Although it is often claimed that 90 percent of the people of Canada live within 100 miles of the U.S. border, this assertion is of dubious validity. Reputable sources put the figure closer to 80 percent. Statistics Canada more convincingly claims that 66 percent of Canadians live within 100 kilometers (62 miles) of the border, based on the 2016 census. An excellent population-density map, derived from 2006 census data, depicts this “100 kilometer” limit, which passes through Winnipeg and Quebec City.
Canada Population Density Map
But if Canada’s population is not as clustered along the U.S. border as many people think, the lack of habitation across the northern three-quarters of the county is still noteworthy. Significantly, the zone of extremely low population density – defined on the map posted above as fewer than 0.4 persons per square kilometer – extends to the country’s southern limit in central and western Ontario. Demographically speaking, Canada is thus divided into western and eastern segments. As a result, population-based cartograms of Canada have a strange if not disconcerting shape, depicting the country as two imbalanced lobes. This demographic division is of some relevance to the separatist movements of Alberta and Saskatchewan.
Canada Population Cartogram
Canada Population-Electoral cartogram
As a cartographic experiment, I took the Wikipedia map of the results of Canada’s 2025 election and eliminated most areas of extremely low population density. This electoral map, like most others, exaggerates support for the Conservative Party and Bloc Quebecois, as these two parties find much of their support in agricultural areas of relatively low-density population density.
2025 Electoral Map of the More Densely Settled Areas of Canada
Even though Canada’s population is concentrated in the far south, density of settlement increases along most of the border as one heads north out of the United States. Consider, for example, the map detail posted below (taken from the second map posted below). From “Point A” in sparsely settled northern New York, for example, a trip into Canada would quickly take one into the densely inhabited Montréal metro area, with more than four million inhabitants. A trip south, in contrast, would take one into the sparsely settled Adirondak Mountains. From “Point B” in the wilds of northern Maine, a trip to the north, west, east, southwest, or southeast would take one into more densely inhabited areas in Canada. One would have to travel considerably farther to the south to reach areas of comparable population density in the United States.
Population Density of Northern New England and Maritime Canada map
U.S. and Canada population density map
Canada’s trans-border demographic advantage is also pronounced in the northern Great Plains, as is depicted in the “five-state, three-province” map posted below. Alberta alone has a larger population than Montana, Wyoming, North Dakota, South Dakota, and Nebraska combined. Canada’s Prairie provinces are also more urbanized than the northern Plains states of the United States. Calgary, Edmonton, and even Winnipeg far exceed any cities of the Dakotas, Montana, and Wyoming. It is also notable that the largest cities on the U.S. side of this trans-national region are on or near its eastern flank, unlike those of Canada.
Demographic Patterns of the Prairie Provinces of Canada and the Northern Great Plains in the U.S. map
In southwestern Saskatchewan and southeastern Alberta, rural population density increases markedly a few hundred kilometers north of the U.S. border. It does so largely because soil fertility and moisture also increase in this area, allowing more productive farming. Here one finds – or found, before agricultural settlement – the Aspen Parkland, a transitional biome between the grasslands of the Great Plains and the northern boreal forest (taiga). This agriculturally favorable zone terminates abruptly in north-central Alberta and Saskatchewan, as is reflected on the population density map. Farming is not feasible across most of the northern halves of the Prairie Provinces not because of climate but rather because of soil. But to the northwest of the belt of continuous agricultural settlement lies a secondary zone of fertile Aspen Parkland in the so-called Peace River Country, which is home to roughly a quarter-million persons.
Aspen Parklands and Population Density in Alberta and Saskatchewan map
The main body of the Peace River Country forms a relatively compact zone of settlement, but far to the north lies a smaller farming region, located to the east and southeast of the town of High Level. Approximately 20,000 persons live here, cultivating some 350,000 acres of rich farmland 455 miles (733 kilometers) north of Edmonton. This northernmost extension of the Peace River Country was first settled by farmers after World War II, perhaps making it the final agricultural frontier of North America. It also has two large oil and gas fields and two lumber mills.
Satellite Map of the High Level Region of Northern Alberta
I would like to end this exploration of global demography by posting two paired maps, which show the global Total Fertility Rates (TFR) in 1950 and 2023. The contrasts are interesting and instructive. Note that the color schemes are similar but the scales used are quite different. For 1950, the relatively low-fertility categories depicted in blue begins at a TFR range of 3.5-3.9; the same category in 2023 is depicted in light brown in the relatively high fertility grouping. In 1950, the highest fertility range extended across North Africa and the Middle East and included several Latin American countries. At that time, much of west-central Africa had somewhat lower fertility levels, largely because the burden of disease reduced fecundity. On the 1950 map, Argentina and especially Uruguay stood out in Latin America for their relatively low fertility. In Europe, Kosovo, Bosnia, Northern Macedonia,* and Albania, all of which have large Muslim populations, stood out for their high fertility. Needless to say, all of these patterns have disappeared from the map.
Global Fertility 1950 & 2023 Compared Map
* Kosovo, Bosnia, Northern Macedonia were at the time parts of Yugoslavia.
