Tuesday, November 6, 2012

On Voting

Mencius Moldbug once opined that for a westerner to not believe in democracy in the 21st century is somewhat akin to not believing in God in the 18th century - not so much in terms of the persecution, but in terms of how much it makes you an outlier relative to respectable opinion.

Most people who get all misty-eyed when election day rolls around tend to rarely enunciate why they think democracy is such a good thing, for much the same reasons that Christians in 1700 rarely described why Christianity was a good thing. But the broadest arguments fit into two camps.

The first is that democracy is an instrumental good - voting generally, and universal suffrage specifically, are ways to ensure peaceful good government. Voting itself is neither good nor bad, but it produces much better governments than non-democratic procedures.

The second is that democracy is an inherent good. Having people collectively decide their leaders on a regular basis is the morally correct thing to do, and participating in this process should make one feel virtuous.

I can sympathise with the former argument, although I think it needs some obvious qualifications.

I cannot really believe in the latter argument any more.

In the first place, it seems that the univariate comparison between western democratic countries and third world non-democratic countries vastly overstates the treatment effect of democracy. This is an enormously complicated empirical question that the development economists war over viciously. But even just in terms of the anecdotal discourse, the democracy boosters never seem to consider the harder cases. I'm not even talking about the cases like Singapore or Dubai, which they tend to wave away as despotic, if prosperous. But was the Austrian empire ca. 1900 a despotic and terrible place to live? Hardly. By any measure of its cultural, scientific or literary output, or just general standards of living, it seemed quite pleasant indeed for the time, although it certainly wasn't democratic. Or for a modern example, would Lichtenstein be meaningfully improved by transforming it to a democracy? It's hard to see how.

Democracy boosters also never seem to want to talk much about the first world cases where democracy is receding. Quick, name an important EU-wide decision made in the last few years that was decided by anything like a popular vote! No rush, I'll go and get a coffee and check up on you when I get back. Are you railing against the EU? Maybe for their economic policy, but what about their internal governance? I don't think so.

I think there are at least two good arguments for democracy as an instrumental good. The first is the analysts consensus forecast problem - the median value of the forecasts from lots of independent analysts tends to be more accurate than the forecasts of most individual analysts. If lots of people all estimate what they think is best for the country and vote on it, the variance of the mean of our estimates is likely much lower than the variance of any one individual. So a democratic process is less likely to screw up by picking an oddball policy.

The problem arises when people aren't voting based on what they think is in the country's interests, but their own. If 51% of people get together by voting to expropriate the remaining 49% (which seems like a fair description of the west today), it's hard to see how the analysts consensus forecast improves this.

The second is the idea of increasing popular support - democracy makes people feel they have a stake in the outcome and a way to vent their grievances, hence there is less civil disruption and fewer coups. I think this definitely has a value, but then again absolute monarchies used to be quite popular at times too, especially when they had a good king (although they'd be highly unpopular now. Again, except Lichtenstein).

But if democracy is justified as an instrumental good, it's surprising how rarely people make the obvious qualifications - that its value will depend greatly on who is voting, and what they're voting for. If the people voting are mainly fools, madmen or thugs, I don't expect the ballot box to transform them into Thomas Jeffersons. If you vote for Hamas, you will get Hamas.

This leads us to the limited moral argument for democracy - that even in the case of bad outcomes, people at least get what they deserve on average. We'll put aside the case of whether the minority getting expropriated deserves their fate for their inability to stop the majority. By this rationale, the Coptic Christians are now being 'deservedly' hounded out of Egypt, just like the Christians were 'deservedly' hounded out of Iraq.

But more generally, should we celebrate when societies are transformed from undeserved good governance to deserved poor governance? Rhodesia was a racist semi-democratic state with a functioning civil society whose benefits flowed mainly to the whites, but whose level of growth was pretty good. When this transitioned to the fully-democratic (at least initially) Zimbabwe, what resulted may or may not be considered less racist (it depends on how you score the massive violence against white farmers), but it's a basket case society that has ruined and impoverished nearly everybody, black and white alike, outside of a tiny ruling elite. So celebrate! They're now 'deservedly' reliant on foreign food aid instead of exporting food to the world.

You see the problem?

Of course, the true believers think that democracy and voting have a more basic inherent moral quality - it's just the right thing to do.

You cannot reason out any system of morality without axioms, so there's not really much to dispute in this statement. I disagree, but your mileage may vary. We are still, however, entitled to ask what shadow value you place on this moral good relative to other moral goods. In other words, how much ruin in Zimbabwe are you willing to tolerate for the fact that they now have universal suffrage, instead of restricted suffrage?

I value the rule of law, and peaceful stable societies. To the extent that democracy produces this, great! To the extent that democracy destroys this, then a pox on democracy.

