceptionalism. We’ve heard a lot, I’m sure in your seminars and elsewhere you’ve talked about this phenomenon of American exceptionalism. Since the 1970s the United States has emerged as a demographic exception to the trends of fertility decline that I illustrated at the beginning of this chat. And I can’t tell you exactly why this is happening, but it’s really pronounced, and it’s worth taking a look at. The story can be told quite quickly by a couple of charts. Slide 31 shows Russia’s population structure in the year 2000 and also how it is projected to be in the year 2025. The red and the blue would be Russia in the year 2000 by age groups. This indentation here is how many people are infected within each of those age groups in the year 2025. You see that Russia’s eaten away at the bottom but replaced a bit at the top. Slide 32 is what Western Europe looked like in the year 2000 and how Western Europe will look in 2025, as imagined by the U.S. Census Bureau. Slide 33 is the United States. Notice the difference? We’ve gone from having almost an upside-down discarded Christmas tree to a narrowing upside down pyramid to more or less a straight block. Since the area of these graphics shows how many people are imagined to exist in this future world, you see a much smaller Russia, a somewhat smaller Western Europe, a distinctly larger United States than the one we know today. What’s going on here?

Part of what’s going on is a divergence in fertility patterns. In Eastern Europe since the end of communism, there’s been a big fertility shock. Many of the countries of Eastern Europe are recording rates of total fertility which, if continued, would imply 1.2, 1.3 births/woman/lifetime.

(Slide 34) In Western Europe, the aggregate would probably now be something like 1.5 births/woman/lifetime, although there’s quite a range from Italy and Spain, at about 1.2, up to France and Ireland, which are now in the 1.8-1.9 range. The U.S. is bouncing around a little under 2.1. It would be the most fertile country in Europe today if it were a country there, to judge by recent TFRs. Part of this is due to changes in marital patterns and extramarital births (slide 35). We bemoan the out-of-wedlock situation in the United States, I think, for many reasons rightly so. But as the demographer Waylon Jennings said, “If you think I look bad, you should have seen the other guy.” The U.S. is the red bar on the far right. Compare that to Western and Eastern Europe. If you wanted to play the ethnic game, in the U.S. the so-called Anglos--the white non-Hispanics--are reporting out-of-wedlock proportions of about 22, 23 now. If you want to match up so-called European populations in the U.S. with so-called European populations in Europe, the U.S. would be much lower actually than most of Europe in this regard now.

Even North America doesn’t look so American, if you follow my drift. (Slide 36) The U.S. and Canada had very similar levels of fertility back in the mid-1970s. Canada looks pretty European now. The U.S. looks pretty American. We’re over 2, they’re headed down toward 1.4. There is a possibility that this increase in American fertility may be a timing effect and that the European decrease may also reflect timing. Which is to say that if women decided that they were going to delay childbearing but end up having two kids anyhow, and they did that en masse in a certain period, you’d see a drop in the recorded year-to-year synthetic rate, with a kind of catch-up later on. That seems to be what happened in the U.S. in the 1970s. Some people suspect this may be what’s happening in Europe and Canada today. But it isn’t over til the fat lady gives birth! At some period of time you have to say that this plausible hypothesis that’s had a generation or two to be under consideration and people are kind of at the end of their game. It is possible that there is some of this change in timing in Europe, but with continuing decreases in fertility being reported by this snapshot rate, that argument becomes increasingly less plausible with each passing month.

And just to emphasize American exceptionalism once again, slide 37 is a graphic showing the different recorded snapshot fertility rates for America’s 50 states and D.C. for the year 2000. Howard Dean’s Vermont is the most European, George W. Bush’s Texas is 2.5. The most interesting note to me is that I think it’s 30, 31 of the states plus the D.C. are at 2.0 or higher. These include places that aren’t terribly known for their ethnic diversity, such as Minnesota. So there is something broad and common to this exceptionalist pattern in the U.S. today.

But part of what we see if we try to disentangle this, this can’t be explained away as ethnic disparity (slide 38). There are differences by reported ethnicity, but they don’t explain away this striking difference. American blacks have a somewhat higher than national average level of fertility, in the sense of being 1 percent higher than the national average; Hispanic Americans are definitely higher, 2.7 is the latest estimate, although within Hispanic U.S. the Cuban population is subreplacement, Puerto Ricans are just barely above replacement. Cuban Americans have higher fertility than Cuban Cubans at this point, I believe. In Japan, similarly, Japan is at about 1.3; Japanese Americans are about 1.2 or 1.1. But our good old-fashioned Anglos, the ones who profess to have ancestors coming from Europe, there are a couple of spots in Europe that have the same sort of fertility level today as reported by Anglo U.S. types. On the whole, the so-called white non-Hispanic fertility level in the U.S. is 4/10th of a birth higher than Europe’s at this point. And when we’re dealing with these sorts of numbers, that makes a big difference.

