The abrupt end of the Chinese Civil War (War and Peace in East Asia, Lecture 12)

Yesterday touched off the next unit of the course, which begins with two big events: a drastic reduction in the number of great powers and, our focus in this lecture, the Communist victory in the Chinese Civil War. For all its shocking length–Paine argues convincingly that the war effectively started in 1911–the war seems to end somewhat abruptly (if not, from a military standpoint, prematurely) in 1949, with

1. cascading defections of KMT (and KMT-loyal units) to the Communist armies
2. a rapid shift in public opinion against the KMT
3. Chiang’s hasty retreat to Taiwan,

all of them occurring well before the prospect of final military defeat loomed too large–which, given our understanding of how most civil wars end, is genuinely puzzling. Why, after seeing off Japan in 1945, and with massive armies still in the field, did the KMT not continue the fight but, in essence, resign from the war? And why was it not just the leadership, but also the soldiery and the citizenry making similar yet largely independent decisions?

Paine’s analysis (Chapter 8) treats these events somewhat separately, as I noted in the first run of this lecture a few years ago, they all seem consistent with a common root cause (I’ve edited the quote a little to clean up the language, btw):

…the key to understanding each of these events is (a) the conversion of the Communists into a conventional military force and (b) the related string of rapid and sizable battlefield victories. This revealed some important information about the likely outcome of a fight to the finish, which for the Nationalists would be long, horrifically costly, and likely not victorious.

As a result, we saw (a) more of the public being willing to side with the Communists, as individuals were more confident that others would be doing the same thing; (b) armies and their generals thinking along similar lines, hoping to preserve their authority and forces intact; and (c), finally, the Nationalist leadership realizing that cutting their losses and retreating to the relative safety of Taiwan was optimal.

In the end, we’ve got a story about battlefield outcomes revealing information that, in the case of the Nationalist retreat, straightforwardly encourage an end to the fighting, but this is pretty conventional (see, inter alia, here, here, and here). What I found interesting was the apparent second-order effects of the same information–that is, raised estimates of the likelihood of ultimate Communist victory–on the public and the military. In both cases, it seems there were those who could be swayed to defect, but only if they weren’t the only ones to do so, and once such a clear public signal emerged of the relative strengths of the belligerents in the civil war, we saw large and massive–even unexpected, from some perspectives–defections that contributed to the rapid collapse Nationalist resistance and the end of the war.

So the same event, as it happens, is at the root of all three of these mostly independent decisions: the loud-and-clear signal that, while the KMT could hold out for quite some time, it was by 1949 unlikely to be able to win at the end of that endeavor. This leads to rapid attempts to cut one’s losses, just as we would expect based on an informational account of how wars end; if you can guess the outcome and you can secure something similar by giving up the fight, we’d expect you to give up the fight.

I also think it’s worth mentioning that wars ending this way—rapidly and while resistance is still feasible—aren’t unprecedented. Paine attributes some of these decisions to cultural features specific to Chinese history, and while that might be a descriptively accurate rendering of how people and soldiers framed and discussed their decisions, the underlying story we told in lecture shows that the outcome isn’t quite so exotic. In fact, two examples of similar events, particularly for the armies in the field, spring to mind: the rapid German collapse in 1918, despite its territory never being conquered, and the evaporation of Iranian willingness to continue the war against Iraq in 1988.

When information from the battlefield both travels widely and portends bad things about the future, leaders, armies, and civilians can all make rather abrupt decisions to withdraw from a war, and it seems like we saw just such a process play out in the final act of a war that produced the politics of the Taiwan Strait as we know it today: two governments, never formally reconciled to the end of a disastrous civil war, divided by a narrow stretch of ocean, and on opposite sides of the current round of great power competition.

Why did Japan surrender in 1945? (War and Peace in East Asia, Lecture 10)

Isolating the causes of Japan’s surrender in August 1945 is a fraught process, because it invokes so many difficult topics, from American firebombing of Japanese cities as soon as its bombers were in range to the nuclear attacks on Hiroshima and Nagasaki to the rapid and brutal Soviet conquest of Manchuria, all framed against Japan’s own atrocities against civilians and soldiers alike in China, its other imperial conquests, and in battles with American forces fighting their way (and mistreating prisoners on their own) towards the Home Islands. Some arguments pride nuclear bombs as the war-ending weapons, some revisionist accounts credit the Soviet invasion of Manchuria, which followed the nuclear attacks and preceded the surrender, but we argued today in class that both claims—often ideological, sometimes for the contrarian or partisan fun of it, and frequently weak on theory—fail to consider a key aspect of the politics of surrender: the personal fates of the top-level civilian and military decision-makers in Japan. Seen in that light, both factors can be seen as necessary components of the surrender.

