Why Doesn't Bloomberg Have a Successor

There is a fascinating story in the Times today about the efforts of New York City business leaders to find someone to run on roughly the same platform as incumbent Mayor Michael Bloomberg, who is barred by term limits for running for a third term These leaders – including superlawyer Marty Lipton, financier Steven Rattner and developer Jerry Speyer – are having difficulty finding a viable candidate who espouses the same set of policies as Bloomberg.

This is amazing. Bloomberg was overwhelming re-elected three years ago and continues to have an astounding level of popularity eight years into his Mayoralty, despite widespread worries of a serious economic downturn in a city that is extremely dependent on Wall Street. And they can’t find anyone with similar credentials to run on the same platform.

What gives?

These titans are running into the basic problem of urban politics, a lack of fit between national political parties and urban affairs. My paper, “Why is There No Partisan Competition in City Council Elections? The Role of Election Law” explores this very question. You can download it here.

First, some ground needs to be cleared. As noted in the Times article, a number of city leaders are worried that Bloomberg’s policies – although popular – won’t be followed in a new administration. Normally, we have a solution for this problem; if a set of policies is popular, voters can just vote for a candidate from the same party as the incumbent if the incumbent isn’t running. The link isn’t necessarily perfect, but it’s quite good. If you liked Bill Clinton, voting for Al Gore was probably a good idea, just as if you like George Bush, it’s quite likely that you’ll get a lot of the same things from John McCain (all protestation to the contrary notwithstanding). This is the genius of party governance – you don’t need to know a lot to vote, only which party you favor. This is a good thing, as voters, well, don’t know a lot (some, like my colleague Ilya Somin, dispute that people even have enough relevant knowledge about parties to vote retrospectively). But to the extent voters know anything, they are able to keep a running tally of which party they like and because the party name is on the ballot, they can use this information to guide their voting decisions. As a result, most voting is partisan voting.

This effect is most strongly felt “down-ballot.” For reasons that I explore in the paper but that are reasonably well-settled, the effect of party on voting is stronger as elections become less prominent. Again, this makes a lot of sense – you might have information other than party about a candidate for
Governor that effects your decision, but unless you are a total wonk, you are not going to have the same type of information about a candidate for Public Advocate or County Assessor. To the extent that preferences about the President track preferences about policy at another level of government, this isn’t a problem.

However, in urban politics, it is unlikely that preferences about national politics closely track preferences about urban policies. The group of people who comprise the Democratic Party at the national level have a relatively closely-linked set of preferences about national policies – there are some diivides, but by finding out that a candidate (or just a person) is a Democrat, you are likely to be able to guess their beliefs about a whole host of issues (abortion, the war in Iraq, progressive income taxation etc.). On local issues, this isn’t the case. There is a bunch of data in my paper I use to show this, but just ask yourself the following question. On issues like Mayoral control of the school system, broken windows policing or using public funding for sports stadia, can you name the Democratic or Republican position? What about on phonics v. whole language techniques for teaching reading, permitting construction of towers in residential neighborhoods, or congestion pricing? Beliefs about local issues are largely, if not entirely, orthogonal to preferences about national ones.

Because parties provide voters with information about candidates, and because this information isn't very good at the local level, voters are left adrift. The bulk of the paper is an effort to explain why local parties are not able to strategically change their public perception for the purposes of city council campaigns. I develop a model that tries to explain the problem, but the basic logic is something like the following. Due to the low information available in local elections, voters will use the information provided on the ballot so long as it is not entirely useless (and, although finding out someone is a Democrat or a Republican doesn't tell you much about their preferences about local issues, it doesn't tell you absolutely zero either, particularly because some issues overlap between local and national politics). Individual candidates have trouble developing brand identities due to the lack of information available in local elections. The party that is in the minority in national elections -- which we would ordinarily expect to come up a new platform in order to compete for votes -- cannot develop an independent local political identity because its membership didn’t sign on to do so. People become Democrats or Republicans on the basis of national, not local issues, and hence do not necessarily share beliefs about local issues, resulting in standard bearers that are inconsistent on local issues. Election laws governing party membership, ballot access, primary elections and campaign financing all make it difficult for parties to change their local perception and for new local-only party entrants. The local majority party is not forced to develop a coherent agenda because it faces no electoral pressure -- it just wins all the time -- and even if it did, it would face the same problems doing so as the minority party. The overall result is non-programatic parties (i.e. parties without consistent platforms on local issues) and one-party dominance, at least as long as one party dominates national elections in the city. (Where national elections are close, so are local elections, as is the case in Indianapolis). The lack of competition is really startling. For instance, despite Bloomberg’s popularity (and Giuliani’s before him), Republicans have only 3 out of 51 seats on the New York City Council (and are happy they’ve been able to keep that many!) Further, non-partisan elections (and primary elections in cities that do have partisan elections) are even less competitive, and for similar reasons -- in partisan elections, there is at least some information about candidates on the ballot, whereas in non-partisan election and primary elections there is none whatsoever. Ultimately, local elections tell us a great deal about which party locals favor for President, but little about anything else. (There are exceptions to this, like very high profile Mayoral races or situations where someone can spend enough money to overcome informational problems, both of which permitted Bloomberg to succeed.)

Regardless of whether I am right about why, this story shows that local political parties do not adapt to adopt even the most popular set of local policy preferences. Leaders and voters who support Bloomberg can’t count on either of the local major political parties to espouse a Bloombergian set of policy positions. The result is that the preferences of locals about local issues – the popular Bloomberg policies – are not likely to be continued after 2009.