What Happened: The Overview
After four long years of Trump, Joe Biden is heading to the White House – despite the GOP’s protestations to the contrary. And though ousting an incumbent president is no small feat, this was a bittersweet election for Democrats. Down ballot, there’s much less to celebrate. A Senate win depends on two special elections in Georgia; the House majority has been cut. And at the state house level, well, it was a bloodbath. Democrats failed to win chambers and lost seats they needed to hold. As a group that prides itself in understanding the data behind the states, we’re writing this to explain what happened. It’s a story told based on the last three years of working in the trenches of the states, but it begins with where we were just a short while ago.
“There’s Something We’d Like to Tell You”
Polling in state house races is much more scarce than national races and typically doesn’t start until mid-summer at the earliest. But from the first polls this cycle, we kept getting the same message from caucus directors: we’d like to talk to you about the latest polling numbers. And when we’d talk, the same perspective would become clear: Republicans were cratering in the suburban districts we needed to win to claim legislative majorities. Now was the time to push the investments toward the districts that would overcome Republican gerrymanders and allow us to take control.
Caucus directors are unsung heroes in the Democratic political space. Working for the state-level Senate or House members, they have to balance between keeping their bosses in office and building a Democratic majority. Throughout the cycle, they leveled with us on what candidates they thought had a chance, and which ones were weak. They clarified where candidates needed outside money like ours and where the candidates were doing just fine, thank you. So, we believed them when they showed us poll after poll, across the country, that had Democrats in striking position in the districts we needed.
And those polls matched what we were hearing outside of the state legislative bubble. Dave Wasserman of the Cook Political Report said that the suburban congressional polling was revealing a potential romp for Democrats and for Biden, inverting trends that he had seen in 2016 that foretold a Trump victory. The polls drove late investments, particularly as grassroots dollars flowed into campaigns and organizations following Justice Ruth Bader Ginsburg’s passing. How bad were the polls? In Texas, two publicly released polls showed Celina Montoya (HD 121) with a seven-point advantage (she lost 54%-46%) and Joanna Cattanach (HD 108) with a five-point advantage (she lost 50%-48%).
While the polling error drove optimistic decision-making this year, there were earlier signs that Democrats might face tough sledding in 2020. In Virginia in 2019, Democrats swept the State Senate and the State House, building on the massive 15-seat state house gain in 2017. But as we wrote on our blog in the wake of those results, “While the tremendous gains Democrats have made over the past two cycles must be celebrated and offer some direction for the future, it will be harder to replicate this success in the challenging environment of 2020. Over the past two years, Democrats have won the “easy” seats, the seats that fundamentally lean blue. Where we’ve failed to make substantial gains is with seats that lean red.”
Across both the Senate and the House in Virginia in 2019, the districts that Democrats flipped had a median partisan lean of D+7. The most conservative district, based on our Legislative District Index, was R+1. Democrats failed to win in multiple other R-leaning districts.
That spelled trouble for 2020.
The Landscape Entering 2020
In 2019, we published our Purple States Report, an attempt to characterize the competitiveness and demographics of the key state legislative districts that would be up in 2019 and 2020. We recognized then that the 2020 state legislative terrain was fundamentally different from the places that ended up seeing big Democratic gains in 2017, 2018, and 2019.
First of all, the 2020 landscape favored Republicans. We estimated that over 60% of the districts that Democrats needed to win to flip chambers leaned Republican. A decade of Republican gerrymanders had created a tremendous advantage within critical state legislative districts even if Democrats positioned themselves well for the cycle.
Secondly, within those districts, the opportunities for gains were tempered by the challenging demographic conditions. Of the 250 key districts we studied, only 44 mirrored the diverse coalition that Democrats had traditionally ridden to victory. Of the ten most Hispanic districts, five were in Florida, where conservative Cubans are predominant, and three were in Texas, where Hispanic voters have traditionally been more conservative.
The remaining districts were predominantly white. While many of those districts had large college-educated white populations, which powered Democratic gains in 2018 at the US House and state house level, more than half had a likely voter population at least 40% non-college white.
