Population Geography

The Cost of Housing and Essential-Worker Relocation in Booming Southwestern Montana

Casual conversations in Bozeman, Montana often turn to housing. Some verge on the tragic. I have spoken with several young couples who spent years saving for a down payment and were finally on the verge of making an offer – and then Covid hit, remote work intensified, and prices soared. As can be seen on the map, housing prices are high in Gallatin County as a whole, but in central Bozeman they approach the level found in California’s Bay Area. As a result, the city faces a mounting labor crisis, particularly for essential workers such as nurses. According to a confidential source, in a single unit of approximately 30 nurses in Bozeman’s Deaconess Hospital, 12 nurses quit in the past 12 months, with 9 citing housing costs as their reason.

Nurses and other financially stressed residents are abandoning Bozeman for more affordable places. Several nurses have relocated to Spokane, Washington, where wages are higher and housing is cheaper. Most have opted for another Montana city. As can be seen in the paired maps, nurses make more money in Yellowstone County (Billings) than in Gallatin County, and the cost of living is much lower.* Trade-offs, of course, must be factored in. Yellowstone County has a much higher crime rate than Gallatin County, and both its natural and cultural amenities are less appealing.



For those who value natural amenities above all and want to remain in Montana, options are more limited. Most of the scenic valleys of western Montana also have high housing costs. As can be seen on the ArcGIS “housing affordability” map, made just before the Covid price surge, Madison County, with fewer than 10,000 residents, was even less affordable than Gallatin. In such a low-population county, it does not take many affluent amenity seekers to skew the market upwards. Housing is less costly, however, in a few western counties, such as Silver Bow (Butte), Deer Lodge (Anaconda), and Lewis & Clark (Helena).

Anecdotal evidence from casual conversations suggests that the favored destination of priced-out Bozeman residents is Helena. Surrounded by mountains, Helena is nestled in a scenic location, and as the state capital it has a descent array of cultural resources. More important, it is relatively affordable and its crime rate is relatively low.





Another popular relocation place is Anaconda in Deer Lodge County, a former copper-smelting city of some 9,000 people. After a long period of decline, Anaconda is again growing. As an extended headline in a recent article in the Montana Free Press (MTFP) reads, “Building on its past, Anaconda draws new residents seeking best of Montana: The long-struggling southwestern Montana town is gaining popularity with recreationists and homebuyers. Can it retain its historic character?” As the article notes:

Anaconda is the last best place of the last best places,” said Vanessa Romero, 39, who moved to Anaconda in 2017 from Boise, Idaho, and is opening a wine shop downtown.

This isn’t the typical Montana discovery tale. Movie stars aren’t buying sprawling ranches here. Tourists aren’t pouring into town, though they are coming at a steady trickle. The rich and famous aren’t flocking to the local ski area. Rather, Anaconda’s newcomers are often young couples and extended families who want a low-pressure lifestyle in a Montana community but have been priced out of other areas. The town is an example of a historically industrial community that is adapting to a recreation economy. …

We’re seeing the refugees of Missoula and Bozeman” and other rapidly growing towns in the West, said Adam Vauthier, executive director of Discover Anaconda, an economic development organization. “The other recreation towns in the vicinity just got so big.”

But will such cities as Anaconda and Helena retain their affordability?  Footloose and often well-compensated Zoom workers also find them attractive. And prices are increasing. As the Montana Free Press(MTFP) article notes, “the Multiple Listing Service shows a median home price of $294,000 in Anaconda, up 36.7% over the same time in 2021.” Some local residents are also concerned about the city’s changing demographic characteristics. As one resident told MTFP reporter Erin Everett, “’We want to get rid of the buffalo touchers’ … referring to visitors who get too close to bison in Yellowstone National Park.”

*As the paired maps also show, Golden Valley County is an even more advantageous location for nurses, but with only 823 residents, employment opportunities are limited.

The Cost of Housing and Essential-Worker Relocation in Booming Southwestern Montana Read More »

What Is a Zoom Town?

Bozeman, Montana is often described as a quintessential “Zoom town,” a city or small town that has experienced explosive growth owing to the relocation of remote workers since the beginning of the Covid pandemic. Bozeman is certainly booming, and many of its new residents do work remotely, usually through Zoom. But how widespread is this phenomenon, and where might other “Zoom Towns” be located?

Although many article have been written recently on Zoom towns, the term remains poorly defined, with most designations based on informal impressions. Maps showing Zoom town locations are all but non-existent. The only map that I was able to find comes from the Ownerly website, specializing in data for homeowners. This map was made to illustrate an article called “Zoom Towns USA: America’s Best Cities for Remote Workers.” Ownerly devised a metric to measure Zoom-town suitability by analyzing:

445 cities across the nation, looking at rent and housing prices, cost of living, safety data, level of broadband and free Wi-Fi coverage. Besides expanding our list of cities by nearly 50% from our 2021 Best Zoom Towns list, we also added new metrics that include cost and availability of childcare, restaurants and coworking spaces.

The resulting map is reproduced here. Ownerly’s “best” Zoomtowns are concentrted in the northeastern quadrant of the country, with Wisconsin and Pennsylvania leading the list. But ideal though they may be, few of the towns and cities on this map are commonly deemed “Zoom towns.”  Consider, for example, a map showing Wikipedia’s “examples of Zoom towns and regions.” Intriguingly, none of these places makes the Ownerly list. Wikipedia’s Zoom towns are concentrated in the West, with California in the lead position.

Other sources have their own lists, often focused on a particular part of the country. Rate.com emphasizes towns in the Rocky Mountain region, calling special attention to Lewistown, Idaho; Walla Walla, Washington,; and Caspar Wyoming. Northwest News, focusing on the Northwest, focuses on Bend, Oregon while also mentioning Washington’s Methow Valley and San Juan Islands; Kelowna in British Columbia; Sandpoint, Idaho; and (yet again) Bozeman Montana. A BBC article on Zoom towns begins with a discussion of Fayetteville, Arkansas while also mentioning Sandpoint, Idaho; Moab, Utah; and Durango, Colorado. The article stresses the location of Zoom towns in “rural enclaves.”

Rural enclaves certainly do not dominate Wikipedia’s list of exemplary Zoom towns. Roughly half of the places on the Wikipedia list are suburbs with populations over 100,000. Such cities were growing quickly before Covid and are within the normal commuting range of their metropolitan cores. Placing such cities in the same category as Aspen, Colorado; Truckee, California; and Bethel, Maine might be a bit misleading.

A rigorous definition of Zoom town would emphasize population growth since 2020 and the percentage of workers working remotely. The latter piece of information is not easily obtainable. Using LinkedIn data, however, a 2021 Make It article claims that:

      These small cities have the highest proportion of remote work applications:

    1. Bend, Oregon: 41.8%
    2. Asheville, North Carolina: 38.7%
    3. Wilmington, Delaware: 35.9%
    4. Johnson City, Tennessee: 35.2%
    5. Eugene, Oregon: 34.9%

These larger cities have the highest proportion of remote work applications:

    1. Cape Coral, Florida: 33.1%
    2. Charleston, South Carolina: 31.6%
    3. Tampa Bay, Florida: 29.6%
    4. Jacksonville, Florida: 29.4%
    5. Orlando, Florida: 29.2%

Based on the information contained in this GeoCurrents post, Bend, Oregon is a good candidate for the title of America’s quintessential Zoom town. Bend was booming, however, well before it was Zooming; its population surged from 20,000 in 1990 to 99,000 in 2020.


What Is a Zoom Town? Read More »

Montana Population Change in Comparative Context

Recent posts have emphasized population decline in the Great Plains of eastern Montana. Comparative analysis shows, however, that eastern Montana has fared better than many other parts of the region. Almost every county in western Nebraska, for example, experienced population decline from 2010 to 2020. Yet Nebraska as a whole saw moderate population growth in the same period. Most of this growth was focused in the Omaha metropolitan area in the far eastern part of the state.

Although outside of the Great Plains region, Illinois also makes an interesting contrast.  In that state, almost every rural county declined, many sharply. Illinois as a whole, along with West Virginia and Mississippi, lost population during this decade.

To the east of eastern Montana, the western Dakotas experienced pronounced population growth from 2010 to 2020. Growth in western North Dakota was almost entirely a result of the oil boom in the Bakken Formation. The Bakken also extends into far eastern Montana, but most of the oil extraction there took place before 2010. In western South Dakota, growth has been focused on the forested Black Hills region.

Montana Population Change in Comparative Context Read More »

Montana’s Changing Population Geography

(Note: The next several posts will focus on the geography of Montana.)

As noted in previous posts, Montana is experiencing a population boom. Three decades ago, however, there was widespread worry about population stagnation and possible decline. Since becoming a state in 1889, Montana has experienced several demographic cycles, each marked by different geographical patterns. Geographer William Wyckoff has extensively documented these changes. In particular, I recommend his essay “Peopling the Last Best Place, 1870-1990,”* published by the Burton K. Wheeler Center at Montana State University at the time of lagging growth and concern about the state’s future. I will briefly summarize some of Wyckoff’s arguments before considering the more recent era of population growth.

The early cycles of boom and bust, Wyckoff argues, were linked to economies of natural resource development, governmental investments, and climatic fluctuations. Prior to 1890, the non-native population was heavily focused in the southwestern part of the state, particularly in a handful of mining districts. As Wyckoff’s map of settlement in 1870 shows, southwest Montana at the time was linked most closely to Utah. A decade before the turn of the century, however, the settlement of the Great Plains in the east accelerated, propelled by railroad development, a series of relatively wet years, and high wheat prices.

