Geo-Ethnic Polarization

A comparative analysis of Eurasia and the west.

According to the CIA Factbook, China is 91.1% Han Chinese.¹ Western dissidents praise China for its homogeneity, and contrast it with America and Europe, where mass immigration has made Gen Alpha a “majority minority” generation.² How accurate is this picture?

In the last few years, it was revealed that China over counted their population by 100 million, or 7% of their population.³ This “data massage” has occurred as recently as 2022.⁴ Why is China lying about their demographics?

Firstly, Chinese local officials have an incentive to pump up the numbers. The greater their population, the greater funding they receive. Secondly, China has a national interest in obscuring its population decline. International investors know that a high fertility rate combined with a highly skilled labor force has a higher potential for future growth than a population in decline. Countries like Japan have suffered economically from a decline in fertility, and mass immigration is often justified on the basis that it will replenish the labor force. If China hides its decline, it becomes more attractive to foreign investment.

China is lying about its quantity of residents, but what about their qualities? Is it possible that China is undercounting its ethnic minorities in a bid to appear more homogeneous, less diverse, and therefore more attractive to investors? If China was being deceptive, how would they hide internal divisions and present a united front?

The frequently touted figure, that 92% of Chinese residents are “Han,” refers to the fact that 92% of Chinese residents speak one of the Hanyu languages. The term Hanyu means “languages of the Han.” The largest Hanyu language is Mandarin, which is the first language of 65% of Hanyu speakers.

When we multiply the percent of China that is Hanyu by the percent of Hanyu who are natively Mandarin, the resulting percentage is 60.4%.

Alternatively, if we accept the Chinese statistics that there are 940 million Chinese Mandarin speakers, and additional 200 million who speak Mandarin as a second language, then the total percentage of Mandarin speakers ranges from 66.7% to 80.9%. This is in line with other statistics claiming that between 70% and 80% of Chinese residents speak Mandarin.

By comparison, 78.3% of American residents speak English at home, and 91.8% of American residents speak English “very well.” On a purely linguistic basis, this makes America more “homogeneous” than China.

China is also more Balkanized than America. Balkanization refers to the geographic concentration or geographic polarization of ethnic minorities, as opposed to geographic diffusion. This can be represented mathematically as follows:

1. Select a given ethnicity.

2. Through a method of trial and error, draw various borders which subdivide the country. Quantify the complexity of the border as C.

3. Find the border where C, the complexity, is the lowest, while P, the percentage of that ethnicity, is highest.

4. The formula for Balkanization is as follows: B = P - C.

This would be the most exact way to measure Balkanization, but it would require a method of drawing lines, measuring the complexity of those lines, and calculating the resulting percentages. By complexity, we mean that if several disconnected lines need to be drawn (such as in the case of isolated ghettos, or islands), this creates complexity and reduces Balkanization. Conversely, the “simpler” the shape, the more Balkanized the country.

The states with the greatest percentage of Black Americans, for example, are Mississippi, Louisiana, Georgia, and Maryland. These states are 32% to 38% Black.

There are 104 counties in the United States which are at least 50% Black. Since these counties are concentrated in the south east, they are on average smaller in area than counties in the north or west. However, even with the acknowledgement of this overestimation, this means that, at most, 3.3% of American land is majority Black. This is a far different number from the percentage of Blacks, which is 14%.

Using these two numbers, we can create a simpler metric of Balkanization:

1. Assess the total number of provinces, states, or counties which are “majority minority.” Since this has to do with geography and land area, call this A.

2. Divide A by T, which is the total land area of the country. A/T is the percentage of land that is “majority minority.”

3. Finally, divide A/T by M, where M is the total percentage of the country of that ethnicity.

4. The simple formula for Balkanization is: B = (A/T)/M

When an ethnic minority has a majority in a disproportionately large area of a country, this increases B. When an ethnic minority is evenly distributed throughout a country, this can bring B to 0, if there are no areas where that minority forms a majority. Following this simple formula, the value of Balkanization for Black Americans is 0.23.

