As I noted in my last post, there have been a few efforts to explain the lack of partisan competition in city council elections. Probably the most intuitive explanation is that, in most big cities in the country, most residents vote for Democratic candidates for President, so why shouldn't they vote exclusively for Democratic candidates for City Council?

It is true that in almost all major American cities, one party is dominant. However, as I argue here, this, it turns out, can't explain the lack of competition either, at least on its own. In order for national political preferences to explain the lack of competition at the local level, we would need a reason why the party that is in the minority locally doesn't change its stripes -- at the local level -- to become more attractive. That's what Downs would predict, yet it doesn't happen at the local level.

One reason that a local party might not do this is that national parties exert extremely strong control over local branches to stop them from modifying their platforms in order to compete. It is a bit unclear why a national party would do this -- after all, they generally want their local branches to succeed. That said, there certainly are some disciplining forces, notably the limits local dissenters have when trying to climb the electoral ladder.

However, even if parties do exert control over their local branches, stopping them from deviating from the national party too dramatically, local segregation by national party preferences could explain the complete and total lack of local party competition in City Council elections only if preferences on national issues and local issues are extremely consistent. If they are not, the majority party's coalition would fracture occasionally, leading a group of dissident majority party members into the hands of the local minority party now and again. When combined with the ability of the local minority party to deviate for national dictates somewhat, national preferences could only explain the lack of local competition if Democrats and Republicans form clear and very distinct groups at the local level on local issues.

Although the data is a bit spotty, what information we do have shows that preferences on national issues are not strongly tied to preferences on local issues. In the paper, I present all sorts of evidence -- polling data, studies of turnout -- that shows this. One notable piece of evidence comes from newspaper editorial boards. These boards can be presumed to have informed, complete preferences on political issues. If they have different endorsement patterns in national and local elections, it would suggest that national party preferences do not predict local policy preferences all that well. And indeed this is the case. Newspapers with heavy Democratic tilts in federal elections -- like the New York Times or Newsday -- have substantially less pronounced tilts at the local level. On the other hand, the Daily News, which has been consistently pro-Republican in New York City local elections, has a very slight pro-Democratic endorsement pattern in federal elections.

Further, the "they're all in one party anyway" concept can't explain the divergence between executive and legislative levels of competitiveness. And, moreover, the idea that national and local preferences track one another closely just doesn't pass the sniff test. On key local issues, one cannot identify the Democratic or Republican position. Mayoral control of the schools? Broken Windows policing? Using zoning policy to limit substantially the heights of buildings or to require large minimum lot sizes for houses? Bond issues for stadiums? This is the stuff of local politics, and there is little consistency across party members on these and other key urban issues.

Preferences on local issues just don't track preferences on national issues very closely. And as a result, they cannot explain the total lack of partisan competition in City Council elections -- at least by themselves. Tomorrow, I'll attempt to sketch an understanding that does.