Two additional lectures on global demography have been uploaded on the GeoCurrents YouTube channel. The first covers the period from 700 to 1500, focusing on the Black Death of the mid-fourteenth century and its repercussions. The second lecture covers the period from 1500 to 1700. The first part of this lecture focuses on the depopulation of the Americas that occurred after 1500, due mainly to the introduction of infectious diseases from the Eastern Hemisphere. The second part of the lecture mostly covers the reduction of population that occurred in Europe and China in the mid-seventeenth century. This demographic downturn is generally linked to political instability and the effects of the little ice age.
Two new lectures in the series on global demography have been uploaded on the GeoCurrents YouTube channel. The first, number 5 in the series, looks at the ancient world, and the second (#6) looks at late antiquity and the early medieval period (roughly 400-750 CE). The second of these lectures focuses on the Mediterranean world, northern India, and China. These lectures can be found here and here.
In a recent lecture on declining birthrates, I mentioned pronatalist worries about a corresponding decline in our ability to solve problems and address crises, due mainly to the aging of the population. Creative thinking, according to this argument, is primarily an attribute of youth. Most environmentalists find this thesis unconvincing if not absurd. Creativity, they contend, is not limited to youth, and that even if it were, there will still be enough young creative minds even in countries with the lowest birthrates.
The idea that an aging society might be hobbled by a decline of creativity and vigor is not a fringe position, nor is it restricted to conservative pronatalists. Consider, for example, this recent passage from liberal New York Times columnist Nicholas Kristof:
In an aging and perhaps enfeebled world, Africa will also be a continent of youth — arguably making it comparatively vigorous and more of a hotbed for entrepreneurship and for music and popular culture. In a sign of increasing cultural influence, Africans in recent years have won both the Booker Prize and the Nobel Prize in Literature.
The scientific literature on age and creativity is large but no definitive answers have been reached. Some evidence suggests that creativity tends to decline with age in some domains but not in others. In fields in which raw mental power, along with the ability to think outside of established frameworks, are essential, breakthroughs often come early. Mathematics and physics are prime examples: Issac Newton developed calculus in his early and mid-twenties; Alfred Einstein was 26 when he wrote four groundbreaking papers, one of which established special relativity. In contrast, in fields in which the accumulation of knowledge and the honing of judgement are crucial, such as history and literature, age can be an advantage. This distinction is sometimes framed in terms of the difference between “fluid intelligence” and “crystalized intelligence.”
Songwriting is an intriguing area for musing on this question, as it combines the distinctive domains of musical composition and poetry. Poetry might seem to be good example of a field in which experience can be beneficial, but a vast trove of acclaimed poetry has been composed by the young. (John Keats died at 26, Percy Shelley at 29, and even Lord Byron made it only to 36.) Musical composition, in contrast, might seem favor the young, but some of the world’s most celebrated composers retained their powers into advanced age, including Johann Sebastian Bach, Ludwig van Beethoven, and Giuseppe Verdi. All told, one might expect some decline in creativity among songwriters. But – yet again – examples to the contrary are not hard to find. Leonard Cohen recorded his most beloved song, “Halleluiah,” in his late 40s, and his three final and widely celebrated albums were produced when he was in his late seventies and early eighties.
Yet when I casually consider the works of songwriters who came to fame in the 1960s and ‘70s and remained active for decades, most seem to have experienced some dimming of their creative spark as they aged. At 82, Paul McCartney continues to produce quality work, and he recently announced that he been “working on a lot of songs” that he hopes to finish and record in 2025. But I doubt that many critics would claim that his post-Beatles output, produced over 54 years, shows more creativity than the songs that he recorded with the Beatles over an eight-year period. (McCartney did, of course, share songwriting credit with the equally brilliant John Lennon in his early and mid-twenties, but most of the Beatles songs that he sang were essentially of his own creation.) As creativity declines, moreover, songwriting effort seems to increase. If I recall correctly, Paul Simon once said that in his youth songs simply poured out, whereas the older he became, the harder he had to work.
Bob Dylan makes an excellent case to consider the relationship between age and songwriting creativity. Dylan is widely regarded as the greatest songwriter of our time – if not of all time – and his output is vast (over 600 songs). He has worked in many genres yet is highly distinctive. He also continues to receive critical acclaim for his new songs. In 2021, 80 relatively well-known musicians were asked to name their favorite Dylan song to honor his 80th birthday, and five picked something from his 2020 album, Rough and Rowdy Ways. But as creative as he still is, I do see a descent over time. I also think that most critics and fans would agree.
This hypothesis might be testable, and least to some minimal degree. Many assessments have been made of Dylan’s greatest songs, by critics and others. The songs that are so acclaimed might serve as a reasonable proxy for his creativity, allowing one to graph his best output against his age (see the figures posted below).
Although the sources that I consulted vary significantly in the songs that they selected, the general pattern is clear. Dylan – in the imaginations of those who assess him – peaked in his early and mid-twenties. (As Joan Baez memorably put it, he “burst on the scene already a legend.”) The ranking of his work then plummets when he reached his late twenties, only to rebounded in his early and mid-thirties. But according to the critical consensus, he would never regain his previous heights. By his 40s, according to most sources, Dylan was writing few if any great songs. Although many critics have opined that he “returned to form” in his mid-50s, few of his songs from that period – and beyond – make the lists of his best. The general consensus seems to be that Dylan’s output from his early and mid-twenties outweighs everything that came later.
