I was looking at a thread on the orange website and I was surprised by people’s shortsightedness. They’re talking about how the stock market remains high despite the pandemic’s damage to the economy, but their spectrum of possible outcomes seems to stop at the Great Depression.
I see a lot of people in the US trying to understand the current disaster through the lens of the last big disasters that happened in the US: the US involvement in the Vietnam War, the US involvement in WWII, the Great Depression, the 1918 flu. Of course I don’t know what is going to happen — and if I did, I wouldn’t want to attract the resulting attention — but I think you should take a wider perspective. This might be an OCP, more like when Sherman marched to the sea or Cortés rode into Tenochtitlán. It might be more like Cambodia’s experience of the Vietnam War than the US’s. Things may change more than you expect. The US may not survive.
As Annalee Newitz writes in the New York Times of the Bronze Age Collapse:
When their cities were swallowed by fire, the Bronze Age ruling classes lost everything, including the subjects they once controlled. Greece’s population dropped by roughly 50 percent during this time, probably because of a combination of war, drought and migration, according to Sarah Murray, a classics professor at the University of Toronto and author of “The Collapse of the Mycenaean Economy.” Mr. Cline believes that plagues may have driven people into the hinterlands, too.
It’s hard to estimate the probability of such a country-destroying disaster, so it’s tempting to just dismiss the possibility out of hand as outlandish — nothing like that could possibly happen, since the US has remained stable and indeed grown in power in our lifetimes, in our parents’ lifetimes, in our grandparents’ lifetimes, even in our great-grandparents’ lifetimes. It’s tempting to assume that an edifice that has thus stood the test of time will endure forever. Moreover, since the United Nations won World War II, the world has experienced a historically unusual period of relative peace, the Pax Americana, sometimes called the American Century.
But periods of peace and countries do not endure forever; they are wracked by invasions, revolutions, military coups, and simple collapse. What’s the base rate of such events?
Let’s consider specifically the kind of events that upend the established order in a country and consign the rich and powerful to poverty and death. Again, it’s hard to measure precisely, but we can get within an order of magnitude. In recent centuries, most countries experience such a major upheaval about once every century or two. Once a decade is clearly too often, and once a millennium is too rare.
Let’s consider recent events worldwide.
At the beginning of 2020, most countries still had the same way of life they had in 2010 and indeed in 2000, and investments made in 2000 were still secure in 2020. We can enumerate the exceptions: Egypt, Tunis, Iraq, Congo, Liberia, Syria, Afghanistan, Sudan (if you live in Darfur), the Rohingya regions of Myanmar and India, parts of the Niger delta, Yemen, the Crimea, Venezuela, Libya, and arguably Hungary, Mexico, Bolivia, Ukraine, Mali, Honduras, Kyrgyzstan, Ecuador, and Turkey. That’s somewhere around 10 to 30 countries, although at the border, it’s pretty fuzzy.
That’s out of about 200 countries (again, pretty fuzzy), so we’re looking at a rate of around 0.5% to 1% per year.
But is that really fair? Afghanistan has been a mess for generations, and the Crimea for centuries. Perhaps some polities, like Switzerland and the Roman Republic and Empire, are very stable, while others, like Afghanistan, are very unstable. (Note, though, that Rome suffered its share of disastrous revolutions and sackings before it finally fell.)
That’s as may be, but for now we’re just trying to establish the base rate. Later we can work out how large and frequent to expect local deviations to be, and in which direction.
Let’s take a longitudinal survey looking at the histories of particular countries. If you look back in the history of any given country, you mostly only have to look back a few decades to the last such event, maybe a century or two. Let’s look at every 17th country from Wikipedia’s list of countries by water use.
India? Partition, in 1947.
Italy?† Following the Holocaust (killing some 15% of its Jewish population), in 1943 they were defeated in World War II, their prime minister hung upside down from a gas station, their king forced to abdicate, and much of their territory given to Yugoslavia and the United Nations; over the next 40 years they had successive economic crises and terrorist massacres.
