1 Introduction

In contrast to the current enthusiastic embracing of the concept of sustainability and adaptation, the idea of “progress” has, in recent times, fallen out of vogue. Landscape paintings during the heyday of the industrial revolution often featured steamboats belching plumes of coal smoke into the air. Today, postmodernist thought questions whether the seemingly inexorable directions of change in human affairs necessarily point toward improved states of being.

A book of commissioned essays entitled Progress: Geographical Essays (Sack 2002) attempted to assess the concept of progress in various branches of scholarship in geography and, more broadly, the relationships between geography and progress. In the introduction to the book, editor Robert Sack notes: “Contemporary science rarely thinks of nature in terms of design. Nature is constantly changing, but, generally speaking, it has no preferred direction or goal” (p. xi).

In the book’s first essay, Thomas Vale assesses whether ideals of progress exist in physical geographic research. He cites Stephen Jay Gould. Gould contends that the concept of progress (together with adaptationism, determinism, and gradualism) constitutes one of the four biases of modern science. In evolutionary history, Gould has been a proponent for a view of directionless diversification rather than of “linear, directional pathways and development toward ‘higher’ or ‘better’ forms” (Vale, in Sack, p. 2).

Progress: Geographical Essays does not specifically treat population geographic or regional science research or the interconnections of demographic change and the economic issues at the forefront of contemporary regional science research. It seems to me, however, that many of the theories of demography intrinsically embed concepts of progress. And, similarly, much of our thinking about the world’s geography of population takes as largely non-problematic the connections between demographic change and regional economic development trends. There are clear and obvious connections between the basic quantities studied in population geography – that is, demographic indicators – and notions of quality of life (QOL) or human well-being. Indeed, social accounts often directly deploy basic demographic indicators as constituent components of QOL.

Is a concept of progress intrinsically less problematic for population geography and regional science than it is for other branches of the social and natural sciences? A critical theorist’s stance on this question might be that population geographers have been troglodytes for being so accepting of such a directionally rooted paradigm. The counter assertion would be that – for the overwhelming majority of the world’s now more than seven billion people – demographic progress is not merely some ethereal academic construct; it is, rather, a concrete goal toward which groups of human beings aspire, thus inspiring their collective actions. It seems to me that given the normative focus of much regional science – quantitative analysis undertaken to contribute to better economic and social policy – a notion of social welfare, betterment, advancement, improvement, or progress must almost necessarily lie at the heart of our enterprise. And many of the key variables of demography lend themselves, directly or indirectly, to attempts to measure the quality of life.

It is hard to argue that it is not progress when a country’s infant mortality is lowered. Nor would many assert that the long-term trend of increasing average life expectancy at birth is not, on balance, all to the individual and communal good. And what of the strong empirical evidence (provided in case after case) that reductions in a country’s fertility – at least reductions down from levels substantially in excess of the replacement total fertility rate (TFR) of ca. 2.1 babies per woman – have seemingly gone hand-in-hand with increased economic empowerment of women and thence rising material prosperity?

The heritage of transition theories so pervasive in demography is, I think, also reflected in the style and topics of regional science research carried out to date. Not coincidentally, regional science was founded and came of age concurrently with the formulation by demographers of theories of demographic transition. Both were products of the post-World War II period in North America, Europe, and Japan, where economies were rapidly developing concomitant with these regions moving through what has come to be called the third or late-expanding stage of demographic transition. The post-World War II baby booms experienced in these countries generated youthful population age structures that, when coupled with the subsequent shift in preferences for smaller families, provided the labor forces and consumer demand for an unprecedented period of long-term economic expansion.

As the boomer generations of today’s most developed countries (MDCs) reached working ages, the advancement of material prosperity was sustained as their capitalistic economies reaped a so-called demographic dividend. Such a dividend is the possibility for accelerated economic growth that a country may reap as a result of a lagged change in its population age structure after mortality and fertility levels are brought down. With fewer births, a country’s young dependent population for a period of time grows smaller in relation to its working-age population. With fewer children to support, and as women increasingly enter the labor wage economy, a country has a window of opportunity for rapid economic growth, presuming the right social and economic policies are developed and wise investments made (Grimner and Bremner 2012).

The rising tide of material prosperity and rapid technological innovation after World War II drove academic interest in enhancing and developing new quantitative tools for studying and advancing economic and social well-being. Now, as demographic transitions and subsequent demographic dividends are available to other parts of the world, for the MDC countries, a critical focus becomes how to sustain one’s own society’s economic progress while sharing a world in which former economic disparities are much reduced. The regional science focus on normative science is shared with the field of demography, where it developed from a long history of concern with accurate measurement and numerical methods for estimation, projection, and analysis.

Given the nature of this initial impetus for the founding of regional science as an academic endeavor and the inherent interconnectivity of population geography with demography, it should not be surprising that adherents to these two fields have been somewhat resistant to the bufferings of postmodernist thought. Questioning of scientific approaches has come into vogue, in part, because of the changed global economic landscape driven by demographic transition and the spotlight it has put on marginalized populations in an era when returns to the owners of capital have soared. It thus seems critical, for the future of the regional science movement, to focus on understanding how, once most all countries’ demographic transitions are completed, can ongoing enhancement of quality of life be sustained in all regions of the world? At present, it is the formerly less developed countries that are being given the potential to reap their demographic dividends and rise on the development spectrum. Ultimately, however, once all these finite-period population-driven labor effects have been expended, will the large demographic divide characteristic of the demographic transition period of human history become a relic of the past? Or will other comparative advantages lead, again, to a world of “haves” and “have nots”?