In the west today, it seems about a zero NPV proposition. Like all NPV calculations, it depends on what the alternative is. Transitioning from modern Britain to North Korea would be a huge step backwards, but is that really the relevant counterfactual? Europe is slowly becoming less democratic each day and nobody seems to much notice or care.

To the extent that democracy works in the west, it seems mainly because the west has cultural values that support peaceful, stable government, and they vote accordingly. I celebrate this fact, but I think it would lead to nearly equally good outcomes if they didn't vote.

This doesn't fit neatly on a sticker that you can put on your chest after leaving the polling booth. Then again, not much sensible advice ever does.


  1. Democracy is not perfect. Everybody knows that. What you haven't told us is what you'd replace it with (in detail).

    It's true that democracy doesn't magically turn countries into prosperous communities. There's no powerful unconditional democracy "treatment effect". You can give the vote to corrupt, divided societies and you won't get anything good out of it.

    Rather, there is a complex interplay between political and economic institutions: there's two-way causality that goes back centuries. Western democracies are, without fail, democratic. Even if you think that democracy didn't help at all (no "treatment effect"), but it was their "culture" (whatever that means) that led them to prosperity, those countries didn't flip a coin and decided to be democratic. Rather, democracy was the culmination of a long historical process: for all intents and purposes, the countries with "successful culture" picked democracy. So, even in the case in which democracy is pure selection, the selection is informative about what people in successful cultures want. That should tell you something.

    The reality is that, as I said above, the interaction between politics and economics is very complex. But if you want to boil it down to a few sentences here it is: democracy is something the middle class wants. That's how it came about in ancient Greece: former peasants became merchants and, after a while, trading all over the Mediterranean, got rich. When that happened, their political power became out of balance with their economic power. That cannot last long, no matter what, and democracy was born. The same is true in modern Europe: democracy started in the medieval and early modern cities which became powerful though trade and had a strong merchant (middle) class. When China grows enough to have a strong middle class, democracy will come there, as well, and the Communist party is aware of it and preparing for it. In short, democracy is a prize that the middle class gets when there has been enough development to sustain a strong middle class. By the way, the same is true for the welfare state. As countries become richer, their citizens require more insurance from society (i.e. each other). That's why democracy and the welfare state typically go hand by hand. Democracy and the welfare state, in turn, help cool down passions and typically help already successful countries prosper even more by ensuring social peace and stability. And that, in a nutshell, is the two-way causality I was talking about. And why introducing democracy by force, without allowing it grow organically, is not a good idea.

  2. Correction: "Western democracies are, without fail, prosperous".

  3. First of all, the question of whether democracy is an inherent moral good doesn't seem to depend much on the question of what you'd replace it with (although I'd grant you that this clearly weighs very heavily on the value of democracy as an instrumental good).

    I don't have a convincing complete theory of politics to give in a single blog post, but there's a few things to think about for what you might replace democracy with. Mencius Moldbug has a lot of interesting ideas, including sovereign corporations. Think the Dutch East India company but with modern corporate governance. If you can somehow secure freedom of entry and exit (a big if), I'd be most interested to see it tried, like the Charter Cities model. Would I like to replace all of the US with it, untested? Certainly not. Would I be up for trying it out on a small patch of South Dakota or Honduras and seeing how it works? Absolutely.

    But if that's too pie-in-the-sky for you, there's other options to observe in the world. One-party states with capitalist leanings (Singapore, Dubai) don't seem to be doing too badly. True monarchy in Lichtenstein doesn't seem to work too badly there either (if you want a hilarious argument for Arnie as King of the US, try here). The EU model of government by bureaucracy is not inspiring, but trundles along okay.

    Are any of them obviously better? Probably not, but some of them seem about as good overall, albeit with different strengths and weaknesses - that's why it's about a zero NPV proposition in my mind.

    Rather, democracy was the culmination of a long historical process: for all intents and purposes, the countries with "successful culture" picked democracy

    This seems like a good model of the second half of the 20th century, particularly places like South Korea, Taiwan and Chile. But wouldn't you concede that it's a pretty bad model of the first half of the 20th century? The Austrian Empire, the Ottoman Empire, Nazi Germany and Imperial Japan didn't 'choose' to become democratic, they were defeated militarily and had it imposed on them. It's easy to imagine either WW1 or WW2 going the other way.

    In the case of Austria in 1900, it had a successful culture, as measured by the vibrancy of the society's contributions to dozens of areas, but not a strong push for democracy. Germany in 1925 was highly civilised, and actively moved away from democracy. If either of these countries had prevailed militarily, the seemingly inevitable link between economic growth and eventual democracy may not have seemed nearly as inevitable after all.

    Democracy and the welfare state, in turn, help cool down passions and typically help already successful countries prosper even more by ensuring social peace and stability.

    I'll grant you that they've likely had that effect thus far. Although if you look at the long-term budget prospects of nearly all western nations and think that the final chapter has already been written on the conclusion that 'welfare ensures stability', I would respectfully disagree.