A Time to Kill, A Time to Heal

Why do governments sometimes engage in mass killings? Mass killings could help governments to suppress the opposition–that seems obvious–but it’s also true that mass killings can create blowback and further stiffen the opposition’s resolve. Uzonyi and Hanania offer a simple theory and some clarification:

We argue that government mass killing during war reduces opportunities for the opposition to return to military conflict in the future. This allows for longer periods of post-conflict peace. However, government atrocities that begin after the end of a civil war create new grievances without diminishing the ability of opponents to fight. This makes a faster return to conflict more likely. Statistical analysis of all civil wars between 1946 and 2006 strongly supports our arguments, even when we account for selection effects regarding when governments are more likely to engage in mass killing. These results reveal that both during-war and post-war tactics influence civil war recurrence, but that the same tactic can produce different effects depending on the timing of its use.

Essentially the authors are arguing that civil wars sometimes end when one side decisively wins. Not surprising but how about this for an uncomfortable thought:

We stress that mass killing is a grizzly and morally appalling
tactic. But it does appear to keep a country at peace for a
longer duration once a conflict ends. If the international
community disrupts these effects of mass killing, it may be
inadvertently increasing the likelihood that civil war will recur. Thus, if the international community chooses to intervene in conflicts to protect civilians, member states must also
be willing to remain in the country over the long term to
help the government and opposition groups refrain from returning to war. Unfortunately, few states have demonstrated
an appetite for such long-term commitments


I sense that these guys could get some really unsettling calls about consulting gigs.

Nope. The Chinese are already handling them.

See Sri Lanka.

wrong---do real research on that--worthy of this sight----then you will see it is all 100% made up/estimated figures.

yes the videos are real---but they show less than a dozen dead, most of them probably tigers plus their female Lord Haw Haw.

and two officers that surrendered and one 12 year old boy. Yes that is wrong--yes that is not moral but it is not the mass killing of 30,000 to 40,000 civilians like every non sri lankan says it was---how do i know?----well i was there---i was on that spit of land and i was the first person on that site after the last shell---absolutely the first person there without a horse in the race.

Sorry Alex what's so insightful about this ??

Robert Stacy McCain suggested a decade ago (to the consternation of bien pensants) that the solution to the Arab-Israeli conflict was for Israel to recapitulate Sherman's march through Georgia, from the Jordan River to the coast, on the theory that only when an enemy knows he is defeated does the conflict end. Mr. McCain doesn't pretend to be an academic social researcher (much less an economist playing a military professional).

Sherman's march left people alive and very pissed off. That's kind of the opposite of what the author is talking about.

As in, even today, the South is still refighting the civil war it started and lost.

The destruction of the South was just not bloody enough to deliver real peace 150 years later.

On the other hand, the mass murder of native people's was more successful, though complicated by the US taking so much land from first Spain and then Mexico. Spain assimilated into the native people's creating a new culture that claims Latin America and most of the US West. They have simply continued their cultural ways of moving freely on this territory much as they did going back 10,000 years and then 500 years.

Trump sees them as the biggest enemy because they have not conceded defeat after the 1965 law that was supposed to keep most out by creating new quotas which required cutting quotas for the Northern Europeans Trump wants. Those quotas were put in place to keep out Southern Europeans, which were too much like the Latinos freely moving across US southern borders. The "bad deal" trump hates from 1965 was not fixed by the Reagan "bad deal".

Worse, Clinton building hundreds of miles of border wall turned millions of migrant workers crossing both North and South borders into permanent resident immigrants. It's gone down hill since.

Basically, the war started after the Civil War with Mexico by many of the losers from the South has still not been resolved, and that one ended by granting Mexicans rights as US citizens. A bad deal forced on Mexico, but also on the US because the spilling of blood so soon after the Civil War blood bath was not acceptable.

"As in, even today, the South is still refighting the civil war it started and lost.

The destruction of the South was just not bloody enough to deliver real peace 150 years later"

The US Civil War still rages on in the delusional mind of Mulp.

"Worse, Clinton building hundreds of miles of border wall"

You are correct that Clinton built hundreds of miles of border wall. Oddly enough it wasn't racist when Clinton built a border wall.

Despite these issues, the U.S.A. has pretty clearly been internally at peace since the civil war ended, with no serious threats of another civil war or the union breaking up.

Also, many historians place a lot of the blame for post-war resentment in the south on subsequent reconstruction policies, rather than the way the war was fought.

Sherman’s march left people alive and very pissed off. That’s kind of the opposite of what the author is talking about.