To sum it up, most observers agree that Dylan has experienced creative decline with age. But it is equally important to note that the downward slope has not been even, and that he has remained an impressively creative songwriter to this day. Youth helps, it would seem, but other factor are always at play. Dylan’s artistic decline in his late twenties, for example, may have had less to do with a drop in creativity than with devotion to family life. Consider, for example, these sweet but unremarkable lyrics from “Sign in the Window” (1970):
Build me a cabin in Utah
Marry me a wife, catch rainbow trout
Have a bunch of kids who call me “Pa”
That must be what it’s all about
That must be what it’s all about
The first graph is derived from asking ChatGPT to list Dylan’s “best songs by critical consensus.” I start with AI because it tends to give a bland overarching view, derived from numerous sources. The list that it provided clearly shows the pattern that I described above, albeit without the “long tail” found among the more knowledgeable and perceptive critics. The next figure purportedly graphs the collective assessment of “Dylan fans.” The same pattern is seen, although in simplified form. Ditto for the next two graphs, each of which is limited to his ten best songs.
The next set of figures are derived from longer lists. That of ShortList is somewhat unusual in showing a clear peak in his mid-twenties, relatively little resurgence in his thirties, and a long tail as he aged. One37PM, in contrast, has Dylan peaking in his early twenties. American Songwriter does so as well, and includes only one song written after his mid-thirties. Music Grotto also has an attenuated tail, but does contain one recent song (“Murder Most Foul”). In contrast, Paste includes many songs written after 1980. It also shows a smaller-than-usual decline in his late twenties, and rates the work of his early and mid-thirties at a higher-than-average level.
Rolling Stone has the most comprehensive list, as is fitting for a publication named after one of Dylan’s most famous songs (“Like a Rolling Stone,” which it places in the top slot). The overall pattern that it depicts, however, is fairly typical. Rolling Stone does, however, include a higher percentage of songs written in the early 1980s than is usual, as this is generally seen as the low point in his career. (The songs in question are: “The Groom’s Still Waiting at the Altar” [1980]; “Someone’s Got a Hold of My Heart” [1983]; “Sweetheart Like You” [1983]; “Jokerman” [1983]; and “Blind Willie McTell” [1983].)
Dylan’s 1997 album Time Out of Mind is regarded by many critics as marking a significant reappearance of excellence after years of disappointing ventures. As the Wikipedia article on the album notes:
For many fans and critics, the album marked Dylan’s artistic comeback after he appeared to struggle with his musical identity throughout the 1980s … .Time Out of Mind is one of Dylan’s most acclaimed albums, and it went on to win three Grammy Awards including Album of the Year in 1998.
But while a few songs from Time Out of Mind appear on several of the longer lists graphed below, the period in which it was written does not really stand out on these figures. There may have been a “return to form,” but it was evidently nothing like the form that he had reached in his early and mid-twenties.
(Several clarifications must be said about the graphs posted below. First, the age categories are based on the year in which he wrote the song in question, not the year in which it was recorded or released. Songs on the celebrated album Blood of the Tracks are conventionally dated to 1975, but most were written in 1974. The time brackets that I employ, dividing the decades of Dylan’s career into three segments, are of unequal length: four years for the “early” portion of each 10-year span, and three years for the “mid” and “late” portions. (I think it is safe to assume that most of us regard persons aged 20, 21, 23, and 23, as being in their “early” twenties, with “mid-twenties” reserved for those aged 24, 25, and 26, and so on.) Also to note is the fact that breaks on the graph often fail to correspond with breaks in Dylan’s style. His early electric songs, such as “Subterranean Homesick Blues” and “Maggie’s Farm,” are graphed as being from his early twenties, as they were written in 1964 when he was 23, although they were released in 1965 and are associated with his mid-twenties phase, the origin of which is chronicled in the second half of the recent film, A Complete Unknown.)
ChatGPT List of Best Bob Dylan Songs Graphed by Dylan’s Age
StudyFinds List of Best Bob Dylan Songs Graphed by Dylan’s Age
UCR List of Best Bob Dylan Songs Graphed by Dylan’s Age
Singersroom List of Best Bob Dylan Songs Graphed by Dylan’s Age
ShortList List of Best Bob Dylan Songs Graphed by Dylan’s Age
One37PM List of Best Bob Dylan Songs Graphed by Dylan’s Age
American Songwriter of Best Bob Dylan Songs Graphed by Dylan’s Age
Music Grotto List of Best Bob Dylan Songs Graphed by Dylan’s Age
Paste List of Best Bob Dylan Songs Graphed by Dylan’s Age
Rolling Stone List of Best Bob Dylan Songs Graphed by Dylan’s Age
The fourth GeoCurrents lecture on global demography has been uploaded on YouTube and can be found here. It examines population issues that arose after the development of agriculture, as well as those following the initiation of horse-based nomadic pastoralism. It first looks at the spread of neolithic farming into Europe, noting that the expanding agricultural populations largely replaced the preexisting hunter-gatherers, who were evidently restricted to a few areas in which crop-growing was not feasible. It then turns to the oddity of the resurgence of genetic material from hunter-gatherer peoples that occurred in the later neolithic period in Europe around 4000 BCE. The argument is then advanced that new agricultural populations usually overwhelm hunter-gatherer population in areas conducive to farming, but that hunter-gatherers sometimes expand out of such “refugia” into much more densely populated agricultural areas, where they invariably take up agriculture themselves. Three examples are then discussed, using historical linguistics as the primary base of evidence. Two first two look at instances of the spread of farmers and the retreat of hunter-gatherers: the Bantu expansion in Africa and the Austronesian expansion in Southeast Asia, the Pacific, and the Indian Ocean. The third example looks at a hunter-gatherer expansion into an agricultural zone, that of the Uto-Aztecan-speaking “Chichimeca” people who pushed into Mesoamerica and gave rise to the Aztec Empire.