Syria? Currently on fire.
Kyrgyzstan? The Soviet Union collapsed there in 1991, ushering in decades of poverty (22% of the population is still below the poverty line), the capital was looted during a popular uprising in 2005, the mafia keeps assassinating parliamentarians, the president fled in 2010, and the interim president requested an invasion from Russia to put down an incipient civil war.
Cambodia? The Khmer Rouge killed all the intellectuals and a quarter of the population in 1975–8.
The United Arab Emirates? The current state there was established when the British blew up Ras al-Khaimah in 1819, conquering the country.
Oman? In the 18th century it was the preeminent power in the Indian ocean. After centuries of decline, Britain bombed the shit out of it from 1957 to 1959 in order to bring the Imamate of Oman (incidentally, one of those thousand-plus-year-long states like Rome) under the power of its ally the Sultanate of Oman.
Suriname?‡ It had a civil war in 1986–9 in the wake of the bloody 1980 coup (whose winner is the current president, though recently sentenced to 20 years in prison), and although the genocide of the Americans during the colonial era was far less complete there than elsewhere in America, only 4% of Suriname’s people today speak an American language and less than 2% practice an American religion.
Qatar? Doha and al-Wakra were sacked and looted by forces from Bahrain and Abu Dhabi in 1867, following which point Qatari sovereignty was established.
Papua New Guinea? They were a major battleground of World War II in 1942–5.
So in sorted order the last time there was a country-destroying catastrophe in these countries were 1819, 1867, 1943, 1945, 1947, 1959, 1978, 1989, 2010, and 2020, with a mean date of 1947. We can see some clustering there: two of the countries were destroyed in World War II, and India’s destruction (and subsequent glorious rebirth) as part of the collapse of the British Empire was surely related to World War II as well. Whatever the distribution and clustering of these catastrophes within any given country, we should expect that the distribution of intervals since the most recent catastrophe is the same as the distribution of intervals until the next one.
In particular, this controls for the problem of instability clumpiness: like unstable servers, unstable countries tend to remain unstable for decades, with one crisis or collapse rapidly following another, so if we just count the collapse events in the world over some period of time, we will get an unrealistically high number. Today, for example, Oman has been stable since 1959, but its previous century was riven with intrigues, secessions, civil wars, truces, invasions, and gradual subjugation by colonialist British boots.
This suggests an average time-since-violent-collapse (and thus also time-until-violent-collapse) of some 73 years, with a fairly smooth distribution containing a significant number of countries going out to 200 years of stability or so.
What about the USA? Does it not have 240 years of stability? Only from the point of view of the Northerners; the catastrophic depredations of the Civil War in 1861–5 reduced the Southern states, which previously included the richest part of the country, to a poverty from which they have not recovered 155 years later, although of course the poorest people in the South were thus immeasurably enriched. (Please note, I am not arguing that the Civil War should not have happened; I am merely saying that if you were a wealthy investor in the Confederacy, you would likely be ruined by the war.) It was no picnic for the North, who suffered some 800'000 casualties, some 10% of its fighting-age men.
Even so, 155 years of stability puts the US in the tail of our empirical distribution, bested only by the UAE in my sample above. Still, it should give us some pause that the US spent four of its 240 years at war with itself. Stability is not to be taken for granted.
† Oops, now I realize Italy is one country early. I guess I’ll go with it.
‡ Oops, now I realize Suriname is one country early.
These figures give especial weight to the 20th and early 21st centuries. To some extent, this is justifiable: there are secular trends that change what is possible and what is probable, so events in the 16th century perhaps have less bearing on what could happen in the 21st century than events in the 20th do. On the other hand, we should be alert to the possibility that a short recent period, like 2000 to present, 1980 to present, or 1940 to present, is really representative of what is to come — it might happen to be anomalously stable or anomalously unstable, in a way that might not continue to hold true in the next decades.