While such big-picture economic prognosis lies beyond my expertise or ambitions in this paper, inexorable, big, and fairly predictable demographic trends have been, and will continue to be, underway for some time. Most countries’ populations are aging, and their age compositions shifting. But the demographic divide that has characterized the world over the past century or more ensures it will still be several generations before a convergence of age structures around the world becomes a possible future system state. In any of several possible such future, post-demographic transition worlds, can there be a continuing role for a notion of demographic progress? Can demographic progress be an empirically sustainable paradigm over the long term for scholarship in demography and regional science? Or, like modernity, does demographic progress contain the seeds of its own nihilism? Transition theories imply a state of completion. But as the countries of Europe, Japan, and others are now wondering, what comes afterward? And is, in fact, the direction of demographic evolution the wiping out of economic and demographic differences among rich and poor nations? Or does global capitalism inherently promote the continuance of inequities in nation-scale demographic outcomes as a result of the nature of the transnational competition for access to resources and material prosperity?

In this essay, I won’t provide any firm answers to all these broad questions! Rather, I shall share only some tentative thoughts, buttressed by the results of a simple exploratory empirical exercise. In the third section of the paper, I shall look at the changes within a set of selected indicators for the world’s countries over a two-decade period of time up through the tech boom that ended in the early 2000s. I shall then, more cursorily, look at the subsequent period up to the present. Before getting into that exercise, however, I wish first to briefly survey the progressive elements of our basic transition theories, because such theories form the backdrop against which we shall examine the results of recent empirical evidence on the direction of change in world demographic patterns.

2 The Role of Progress in Theories of the Geography of Population Change

In assessing change, the geographical analyst of population makes frequent use of the so-called components of population change, which are summarized in the basic demographic accounting equation (sometimes called, simply, “the demographic equation”):

$$ \Delta P=B-D+NM. $$

The change in the total population of an area of the earth is the difference between births and deaths (this difference termed also “natural increase” or, more properly these days, “natural change”) plus net migration: the difference between the number of in-migrants and out-migrants for the area. Three major transition theories summarize the long-term processes countries appear to experience in these constituent components of overall population change. The concept of progress is embedded in each of these three theories.

The science of demography began with the systematic study of mortality. The epidemiological transition summarizes changes over time in the societal risk of death (Omran 1971, 1977). The epidemiological transition theory is closely related to the mortality component of the overall demographic transition theory, which summarizes change over time of both death and birth rates, i.e., of a country’s natural increase.Omran (1982) summarizes the theory to be constituted of five propositions. The first is simply the premise that mortality is a fundamental factor in population dynamics. The second, however, asserts that:

…a long-term shift occurs in mortality and disease patterns whereby pandemics of infection are progressively (but not completely) displaced by degenerative and man-made diseases as the leading causes of death. (Omran 1982, p. 172)

The theory typically posits three stages in this progressive process: (1) the age of pestilence and famine, (2) the age of receding pandemics, and (3) the age of degenerative and man-made diseases.

Omran’s third proposition is that, in terms of the enhancement of life spans, the transition tends to “favor” the young over the old and females over males. The fourth proposition asserts that the transitions occurring prior to the twentieth century in the more developed countries:

…have a closer association with rising standards of living and improved nutrition than with medical progress. In contrast, the twentieth-century transitions (i.e., in the less-developed countries) are initiated by medical progress, organized health care, and disease control programs that are usually internationally assisted and financed, and thus largely independent of the socioeconomic level of the country. Further maturation of the transition, however, depends on a beneficial synergy of health care progress and socioeconomic development. (Omran 1982, p. 172)

Note, here, his observation regarding the sustainability of the transition, e.g., of demographic “progress” with respect to the enhancement of life spans.

Finally, Omran’s fifth proposition builds from the above observations to recognize a “basic” or “classical” model (derived from the historical experience of today’s more developed countries) and three variants: (a) an “accelerated” version to account for differences in Japan, Eastern Europe, and the former Soviet Union, (b) a “delayed” one for most Third World countries, and (c) a “transitional variant of the delayed model” to describe the experiences of countries that have recently rapidly developed, such as Taiwan, South Korea, Singapore, Hong Kong, Sri Lanka, Mauritius, Jamaica, and China.

Whereas the epidemiological transition theory does not posit a single pattern of change for all countries to pass through, there is obviously an inherently assumed unidirectionality of change, with tight linkages between mortality, scientific advances, economic development, and forces of globalization.

Notestein (1945) coined the term “the demographic transition” to describe the demographic history of Western Europe. The characteristic drops that occurred first in death rates and then in birth rates can be summarized by a pair of overlain, reversed S-curves. The lag between the time of the declines in mortality and those of fertility, however, resulted in a transition period of rapid population growth.

In addition to the resultant drops in mortality that stem from the epidemiological transition, demographic transition theory is dependent on fertility decline occurring concomitant with “modernization” of a country’s lifestyle. Conventionally, the theory has involved a country passing through four stages: (1) the high stationary, (2) the early expanding, (3) the late expanding, and (4) the low stationary. Explosive population growth is brought about by the divergence of birth and death rates during Stage 2 and the considerable population “momentum” due to young age structures that perpetuates high growth rates even as birth and death rates are converging during Stage 3.

