Counties expelling Blacks after ‘Brown v. Board of Education’

Reading the late James Löwen’s Sundown Towns: A Hidden Dimension of American Racism reveals that, little-known to the casual historian, the vast majority of the rural North, essentially all the rural and even small-city West, and large parts of the nonplantation South completely excluded African-Americans. Some of this has been discussed on previous posts in this blog here, here and here; however, for this post my goal is to see how many counties had their black population expelled after Brown v. Board of Education. Löwen used that fact to show that racial attitudes in white rural America did not really change after 1940, although election statistics suggest that at least for a while opposition to the civil rights movement of that era was weak in the rural North and nonplantation South.

In my older post about attempting to correlate sundown status with long-term GOP voting in the nonplantation South and some parts of the Midwest, one does see at least one county — McCreary County in Kentucky— that definitely expelled its black population following Brown v. Board of Education. For this post I have looked at the nonplantation South — defined, as in older posts, as the seven states of Texas, Oklahoma, Arkansas, Missouri, Tennessee, Kentucky and West Virginia — and examined:

1. all counties in the nonplantation South which had at least eleven blacks in 1930 but fewer than five black households in 1970
• a few counties in Kentucky and West Virginia discussed in earlier posts will not be covered as doing so would be repetitive
2. other counties in the nonplantation South which had fewer than 25 black households in 1970 and which show evidence of major declines after 1930
3. three counties in western Virginia — a state belonging mainly to the plantation South— where I have gathered via population figures strong evidence for the expulsion of blacks after Brown

I have assumed that if a county expelled its black population as a direct result of the Brown v. Board of Education decision:

1. it would show a very large fall in its black population between the 1950 and 1960 censuses
1. I have used a cut-off of 50 percent fall between 1950 and 1960 for any likely or possible case
2. I have also used a cut-off of at least 30 black residents in 1950 , and fewer than 50 in 1960, for a likely case
3. all assessments of likely expulsion of blacks after Brown are, however, influenced by my own discretion — although I have not examined precise details for any county

Counties Previously Undiscussed That May Have Expelled Blacks After Brown v. Board of Education:

County		1930	1940	1950	1960	1970 households	1950-1960 change
Benton	Arkansas	88	46	11	23	5	109.09%
Carroll	Arkansas	25	8	5	6	0	20.00%
Greene	Arkansas	22	71	26	14	5	-46.15%
Madison	Arkansas	16	15	2	3	0	50.00%
Montgomery	Arkansas	136	90	14	20	3	42.86%
Scott	Arkansas	13	454	70	2	0	-97.14%
Van Buren	Arkansas	121	137	50	95	17	90.00%
Bracken	Kentucky	222	222	129	63	20	-51.16%
Breathitt	Kentucky	203	128	79	49	5	-37.97%
Carlisle	Kentucky	294	218	104	84	19	-19.23%
Carter	Kentucky	53	48	29	11	5	-62.07%
Grant	Kentucky	184	143	77	59	15	-23.38%
Johnson	Kentucky	46	14	3	1	1	-66.67%
Lawrence	Kentucky	124	95	69	41	15	-40.58%
Livingston	Kentucky	350	224	111	78	18	-29.73%
Marshall	Kentucky	62	155	11	10	0	-9.09%
McLean	Kentucky	315	247	99	60	15	-39.39%
Menifee	Kentucky	34	24	19	11	3	-42.11%
Robertson	Kentucky	46	40	31	18	5	-41.94%
Trimble	Kentucky	27	22	12	3	0	-75.00%
Atchison	Missouri	16	1	3	3	3	0.00%
Benton	Missouri	106	97	31	22	2	-29.03%
Caldwell	Missouri	102	45	11	5	0	-54.55%
Clark	Missouri	23	26	11	8	0	-27.27%
Daviess	Missouri	110	90	55	30	4	-45.45%
DeKalb	Missouri	38	19	10	0	0	-100.00%
Grundy	Missouri	85	75	35	18	6	-48.57%
Holt	Missouri	36	31	9	2	0	-77.78%
Iron	Missouri	129	132	85	48	15	-43.53%
Knox	Missouri	94	66	27	18	3	-33.33%
Miller	Missouri	89	54	17	20	10	17.65%
Nodaway	Missouri	95	33	19	8	4	-57.89%
Osage	Missouri	74	61	21	14	0	-33.33%
Perry	Missouri	99	75	30	9	4	-70.00%
Saint Clair	Missouri	134	173	55	36	7	-34.55%
Sainte Genevieve	Missouri	342	205	142	70	17	-50.70%
Scotland	Missouri	17	8	3	4	0	33.33%
Sullivan	Missouri	26	17	8	2	0	-75.00%
Vernon	Missouri	76	49	27	15	3	-44.44%
Dewey	Oklahoma	22	5	12	0	0	-100.00%
Latimer	Oklahoma	385	235	166	120	20	-27.71%
Mayes	Oklahoma	531	356	153	105	18	-31.37%
Roger Mills	Oklahoma	95	16	11	1	0	-90.91%
Grundy	Tennessee	155	66	43	15	2	-65.12%
Jackson	Tennessee	234	210	103	34	4	-66.99%
Overton	Tennessee	136	116	83	77	14	-7.23%
Polk	Tennessee	152	83	76	28	0	-63.16%
Sequatchie	Tennessee	11	8	3	0	0	-100.00%
Van Buren	Tennessee	65	81	23	31	6	34.78%
Archer	Texas	19	16	16	19	3	18.75%
Bandera	Texas	11	3	15	19	3	26.67%
Callahan	Texas	38	20	7	2	0	-71.43%
Coke	Texas	66	38	5	4	0	-20.00%
Concho	Texas	82	27	10	3	0	-70.00%
Glasscock	Texas	14	20	11	13	3	18.18%
Mason	Texas	57	100	64	23	3	-64.06%
Carroll	Virginia	374	332	392	74	24	-81.12%
Dickenson	Virginia	302	548	319	149	22	-53.29%
Highland	Virginia	153	86	118	18	6	-84.75%
Tucker	West Virginia	77	68	46	20	2	-56.52%