To see how Balkanization values could increase due to migration, consider two “ethno-state” scenarios:

1. The “Little Africa” Scenario: Take the states of Mississippi, Louisiana, Alabama, South Carolina, Virginia, North Carolina, Georgia. All together, these states have a population of around 49 million, which is slightly more than the number of Black Americans, around 48 million. The total land area of these states is roughly 9% of the total area of the United States. If all Black Americans migrated to this area, and all white Americans migrated out of this area, then B would be .09/.136, or 0.66.

2. The “Big Africa” Scenario: If Black Americans were granted an ethno-state which also included Texas, Florida, Missouri, and Arkansas, then A/T would be 25.6%, and M remains 13.6%. Thus, B would equal 1.88.

Like Black Americans, Hispanic Americans fail to reach a majority in any American state. Furthermore, in New Mexico, which is 48% “Hispanic,” the Hispanic identity can be broken down further, with the largest ethnicity being Mexican, at only 32%.⁵

Ignoring this breakdown between Hispanics and Mexicans, and assuming the cohesiveness of a “Hispanic ethnicity,”⁶ we can again count the number of counties with a Hispanic majority. This comes out to 179 total, although 78 of these are in Puerto Rico, meaning that only 101 counties in the mainland are Hispanic majority.

Many of these counties are in the American southwest, which is largely a desert. However, in this analysis, no attempt is made to differentiate between “good land” or “bad land” when assessing Balkanization.

Similarly to Black Americans, the A/T of Hispanic Americans is only 3.2%. Given that the total number of Hispanics is 18.9%, this results in a B value of 0.17, which is smaller than the Black Balkanization value of .23. This implies that Black Americans more Balkanized than Hispanic Americans, and Hispanic Americans are more distributed throughout the population.⁷

This slightly contradicts public school data, which show that Hispanics are more likely to go to a Hispanic majority school than Blacks. However, part of this may be because of bussing and desegregation efforts and programs which began in the 1960s and which are still in effect nationwide. By contrast, there may be less effective policies focused on "bussing" Hispanic students.⁸

Asian Americans are more diverse than Hispanics in that they have no common religion (Hispanics are generally Catholic or Protestant), no common language, and no common culture. Chinese Americans tend to concentrate in New England (especially Maine), Filipinos in the west coast (especially southern California), and Vietnamese in south-central (especially Louisiana).

Asian Americans make up 56.6% of Hawaii, but this classification of “Asian” includes “Asian and Pacific Islander,” the largest of which is native Hawaiian, at only 21.8%.

The roughly 20 million Asian Americans are general divided into Chinese Americans (5 million), Indian Americans (4.3 million), and Filipino Americans (4 million), with the remaining 7.7 million belonging to other groups.

Only 9% of Asians claim that Asian Americans belong to a “common culture.” Only 72% of Asians say that they look Asian/Chinese, with 20% claiming to appear Hispanic, White, Middle Eastern, or Multi-racial. This effect increases over time, with 3rd generation Asians only appearing 47% Asian/Chinese, and 39% claiming to be perceived as Caucasian, Hispanic, or Multi-racial. Finally, only 34% of Asians use the word “Asian” to self identify, with the majority preferring ethnic terms like “Japanese,” or “Japanese American.”

Considering these facts, it would be inaccurate to claim that “Honolulu has a majority Asian identity,” since the category of “Asian and Pacific Islander” is not an effective or cohesive identity. Outside of Honolulu, there are no Asian majority counties in America, and therefore, Asians have a Balkanization value of 0.

There are 7 states where whites are a minority: Hawaii, California, New Mexico, Texas, Nevada, Maryland, and Georgia. Adding these up and subtracting them from the total area (A/T) means that whites hold majorities in 78.9% of land in America. This is similar to the number of white students who report attending white majority schools.⁹ If we divide this by M=59.3%, the total percent of whites as a share of population, the value of white Balkanization is 1.33.