But as the map shows, Montana’s population in 1900 was still concentrated in the west, particularly in the four labeled counties. Far above the others was Silver Bow, containing the extraordinarily productive and profitable copper mines of Butte. Neighboring Deer Lodge, a copper-smelting hub, also stands out (its shuttered Anaconda Smelter Stack, at 585 feet, is the tallest surviving masonry structure in the world). Relatively large population clusters were also found in Lewis & Clark County, another mining center, and Cascade County, an emerging energy (hydropower) and transportation hub. Helena, in the former county, boasted of having the world’s highest concentration of millionaires (50 in a population of 12,000 in 1888). Cascade’s county seat of Great Falls exploded from 3,979 residents in 1890 to 14,930 in 1900. Gallatin County in southwest, with 9,533 residents, was the main agricultural region supporting the mining economy. As can be seen in Wyckoff’s second map, the Butte mining district had become the state’s demographic core by 1890.


In the first two decades of the twentieth century, agricultural expansion transformed the Great Plains, at least in areas accessible to the rapidly growing rail network. World War I was an especially promising time. As a result, many eastern Montana counties were subdivided. Ironically,  the demographic cycle went into reverse just as new counties were proliferating. Drought struck in the late teens and returned periodically in the 1920s, a decade that did not roar in Montana; the state’s population dropped by 2.1 percent during this period. As Wyckoff notes, over half of Montana’s banks failed between 1920 and 1926.

The agricultural boom of the period from 1890 to 1920 resulted in a far more geographically balanced population structure for the state as a whole. As can be seen on the 1930 population map, the mining district of the southwest were no longer dominant. We also see the rise of Yellowstone County and its key city of Billings, an emerging agricultural, energy, and transportation hub. Billings would surpass Great Falls as Montana’s largest city around 1970, and today has almost twice Great Fall’s population.

Moderate population growth returned to Montana in the post WWII period, with the 1950s and 1970s posting strong gains. With the exceptions of Cascade and Yellowstone countries, however, the Great Plains continued to stagnate. Agricultural mechanization reduced the need for labor, and the region’s harsh climate and remote location discouraged settlement. As Wyckoff notes, population growth in the post-war period was propelled by the forestry industries in the northwest, energy in the greater Billings and Great Falls area, and the early emergence of natural-amenities development in some of the more scenic valleys in the mountainous west.

In the 1980s, Montana’s population growth rate fell to 1.6 percent. As Wyckoff writes, “More than two-thirds of Montana’s 56 counties lost population, and the state’s struggling economy since 1980 prompted over 50,000 residents to leave.” Stagnation in a time of national economic growth generated concern, leading Montana State University to commission the study in which Wyckoff’s essay appears. But the situation would soon turn around. Montana’s population grew by almost 13 percent in the 1990s and by almost 10 percent in each of the first two decades of the new century. Much of this growth was focused on amenity-rich areas in the west. Of particular note are the two college towns of Missoula and Bozeman, home, respectively, of the University of Montana and Montana State University. Missoula County grew by almost 22 percent in the 1990s, while Gallatin County, where Bozeman is located, posted a gain of more than 34 percent. As Montana State University increasingly emphasized science and engineering, Bozeman emerged as a regional tech hub, contributing to population gains of more than 30 percent in Gallatin County during both the 2000-2010 and 2010-2020 periods.

The 2020 map of population by county again reveals a stark demographic divide, especially when contrasted with the map of 1930. With the exceptions of Yellowstone (Billings) and Cascade (Great Falls) counties, the Great Plains remains very sparsely settled. Most of its counties had larger populations in 1920 than they do today, and several have fallen below 1,000 residents. Population is now concentrated in the west, but even in western Montana most countries have small populations and have experienced little growth. That situation is perhaps changing, however, as exorbitant real estate costs lead many people to seek housing outside of the region’s booming cities, especially Bozeman. Many other parts of western and central Montana have excellent natural amenities, and thus appeal to those who value outdoor experiences.

*In “Population Decline in Montana,” by Patrick C. Jobes, Craig Wilson, and William Wyckoff. Published by the Burton K,. Wheeler Center, Montana State University, Bozeman, Montana. April 30 and May 1, 1991

Montana’s Changing Population Geography Read More »

Recent Population Growth — and Decline — in Montana

(Note: As I am spending the summer in Montana to be with my granddaughters and their parents, a number of forthcoming post will focus on the state.)

As can be seen in the map on the left, Montana is a booming state, posting the third largest rate of population growth (in percentage terms) in the United States from 2020 to 2021. In the 2022 election, Montana will gain a second seat in the U.S. House of Representatives, which it had lost in 1993 due to lagging population growth at the time. Montana is still a sparsely settled state, with the third lowest population density in the country (after Alaska and Wyoming). Its population surge, moreover, is relatively recent. As can be seen in the second map, from 2010 to 2020 it experienced a moderate rate of growth (9.6%), higher than the national average (7.4%), but well below those of neighboring Idaho (17.3%) and North Dakota (15.8%), as well as nearby Utah (18.4%).





As might be expected, Montana’s recent population expansion is very unevenly distributed. Whereas some counties are experiencing sizzling grown, others continue to decline. Growth is concentrated in the mountainous west, whereas decline is pronounced across much of the Great Plains in the east. But as can be seen in the map on the left, two far-eastern counties (Richland and Carter) experienced rapid growth from 2010 to 2020, defying the regional norm. This map, however, is somewhat misleading; as Richland and Carter are sparsely settled countries, small increases in absolute numbers translate into rapid proportional growth. A gain of a mere 255 residents in Carter resulted in a 22% growth rate over the decade. Previously, the tendency had been one of steady decline. In 1920 the county had 3,972 residents, dropping to 1,415 by 2020.


The growth in Richland County from 2010 to 2020 is easily explained; adjacent to the booming oilfields of eastern North Dakota, it became something of a bedroom community for housing-short Williston, ND. But as the oil boom has receded, so too has Richland County. As can be seen in some of the maps posted below, the county lost population from July 2020 to July 2021. Explaining the growth in Carter, one of the most remote counties in the lower 48 states, is more challenging. A recent article in the excellent Montana Free Press, however, is quite helpful. A newly paved road improved access to eastern South Dakota, another area of recent population expansion (due in part to the natural amenities of the Back Hills region). The article also cites a healthy county budget, owing in part to transit fees from energy pipelines, and, unlikely as it might seem, dinosaur tourism. Ekalaka, Carter’s county seat, is evidently a high point on Montana’s so-called Dinosaur Trail.

To help readers make sense of changing population patterns in Montana, I have made several versions of my population-change-by-county maps (posted below). As can be seen in one of these iterations, growth has been concentrated in and around Montana’s largest cities, although Great Falls has lagged behind. Later posts will explore some of these patterns in more detail.

The most recent census data, covering the period from July 2020 to July 2021, shows a continuation of most of the trends seen in the 2010-2020 period. Although most Montana countries grew sharply during this time of COVID, the northern Great Plains continued in its seemingly inexorable decline. All of Montana’s larger cities, except Great Fall and Butte, saw rapid growth. So did Ravalli County in the scenic Bitterroot Valley, a zone of high rural population density (by Montana standards). Also of note is the growth rate of Flathead County in the northwest surpassing that of Gallatin County (which includes Bozeman) in the southwest. This somewhat surprising; as an expanded headline in the Montana Free Press notes, “Montana’s fastest-growing city last year? It wasn’t Bozeman. New Census Bureau estimates chart Montana’s population shifts during the first full year of the COVID-19 pandemic. Kalispell led the pack.” This article, by Eric Dietrich, is well worth reading, especially for Dietrich’s superb cartography. (Note that there are a few discrepancies between his map and mine; although I have re-checked the data, his figures may be more accurate.)





The competition between Bozeman and Kalispell is interesting. Intriguingly, Kalispell is Montana’s only significant city with no “blue” (Democratic voting) precincts, although the nearby tourism-oriented town of Whitefish is a different matter. Bozeman, in contrast, is markedly blue, as is Gallatin County as a whole. Gallatin’s smaller towns and most of its rural areas, however, are decidedly red. I have posted details from one of Eric Dietrich’s excellent Montana electoral maps, also published in the Montana Free Press. Later posts will explore Montana’s changing electoral geography in more detail.










Recent Population Growth — and Decline — in Montana Read More »

Are Small Towns Really Urban Places? Eastern Montana According to the U.S. Census Bureau

When teaching population geography, I always discuss urbanization rates. Perceptive students sometime ask about the official statistics that I use, wanting to know how “urban” is defined. When I tell them that that U.S. Census Bureau classifies all settlements with a population of 2,500 or more as urban, they look surprised and sometimes question the designation. That threshold, selected in 1910, seems far too low, not fitting the current concept of urban settlement.

Eastern Montana is a good place to examine the mismatch between the official designation and the popular view of what constitutes an urban place. As can be seen on the map posted here, the eastern quarter of the state is one of the most sparsely populated parts of the contiguous United States, with most census tracks having fewer than one person per square mile. A Purdue University study gave most of the counties in the region a high “relative rurality score.” A more recent U.S. Census report classifies most counties in eastern Montana as “100 per cent rural.” Yet the same report also placed four of the region’s counties in the “mostly urban” category. It did so because most of their residents live in the county seats, small towns that exceed the 2,500 urban threshold.






















One of these counties is Dawson, a sparsely settled place with only 3.4 people per square mile (1.4/km2) and fewer than ten thousand total residents. But more half of them live in Glendive (population 4,873) and adjacent West Glendive (population 1,998). This distributional pattern makes Dawson a “mostly urban” county, at least according to the Census Bureau.









It is probably safe to say that most observers would reject the idea that the small and declining town of Glendive constitutes an “urban cluster,” as it is designated by the Census Bureau. But that does not mean that it is rural place. Intermediate terms are needed. Fortunately, several exist, including “suburb,” “exurb,” and “small town.” Glendive clearly fits the latter category. As a typical small American town, it shows far greater cultural affinity with rural areas than it does with urban ones. It does not, for example, exhibit anything like an urban voting pattern. In 2020, Glendive gave roughly 70 percent of its votes to Donald Trump. But as can be seen in the map posted here, a minor urban effect is visible: rural areas of Dawson County gave Trump more than 80 percent of their votes.