There was no insurgency in the Southern United States after 1865. The closest you get to that would be the political violence associated with the original Klan (1870-77)

Yes, letting the descendants of the Arab invaders stay in Judea and Samaria after 1967 was a fundamental mistake.

Germans were expelled from East Prussia and the Sudetenland after WW2 to remove the possibility of revanchist sentiment. Their ancestors had lived in those areas for centuries as well.

Sherman's March Through Georgia did not involve mass killings. Plenty of theft, a fair amount of arson and vandalism, but very little in the way of homicide.

>Sorry Alex what’s so insightful about this ??

No kidding. The guy is trying to apply logic to MASS KILLINGS. "Hey everyone, I've found a study showing that they might not be a great idea! People should read this and try to knock it off!"

Mass killings happen when people are over-motivated by rage, and have way more power than they can handle or deserve, and have reason to believe they will never be held accountable for anything they do.

For example, the New York Time Editorial Board for the last fifteen months. I suggest you keep an eye on those people.

My reading of it was the mass killings, when done at the right time, had benefits.

Machiavelli, the Prince.

“When you come to power, and have to go kill, do all your killing at once, rather than spread it out over a long period of time.

People will accept the former but not the latter.”

(My paraphrase.)

The only case i can think of that comes close is the 400 years of mass murder of native people in the US. Nature, in the form of disease helped, but mass murder by US governments continued to at least 1900. Far more successful than Hitler, or the Zionist terrorists who drove out the British who for 30 years tried to bring peace in the British Mandate for Palestine.

The only case i can think of that comes close is the 400 years of mass murder of native people in the US. N

Except that never happened outside your imagination. Deaths from all causes on the trail of tears amounted to perhaps 4,000. Wounded Knee claimed about 300 lives.

Maybe this will help you.


It was a war, not "mass murder". Indians were inferior in both numbers and resources to the European invaders so they lost.

Plenty of US soldiers were killed in battle. White civilians were killed as well.

If you are a utilitarian, you may develop a rule to allow mass killings.
That's very disturbing, and insightful.

The good of the many out way the good of the few. If you have to starve a few million peasants in order to collectivize the farms, it's all for the public good.

Not true. Respect for human rights produces lots of indirect benefits. Most utilitarians have always known this.

@JWatts is referring to specific historical events where a utilitarian approach to achieving ideological ends led to the deaths of millions. That can't be simply rejected without argument.

It may well be true that the best utilitarian arguments favor human rights (most of the time). But not all arguments are the best arguments. Not all utilitarians are "most utilitarians". There is often considerable disagreement about what the goals are and how best to achieve those goals. And one can almost certainly conceive of scenarios where the utilitarian "benefit" of killing lots of people outweighs the indirect benefits of human rights.

Utilitarianism removes moral barriers to mass killing, and that is hugely consequential, even if some other barriers sometimes remain.

What isn't?

Here is the famous essay by John Stuart Mill on intervention/non-intervention: http://www.libertarian.co.uk/lapubs/forep/forep008.

Of course, it's the ethnically/religiously diverse nations created from former colonies that have given rise to brutal and seemingly endless civil wars. My observation is that diversity can be stabilizing provided that there is a balance of power (i.e., approximately equal numbers in each group). Balance of power was the central feature of Americ's post-WWII strategy for maintaining the peace in Europe, and it has mostly succeeded. The absence of a balance of power has produced the worst in civil wars, Syria for example (which is roughly 85% Sunni Muslim and 15% Shiite (Alawite) Muslim and has a minority (Shiite) government). On the other hand, there's Lebanon, which is almost evenly split between Sunni and Shiite Muslims and for the most part has been stable notwithstanding its location. One might view this as counter-intuitive, that a roughly equal number in each group would actually promote instability and civil war since each side has greater ability to sustain a long war. On the other hand, a balance means that each side has more to lose from a long civil war, and the absence of a minority means there's less likely to be deep sentiments of grievance.

It hasn't occurred to Rayward that there's both a Catholic and Eastern Orthodox population in Lebanon.

And the Christians were, not that long ago, the majority

There's a lot of international relations theory and findings which say equal numbers gets you a hyper-regulated "truce" between the ethnic blocs. It also suggests that, sooner or later, demographic shifts break the 50-50 split and a terrible, bloody civil war often ensues to re-allocate the spoils based on the new division of power.

Balance of power is unstable in a binary system.

Lebanon had a civil war for a long time (longer than Syria), and has still had political assassinations and paramilitaries stronger than the national army recently.

