A week ago, we explored the historically precarious position the Democratic Party is currently in as it relates to the balance of power in the 99 state legislative chambers. At present, the Democrats control just 31 of those 99 chambers, and have exclusive control of just eleven state legislatures: California, Connecticut, Delaware, Hawaii, Illinois, Maryland, Massachusetts, New Jersey, Oregon, Rhode Island, and Vermont.
Meanwhile, the GOP not only controls more than two-thirds of the state legislative chambers in America, they have exclusive control of the state’s legislative levers of powers in eight states where Barack Obama twice carried the state: Florida, Michigan, Nevada, New Hampshire, Ohio, Pennsylvania, Virginia, and Wisconsin.
(Note—Thanks to several readers who pointed out that in three of those states, the Democrats have some leverage courtesy of holding the governor’s mansion)
This disparity is both jarring and terribly problematic, given how much policy that impacts our day-to-day lives is not carved out in the District of Columbia, but rather in cities like Carson City, Columbus, Harrisburg, and Madison. While the overwhelming bulk of the public conversation is devoted to the twists and turns of the presidential sweepstakes (admittedly, a wildly entertaining story), the real story about the policy direction of America is being waged in relative silence.
Democratic supporters are very quick to blame the disastrous 2010 election, which gave the Republican Party license to gerrymander state legislative maps from coast-to-coast. To be sure, it is a subject that has received a lot of attention on Daily Kos Elections: both in articles I have written, and articles written by colleagues like Stephen Wolf and Jeff Singer. And, without question, it will continue to be an intense topic of conversation around these parts.
But there is more to the gradual decline of Democratic support at the state legislative level than mere gerrymandering. This week, we explore the possibility that Democrats are hamstrung, even if slightly, by a tendency of their “soft” supporters being more willing to reach across the aisle and support legislative Republicans than the converse. Indeed, we have heard much about asymmetric polarization. The decline in split-ticket voting (which has been well documented), it appears, may be happening asymmetrically, as well.
To explore the theory, we looked at two sets of data: one was the absolutely invaluable presidential elections by LD data set put together by the Daily Kos Elections staff. That gave us presidential data for a total of 47 states. The exceptions were Mississippi, Alabama, and Maryland. In advance of this project, I did a cursory, back-of-the-envelope calculation for Mississippi. The other two states proved impossible, but that still gives us data for over 7,000 state legislative districts in 48 states. What’s more: since Alabama is GOP-heavy, and Maryland is Democratic-dominated, their impact on this analysis would likely be pretty much a wash.
The second set of data, which was easy to find but time consuming, was to look at which party occupies each state legislative seat. Since there are over 7,000 of them, this took … a while. It also requires a caveat—this is probably more than 99 percent correct, but there’s no chance that it is 100 percent correct. The reason is because with that many public officers, someone is resigning or getting elected in a special election more or less daily. Case in point: just this past Monday, an Arizona state senator quixotically switched to the GOP, which actually changed our numbers slightly.
Having said all that, the numbers were crunched, and the results were pretty fascinating.
First of all, with over 7,000 state legislative seats (7,055, for those scoring at home), the overwhelming majority of them were seats where the party occupying that district (or most recently occupying that district, for vacancies) was also the party that carried that district in the 2012 presidential election. Specifically, of those seats, only 942 of them (13.4 percent) of those seats are “split ticket” legislative districts.
The next question, of course, was to explore which party was doing the ticket splitting. In this case, there was quite a disparity:
| Partisan composition |
number of state lege seats |
% of state lege seats |
| Same party prez/lege |
6113 |
86.6 |
| d prez/r lege |
583 |
8.3 |
| R PREZ/D LEGE |
359 |
5.1 |
To put this data another way, of the “split ticket” legislative districts across the country, 62 percent of them were Republican officeholders occupying seats carried by President Obama in 2012. Just 38 percent, conversely, were Democratic officeholders in seats carried by Mitt Romney in 2012.
Aggravating the dilemma for the Democrats is that they are typically the beneficiaries of “split ticket” seats in state legislatures that are so disproportionately Republican that their ability to pick off those legislative chambers is not changed at all by their ability to pick off a handful of Romney seats.