In 1910, when the urbanization rate in the United States stood at only 45.6 percent (with Montana recording 35.5 percent), it made sense to classify small towns as urban places. It no longer does. Depicting Dawson County now as “mostly urban” is misleading, based on an antiquated classification system.

Are Small Towns Really Urban Places? Eastern Montana According to the U.S. Census Bureau Read More »

Total Fertility Rates by Country, 1950 and 2015

(Note to readers: The distribution of free customizable base-maps will recommence later this week)

TFR 2010-2015 World MapI was recently asked to make a world map of the Total Fertility Rate (TFR) using World Bank data for the period 2010-2015. The same data sheet includes TFR figures for each five-year period starting in 1950. The contrast between fertility rates in 1950 and 2015 is so striking that I could not resist the temptation to make contrasting maps.




TFR 1950-1955 World MapAs can be seen, the overall drop in the human fertility rate has been pronounced. In countries ranging from China, to Iran, to Libya, to Brazil it has been nothing short of extraordinary.

TFR 1950-2015 World MapsTropical Africa, and especially the western portion of the region (“Middle Africa,” according to the World Bank), forms the major exception to this pattern. Here, several counties had higher TFR figures in 2010-2015 than they did in 1950-1955.   In DR Congo, for example, the respective numbers are 6.15 Gabon Fertility Graphand 5.98. Gabon also ended up in a higher category in the 2010-2015 map than on the 1950-1955 map, although the actual difference between its TFR in these two periods is not statistically significant: 4.00 and 3.99. As the graph shows, Gabon’s TFR did not stay the same during this period, but rather rose gradually and then began to decline gradually. Similar graphs can be found for other countries in “Middle Africa.”

It is quite significant that extremely high fertility figures are now mostly confined to tropical Africa, with only a few exceptions (such as Afghanistan and East Timor). It also seem to me that this phenomenon has been under-reported. Certainly most of my own students come into class with the impression that high fertility rates characterize most of the word’s less-developed countries.


Total Fertility Rates by Country, 1950 and 2015 Read More »

Population History, Population Density, and Cultural Values in the Philippines


Philippines Fertility Rate ChartWith 103 million inhabitants, the Philippines now ranks 12th on the list of countries by population. Owing to its moderately elevated birthrate, its population is still expanding at a brisk rate, outpacing those of its Southeast Asian neighbors. Steady population growth has been a feature of the country for some time. In 1903, the archipelago held fewer than eight million people, by 1940 its population had doubled, and by 1990 it had reached 60 million.



Philippines Popluation Density 1939 MapThe distribution of the inhabitants of the Philippine has historically been decidedly uneven. Some areas have been thickly settled for centuries, but vast expanses long remained open. As can be seen in the map posted to the left, in 1939 relatively few people lived in Mindoro, Palawan, northern Luzon (with the exception of the narrow Ilocos Coast of the northwest), and Mindanao (with the exception of the north-central coast). Central Luzon and the central Visayas islands (Cebu, especially) were densely populated. Philippines Religion 1890 MapThese same patterns are reflected in the map of religion from the late Spanish period. As can be seen, in the late 1800s most of northern Luzon, Mindoro, Palawan, and even Mindanao were inhabited primarily by animist tribal populations, most of which were small in numbers (although not all, see below).


Luzon Central Plain MapSome of the patterns seen on the 1939 Philippine population map were of long-standing, but others had emerged more recently. The island of Mindoro, for example, had been a lightly settled land since the early Spanish period, although prior to that it had been an important trading hub; evidently, struggles between the Spaniards and the Muslims of the south resulted in partial depopulation followed by the spread of malaria. The Central Plain of Luzon, now the agricultural heartland of the country, tells a different tale. Up to the mid and late 1800s, its central region was Ilocano Migrations Mapstill heavily forested. As Marshal McLennan’s maps show, only its southern and northern margins were settled in 1837 by farming communities, those of the Kapampangan and Pangasinan peoples respectively. Starting at about that time, however, a major stream of migrants from the densely populated Ilocos Coast began to flow southward into the Central Plain. Somewhat later, Tagalog-speaking people began to move north into the same belt of fertile lowlands. By the mid-20th century, the formerly peripheral central zone of the Central Plain had been added to the core region of the Philippines.

Philippines Internal Migration MapThe Central Plain of Luzon was by no means the only region of the Philippines to see large-scale migrations in the early 1900s. Ilocanos from northwestern Luzon were also moving into the Cagayan Valley of northeastern Luzon and other places as well, including Hawaii and California (the vast majority of Filipino migrants to the United States in the first half of the 20th century were from the Ilocos region). Other densely settled areas, such Cebu, Bohol, and southeastern Panay in the Visayas Island, also sent out large numbers of migrants, most of whom settled in Mindanao. A seemingly odd feature on Spencer and Wernstedt’s map of internal Philippine migration from 1948 to 1960 is the high level of emigration from Samar in the eastern Visayas, as Samar has never been a densely populated island. But it was—and is—a very poor place, another factor that often encourages people to seek better opportunities elsewhere.

Philippines Popluation Density MapOwing to this long history of internal migration, the population distribution of the Philippines has become somewhat more even. Mindanao is no longer a frontier zone, and is now moderately populated by Philippine standards. To be sure, Palawan, Mindoro, and much of the Cordillera of northern Luzon still have relatively few inhabitants, but they also lack extensive lowlands that attract agricultural settlers. Central Luzon, on the other hand, has intensified its position as the demographic core of the country, a process closely associated with the explosive growth metro Manila, which now counts a population of almost 12 million. Today, the focus of Filipino migration is foreign rather than internal. An estimated 10.2 million persons born in the Philippines currently live abroad, representing around 10 percent of the country’s population.


Philippines Poplulation Growth MapOn the population density map of 2010 (above), the Ilocos Coast of northwestern Luzon appears as a zone of low to moderate population density. This standing seems surprising, as this area was noted for its dense settlement dating back to the time of the Spanish conquest in the late 1500s. The mapping here is a little bit deceiving, however, as the three provinces of the region all include rugged highlands as well as the narrow coastal plain. But it is true that the Ilocos Coast is no longer very demographically distinctive. Many Ilocanos, as we have seen, settled elsewhere, and in their homeland their fertility rate dropped, doing so sooner than almost all other areas of the country. As a result, the Ilocos region has seen relatively little population growth in recent decades.

In the Philippines, Ilocanos have a reputation for being hard working, thrifty, and somewhat stingy. As a result of this stereotype, they have been occasionally deemed the “Yankees of the Philippines.” To the extent that this common view is is accurate, the industriousness and frugality of the Ilocanos can be partially explained by the powerful theory of the Danish economist Ester Bosurup (1910-1999). In her classic anti-Malthusian work, The Conditions of Agricultural Growth: The Economics of Agrarian Change under Population Pressure, Boserup argued that population pressure, which the Ilocanos long faced, forces agricultural communities to engage in ever more frequent and intensive cropping cycles. As this process generates diminishing returns over time, people living under such conditions are forced to develop a strong work ethic. Although Boserup framed her theory in cultural terms, a number of geneticists now claim that there is a substantial genetic component to this process as well.* This notion is, of course, extraordinarily controversial, and I remain unconvinced although open-minded. But I am convinced that Boserup figured out one of the most important dynamics of human history, and I also think that she should occupy a much more prominent position in intellectual history than she does.

In 2012, The Inquirer, a major Philippine newspaper, reported on a modest test of the supposed frugality of the Ilocanos based on an examination of savings rates by region in the Philippines. As the report concluded:

Finally, Ilocanos have statistical proof to show that their detractors are wrong in calling them the most tightfisted Filipinos. A study by the National Statistical Coordination Board (NSCB) shows that Ilocanos now rank only sixth among the prolific savers in the country’s 17 regions.

Ifugao Rice TerracesIntriguingly, the same study found that the Cagayan region of northeastern Luzon—populated almost entirely by Ilocanos—ranks as the second thriftiest region of the Philippines. The most savings-oriented Filipinos by a wide margin, however, are the highlanders of the Cordillera of Northern Luzon. The Inquirer article notes that most people of this region also speak Ilocano, but in actuality they do so as their second (or third, or fourth) language, and they certainly do not belong to the Ilocano ethnic group. But we can again turn to Boserup to gain insight on this finding. The ancestors of most Cordilleran groups migrated out of the lowlands into the highlands in order to escape Spanish taxes and to continue to worship their ancestors and practice their indigenous religion. There they had to undertake herculean labors to wrest a living from the rugged landscape, constructing some of the world’s most impressive agricultural terraces in the process. The Cordilleran people, in general, had to develop a strong work ethic to survive under such circumstances.

Owing in part to the highlanders’ predilection for intensive labor, and in part to their own prejudice against Hispanicized cultures, American administrators in the Philippines in the early 1900s often favored the mountaineers over the more civilized Filipino lowlanders. This attitude is explored by the historian Frank Jenista in his 1987 book White Apos: American Governors on the Cordillera Central. As Jenista frames the issue in his preface:

The purpose of this study is to examine the Philippine experience, the fascinating but hitherto untold story of the interaction between the American colonial authorities and the independent headhunting, terrace-building people of Luzon’s Gran Cordillera Central…. It focuses primarily on the province of Ifugao, for the American presence there was longest and best remembered. Despite the great cultural differences between the turn-of-the-century Ifugaos and Americans, the oral accounts which make up this story portray for us an intriguing relationship—a mutually satisfactory symbiosis due in large measure to an unexpected congruence of important cultural values…

*For an overview, see Nicholas Wade, A Troublesome Inheritance: Genes, Race, and Human History.



