Frequent readers of Daily Kos Elections are already very well acquainted with our intense (some might say “interminable”) focus on the impact of the redistricting process on the quality (and in some cases, quantity) of our elections.
One aspect of the process that has done real violence to both the quantity and quality of those elections in many states has been the impact that creative redistricting efforts (which we commonly call gerrymanders) have had on minimizing the number of competitive districts where both parties have at least a puncher’s chance of earning election.
In this week’s look at how redistricting has impacted the electoral process at the state legislative level (feel free to review some of the other pieces of this series of articles), we look at competitiveness. In other words, we will look at virtually every state in the Union, and we will see what proportion of their state legislative seats are legitimately up for grabs in any given election cycle. As we go beyond the jump, we will learn that gerrymandering is not the only villain as we seek to understand the general lack of electoral competitiveness, but it is—very clearly—a key factor in the hardening state of play in state legislative politics during this decade.
Let’s start with a simple caveat, one which should be universally agreed upon: Not every state will have equally competitive elections. This should be obvious—after all, the politics of Ohio are a little more evenly split along partisan lines than Hawaii. Florida, without question, should be a little more competitive than Wyoming.
So, to look at the competitive districts question, we need to start by doing two things. The first thing is to define what we mean by a competitive district. Off the top, let’s take incumbency out of the equation by only looking at the presidential electoral performance of these districts. After all, there are state legislators who have held their positions since shortly after the Earth cooled, and they’re not going anywhere soon, regardless of how their districts vote in other contests. They are institutions, and largely immune from the winds of political change.
So, what we do instead is use the invaluable set of results, breaking down the presidential results in 2012 by each state legislative district compiled by my colleagues here at Daily Kos Elections. That gives us 48 states (left out of the equation, for now, are Alabama and Maryland).
Then, to define a competitive state legislative district, we took any marginal presidential result within that district. For our purposes here, let’s define “marginal” as any result within five percentage points of the national outcome. To make life easier for me (I was looking at 7000 districts, after all), we went with any district where President Obama scored between 46.00 and 55.99 percent of the vote (the President’s overall total was 51.01 percent).
The second task was to look at how competitive each state actually was at the presidential level. After all, it stands to reason that a state that was lopsided for one candidate or another would be highly likely to be lopsided at the legislative level, as well. Which, of course, is provably true—see the legislative balance of power in Hawaii and Wyoming, for example.
For the curious, here were the dozen “closest states” in the Union, when retrofitted to the final national result. This is not the twelve closest states based on margin of victory in that state, since Obama won by four points, not in a dead heat.
So, in other words, North Carolina might be one of the “closest” races in terms of who carried it on Election Day 2012, but when compared to the national result, it was only the 12th ranked state.
2012 Presidential Election By State, Closest to National Average
State |
2012 Obama percentage |
deviation from national average |
Virginia |
51.16 |
0.15 |
ohio |
50.58 |
0.43 |
colorado |
51.45 |
0.44 |
pennsylvania |
51.95 |
0.94 |
new hampshire |
51.98 |
0.97 |
iowa |
51.99 |
0.98 |
florida |
49.90 |
1.11 |
nevada |
52.36 |
1.35 |
minnesota |
52.65 |
1.64 |
wisconsin |
52.83 |
1.82 |
new mexico |
52.99 |
1.98 |
north carolina |
48.35 |
2.66 |
In theory, the 10 states with the most competitive districts should largely come from this list. After all, the states that were closer to dead even on Election Day should be more likely to have closely contested districts than a state that was the site of a 2012 presidential rout.
Except that’s not really what happened.
The majority of the states in the top dozen, when measured by the proportion of state legislative districts that fell into that “competitive” range established earlier, were not in the top 12 states in terms of presidential competitiveness.
Indeed, one of those dozen “close” Obama/Romney states listed above ranked 34th in terms of competitive legislative districts, with only 8.24 percent of its districts falling into that 46-55 percent Obama window.
Here are the most competitive states, at the legislative level, based on our Presidential vote-by-LD calculations:
Percent of "Competitive" State Legislative Districts, BASED ON THE 2012 Election
STATE |
LEGE SEATS WITHIN “COMPETITIVE” RANGE |
PRESIDENTIAL COMPETITIVENESS RANK |
Maine |
45.16% |
16th |
IOWA |
41.33% |
6th |
new hampshire |
41.04% |
5th
|
connecticut |
39.04% |
22nd |
minnesota |
35.32% |
9th |
massachusetts |
33.50% |
29th |
washington |
32.65% |
15th |
new york |
32.39% |
37th |
michigan |
31.76% |
13th |
oregon |
31.11% |
14th |
FLorida |
30.00% |
7th |
wisconsin |
28.79% |
10th |
An astute set of eyes will immediately see a non-gerrymandering tendency here. Four of the top six states are located in the same region of the country—New England. And, for what it’s worth, the two that aren't on this list are so overwhelmingly lopsided (Vermont and Rhode Island) that it is virtually impossible to forge competitive districts there.
New England, by and large, has a homogeneity that is missing in many of the other states that command our attention on Election Day. Iowa, Minnesota, and Washington are also not exactly states that are known for their rich ethnic diversity. To put it another way, these states all have a commonality: They are overwhelmingly white states that have substantial enough liberal white voting bases to make them blue (or at least purple) at the statewide level. They lack the substantial blocs of Latino and African-American voters that can be found in a number of other blue/purple states.
