part one—re-evaluating the landscape post-2018
If one were solely to base their assessments of the 2020 electoral map on 2016 results, the prediction I offered above the fold would seem optimistic at best. While Minnesota and Texas were moving in opposite directions, they were both moving. As mentioned above, Texas moved roughly seven points in the Democratic direction between the 2012 and the 2016 presidential elections. But, at the same time, Minnesota also moved considerably between those elections, with a more than six-point shift toward the GOP.
But what should make Democrats more optimistic about Texas, and Republicans less bullish about Minnesota, is what we saw in 2018. For an election that analysts on election night desperately tried to portray as anything other than a wave, the data for Democrats on that night was unremittingly good, save for the most favorable Senate map for Republicans in generations.
But even accounting for the national climate, 2018 was abnormally good for Democrats in both Texas and Minnesota.
Start with Texas, where the focus in 2018 was almost entirely on Democrat Beto O’Rourke’s extraordinarily strong Senate campaign, one that ended in a near-upset over Sen. Ted Cruz. The focus on Beto was in no small part deserved. And it should absolutely be used as a cudgel against the litany of so-called smart political thinkers who decried his historic fundraising efforts in an “unwinnable” cause.
But the thing about Texas in 2018 is that the best news for Democrats was the breadth of their competitiveness, not just the outcome of a single race. Beto’s money and his campaign presence were undoubtedly contributing factors, but there were eye-popping headlines across the board for Democrats in the Lone Star State. They built their new majority in the U.S. House by picking off two seats in suburban Texas. Republicans also saw their state House of Representatives majority (which had been viewed as insurmountable) cut in half, leaving Democrats with the very real prospect of seizing the state House majority in 2020, given that they only need to pick up nine seats (they nabbed 12 last year), and that Beto O’Rourke actually carried a scant majority of the state House seats in his losing effort last year (76-74).
But perhaps the most eye-popping Texas electoral statistic largely went unnoticed in the wake of the 2018 election. Oftentimes, down-ballot statewide elections (for constitutional offices such as lieutenant governor and attorney general) tell us a great deal about the generic lean of a state. The rationale for this is really quite simple: Campaigns for these offices are often outshone by the big statewide races (governor/senator), and voting based on partisan preference is considerably more common. With that in mind, consider the following:
This seismic shift of 15 points in the direction of Democratic candidates (and the very similar results in the lesser-known statewide races) is a greater one than we saw between the two presidential contests of 2012 and 2016 (the aforementioned 7-point shift), and it is also a greater shift in the performance of Sen. Ted Cruz between 2012 and 2018 (12 points). That this “row office” shift was greater than the shift in Cruz’s fortunes is especially notable when you consider that, while Cruz faced an extremely well-funded foe in 2018, his challenger in 2012 was a well-regarded but lightly funded former state legislator, Paul Sadler.
The collapse of what seemed to be a semipermanent and substantial GOP edge in these offices got lost in the shuffle, mainly because none of those huge shifts actually resulted in a shift of partisan control of an office. That said, it is worth evaluating, because it shows a rapid change in voter preferences. If nothing else, it tells us that something is happening in Texas. And it’s happening faster than we thought.
Meanwhile, 800 miles to the north, Republicans began the 2017-2018 cycle fairly bullish about their prospects in Minnesota. The 2016 elections had been a nearly unmitigated success for the red team in Minnesota. In addition to Trump nearly claiming the state’s 10 electoral votes, Republicans seized control of the state Senate (turning a 39-28 Democratic majority into a razor-thin 34-33 Republican one), and expanding on the state House majority they had won in 2014. What’s more, Democrats were initially optimistic about their prospects in the suburban U.S. House seats of MN-02 and MN-03, but lost both. Instead of hailing key wins in the House, Democrats had to wait late into the night to see if they had held on to their seats in MN-01 and MN-08, both of which stayed Democratic by less than a percentage point.
By the spring of 2018, Republicans had even more reasons for optimism. It started in March 2017, when Rep. Tim Walz of the swing-y seat in MN-01 announced his retirement from the House in anticipation of a gubernatorial bid. That was followed by the resignation of Democratic Sen. Al Franken, which would necessitate a special election in 2018 for his seat. In February of 2018, Democrats lost another vulnerable House incumbent when Rep. Rick Nolan in MN-08 announced his retirement. Finally, in April, former two-term Republican Gov. Tim Pawlenty announced that he would run for governor, evidently sensing opportunity to return to politics in a state he had won when the general lean was a bit more toward the blue.
As it happens, that April announcement was the high-water mark for the GOP in Minnesota. In the August primary, Pawlenty was upset by 2014 nominee Jeff Johnson, who rather easily (53-44) eclipsed the former governor. The real reckoning came in November, however. While the GOP, as a number of people expected, picked up both MN-01 and MN-08 from the Democrats, those gains were offset by Democrats seizing the two suburban seats they had missed out on in 2016. Meanwhile, Democrat Tim Walz not only defeated Republican Jeff Johnson for governor, but he did so by the largest margin of victory in a Minnesota gubernatorial race since the 1994 re-election bid of moderate GOP Gov. Arne Carlson. Meanwhile, despite the ability of Republicans to recruit a decent challenger (state legislator Karin Housley), the election bid of appointed Democratic U.S. Sen. Tina Smith was never in doubt, as she cruised to a 53-42 win. Democrats also gained 18 seats in the state House of Representatives to reclaim the majority in that chamber (the state Senate is not up for election until 2020).