The next part of the lecture turns to population crashes that sometimes occurred in early agricultural societies, whether from disease, weather fluctuations and climate change, soil erosion and environmental degradation more generally, or the violent influx of new human groups. The main example investigated is the decline of European populations in the late neolithic, which was marked by decreasing health, the appearance of the bubonic plague, and the arrival of Indo-European-speaking peoples with a partly pastoral way of live. The lecture ends with a discussion of the population surge and geographical expansion of the early Indo-Europeans, who had created a new way of life in Central Eurasia centered on horse-based nomadic pastoralism.
The third lecture in the current series on global demography has been recorded and uploaded on the GeoCurrents YouTube channel and can be found here.
This lecture covers demographic patterns and processes from the origin of modern humans some 200,000 years ago to the initial development of agriculture around 11,000 years ago. I begin by advising those interested in this topic to watch a YouTube video of Dwarkesh Patel interviewing David Reich, a leading expert on the genetic history of humankind. Reich argues, in essence, that the more we know the less certain we become about the overarching narrative. I then outline the current understanding of the spread of anatomically modern people out of Africa roughly 70,000 years ago and their interactions with the archaic human groups of Eurasia, the Neanderthals and Denisovans. The lecture then goes on to explore such demographically significant issues as the Toba Event 74,000 years ago. The explosion of the massive Toba volcano in Sumatra has been linked to the near extinction of humankind, although recent research indicates that its effects were not nearly so severe. At any rate, rough estimates are given for human populations at various stages of our early history.
David Reich’s Fascinating Speculations
The discussion then turns to the expansion of humans to new lands, first to Australia, circa 50,000 years ago, and then to the Americas. I look at the controversies surrounding the original inhabitation of the Americas. I argue that the first people in the Western Hemisphere probably followed the coast and remained somewhat tied to it. A substantial amount of evidence indicates that they were later followed by another wave at the end of the last glacial period years that came from Siberia over the Bering “Land Bridge.” A discussion of the Pleistocene megafauna (mammoths, mastodons, giant ground sloths, etc.) that these newcomers encountered follows. I conclude this section by considering the extinction of most large mammals in North and South America, whether from climate change or human overkill – or some combination of the two.
Pleistocene Australia Map
The lecture then turns to the abrupt climate change that marked the transition from the Pleistocene to the Holocene epochs around 11,700 years. As the globe warmed and the ice caps retreated, vegetational patterns rapidly shifted, forcing people to adapt to new conditions. In the Near East, agriculture began to emerge at this time. I end the lecture by discussing the demographic consequences of the development of agriculture – higher fertility rates and much larger local populations – as well as the health costs that the transition entailed.
The second GeoCurrents YouTube video has been posted, which examines the current debate between anti-natalists, who think that the world is severely overpopulated and therefore want to reduce birthrates, and pro-natalist, who are concerned about plunging fertility and therefore want to increase birthrates.
I do not take a side in this debate on this video. Instead, I try to faithfully outline the arguments of each camp, while supplying information, as accurately as I can, that illustrates or supports the various positions taken.
My main goal in teaching (although not in writing), is to provide information and context that help students reach their own informed positions of issues that are subject to debate.
The lecture becomes rather intricate in places, as I strive to outline arguments, counter-arguments, rejoinders to such counter-arguments, and so, as can be seen in the two slides posted below.
I have posted the first of an estimated 16 illustrated lectures on global population on the GeoCurrents YouTube channel. These lectures are derived from a class that I am currently teaching in Stanford University’s Continuing Studies Program (adult education) called “Population Explosion or Birth Dearth? Understanding Global Demography.
The course is framed around the contentious debate between anti-natalists, who think that the world is severely overpopulated and therefore want to reduce birthrates, and pro-natalists, who are worried about declining fertility and therefore want to increase birthrates. As in all my classes, I strive to retain a neutral position and explain the positions of both camps without bias.
Pro-Natalism vs Anti-Natalism graphic
I begin with the anti-natalist position, noting that the world population is increasing by around 220,000 persons per day. I then outline the high fertility rates found in most of Africa and in countries such as Afghanistan, exploring the growing environmental and other problems that these places are facing. I then note that while anti-natalist admit that the global population will almost certainly plateau and then decline, they believe that such a transition will be highly beneficial and should be hastened.