This would seem, at least on some level, to give credence to the geographic sorting theory that states that Republican control of the state legislative balance of power (and that of Congress, for that matter) is based not upon nefarious redistricting shenanigans, but rather on the basic fact that core Democratic voting groups tend to cluster geographically.
However, geographic sorting, or similar demographics-based arguments on the redistricting question, cannot explain everything.
They cannot, for example, explain why two fairly diverse states that did not have Republicans holding the redistricting pen (Nevada and New Mexico) yielded more competitive districts than two states which were more competitive on Election Day at the presidential level, and yet had Republicans controlling the process (Pennsylvania and Ohio).
And, above all else, there is North Carolina. As mentioned earlier, North Carolina was among the closest states on Election Day 2012, with Barack Obama losing the state by just two percentage points. Despite that, just one in 12 of the state legislative districts in the state falls within the parameters of a “competitive” district. Idaho, where President Obama snagged just 32 percent of the vote, had a greater proportion of competitive state legislative seats than did North Carolina.
Consider North Carolina’s comical GOP map through the lens of history.
In 2012, the Democratic nominee (President Obama) ran 2.66 points behind his national average in the Tar Heel State. Despite that, the Republicans held a combined 110-60 advantage in the two chambers of the North Carolina legislature after the 2012 elections. The GOP was carrying the state with 50.3 percent of the vote in the presidential election, but held 64.7 percent of the state’s legislative seats.
Contrast that with 2004, when the Democrats ran 4.68 percent behind their national average in the presidential election (despite native son John Edwards being on the ticket). After that presidential election, however, it was the Democrats that held a 92-78 edge in the state legislature, controlling both houses.
Neither geographic sorting, nor the natural movement of southern white voters to the GOP, can be credibly blamed for a 32-seat shift in just an eight-year span. Voters of color in North Carolina, presumably, did not all migrate suddenly during the Bush years or the early Obama years. And Obama lost white voters in 2012 in North Carolina by a 68-31 margin, which was actually marginally better than John Kerry had managed in 2004 (Kerry had lost them 73-27).
Determining that North Carolina is a hideous joke of a gerrymander is not plowing new ground. Since the map was drawn nearly a half decade ago, we have known that the map (drawn by Republicans because of a quirk in state law that denied the governor, then a Democrat, the right to effectively veto the map because the override would have been via majority vote) favored the GOP to a nearly absurd level. Our own Jeff Singer showed us as much last summer when he looked at a different metric for measuring gerrymanders—the presidential performance of the state’s “median district.”
North Carolina is certainly the most egregious example, but it is not alone. Virginia (where the GOP-led House of Delegates and the then-Democratic state Senate drew their own maps amid a deadlock) and Pennsylvania created far fewer competitive districts than the overall competitiveness of the states would seem to dictate. The same was true for Ohio.
However, even in states that look nominally competitive according to our metric, clever crafting of maps can create a distinctly partisan edge.
In Michigan, for example, President Obama won 54 percent of the vote, but the GOP enjoys an enormous 27-11 edge in the state Senate. This seems unbelievable, until one considers that Mitt Romney actually managed to carry the majority of the state’s Senate districts. How is this possible? Well, it might have something to do with the fact that the Republican legislature, who concocted the map, managed to create a structure where a majority of the state’s Senate seats were Republican, but only marginally so. Specifically: 21 of the state’s 38 Senate districts are seats where President Obama received between 42-50 percent of the vote.
Perhaps the best example of this clever cartography was in the state of Wisconsin, where Republicans had total control over the redistricting process. On paper, it looked like the GOP did a reasonably fair job of maintaining competitiveness in their legislative mapping. In our analysis above, you can see that nearly 29 percent of the state’s legislative districts fell within our parameters of a “competitive” district, which ranked them 12th overall. Since the state was the 10th closest to the national average on Election night 2012, that seems, on the surface, to be “fair.”
But a deeper dive inside the Wisconsin map shows an incredibly shrewd approach to the map. Virtually all of those “competitive” districts are somewhat close in range, but they favor the GOP ever so slightly. Only five House districts, for example, came in with a presidential performance between 50-55 percent for President Obama. But 32 House districts, nearly one-third of the chamber, came in with a presidential performance between 50-55 percent for challenger Mitt Romney.
What Wisconsin Republicans were counting on was two things: The absence of a huge Democratic wave election in their state, and the presence of a genuine inelasticity in the voting habits of their state’s voters. Through two elections with their map, they are two-for-two on both counts. Control of the state House remains solidly in Republican hands, propelled in part by the fact that not one of those marginally GOP-performing districts is represented by a Democratic representative. (By contrast, four of the five seats that edge Democratic presently have a Republican representative).
Perhaps most agonizingly for those eager to see a Democratic resurgence, 15 of the 32 marginally Republican Wisconsin seats did not even see a Democratic challenger in 2014. This is another huge problem for the Democratic Party, as they look forward to trying to edge back towards a more equitable balance of power downballot. And that is a problem we will explore next week.
(Anyone interested in looking at the data on competitive districts and presidential performance utilized in this article can click here.)
(Anyone interested in how the redistricting process was conducted in each state can examine this very helpful graphic created by my colleague Stephen Wolf.)