Population History, Population Density, and Cultural Values in the Philippines Read More »

Mapping Crime and Substance Abuse in Russia


In the previous post, I examined regional differences in demographic issues across Russia. As many sources note, alcoholism is one of the biggest factors contributing to low life expectancy and high rate of death from non-natural causes. In fact, Russia ranks at the top in terms of both alcohol consumption (especially by men), as discussed in detail in my earlier post. Russia and the neighboring FSU countries also top the charts in percentage of deaths attributable to alcohol. According to the World Health Organization report, such a high level of alcohol-related deaths results not only from what and how much people drink (e.g. drinking spirits causes more alcoholism and alcohol-related deaths than drinking wine or beer), but also how they drink. The WHO report includes data on what the authors refer to as “Patterns of Drinking Score” (PDS), which is “based on an array of drinking attributes, which are weighted differentially in order to provide the PDS on a scale from 1 to 5”; the attributes include the usual quantity of alcohol consumed per occasion; festive drinking; proportion of drinking events, when drinkers get drunk; proportion of drinkers who drink daily or nearly daily; and drinking with meals and drinking in public places. Some countries with high per capita alcohol consumption have low PDS scores (particularly, countries in southern and western Europe), while Russia has one of the highest PDS scores worldwide.

A combination of what, how much, and how people drink—as well as who is doing the drinking—creates the observed patterns of alcoholism within Russia as well. (The data on alcoholism and drug abuse, discussed below, comes from the Wikipedia.) The prevalence of alcoholism differs from region to region even more than other social indicators considered in previous posts. The range between the highest and lowest alcoholism levels is one of three orders of magnitude, from less than one case reported per 100,000 population in Ingushetia to nearly 590 cases per 100,000 in Chukotka. Overall geographical patterns are quite clear: the highest levels of alcoholism are registered mostly across the Far North (and some of the regions of the Far East, which despite their relatively low latitude qualify legally as the “Far North”), while the lowest levels of alcoholism are found in the Caucasus. According to the alcoholizm.com website, alcohol consumption is also the lowest in the north Caucasus, especially in Chechnya, Ingushetia, and Dagestan, and the highest in the harshest areas of the Far North/Far East (particularly in the Jewish Autonomous oblast, Nenets Autonomous Okrug, Magadan, and Kamchatka). However, they do not list the specific figures of alcohol consumption in various regions and I could not find them elsewhere, but I suspect that the quantity of alcohol consumed per capita in the northeast Caucasus and in the Far North does not differ by three orders of magnitude. Much more significant is the type of alcoholic drink being consumed: vodka is the most common choice in the Far North, whereas the Caucasus region has a long history of viniculture going back 8,000 years. Besides viniculture, the Caucasus is well-known for its venerable tradition of wine-drinking with meals, accompanied by elaborate toasts, which slows down the consumption and metabolism of alcohol. Another important factor is genetically based difficulty that people from many indigenous groups have in breaking down alcohol. This genetic propensity causes particularly high levels of alcoholism in such areas of northern Russia as Nenets Autonomous Okrug, Sakha Republic, and Chukotka.


Unlike alcoholism, drug abuse is considerably less common than alcoholism across the Russian Federation: all but four federal subjects register over 50 cases of alcoholism per 100,000, whereas none reports more than 50 cases of drug abuse per 100,000, and fewer than a dozen register over 30 cases. Geographical patterns of drug abuse are less obvious and harder to explain than those of alcoholism. Particularly high levels of drug abuse (over 30 cases per 100,000) are found in Murmansk, Chelyabinsk, Sverdlovsk, Kurgan, Novosibirsk, Kemero, Amur, and Sakhalin oblasts, and in Primorsky Krai. One common denominator is that all of these regions had a high level of industrial urban development during the Soviet period but have experienced significant economic stagnation since the fall of the USSR. One area that does not fit this generalization is Sakhalin, which is still a major area of natural resource extraction and processing. It would be interesting to know if any of our readers can think of another explanation for these patterns of drug abuse.


Interestingly, neither alcoholism nor drug abuse correlate very closely with crime rates, although there is a connection. Of the areas where drug abuse is most prevalent, only Sakhalin and Kemerovo oblasts also exhibit high crime rates (over 2500 cases per 100,000), while several others—Murmansk, Sverdlovsk, and Novosibirsk—have an average level of crime. Alcoholism correlates with crime even less, as only Magadan and Sakhalin oblasts exhibits high levels of both. A number of regions with a high level of alcoholism, particularly Nenets Autonomous Okrug and Sakha Republic, which have a large percentage of indigenous population, exhibit little crime. Other areas with high proportions of indigenous peoples, such as the Caucasus and the Middle Volga, also have particularly low levels of crime. The low level of reported crime in Tula, Ryazan, and Belgorod oblasts is perplexing, however.


When it comes to murder, eastern and especially southern Siberia is clearly more dangerous than European Russia or western Siberia. The “murder capital” of Russia is Tuva, which registered over 35 homicides per 100,000 population. Neighboring Altai Republic has over 25 murders per 100,000 population, as does Zabaikalsky Krai. Curiously, high rates of alcoholism may contribute to the elevated murder rate in such areas as Sakha Republic, Chukotka, and Jewish Autonomous oblast, but other areas where alcoholism is prevalent, such as Karelia and Sakhalin, register only slightly above-average murder rates. Similarly, the prevalence of drug abuse does not correlate with murder rates. Nor do crime rates or murder rates correlate with the level of urbanization, unemployment level, or regional GDP.



Mapping Crime and Substance Abuse in Russia Read More »

Mapping Russia’s Demographic Problems

[Note to readers: customizable maps of Russia are now available in Russian here.]

Much has been written about Russia’s demographic problems, particularly in the 1990s and 2000s. The country as a whole is characterized by low birth rates and high abortion rates; high death rates, especially from non-natural causes; rather low life expectancy, especially for men; and skewed sex ratios. This post examines some of these issues, focusing on regional differences across Russia. The GeoCurrents maps presented below are based primarily on data from the Federal State Statistics Service; some of the indicators, such as the percentage of working age adults and of pensioners as well as sex ratios, have been calculated directly from the FSSS data. Additional data comes from the “Children in Russia” publication by the FSSS, available (in Russian) here.



As maps of Russia’s birth rates, death rates, TFR, and natural population growth by federal subject can be found in Wikipedia, we begin by mapping life expectancy (at birth). According to data from the World Bank, the life expectancy of an average Russian male is a whopping 10 years shorter than that of an average Russian female: the figure for men is 66 years (the same as in Kazakhstan, Iraq, and North Korea), while that for women is 76 years (the same as in Iran, Honduras, and Tonga). But as the FSSS data mapped on the left reveals, there are significant differences in life expectancy among Russia’s federal subjects. For example, life expectancy for an average Ingush woman is almost 15 years longer than that of her Chukotkan counterpart (81.32 years vs. 66.42 years). The contrast is even more striking with respect to men: life expectancy for an average Ingush male it is almost 20 years longer than for his Tuvan counterpart (75.97 years and 56.37 years, respectively). Overall, the highest life expectancy, for both genders, is found in the two federal cities (Moscow and Saint Petersburg) and in the northeastern and north-central Caucasus. For both genders, Ingushetia has the highest life expectancy figures, while Dagestan is in the top four (72.31 years for men, 78.82 years for women). North Ossetia ranks 3rd in life expectancy for women (79.06 years), with above-average life expectancy for men (68.46 years). Curiously, Chechnya ranks 4th in male life expectancy (70.23), the ongoing insurgency notwithstanding, while life expectancy for Chechen women is close to average. Neighboring regions of southern Russia also post fairly high life expectancy figures. Moscow City ranks 2nd and Saint Petersburg 5th in life expectancy, for both men and women. As with many other standard-of-living indicators, some of which are discussed in the previous post, the oblasts surrounding the two federal cities present a sharp contrast to the cities themselves: both Moscow and Leningrad oblast post average figures for female life expectancy and below-average figures for male life expectancy.

Outside the Caucasus and the federal cities, life expectancy is shorter, with most regions in Siberia posting lower figures than those of European Russia. There are a few exceptions, however, including higher-than-average figures for both genders in Belgorod oblast in south-central Russia, Tatarstan in the Middle-Volga region, and Khanty-Mansiysk, and higher-than-average figure for men in Yamalo-Nenets Autonomous Okrug. For women, life expectancy is higher than average in a number of regions in south-central Russia (Belgorod, Voronezh, and Tambov oblasts) and Middle-Volga region (Chuvashia and Penza oblast). At the bottom of the ranking, one finds Tuva and Chukotka (and for men, the Jewish Autonomous Oblast): life expectancy in these regions is almost a decade shorter than the country’s average.


A large gap in life expectancy for men and women helps explain the skewed sex ratios. As can be seen from the map of the 2013 FSSS data, most Russian regions have more women than men, with many regions having fewer than 85 males per 100 females (shown in the two darkest shades of red). There are only three regions where there are more men than women (shown in blue): Yamalo-Nenets Autonomous Okrug, Chukotka, and Kamchatka.

