Demographics--Implications for Our Changing World

Griffith Feeney
East-West Center
Honolulu, Hawaii 96848

Demography has been mentioned more than once here, and very pertinently. It's a pleasure, as well as a convenience, to address an audience so well primed for this presentation.

I'm going to give you the big picture, world wide and over the long term. I'm not going to shower you with numbers, and those I do show you will all be in the form of plots. My aim is to give you an understanding of pure demographic dynamics during roughly the century we are now in the middle of: 1950-2050.

I'm not going to say a great deal about implications directly. You have as many and as valuable thoughts about this as I have. Let's take them up in discussion. My most useful contribution here will be to clarify exactly what it is we are talking about the implications of.

World Population Dynamics

Here's a picture of the entire population of the world in 1990, “broken down,” as we demographers say, by age. Why age? Two reasons. First, everything you're interested in changes with age, and changes a lot. Second, age distribution is the key to demographic dynamics in the world today--and in every part of the world today.

Figure 1: Age Distribution of World Population in 1990

Figure 1
Age Distribution of World Population in 1990


The horizontal scale shows the range of ages in our life span. The vertical scale shows numbers of persons in millions. Each plotted point--each solid dot--shows the number of persons in a given year age group.

The far left dot, for example, shows that there are about 620 million persons between the ages of zero and four years. The far right dot shows that there were hardly any of us 80 years old and older, a mere few tens of millions.

This is the way we were, as a population, in the year 1990. What do we see? Nothing subtle. A lot more of us were young than old—2.2 billion under age 20 as compared with a mere 326 million aged 65 and over.

Okay. Now, two questions. First, why? Second, what (if anything) does this tell us about the future?

First the why. Migration aside—and it is a tertiary factor—the shape of any age distribution is determined by two factors, mortality and population growth. The first influence is obvious:

the further out we go in age, the fewer people survive. The second influence is less obvious. Old people were born longer ago than young people. In a growing population, fewer people were born in the more distant past than in the more recent past. We have fewer old people in a rapidly growing population because they are the survivors of a time when there were fewer births.

What does this age distribution tell us about the future? A great deal, if we know how to interpret it. I want you to study this picture and think about how it will change between 1990 and 1995. To help, I'm going to let out a trade secret, the most important fact in demography. The most important fact in demography is that we all get one year older every year

What that means is that all the 0-4 year olds in 1990 will be 5-9 year olds in 1995, all the 5-9 year olds will be 10-14 year olds, and so on. All? Well, not quite all. There is an important qualification: we all get one year older every year if we survive.

When we're young, our changes of survival are very high, so those points move five years to the right and only slightly down. When we're older, our chances of survival decline, so these points move five years to the right and down rather more.

To make this concrete, let's look at what actually happened between 1990 and 1995. Here's the picture.

Figure 2: Age Distribution of World Population in 1990 and 1995

Figure 2
Age Distribution of World Population:
1995 compared with 1990


To help you see the change, I've shown not only the age distribution in 1995, but also a shadow of the age distribution in 1990. If you look at this closely, you'll see that the points mostly move over five years to the right. They move down too, but not by very much. To help you see, I've drawn in two reference lines, for persons aged 25-29 in 1990 and persons aged 65-69 in 1990. The younger group looses very few people over the five year period. The older group looses more, but the vast majority remain.

Something very important is emerging here, but five years is not a long enough period to show it clearly. Let's look forward 20 years from 1990, to the year 2010. Before I put up the transparency, I want you to guess what it's going to look like. Just mentally move these points (point) to the right by 20 years.

Here's the picture.

Figure 3: Age Distribution of World Population in 2010 (Compared with 1990

Figure 3
Age Distribution of World Population:
2010 compared with 1990


Again, the 1990 distribution is shadowed for comparison. There are several things going on here. The number of people 65 plus increased by about 180 million. The number of people aged 20-64 increased by about 1.2 billion. And the number aged 0-19 increased by about 360 million.

Where did most of the growth occur? Among persons over age 20. Think about that. Where did this growth come from? Could it have been reduced by effective family planning? Of course not. These people where already born in 1990. What a minute. We can't get more people that were born, what's going on here?

What's going on here is the shape of the age distribution. It was tilted down to begin with, and it moved to the right, and when you put those two things together you get growth. The most important influence on population growth is not fertility, and it's not mortality. The most important factor in world population growth today is age distribution--together with relatively high life expectancy, which means that the points on these plots don't move down very much as they move to the right.

