Tuesday, October 28, 2008

Election 08: Measure 65

If you are like me, this measure is the one that has delayed me sending in my ballot. I just can't decide how I feel about this and as an economist, there is not a whole lot in economic theory that seems to help. But here is my best attempt. Comment away.

The basic premise of the species homo-economicus is that we are essentially rational utility maximizers. This creates a problem if we think about the marginal impact of a single vote (basically zero) versus the cost of voting (not zero). But if we expand our understanding of the typical voter's utility function, it is not hard to believe that a sense of civic duty, a desire to play a part no matter how small, etc. creates benefits beyond just the ability to determine the outcome of elections. So the basic theory is pretty simple - the voters that determine elections are the ones in the middle and the politician that can appeal to the greater fraction of these middle voters will win.

This makes a lot of sense if you take as given that most people will vote or at least that the proportion of people that will vote is relatively constant throughout the political spectrum. But this assumption is not valid. So economic theory has to incorporate the fact that there are particular issues that will motivate particularly non-centrist voters and will change the proportion of populations that actually vote throughout the political spectrum. This assumption appears empirically valid. The success of Rovian political theory seems to show that voters can become more motivated through political appeals to the things that they care deeply about.

To this I'll add the somewhat controversial claim that we are, to ever increasing degrees, self-segregating along political lines. Take Oregon; the increasing proportion of registered Democrats in the state and the overwhelmingly democratic Willamette valley seem to suggest that Oregon is destined to become heavily democratic.

So what would an open primary look like in such a state? Well in an open primary, standard political economic theory would seem to suggest that the fighting would be over who could be the most popular center-left candidate. But with more than one of these battling over, say 60% of the votes, one could easily imagine one or two center-right candidates finding a relatively easy path to the final. A more modern theory might predict a number of hot-button issue candidates trying the mobilization strategy which, if successful could leave the center behind in a general election.

I have seen arguments that we shouldn't worry because in the state of Washington every single open primary they have had has led to a Democrat v. Republican general election. Is this good news? If it is then we should not change things because the current system delivers basically the same thing. Is it bad news? Well imagine a general election with two democrats. What would be the optimal strategy: pander to the base and mobilize, or try and capture the most right-leaning votes you can. I can imagine both could be winning strategies. Also, the fact that Washington has yet to have a same party general election may simply be the artifact of a two party system that is one, well-financed, and two, deeply ingrained in the psyches of voters.

So here is the rub - does this make our political process more fringe driven or less and does it matter? I think the answer to the first question depends on the marginal impact of mobilizing the base relative to the marginal impact of capturing the middle. I worry that as we become more self-segregated the marginal impact of mobilizing the fringe will begin to dominate. I think the answer to the last question is yes, it does matter. The health of a representative democracy, I believe, depends directly on how effectively the interests of the entire society are represented. And I guess I am as yet unconvinced that our two party system, for all of its obvious faults, is not the best way to ensure this happens. I know that this is an academic's response, but I think I'll wait for more evidence before I vote for a change.

But ask me again tomorrow.... any thoughts that could help me out?


Colin Dailey said...

Hello Patrick,

There have been a number of serious reporting errors regarding Washington State's experiment with its first-ever top-two primary (held on August 19, 2008).

For example, you wrote:

"I have seen arguments that we shouldn't worry because in the state of Washington every single open primary they have had has led to a Democrat v. Republican general election..."

I've heard this as well, but it is incorrect. For starters, they have only had one top-two primary (August 19, 2008). Second, Washington's November general election ballot will have 8 one-party races. They are as follows:

* There will be two Democrats running against each other in districts 11, 22, 27, 36, and 46.
* There will be Republicans running against each other in districts 7, 8, and 12.

Another error in the press coverage is the claim that Washington's first-ever top-two primary had the highest voter turnout since 1972. This too is incorrect. The actual voter turnout was 42.5%, which was less than the year previous (which did not use a top-two primary) at 45.1%.

Many of Oregon's business associations and media outlets signed on early to support measure 65, which may explain their exuberance for casting Washington's experiment with a top-two in favorable light.

But, as this is an economics blog, how would this measure affect the economics of campaigning and elections?

We can't be certain, but past instances of the top-two primary show that it increases the money involved in politics. The primary becomes much more important than ever before. Not only do candidates have to survive a free-for-all primary with plenty of spoilers on either side, but they also have to worry about how they fare against their strongest opponent (from the other major party - a concern they haven't had to worry about in past primary nominations). Campaigns start earlier, fundraise more, spend more on tv, print mailers, and radio advertisements. Also, outside spending increases, as outside organizations (that normally wait until the general election to bombard the state with their media) chime in early for the new R vs. D primary.

A good (near perfect) example of this is the difference in campaign spending between the 2008 and 2004 Governors race in Washington State between Rossi and Gregoire.

Same candidates. The campaigns started earlier, raised much more money. Total money raised nearly doubled. Outside spending went up by by millions of dollars. (All this can be easily fact checked online at the Washington Secretary of State's election division.)

I don't believe it's all cut and dry on this ballot measure. But from what I heard at the Portland City Club (and read in their report, the most comprehensive, unbiased report yet) it seems there are many legitimate concerns about a top-two primary.

Hope this helps (or at least adds to the debate).

Thanks for writing.

- Colin

Patrick Emerson said...

Thanks for the comments Colin, I am interested to hear your take. I was less concerned about election spending than thinking about how mechanism design affects diversity of opinion in political races. I appreciate your perspective and thank you to putting in your two cents.

chris farrell said...

Consider this: most districts are uncompetitive, with only one person, the Democrat or Republican, having any chance of winning. Once they get past the primary, they are assured of winning. For that reason, they make sure they are as far left or right as possible so that they are picked in the primary. This makes for excessive partisanship, because by the time they are elected, they know they owe their position to following party lines, not to the will of the people.
Contrast that with the situation in which, in a heavily Republican district, after measure 65 passes, we end up with two Republicans as the top vote-getters going into the general election. In that situation, there is substantial competition between the two, and the one that will win is the one that can appeal to the largest slice of the general public. Measure 65 will reduce partisanship gridlock in the state governement and it will force politicians to really look at the interests of everybody, rather than just looking at the interests of the members of their party that vote in the primary election.
Of course, many districts are competitive for both Democrats and Republicans, ...and nothing much would change there. It's the districts that are heavily weighted Democrat or Republican in which measure 65 would do some good.