Research has focused on triggers for fertility declines and the relevance of the Western Europe experience to the contemporary settings of lesser developed countries (Teitelbaum 1975). Economic theory has contributed a basic framework to analyze the microfoundations of such declines (e.g., Easterlin 1975), while the roles of cultural factors have been highlighted in empirical studies as extremely significant influences on the onset and spread of the fertility decline (Knobel and van de Walle 1982). Despite a large number of caveats, the basic thrust of demographic transition theory is the irreversibility of family limitation practices and thus fertility decline. Examine, for instance, this summary of the European demographic experience in terms of four main findings:

(1) Fertility declines took place under a wide variety of social, economic, and demographic conditions. (2) The practice of family limitation was largely absent (and probably unknown) among broad segments of the population prior to the decline in fertility, even though a substantial portion of births may have been unwanted. (3) Increases in the practice of family limitation and the decline of marital fertility, once underway, were essentially irreversible processes. (4) Cultural settings influenced the onset and spread of fertility decline independently of socioeconomic conditions. (Knobel and van de Walle 1982, p. 268)

The third and final transition theory concerns the characteristic evolution of migration patterns within countries. It stems from the logical linkages between industrialization, economic development, and urbanization, perhaps first summarized in a coherent way by Ernst Georg Ravenstein in his classic papers on the “laws of migration.” Ravenstein provided a series of cogent observations concerning urbanization and the geography of the flows of human capital underway in Great Britain during the epoch of the industrial revolution (Ravenstein 1885, 1889).

The longer-term trends in the internal migration experiences of countries were summarized by Zelinsky (1971) in his “Hypothesis of the Mobility Transition.” As explained by Weeks (1999):

The demographic transition helped to unleash migration … migration is a ready adaptation that humans (or animals for that matter) can make to the pressure on local resources generated by population increase… Because the demographic transition occurred historically in the context of economic development, which involves the centralization of functions in cities, migrants have been drawn to cities, and urbanization is an important part of the migration transition (p. 238).

As in the other two transition theories, in Zelinsky’s model, a country is posited to undergo a progression of changes in three waves of major movement: (1) rural to rural, which involves the settlement of frontier regions; (2) rural to urban, i.e., urbanization; and (3) urban to urban, which becomes the dominant mode once a fully developed urban system has been achieved.

Shortly after Zelinsky formulated the mobility transition hypothesis, considerable attention came to be focused on a newly postulated fourth mode of movement: urban to rural. Newly extended migration transition theories have more recently been developed with reference to “counter-urbanization” and the changing economic roles of cities at different levels of national hierarchies (e.g., Geyer and Kontuly 1996).

All three transition theories share a focus on the dynamic interrelationships of microscale demographic behavior and macroscale processes of economic systems, which are brought about as a result of accumulating scientific knowledge and the technology changes that it facilitates. In the case of all three theories, the state-of-the-art research focuses on the question: What happens next? All three are essentially post hoc descriptions of events that have occurred in a part of the world being extended to countries in other areas. While historical progress is largely uncontested in these theories, interesting questions arise as to the sustainability of the embedded notions of progress once the transitions, as they were originally conceived, become completed.

Let’s turn, then, to a simple assessment of two recent decades of experience through the lens of demographic progress. The time period chosen for my analysis, 1982–2002, is somewhat arbitrary, just as would any time period for researching a set of continuously evolving time trends (Boulding 1985). The analysis was originally carried out preparatory to my (unpublished) 2003 Pacific Regional Science Conference Organization (PRSCO) Presidential Address and contemporaneous with the appearance of the Sack (2002) book, both of which provided the impetus for my thinking on these matters. The year 2002 is also, however, the most fortuitous one for our purposes here, marking the proximate end of the “tech boom” period in the USA and other highly developed countries. By then, too, the window of opportunity known as the “demographic dividend” had passed to elsewhere in the world. In 2002 the first members of the US baby boom generation (those beginning with the 1946 cohort) were entering their late 50s, and the last members (born in 1964) were into their late 30s and had already been absorbed into the labor force. It could be argued that, by 2002, not only was the demographic transition within MDCs completed, so too were the subsequent demographically driven labor force effects. The period 1982–2002 was equally interesting for our purposes here, when the “sweet spot” in population age structure was starting to be reached by a number of highly populous countries within the middle tier of development status, including the world’s two most populous: China and India.

3 Demographic Status and Progress of Countries Around the World

Truly comparable, exact international population data for any given point in time do not exist. The Population Reference Bureau, in its annual World Population Data Sheets, attempts to provide a set of indicators that are as consistent, as feasible, given the different years when countries enumerate their populations, as well as the variations in basic data definitions and reporting practices employed by countries across the world. These data cannot be used reliably as an annual time series, but they are suitable for longer-term, general comparison purposes, such as here. For the analysis that follows, I used data from Population Reference Bureau (1982, 2002, 2015).

To examine the recent course of demographic change for the world’s countries, I decided to focus on eight key demographic indicators, those shown in Table 1.

Table 1 Demographic indicators used in the study
Full size table

In choosing these, I limited myself to indicators widely available for the approximate end points of the 1982–2002 study period, and I excluded countries that had total 2002 populations of less than one million. Unfortunately, countries that split up during the two decades (the former Soviet Union, Czechoslovakia, and Yugoslavia) also had to be excluded from my database because of the lack of ca. 1982 information for the new, smaller-bounded units.Footnote 1 Conversely, I was able to include countries that had unified by 2002 (namely, East and West Germany and North and South Yemen). I used total populations of the formerly separated countries to combine, on a pro rata basis, their 1982 demographic indicators.

3.1 Mortality and Longevity

Let us begin our examination of demographic progress around the world with mortality and longevity statistics. The simplest of these is the crude death rate (CDR), which is simply the number of deaths recorded in a population per 1000 people alive at the midpoint of the period. The original demographic transition theory was framed in terms of the CDR and the similarly defined crude birth rate (CBR) statistic.