Conclusion:

The census figures above suggest that no more than eight counties not discussed in earlier post are likely to have expelled their black populations as a direct result of Brown v. Board of Education.

One — Grundy County in Tennessee, historically home of the most militant unions anywhere in the South and also home of a famous school training civil rights activists — is documented in Sundown Towns.

Two others — Highland County in Virginia and Tucker County in West Virginia — were the “inspiration” for this survey inasmuch as Löwen did not discuss either in Sundown Towns, but my mapping of counties with few or no blacks in the 1970 and 2010 censuses clearly showed these as quite distinct, yet mapping from 1930 showed they previously had black populations. A fourth — Jackson County in Tennessee, which voted even for George McGovern — is documented on Löwen’s website at Tougaloo College, but not in Sundown Towns itself.

Most counties tabulated above either did not show the necessary decline in black population over time, or got rid of their black populations well before Brown. Löwen did note counties that got rid of their African–Americans during the 1930s, and several here — Roger Mills County in Oklahoma, Concho County in Texas, and Johnson County in Kentucky — appear very likely to have done so looking at the table, whilst some others in the table above could have banned new black residents during that decade.

On the whole, the number of expulsions of blacks as a direct result of Brown is not that large, although the table must understate the total. such expulsions do prove how little white attitudes had changed in a nonplantation South whose voting patterns remained dominated by Civil War party loyalties.

The super-polluter social contrast

In my previous post, I noted how the US states of Wyoming and North Dakota rival the three super-rentier Gulf petromonarchies of Kuwait, Qatar and the United Arab Emirates as “super-polluters”. “Super-polluters” are political entities whose [extraction-based] greenhouse gas emissions per citizen are (roughly) at least 1½ orders of magnitude greater than the global average. Put plainly, citizens of super-polluters profit immensely from runaway global warming, via rents and/or job opportunities unavailable in states without these resources, and via not having to pay taxes.