So we had two questions to answer in class today. First, why did Japan surrender when it did, in August of 1945, despite knowing that the war was as good as lost many months before (after all, its initial gambit based on 50-50 odds of defeating the United States in a short war clearly didn’t pay off)? Second, what ultimately drove it to surrender? As it happens, and as Paine argues pretty convincingly in Chapter 7, Hirohito’s decision to accept the Potsdam Declaration and capitulate (to all the Allies except the Soviet Union, which really did have a lot of trouble convincing people that it was the power least shitty to surrender to) was made possible by both the American attacks on Japanese cities (conventional and nuclear, though the former killed far more in total) and the Soviet destruction of Japanese military might on the Asian mainland.

Hein Goemans argues that leaders in some regimes, particularly those that mix democratic and autocratic institutions in such a way as to guarantee that losing office is both probable and deadly after failing to bring home the spoils of war, have strong incentives to fight on even in the face of terrible odds on the battlefield. Indeed, Japan’s civilian leaders, who competed with the military for control over the war effort by 1945, “feared a social revolution from below if the war continued and an army rebellion if Japan surrendered” (Paine, p212), meaning that and end to the war would require either (a) total victory, ensuring the survival of both civilian and military elites, or (b) a catastrophic event that made defeat inevitable and surrender no longer the worst option. In other words, for Japan to surrender, both civilian and military elites would have to be convinced that fighting on no longer held out any prospect of saving their skins; otherwise, fighting would represent a “gamble for resurrection,” a chance (however tiny) of avoiding the certain grim fate that would certainly follow surrender: exile, jail, or death.

Paine argues that the first step in this process was the bombing campaign against Japanese cities, capped off with (though not ended by) the nuclear attacks on Hiroshima and Nagasaki, which convinced the civilian leadership that war no longer held out the prospect of survival. But it was the Soviet invasion of Manchuria, which hurled an army of 1.5 million soldiers against a Japanese army starved of resources after years of American submarine attacks against Japan’s merchant marine, that convinced the military of the same. The combination of both events, one imperiling the civilian leadership, the other imperiling the military, was sufficient to prompt the emperor to announce Japan’s acceptance of the Potsdam Declaration—a nominal unconditional surrender that nonetheless left the Emperor’s role intact and, a concern for the army especially, prevented a Soviet occupation of northern Japan. (In fact, Japan’s refusal to surrender to the Soviets has ensured that the latter’s territorial gains at the end of the war—none of which were previously held by the Russian state—remain disputed in a region full of territorial disagreements.)

So which event—the nukes or the Red Army—precipitated the Japanese surrender? From a logical standpoint, it’s impossible to say. Absent a Soviet menace to its position on the Asian mainland, it’s hard to see the army surrendering, especially after it not only refused to do so in the wake of the nuclear attacks but attempted a coup to prevent the civilians from surrendering after Nagasaki. Absent the nuclear attacks, on the other hand, it’s hard to see an otherwise insulated civilian leadership, confident in the army’s ability to bleed the Americans white if they attempted an amphibious invasion of the Home Islands, willing to throw in the towel when the prospect loomed of forcing the Americans into a negotiated settlement after drawing them into the equivalent of many more Okinawas (a seriously costly victory) just to make headway in Operation Downfall, the massive planned operation to invade and complete the conquest of Japan.

Thus, the nuclear attacks, the Red Army, and—of course—Chiang Kai-Shek’s years-long resistance, tying down massive numbers of Japanese troops on the mainland, all contributed to both why and when Japan surrendered. Take away one element, and it’s difficult to see the other two being enough to cause Japan to surrender when and under what terms it did—meaning that the debate over which particular factor was primarily responsible is fairly beside the point.

Why did the Great War last so long? (WWI in Real Time, Lecture 24)

When it comes to international politics, World War I pretty much has it all. That’s a blessing in some ways—there’s lots to be puzzled by, to explain, and to learn from—but in others, it’s not so great: any one event looks overdetermined when you look closely enough, and it’s inferentially impossible to prize some explanations over others. That said, we spent some time today talking about why the war lasted as long as it did. We finish the “real time” part of the course in December 1914/2014, but there were/are 46 more months of war to go. Why, after Germany’s defeat at the Marne, the onset of trench warfare on the Western Front, and the stalemating of the East, did the war continue until 1918? After all, while Germany suspected it was beaten in December 1914, it pretty much knew that it was in December 1917…and yet the war dragged on.

There’s a lot to draw on here, and if you want some accessible treatments of war termination in general check out this and this (both of which figure prominently in today’s story). To summarize the discussion, though, we emphasized that wars often end when fighting solves the problem that stood in the way of a settlement in the first place. When wars are driven by commitment problems, for example, they tend to end when either (a) commitments are made credible by removing incentives or abilities to renege or (b) the source of incentives to renegotiate (like rising power) are eliminated. This, of course, tends to push states towards particularly extreme war aims, from dismembering states to “ending” them as great powers to replacing their governments. On the other hand, when wars begin with states disagreeing about how a total war would play out, they end once fighting produces sufficient agreement that both sides can agree to save the costs of further fighting. (This happens a lot; there’s a damned good reason the vast majority of wars end short of what we talk about colloquially as military “victory.”)

