22 February 2008

Why I Might Vote for Obama Anyway

Why, despite saying that I "can't" vote for Obama, I might.

There are two types of political scientists: those who think there is a duty to vote, and those who think voting is irrational. I'm in the latter camp. The argument for a duty to vote was elegantly stated by H.B. Mayo.
The usual argument that voting is a duty of all eligible to vote runs something like this. A successful democracy depends upon widespread interest and participation in politics, among which voting is an essential part. Deliberately to refrain from taking such an interest, and from voting, is a kind of implied anarchy; it is to refuse one's political responsibility while enjoying the benefits of political society. A right of non-voting, if widely exercised, would hasten the end of a democratic government; non-voting is a mode of action impossible to universalize in a democracy, and so fails to meet Kant's test for the categorical imperative. (citation below)
But that's a very normative argument, and it fails to refute the positive argument. The critical component of the positive argument is the undeniable fact that my vote won't make a difference in the election--that is, the expected utility of voting, if the value sought is "effect on outcome," or "getting the policies I want," is precisely zero (in a presidential, or any other large turnouot election); or, put another way, voting has little to no instrumental value. (There is a vast literature on this, which is all footnotes to Anthony Downs.)

Some people find that disturbing--I find it freeing. Some people worry that a vote for a third-party is a wasted vote, but I tell my students that since your vote can't change the election, it's real value to the individual voter is it's expressive value--that is, voting can have "intrinsic consumption" value. Therefore, you waste your vote whenever you vote for someone who is not your top choice, because you have reduced it's real value to yourself. (Of course that doesn't hold true in cases where your vote can make a difference--then strategic voting, casting a vote for your less preferred alternative, to block your least preferred alternative from winning, is entirely rational. And if your expressive value comes just from the act of participating, of supporting democracy, then you could vote randomly and still capture the intrinsic consumption value.)

That's why I regularly vote Libertarian. Sure, the Libertarian candidates are invariably nutcases, but I express my preference for less government regulation that way. (In my dreams, enough people vote Libertarian that one of the major parties takes notice and co-opts us by shifting away from their love of government regulation to a moderate libertarianism).

And that's why I might vote for Obama (assuming he's the candidate). In fact, I think we have much better choices for president this year than we have for many years. McCain has far more foreign policy understanding than Bush, Kerry, Gore, B. Clinton, Michael Dukakis, Reagan, or Carter did when they ran for office. H. Clinton knows what the presidency is really like, from her eight years in the White House, and wisely chose to serve on the Senate Foreign Relations and Armed Services committe, to gain crucial knowledge about those issues, so vital to the presidency. Obama lacks that experience, but appears to have real integrity, not just a political pretense of it. I think he's a man of honor, and after Bush, Jr., and Clinton, I'm ready for that. Better a man who honorably makes mistakes, than one who dishonorably does. And I hope, reasonably I think, that he's smart enouogh to surround himself with experienced people.

I honestly could be comfortable casting a vote for any one of these people, so the question is, which would provide the greatest expressive value for my vote? Foreign policy-wise, John McCain. But breaking the gender and color barriers are also important to me, and on election day I might vote for the Democratic candidate just to express that value. If my vote could tip the election, I'd vote for McCain. But my preference for him is not nearly strong enough to automatically outweigh these other issues.

Mayo, H.B. 1959. "A Note on the Alleged Duty to Vote." Journal of Politics 21(1): 319-323.


James K said...

One of the most annoying things about a 2-party system is that the bundles of policy and expressive variables are so big that it is hard to choose which bundle you prefer. That fact that neither side has much to offer someone with libertarian preferences doesn't help much either.

James Hanley said...

Yeah, the two-party system creates an unavoidable adverse-selection problem.

It's virtue is that it creates a great deal of political stability. I think a sub-optimal stability is preferable to significant instability. Citizens, including business-owners, can adapt to most systems, if they remain stable, and make plans for the future. When the rules change too much and too frequently, it's, at best, harmful to business planning and economic growth, and, at worst, precursor to revolution.