Map of Russia Sex Ratio

The map above can be instructively compared with an earlier map, based on the 2002 census data. The two maps are constructed so that the data is binned in the same way, although the lowest (most female-dominated) category of the older map has been broken down into two categories in the newer map. A comparison of the two maps reveals that the situation did not improve in the decade or so separating the two maps; on the contrary, many regions became even more skewed in the female direction in this period. A good example comes from what is today Krasnoyarsk Krai. In 2002, Krasnoyarsk region consisted of three administrative units: Evenk autonomous district with a ratio of 1.007 (i.e. more men than women), Taimyr (Dolgano-Nenets) autonomous district with a ratio of 0.948, and Krasnoyarsk territory with a ratio of 0.889. In 2013, the entire Krasnoyarsk Krai, amalgamated from these three regions, had a ratio of 0.875, which was more female-biased than the ratio in any of its constituent parts in 2002. Similarly, in most regions in Siberia and the Urals (i.e. Zabaikalsky, Primorsky, Khabarovsk, and Altai Krais; Amur, Irkutsk, Kemerovo, Tyumen, Sverdlovsk, Kurgan, and Chelyabinsk oblasts, Jewish Autonomous Oblast, Khakassia and Sakhalin), sex ratios also became more skewed towards women. The sex ratio of Murmansk oblast followed the Siberian trend in becoming more skewed towards women. Movement towards a more balanced ratio have been registered only in a small number of regions: Novosibirsk (from 0.866 to 0.872), Ulyanovsk (from 0.856 to 0.878), Belgorod (from 0.848 to 0.854), Tambov (from 0.844 to 0.859), Kaluga (from 0.841 to 0.859), and Moscow oblasts (from 0.850 to 0.858), Chuvashia (from 0.863 to 0.872), and Mari El (from 0.869 to 0.872). In the north Caucasus and southern Russia, the overall tendency has also been towards more skewed sex ratios, the most striking case being that of Ingushetia, where the sex ratio dropped from 0.876 to 0.819, the fourth lowest in Russia. The lowest sex ratio is found in Yaroslavl oblast, whereas Ivanovo oblast—whose capital has been known unofficially as “the city of brides”—ranks second from the bottom.



Regional differences in the age structure of the population are also significant—and help explain the demographic and economic situations in the various regions. Let’s begin by examining the youngest segment of the population, children under age 17 (the data comes from the “Children in Russia” publication by the FSSS). Clear geographical patterns can be seen on the map on the left: children constitute a larger proportion of the population (25% or more) in eastern and southern Siberia (but generally not the Russian Far East), northeastern Caucasus, and Nenets Autonomous Okrug. A comparison with the second map posted on the left shows that areas with a higher proportion of children are generally those also with a higher proportion of indigenous peoples (i.e. a lower proportion of ethnic Russians). This correlation is confirmed by the figures listed in the Wikipedia article on the demographics of Russia: ethnic Russians have the country’s second-lowest fertility rate. (Russia’s Jews have the lowest fertility figure of all ethnic groups.) As can be expected from the high birth rates in these three regions, Chechnya, Ingushetia, and Tuva rank at the top in the percentage of children, with about a third or more of their populations under the age of 17. In contrast, in most of European Russia (and a few regions in Siberia) children constitute less than 20% of the population, and in most oblasts in central Russia the figure is below 17%. Unsurprisingly, the lowest percentage of children is found in the two federal cities: 14.3% in Moscow and 14.4% in Saint Petersburg.


At the other end of the age spectrum are the pensioners, a group that is now raising a lot of concerns. Despite Russia’s relatively low life expectancy (especially for men, as discussed above), the percentage of pensioners has been growing for some time. According to a paper by Julie DaVanzo and David Adamson “Russia’s Demographic “Crisis”: How Real Is It?” (published in 1997), “between 1959 and 1990, the number of persons aged 60 and over doubled… [by] the beginning of the 1990s, reach[ing] 16 percent”. Since then, the trend has continued, as noted by the Ministry of Labor and Social Protection. Russia currently has one of the lowest retirement age thresholds: according to the Pension Fund website, “men older than 60 and women older than 55 qualify for [the old age] pension” (people living in the Far North and other harsh regions have an even lower retirement age threshold). However, recently there has been much discussion about the possibility of raising the retirement age to 65, possibly as early as in 2016. One of the main aims of this proposal is to alleviate the shortages of pension funds that resulted from Russia’s current economic woes. These problems may also lead to a potential decrease in pensions paid to current retirees, as reported by Gazeta.ru.

As with other demographic indicators, the percentage of pensioners differs widely from region to region. Only three regions—all of them in the north Caucasus—had fewer than 22% retirees in 2013, and half a dozen others (including Moscow City, the two autonomous okrugs in Western Siberia and three additional regions in the north Caucasus) posted figures below 25%. Conversely, the highest percentages of pensioners are found in northern European Russia (Republic of Karelia, Arkhangelsk oblast, and Komi Republic), four oblasts south of Moscow (Bryansk, Oryol, Tula, and Ryazan), as well as in Kurgan oblast and on Sakhalin. (Curiously, a lower percentage of pensioners does not correlate closely with high GDP: for example, Yamalo-Nenets and Khanty-Mansiysk Autonomous Okrugs have relatively few pensioners, whereas Sakhalin ranks 7th highest in the percentage of pensioners.) The two federal cities, especially Moscow, have low percentages of pensioners.


To conclude our discussion of the age structure, let’s consider the map of the percentage of working age adults. As with the other indicators, regional differences are quite pronounced and are due to different factors. For example, Kurgan oblast has the lowest percentage of working age adults (55.3%), and Chechnya ranks 3rd lowest (with 56.3%), but the population structures in those two regions are quite different: Kurgan oblast has a high percentage of pensioners, whereas Chechnya ranks highest in the percentage of children. At the other end of the spectrum, one finds such regions of high GDP as the Yamalo-Nenets and Khanty-Mansiysk Autonomous Okrugs and Chukotka, as well as adjacent Kamchatka and Magadan oblast. However, despite similar percentages of working age adults, the overall population structure of these regions is distinctive: Chukotka, for instance has substantially more children and fewer pensioners than Magadan oblast. The latter region has experienced a significant depopulation trend, which affects the younger adult population more than the older people, resulting in a disproportionately aging population. Moreover, this trend feeds itself: even without taking into account the out-migration, an aging population results over time in lower birth rates.


The shortages of working age adults in regions such as Chechnya and Tuva are further exacerbated by high levels of unemployment, as can be seen from the map on the left: 26.9% and 19.3%, respectively. The highest unemployment figure, 43.7%, comes from Ingushetia. Elevated unemployment rates (over 10%) are found also in several other regions in the North Caucasus (Dagestan, Kalmykia, Karachay-Cherkessia) and southern Siberia (Altai Republic, Zabaikalsk Krai). Unsurprisingly, these regions also have the country’s lowest GDP figures. In the following (and last) post on the regional differences across Russia, we will examine unemployment patterns in the context of substance abuse and crime rates.





Mapping Russia’s Demographic Problems Read More »

Mapping Regional Differences in Economic and Social Development in Russia—A GeoCurrents Mini-Atlas

Generalized indicators of economic and social/human development, such as GDP per capita or HDI, typically place Russia into a medium-high category. However, such ratings overlook regional differences in economic and social development, which are highly pronounced in Russia. To examine these regional patterns, GeoCurrents has created a mini-atlas of Russia, designed using GeoCurrents customizable maps, which are available for free download. These maps examine a wide range of topics, from food consumption to alcoholism, and from crime rate to healthcare; additional maps cover issues that help explain regional patterns in development, such as the age structure and ethnic composition of the population. Unless indicated otherwise, the data comes from the Federal State Statistics Service, and refers to the year 2013. Since the data offered by the FSSS is presented in 83 Word files, one for each federal subject, we have re-organized the data into one Excel file (available for download here: Rosstat_data); some of the measures, such as the percentage of working age adults or of pensioners and sex ratios, have been calculated based on the FSSS data. Additional data comes from the “Children in Russia” publication by the FSSS, available (in Russian) here; this document, published in 2009, contains data from the preceding year. Some other data come from Wikipedia and refer to 2010 or 2013. Unfortunately, we have not been able to obtain data from a more recent date, particularly from after the annexation of Crimea in March 2014; if any of our readers know of such publicly available data, in English or Russian, please let us know.

Russia_Living_space_2013We’ll begin by looking at two rather unusual measures of the standard of living: the availability of living space and food consumption. Although Russia is a large and sparsely populated country, the availability of residential housing has long been a problem. As can be seen from the map on the left, residents of central Russian oblasts have more living space per capita than average, with inhabitants of Tver oblast enjoying an average of 29 sq. meters (312 sq. feet) per person. The only exception here is Moscow City, where an average resident has only 19.2 sq. meters (207 sq. feet) of living space, reminding one of Mikhail Bulgakov’s lament about Moscovites written some 75 years ago: “mercy sometimes knocks at their hearts…ordinary people… only the housing problem has corrupted them…” (Master and Margarita). While residents of Northern European Russia, the Volga region, and the Far East (Chukotka, Kamchatka, Sakhalin) have fairly ample living space, the North Caucasus and most of Siberia offer an average citizen more crowded housing. Dagestan, Kabardino-Balkaria, and Chechnya have less than 20 sq. meters (215 sq. feet) of living space per capita, while Ingushetia posted the second-lowest figure in all of Russia: 13.5 sq. meters (145 sq. feet). A notable exceptions here is North Ossetia-Alania, with the figure of 26.9 sq. meters (290 sq. feet) of living space.

Similarly, several Siberian regions, such as Khanty-Mansiysk and Yamalo-Nenets Autonomous Okrugs, and Altai Republic (not to be confused with Altai Krai), have less than 20 sq. meters (215 sq. feet) of living space per capita. Particularly striking is the situation in Tuva: 12.9 sq. meters (139 sq. feet) per capita. As we shall see in subsequent posts, Tuva is found at the bottom of many development rankings. As mentioned above, the Far East overall has more residential housing per capita, although differences between, on the one hand, Primorsky Krai and Jewish Autonomous oblast, with less than 22 sq. meters (237 sq. feet) per capita, and Magadan oblast, with its ample 29 sq. meters (312 sq. feet) per person, is striking. However, in the case of Magadan, the higher availability of residential housing may be a symptom not of a higher standard of living, as one might think, but actually of a lower standard of living: since the dissolution of the Soviet Union, Magadan oblast has a significant depopulation trend, and as we shall see in subsequent posts, many other indicators of human development there paint a grim picture, which helps explains this trend.