Turnout in the midterm elections in Texas was absolutely massive. Turnout in 2018 was an almost inconceivable 93% of the presidential election turnout in 2016. Indeed, there were a handful of counties in Texas that actually saw higher turnout in 2018 than they saw in 2016.
One of those was Williamson County, and it serves as an excellent case study in why Democrats should feel some confidence about their future prospects in the Lone Star State. Williamson County lies just north of Austin, and, along with Hays County (just to the south of Austin), it is among the fastest-growing counties in America. Williamson County, after the steady realignment of the South from the 1960s onward, was one of the anchors of Republican strength in the state. In the 2000 election, Texas Gov. George W. Bush annihilated Al Gore in Williamson County, 68-28. Even without a Texan on the ballot, though, Democrats continued to struggle in Williamson County: Mitt Romney easily bested President Obama here by a 59-38 margin in 2012.
Then, in 2016, Williamson County swung sharply toward Democrats—even more sharply than the state as a whole. Hillary Clinton still trailed in Williamson, but by only a 51-41 margin, effectively halving the Republican advantage here.
In 2018, Williamson County went for the Democrat, Beto O’Rourke, by a 51-48 margin. More importantly, the county was one of 12 that actually saw their turnout double compared with the previous midterm election in 2014. O’Rourke carried seven of the 12 counties, and of the five Cruz counties where the turnout doubled, the Republican margin was cut by over 20 points (versus Cruz’s 2012 performance) in three of them—Brazos, Collin, and Denton.
Therein lies a foundational crisis in the Republican electoral calculus in Texas. It must be understood that, in spite of the red-state status of Texas over the past few generations, there have been substantial pockets of Democratic support all along. As is true in most places, those pockets today are in well-populated urban counties: Dallas, Harris (Houston), Bexar (San Antonio) and Travis (Austin). As has been true, to varying extents, the blueness of these urban areas has been a somewhat recent development (especially so in Harris County), but even when Democrats were losing them in the 1980s and 1990s, it wasn't normally by outsized margins. Today, all four are blue by double digits at the presidential level. Marry those four large urban counties with the largely Latino counties in the Rio Grande Valley, and there are a lot of Democratic votes in the state.
But what has saved the GOP in Texas has been a potent combination of its own growing and now almost universal support in the predominantly white rural corners of the state, and overwhelming support in the mostly white suburban counties that surround those urban centers, which with each passing election become a greater proportion of the overall electorate.
Let’s start with the rural counties. Remember that Texas has 254 counties, and a vast number of them are small, white, and (of recent vintage) monolithically Republican. Take my mother’s birth county: Navarro County, which lies between Dallas and Waco. Navarro County was a 73-24 Trump county in 2016, and Ted Cruz did nearly as well in 2018, carrying the county 72-27. The times, they are a-changing: Thirty years earlier, Michael Dukakis carried Navarro County for the Democrats. But, alas for Republicans, these counties are not growing quickly. Navarro County had 13,232 people turn out to vote for president when Dukakis carried the state in 1988. When Trump dominated the county in 2016? There were 16,433 people that voted for president, a somewhat soft increase of 24%. By way of comparison, statewide presidential turnout increased 66% between 1988 and 2016. This is a dilemma for Republicans.
Another dilemma for Republicans? The times are a-changing as well in those suburban counties that have been essential to their margins of victory. Beto O’Rourke didn't just carry the four big urban counties and the smattering of Latino-dominated counties along the Rio Grande. He also won Fort Bend County (west of Houston), Hays and Williamson Counties (surrounding Austin), and even claimed Tarrant County (Fort Worth). And this was not just a Beto-centric phenomenon. To put it in a slightly larger context, consider the following:
To put this in clearer terms, let’s take Dallas County out of the equation. In 2012, the combined Republican presidential advantage in Tarrant, Denton, and Collin Counties was just shy of 268,000 votes. By 2016, that margin had been whittled down to just 177,632 votes. In 2018, Ted Cruz beat Beto O’Rourke in these three counties by 41,598 votes. Given the way that the blue urban centers of Texas are growing even more deeply blue (Dallas County went from D+15 in 2008 to D+26 in 2016), the GOP cannot afford further erosion in these suburban pockets, especially since that’s where the real electoral turnout growth has been seen over the past several cycles.
Minnesota did not have the meteoric turnout growth that Texas did between midterms. That is considerably harder to accomplish in Minnesota, where voter turnout almost always is sky-high (and, indeed, was the highest in the nation in 2018). Despite the absence of such an increase in turnout over a four-year span, there are still lessons to be learned from the 2018 turnout in the state.
