The purple-state problem
Swing states might be in the ideological center, but data shows they may be closer to one end than the other.
Most close readers of this column know that I see our nation’s political divisions as so great as to make the last line in the Pledge of Allegiance, that we are “one nation” and “indivisible,” hardly descriptive.
While some states or congressional districts are characterized as red for Republicans or blue for Democrats, purple states or districts are thought of as fair fights where neither party has a distinct advantage. In those purple regions, it’s often thought that to win, a candidate must do well with voters near the ideological center.
It is tempting—and more often right than wrong—to generalize purple states and districts as roughly midway between Republicans on the right and Democrats on the left, while also acknowledging growing populism on each end of the ideological spectrum, though not necessarily on every subject or in every state.
One of the few truly blue-chip quality polling operations in the country—I am a poll snob and think most public polls are varying degrees of rubbish—is the Pew Research Center.
Last month, Pew released a study looking at the national political landscape 100 days before the midterm election. President Trump’s approval rating among 3,554 adults interviewed showed 34 percent approving his job in office and 64 percent disapproving. Forty-three percent of the 2,628 registered voters indicated a preference for Democrats for the U.S. House, to 37 percent for Republicans.
The terrific political research team at Pew was kind enough to break out their results, detailing the 1,477 adults (1,039 voters) in 24 red states, the 1,465 adults (1,122 voters) in 19 blue states, and the 606 adults (462 voters) in seven purple states.
Looking at these data, two points jumped off the page. The first was that Republicans were surprisingly weak within their 24 base red states, leading by only 4 points, 39 to 35 percent, and leaving 26 percent up for grabs. Those interviewed in 19 blue states favored Democrats by a whopping 16 points, 50 to 34 percent, with 16 points on the table. The two parties were tied at 42 percent in purple states.
Many would focus on how much stronger Democrats are in blue areas than Republicans are in red areas. But this means that Democrats are “wasting” a lot of support. In effect, they’re running up the score in blue states, where there are few GOP seats to flip, and inflating their national showing. In purple states, where Democrats want an edge in a politically favorable environment, the support levels are even.
That leads to a second and closely related finding. Across several questions, voters in purple states are more closely aligned with red staters than with blue staters.
The Pew folks note that some of the difference might stem from an oversampling of Hispanic and Asian voters as researchers look more closely at their voting behavior. But in the national numbers, they were weighted to their normal proportions in the electorate. It’s also important to note that only 462 voters were surveyed in purple states, compared with 1,039 in red and 1,122 in blue states, making the margin of error in the purple states. Maybe one or both of these factors are important; or maybe not.
Another explanation is that the makeup of the electorate in blue states is far different than the rest of the country. Blue states tend to have much larger shares of urban, major metropolitan areas and nearby suburbs, while purple and red states tend to have plenty in rural areas, small towns, and exurbs. Blue states hold a much larger share of people with college and advanced degrees, who skew increasingly Democratic. Those same blue states have smaller shares of non-college-educated whites, a group that Democrats are increasingly struggling with. And unionized workers in blue states are more likely to belong to the public sector, whose members are more likely to vote Democratic than their counterparts in construction or manufacturing unions.
In the 24 states that make up Red America, states that voted for Trump in each of the last three presidential elections, Republicans hold all 48 U.S. Senate seats and 22 of 24 governorships—Govs. Laura Kelly and Andy Beshear break the pattern in Kansas and Kentucky, respectively. About 197 (45 percent) of all House districts can be categorized as red, with presidential election patterns of voting more than 5 percentage points more Republican than the country as a whole, according to the revised Cook Political Report Partisan Voting Index (PVI), re-released last week to reflect redistricting in ten states since 2024.
Similarly, in the 19 states that constitute Blue America, which voted Democratic in each of the last three presidential elections, Democrats hold 37 of 38 Senate seats, with Susan Collins in Maine the exception, and 17 out of 19 governorships—all but New Hampshire and Vermont due to Govs. Kelly Ayotte and Phil Scott. Note that the three blue states with a Republican senator or governor are the three states that make up northern New England, where the GOP tends to be less MAGA-oriented than elsewhere. Of all 435 House districts, 161 (37 percent) could be considered blue, Democratic-tilting districts.
That leaves the seven swing states of Purple America: Arizona, Georgia, Michigan, North Carolina, Nevada, Pennsylvania, and Wisconsin. In 2016, Trump carried six out of seven—all but Nevada. Former President Biden won six out of seven in 2020—all but North Carolina. In 2024, Trump swept all seven. Of the 14 Senate seats in these seven states, Democrats hold 10 and Republicans hold four. Democrats hold five of the governorships, and Republicans hold two. And 77 House districts, 18 percent, can be categorized as purple districts, voting no more than 5 percentage points in favor of either party than the country.
Since Trump was first sworn into office, there have been 21 Senate elections in those seven swing states; Democrats won 17, to Republicans’ four. In that same time, Democrats have won 10 of the 14 gubernatorial races in those seven states.
A good case can be made that since Trump took office, a disproportionate share of Republican nominations have gone to MAGA-style Republicans, ones more at home in red states and districts than in purple states and districts where less ideological candidates might run stronger. Less-MAGA Republican candidates who may have been strong in a general election may also have chosen not to run because they’d struggle against more MAGA opponents in primaries. This could become quite important, but with the shoe on the other foot. Democrats are now nominating candidates in some swing constituencies who would be stronger and more appropriate in bluer places.
The Michigan Democratic Senate primary tomorrow and the Wisconsin gubernatorial next week could become examples of this: Democrats nominating blue candidates in purple places.
Much of the macro-political analysis of the U.S. House midterms these days is driven by a focus on results of generic congressional ballot test polling, a metric that I have followed and used to an extent for more than four decades. But there would seem to be an overreliance on the findings of that question, even though the footprint of competitive House races is so much smaller than it used to be.
In an old joke, a man with one foot in a pot of boiling water and the other in a bucket of ice is asked how he felt. “Very comfortable, on average!” Generic ballot tests now serve a similar purpose, dipping into boiling-red regions and freezing-blue regions, masking as much as they reveal. A generic ballot test might be a good measurement if a district is a generic one, but these districts are increasingly anything but.