Africa TFR Map
I then move on to the pro-natalist position, using Elon Musk as a prime example of such thinking. I outline the plummeting birthrates of a number of countries and explore the implications of the hyper-low fertility found in South Korea. I then explore in some detail the recent fertility transitions of the world’s two demographic giants, China and India. I also show that fertility rates in Africa are now rapidly dropping.
Elon Musk and Pro-Natalism
The next segment of the lecture looks at a few countries that defy these global trends and offers a few possible explanations for their exceptionality. Most of Central Asia, for example, has seen a fertility rebound since around 2005, and is now above the replacement rates. Israel is a more commonly noted exception, as it is the only wealthy country with a fertility rate above the replacement level. I also explore the data on fertility and religion – and religiosity – in Israel.
Mongolia Fertility Rate Graph
The lecture ends with a brief consideration of the accuracy of the fertility figures that I use, contrasting data from the United Nations with information from a popular X-account called Birth Gauge. I also consider such issues as the “tempo effect,” in which fertility rates appear to be dropping faster than they actually are due to the fact that many women in developed countries are choosing to delay childbearing. But, as I note, countries with the highest mean age of childbearing and those with the lowest mean age of childbearing are all characterized by fertility rates at or below the replacement level. In countries with high fertility rates, in contrast, women typically have children over the span of many years.
A recent GeoCurrents post on Utah’s declining birth rate included maps of the Total Fertility Rate (TFR) of the U.S. by state in 2008 and 2022. Comparing the state-by-state data from these two years yields a map of total fertility change during this period (posted below). As this map shows, the fertility drop was, in general, more pronounced in the West than in the rest of the country. In contrast, it was less pronounced in the Midwest, along with a couple of adjacent states (Kentucky and West Virginia).
Changes in U.S. Fertility Rate by State, 2008-2022 map
This regional distinction made me wonder about the possible role of urbanization in changing fertility patterns. The two states with the most pronounced TFR drops, Utah and Nevada, are both characterized by high urbanization rates, whereas states with less pronounced declines, such as North Dakota, Iowa, West Virginia, and Kentucky, have lower urbanization rates.
The Wikipedia map of urbanization by state, however, shows that the correlation is relatively weak. Mississippi, for example, had a significantly greater fertility drops than Louisiana, yet its urbanization rate is much lower. It is intriguing, however, that Maine, the eastern state with the lowest fertility drop, also has one of the country’s lowest urbanization rates. The least urbanized state, Vermont, also had a relatively low fertility decline in this period. But the TFR of Vermont is still the lowest in the country (1.35), which was also the case in 2008 (when in was 1.67).
U.S. Urbanization Rate by State, map
Changing fertility patterns is a complex issue, one that continually challenges – if not bedevils – professional demographers. Their problems in projecting changes in this regard are glaringly evident in the Financial Times’ graph posted below. As a well-known quip of uncertain origin reminds us, “It is difficult to make predictions, especially about the future.”*
* Attributed by different sources to Niels Bohr, Samuel Goldwyn, K.K. Steincke, Robert Storm Petersen, Yogi Berra, Mark Twain, Nostradamus and “Anonymous.”
On the U.S. Total Fertility Rate (TFR) map of 2008, Utah clearly stands out. At the time, the state’s TFR was 2.60, well above the national average of 2.08. Utah’s relatively high birth rate has usually been linked to the demographic dominance of members of the Church of Jesus Christ of Latter-day Saints (LDS), colloquially called the Mormon Church. As of 2020, Mormons constituted a little over 60 percent of the population of Utah. Significantly, second-place Idaho, with a TFR of 2.47 in 2008, also has a large LDS population, roughly a quarter of the population. As noted by demographer Tim Heaton in 1988, when the U.S. fertility rate was plummeting in the 1970s, that of Utah was increasing.
U.S. Total Fertility Rate 2008 Map
The Mormon church has long encouraged large families. My own paternal grandmother came from an LDS polygamous family of 19 children, born of two sister wives. (As a result, she had quite a few half siblings who were also cousins.) Although I was not raised in the church, I did attend extended-family reunions, where my relatives seemed countless. One year, my first cousins and I got in a fight with another group of boys who turned out to be second cousins attending the same event, held in a large public park. The adults who broke up the scuffle found it quite amusing.
Although the Mormon fertility rate in the U.S. remains higher than average, it has declined considerably in recent years. An article entitled “The Incredible Shrinking Mormon American Family” notes that the TFR of the faith declined from 3.31 in 1981, when the national rate was 1.81, to 2.42 in 2014. Although recent information is not readily available, it seems safe to assume that the decline has continued. The TFR of Utah in 2022 waS 1.85, only a little above the national average.
2022 U.S. Total Fertility Rate Map
According to the author of the “Incredible Shrinking Mormon American Family,” changes in church teachings are largely responsible for this decline. As the author notes:
Gradually in the 1980s and ′90s, Latter-day Saint leaders stopped overtly preaching against birth control, even though they still promoted the importance of children. In the 1998 Handbook, contraception is considered to be a matter between the couple and the Lord, and church members are advised not to judge one another about it. (As if.) That don’t-ask-don’t-tell approach has effectively been the church’s policy for the past 20 years.