By totaling Balkanization values for all subgroups, and running simulations, we can compare theoretical values of Total B. Max Balkanization with 4 subgroups (a situation where each group has exactly 25% of the land and 25% of the population) produces a total B percentage of 400%. Therefore, given X number of groups, the maximum B/X=1+X/100. This is how the Bifurcated B/X = 1.02, while “Max Balkanization” is only 1. Conversely, if we look at the “Homogenous” simulation, we can see that B/X approaches 0 as X becomes large.

One criticism of this model of Balkanization is found by looking at the multi-diffusion model. In this model, each land area, town, or city is perfectly divided between four groups. Such a city would have no areas or neighborhoods which had any clear majority. In this model, such a city would have a B/X of 0. In reality, humans self segregate at the level of cities. What this measure of Balkanization is meant to measure is national self-segregation on a geographic basis, which is relevant for potential ethnic conflicts.

In order to adjust for the low B values of the multi-diffusion model, the maximum group was selected (in this case, white), then, it was added to the B/X value to make B/X+m, where “m” stands for “max.” This B/X+m was then normalized by 0.67 so that the homogenous model would produce a B/X+m value of 0.

B/X+m values were as follows:

Still, the multi-diffusion model continued to have a lower B/X+m value than the mono-diffusion model.

China’s Balkanization

Now that actual and theoretical B/X+m values have been calculated for America, it will be possible to compare them to China.

First of all, China has 117 autonomous counties and three autonomous banners, out of a total number of 2,851 county-level divisions. Without accounting for the variation in the size of counties (autonomous counties tend to be bigger, with the autonomous banners being the biggest), and without accounting for which ones have “good land” and which ones have desert, it can be estimated that 4.1% of the land in China is minority-majority. This is not very different from America, where 6.5% of land is minority-majority.

One comparison that could be made is between China’s ethnic minorities and native Americans. Native Americans are assigned the equivalent of “autonomous reservations,” and own 2% of all American land, collectively. However, the reason for using “counties” rather than “acreage” is because counties with worse land tend to be bigger, while counties with better land tend to be smaller. As a result, native Americans only have 28 counties out of 3,143, which is less than 0.89% of all counties.

If the Chinese A/T is 0.041, and the number of non-Mandarin is 20%, then the Chinese B value is 0.82

Russia’s Balkanization

Russia has 85 Federal Subjects, and out of these, 24 are Republics, which have their own constitution, and national language; 1 is an autonomous oblast; and 4 are autonomous okrugs. All together, that means out of the 85 Russian divisions, 29 are designated as belonging to majority-minorities, which gives Russian minorities an A/T of 0.34. If the number of non-Russkiye is 20%, then the B value of Russia is 6.8.

The final scores show interesting results. By pure B values, America has not become any more Balkanized than it was in 1960. This is due to the fact that Black Americans have not become any more geographically concentrated than they were in 1960, and Hispanic Americans tend to diffuse throughout the continent, being attracted to cities, rather than migrating solely to the south west. Due to the greater geographic concentration of both Russian and Chinese ethnic groups, those countries can be described, under a certain perspective, as more Balkanized than America.

B/X values have not changed in America since 1960, and by this measure, America is slightly less Balkanized than China, but one third as Balkanized as Russia.

B/X+m values paint a different story. When we account for “how dominant is the titular nation?”, represented by “m,” we find that America has become twice as Balkanized since 1960. Interestingly, America in 1960 was actually less Balkanized than China. Finally, Russia is twice as Balkanized as America.

Summary

It is simple and easy to measure ethnic proportions within a population. This allows us to assess how diverse a country is, which according to Charles Murray, can lower social trust. However, Balkanization describes a separate phenomenon. Balkanization, or Geo-Ethnic Polarization, (GEP) describes the level of concentration of ethnic distinction, as it is distributed over the land area of a country.

GEP is important because in the classic case of the Balkans, it allowed for the war between Serbs, Croats, Albanians, Bosnians, and so on. If those populations were completely intermixed, without any one area having a predominant majority, the Yugoslav War would have not been possible.