Several important conclusions follow.

(1) Most of the population growth we are looking at in the next 20 years has already happened. It's there in the age distribution. It is cannot be affected by "population control" efforts, no matter how vigorous. Birth control can prevent births, but it can't prevent people already born from surviving.

(2) This population growth is highly predictable. The only thing that can prevent it is death on an extraordinary scale. Does your business have a contingency plan for calamities that kill a billion people over the next 15 years? No? Well then you can count on a large increase in the population of adults. This is, in Peter Drucker's excellent phrase, "the future that has already happened."

(3) The flip side of this is that the projected numbers of persons under age 20, which depend mostly on forecast numbers of births, are much less predictable. Reality could turn out to be substantially higher or lower. Not all forecasts are created equal, and one advantage of looking at age distributions over total population numbers is that you can sort out the reasonably certain from the more speculative.

(4) These population increases are all of adults, i.e., potentially, at least, consumers. If a modest fraction of them convert to consumers--and the rest don't gum up the works--Bruce has his 300 million. The qualification is what Gene cautioned us not to ignore.

It's worth pushing this once more, this time all the way out to the middle of the next century.

Figure 4: Age Distribution of World Population in 2050 (Compared with 1990)

Figure 4
Age Distribution of World Population:
2050 compared with 1990


What we see here is world population aging, the shift from an age distribution with large numbers of young persons and small numbers of old persons to an age distribution with nearly the same number of persons at all but the oldest ages.

There is a significant complication in this picture for 2050, however. We've looked 55 years ahead, here, and that means that all the numbers of persons under age 55 are based on projected numbers of births, rather than on persons living in 1995. That in turn means that the whole picture is a good deal more speculative than the previous ones. Useful conclusions require more effort, for we have to think through how things play out if the picture isn't on target.

Suppose the picture is not on target. There are several ways this might happen. We could have fluctuations in births about the indicated level, about 650 persons every five years. In terms of aging that ends up being more or less the same thing.

Or, we could have substantially fewer births than anticipated. The left side of the plot is lower. The magnitude of aging would be even greater than what is shown here.

Or, finally, we have substantially more births than anticipated. This is the only scenario that gives us less aging. Note that this lessor aging goes together with higher birth rates, faster population growth, and a much larger population.

Where will we come out? I can't give you evidence in this short talk, but the bottom line is: we may not have this much aging, but we will certainly have a lot of aging, and more aging is as likely as less aging

I promised to give you the big picture. Here it is. Never before in history have human societies lived with large numbers old persons. In the more distant past, very high mortality kept the numbers low. More recently, mortality decline and the consequent acceleration of population growth has kept numbers of old persons relatively low.

The old age distributions we face in the future really are something new on the face of the earth. We—and our parents and grandparents and children and grandchildren—are living through an historical transition that has never occurred before and will never happen again. At least we should all hope it will not happen again. Why?

There are only two ways we could return to the "young" populations of the past: a radical deterioration in health conditions and consequent return to high death rates, or renewed rapid population growth.

The United States Baby Boom

Now that you have a good feeling for this device of plotting age distributions, I'd like to use it to help you visualize the US baby boom, because we need to clarify the relation between aging in general and the aging of the baby boomers. Here's the age distribution of the US population in 1950.

Figure 5: Age Distribution of the United States in 1950

Figure 5
Age Distribution of the United States in 1950


The dip here (15-19) corresponds to 1930-34, the depths of the great depression. The rise to the left is the beginning of the baby boom, which started in the mid-1030s, in fact, not post-war, though there are some wartime wiggles in annual numbers of births that confuse the issue

Move forward 20 years to 1970 and you get this picture.

Figure 6: Age Distribution of the United States in 1970

Figure 6
Age Distribution of the United States in 1970


The peak here (10-14) is the peak during 1955-59 and the boom drops off to the left. Notice the age scale inverts the time scale. Moving up in age means moving back in time.

We know that in another 20 years the boom age groups will have aged 20 years, thus the age distribution in 1990.

Figure 7: Age Distribution of the United States in 1990

Figure 7
Age Distribution of the United States in 1990


Note the great gap between the pre-boom and the boom numbers, roughly 10 versus 20 million. Notice also that the peak boom cohort is still in it's early 30s.