I am somewhat skeptical of this argument. The prohibition on contraception in the Roman Catholic Church, after all, has not changed, yet the American Catholic fertility rate dropped before that of the Mormons. It is quite possible, however, that LDS families are more inclined to follow church teachings than Catholic ones.
Tim Heaton argued in 1988 that the Mormon family of the time was quite distinctive from the modal American family, being characterized by what he referred to as the “Four C’s,” namely (premarital) chastity, conjugality, children, and chauvinism (the final term referring to a belief in traditional gender roles). But he also argued that these distinctive attributes were changing and would probably continue to do so. As Heaton perceptive argued:
The ideology of the LDS church has been remarkably flexible in accommodating social change The same central doctrine of eternal marriage was used to sanction polygamy in the nineteenth century as is currently used to promote the family patterns described above. As the church has spread to more culturally diverse areas and as new social trends have been adopted by the LDS membership, policies and practices have modified accordingly.
Another possible reason for the declining LDS fertility rate is the gradual convergence of Mormon social and cultural practices with those of the American population at large. Distinctive religious groups in the U.S. that have maintained high or ultra-high fertility rates – the Amish, Old-Order Mennonites, Hutterites, and Haredi Jews – have, in contrast, remained far more distinctive, and their ideologies have not been “remarkably flexible in accommodating social change.”
But purely material factors may also play a role. Housing prices are now elevated in Utah (see the map below), making it difficult for many young couples to have children. Perhaps not coincidentally, in the states with the highest current fertility rates, South Dakota and Nebraska, housing remains much more affordable. But housing prices are even lower in West Virginia, and its TFR is roughly that of the national average. The situation is, to say the least, complicated.
According to the United Nations Population Fund, China’s 2024 Total Fertility Rate (TFR: the expected number of children per woman) was 1.2. Among sovereign states, only Singapore and South Korea are ranked below China (at 1.1 and 0.9 respectively). Some sources, however, contend that China’s fertility rate is considerably lower. Database Earth, for example, gives a current figure of 1.02. Conversely, other sites report higher figures. Macrotrends claims that China’s TFR is currently over 1.7 – and that is has been slowly increasing for a quarter century (see the graph posted below). Macrotrends’ analysis is ostensibly based on UN data, specifically from its “World Population Prospects 2024” website. But the graph of China’s TFR on this UN website does not square with that of Macrotrends. It does show China’s TFR slowly increasing from the late 1990s to the late 2010s, eventually reaching around 1.7, but it then depicts a sharp drop-off that has continued to the present. The same UN graph, however, then projects China’s TFR as slowly increasing over the rest of this century at various estimated rates, although likely remaining below replacement level. Evidently, there is a lot of unacknowledged uncertainty about global fertility rates, especially when it comes to future projections. Overall, previous UN projections have severely underestimated the pace of global fertility decline.
China 2024 TFR Database Earth
Macrotrends China TFR graph
China UN TFR projection graph
One of the more prominent sources arguing for fertility rates below those reported by the UN is the data-rich X account called Birth Gauge. In October 2024, Elon Musk referenced Birth Gauge in warning that low birth rates “will lead to mass extinction of entire nations.” The author of Birth Gauge uses information on birth numbers reported by individual countries to calculate up-to-date TFR figures. I cannot assess the accuracy of these assertions, but I must say that the author’s estimate of the Philippines’ TFR falling to 1.3 in 2024 seems unlikely (see the table below); the UN, after all, still pegs the Philippines’ TFR at 2.7.
Birth Gauge 2024 TFR Data Table
Birth Gauge TFR figures are usually but not always below those of the UN. Most of its higher numbers pertain to Central Asia. For example, it posts a 2024 figure of 3.57 for Kyrgyzstan, as opposed to the UN’s 2.9. Also of note is Birth Gauge’s admission, given in an April 7, 2023 X post, that actual TFR figure may end up being higher than what it and other sources report: “When women shift their births to higher ages, the conventional TFR underestimates how many children they will really have in the end, which is the ‘tempo effect.’”
Birth Gauge pegs China’s 2024 TFR at 1.10, a slight uptick from its 2023 figure of 1.02. It provides a useful data table of China’s estimated TFR figures for 2023 by region, which I have converted into a map (see below). As can be seen, all China’s province-level administrative districts are evidently below the replacement level. Only Tibet, with a reported figure of 1.97, comes close. As a comparison with the second map posted below shows, which depicts per capita GDP by region, fertility patterns in China have little connection with economic productivity. Extremely low birth rates are found in some of China’s wealthiest provinces (Jiangsu) and in some of its poorest (Heilongjiang). Regions with relatively high fertility, however, do tend to be less economically productive than average. Intriguingly, the coast/interior dichotomy, prominent on the GDP map, does not appear on the TFR map. The latter map, however, does show a muted differentiation between the lower-fertility north and the higher-fertility south. Particularly notable are the extraordinarily low birth rates in China’s northeastern “rust belt” (Manchuria). This was the most economically productive part of China in the mid-twentieth century, noted for its heavy industry. It did not, however, share proportionally in the economic boom of the 1990s and early 2000s, and is now known for its depressed conditions.