While these factors suggested to us early on that 2020 would not be a walk in the park, there were signs from 2018 that Democrats could make further gains. The 2018 cycle was a big night for Democrats, netting over 300 seats. And while, as we told the New York Times, these gains were largely in places Democrats shouldn’t have lost, they included nine pickups in the Pennsylvania House and 12 pickups in the Texas House. These gains happened despite two fundamental weaknesses: fundraising and recruitment.
In the wake of the 2016 election, a flurry of groups emerged in the state legislative space, ourselves included. And while that meant that there was some new money in state politics for Democrats, in 2018 it was pretty uneven. In our analysis of the state legislative candidates we worked with in 2018, unsuccessful candidates in competitive districts were outraised by their GOP opponents by $250,000 on average. We contend that the absence of candidate-level fundraising mattered even in places where so-called “independent expenditures” (or IEs) were active. Without the candidate’s ability to convey their own district-specific message, uncoordinated IEs can be minimally effective, at best, or muddy the waters, at worst. Anecdotally, candidates we endorsed in 2019 in Virginia who lost competitive races spoke to us after the election about how Democratic IEs in their district pushed messages that were unhelpful and counter to the messaging the candidate was trying to get out there to voters.
Candidate recruitment also emerged as a new focus after 2016. Groups like Run for Something, Emerge, and the National Democratic Training Committee placed resources into the recruitment work. But there were many misses on that front in 2018, with weak candidates (or no candidate) representing the Democratic Party in dozens of competitive districts in the key tossup states.
Those failings in 2018 were eminently fixable, giving us hope that with more focus than ever before, the state legislative results would be meaningfully different in 2020 – in a good way. Still, entering the cycle we were cautious. Of the 24 chambers up in 2020 that we studied in the Purple States Report, we estimated that only four leaned Democratic in the leadup to November.
The Bloody Details
Buoyed by the encouraging polling at all levels, we entered Tuesday, November 3 with high hopes, expecting big gains, even if they took some time to materialize. It wasn’t just the polling that offered encouragement. We had largely overcome the two hurdles that had bedeviled us in 2018.
Our candidates overcame many of the fundraising deficits we had seen in 2018. We’ll have a further fundraising analysis once the final reports are due at the end of the year, but generally speaking candidates in competitive districts had the resources they needed to compete. Some of the strongest fundraisers raised mid-to-high six figures, practically unheard of for Democrats in these states prior to this cycle.
And we had good candidates recruited in a wide range of competitive districts. While we noted that poor candidate recruitment was a hurdle in 2018, EveryDistrict endorsed over 100 strong candidates in nine states across the country. In the most competitive states, enough candidates with good campaign plans and good stories for the district to flip the chambers were in position this cycle.
Our predictions before the election, along with those of other commentators (see, for example, here), all indicated that Democrats were favored to make gains. As the night and week wore on, however, the grim reality set in. Rather than making large gains, nearly everywhere that we looked, Democrats lost seats.
The results of the election are described in the table below for the key state legislative chambers across the country. Results where we are still awaiting final results of recounts and absentee ballots are denoted in parentheses.
Table. 2020 State Legislative Results
|State||Chamber||Pre-20 D||Pre-20 R||Post-20 D||Post-20 R||Uncalled||D Net|
|Florida||Senate||17||23||16||23||1||(-1 to 0)|
|Kansas||Senate||11||29||10||29||1||(-1 to 0)|
|House||75||59||69||64||1||(-6 to -5)|
|Ohio||Senate||9||24||8||24||1||(-1 to 0)|
|Pennsylvania||Senate||21||29||20||29||1||(-1 to 0)|
Unpacking the Devastation
So, what happened? How did we go from actively expecting to break through in multiple states to losing ground almost everywhere? Some analysis has focused on tactics – we did not knock doors during the worst of the pandemic. AOC’s critique of many congressional Democrats’ digital programs may well apply to some state legislative candidates. Our assessment of tactics found that many campaigns still lacked the strategies to identify the key swing voters in their districts.
Moving beyond tactics, we think larger structural effects were at work in the results this year. It’s still very early and there’s a lot of data that needs to be gathered to understand the full picture. But looking at the data we do have, it appears clear that the gerrymandered terrain on which Democrats had to work exposed them to unique challenges that left them vulnerable in the conditions created in the 2020 election. Here are the four trends that we think mattered most.