Patrick said...

James -

"The critical component of the positive argument is the undeniable fact that my vote won't make a difference in the election--that is, the expected utility of voting, if the value sought is "effect on outcome," or "getting the policies I want," is precisely zero (in a presidential, or any other large turnouot election); or, put another way, voting has little to no instrumental value."

But this isn't true. Your vote's utility isn't exactly 0, and it DOES have instrumental value, even if that value is very small.

The logical conclusion of that argument is that if nobody voted, we'd still get the politicians we want, which I don't think is true. I mean, it would pretty much be the point at which the system breaks down. How do you declare a winner if all the candidates have 0 votes?

I'm probably missing something since this is your specialty, but if I am I'd love to hear it. Logically I can't wrap my brain around the idea that every vote has 0 value (keep in mind I'm an engineer by trade, so that has alot to do with what informs my view of the subject).

James Hanley said...


It all depends on what I want to accomplish with my vote. If what I want to accomplish is to affect the outcome, that's instrumental utility--I'm trying to use my vote to cause some change.

But When approximately 100 million other votes are being cast, the chances of my vote making a difference are vanishingly remote--it not actually zero, then so close it's not worth distinguishing the difference--too damn many zeros after the decimal point but before you get to the one. So, perhaps non-zero, yes, but not by much.

And if you factor in the costs of voting--thinking about candidates, taking the time to register, and the time to vote--then theoretically the value of casting a vote could be negative. That's if your goal is to affect the outcome.

But that doesn't mean nothing would change if nobody voted. The instrumental value is dependent on how many votes are being cast.

More precisely, the expected value (EV). That is, the actual value of making the difference could be huge (I'd get great satisfaction out of determining who's our next president!), but when multiplied by the probability of that outcome occuring, we get the EV.

So, if the instrumental value is V, and the probability in a group of 10 voters is P1, then the EV is P1(V). But call the probability in a group of 100 million voters P1-Z. Then P1-Z(V) < P1(V). And at the upper bound, in the case of me being the only voter, we have 1.0(V).

But if I think of my vote in terms of expressive value, which is purely internal, then my probability is always 1.0, even though the Value may be V-Q. Then 1.0(V-Q) > P1-Z(V).

So, yes, I was imprecise in saying zero value for instrumental voting. Hope this explanation helps. If I totally befuddled the issue, please let me know.

James K said...

One way to combat the instabiltiy problem is to have two major parties and several minor parties. That increases the dividedness of government and offers more policy alternatives, but still allows the whole apparatus to function.

That's how it is in New Zealand, 2 major parties and 6 minors. That would probably be the outcome if the US adoped proportional representation as well, The Republicans and Democrats would be diminished, but would retain much of their dominance (path dependence is hard to overcome in politics).

James Hanley said...

The U.S. actually could adopt some degree of proportional rep without any changes to the U.S. Constitution. Any state could do away with the single-member district system for the House of Representatives. That wouldn't have much effect in the handful of states that have just 1-3 Representatives, but in states like mine, with 15 Reps, a single 15-member district would surely bring out some successful third-party action. And certainly in those states with 20+ (and California with it's 56 or so).

The selection of the President, and the selection of Senators in single-member statewide districts, would limit the reach of those parties. So it might create the kind of stability, with greater voice, that you're speaking of.

James K said...

That's interesting, but I suspect there is a collective action problem. A state with proportional representiation dilutes its influence.

Having said that, with only 50 states (and you wouldn't need all of them to participate), there's no reason why this should be insurmountable.

The states could try a "pairing" agreement where a blue state agrees to proportionalise if an equivalent size of red states also agree and vice versa.

James Hanley said...

"The states could try a "pairing" agreement where a blue state agrees to proportionalise if an equivalent size of red states also agree"

But both the blues and the reds would lose out, because then we would elect some greens. Also some libertarians, whatever color they are. So the blues and reds already are cooperating to achieve their common interest.

James K said...

Yes, this is the problem when cartels have the power to set their own laws.