Most demographers agree that the Total Fertility Rate (TFR) of India has dropped below the replacement rate of 2.1 to 2.2 children per woman. The United Nations Population Fund places it at 2.0 for 2024, while Database Earth puts it at 1.94. As the graph from the latter organization (below) shows, India’s fertility decline has occurred at a relatively even pace since peaking at 6.0 in 1964.
India Total Fertility Rate Decline graph
Such figures, however, mask significant regional variation over the vast country. Roughly a decade ago, I made a map of Indian demographic patterns contrasting the low fertility present at the time (2012) in the south and far north with the moderately high levels of fertility found in the north-center and the northeastern periphery. A series of new maps that I have just made, based on 2024 data, show a more muted regional pattern. Currently, only one large state, Bihar, has a TFR above 2.5, and only two additional large states, Uttar Pradesh and Jharkhand, remain above the replacement rate. (Provided, of course, that the data are accurate.)
India Total Fertility Rate by State 2012 map
Although north-central and far northeastern India still have higher fertility than the rest of the country (see the map below), a general convergence has taken place. When I began teaching world geography in the late 1980s, India’s southwestern state of Kerala grouped with Sri Lanka as the only polities in South Asia that had moderately low human fertility. By 2012, Kerala, with a TFR of 1.7, was well below the replacement level – as was neighboring Tamil Nadu and several other southern Indian states, as well as a few states in other parts of the country. Today, the southeastern half of India is significantly below the replacement rate, while parts of the far north have entered a regime of extremely low fertility. India’s Himalayan state of Sikkim, for example, posted a TFR of 1.1 children per woman in 2024. If Sikkim were an independent country, as it essentially was before annexation by India in 1975, it would be tied with Singapore in the second lowest slot, ahead of only South Korea.
India 2024 Total Fertility Rate Replacement Level Map
India 2024 Total Fertility Rate by State Map
As the next map in the series shows, the regions of relatively high fertility in India’s north-center and far northeast experienced sharper drops in fertility between 2012 and 2024 than the rest of the country, illustrating the continuation of the convergence process. The large state of Madhya Pradesh (population 88 million), for example, saw its TFR drop from 3.0 to 2.0, a remarkably fast decline. The mostly Christan (Baptist) state of Nagaland in the far northeast experienced a larger drop, from 3.3 to 1.7. In contrast, Kerala and Tamil Nadu in the far south posted modest increases, from 1.7 to 1.8 in both states. This development indicates that southern India is unlikely to descend into a regime of extremely low fertility.
India Change in TFR 2012-2024 Map
The final map in the series puts the fertility rates of Indian states in regional context by including the 2024 TFR data for nearly countries. As can be see, Sri Lanka, Nepal, Bhutan, and Bangladesh are all below the replacement rate, while Myanmar (Burma) is essentially at that level (2.1). It is a different story, however, in the northeast, where Pakistan (3.3) and Afghanistan (4.3) remain well above the replacement rate.
A recent article in Newgeography by demographer Wendell Cox analyzes net domestic migration from one state to another from July 2020 to July 2024. Aptly titled “United States Moves South,” the article includes an intriguing data table outlining the current trends. In this short period, the South (as defined by the U.S. Census Bureau) gained 2,658,499 domestic migrants at the expense of the rest of the country, while the Midwest lost 494,057, the West lost 978,415, and the Northeast lost 1,186,027. As the author also notes, these gains and losses were unevenly distributed across each of these four large regions. Most the migration from the Midwest, for example, came from the more industrial eastern half of the region, a subregion designated “East North Central” by the Census Bureau (Ohio, Michigan, Indiana, Illinois, and Wisconsin).
U.S. Census Regions and Divisions Map
The pattern is more intricate at the level of the individual states. As a crude map of the raw data shows (see below), most of the migration out of the eastern Midwest came from one state, Illinois, which posted a figure of -395,721. As the map also indicates, one state – California – is responsible for most of the outflow from the West. A somewhat similar situation is found in the Northeast, with New York losing a whopping 893,537 residents. As can also be seen, five southern states each gained more than 200,000 domestic migrants: Florida (810,122), Texas (693,979), North Carolina (383,851), South Carolina (299,962), and Tennessee (237,368). In the same region, however, Mississippi and Louisiana saw substantial outmigration (20,099 and 124,444 respectively).
U.S. Net Domestic Migration 2020-2024 Map
U.S. Net Domestic Migration 2020-2024 Population-Proportional Map
It is not surprising that the most populous states – California, Texas, Florida, and New York – would post large absolute figures, whether positive or negative. But a different picture emerges when we look at domestic migration from 2020 to 2024 as a percentage of total state population. As this map (posted above) shows, the interior West vies with the Southeast as a magnet for domestic migrants. In proportional terms, Idaho was the top attractor, and Montana came in third (after South Carolina).
As the map also shows, domestic migration from 2020 to 2024 was not really focused on the so-called Sun Belt. Several states and areas in the southern portion of the country experienced net outmigration: Mississippi, Louisiana, New Mexico, Hawaii, and especially southern California. By the same token, several “Frost Belt” states in the north gained migrants: Idaho, Montana, Wyoming, South Dakota, Maine, New Hampshire, and Vermont.
