Lessons from the Election, Part 1: The Arkansas Legislature and the Decline of State Democracy

I’m going to do a few post-election blog posts on what the results can tell us about some of structural changes we’ve seen in American politics and their implications for constitutional and election law.   The theme of each of them is going to be that both popular and scholarly understandings of issues in politics – be they about federalism, campaign money, or Congressional commissions – have not caught up with our modern, ideologically coherent, highly polarized political parties.

The place to start is Little Rock, AK, where voters gave Republicans control of the state senate for the first time since Reconstruction and where control of the state house of representatives is still up in the air.   With Arkansas going Republican, there are now zero Democratic controlled legislatures in the deep south.  At the same time, in New York, the Republicans lost control of the State Senate for only the second time in my lifetime (although they may win it back, sort of, through party switching.)  Here’s a map of state legislative control, provided by the really excellent political scientist Steve Rogers:


Look familiar? Outside of the industrial Midwest and Florida, it looks a lot like the Obama/Romney map.  (Notably, if you want to understand how Republicans held the House of Representatives despite Democrats getting many more votes in House races, look to the same states and the effective gerrymanders engaged in by legislatures in states like Ohio and Pennsylvania, where you see such differences.)

With the fall of the Arkansas Democrats and the New York Republicans, what you can see is that the legislative vote follows a party’s popularity in the state on national issues.  Rogers also shows that shocks in party support that only affect national policy affect state legislative elections as well.   He writes: “Presidential evaluations and the national economy matter much more for state legislators’ elections than state-level economic conditions, state policy outcomes, or voters’ assessments of the legislature.”   On election day, this trend was confirmed in a big way.

For more on why this occurs, and what this tells us about American federalism, continue below the fold

Why this occurs is a difficult question to answer.  After all, any good Downsian should expect a party that is losing elections at a level of government to change its policy positions on issues relevant to that level of government to those supported by the median voter.  What’s curious is that this doesn’t seem to happen and even where it does happen, it doesn’t seem to work.

I have argued that failures of local parties to adapt to the local median voter can be explained by a group of election law rules that require, among other things, for primaries to be held among partisan voters that decided which party to support on national issues (and made it difficult for voters to switch parties between elections)  As there is no necessary correlation in issue stances in national and local parties, the result is that there is no coherence in candidate stances on local issues, making it hard to build a state or local party brand.  The result is party brands that are “mismatched” with the issues and median voter at the state levelChris Elmendorf and I argue that even when this is overcome (which happens sometimes, as Boris Shor and Nolan McCarty have shown), there are reasons to believe that it is difficult to get rationally ignorant voters to perceive state parties as distinct from national ones.   Chris and I suggest a variety of tools to improve state voter performance, from narrow ones like indicating on the ballot which party controls the state legislature, a fact known to make state voters more responsive to state issues, to aggressive ones like having high profile executives (like Mayors) make on-ballot endorsements in non-partisan elections, leveraging well-known individuals into quasi-party brands.

But regardless what you think of our proposals, the problem remains.  Having ideologically distinct and polarized parties improves voter performance in national elections, as Lau and Redlawsk have shown, by giving voters clear choices.  The dominance of these national parties in the minds of voters, however (and the absence of state parties that are similarly competitive), has meant that state elections are frequently little more than non-binding referenda on national policies.  In 2014, the best way to understand state legislative voting will be as a referenda on President Obama, not as a series of local decisions about the performance of local politicians.

I once joked to a class that leading academic and political supporters of “our federalism” do not seem to spend very much time in Albany, NY, Little Rock, AK or Springfield, IL.  The quality of debate one finds in these state legislatures is consistent with the finding that state legislatures are not held accountable for their performance.  Of course there are other arguments for federalism, from Tieboutian sorting through Feely and Rubin-style protection of distinct political identities (I make these arguments all the time!).  But some beliefs about federalism — certainly ones that get mentioned on the campaign trail — are centered on the belief that policies get more representative the closer they get to the people.  This is difficult to square with the finding that state and local races are largely determined by national political forces, from war to federal taxes.  States lack political parties at the state level differentiated on state issues, understood by voters as differentiated on state issues and taking stances that appeal to the median state voter.  And this means that voters in state election, with little information about individual legislators, have trouble holding state legislators accountable through the ballot box.

On election day, we were reminded that state legislative voting is mostly about national parties.  And this is very problematic for supporters of American federalism.



You may also like...

2 Responses

  1. I. Glenn Cohen says:

    Great post David! I was curious what you think this means for the economic/social divide on possible issues. The quote you take from Rogers seems to focus on *economic* assessments, but I am wondering if the same goes for social issues as well, such as gay marriage, abortion, etc.

  2. David Schleicher says:

    Interesting. I don’t have any reason to believe social issues are different in this respect, although I’d have to check if anyone has broken them out and studied them separately. Voter assessments of state legislatures certainly should capture social issues as well as economic ones. My instinct is certainly that there is little reason to believe that state policies on social issues are more responsive to voter preferences than state policies on anything else. But perhaps they are sometimes high-profile enough that voters do take note and politicians are able to differentiate themselves from the national party brand by taking popular stances on such issues (although one of the mechanisms discussed above, primaries among voters with stances defined on national lines, seems to have done a few of the state senators who voted for marriage equality in NY state, which is generally popular but not among republican primary voters (and even led to a 3rd party challenge in Dutchess county).