GEP also enabled Czechoslovakia to split cleanly and peacefully into two states: Czechia and Slovakia. While this may sound like a good thing, it is generally not good for great powers when their provinces being to separate. Czechoslovakia was never going to be a major power in Europe, and it has historically been dominated by the Soviet Union, and now is dominated by Germany via the EU. China or Russia on the other hand would have a lot to lose from a “velvet revolution.” This is why Russia cannot allow for the loss of Ukraine or the loss of Chechnya, and China cannot allow for the independence of Tibet or Xinjiang. It was similar reasoning that compelled America to resist Confederate separation in the Civil War. Small states may be able to peacefully split apart, but for large empires, this is a threat to the legitimacy of the state, and cannot be tolerated.

The clear and present danger of GEP explains why China is actively flooding Tibet and Xinjiang with Han Chinese. If China is successful in this policy, it will reduce GEP and the threat of independence movements.

There are many ways to measure ethnicity, and one of the oldest is by language. The Greeks referred to foreigners as barbarians because “barbar” was the Greek onomatopoeia equivalent to “blah blah.” By the measure of language, the west is much more united than Eurasia.

Far right dissidents theorize that social trust is determined by racial homogeneity. Therefore, they surmise, China will win over America, because China is all Asian, whereas America is racially diverse. However, historically, America and the European Empires since 1492 all had colonies or territories which were non-white, and they were able to dominate “racially pure” China. The question, then, is less about how “racially pure” a state is, and whether it is able to effectively manage internal ethnic conflict.

In other words, mass immigration is creating a quantitative strain on the American system, but qualitatively, if America is able to assimilate or neutralize foreign ethnic identities, it will remain competitive against systems which antagonize or incite their minorities. America, historically, has succeeded in neutralizing or ameliorating the ethnic distinctions between Anglos and Irish, Italians and Germans, Greeks and French. Will Russia and China achieve similar feats, integrating their minority populations? Furthermore, can Russia and China achieve a similar level of cross-cultural admiration and mutual understanding as compared with America and Europe?

Again, on a purely racial basis, America and Europe have much more in common than Russia, Iran, and China. If racial homogeneity determines the ability to cooperate, then America may degrade in its internal capacities, but it should not become alienated from Europe. By contrast, the alliance between Russia, Iran, and China is purely contingent on their mutual hatred of America, and there is no ethnic or genetic ties between the three countries, let alone religious, cultural, or linguistic ties.

While it is possible that the loss of ethnic homogeneity will increase social alienation and decrease social trust, it is not likely to impede American and European diplomats from working together as a geopolitical block. Meanwhile, in China and Russia, if the ethnic theory of cooperation is accurate, that alliance is fragile and ultimately doomed.

The irony of the pseudo-Duginist “nationalism for all peoples” is that it was used by JFK and liberal anti-communists to use as propaganda against the Soviet Union. It was earlier used by FDR in World War II, and by Wilson in World War I. It is a fundamentally liberal idea, based on ideas of sovereignty, national independence, anti-imperialism, and human rights. What is the right to national self-determination if not the purest expression of universal “human rights”?

Eurasianist or multi-polar propaganda which promises to liberate the world from American hegemony operates on the same logic as anti-racist propaganda which promises to liberate the world from whiteness. Remove the center, and it cannot hold. Every system, every empire, and every state needs a central ethnic, ideological, religious, linguistic, and cultural core. No matter how diverse a country is, if it has an elite core (such as the Normans in Britain), the petty squabbles of diverse groups can be overcome. What is necessary is a united vision from the top. China and Russia, by contrast, cannot come together with a united leadership or a common vision. The lack of a common vision was also what destroyed the American-Soviet alliance after the defeat of Germany.

China and Russia have myriad other flaws besides GEP. Low birthrates, low relative verbal IQ, and an aging population paint a bleak future for the Eurasian alliance. In spite of this, Eurasianists claim that Russia and China are homogenous countries, and this cultural unity will give them the advantage over the west and its mass immigration. However, Russia’s immigration rate is 50% of America’s — are Islamic Kazakhs really much easier to integrate than Catholic Mexicans? The difference between America and Eurasia is not black and white, but a question of degree. GEP, like all other problems of administration, can be counted as a liability, but it is not the only factor to be considered.