What is strange when comparing the two sets of “super-polluters is that”:

1. both groups have a very restricted number of local “citizens”, which accounts for the super-polluter status, but
2. the Gulf petromonarchies have imported huge amounts of expatriate labour, but foreign labour is remarkably absent from the worst-polluting US states
• Wyoming and North Dakota have for many decades been amongst the bottom six US states in terms of proportion born outside the country, and this has not changed even with the oil boom and resultant glut that occurred in the 2010s.
• Moreover, Montana and West Virginia — two other very large polluters — are the very lowest two states by proportion born outside the US
• Alaska and New Mexico, two other great polluters (both among the most sui generis US states it should be emphasised), have larger proportions born overseas but these are still below to the US mean

In his landmark 2005 Sundown Towns, the late James Löwen noted that North Dakota — the second worst polluting state in the US today — was a sundown state for blacks during minimally the first two-thirds of the twentieth century. Census data from 1970 suggest the only place in the state a black family might live remotely safely was the military town of Minot. The exclusion of blacks in the rural Northwest was in fact so complete as to make research virtually impossible. The same is true of Montana, and marginally less so Wyoming, as can be seen below:

Counties with no or very few black households in 1970 mapped. Taken from https://www.census.gov/library/publications/1973/dec/population-volume-1.html. Note the huge area without black households in the Plains — and note that essentially all the rural Northwest was a sundown area for blacks.

The explanation can, I think, be found in the 2007 thesis ‘Measuring Ancient Inequality’ by Branko Milanovic, Peter H. Lindert and Jeffrey G. Williamson. Although in a revised edition of the thesis that team says it is not likely that any society’s subsistence minimum can be lower than $PPP 300 [1990 US dollars] and that it is possible that in the most exploitative societies a proportion of the population could not afford the subsistence minimum, what is critical is how the team noted the likelihood of lower subsistence minima for tropical and subtropical climates compared to temperate zones. It is likely that in climates as cold as North Dakota’s, requiring extreme shelter, clothing and food, the subsistence minimum would be not merely greater than the $PPP 300 used in their second edition, but significantly greater than the $PPP 400 used by Angus Maddison in The World Economy: Historical Statistics. No society in so cold a climate is analysed and my earlier post aims to explain why finding one is naturally difficult if not impossible. Although the team implies that one needs to be cautious re differences in subsistence minima, it is logical and likely that the subsistence minimum in North Dakota, Wyoming or Montana would exceed $PPP 500 and could reach $PPP 600.

The fact that bosses in Wyoming, Montana or North Dakota must provide workers a substantially higher subsistence income than in the Persian Gulf undoubtedly affects ruling class thought. Whereas in the Persian Gulf the ruling class brings in vast numbers of domestic and other low-skill workers, in the above-mentioned US states this would be too expensive as they require so much protection from frigid weather. Hence, the ruling classes of these cold super-polluters desire to completely exclude undesirable people — most especially and emphatically blacks — to limit the numbers of the lower class they must sustain at the elevated minimum. This, paradoxically, substantially lowers prices for those able to live in these regions below those of many hotter climates. Fewer buyers means less competition, with the result that buyers become more able to exert downward price pressure. This can be seen in how, despite the almost certain high subsistence minima, one’s dollar goes further in North Dakota than in Florida or Arizona.

Purchasing power of $100 in US states compared (2024). The subsistence minimum theory of Milanovic, Williamson and Lindert (2007, 2008) would predict a steady decrease the cooler the climate; however, historic sundown laws and related factors distort this

The basic conclusion is that climate affects the social structure of super-polluting states greatly. Only when subsistence minima are sufficiently low will the wealthy use their wealth to invest in exploitative labour practices: otherwise they are more likely to simply exclude others from it.

Super-Republican and super-polluters

I have been familiar for several years with the existence of “super-polluter” petrostates whose per capita extraction-based greenhouse gas emissions are hundred of times the global average. Economically, these “super-polluter” ruling classes who sell the fossil fuels profit from runaway ecological distraction caused by unlimited fossil fuel usages to an extent orders of magnitude larger than the average global ruling class. The wealth of these ruling classes, allows them to dictate policy globally to a very high degree — most especially to prevent demands for ecological justice and even for anything except carte blanche to emit however much greenhouse gases their rulers can profit from. The super-polluters of the Persian Gulf — Kuwait, Qaṭar and the UAE — are well documented by emissions figures. They become especially egregious if we debit zero emissions to expatriates and measure emissions per citizen, a perspective implied correct by Arun Saldanha’s ‘A Date with Destiny: Racial Capitalism and the Beginnings of the Anthropocene’. By such a calculation, Qaṭar’s per-citizen emissions are 450 times(!) the global average, and those of Kuwait and the UAE around seventy times.

