Totalitarianism is not a very useful political category. Authoritarianism is a preferred concept. In general democracies tend to have larger and more effective bureaucracies. China and the Soviet Union are outliers in this regard, inaccurately suggesting that authoritarian states are necessarily large and interventionist. They are usually much less competent.
Authoritarian states can emerge from democracies. The following risk factors are observed
Young democracies
Presidential systems rather than parliamentary
Poorer countries
Countries with large natural resources. This is well established
Weak democracies are sometimes created to protect the outgoing elites. Examples include Lebanon, Burma, Hungary, and (sort of) the US. The resulting democracies are less successful at creating legitimacy and may backslide more often. This theory is debated. See https://faculty.washington.edu/vmenaldo/Articles in Journals/BJPS Article.pdf
Because the US is a presidential model with many veto points and FPTP, it is more likely to have a coup. This makes it unusual among long established democracies. Japan is also a younger democracy (first regime change in 1994).
I want to know what my prior should be for a US coup. Sure, perhaps these factors you cite make it more likely to have one than it would otherwise and ceterus paribus, but what’s the base rate of coups in a multi-century, presidential, wealthy, resource-rich democracy?
Assigning a base rate here is difficult. We know presidential systems have more coups, and there are very few multi-century presidential systems. If your base rate is based on only those factors its low because of New Zealand and Sweden and the UK, which almost never have divided legislatures or divided judiciaries. This is a real problem—all the democracies that last as long as us look different. The democracies that are most like us had coups long ago.
If you ignore that problem, the base rate is like .3%. If your reference class is presidential democracies, then your base rate is more like 3%.
Chile had lots of other risk factors:
Of Chile’s three neighbours, two experienced 7 or more coup attempts in 1950-89. The other, Peru, experienced 5.
Executive and parliament not just divided, the legislature in coalition against the executive
President elected with just 36% of the vote
Riots and protests were common.
Escalating political violence
Inflation 140%/year
Judiciary publically criticizing the executive
Failed coup just one month prior
Economic contraction
All of those combined I say make coups quite likely. Over the 5 year period from 71 to 76, maybe 25%.
Do the data speak about the relationship between coups and federal systems? In the US, there is more than one level of fundamental government in play, even though they use similar models. I wonder if this helps to explain our weird longevity.
I just did a very quick search. The literature focuses really heavily on the relationship between federalism and interethnic violence at the national level (if we give tribe B their own province, are they more or less likely to launch a coup/civil war). Your question is addressed much less often, but if I had the time to dig I could find something. One note—among non-democratic states I doubt a relationship. Soviet Union was federal and high-coup.
In the US case, I strongly agree with your explanation. There are two plausible mechanisms.
The states would resist any coup in distant Washington. GW and TJ could not name themselves kings because the states had much larger armies. Similarly like Macron and Merkel cannot take over Europe by couping the EU. Biggest reason.
Any faction has a reduced incentive to launch a coup. This is more subtle, but it explains the large divergence in regime length in the Christian and Muslim world from 1,000 AD on (because Christian feudalism is “federal”). Each faction controls the wealth of a state/province/barony and has rich opportunities for rent seeking there. They can increase their rent-seeking by couping the capital, but the increase is actually low. They will still have to share with the states, and they already control their base. So the incentive for each faction to coup is much lower.
Imagine, by comparison, being an Ottoman Mamluke. Choose not to coup—you have 0 wealth. Win the coup, you get all the wealth. Huge incentive to take risks.
Caveat—not all coups are about rent-seeking. Actors may launch a coup to avert a national crisis, like the many coup attempts against Hitler. These are a minority (although everyone pretends they aren’t rent seeking).
Totalitarianism is not a very useful political category. Authoritarianism is a preferred concept. In general democracies tend to have larger and more effective bureaucracies. China and the Soviet Union are outliers in this regard, inaccurately suggesting that authoritarian states are necessarily large and interventionist. They are usually much less competent.
Authoritarian states can emerge from democracies. The following risk factors are observed
Young democracies
Presidential systems rather than parliamentary
Poorer countries
Countries with large natural resources. This is well established
Weak democracies are sometimes created to protect the outgoing elites. Examples include Lebanon, Burma, Hungary, and (sort of) the US. The resulting democracies are less successful at creating legitimacy and may backslide more often. This theory is debated. See https://faculty.washington.edu/vmenaldo/Articles in Journals/BJPS Article.pdf
There’s coupcast model. It’s not very good https://oefresearch.org/activities/coup-cast
Because the US is a presidential model with many veto points and FPTP, it is more likely to have a coup. This makes it unusual among long established democracies. Japan is also a younger democracy (first regime change in 1994).
I want to know what my prior should be for a US coup. Sure, perhaps these factors you cite make it more likely to have one than it would otherwise and ceterus paribus, but what’s the base rate of coups in a multi-century, presidential, wealthy, resource-rich democracy?
Assigning a base rate here is difficult. We know presidential systems have more coups, and there are very few multi-century presidential systems. If your base rate is based on only those factors its low because of New Zealand and Sweden and the UK, which almost never have divided legislatures or divided judiciaries. This is a real problem—all the democracies that last as long as us look different. The democracies that are most like us had coups long ago.
If you ignore that problem, the base rate is like .3%. If your reference class is presidential democracies, then your base rate is more like 3%.
Chile had lots of other risk factors:
Of Chile’s three neighbours, two experienced 7 or more coup attempts in 1950-89. The other, Peru, experienced 5. Executive and parliament not just divided, the legislature in coalition against the executive President elected with just 36% of the vote Riots and protests were common. Escalating political violence Inflation 140%/year Judiciary publically criticizing the executive Failed coup just one month prior Economic contraction
All of those combined I say make coups quite likely. Over the 5 year period from 71 to 76, maybe 25%.
Do the data speak about the relationship between coups and federal systems? In the US, there is more than one level of fundamental government in play, even though they use similar models. I wonder if this helps to explain our weird longevity.
I just did a very quick search. The literature focuses really heavily on the relationship between federalism and interethnic violence at the national level (if we give tribe B their own province, are they more or less likely to launch a coup/civil war). Your question is addressed much less often, but if I had the time to dig I could find something. One note—among non-democratic states I doubt a relationship. Soviet Union was federal and high-coup.
In the US case, I strongly agree with your explanation. There are two plausible mechanisms.
The states would resist any coup in distant Washington. GW and TJ could not name themselves kings because the states had much larger armies. Similarly like Macron and Merkel cannot take over Europe by couping the EU. Biggest reason.
Any faction has a reduced incentive to launch a coup. This is more subtle, but it explains the large divergence in regime length in the Christian and Muslim world from 1,000 AD on (because Christian feudalism is “federal”). Each faction controls the wealth of a state/province/barony and has rich opportunities for rent seeking there. They can increase their rent-seeking by couping the capital, but the increase is actually low. They will still have to share with the states, and they already control their base. So the incentive for each faction to coup is much lower.
Imagine, by comparison, being an Ottoman Mamluke. Choose not to coup—you have 0 wealth. Win the coup, you get all the wealth. Huge incentive to take risks.
Caveat—not all coups are about rent-seeking. Actors may launch a coup to avert a national crisis, like the many coup attempts against Hitler. These are a minority (although everyone pretends they aren’t rent seeking).