Political Theory, politics, US Politics

Is there a versioning effect in elections?

Over at Crooked Timber, Henry Farrell is wondering why Obama’s lead over McCain is bigger in polls where more left-wing Presidential candidates are included – A puzzle about the polls. Comment #1 by a certain Cryptic Ned is proposing an interesting theory: versioning also works in elections.

Versioning is a term used in marketing where it describes a product/pricing strategy in which one combination is significantly more expensive only to make another one look like a bargain – think of your favorite fast food chain’s size/price combinations for soft drinks. Or, the other way around, think of my ever reliable method for choosing great bottles of Bordeaux if you don’t know the wines – buy the second cheapest bottle that fulfills your measurable quality standards, as the cheapest one is probably only on offer to make the second cheapest look better and should not actually be consumed.

But is versioning applicable to electoral decision making? Could having more extreme candidates on one side of the marginal vote lead to a higher vote for that side’s more centrist candidate instead of (usually assumed, arithmetically plausible, or even statistically evident) opposite?

Definitely an interesting thought. But let’s still hope that there won’t be another Florida recount and this will remain a purely theoretical question, at least for this US Presidential election.