Now why is that? I don’t know. But I am intrigued by the lack of materialist explanations for it. Income differences between Western Europe and the United States aren’t tremendous. Differences in vacation and free time, well actually in Western Europe there’s more free time and vacation then there is in the United States. The Americans are the workaholics. You’d think that would favor higher levels of fertility in the place with a little bit more leisure. Social support? The infrastructure is greater in Western Europe than it is in the United States. So how do we explain this? Well, possibly it has to do with attitudes and outlook. Possibly it has something to do with this. And there are big revealed differences in attitude and outlook between U.S. Americans and Europeans. I just took a couple of these out of a recent survey that was run in the Economist (slides 39-41) because they’re kind of striking. “Which is more important for government? To provide freedom or to guarantee one’s needs?” The U.S. has a completely different set of responses than Germany, France, Britain, Italy. “Religion plays a very important role in my life.” U.S. versus these countries, if you look at other countries in Western Europe, the numbers wouldn’t be that different. You’d have the same sort of disparity. “Are you proud to be an [American, Briton, French, Italian, German]?” Again, patriotism, religiosity, the individualism. If we had a graphic for “Are you optimistic about the future?” I think we would see a similar sort of distinction between the U.S. and many of the countries in Europe.

[Responding to a question about whether optimism/birth rates are linked--Now, I can’t prove it, but it’s an interesting conjunction, the conjunction I find most worth further pursuit. . . . If you did the educational attainment by years in school, you’d probably have an even more striking disparity, since Americans tend to be overeducated. We spend an awful lot of years in school, compared to Europeans. Europeans may be able to multiply and divide and read and compose better than us, depending on what grade level you’re looking at.]

One of the consequences of having almost replacement fertility as opposed to distinctly subreplacement facility is that the U.S. does not require arithmetically large flows of newcomers to stabilize its population over the generations to come (slides 42-44). We actually have, as you know, quite a large flow of immigration from abroad, larger than we knew until the large census. But these numbers were calculated by the UN Population Division to see what sorts of flows would be necessary over a 50 or 55-year period to keep populations from declining or to keep the so-called working age population (15-64 years of age) stable. And you can see here that for the EU, you need about 2.5 million people/year to stabilize the population, about 4.3 million to keep the workforce as so defined stable, as opposed to 700,000 that were coming in a year in the late ‘90s. Way bigger for the U.S., which has larger current inflows than is needed for this. Part of the consequence of inflow is that you get a different composition of descendants from what you’d have had before. For the EU, if you were stabilizing the population through inflows, you’d end up in 2050 with about a sixth of the population being the descendants of these new immigrants. If you lived in Germany, it would be almost 30 percent, the U.S. much smaller, even though we have much higher immigration. Our population will, because of these exceptionalist trends, age much more slowly than Western Europe’s or Russia’s. And we won’t fall off the globe the same way that Western Europe and Russia are poised to fall off the globe as a portion of total world population. The red line heading down toward zero from the top corner (slide 45) is Europe from 1950 to 2050 in these admittedly sci-fi long-term projections. The yellow line that’s going from 4 percent to 1 percent of world population is Russia as envisioned in these projections. The U.S. proportion of world population drops from 6 percent to about 4.5 percent around now and then by these projections stays there as long as I’m going to be around! Interestingly enough, if you try to see how different population totals in the world would look (slide 46), the U.S. goes from being, with these redone borders, chopping Russia out of the Soviet Union and putting Western Europe together into a single entity, the U.S. goes from being the 4th largest entity in 1950 to the 4th largest now and the 3rd largest in 2050. Western Europe becomes the 4th largest, just edging out Indonesia and Nigeria. Russia is down below Ethiopia and I think right above Vietnam, if the chart were a little bit bigger, as the 16th or 17th largest country in the world in 2050.

So these are the four unexpected surprises. They are all immediately affected by human behavior, values and decisions. It’s just that these values, behavior patterns, and decisions look a little bit different from the standard demographic, globalist template.

Nicholas Eberstadt is the Henry Wendt Scholar at AEI.