Moving Borders

Much of modern geopolitics is based on the idea of territorial integrity. A state has these borders, and they are not to be changed. When we invaded Iraq, our stated goal was to ensure a unified country was left behind. When countries try to change the borders of other countries, the international community freaks out, and often lots of people die.

Territorial integrity, like most political concepts, is a nice fiction. There are still border changes. There always have been, and always will be. We're naive to think that the arbitrary lines we've drawn on the map this century will last the rest of time, as if they're different from the arbitrary lines of the past.

Let's look at Iraq for an example. Why is Iraq falling apart right now? Because fundamentally, Iraq has no reason to exist as a single country. There are three major groups, all with their own interests and territory, and none of whom really have a desire to be unified with each other. (Overly broad general statements, obviously.) Iraq came into existence because foreign powers drew the borders, and it's stayed in existence by brute force of the central government. With that force gone, there's nothing holding that country together. The same phenomenon is pulling Syria apart.

Most border changes come about due to war. And war, we would all agree, is an expensive and risky proposition, best avoided. There may be circumstances where war is the best available option, but when it comes to the kinds of wars that lead to border changes, that's not generally the case. We should try to find a better general solution.

(I should say at this point that I am not, with this writing, advocating any particular instance of border change. I am interested in systems that work, and in keeping people alive, not in achieving any particular political end. This should not be interpreted as commentary on any particular revolution, war, or secession movement, past or present. If I choose to comment on those things, I will do so in an unmistakeable manner.)

There's a correlation between war and border change, but which causes which? Does the war cause the border change? Does the desire for border change cause the war? Or is there some third factor that causes both? Clearly, war is a consequence, not the root cause. In many cases, people resort to violence when all peaceful means of achieving their desires have been cut off. (This is one of the reasons why democracy is more stable than totalitarianism; you can't deny people what they want forever.) To avoid violence, therefore, we need some means for people to have what they want. It must be possible for the will of the people to be recognized and executed, even if that means redrawing the map.

Now, redrawing national borders shouldn't be taken on lightly. We can't be talking about a unilateral, simple majority, one-time vote. If secession is too easy, the world becomes unstable instantly. But the barrier also can't be insurmountable, or the pressure buildup leads to the same instability. Wherever we draw the line, it needs to be fixed, so people on both sides know what to expect, and can't complain about the rules being rigged.

Since self-determination is the principle people are typically fighting over, obviously we need to hold a poll. Suppose a part of a state wants to secede. We poll everyone inside the region in question, to find out what they want. But they're not the only interested party; we should also poll the rest of the country. (For multi-sided issues like Iraq, or areas that want to leave one country and join another, we'd need to expand this. Things can get exponentially complex. Right now we'll just address the simplest case.) We have to be sure that the poll is valid, that votes are counted accurately and cast without coercion. Uninvolved international observers are required. There would also have to be some agreement on who gets to vote, age limits, sex, citizenship, that sort of thing. There's going to have to be some serious negotiation going on before the actual vote is held.

But what are the people actually voting on? Obviously if we're talking about redrawing national borders there are an infinitude of possible outcomes, and while only some finite set of them are of interest, it could still be a very long list. There's going to have to be a lot of discussion beforehand about what possibilities are of interest, and the list will have to include the status quo ante.

That means plurality voting won't work; it's fundamentally broken for more than two ballot options. We need either some sort of ranked-choice system, or approval voting. I prefer approval voting for American elections, but in a case like this a Condorcet system might be more appropriate. However, approval voting has one significant advantage over ranked systems: approval can clearly show when no option is acceptable to a majority of the voters. That's useful information in a situation like this.

So we can hold a vote, and gather the opinions of all relevant people in a detailed and accurate fashion. What now? Obviously you can't just treat them as one big vote; the whole point of this is to give the majority in a region a way to win out over the rest of their country without violence. So we need a system for combining the two sets of preferences into a single outcome, in such a way that the votes in the seceding region are equal to the votes outside that region.

The simplest idea is to weight the regions by population. If the main body of the country has ten times the population of the area that wants to secede, give the area that wants to secede ten times the weight. Then combine all the votes into a single count. Unfortunately, the simple idea leads to some absurdities. By this standard, if a region of one person wanted to secede, it would take a unanimous vote of the rest of the country to stop them! Clearly that's not stable. So we reduce the weight by some negotiable factor. This way the people doing the leaving get more weight, but the people being left still count for something. It tends to force people to find an acceptable middle ground.

(This gets more complicated still if the area that wants to secede isn't clearly defined, and that's one of the issues being voted on! Presumably you'd have to evaluate each choice of borders with different weights, depending on who would be on which side of the line. To fully develop this set of rules for every possible circumstance is clearly impossible.)

And this is all theory. It's nice to have a mechanism that could be implemented. Making people accept it, that's a whole different problem. You can hold all the polls you want, but some people don't really care about the outcome of polls. But not everywhere is like that! If such a system were in place, how many secession movements would we see try it in more peaceful areas? How much better would it be for the central governments in those countries? After all, a successful and popular movement that's suppressed by the government makes that government less legitimate. And an unpopular movement that's clearly identified as just being a few troublemakers also helps. There's no downside.

Unless, of course, you're just particularly attached to keeping the map from changing for the rest of time. I, for one, am not.