While reasonably useful for studying trends over time, the CDR can be a misleading indicator to deploy for intercountry comparisons. It is strongly affected by age composition. Mexico, for example, with its still child-dominated age structure, currently has a lower CDR than does the USA even though Mexicans, at all ages, have higher age-specific probabilities of dying. A composite indicator constructed from age-specific death rates would be preferable for such comparisons, but age-specific information is not available for all countries.

The infant mortality rate (IMR) is an age-specific death statistic that is almost universally available and consistently reported. Surely demographic progress is evidenced when a country achieves greater success in preventing unnecessary deaths of its babies! The IMR is a telling indicator of social welfare; it is sometimes used as a proxy for both the level and degree of diffusion throughout a country of quality medical care, nutrition, and sanitation. With just 2.2 infant deaths per 1000 live births, Singapore registered the world’s lowest infant mortality in 2002; it was followed by Hong Kong, Japan, Sweden, Finland, and Norway – all with rates below 4.0. In 2002 the countries with the world’s highest infant mortality were Sierra Leone and Afghanistan, where rates exceeded 150.

As can be seen in Table 2, substantial progress was, in fact, being made around the world in the likelihood of infants living to their first birthdays. Of the 130 countries in the data set, 124 achieved lower IMRs by 2002! Offsetting this encouraging news, however, was that fact that, on average, the percentage reductions were greater in the more developed than the less developed countries (LDCs), despite the much higher 1982 rates of LDCs than MDCs.

Table 2 Change in the infant mortality rate for world countries, 1982–2002
Full size table

Among the world’s dozen most populous countries (a list that excludes, for data limitation reasons mentioned, Russia), two of the world’s most highly developed countries, Japan and Germany, and two currently at intermediate levels of development, Brazil and Mexico, led the pack of those achieving the greatest percentage reductions. The USA trailed most MDCs on this measure due to its high degree of inequality of access to high-quality medical care.

In terms of infant mortality, the reductions achieved – even in those countries that already had the lowest rates two decades earlier – suggest that demographic progress, as measured by this indicator, remained an attainable and worthy goal for most countries in the near-term future. Continued medical advances, better pre- and postnatal care, and improved nutrition and sanitation all increase the percentages of infant deaths that are considered preventable. But the 1982–2002 analysis clearly shows that the most potential for sustained demographic progress is to further the diffusion around the globe of “best practices.”

The experience since 2002 is that sizable further progress has, indeed, been made. From 2002 to 2015, the reductions in infant mortality did not simply continue apace, but actually sped up! The worldwide IMR dropped from 54 to 37 infant deaths per 1000, which is an annualized decrease of 2.4 percentage points versus the 1.8 percentage point per year improvement registered over the 1982–2002 period. By 2015, the rate across the MDCs had been brought down to 5 and that for all LDCs combined to 40.

Almost as non-problematic as not wanting babies to die is the nearly universal interest in achieving greater human longevity. Another telling indicator with respect to many of the same factors as infant mortality (nutrition, health care, sanitation, and so forth) is average life expectancy at birth. In 2002, Japan (at 81 years) ranked #1, and Australia, Italy, Sweden, and Switzerland (tied at 80 years) ranked #2 among all world countries for longevity of their citizens. At only 37 years, Zambia’s people had the shortest life expectancy.

Table 3 summarizes the 1982–2002 progress in extending life spans around the world. The greatest percentage gain was in Oman, and, among the most populous countries, truly spectacular increases took place in Indonesia, India, Bangladesh, and Pakistan. In the span of less than a single generation in these four countries, which together accounted for almost exactly a quarter of the world’s population, average expected years of life were extended by more than 20 %!

Table 3 Change in life expectancy at birth for world countries, 1982–2002
Full size table

In a large part of the world, however, the AIDS epidemic was substantially setting back progress that otherwise should have occurred. Life expectancy at birth went down from 1982 to 2002 in 15 African countries and in Haiti. In 11 of the 15 African cases, life spans were reduced by more than 15 %, and they went down by more than 20 % in Zimbabwe and Zambia.

As with infant mortality, the period since 2002 has been one of remarkable, continued improvement. Worldwide average life expectancy at birth continued to rise, reaching 71 years in 2015. People born in 2015 in MDCs could expect to live 3 years longer than in 2002, while those in LDCs gained, on average, four more years of expected longevity. The differential between those living in less and more developed parts of the world has also substantially narrowed. Back in 1982, those born in LDCs could expect to live only 79 % as long as those born in MDCs. By 2015, however, life expectancy in LDCs was 87 % of that found in MDCs.

Most of the world’s countries have by now passed through Stage 2 of the demographic transition, which is the period when the most rapid drops in mortality take place. Whereas we may be bumping into physiological constraints on longevity in MDC contexts, as with infant mortality, considerable room exists for extending the life spans of the majority of the world’s people.

3.2 Fertility

Turning from mortality to fertility, the concept of demographic progress becomes substantially more complex. During the last two-decade 1982–2002 study period, a downward trend in childbearing propensity was taking place in most every country of the world regardless of preexisting typical family sizes. From an environmental sustainability perspective, for the long term, a lower rate of planetary population growth is most welcome news. In the shorter term, however, the rise in material prosperity for the majority of the world’s people that accompanies family size reductions poses challenges.

Table 4 shows the situation with respect to total fertility rates, which are the average numbers of children to be born per woman if age-specific period fertility rates were to apply throughout the childbearing years. The TFR of the world as a whole dropped dramatically in the 20 years, 1982–2002, from almost four children per woman down to fewer than three. Most of the more developed countries by 2002 had rates below the long-term “replacement level” of 2.1, whereas the typical rate for less developed countries was close to 3.