If you can't guess what's coming next before I put up the transparency, I'll never make a demographer of you. We've into the future now, but it's a future (as I've explained) that we know with considerable confidence. It's 2010 and the age distribution will look like this.

Figure 8: Age Distribution of the United States in 2010

Figure 8
Age Distribution of the United States in 2010


We might note in passing that while we've been cruising down this demographic highway the population has slipsided it's way up to 300 million, just double what we began with in 1950. Mortality is beginning to play a role here, flatting the profile of the boomers. Deciding exactly which age groups should be called "boomers" gets a bit dicey here, but however we call it, most of them have yet to reach retirement.

By 2030, however, they are largely into retirement, and mortality has further muted the profile.

Figure 10: Age Distribution of the United States in 2030

Figure 9
Age Distribution of the United States in 2030


The shape of the age distribution has now become more or less what demographer's call stationary. As long as the number of births to the population each year remains roughly constant, the distribution won't change much. A nearly century long transition is essentially complete, and we have massively large number of old persons relative to young persons from what we (or any other society) has experienced in the past.

One important lesson of this description of the US baby boom is to distinguish the dynamics of the baby boom itself, which is a transient (if long term) phenomenon, with the long term dynamics of aging populations. These two phenomena are entirely independent: long term aging would be occurring even if there had been no baby boom-as it is occurring in the little dragons of East Asia, which had no baby boom but are inexorably aging. And we could have baby booms—and to some limited historical extent have had them-without the long term aging trend. Let's not confuse ourselves by mixing up these two phenomena.

Conclusion

I want to conclude with a few comments on the old age support aspects of population aging. First, some numbers that defined the issue in quantitative terms, for Japan, where aging has proceeded most rapidly.

Figure 10: Population Aging in Japan: 1950-2000

Figure 10
Population Aging in Japan: 1950-2000


This figure shows the ratio of persons aged 20-64 ("workers") to persons aged 65+ ("retired persons") for Japan. In 1950 this ratio was 10-- 10 workers for each retired person. By 1995, this had fallen in half to 5 workers for each retired person.

Figure 11: Population Aging in Japan: 2000-2050

Figure 11
Population Aging in Japan: 2000-2050


By 2015--figures for which are for all practical purposes a done deal, even though it is 20 years in the future, its down to about 3.5, and another decade brings it down to 2.

In perspective, a decline from 10 to 2. It might not go so low, if Japanese families have more births than these figures postulate. On the other hand, if Japanese families have even fewer births, the ratios will be lower. The terminal figure of 2 is a good working hypothesis for what the country will be up against.

There isn't time to go into details here, but broadly speaking the trend we see in Japan is being duplicated in other countries. We are in a transition from a world in which the ratio of persons aged 20-64 to persons aged 65+ is ten or more to a world in which is ratio is about two.

We can go into implications in the discussion. I do however want to make a few very brief points on the long term social response to aging before I conclude.

Does the aging of world population pose a problem? Absolutely. Does it pose a major political conflict between generations? Absolutely. But in recognizing this, let's also recognize several very special features of the issue and it's broad sociopolitical context.

(1) The problem is universal and it won't go away except on terms that everyone would find worse that the problem itself. There being no way to explain away the problem, society will eventually-and let us hope sooner rather than later-face it. And the worst enemy in human affairs is often not so much the severity of the problem itself as the refusal to recognize and address it.

(2) This is a very special political conflict for several reasons. First, they are our parents; or they are our children, as the case may be. Not all old people have children, but most of them do, and the bond between parents and children is the strongest of all emotional bonds. That won't solve the problem, heaven knows, but its a factor not present in, for example Bosnia, the Middle East, or Northern Ireland.

(3) Even if you don't buy this argument there's another that you surly will. This is a political conflict in which we all hope to be on both sides. This may not be clear to very young adults, but it becomes clearer to us as we age—doesn't it (!). Nowhere is it truer that "we have met the enemy, and they are us."

(4) Finally, let's point out that a retirement age of 65 was engraved in stone tablets before Moses. Humanity has always known that, once you reach 65, you're home free, right? Your entitled to a comfortable retirement for the rest of your life.

Obviously not. The striking thing about age 65 is how arbitrary it is, when placed in historical context. To say that the aging problem can be solved merely by redefining what we mean by old is obviously an oversimplification. At the same time, however, I suggest to you that it is, at the end of the day, the only way out. It will happen because it must happen, and our job as a society will be to figure out how to make it happen in the best, fairest possible way.