Last year, 10 counties experienced turnout that was at least 135% of the turnout for the prior midterm in 2014. The 10 counties split evenly in their voter selections in 2016: Five counties were carried by Donald Trump, and five counties were carried by Hillary Clinton. So at first blush it wouldn’t appear that the turnout spike in Minnesota had much probative value.
But, notably, Tim Walz won eight of the 10 counties that saw this turnout spike, and he won six of those by double digits. And in the special election that originally had Republicans salivating, Tina Smith also won seven of the 10 high-turnout counties (barely losing Anoka County, home to the northern suburbs of the Twin Cities).
Where the real lesson lies, it would seem, is at the opposite end of the spectrum. While Minnesota turnout was robust (up 31% from 2014), there were six counties in the state (out of 87) that saw turnout actually decline between the two midterm elections. There was not much that wedded them together—they were all relatively small, and ranged from the northernmost to southernmost corners of the state. As it happens, they really had only one commonality—Donald Trump carried all six of them, and he carried each of them by a margin of 20 points or more. What’s more, those same six counties also had anemic turnout growth while giving Trump outsized margins of victory, with five of the six also dipping in turnout between 2012 and 2016.
What does this mean? In a nutshell, it means that Trump’s gains in the state might be close to maxed out, because where he did the best also happens to be the places where the population in the state is stagnant. Meanwhile, the places where turnout growth in 2018 was the most substantial (in the state that had the best voter turnout in the nation) were, with only a couple of exceptions, in the urban and suburban parts of the Twin Cities. That combination is going to make it hard (though not impossible) for Trump to repeat his 2016 performance, let alone close the gap.
Also, as I noted in a postmortem on Virginia back in early 2018, rapid voter growth in heavily populated (and blue-leaning) areas is exceedingly tough for a Republican campaign to offset:
“…that’s not even taking into account Fairfax County, which saw Northam defeat Gillespie by a 37-point margin, and saw turnout at nearly 123 percent of the McAuliffe/Cuccinelli race. To put that in perspective, that means there were nearly 70,000 more voters in Fairfax County this year than there were in 2013.
To allow for even more perspective, take all of the larger localities that voted for Ed Gillespie. There were 33 of them, and their combined turnout saw only about 71,000 more voters hit the polls in 2017. In other words, all of the turnout increase in the red counties/cities in Virginia were essentially offset by Fairfax County alone.”
A similar dynamic was seem in Minnesota. The five Trump counties that ranked in the top 10 for turnout growth between midterms saw a raw vote increase of just over 82,000 votes. That growth was offset—and then some—solely by the turnout increase in Minneapolis’ Hennepin County, which saw an increase of over 180,000 votes all on its own.
I feel compelled to point this out before we conclude: Readers should bear in mind that, while there is ample data that looks very good for the Democrats in this piece, every election brings results that defy the data that came before it (see: Michigan, 2016).
That said, there is a lot here to indicate that Democratic fortunes are on the rise in Texas, even if that does not necessarily portend a blue Texas in 2020. The directional indicators are not great for Republicans, especially based on what we saw in both 2016 and 2018. The rural counties at this point are more or less cashed out. There just isn’t much farther for Democrats to fall in those places. And, in addition, voter turnout growth in those rural counties is comparably weak. The 100 smallest counties in Texas (based on 2018 voter turnout) saw midterm turnout that was 84% of the presidential voting turnout in 2016. While there is a smattering of heavily Latino border counties in that mix, this is largely GOP turf—if you average the 2018 Democratic performance (governor plus Senate) in these counties, you get just 21% Democratic support. Meanwhile, when you consider the 10 largest counties, which are considerably more blue (54.5% Democratic performance), we see turnout that was 96% of the 2016 presidential turnout.
A similar dynamic is visible in Minnesota. That Trump and the GOP have made inroads in parts of the state is undeniable. His dilemma is that those are not by and large areas where the voting pool is growing. What’s worse, some of the places where the GOP can afford Democratic growth the least were the places that were most inspired to turn out in the midterms, an election nationalized in no small part as a referendum on Trump. Hennepin County (Minneapolis) ranked 44th out of 87 counties in terms of four-year turnout growth between 2012 and 2016. By that same metric, between the two midterms? It ranked third. Similarly, Ramsey County (St. Paul) went from 65th in turnout growth between the presidential elections to seventh between the midterms. Translation: The Trump presidency has clearly awakened dormant voters in the two bluest and most urban counties in the state.
The dilemma for Republicans is that they have clearly turned off voters in urban and suburban areas, and those locales are where the voter participation growth is to be found. They’re getting dangerously close to maximizing their margins in rural areas, while Democrats still have plenty of room for growth in those high-turnout suburban counties. Meanwhile, turnout spikes in places such as Hennepin County, Minnesota, and Travis County, Texas, show that Trumpism has awakened dark-blue urban centers as well. For the GOP to continue its presidential winning streak in Texas, and to have any hope of ending the Democratic presidential winning streak in Minnesota, it is going to have to conjure up votes in rural areas, somehow reverse the trend and stop the bleeding in the suburbs, or find a way to put the urban centers back to sleep. Any of those things is possible, but none of them, as we sit here today, feels like a likely outcome.