Do we have strong reasons for believing this to be the case?
I don’t think we do. On one hand the Pax Americana reduced the number of large wars, but the Cold War also destabilized countries — this was a factor in the Suriname coup mentioned above and the destruction of the Imamate of Oman, for example, and Cambodia was of course only able to get away with its abuses because so many saw them as needed measures that only affected those with privilege, anyway. Kyrgyzstan’s collapse was a result of the Cold War’s end, although it’s done worse than many former Soviet republics — would it have been stabler and safer without the Pax Americana and consequent Cold War? Perhaps. Or perhaps it would have been more unstable and more dangerous.
A different objection, which I hadn’t thought of until berndj raised it, is that the weightings above are biased toward small countries. Any particular country on the sampled list had an 0.5% chance of being Cambodia and an 0.5% chance of being the UAE, but their populations are only 15 million and 10 million, respectively, so a randomly chosen living person only has a chance of 0.2% or 0.1% chance, respectively, of happening to live there. But failure rate probably is not independent of country size! It could easily be that large countries tend to collapse much more often than small countries, or much less often. If you carried out my survey on a planet consisting of 199 tiny countries that each collapse every year on average, and one giant country containing 99% of the world’s population that only collapses every ten thousand years, you’d incorrectly conclude that people’s mean time to living through a collapse was a year. So we should take another look at large countries, even though we can’t take a large sample of them.
The fact that India, the largest country on the list, with 18% of the world population, happens to have most recently collapsed (with genocidal massacres with and masses of refugees) precisely at the mean date of the survey, 1947, might be a coincidence.
The other largest countries are China, with 1.4 billion humans (also 18% of the world population); USA, with 331 million (4.2%); Indonesia, with 270 million (3.4%); Pakistan, with 221 million (2.8%); Brazil, with 212 million (2.7%); and Nigeria, which at 206 million (2.6%) pushes us over the 50% mark. I think these countries’ most recent country-destroying disasters, where times were harder than the US’s Great Depression, the ruling classes lost everything, the rich and powerful were reduced to poverty, there was widespread violence, and mass emigration ensued, were, respectively: 1949, or 1958, if you count the Great Leap Forward; 1865; 1975–1999, if you count the invasion of East Timor, or 1965–6 otherwise; 1947 again, same disaster as India; arguably the military dictatorship and guerilla warfare following the coup against Goulart in 1964, though the resulting countrywide impoverishment was slow and prolonged, but otherwise the tumultuous 1889–1930 First Brazilian Republic, or the 1864–1870 War of the Triple Alliance, or the 18th-century invasion by Portugal; and one of the 1993–8 Abacha dictatorship, the 1976 coup, or the 1967–70 civil war.
So, based on tentative dates of 1865, 1930, 1947, 1947, 1949, 1965–6, and 1976 for the most recent country-destroying collapses in the seven largest countries that house the majority of Earth's human population, there doesn't seems to be strong evidence that large countries are either especially stable or especially unstable. Of course, after a collapse a country might break into smaller pieces, as the USSR did, so perhaps we care more about countries that used to be part of big countries — but we've already covered those by sampling small countries. The USA's recent stability just looks like an outlier.
How about USAmerican exceptionalism? Do we have strong reasons for thinking that the US is far more stable than other countries? I don’t know that we have strong evidence either way. The US is still the world hegemon, and hegemons tend not to be invaded by the countries they dominate. But their economy frequently depends on their hegemonic status, which is fairly fragile, and its loss can precipitate major upheaval — even when internal power struggles don’t.
So I don’t think there’s a strong justification for thinking that the US’s risk of collapse in average years is significantly different from the 1% or so from the above.