The question of whether other super-polluters exists at a more local level has interested me recently. The fifty states of the United States provide a unique opportunity to examine this possibility, because I have long known of statewide breakdowns of fossil fuel production for coal, oil and natural gas, which can be used to approximate state-by state cumulative and present extraction-based greenhouse gas emissions.

Whilst it was too difficult for me to print all the figures, I have shown for each US state details of:

1. percentage of cumulative US extraction-based emissions
2. percentage of global cumulative extraction-based emissions
3. percent of current US population
4. percent of current US emissions
5. percent or ratio to total US extraction-based emissions per capita
1. I have used percent where the ratio is less than one, and ratio if it is greater
6. percent or ratio to global extraction-based emissions per capita

It might be noted that there are small extraction-based emissions from other sources which I did not calculate for lack of time and unavailability of data. (Extraction-based emissions are not zero for polities producing no fossil fuels currently or cumulatively, which includes a third of US states). Nevertheless, this table should be a clue as to which US state ruling classes are the biggest profiteers from global warming.

I have also shaded “red wall” states in red and “blue wall” states in blue to see how much US political divides relate to actual greenhouse gas emissions.

US Extraction-Based Emissions by State:

State	Percentage of US cumulative emissions	Percentage of global cumulative emissions	Percent of US population (2020s)	Percent of Present US emissions	Percent of or ratio to US per capita emissions	Percent of or ratio to global per capita emissions
Alabama	1.65%	0.36%	1.52%	0.63%	41.526%	1.356784592
Alaska	2.29%	0.50%	0.22%	1.63%	7.463	24.38419474
Arizona	0.31%	0.07%	2.23%	0.00%	0.012%	0.04%
Arkansas	0.46%	0.10%	0.91%	0.47%	51.978%	1.698253119
California	3.59%	0.78%	11.62%	1.08%	9.294%	30.36%
Colorado	2.02%	0.44%	1.76%	3.67%	2.089	6.824671159
Connecticut			1.08%
Delaware			0.31%
Florida	0.08%	0.02%	6.89%	0.01%	0.149%	0.49%
Georgia	0.01%	0.00%	3.29%	0.00%	0.000%	0.00%
Hawaii			0.43%
Idaho	0.05%	0.01%	0.59%	0.00%	0.269%	0.88%
Illinois	4.83%	1.05%	3.74%	1.88%	50.252%	1.641858174
Indiana	1.95%	0.42%	2.04%	1.13%	55.179%	1.802853651
Iowa	0.00%	0.00%	0.96%	0.00%	0.000%	0.00%
Kansas	1.30%	0.28%	0.88%	0.37%	42.098%	1.375445865
Kentucky	6.69%	1.45%	1.35%	1.45%	1.071	3.500860065
Louisiana	5.64%	1.22%	1.35%	3.66%	2.700	8.821153759
Maine			0.41%
Maryland	0.29%	0.06%	1.85%	0.06%	3.498%	11.43%
Massachusetts			2.10%
Michigan	0.27%	0.06%	2.99%	0.11%	3.540%	11.57%
Minnesota			1.71%
Mississippi	0.44%	0.10%	0.87%	0.29%	16.721%	54.63%
Missouri	0.04%	0.01%	1.84%	0.00%	0.220%	0.72%
Montana	1.56%	0.34%	0.34%	1.54%	4.607	15.05241983
Nebraska	0.06%	0.01%	0.59%	0.01%	2.153%	7.03%
Nevada	0.01%	0.00%	0.96%	0.00%	0.184%	0.60%
New Hampshire			0.42%
New Jersey			2.80%
New Mexico	2.70%	0.59%	0.63%	7.13%	11.355	37.10152528
New York	0.02%	0.01%	5.85%	0.01%	0.194%	0.63%
North Carolina			3.25%
North Dakota	1.72%	0.37%	0.23%	5.23%	22.297	72.84921958
Ohio	3.05%	0.66%	3.50%	2.52%	72.055%	2.35423961
Oklahoma	3.52%	0.76%	1.21%	3.69%	3.057	9.987822025
Oregon	0.00%	0.00%	1.26%	0.20%	15.952%	52.12%
Pennsylvania	12.84%	2.78%	3.85%	9.41%	2.441	7.974852848
Rhode Island			0.33%
South Carolina			1.61%
South Dakota	0.01%	0.00%	0.27%	0.01%	2.777%	9.07%
Tennessee	0.46%	0.10%	2.13%	0.00%	0.219%	0.71%
Texas	16.23%	3.52%	9.22%	24.74%	2.683	8.767608641
Utah	1.15%	0.25%	1.03%	1.10%	1.070	3.495456591
Vermont			0.19%
Virginia	1.82%	0.40%	2.60%	0.61%	23.374%	76.37%
Washington	0.07%	0.02%	2.34%	0.00%	0.000%	0.00%
West Virginia	10.05%	2.18%	0.52%	6.81%	13.060	42.66954766
Wisconsin			1.76%
Wyoming	10.11%	2.19%	0.17%	13.47%	77.804	254.2074428
Federal offshore	2.71%	0.59%		7.08%