Yes, this is particularly stupid even for rayward. "Beirut" was the name of two different games that kids my generation grew up with, in both cases because of the resemblance to bombings.

"On the other hand, there’s Lebanon, which is almost evenly split between Sunni and Shiite Muslims and for the most part has been stable notwithstanding its location."

This is so counter-factual one wouldn't know where to begin. I showed it to a Lebanese (Catholic) resident--I heard him spit out his coffee around the world.

I've decided to put a 98% presumptively dubious rating on all Rayward's posts.

>>mass killing is a grizzly

grisly, not grizzly. Dammit!

Mass Kiing is a grizzly bear who ets people I don't like.

The civil war started over the spelling of a word . . .

"We argue that government mass killing during war reduces opportunities for the opposition to return to military conflict in the future." Why didn't I think of that?

Don't feel bad, it eluded Obama for eight years.

I don't know why, but I know you didn't, and you wouldn't

It works. False flag with fake beards is especially good if you can drug your people enough so they do it. "Look at how brutal the insurgents are! We the state are here to protect you from this excessive violence! Fortunately, after this latest massacre, France and the US will give us even more money to protect you against Islamist barbarians!". Read "Qui a tué à Benthala" by Nesroullah Yous. Appalling.


'Why do governments sometimes engage in mass killings?'

Well, that framing turns out to be quite unsatisfying, as the largest mass killings in human history had nothing to do with civil wars. Stalin, Mao, and Hitler all had different reasons for their mass killings, but not a single one of them was engaged in an actual civil war during the periods that millions of people were killed under the governments they led.

Russia & China both had extremely bloody civil wars.

Yes, but Stalin's and Mao's bloodiest periods came well after those civil wars were decided.

Re: Stalin. Perhaps I t’s only after you win that you can weed out the trouble makers (which for him was also whole classes, not just individuals)?

Lenin won the civil war, not Stalin.

Stalin was second in command of the Red Army during the Russian civil war, Leon Trotsky being founder of the Red Army and first in command.

You think it's easier to mass kill people before they start a civil war (or after the civil war is won)? When they don't have weapons and aren't organized?

I think in both Mao's and Stalin's case, civil war plays no role in how they used their power, resulting in mass death. Neither the Russian nor Chinese peasantry have been historically noted for engaging in civil war. (That the Chinese can have extended periods like that of the warring states or warlords is a separate subject.)

The Taiping Rebellion was largely an enormous peasant mob that almost overthrew the Qing Dynasty..

Do not blame Prof. Tabarrok for the spelling error in the excerpt cited - it is not his error. And using [sic} would be snarkier than Prof. Tabarrok's usual MR style.

At least some of us could show off our super-duper-intelligence by figuring out what the author really meant.

Uh, what's the Straussian read of this?

That Econ PhD's command of English is so crude that howlers like this make it all the way to print?

To the Oxford University Press, no less?

Good Lord.

I wonder how much of this effect is just eliminating a lot of the bad people.

Like, think of the casualties in WW2-era Japan and Germany. They were soundly defeated, many of the crazy fanatics were killed in combat (and afterwards, the remainder were comprehensively removed from decent society and political power), and both countries were profoundly peaceful, tolerant, and pacifist for a long time after.

"Unfortunately, few states have demonstrated an appetite for such long-term commitments."

The burden of proof required before implementing this policy needs to be overwhelming. Is this a satire?

Interesting to start at 1946 rather than 1939, the end of the Spanish Civil War, when Franco rounded up and assassinated tens of thousands of opposition soldiers and leaders. Spain has been at peace since.

“The victor is not victorious if the vanquished does not consider himself so” ?

Yet different means of mass killing may produce different results?

"Death squads" and other means of mass murder make the government that uses such methods an international pariah; deliberate mass starvation via embargo or other means (e.g., Nigerian Civil War) seems to produce much smaller moral condemnation.

In any case, this seems a subset of the "productive war" hypothesis: that a war may be considered productive if/when it significantly reduces the probability of further violent conflict. As might happen if someone gains the means to consolidate a number of independent, often warring states into a single nation or empire, for example.

In any case, the authors might find more productive ground in exploring the wisdom and utility of limited wars, ones in which states deliberately choose to avoid either a decisive victory, or a decisive peace following the limited war. For a limited war may be limited in scope and objectives, yet such a war may prove to be unlimited in duration. And how productive is never-ending, low-level war compared with alternatives?

This might be a slightly new dataset but this is a standard result in political science.

Comments for this post are closed