Table 4 Change in the total fertility rate for world countries, 1982–2002
Full size table

The 1982–2002 period was a significant one for many countries on the path toward completing their demographic transitions. Among the 130 countries in the data base, fertility rates dropped over those 20 years in 110 cases; rises were experienced in only 18 countries. Among the world’s most populous countries, the USA was the only one that registered a rising total fertility rate. This was mostly attributable to high levels of immigration, with the US foreign-born having significantly higher fertility than the native-born.

So how to define demographic progress with respect to changing fertility levels? If progress is a march toward “the greatest good for the greatest numbers,” does increasing the world’s population through natural increase represent progress? Or, as has more commonly been the prevailing view during modern times, does progress lie in achieving the greatest gains in quality of life for the planet’s residents? Maximizing the increase of material prosperity and thence QOL seems to be achieved when a country brings its population growth rate down to a level where economic expansion is not stifled by excessive dependency, when the productive capacity of its women is harnessed in the labor force rather than exclusively in bearing and caring for youths. Children are, of course, the labor force populations of the future, but, equally, an overly high proportion of children in a country’s population can stifle development.

It is during Stage 3 of the demographic transition when birth rates lower. Stage 4 has conventionally been defined as one when a new equilibrium is reached and maintained. Births and deaths were foreseen to come into a new balance at levels far below those of Stage 1, which pertains to most of human history, when unchecked mortality had kept natural increase to essentially zero, despite fertility levels close to the upper limits of fecundity. Stage 4 was similarly posited to be associated with no natural increase and was thus dubbed the low stationary stage. However, a remarkable aspect of our 1982–2002 study period was that in most all of the world’s most developed countries, TFRs plummeted to well below the replacement level of 2.1, raising cause for alarm. In the long run, a total fertility rate below 2.1 is non-sustainable: the country will become depopulated, or, if experienced for the world as a whole, our species would ultimately go extinct. Both high and low TFRs, benchmarked around the 2.1 “replacement level,” can be considered unsustainable over the long course.

So how, then, do we apply the concept of sustainable progress to fertility trends? Is progress not then simply measurable by unidirectional reductions from former levels but, rather, by a narrowing of the difference in either direction between a country’s current level and that magic number of 2.1? Or would the inhabitants of some parts of the earth, and the planet itself, perhaps be better off with either more or fewer inhabitants than implied by a TFR of 2.1 and zero natural increase? This is an interesting unresolved question at this juncture of human history. On some parts of the planet, birth limitation is pursued as a primary societal goal, while, elsewhere, pro-natal measures enjoy strong support to ensure the sustainability of national cultures and economies.

3.3 Age Dependency

Age dependency ratios are interesting measures to examine in the context of assessing potentials for future economic development and for enhanced or sustained quality of life. Tables 5 and 6 present the data for the 1982–2002 period changes in youth and elderly dependency ratios.

Table 5 Change in the youth dependency ratio for world countries, 1982–2002
Full size table
Table 6 Change in the elderly dependency ratio for world countries, 1982–2002
Full size table

As fertility rates have dropped, the world’s youth dependency ratio (YDR) has dropped also. As shown in Table 5, the decrease has been the greatest overall in LDCs rather than MDCs. Japan at just over 20 % had the world’s lowest ratio in 2002, whereas YDRs for Uganda, Niger, and Burkina Faso exceeded 100 (meaning there were more children in the 15 1-year age groups of 0–14 than there were adults in the 40 years from age 15 to 64, traditionally considered to be those of labor force participation).

Since 2002, youth dependency worldwide has continued to plummet. By 2015 it stood at 39.4, less than two thirds of its 1982 level. How should we measure progress with respect to changing youth dependency ratios? The name itself suggests that reductions might normally be desirable to achieve future greater per capita prosperity. But are YDRs in the low 20s, such as now found in Japan, Italy, Spain, Greece, Portugal, and Germany, a harbinger of future declines in material well-being and decreased QOL for those countries’ citizens? Is there an optimal age structure and thus a target YDR for a country to seek?

The question of a sustainable family size to sustain a vigorous labor force is even more complicated by the fact that reductions in fertility and increased life expectancy are now beginning to contribute to rising elderly dependency ratios (EDRs) around the world. Elderly dependency is of concern most especially in those MDCs where reduced levels of childbearing have been the norm for the longest times.

In 2002, Italy and Japan had the highest EDRs in the World. As shown in Table 6, Japan’s percentage increase was almost 100 % over just the two decades since 1982. By 2015 Japan’s EDR was twice its YDR: for every 100 people aged 15–64, there were 42.6 who were 65 or over.

In the majority of less developed countries, rising EDRs probably do not pose much of an obstacle for these countries’ near-term aspirations for improving well-being. (China, now on the cusp of entering the ranks of the MDCs, is a major exception.) Rather they reflect the positive aspect of enhanced life expectancy.

3.4 Female Labor Force Participation

In addition to age composition changes, the other major controlling factor for many countries’ changing productivity situations has been rising female labor force participation. Table 7 summarizes these trends over the primary study period.

Table 7 Change in the female force participation rate for world countries, 1982–2002
Full size table

Worldwide, female labor force participation increased moderately between 1082 and 2002, going from 57 to 61 % of women of labor force age. What’s more, the rate of increase was almost twice as great in MDCs than in LDCs.

The countries with the largest percentage additions of women to their workforces were certain rapidly developing countries where participation in 1982 had been at considerably below the world average. For example, among the most populous countries, Mexico, Pakistan, and Brazil posted the biggest percentage increases. Mexico, however, remained in the bottom third of all countries in terms of its actual rate.

