The Market for Votes in Afghanistan

Human beings aren't walking statisticians, which is why, all too often, they undervalue their slice of shared assets. The franchise in a body politic is one such asset, and as the New York Times reported yesterday, "[e]ven by the standards of a country rated as one of the poorest in the world, Afghans seem to be selling their votes cheap.”

The article goes on to explain how, with as few as 2500 votes often needed to clinch an election, priced at roughly $5-$10 each, a mere $25k outlay can land you comfortably in the state house. I say mere, because “the lucrative sinecure of a seat in Parliament [...] not only comes with a healthy salary—about $2,200 a month gross—but tremendous opportunities for graft.” Now, given, terms in the Afghan parliament aren't set: they begin and end with votes of confidence. So, it's not like you're buying those perks for the full five-year maximum for a Parliamentary term. It could be five years; it could be one. But even on the conservative assumption that your tenure's going to last, say, two–that's a huge return on $25k! Think about it: $52,800 in guaranteed income, a two-year opportunity to build momentum for future elections, graft. Even assuming that incumbency only predisposes you to a 10% advantage the next time you run, and you only take a couple thousand-dollar kickbacks, that's $25k spent on $60k in earnings. So the question becomes: why the bargain? Couldn't the sellers be demanding more?

I think the answer is immensely multivariate, and teasing out each and every factor is beyond my means—not to mention assigning all of them explanatory weight! But a few factors do seem to bubble to the surface.

First of all, supply is completely rigid. No matter how much demand varies, supply will always equal an amount equivalent to the sum of voter-eligible Afghanis. Actually, let me qualify that: the proportion of voter-eligible Afghanis who would ever consider selling their vote in the first place. (I imagine there's at least some percentage unwilling to sell for any price.) Either way, this is textbook supply inelasticity. Now, given that only a minuscule fraction of these votes are actually in demand (that meagre 2500-per-district-per-candidate), we've got a ferocious bid war brewing between sellers to find a buyer. It's a classic race to the bottom.

Second, as the article points out, sellers wouldn't have much bargaining power even in the absence of a bid war. Most of these people aren't exactly sold on their corruption-riddled charade of an electoral process—nowhere near enough to plan on participating. And add to that the (very real) possibility risking life-and-limb for doing so? Voilà, an “asset” (the franchise) which very truly costs them nothing to sell. They know it; their buyers know it. Think about that in the context of any commodity. I'm reminded of whenever I use the “you're going to throw it out anyway” argument when I'm trying to score cheap doughnuts from the local convenience store. It often works, but not because I'm buying in an oversaturated market. It works because the doughnut on display would costs nothing for Quik Stop to give away. Why not make a dime off me instead?

But third, and most intriguing to me, is the impact of cognitive bias. Humans consistently underevaluate the amount they pay for public goods. This has to be operating somewhere in the background here: even in a healthy democracy, where John Q. Public does vote and would internalise a cost in giving that up, he'd probably be imputing a loss less than the price he pays for being able to cast that ballot. Since it's well documented that all of us exhibit a disposition bias, John Q. Public then ends up being amenable toward otherwise-unattractive offers. He'd be a pawn to cognitive forces operating systematically to his disadvantage. And if this dynamic would operate on Mr Public, having uninhibited access to all the information necessary to be an informed seller, we can only presume it applies in at least equal force to Afghanis.