Lastly, it would be relevant for future analysis to expand the concept of GEP to GIP — Geo-Identitarian Polarization. This includes, for example, political ideologies. Before the Civil War, America had high levels of GIP, with slave owners dominating the south, and abolitionists agitating in the north. High levels of geographic polarization made war much more likely. By contrast, Democrats and Republicans today are diffused throughout America. For example, the state with the highest percentage of Republicans, Wyoming, has 57% who identify or lean Republican.¹⁰ The state with the highest percentage of Democrats, Vermont, has the same exact percentage who identify or lean Democrat. After Vermont comes Massachusetts, which just had a Republican governor last year. Even in Vermont, Trump won 31% of the vote in 2020.

By contrast, take for example the 1948 election. This was just 3 years after World War II, when the country was supposedly demographically homogeneous and united by the war effort. Yet in Mississippi, 87% of voters went for Strom Thurmond. Those sort of landslides seem impossible today, where even Wyoming only went 70% for Trump. College education seems to play a role in creating a wide geographic spread of ideology. Russia and China don’t deal with political polarization, but the fragility of one-party rule may end up being a disadvantage.

The American model is to enforce ideological consistency in the heart, through effective propaganda and moralization. The Russia and Chinese model is to enforce loyalty through the head, by punishing dissent. Which is more fragile?

One of the greatest American political achievements of the last 60 years has been to significantly erode GIP, which was a long-standing issue going back to the American Civil War. Although polarization seems to be increasing, due to lack of GIP, a civil war is less likely than before.

Although mass immigration may be too late to turn back, America can still reduce GEP and GIP through strategic urban planning. America could reduce geographic polarization in rural areas by building new universities and offering free or reduced housing to minorities in those areas. At the same time, America could reduce geographic polarization in its cities and academia by abolishing DEI-style affirmative action policies, and removing welfare benefits from urban areas. These strategies for geographic diffusion are worth discussing further in another essay.

1. https://www.cia.gov/the-world-factbook/countries/china/

2. https://www.theguardian.com/us-news/2023/aug/08/gen-z-americans-white-majority-study

3. https://reuters.com/world/china/researcher-questions-chinas-population-data-says-it-may-be-lower-2021-12-03/

4. https://www.project-syndicate.org/commentary/chinese-population-smaller-than-stated-and-shrinking-fast-by-yi-fuxian-2022-07

5. https://data.census.gov/table/ACSDT1Y2019.B03001?q=B03001:%20HISPANIC%20OR%20LATINO%20ORIGIN%20BY%20SPECIFIC%20ORIGIN&g=040XX00US35&hidePreview=true&tid=ACSDT1Y2019.B03001

6. Cuban Americans and white Hispanics vote very differently from Mexican Americans, and have a very different identity from Puerto Ricans, El Salvadorians, Ecuadorians, Brazilians, Colombians, and so on. It is likely that the current “Hispanic” block will continue to differentiate and fragment, self-sorting based on assimilation into the larger white population. Furthermore, in Mexico itself, “39.4% de los hombres se identifican con los tonos de piel más claros”: https://www.conapred.org.mx/documentos_cedoc/21_Marzo_DiaIntElimDiscRacial_INACCSS.pdf

7. If we account for the percentage of Hispanics which are Mexican, 61.6%, then the percentage of Mexican Americans is only 11.7%. There is a census dataset for Mexican majority counties, but I was not able to figure out how to access the table: https://www.socialexplorer.com/blog/post/texas-dominates-list-of-counties-with-highest-percentage-of-mexican-american-population-according-to-recent-survey-celebrate-cinco-de-mayo-with-social-explorer-8217-s-mapping-tools-13201

8. https://www.pewresearch.org/short-reads/2021/12/15/u-s-public-school-students-often-go-to-schools-where-at-least-half-of-their-peers-are-the-same-race-or-ethnicity/

9. https://www.pewresearch.org/short-reads/2021/12/15/u-s-public-school-students-often-go-to-schools-where-at-least-half-of-their-peers-are-the-same-race-or-ethnicity/

10. https://www.pewresearch.org/religion/religious-landscape-study/compare/party-affiliation/by/state/