Among MDCs, the female participation increase in the USA is notable, going from one percentage point below the MDC average in 1982 to a full five points above it by 2002. Nonetheless, among MDCs, by 2002 the USA still trailed Canada (72 %) and the Scandinavian countries (Finland, 73; Norway, 74; Denmark, 77; and Sweden, 82).

If gender equity in the workforce is taken as a reflection of social progress rather than economic necessity, the last two generations of working-age women in some countries have made progress by taking wage jobs in greater proportions than their mothers. But, in many parts of the world, there remains the potential for many more women to pursue workforce careers, and, for those employed, there is certainly a need to achieve better rewards and conditions for their wage labor.

3.5 Migration and Urbanization

Let’s turn now to the migration component of population change. Extraordinarily high rates of urbanization have been the norm in many LDCs. More developed countries, on the other hand, seem to have entered a period of oscillating waves of movement up their urban hierarchies and of counter-urbanization down them (see, for instance, Kontuly 1988; Geyer and Kontuly 1996; Plane et al. 2005). Considerable regional science research has been conducted since the first “turnaround” phenomena were noted in the 1970s (e.g., Vining and Kontuly 1978; Vining et al. 1981).

Care should be taken in interpreting data for the proportions of countries’ populations who inhabit urban and rural areas, because statistical practices differ greatly around the world with regard to both minimum threshold population sizes and how urban areas are bounded. Table 8 strongly illustrates, however, the impressive magnitude of the increases over the two decades in the study period in LDCs. Among the world’s largest countries, the most populous of all – China – has had the greatest relative urbanization. Note, however, that back in 2002 it was still lagging by two percentage points the average level for LDCs in general. By 2015, however, China’s 55 % urban percentage exceeded both the world and LDC averages (which, by then, stood at 53 % and 48 %, respectively). Bangladesh, Indonesia, and Nigeria have also experienced massive rural-to-urban internal population redistribution, which has resulted in the hyper growth rates of their primate megacities.

Table 8 Change in the urban percentage of the population for world countries, 1982–2002
Full size table

Does the urbanization being experienced in most LDCs, or the counter-urbanization sometimes now found in MDCs, represent demographic progress? Or do both of these trends? For the most part, internal or domestic migration, more than other demographic phenomena, represents the freewill decision-making behavior of individuals who “vote with their feet.” The fact that so many people have been on the move within countries around the world might be interpreted as yet another sign of demographic progress and of market forces to establish altered settlement patterns in response to changed economic realities.

The last demographic indicator that I had wished to report was one for foreign immigration. Unfortunately, international migration statistics are extremely problematic. I report in Table 9 selected summary statistics derived from estimates given in the 2002 World Factbook, a website of statistical data maintained by the US Central Intelligence Agency (CIA).

Table 9 Rates of net immigration compared to rates of natural increase for world countries, 2002
Full size table

Statistics on net foreign immigration do, I think, help round out the demographic profiles of the world’s countries. For most countries, foreign immigration is negligible in comparison to natural increase. The CIA reports essentially zero net immigration for 35 of the countries on my database, net immigration for 37, and net emigration for 56. Just as in the case of internal migration, there appears to be a tendency for the destinations of in-migration to be more spatially focused than the source areas of out-migration. Only a dozen countries have annual net emigration at or exceeding 2 per 1000, whereas there are half again as many immigrant receptor countries with rates at or above 2 per 1000.

For the world’s most populous countries, note that for the emigrant senders, Mexico, Pakistan, Bangladesh, China, and Indonesia, the 2002 rates of outflux were relatively insignificant for balancing these countries’ still fairly high rates of natural increase. Although not very effectively staunching population growth, in terms of the exportation of human capital, however, positively selective out-migration may pose policy concern. On the other hand, these are countries that still have momentum for growth due to their young age structures. Emigration, which is highly focused on the young labor force ages, can provide some relief from labor market entrants swamping the ability of their expanding economies to provide new jobs.

Note that for the two immigrant-receiving countries on the most populous list, Germany and the USA, immigration in 2002 was an extremely important constituent component of population change. By 2015, US net immigration was substantially down from the levels recorded over the 1982–2002 study period. Germany, however, is now grappling with trying to accommodate a massive influx of refugees. In Germany’s case, immigration offsets population loss that would otherwise occur from natural decrease, whereas for the USA, even during the high immigration period of 1982–2002, it represented no more than half the level of positive natural increase.

In contrast to Germany, Japan has, to date, been less sure about using immigration to counterbalance its natural decrease, but it will be interesting to see how that dynamic plays out over the next couple of decades. One observation I would hazard about the future of international migration is that I would expect emigration and immigration to play an increasingly important role as each country tries to define the rather complex equation of what constitutes its own version of ongoing and sustainable demographic progress.

4 Composite Indexes and World Demographic Country Groupings

We have now examined the values and rates of change of eight demographic indicators for the world’s countries. What are the general conclusions to be drawn from all this information? As has been seen, the picture is very different around the globe during this stage of world history when the majority of developing countries are well along the path to completing their demographic transitions, whereas the most developed countries, having completed there’s, are starting to come to terms with the new realities of actual or pending natural decrease.

In an attempt to group countries and summarize the geographic dimensions of the aforementioned indexes of demographic status and their trends, I submitted my variables to two different factor analyses. In both cases, I calculated factor scores and examined how world countries placed on those. I then cluster-analyzed both the standardized values of the original data and the factor scores to study how the world’s countries group together demographically. Which countries exhibit the most similar overall levels of demographic status and which seem to be following the most similar trends?Footnote 2

My first factor analysis was carried out using the eight demographic indicators measured at their year 2002 values. Principle components analysis was used as the extraction method. Based on an examination of eigen values and a scree plot, I determined that three factors, accounting for 89 % of overall variance, should be extractedFootnote 3; these were then submitted to a Varimax rotation. The rotated factor loadings are presented in Table 10.