To illustrate this, let’s imagine a hypothetical scenario. Let’s imagine that all 7,000-plus districts voted with their presidential nominees. If that were to happen, and we had a “straight-ticket” downballot distribution of seats, the Democrats would claim/re-claim the following legislative chambers:
Colorado Senate, Iowa House of Representatives, Maine Senate, Minnesota House of Representatives, Nevada House of Representatives, Nevada Senate, New Hampshire House of Representatives, New Mexico House of Representatives, New York Senate, Virginia Senate, Washington Senate
Meanwhile, the Republicans would also be the beneficiaries of this “straight-ticket” scenario. They would claim/re-claim … the Kentucky House of Representatives. That’s it.
So “straight-ticket” voting alone would reduce the GOP legislative chambers majority from 68-31 all the way down to 58-41. And that’s without redrawing a single of the gerrymanders the GOP was able to fashion in the post-2010 redistricting.
Of course, this is not a perfect comparison. Smart observers could immediately point to two issues right up front: (1) Obama won by four points (51-47), so by definition there is a likelihood of there being more Obama/R seats, because Romney had a lower vote share right off the top; (2) The bulk of these state legislators were elected in 2014, which was (as we well know) an absolutely brutal year for Democrats. Therefore, it is a certainty that the “split ticket” disparity was not as sharp a differential post-2012 as it was post-2014.
Those fair critiques aside, neither of them would overturn the basic premise. Even if the analysis was done from a “neutral” presidential position (i.e. only counting Obama wins of 51 percent or more as “Obama districts), and even if it was done after the 2012 legislative elections, there would still be more Obama/GOP state legislative seats nationally than Romney/Democratic ones.
So does this mean that Democratic voters are just more naturally bipartisan, while Republicans are cold ideologues that only vote the party line? Well, the evidence there is, at best, mixed.
Exit polling data, while not necessarily an unimpeachable metric, does not tell us that there is a huge “split ticket” gap between the parties. In 2014, registered Democrats were slightly more likely to flip sides (7 percent) than Republicans were (5 percent). The same disparity, down to the percent, can be found in the 2010 exit polling. Of course, these were also two lousy cycles for the Democrats, so disproportionate defections should shock no one. In fact, in the one decent midterm cycle of recent vintage for the blue team (2006), it was Republicans (8 percent) that were slightly more likely to defect than Democrats (7 percent).
Looking at the more specific case presented in our study, examining the presidential exit polling tells us that in two of the last three cycles, voters for the GOP presidential candidate have been slightly more likely to support Democratic House candidates than the converse (exit poll data, not surprisingly, is not so localized that it asks about state legislative votes). The disparity was particularly sharp in 2008, when roughly one-in-eight McCain voters said they voted Democratic downballot (conversely, just 8 percent of Obama voters said they went GOP for the U.S. House). In 2004 and 2012, the margins were considerably narrower, with 2012 being the one recent presidential cycle where the Obama voters defected to the GOP at a marginally greater clip than Romney voters defected to the Democrats.
If it is not a sharp shift in split tickets, what accounts for the dramatically higher number of Obama/GOP state legislative seats?
The more obvious culprit, as alluded to earlier in the essay, seems to be a much-discussed dilemma for the Democrats: disparities in turnout between presidential and midterm cycles. It’s possible that a bunch of Obama voters became Republican devotees in a two-year span. But it seems a lot more likely that a group of “soft” Democratic voters were motivated enough to turn out to re-elect Barack Obama in 2012, but could not be bothered to turn out in 2014. That subject has been a particular interest of mine, as I studied it after the 2014 midterm debacle, focusing on two western states (Nevada and Colorado) where the “turnout gap” may well have proved fatal to Democratic ambitions.
But the turnout gap existed ten years ago, too. Perhaps not as sharply as it does now, but it was still apparent. And the Republicans managed to get their asses handed to them in the 2006 midterms. So, merely pointing out that tanking Democratic turnout in 2010/2014 hurt them legislatively does not necessarily condemn the party to a permanent minority in legislative chambers across the nation.
All that said, the impact of gerrymandering, at least to some extent, is undeniable here. Consider, even if there was a straight-ticket vote in Wisconsin, not only would the GOP retain an 18-17 Senate majority, they would still enjoy a 56-43 majority in the House of Representatives. Which seems a bit incongruent, given that President Obama rather easily carried the state. Geographic sorting (the tendency of Democrats to cluster together) can account for some of that, of course. Madison and Milwaukee do account for an outsized share of Wisconsin Democrats. But, at the end of the day, there are clear examples of states where clever Republican cartography has made it nearly impossible for the Democrats to compete. This is a subject we will explore next weekend.