This is an unusual time. An economist at the Federal Reserve has projected 34% unemployment in the second quarter of 2020, which is higher than the peak of the Great Depression — and that’s six months into the covid pandemic, not three years in. Unemployment insurance claims are orders of magnitude above past records. Last month New York City started digging mass graves for the overload of coffins as its covid infection rate peaked. Also, the US elected a reality TV host as President, and he fired its pandemic preparedness team before the pandemic. Last month, the Yugoslavian he appointed chairwoman of the FDIC he appointed published a video begging the public to “please, keep your money in a…bank.” The Michigan legislature just shut down this week to avoid getting shot by protestors armed with military rifles, encouraged by the President.
Some wag quipped that it’s like having the 1918 flu, the 1929 stock market crash, and Warren G. Harding’s presidential incompetence all at once.
This is not normal.
It’s hard to predict what will happen. Right now, the chance of any kind of rare event is significantly increased because of the covid pandemic, even — perhaps especially — in the US. Moreover, events involving chaos and discord are especially favored.
So the chance of a US collapse is higher this year than its average 1%. Let’s say it’s 20%.
I don’t venture to guess what a US collapse looks like. Typically things like famines and plagues don’t directly topple governments or end cultures; they undermine their economic strength and political legitimacy, making it easier for other forces to assert themselves.
But what groups might be players? Recent new mass movements within the US include the Tea Party and Occupy, but the military (3.2 million employees of the US DoD) is in a better position to take over if the civil state fails. (The US police force is deliberately fragmented to reduce the chance of this; so is the military, but much less so.) Both the Tea Party and Occupy support positions with broad-based popular support.
The Mormons number some 6.6 million in the US, far more than the military, and have always planned to take over government in the US if given the opportunity, in order to build a utopian society they call “Zion”. Economically and organizationally, they are well-prepared for hard times, and if there is a famine, the Mormons may be especially well prepared, because each family is required to store a three months’ supply of food, water, and other essentials, a practice known as Family Home Storage. Mormon communities in Mexico have resisted the incursions of drug gangs with some success. I think it’s unlikely that the Mormons’ dormant plans to assume temporal power will be put into motion unless society is in frank collapse, because I don’t think they have either the firepower or the moral force to effectively maintain control.
Drug gangs in the US already have functional apparatus for projection of force and have geographically widespread networks and functional countermeasures against the police, and over the border in Mexico have achieved substantial, though incomplete, independence from the Mexican state. However, drug gangs generally lack broad-based support in the population in the US, unlike in Mexico, and suffer from serious prejudice, much of which is racist in nature and thus not easily overcome by a change of circumstances.
Many large companies in the US have substantial material resources, well-exercised command hierarchies, committed workforces, and in many cases continuity-of-business plans for disasters. A few even have existing security forces. It’s plausible to think that Walmart (2.2 million employees), Amazon (647k employees), CVS (295k employees), AT&T (254k employees), Ford (199k employees), or Alphabet (99k employees) might be able to take on the burden of protecting their assets and employees without a functioning government. Walmart might have a hard time due to its low profit margins (US$3k/employee), but Amazon (US$15k), AT&T (US$76k), or Ford (US$18k) might be able to take the hit without collapsing; Alphabet (US$311k profit per employee) easily could, and indeed Alphabet has often come under fire for providing public services like food, laundry, and transportation to its employees. (CVS is currently losing money.)
Alphabet is in a unique position to defend itself from security threats, since no potential foe can operate without its services at present, leaving them exposed to intelligence gathering.
There are another several dozen companies in the US with over 100k employees: Accenture (515k), Kroger (453k), Home Depot (413k), Berkshire (though that’s a conglomerate) (389k), IBM (381k), UPS (365k), FedEx (359k), the USPS (500+k), and so on. Large defense contractors include GE (283k), Boeing (153k), Honeywell (115k), Lockheed (102k), General Dynamics (101k), and Northrop (83k). Other telecoms include Comcast (184k), Verizon (145k), and Charter (99k); I mention these because availability of telecommunication is crucial to viability of any geographically distributed organization.
It’s easy to imagine a consortium of these big companies entering into a security cooperation arrangement with one another in order to be able to continue operating, and big defense contractors can probably count on support from any such consortium.