The table above shows a remarkable concentration of US greenhouse gas emissions, and a major political divide:

• Four states account for fifty percent of cumulative US greenhouse gas emissions
• Five states account for 61 percent of current US greenhouse gas emissions
• “Red Wall” states have 22 percent of the US population but produce 64 percent of emissions
• “Blue Wall” states have 34 percent of the population but produce only 9 percent of emissions
• as noted earlier, this figure is an understatement, but even so the “blue wall” states undoubtedly produce low emissions relative to population — lower than the global average

There is definite evidence of “super-polluter” states in Wyoming and perhaps North Dakota, West Virginia and New Mexico. Wyoming is a definite super-polluter, with per capita emissions 250 times the global average, half that of Qaṭar. It is the least populous US state yet the second-largest greenhouse gas emitter and third-largest cumulative emitter! North Dakota has per capita emissions comparable to Kuwait or the UAE, and West Virginia and New Mexico about half that. Excluding New Mexico, these have constituted Trump’s three best states in all his presidential campaigns, which is extremely telling. New Mexico is a complicated and very much sui generis state, which Trump did not win in any of his three campaigns and did not even trend Republican in 2024. It has a countercultural mountain culture akin to Colorado or New England in its highest lands, politically detached from the pollution-producing northwest and Plains regions. In the other three the political influence of fossil fuels is as hegemonic as in the Gulf oil monarchies, and this is also likely to be true in Texas and Oklahoma. Even more than with the Gulf states globally, the tax-free or low-tax policies of super-polluting US states limit the range of policies possible elsewhere in the country, as business relocation is easier.

An overview of Wikipedia citation statistics

Although I have known of Wikipedia’s citation templates for some time — it is true that when I first made edits there I either did not use them or they had not been created — it is only recently that I have studied them in detail.

What one might call the “big four” — {{cite book}}, {{cite journal}}, {{cite news}}, {{cite web}} — have been familiar to me for some years now, because they can be used directly when adding a citation without copying the blank template from the appropriate linked site below. It is only recently though that I have attempted to look at all the templates in “Citation style 1”, and to see if and where I can use them.

A few days ago, I edited an article on the geography of Antarctica and did not know what to do with publications of the Geological Society of London, and of “Scientific Reports”. Presuming by the title that they must be reports of some sort, and seeing they did not fit the criteria to use {{cite report}} (a template I had used before discussing civil rights politics),  I put them under {{cite tech report}} — a rarely-used template found in only a little over two thousand Wikipedia articles (vis-à-vis over a million for the “big four”).

However, re-reading the template for {{cite tech report}}, it was clear to me that the articles I had cited therewith on Geography of Antarctica did not fit the criteria for {{cite tech report}}. They seemed to be closer to {{cite conference}} or {{cite journal}}, although I know nothing about what conference proceedings are.