Our first step was to see if these stories helped us explain why the war lasted so long, and they got us pretty far. Germany, for its part, was fighting a preventive war against Russia and its French ally, the solution for which was ending their run as great powers. The British, likewise, were fighting to make Germany’s commitment not to dominate the Continent credible. Both, of course, drive war aims towards totality, because solving them requires the ability to dictate some pretty harsh terms—terms you can only successfully dictate when your opponent is well and truly prostrate. Defeating an army in the field, especially when prevailing technology and doctrine favor the defense, simply takes a long time, and that helps explain why the war was so long. Total-ish aims didn’t come out of nowhere (though plenty of scholarship acts like they do), and here they emerged from commitment problems. Nice. Okay. One factor in favor of a long war.

But what about information problems? Some work argues that information problems can’t be a compelling account of long wars (here, here), but—especially in the case of WWI—I’m not sure that’s so hard and fast. Consider what our belligerents were uncertain about: the ability to hold out, to wage an attritional conflict by sustaining a war effort that drew over and over from a limited pool of labor and manpower, to bring (and keep) the whole of the population into the war effort, to outlast the other on the way to a “peace of exhaustion.” As opposed to valuations of the prize, per-battle chances of winning, or per-battle costs of fighting, it seems that the only way to prove how long one can hold out is to…well, hold out. In that sense, uncertainty over which side could sustain mobilization likely kept the war long as well—fighting before full mobilization wouldn’t be revelatory (and those involved knew this), but after that (1916, roughly), verbal claims of one’s ability to outlast the other simply couldn’t be credible. Fighting on was the only way to demonstrate that one could fight on, and that likely lengthened the war as well.

However, one can argue that this only gets us so far. By late 1917, with American troops arriving en masse to negate whatever advantages Germany won from the Russian separate peace, the game was clearly up; even gambling to try to hold on to parts of Belgium was likely a dead letter. Why did Germany still fight on? One possible answer is that the German elite was “gambling for resurrection“; expecting a pretty grim personal fate if they settled short of victory, they opted to throw everything into a pair of desperate gambles (unrestricted submarine warfare, then Operation Michael) that bet everything on slim chances of victory. After all, if peace would be disastrous, while the worst possible outcome of fighting on was also a disaster—why not fight and hope for a slim chance of survival?

It’s possible, then, that the commitment problems that drove belligerents to seek military victory, as well as the informational obstacles to judging the relative chances of success in a war of attrition, could explain why the war was a long one. However, why it ended in 1918—and in the way that it did, with a precipitous bottom-up collapse of the will to resist on the front lines—might require an appreciation of the fates awaiting the German leadership (at this point, dominated by the Army) if they did what in 1917 an outside observer might expect of them: settling on Allied terms.

The end of the Chinese Civil War (War and Peace in East Asia, Lecture 12)

After a brief blogging break (giving a talk, then giving an exam), Part II of the course got started today, where we set the stage for the rest of the narrative by examining the end of the Chinese Civil War.

In what was (sadly, if you ask me) be the last reading by Paine for the course (Chapter 8), we were treated to the puzzle of why the war ended when and how it did—that is, while the Nationalists could still muster substantial fighting forces and with a rapid Nationalist retreat to Taiwan. Paine identifies, yet explains separately, three events that are part of the puzzle:

1. Increasing defections of army units to the Communist side
2. A public rapidly shifting its support to the Communists
3. A hasty retreat to Taiwan before total military defeat

The task in class today was to see whether we couldn’t explain all three of these things with the same moving parts—and with fewer than Paine does. For the most part, I think we succeeded, because the key to understanding each of these events is, I think, the conversion of the Communists into a conventional military force and the related string of rapid and sizable battlefield victories. This, we concluded, revealed some important information about the likely outcome of a fight to the finish—in particular, that the Nationalist prospects in such a fight would be none too great.

As a result, we saw (a) more of the public being willing to side with the Communists, as individuals were more confident that others would be doing the same thing; (b) military units and their generals thinking along similar lines, hoping to preserve their authority and forces intact; and (c), finally, the Nationalist leadership realizing that cutting their losses and retreating to the relative safety of Taiwan was optimal.

In the end, we’ve got a story about battlefield outcomes revealing information that, in the case of the Nationalist retreat straightforwardly encourage an end to the fighting, but this is pretty conventional (see, inter alia, here, here, and here). What I found interesting was the apparent second-order effects of the same information—that is, raised estimates of the likelihood of ultimate communist victory—on the public and the military. In both cases, it seems there were those who could be swayed to defect, but only if they weren’t the only one to do so, and once such a clear public signal emerged of the relative strengths of the belligerents in the civil war, we saw large and massive—even unexpected, from some perspectives—defections that contributed to the rapid collapse Nationalist resistance and the end of the war.

I wonder if any other work has looked at these second order effects…