Consumption of different foodstuffs, particularly meat and dairy, which tend to be the pricier components of the Russian diet, is another interesting topic. According to Rosstat data cited in an article in Kommersant.ru, the type of food consumed in largest per capita quantity is dairy: an average Russian consumes Russia_titular_ethnicity_2010over 200 kg (440 lbs) of it a year. (The most popular type of dairy is 3.2% milk and yoghurts.) Meat, however, takes the third place in the Russian diet: an average Russian citizen consumes 75-80 kg (165-176 lbs) of meat annually, which is less than the average annual consumption of bread and other grain-based foods. Russia_Percentage_ethnic_Russians_2010However, there are significant differences in the amount of meat and dairy consumed in different regions. For example, residents of Kalmykia consumed more than twice as much meat per capita as residents of Dagestan (114 kg vs. 40 kg). As for dairy, per capita consumption in Tatarstan is more than 3.5 times greater than Chukotka.

The geographical patterns of meat and dairy consumption can be explained only in part by economic factors, as they seem to correlate more closely with culinary traditions. For example, higher meat consumption correlates well with the presence of traditionally semi-nomadic, Turkic- and Mongolic-speaking peoples: Kalmyks (Mongolic), Sakha (Turkic), and smaller Turkic-speaking groups in Altai Republic. As can be seen from the map of ethnic composition, these regions have substantial populations of their titular ethnicities and lower percentages of ethnic Russians. But this pattern does not work elsewhere; thus, Chuvashia, Tatarstan, and Tuva also feature a significant Turkic population yet have much lower figures for meat consumption. Economic factors may play a more prominent role here. But economics does not tell the whole story either, as such high GDP areas as the three Autonomous Okrugs (Nenets, Yamalo-Nenets, and Khanty-Mansiysk AOs) and Tyumen oblast have some of the lowest meat consumption figures. I find especially perplexing the low figure in Chukotka—merely 51 kg (112 lbs) per person per year, less than half of the amount of meat consumed by an average Kalmykian—because Chukotka is both economically productive and has a substantial indigenous population, which traditionally lives on reindeer and seal.* It is much easier to explain similarly low meat consumption in Northeastern Caucasus—Dagestan (40 kg), Ingushetia (54 kg), and Chechnya (58 kg)—a region of both low GDP and a culinary tradition of supplementing meat (mostly lamb and goat, as well as poultry) with a lot of fruits and vegetables (for more on the cuisine of different parts of the Caucasus, see here and here).

Russia_Milk_dairy_consumption_2013As for dairy, one finds higher levels of milk consumption in (some of the) steppe regions, including Tatarstan (364 kg per capita per year), Bashkortostan (312 kg), Orenburg oblast (308 kg), parts of southwestern Siberia (esp. Altai Krai, 335 kg, and Omsk oblast, 301 kg), as well as in Sakha Republic (281 kg)—all areas where reliance on milk has been an important feature of traditional cuisine of cattle- and horse-raising semi-nomadic indigenous groups. However, milk has not been a staple for reindeer pastoralist groups: Evens, Evenkis, Nenets, Chukchi—so even today dairy consumption in their traditional areas remains fairly low. Another area which registers higher-than-average dairy consumption is St. Petersburg (315 kg) and the surrounding Leningrad oblast (293 kg), which probably goes back to the high number of dairy-producing sovkhoz (state-owned farms) during the Soviet era.

More perplexing are the relatively low figures of dairy consumption in four neighboring oblasts in north-central Russia: Yaroslavl (246 kg), Tver (243 kg), Vologda (236 kg), and Kostroma (194 kg). These regions are traditionally renowned for their specialty butter (Vologda) and cheeses (Yaroslavl, Tver, and Kostroma), so one might expect higher dairy-consumption figures. Historically, Russians produced and consumed “white” or “farmer’s cheese” but not “yellow” or “hard cheeses”, which first came to Russia from Holland with Peter the Great. According to moloko.cc website, the first cheese-making facility in Russia was opened in 1795 in Tver gubernia (now, oblast) in the estate of Prince Meschersky. The first large-scale cheese-making factory was also opened in Tver gubernia in 1866 by Nikolai Vereschagin (brother of famous artist). Cheese-making then spread to Yaroslavl gubernia, where local specialty cheeses were developed: Yaroslavsky, Uglichsky, Poshekhonsky cheeses (the latter two are named after the towns where they were first made: Uglich and Poshekhonye). In 1878, a first cheese-making facility opened in Kostroma by Vladimir Blandov; according to the Wikipedia, by 1912 Kostroma gubernia boasted 120 cheese-making factories in which a variety of cheeses, including the specialty Kostromskoy cheese, were being produced. Kostroma became an unofficial “cheese-making capital of Russia”, notes Vkusnoblog.net. What, then, explains the decline in local cheese-making and dairy consumption in this area? The answer seems to be Soviet food policy. During the communist era, regional specialty cheeses were turned into standardized recipes, mass-produced in factories all around the country, undermining the local specialization. Since the 1990s, some local artisanal cheese making has been revived, but most small local producers have not been able to complete with larger domestic factories and foreign imports. Kostromskoy and Poshekhonsky cheeses, for example, gave way to imported brie and camembert. Russia has imposed sanctions on the importation of many foreign foodstuffs, but it remains to be seen what effect these measures would have on local cheese production. (I thank Sonia Melnikova-Raich for a helpful discussion of this topic.)

Russia_Physicians_2013The rest of this post examines figures and maps concerning healthcare infrastructure. Overall, Russia ranks very high in physician density and the number of hospital beds per capita, but quite low in nurse density. Regional differences in these indicators are quite pronounced, however, and some of the geographical patterns are rather baffling. For example, unsurprisingly, Saint Petersburg boasts the highest physician density (81.2 physicians per 10,000 population), whereas Vladimir, Tambov, Tula, and Vologda oblasts in central Russia are served by fewer than 35 physicians per 10,000. One might expect Saint Petersburg to lag behind Moscow in this measure, but the figure in Moscow City is actually much lower (68.6). This contrast is probably related to the rapid population expansion in Moscow in the last two decades, something that did not occur in Saint Petersburg (for illustrative population graphs, see here); Moscow’s health infrastructure simply could not keep up with that demands of the growing population. (Saint Petersburg also has more nurses per capita and substantially more hospital beds per capita than Moscow.) Overall, Siberia’s population is served by more physicians per capita than that of European Russia and the southern Urals, although there are exceptions: Khakassia and Jewish Autonomous oblast have fewer than 40 physicians per 10,000, and Kurgan oblast is served by merely 30.2 physicians per 10,000. Another geographical pattern that stands out is the disparity in the concentration of physicians between major cities and their surrounding oblasts (Saint Petersburg: 81.2; Leningrad oblast: 34.5; Moscow City: 68.6; Moscow oblast: 39). Sharp contrasts between neighboring federal subjects are found elsewhere as well: Vladimir oblast (33.9) and Yaroslavl oblast (58), Vologda oblast (34.7) and Arkhangelsk oblast (54.5), Volgograd oblast (48.2) and Astrakhan oblast (65.8), Jewish Autonomous oblast (37.7) and Amur oblast (60.6), North Ossetia-Alania (71.7) and Ingushetia (37.7). The high physician density in Astrakhan oblast and North Ossetia-Alania is perplexing in and of itself.

Russia_Nurses_2013As for nursing personnel, higher nurse density (over than 130 nurses per 10,000 population) is found across the Russian Far North and in parts of the Altai region, which are generally areas of lower population density. The highest figures are found in Magadan oblast (151.3), Chukotka (151.1), and Komi Republic (146.6). In contrast, lower figures (fewer than 100 nurses per 10,000 population) characterize most of European Russia, the North Caucasus region, southwestern Siberia, and the southern part of the Far East. The shortage of nurses is experienced in federal cities (Moscow City: 97.9; Saint Petersburg: 98.4) and even more acutely in the surrounding oblasts (Moscow oblast: 76.7; Leningrad oblast: 73); Leningrad oblast has the lowest figure in all of Russia. Another area where nurses are in short supply is the North Caucasus economic region (Krasnodar Krai: 88.1; Dagestan: 82.1; Ingushetia: 77.1; Chechnya: 73.2). As with physician density, sharp contrasts are observed in some cases between neighboring federal subjects: Leningrad oblast (73) and Karelia (123.6), Samara oblast (91.7) and Ulyanovsk oblast (127.7), Zabaikalsky Krai (114.4) and Magadan oblast (151.3).

Russia_Hospital_beds_2013Finally, Russia ranks 3rd in the world (after Japan and Korea) with respect to the availability of hospital beds per capita; unsurprisingly, these three countries top the charts in terms of average length of hospital stays (an average Russian patient stays in hospital for 13.6 days; compare to 4.9 days in the United States). But yet again, regional variation in Russia is quite pronounced, with 149 beds per 10,000 population in Chukotka, but only 46 beds per 10,000 population in Ingushetia. The availability of hospital beds correlates somewhat with nurse density, though far from perfectly. There are more hospital beds per capita (over 120 per 10,000 population) in Siberia (especially, in Eastern Siberia and the Far East) and in parts of the European North (especially, in Nenets Autonomous Okrug and Murmansk oblast). Besides Chukotka, the highest figures are found in Magadan oblast and Tuva (both 136), and Kamchatka and Sakhalin (both 129). Lower figures (fewer than 90 hospital beds per 10,000 population) characterize much of European Russia, the Mid-Volga region and southern Urals, parts of Western and Southern Siberia, and the North Caucasus region. As with physician density, federal cities have higher figures than the surrounding oblasts (Moscow City: 85; Moscow oblast: 79; Saint Petersburg: 92; Leningrad oblast: 69); however, even the two cities do not boast particularly high figures. As with the other healthcare indicators, sharp contrasts are found between neighboring regions, such as Lipetsk oblast (79) and Oryol oblast (101), or Altai Republic (80) and Tuva (136).