The problems I had with this made me both message my brother for some discussion and to actually tabulate the frequencies of the various {{cite...}} templates, which I have done below, alongside percentages of Wikipedia pages used and how the template appears in links (if it does do so).

Frequencies and Appearance in Reference Texts of All Style 1 Wikipedia Templates:

Template	# of Wikipedia pages	% of Wikipedia pages	Template in reference text
{{cite arXiv}}	5,865	0.0099%	in link
{{cite AV media}}	61,062	0.10%	as |type=
{{cite AV media notes}}	29,630	0.050%	as |type=
{{cite bioRxiv}}	415	0.00070%	in link
{{cite book}}	1,763,420	3.0%
{{cite CiteSeerX}}	427	0.00072%	in link
{{cite conference}}	19,030	0.032%
{{cite document}}	1,653	0.0028%	as |type=
{{cite encyclopedia}}	219,847	0.37%
{{cite episode}}	18,233	0.031%	as |number=
{{cite interview}}	9,492	0.016%	(Interview)
{{cite journal}}	1,094,239	1.9%
{{cite magazine}}	338,377	0.57%
{{cite mailing list}}	811	0.0014%	(Mailing list)
{{cite map}}	46,609	0.079%	(Map)
{{cite medRxiv}}	159	0.00027%	in link
{{cite news}}	1,734,008	2.9%
{{cite newsgroup}}	653	0.0011%
{{cite podcast}}	5,158	0.0087%	(Podcast)
{{cite press release}}	76,594	0.13%	(Press release)
{{cite report}}	61,348	0.10%	(Report)
{{cite serial}}	279	0.00047%
{{cite sign}}	833	0.0014%	as |medium=
{{cite speech}}	1,426	0.0024%	(Speech)
{{cite SSRN}}	557	0.00094%	in link
{{cite tech report}}	2,351	0.0040%	(Technical report)
{{cite thesis}}	41,638	0.071%	(Thesis)
{{cite web}}	5,013,194	8.5%

Why climate change helps the rich — evidence of a nondoctrinaire kind

Although publications like Red Flag Magazine have long demonstrated that capitalism is entirely incapable by its very nature of ameliorating, let alone solving the climate crisis — and are perfectly effective even with deeply flawed and outdated geopolitical assumptions — Red Flag have now become truly open that the fight to prevent runaway climate change is purely and simply class struggle against the wealthy. In ‘Tropical Cyclone Alfred and the politics of ‘keeping politics out of it’’ and ‘Why Only Socialism Can Save Us from Climate Catastrophe’, alongside articles I could not find or were rewritten, Red Flag demonstrate that climate change is purely and simply class struggle on a global scale.

What is interesting is that Harvard University’s Branko Milanovic, Peter H. Lindert and Jeffrey G. Williamson, so early as 2007 in their thesis ‘Measuring Ancient Inequality’, made a strong implication that the capitalist class does have much to gain from global warming.

‘Measuring Ancient Inequality’ is a study of inequality in preindustrial societies. The critical issue the thesis notes is how the presence of a “subsistence minimum” below which human survival is impossible creates an Inequality Possibility Frontier (abbreviated to IPF) at a Gini index that appears low by modern industrial standards. Hence, in these ancient societies, even if extraction by the ruling class was the maximum possible, it only permitted seemingly modest inequality before the majority could not survive.

Milanovic, Lindert and Williamson use the symbol ε for the proportion of the population belonging to the ruling class, and s for the substance minimum in 1990 US dollars. The initially assumed value for the subsistence minimum is $PPP 400 — a value based upon the work of Angus Maddison in The World Economy: Historical Statistics from 2003.

Discussing their results, Milanovic, Lindert and Williamson note on pages 15 and 16 that:

“If we used Maddison’s subsistence level of $400, then four estimated Ginis would be significantly greater than the maximum Gini (at their level of income) implied by the IPF: three of these are based on data from India, and the fourth is from Nueva España.[The 1752 Old Castille is also slightly above the IPF.] Recalling our definition of the IPF, these four cases can only be explained by one or more of these five possibilities: (i) a portion of the population cannot even afford the subsistence minimum, (ii) the actual ε is much smaller than the assumed ε=0.001, (iii) inequality within the rich classes is very large, (iv) our estimate of inequality is too high, and/or (v) the subsistence minimum is overestimated. We have already analyzed and dismissed the first three possibilities. The fourth possibility is unlikely: since our estimates of inequality are based only on a few classes, they are likely to be biased downwards, not upwards. The last possibility offers the more likely explanation. It could well be that the subsistence minimum was less than $PPP 400 for some societies. In particular, this is likely to be the case for subtropical or tropical regions where calorie, housing and clothing needs are considerably less than those in temperate climates....”