Table 10 Varimax rotated factor loading matrix derived from 8 demographic indicators for 130 world countries in 2002
Full size table

As can be seen, all but two of the variables load on the first factor, which, based on the signs for the loadings of the specific variables, I have termed “underdevelopment.” High positive scores on this factor are associated with high youth dependency ratios, high total fertility rates, and high infant mortality rates, as well as with low levels of life expectancy at birth, low elderly dependency ratios, and low urban population shares.

The single variable with a primary loading on Factor 2 is the female labor force participation rate. The positive sign indicates that countries with high positive factor scores are those with high percentages of working-age women in the labor force. The extraction of a unique factor suggests that cultural norms can trump developmental levels in arraying countries along the dimension of female labor force participation.

Factor 3 contains the loading for the remaining variable that is not strongly correlated with the others: net foreign immigration.

Country scores on Factors 2 and 3 pretty much mirror the single variable that loads on the factors, and the patterns of those indicators were already discussed in Sect. 3 of the chapter. The scores on the “underdevelopment” factor are, however, quite revealing. They provide, in essence, a demographic spectrum of world countries, showing where they stood in 2002 with respect to the primary recent demographic trends in mortality, fertility, age structure, and urbanization.

Table 11 gives the factor scores for countries at the opposite poles of this spectrum. Included are all those having factor scores, whether positive or negative, of greater than one standard deviation.

Table 11 Countries with the highest positive and highest negative factor scoresa on the composite demographic index: “underdevelopment”
Full size table

In addition to examining factor scores, I experimented with a variety of clustering methods to explore how the world’s countries split out into demographic blocs. I tried, first, a hierarchical clustering of the z-scores of the original eight demographic indicator variables. Because of missing data for two countries, the sample size is 128.

The first countries to combine in this clustering are Spain and Greece. At the next step, the USA joins together with New Zealand. Shortly afterward (at the 119-cluster stage) Australia joins New Zealand and the USA. This is in fact the only three-country cluster to emerge throughout the early stages of the clustering. Still fairly early in the process (at 102 clusters), Canada also enters this same Anglophone cluster. Other countries paired up in the early steps (125–120 clusters, respectively) are Angola⇔Mali, Honduras⇔Guatemala, Austria⇔Netherlands, Thailand⇔China, Norway⇔Denmark, and Colombia⇔Turkey. Although it would seem that cultural and geographic proximity leads to similar demographic profiles, similar economic development levels can, as in the case of Colombia and Turkey, outweigh physical or cultural proximity.

Continuing with the 118- to 115-cluster steps, the pairings are Senegal⇔Mauri- tania, Sri Lanka⇔Mauritius, Hungary⇔Albania, but then France⇔Japan. By the way, the MDC cluster containing France does not join the one with the USA until the world is divided into only 41 total clusters.

The latter stages of this clustering are not all that interesting, in that the world still shows the legacy of the demographic divide, with essentially two large and largely expected groupings. The more developed countries are in one cluster, and all the others (except three outliersFootnote 4) are in an even larger LDC grouping. Because six out of the eight original variables are all correlated with economic development levels, it should not be surprising that the endgame of a hierarchical clustering produces blocs of countries at similar stages of development.

An alternative to clustering the original variables in a data set is to cluster factor scores. In the present case, this will give greater weight to the cultural differences underlying female labor force participation and to foreign immigration as sorting variables relative to the other six indicators (assuming one chooses not to weight the factor scores by, say, eigenvalues before submitting them to the clustering procedure). When the scores on the three-factor scores were clustered, however, the results were not radically different from those based on the z-scores of the original variables.

To experiment further, I tried using a K-means procedure to obtain separate optimal groupings of world countries for successively larger numbers of clusters. In carrying out clustering in that fashion, different blocs of countries are allowed to emerge depending on how many divisions of the world are specified. In a hierarchical clustering, once two countries become joined, they are never separated. That is not the case when optimal clusters are found at each step.

The two-cluster solution is not a very interesting one since only three outlier countriesFootnote 5 are split off from all others. In the three-cluster case, most of the world is bifurcated into a (49-country) least developed country group and an (77-country) intermediate to more developed one, with the third cluster containing just two outliers.Footnote 6 At the four-cluster stage, medium developed and more developed countries form into separate clusters, and three similar large groupings are the pervasive feature of the optimal five-, six-, and seven-cluster groupings (which differ, largely, in the composition of “breakaway” clusters that contain just a few outlier countries in each). The eight- and nine-cluster groupings feature four main blocs of countries. Space limitations do not permit showing the results of each of these optimal clusterings.

Figure 1 shows the ten-cluster solution. Note the separate group of 7 Middle Eastern countries (not including Iran), along with the 12 member group that includes the world’s remaining Communist countries (China, Vietnam, North Korea, Mongolia, and Cuba) clustered with a number of members of the former Soviet bloc, plus Thailand, Albania, Portugal, and Jamaica.

Fig. 1
figure 1

A ten-cluster optimal demographic groupings of world countries based on the three-factor classification

Full size image

To study how the changes in the levels of the demographic indicators further help to classify countries, I carried out a second factor analysis using a larger, 15-variable data set that included the 8 demographic indicators at their 2002 values and the 7 change variables for 1982–2002 time period. Again, I used principle components analysis. This time, examination of the eigenvalues led me to extract seven factors,Footnote 7 which together accounted for 90.2 % of the variance in the data. These were then submitted to a varimax rotation, producing the factor loadings displayed in Table 12.