Overall, it should be noted that the per-capita healthcare infrastructure does not correlate with the region’s GDP.** For example, physician density is expectedly high in richer federal cities and in Chukotka, but it is fairly low in other high GDP areas, especially in Nenets Autonomous Okrug. Similarly, nurse density is predictably high in Chukotka, Sakhalin, Yamalo-Nenets and Khanty-Mansiysk Autonomous Okrugs, but surprisingly low in other high GDP areas, particularly in the federal cities and in Tyumen oblast. Likewise, the availability of hospital beds per capita is unsurprisingly high in such rich regions as Chukotka, Sakhalin, and Nenets Autonomous Okrug, but low in others, especially in Tyumen oblast and Khanty-Mansiysk Autonomous Okrug. Nor is there a close correlation between these three indicators. For instance, federal cities are characterized by a high level of physicians per capita but few nurses; conversely, there are few physicians but many nurses in Kurgan oblast. Similarly, there is no correlation between the numbers of nurses and hospital beds per capita: for example, Khanty-Mansiysk Autonomous Okrug and Altai Republic have more nurses than hospital beds (1.83 and 1.69 nurses per hospital bed, respectively), whereas in Primorsky Krai and Tomsk oblast there are fewer nurses than hospital beds (0.83 and 0.93 nurses per hospital bed, respectively).



*Figures for meat consumption refer to “meat and meat products, including offal of category II and raw animal fat”. According to the Wikipedia, “offal of category II” includes heads (without tongues), feet, lungs, ears, pigs’ tails, lips, larynxes, thyroid glands, esophagus meat, and stomachs. Tongues, livers, kidneys, brains, hearts, beef udders, diaphragms, and beef and mutton tails are considered “offal of category I”.

**Of course, quantitative measures of health infrastructure say nothing about its quality. Much has been written (especially, in Russian-language blogosphere) about the pitiful state of many Russian hospitals. Recently, two lethal incidents that happened in the 2nd city hospital in Belgorod have brought this point home. In the first incident, a doctor pounded a patient to death; a video of the incident caught on security camera is rather difficult to watch. Two weeks later, an 84-year old patient fell from a 4th floor window of the same hospital; whether he committed suicide, or was pushed, or whether it was an unfortunate accident remains to be seen.


Mapping Regional Differences in Economic and Social Development in Russia—A GeoCurrents Mini-Atlas Read More »

Using GC Customizable Maps in the Classroom: Population Density in California

The customizable maps that GeoCurrents is releasing to the public have many potential classroom uses, as this post will seek to demonstrate. Manipulating such maps is a good way to learn some of the fundamental elements of cartography, and can be useful as well for gaining basic geographical knowledge. It is one thing to merely look at a map, and quite another to actively engage with it.

My example today concerns the population density of California. With 39 million inhabitants, California is by far the most populous state in the USA, with 12 million more residents than second-place Texas. But California is not particularly densely populated, ranking only 11th on this score among U.S. states. But all such measurements obscure the extremely uneven nature of the distribution of California’s people. Large expanses of the state are almost uninhabited.

California Population Density 1My starting point for a classroom exercise on this matter would be a customizable map of the counties of California coupled with a ranked list of those countries by population density (see the map to the left). I would then ask the students to construct a population density map of California based on this data, using whatever color scheme and break-points in the data that they see fit. If one uses an unlabeled base map, as shown here, students will have to figure out where each county is located before they assign it to a color category. An easier alternative assignment would be to begin with a labeled map. In this case, however, the small counties (such as Sutter, Alameda, or Orange) can be difficult to color, as when one “clicks” on them California County Namesone will actual click on the name-tag rather than the shape. These name-tags, however, can be dragged away and then replaced after the color has been assigned. If the map is filled in without the county names, the labels can be restored by using the “select all” feature on a county-name-only map (posted to the left), copying what has been selected, and then pasting it on to the colored population density map. Black-letter names placed on dark-colored counties will then have to changed to white to make them legible. (The county-name-only map is available for download in the PowerPoint and Keynote files this post, along with several other maps used in this post.)

California Population Density Map 1It would, of course, be useful at some point to provide students with a model of a map constructed in such a manner, as with the one posted to the left. Here one can see how I categorized the raw data, splitting it into 11 groups. I usually try to limit such divisions to no more than nine, for the simple reason that Apple’s Keynote program provides a handy 9X12 color Keynote Color Matrixmatrix, with a 10th row for white-black (see the illustration). In this case, however, the vast differences in the values being mapped—from 1.6 to 3,472—seemed to demand more categories. I thus used black for the highest category, and for the lowest I took the lightest color in the sequence California Population Density Map 2and set the opacity to 50%, lightening it still further. An alternative for a map with numerous categories is to use a two-color scheme, as illustrated to the left. I prefer this kind of map whenever the value range is large, as it emphasizes extremes. Many people tell me, however, that they find such maps difficult to interpret. Another way to signpost extremes is to provide the actual figures on the map itself, which I have done here, albeit using total population rather than density. As can be seen, the populations of California’s counties vary by roughly four orders of magnitude, ranging from a little over 1,000 in Alpine to more than 10 million in Los Angeles.



California Population Density Map 3Once the population-density-by-county mapping exercise has been completed, it is important to point out its limitations. This step can be accomplished by contrasting the maps that the students have made with a map showing population density at a much finer level of analysis. As can be readily appreciated by viewing the map posted to the left, it is somewhat misleading to focus on overall population density figures for many if not most California counties. This issue is especially acute for a huge county such as San Bernardino, which is larger than nine U.S. states. As can be seen on the map, the vast majority of San Bernardino’s two million residents live in the far southwestern corner of the county. The rest of the county is sparsely inhabited desert.



Ventura County TopographyMany California counties are characterized by such highly uneven population distribution patterns. Los Angeles County, for example, has an extraordinarily sharp north-south population gradient. In neighboring Ventura, almost no one one lives in the entire northern half of the county. In most cases, such patterns are easily explained on the basis of physical geography. To illustrate this, I placed an expanded section of the population density map showing Ventura County hidden below a depiction of the Ventura County Population Densitycounty’s topography. I then roughly outlined (in black) the flat areas of the county, something that students could easily do in a classroom exercise. If the opacity of the topography map is then set to zero, as in the last map posted here, the population density map is revealed, showing the correlation between settlement patterns and landforms.

But even on the higher resolution population density map, certain mysteries remain. Why, for example, would population density abruptly jump just to the north of Ventura County in an equally rugged section of Kern County? The answer here, again, is found in the degree of resolution. This map may seem to be finely divided, but in actuality in entails a large degree of spatial aggregation. Such aggregation is evident in the fact that the color categories often follow county boundaries exactly, which is almost never the case in regard to the actual distribution of the population. The main lesson here is that no map can ever be perfect unless it is made on a 1:1 scale, in which case it would also be perfectly useless.



California Population Customizable Maps (PowerPoint)

California Population Customizable Maps (Keynote)

Using GC Customizable Maps in the Classroom: Population Density in California Read More »

Customizable Maps of the United States, and U.S. Population Growth

New sets of customizable maps of the United States are now available for download, in both PowerPoint and Keynote formats (see the end of this post). Three maps are included in each presentation-software set. The first simply has the outlines of the states, as well as the District of Columbia (note that Alaska and Hawaii are mapped out-of-scale and in the wrong locations). The second map includes the names of the states as well. The names of the small states of the northeast are placed near but not on the state shapes, due merely to a lack of room. (I did not include lines linking the names to the shapes, as this did not seem necessary.) The final map in each set includes the physical-political map that was used to make these customizable maps. The outlines of each state have been traced in here, and hence can be easily manipulated. The three maps in each set are also located at the same place on each screen. As a result, as long as the maps are not moved, elements can be transferred from one map to another and remain in the correct location.

USA 2010-2015 Population Change MapTo illustrate what can be done with these customizable base maps, I have made a map of population change in the United States from April 2010 to July 2015, based on data that was just released by the U.S. Census Bureau. As can be seen on the map, West Virginia was the only state to lose population during this period, largely as a result of the continuing decline of coal mining. North Dakota, on the other hand, shows the largest proportional gain, due mostly to the fracking boom in oil and natural gas. The phenomenon, however, has virtually come to an end, owing to the drop in the global price of oil.

Population growth in the United States continues to be strongly geographically pattered. New England saw little expansion, with Vermont adding only 297 residents. Massachusetts saw the largest absolute and proportional growth in this region, although its 3.8% growth rate over the period was lower than that of the United States as a whole, which came in at 4.1%. The Mid-Atlantic States also saw little growth, with the exception of Delaware and the District of Columbia (Washington D.C.). New York State did add over 400,000 residents, but its growth rate was only 2.2%. Below-average growth also characterized the Great Lakes states. Here the fastest growing state, Minnesota, saw it population expand by only 3.5%.

The South Atlantic states registered strong population growth, with Florida alone adding almost a million and a half residents. Growth was much slower in the south-central states, with the population of Mississippi expanding by less than one percent. Further west, Texas posted the largest absolute gain, at 2.3 million, and the third largest proportional rate of expansion. Growth was moderate in most of the Great Plains, with the exception of booming North Dakota. The West experienced relatively rapid growth overall, but with significant variability from state to state. Colorado, for example, grew by 8.5%, whereas New Mexico, a much poorer state, grew by only 1.3%

Red States Blue StatesOverall, states that usually support the Republican Party grew faster than states that generally support the Democratic Party. One exception here is the Pacific Coast, a strongly “blue” (or Democratic-voting) region that exhibited faster-than-average growth. Several “purple,” or swing-voting, states also expanded rapidly. Examples here include Florida, Nevada, and Colorado.

Click on the links below to download the customizable map sets:

USA Customizable Map (PowerPoint)

USA Customizable Map (Keynote)


Customizable Maps of the United States, and U.S. Population Growth Read More »

The Recent Gilbertese Settlement of the Line Islands

Map of KiribatiIt is difficult to convey the immensity and emptiness of the Republic of Kiribati. The country extends across more than 3.5 million square kilometers (1,351,000 sq mi) of oceanic space, an area considerably larger than India. The distance between its western and eastern islands is comparable to the distance across the United States. Yet Kiribati contains only 800 square kilometers (310 sq mi) of land, an area slightly larger than the city-state of Singapore, and considerably smaller than the city of Los Angeles.