“If the IPF is drawn under the S=$300 assumption, it shifts the frontier upwards enough to encompass at or below it all our estimated inequalities, with the possible (and modest) exceptions of Moghul India and Nueva España.”

If we invert what ‘Measuring Ancient Inequality’ says above about tropical and subtropical societies, we would logically conclude that in genuinely cold climates with extremely high food, shelter and clothing needs, subsistence minima might be substantially greater than $PPP 400. We might also suppose that in the very hottest regions with the most minimal such demands, the subsistence minimum could be still lower than $PPP 300. The initial ‘Measuring Ancient Inequality’ said nothing about how high or low subsistence minima could potentially be. A revised version from 2008, which incudes additional societies, does say it is unlikely that even in the hottest deserts the subsistence minimum could be less than $PPP 300. A 2011 version says:

...violations of the ‘maximum feasible’ definition of the IPF might be due to measurement errors or might reflect the possibility that some people can live below subsistence temporarily. The measurement errors could be of three types: mismeasuring national product per capita, mismeasuring inequality or applying the wrong (too high)  subsistence...[...made more likely by the fact that the extraction ratio is a ratio of two numbers, each calculated with significant amount of uncertainty: Gini, and maximum Gini which depends on estimate of mean income]. In the case of Moghul India and Nueva España, a portion of the population might have been expected to die from hunger, exhaustion due to forced and underpaid labour or lack of elementary shelter. Poor people’s income often does, in any given month, or even year, fall below the minimum and they survive by borrowing or selling their assets. Still, the same individuals cannot, by definition, stay below subsistence for long. The fact that the only two societies in our sample exhibiting a ratio higher than 100%, 1750 Moghul India and 1790 Nueva España, were notoriously exploitative seems consistent with this explanation

If the subsistence minimum were greater than $PPP 400, this would however be difficult to prove. One would need to demonstrate inequality as sufficiently restricted relative to income that a higher requirement becomes exceedingly probable. Even then there always exists the possibility of lesser extraction above a lower minimum. No society in any climate colder than temperate oceanic Western Europe is analysed in ‘Measuring Ancient Inequality’, and partly for reasons I will discuss below, candidates for such are few. [China in 1880 would lie partially within climates likely to have subsistence minima above $PPP 400, but the average subsistence minimum for the whole society would be dominated by the hot southern regions — at least judging by the result of the thesis.]

The above facts acquire deep relevance in light of runaway global warming. Milanovic, Lindert and Williamson imply that a hotter and hotter world will, ipso facto, increase the profits of the capitalists by reducing the absolute minimum wage they are required to provide to workers. Thus, the international capitalist class not merely has zero incentive to ameliorate global warming, but has vast potential gains from making it as bad as possible and permitting lower wages and greater profits. The results also explain the rapid movement in many countries from cooler to hotter regions: capitalist classes were seeking areas where they could gain greater profits, and the lower subsistence minima in hotter regions fit this goal perfectly.

Another fact explained very clearly by ‘Measuring Ancient Inequality’ is why civilisations evolved predominantly in hot deserts. The establishment of a stratified society requires that a surplus be produced, and logically this surplus will initially be small. Thus, it would be critical for the development of a surplus that the society’s subsistence minimum be as low as possible. Hot deserts — especially as they are the most likely region after extremely high mountains to evolve animals with the hierarchical social structures required for domestication — are ideal as the would-be elite is required to give such limited wages. Even more dramatically, the absence of even one origin of civilisation or even of crop agriculture in any latitude beyond the subtropics — European agriculture came via migration from the Middle East — suggests very low subsistence minima were critical for the evolution of stratified societies.