Table 12 Varimax rotated factor loading matrix based on eight demographic indicators (2002) and seven change variables (1982–2002)
Full size table

Like in the case of the 2002 indicators, the first factor in this analysis encompassing both 2002 status and 1982–2002 change variables is a demographically based composite index of “underdevelopment.” Again loading highly and positively on this factor are the youth dependency ratio, the total fertility rate, and the infant mortality rate, whereas the elderly dependency ratio, the life expectancy at birth, and the urban population share load with high, negative values. In addition, two of the change variables have significant secondary loadings on this first rotated factor. Countries with positive scores on Factor 1 tend to have either increases or only small decreases in their infant mortality rates and youth dependency ratios: these are the world’s least developed countries.

The second factor is again associated with those cultural differences that result in differential rates of female labor force participation for countries at equivalent levels of development. Countries with high positive scores on this factor are those that have generally high percentages of women in the work force. Additionally, however, the large negative loading on the change variable for female LFPR indicates that countries with high scores on this factor also have low rates of increase in participation; this is likely because in many cases their rates as of 1982 were already relatively high.

I term Factor 3 “younger age structure” because it involves the change variable for the total fertility rate. Countries having positive scores on this factor are those that actually experienced rising rates of childbearing or ones that failed to reduce their TFRs by very much. Such countries are found both among LDCs and MDCs. Logically enough, this factor also receives the primary (and positive) loading for the variable measuring changes in youth dependency ratios. Further, note that it contains a secondary loading for the TFR variable itself. Most of the countries experiencing big drops in childbearing are LDCs that remained at above-average TFR levels in 2002.Footnote 8

I label Factor 4 “improved survivorship,” because it involves the change variables for the two life span indicators; positive scores on Factor 4 are associated with increases in life expectancy at birth and reductions in the infant mortality rate. Factors 5, 6, and 7 represent “immigration,” “aging,” and “urbanization” influences. These are unique factors that take the primary loadings for, respectively, the net foreign immigration rate, change in the elderly dependency ratio, and change in the urban population share.

5 Conclusions

Progress and sustainability are, I think, very interesting twinned concepts to explore in the context of demographic change. Since the period of the Enlightenment, science has been a paradigm inextricably bound together with notions of progress. Demographic indicators have been used as benchmarks for measuring the efficacy of technological advances in such areas as food security, nutrition, sanitation, and medicine. The demographic transition theories suggest a rather different course of evolution than visualized by Malthus or the Club of Rome. With respect to sustainability, demographic transition theory can be seen as a systemic equilibrium perspective providing countering evidence to the inevitable non-sustainability of population growth.

In a “radical non-Malthusian perspective,” Kleinman (1980) argued that Hardin’s “The Tragedy of the Commons” represents a fundamental misreading of history, and there is not an “inexorable progression to ruin as a result of uncontrolled breeding.” Rather, he ascribes much of the concern with population growth to the fact that it “appears to threaten those who are well off, perhaps because change may be required to accommodate such growth” (p. 266). His conclusion is that:

The welfare of future generations is not likely to depend on the numbers of people. Population, resources and technology maintain a dynamic balance as they change, moved by and compensating for socio-structural stress. The welfare of future generations will depend on a different legacy, the character of the institutional structures and the knowledge developed by preceding generations. Human societies have, in general, been fairly proficient in resolving technological problems: at least they have been more proficient in this area than in others. They have been least proficient in coping with problems of power and inequity. The legacy that would most benefit future generations would be resolution of the contradictions in social and economic organization which make ‘social justice’ sound utopian if not platitudinous while peace, equity, and security are made to appear incompatible with enterprise, productivity, and freedom (p. 273).

My simplistic exploratory data exercise has suggested that, during the current stage of demographic history, the world is still very much divided between two blocs. The first includes those countries that have “completed” the transition and are now facing a somewhat uncertain demographic future with fertility at unprecedented and unsustainably low levels; the second encompasses the majority of countries where demographic progress – as conventionally conceived – seems to be proceeding apace, if not at wholly unprecedented rates.

The population and development challenges of the future will be very much influenced by how societies around the world puzzle out how to structure their geographic interactions with one another. Communications, trade, and immigration should be front-burner issues as the globalization of the world economy and LDC fertility reductions continue to evolve.

Regional science, with its legacy of holistic, systems thinking, and big-picture modeling, with its focus on the connectivity of economic, social, and environmental systems, and with its touchstone concept – the region – is ideally positioned to play a central role in the critical policy debates of the future. To inform future research in regional science during the period while the vast majority of countries complete their demographic transitions and move into new dynamic trajectories with respect to population growth or decline and age composition change, I believe there will be a need to reconceptualize what is meant by demographic progress. It may well need to become more nuanced and more contingent on specific circumstances. Although more problematic to define, the concept of progress should continue to be of central import.

In the decades ahead, humans will be grappling to build new forms of economic and institutional entities to deal with the vastly greater interconnectivity of all parts of the planet as well as with ongoing, rapid technological changes. Future generations will need to harmonize their affairs and adapt to socioeconomic systems on a planet now embarked on a radically altered demographic course. Compared to other parts of the nexus of relevant economic and social trends, population trends derive in large measure from the inexorable aging of each member of our species. Population trends, therefore, are relatively predictable. As drivers of economic models, demographic measures and trends can and should inform a policy-relevant agenda for regional science research. A concept of demographic progress could be critically important in searching to find the creativity and policy means necessary to ensure sustainable human progress.