Wednesday, November 1, 2006

Signal-to-Noise Ratio

Anyone who knows me even slightly knows that I have strong views on many political issues, and that I avidly follow political events. But unless you know me well, you may not know something else about me that may surprise you: I rarely vote.

What?

That’s right. I rarely vote. I don’t believe the conventional wisdom that voting is a duty of every citizen of a democracy — it isn’t; our Constitution grants me the right to vote, but it does not make voting a duty. Even more emphatically I reject the “logic” that every vote counts, and my vote might make the difference — for in nearly every election, the outcome is perfectly predictable.

It is in those exception elections, the ones where the outcome is in some doubt, that I vote. For example, I voted to recall Governor Gray Davis, for I was quite unsure of that election’s outcome and I had an informed opinion (Gray Davis really badly needed to be dumped!).

In most national-level elections, there really isn’t any doubt at all about the outcome. For over 20 years now, between 95% and 99% of all incumbents are re-elected. There’s a good discussion of why this is so here. Depending on whose tally you choose to agree with, in this election cycle there are only 10 or 15 House or Senate elections with any significant doubt about the outcome. All but one or two of these are electing a new Senator or Representative to a seat that’s been vacated. This fact — easily verified by any skeptics out there — belies the notion that candidates win on merit. For how could it be that in 95%+ of the cases, the incumbent is the better candidate?

A while back I read a very interesting study (which unfortunately I can’t locate a link to) that looked at the effect of campaign spending only in Congressional elections without an incumbent. These are nearly the only truly contested Congressional elections, so I was quite interested in this. The most damning result: in those cases where campaign spending differed by more than 20%, the top spender won more than 90% of the time. Do you really believe that the better candidate is so overwhelmingly more likely to spend more money campaigning? Again, the evidence argues that candidates are not elected on their merit.

Here’s a very well-known fact that I suspect few people ever think about: for well over fifty years, 99%+ of all Congressional winners belong to either the Democratic Party or the Republican Party. Our two-party system is so ingrained in us that I’ll bet a lot of people believe it is actually a Constitutionally-mandated situation. But it is not, and many fine candidates agree with neither party’s platform and run either as an independent or with one of the smaller parties. Rarely does one of these candidates win. Does it really make sense that such a tiny fraction of those candidates deserve to win on their merits, when compared to their Republican and Democratic opponents? Of course it doesn’t. Oh, and I know there are all sorts of arguments to be made about how membership in one of the major parties has substantial benefits (to the candidate) — but I don’t give a hoot about any of that. What I’m observing here is a simple fact: a superior candidate running as (say) an independent has a greater chance of getting struck by lightning while winning the lottery than he does of winnning the election.

Merit has little (and perhaps nothing at all) do to with winning Congressional elections. Congressional elections are won by incumbency, party affiliation, and money.

So I don’t bother voting in elections where there’s an incumbent — that incumbent is going to win no matter how I vote. I don’t bother voting in elections where there is no incumbent, but the spending is lopsided — again, my vote isn’t going to make any difference. But if we ever get an national election for my district without an incumbent and with relatively equal spending, I’ll vote.

I’ll also vote when there’s a local or state election of interest. California’s Assembly and Senate seats have a very similar rate of winning for incumbents as the Congress; all the same issues apply. Ditto on the spending. One characteristic of California’s political scene is a bit unusual in the U.S., and may be unfamiliar to many from outside the state: our proposition process. It’s relatively easy for citizens to put a proposition on the ballot — and the results of these propositions are binding. Much of California’s property tax cutting was accomplished this way, via the citizenry bashing their politicians over the head with Proposition 13. Politicians, by and large, hate the proposition process. I love it, of course — and I watch those propositions very carefully. In election cycles with propositions I care about whose outcome is at all uncertain, I’ll vote.

Given the readily observable statistics about voting, the average voter must vote like this: Is there an incumbent? If yes, vote for him. Otherwise, vote for the major party candidate who spent the most money, or for the major party that I vote for no matter what. In the rare cases where the preceding doesn’t tell me how to vote, roll the dice.

With that simple algorithm, I can predict — accurately — the outcome of nearly every Congressional election. Note that the candidate’s merit does not figure in the algorithm at all.

And this tells me that unless an election includes one of those rare no-incumbent, equal-spending races, my vote isn’t going to make any difference at all. There’s a good term for this from the electrical engineering world: the “signal-to-noise ratio”. This is simply a way of describing how the signal strength (for example, of a radio station) compares to the inevitable background noise. In Congressional elections, the “background noise” is the incumbency/spending/party affiliation situation, and the “signal” is my vote. If the signal-to-noise ratio is very low, my vote will not be “heard", and I don’t bother casting it.

What should really concern you, though, isn’t that I don’t usually vote. That’s not the sad story here. The sad story here is that voting algorithm I described above. For how can a democracy flourish when its government is no longer elected on its merits?

No comments:

Post a Comment