A stronger correlation, although still far from perfect, is found between state-level political orientation and domestic migration trends. In general, Republican-voting “red” states gained while Democratic-voting “blue” states lost. Harsh COVID restrictions in 2020 and 2021, along with high taxes and strict regulations, have often been linked to migration out of “blue” states. But again, there are notable exceptions. Republican-voting Alaska, the least-taxed state in the nation,* saw a significant population outflow in this period. Conversely, highly taxed Vermont, by some measures the country’s most leftwing state, had a positive migration rate, although just barely. More striking is Delaware, which posted one of the highest population-proportional in-migration rates. But although a strongly Democratic-voting state, giving Kamala Harris a 15 percent edge over Donald Trump in 2024, Delaware is also noted for its low taxes. According to a Wikipedia table, it had the country’s third lowest “state tax burden” in 2022 (see the map below, derived from the Wikipedia table linked to above). A Delaware Online article attributes the state’s recent population gains to “tax leniency and housing affordability – including homes that are approximately 16% cheaper than Maryland rates and 22% cheaper than New Jersey rates …” Significantly, 13 of the 14 states with the lowest taxation rates had more people moving in than moving out from 2020 to 2024. By the same token, 10 of the 12 states with the highest taxation rates had more people moving out than moving in.
U.S. State Tax burdens Map
West Virginia’s gain of 11,089 domestic migrants from 2020 to 2024 is surprising, as the state is noted for its long history of outmigration. Its population peaked at 2,005,552 in 1950 and now sits at approximately 1,770,000. According to a Wikipedia demographic chart, West Virginia’s lost 3.2 percent of its population between 2010 and 2020. It also supposedly lost another 1.3 percent – 27,737 persons – from 2020 to 2024, a period when it is reported to have gained some 11,089 domestic migrants.
*Despite its low rate of taxation, Alaska does, however, have a high cost of living, due mainly to high levels of expenditure on food and health care.
The Antiplanner blogsite recently ran an interesting and controversial post arguing that South Korea’s extraordinarily low fertility rate is linked to its prevalence of high-rise housing. As the author put it:
South Korea’s high-rise housing and low birthrates are closely related. People don’t have children if they don’t have room for them. High rises are expensive to build so living space is at a premium. Birth rates are declining throughout the developed world, but they have declined the most in countries like South Korea, Russia, and China that have tried to house most of their people in high rises.
The post elicited pushback, with one commenter stating that she saw “not a shred of evidence other than his bald assertion that people in Korea have no room for kids.” Evidence is indeed necessary to support such a claim, but is it available? It is true that some other countries noted for their high-rise housing, most notably Brazil, have also experienced plummeting fertility. But in both Brazil and South Korea, low fertility is also characteristic of rural areas and small towns that are not dominated by high-rise housing, albeit not to the same degree as in large cities covered with apartment towers.
My immediate reaction to this article was to try to devise a geographical test, one that would allow direct comparisons of housing types and fertility rates. Unfortunately, I was not able to find a relevant data source in the time that I allotted myself for the task. The best information that I could find is a list of European countries by people living in detached and semi-detached housing. People not living in such dwellings can generally be assumed to live in apartment (or condominium) blocks, which can be low-rise, mid-rise, or high-rise. Although this would therefore be a poor test of the Antiplanner’s thesis, it nevertheless seemed worth pursuing.
As can be seen in the paired maps below, the correlation between multifamily housing and fertility levels in Europe is weak. It is true that most countries with extremely low fertility have little detached or semi-detached housing, including Greece, Italy, and Spain. By the same token, some countries that have abundant detached or semi-detached housing have relatively high fertility, such as Ireland. But note the exceptions. North Macedonia, for example, has extremely low fertility but a high percentage of people living in detached or semi-detached housing, whereas Estonia shows the opposite pattern.
Since the Antiplanner claims that high-rise housing generates low fertility primarily because of inadequate room for child rearing, a better measurement would be to compare TFR with average living-space per household. I have not, however, been able to find an adequate data set to assess this assertion. A Eurostat graph showing “average number of rooms per person 2021” (size unspecified), however, does not indicate a significant correlation. According to this graph, Malta has the most capacious housing in Europe, with 2.3 rooms per person, yet its TFR, 1.13, is one of the lowest in the world. The same source also indicates that ultra-low fertility Spain has much more spacious housing (2 rooms per person) than relatively high-fertility Romania (1.1 rooms per person).
Culturally informed views about the amount of room necessary to rear a child vary significantly from country to country. In general, the wealthier the society, the more space is considered necessary. Such calculations also vary with employment conditions. I have been told by several young couples that more room is necessary for child rearing than before COVID, as one bedroom must now be reserved for an office that can be devoted to at-home work through Zoom. That belief could be dismissed, however, as a mere rationalization for not having children.
The most interesting finding from the data on detached and semi-detached housing in Europe concerns the geographical differences between these two categories. As the second set of paired maps shows, a few countries that have relatively little detached housing have an abundance of semi-detached housing, particularly the United Kingdom and the Netherlands.