With some 103,000 inhabitants (in 2010), Kiribati has only a moderately dense population. Its 135 people per square kilometer of land (350 per sq mi) places it in the 73rd position among the world’s 244 sovereign states and dependent territories, well below Italy, Germany, and the United Kingdom. But such figures are misleading, as the population of Kiribati is by no means evenly distributed over its far-flung expanse. Most of its residents live on the 16 atolls of the Gilbert Archipelago in the west. Here more than 85,000 people are crowded into roughly 281 square kilometers of low-lying land. On the main atoll of Tarawa, a fast-growing population of some 55,000 is limited to 31 square kilometers. Tarawa’s highest point is three meters above sea level.

The Phoenix Islands in central Kiribati are a different matter. These islands contain 84.5 square kilometers of land but their population is negligible. Only Kanton Island is inhabited, and its population in 2010 was all of 24, having declined from 61 in 2000. Although archeological remains indicate that some of the Phoenix Islands had once been settled, they had no human inhabitants when they were first sighted by Europeans. In the late 1930s, major efforts were made to populate the Phoenix Islands, described by the Wikipedia as “the last attempt at human colonisation within the British Empire.” Imperial agents wanted to reduce over-population in the southern Gilbert Islands and to forestall potential U.S. attempts to gain territory under the Guano Islands Act. The colonization project was abandoned in 1963, however, with the settlers returning to the Gilbert Islands. Low prices for copra, the islands’ only significant export, along with recurrent drought, undermined the scheme.

Pacific Rainfall MapDrought is also a problem in many of the Line Islands, which form Kiribati’s eastern archipelago. The central and southern Line Islands fall within the low-rainfall region of the eastern Pacific, which is generated in large part by the cold waters of the Humboldt Current that are deflected into the equatorial region by the shape of the South American landmass. Partly as a result of meager and uncertain precipitation, the Line Islands—which themselves stretch over 2,350 kilometers of sea-space—were also uninhabited at the time of European discovery. They were all claimed by the United States under the Guano Islands Act, but the U.S. relinquished all of its claims in 1983, with the Congressional ratification of the Treaty of Tarawa (more formally known as the “Kiribati, Treaty of Friendship and Territorial Sovereignty, September 20, 1979,” with a subtitle reading “Treaty of Friendship Between the United States of America and the Republic of Kiribati”).

Line Islands Population growthToday, three of the Line Islands are inhabited: Teraina, or Washington Island, with 1,155 residents (2005), Tabuaeran, or Fanning Island, with 2,539 residents (2005), and Kiritimati, or Christmas Island, with 5,115 residents (2005). The populations of all three islands have increased rapidly in recent decades, owing in part to official Kiribati relocation schemes designed in part to reduce crowding in the Gilbert Islands. Teraina Island was first settled after WWII through the agency of the Burns Philip Copra Company, but its population expanded significantly only after 1990. As noted in an official publication of the Republic of Kiribati:

The population of Teeraina in the 2010 census was 1,690. Compared to the 2005 population of 1,155 and the 2000 population of 1,087, the population is growing very rapidly. The population of Teeraina grew by 535 people between 2005 and 2010, an annual population growth of 7.9%. In percentage terms, Teeraina is the fastest growing island in Kiribati, although the growth is much less significant in terms of absolute numbers.

Kiritimati Island MapGrowth on Kiritimati Island, the giant of not only the Line Islands but also of Kiribati as a whole, has been even more dramatic, its population jumping from 3,431 in 2000 to 5,115 in 2005. With some 388 square kilometers of dry land, Kiritimati Island has a larger terrestrial extent than any other coral atoll in the world.

The vast majority of the people now living in the Line Islands are originally from the Gilbert Archipelago and speak Gilbertese, also known as the Kiribati Language.* As noted in the Wikipedia, “Unlike many in the Pacific region, the Kiribati language is far from extinct, and most speakers use it daily. 97% of those living in Kiribati are able to read in Kiribati, and 80% are able to read English.” Gilbertese has also expanded into nearby countries, with some 5,000 people speaking the language in the Solomon Islands and perhaps another 1,000 doing so in Vanuatu.

Kiribati Fertility LevelThe expansion of the Gilbertese-speaking area has been driven largely by population growth. Although Kiribati’s fertility rate has declined in recent years, it is still well above the replacement rate. Given the small size of the Gilbert Islands, migration to other areas is not surprising. The government of Kiribati is also keen to establish permanent populations in its distant islands in order to cement its control over areas that had been tenuously held. Kiribati also views Kiritimati Island as a potential refuge in the event of a catastrophic sea-level rise. The European Union agrees, and has thus pledged substantial development funds. As noted in a February 2015 article in Radio Australia:

The European Union has just announced a 23 million Euro grant for Kiribati, money that will be used to develop the nation’s largest atoll, Kiritimati Island.

Although it makes up 70 per cent of the country’s landmass, Kiritimati Island was virtually uninhabited for decades and is relatively undeveloped.

By improving facilities on the island, the EU Ambassador to the Pacific Andrew Jacobs says the aim is to reduce the threat posed by climate change to the main population centre of Tarawa.

Kiribati is also concerned about its elevated fertility level. As noted in a World Culture Encyclopedia article on the country:

Population has been growing rapidly since the early 1900s, and overpopulation is a serious concern of the government. While family-planning methods were introduced in 1968 and are delivered free, fertility remains moderately high and large families are culturally valued. Despite government efforts to maintain and improve life on the outer islands, there has been substantial migration to the capital on South Tarawa. There are several thousand I-Kiribati in other countries, most serving as temporary workers.

* The word “Kiribati” itself means “Gilberts” in the Gilbertese language. It is unclear what the indigenous word for the archipelago is, although “Tungaru” has been suggested. Note that in Gilbertese orthography, a “ti” is pronounced as an “s” sound, giving the preferred pronunciation of the country as “keer-ə-bahss.” By the same token, the island of “Kiritimati” is rendered as “kuhris-muh s” or “kəˈrɪsmæs,” the local pronunciation of “Christmas.” It is unclear why the “s” sound in Gilbertese is written as “ti”; I dimly recall reading that the first missionaries brought a typewriter with a broken “s” key, and hence used “ti” as a substitute symbol, but I have been unable to find confirmation


The Recent Gilbertese Settlement of the Line Islands Read More »

The Rural/Urban Divide in Catalonia’s 2015 Election

According to most media sources, the Catalan independence movement scored a major victory in the September 28 regional election, taking 72 out of 135 seats in Catalonia’s parliament (Parlament de Catalunya). More careful reporting, however, noted that the results were actually mixed. In terms of the popular vote, candidates advocating independence gained the support of less than half of the electorate. Had the vote been an actual plebiscite on soverienty, skeptics argue, the motion would have been defeated. But Artur Mas, the leader of the independence movement, offered a different interpretation, claiming that “the Catalan people have spoken”—and have spoken for independence. As he put it, writing in The Guardian:

On 27 September Catalonia’s voters went to the polls and with a record 77.4% turnout gave a win in every single electoral district to the political forces whose campaign promise was, if elected, that they would follow a “roadmap” towards Catalan independence from Spain. Pro-independence lists obtained 48% of the votes and 72 seats out of 135, whereas unionist lists got 39% of the votes and 52 seats. These plebiscitary elections were the only way possible to give the Catalan people the vote on the political future they have long called for, after the Spanish government’s longstanding refusal to allow an independence referendum.

The fact that the pro-independence vote and the Spanish-unionist vote together fall well short of 100 percent indicates the presence of a third option, that of enhanced regional autonomy without actual sovereignty. But this third “regionalist” option, which rests on a mixed sense of Catalan and Spanish identity, was favored by relatively few voters. According to a recent Politico article, this “middle ground” lost support in part “because the campaign was not based on a rational debate on whether it makes economic sense to have full fiscal autonomy or leave the EU, the eurozone or NATO. Rather, it pandered to nationalistic feelings and prejudices…”


Catalonia 2015 Election MapAs mentioned in an Economist article, the pro-independence parties were able to gain control of the regional parliament without winning an outright majority due to “Catalonia’s unequal voting system, which favours less-populated rural areas.” The uneven electoral geography of the contest is clearly evident in a series of maps, posted on the website Saint Brendan’s Island, that show the percentage of the vote taken by the top six parties in each comarca (administrative division). I have amended these maps slightly by providing a crude characterization of the political philosophy of each of these groups (in red), along with their percentage of the vote across Catalonia. The leading contingent, an electoral coalition called “Together for Yes” (Junts pel Sí), is marked as “big tent” on the map because its constituent parties span a fairly wide range of political positions, falling both to the right and the left of center. The much less popular Popular Unity Candidacy party also favors Catalan independence but is situated too far to the left to have joined the “Together for Yes” coalition.


Catalonia Population Density Election MapThe second illustration, which juxtaposes a population density map with an expanded map of the “Together for Yes” vote, clearly shows the urban/rural electoral divide in Catalonia. The region’s most densely populated areas in general gave relatively little support to the independence movement, favoring instead the unionist and regionalist parties. One factor here is the presence of many migrants from other parts of Spain, who not surprisingly tend to support the unionist cause. In Barcelona, Spanish (or Castilian, as most Catalan nationalists insist) is the main language, and although three-quarters of the city’s inhabitants can speak Catalan, fewer than half are able to write in the language. Similar situations are found in the other major urban areas of Catalonia. As noted in the Wikipedia article on the historic city of Lleida: “After some decades without any kind of population growth, it met a massive migration of Andalusians who helped the town undergo a relative demographic growth.”


The Rural/Urban Divide in Catalonia’s 2015 Election Read More »