Voting systems, registration rolls, and voter turnout

I took a class on geographic information systems last week. Neat stuff. My future work will have pretty pictures, even if the content may be lagging.

During the class, I produced a precinct level map of voter turnout in Multnomah County, OR during the 2004 election. (I didn’t save the map, but if anyone requests this, I can quickly produce one: My intent was to see where the votes come from in Multnomah County.

I am curious whether Portland’s city wide election system (no districts) has an impact on the way that city services are provided. As a first cut, I want to see if votes correlate with donations correlate with provision of city services.

What I found was pretty amazing to a jaded political scientist. I knew turnout was high in Oregon, but turnout rates in my county ranged from a low of 78% in some areas to over 95% in others! What the heck is in the water out here?

There is another possibility, though–the voter rolls in Oregon are so clean that they lead us to overestimate participation. Let me remind readers how Oregon maintains its rolls.

  • Approximately 18 days before the election, ballots are sent out to all registered voters. The postal service is prohibited from forwarding this mail.
  • For all ballots that are returned undeliverable, the citizen is coded as “inactive” and a postcard reminder is sent.
  • If the voter does not update the registration record, they remain inactive.

The consequence is that there is very little deadwood on the rolls and individuals are pruned from the rolls very aggressively, even if they have moved within a local area and don’t bother to update their voter registration.

I’m not sure whether this is good or bad in the grand scheme of things, but I’m pretty sure that it results in voter turnout that is higher than other localities. I did a quick back of the envelope calculation, comparing the estimated number of citizens over 18 in Multnomah County (520,156) and the number of registered voters (372,017) = 71.5%.

Why is this relevant? Well, advocates of voting by mail routinely point to Oregon’s high turnout as a reason to adopt the system. It promotes high levels of participation, so they argue. There are other reasons why turnout in Oregon is high (it has a participative culture, it is relatively homogeneous, there are low levels of poverty), but now I wonder whether there are institutional mechanisms that artificially inflate our turnout numbers.

Any help on this out there would be appreciated.