1. Trump-driven turnout swamped in must-win state legislative districts.
The 2020 election was marked by high turnout. At least 150 million people have voted, and Donald Trump has earned at least 9 million more votes than he did in 2016. The geographic concentration of those votes spelled trouble for Democratic candidates across the country.
For example, in Texas HD 67, where our endorsed candidate Lorenzo Sanchez lost 48%-52%, the big GOP turnout swamped any chance of a Democratic victory. As the chart below shows, Lorenzo won enough votes to even overcome a more reasonable turnout increase, getting nearly 4,000 votes more than the GOP incumbent did in an 18-point blowout in the last presidential cycle. The final push of GOP vote meant it didn’t matter.
Table. The Shifts in Texas HD 67
|2020 Candidate||2020 Votes (%)||Change in Party Votes from 2016 (%)|
|Jeff Leach (R)||48,664 (52%)||7,224 (17%)|
|Lorenzo Sanchez (D)||45,289 (48%)||16,253 (56%)|
In Pennsylvania’s SD 15, a hotly contested race in Dauphin and Perry Counties, the continued power of Trump’s name in rural counties, and close alignment up and down the ticket, spelled disaster for George Scott, the Democratic nominee. In the Dauphin (Harrisburg) part of the district, Scott won 53%-47%. In Perry, the rural county included in the district, Scott’s opponent John DiSanto won 75%-25%. Trump improved on his 2016 vote total by 3,000 votes and nearly one percentage point, despite losing the state this time around.
The Trump effect was particularly meaningful, not surprisingly, in the districts with large non-college white populations. The table below compares the historic partisan lean of districts in Iowa and Kansas with their 2020 result and non-college white percentage.
Table. The Big Trump Boost with Non-College Whites
|District||Lean||2020 Result (D-R)||Non-College White Percentage|
|IA SD 8||D+0||48%-52%||60%|
|IA SD 44||D+2||43%-57%||58%|
|IA SD 46||D+2||41%-59%||53%|
|IA HD 55||D+2||46%-54%||55%|
|IA HD 91||D+5||43%-57%||53%|
|KS SD 20||D+3||43%-57%||42%|
|KS HD 60||D+1||40%-60%||45%|
2. The suburban Biden coalition didn’t deliver votes to down ballot Democrats.
In the lead-up to the election, supposedly cratering numbers for Republicans among suburban voters were supposed to bring big gains for down ballot Democrats. That prediction didn’t pan out in two ways. First, in the districts that had already flipped to Hillary in 2016, Republicans held on. At the congressional level, vulnerable incumbents like John Katko (NY) and Brian Fitzpatrick (PA) won re-election. At the state legislative level, one of the most glaring examples of this sort of disappointment was in Pennsylvania’s HD 151. In Montgomery County, which Biden won with 62% of the vote, Democrat Jonathan Kassa lost by six points in a district that Biden will have carried by 20. Biden won 33 precincts in the district while Kassa won 10.
The second type of district that was supposed to carry the day was the next round of suburbs, those that weren’t enamored with Hillary but were tired of the Trump clown show. At the congressional level, that was supposed to include Indiana’s 5th or Missouri’s 2nd, both places where Republicans ended up winning comfortably. In our world, Adam Ericson outside of Wilmington, NC (HD 20) offered a strong campaign built to compete in what had previously been challenging turf. He lost by nearly 11 points as the GOP lean of the community held strong.
Without a clear path in either type of next-on-the-list suburban district, Democrats had limited opportunities for gains.
3. The Latino underperformance hurt in Florida and Texas.
As we noted earlier, the winnable districts with large Latino populations were concentrated in the challenging states of Florida and Texas. Biden’s poor performance with Latinos became clear as returns rolled in on November 3 and he finished with only 56% of the vote in Miami-Dade County and lost a whole host of majority-Hispanic counties in the Rio Grande Valley of Texas.
In the state legislative world, Democrats failed to make meaningful gains in the most Hispanic districts in those two states, or Arizona, across both traditionally Mexican and traditionally Cuban populations. Here are some outcomes in six of the most Hispanic districts.
Table. Democratic Performance in Hispanic State Legislative Districts
|District||2020 Result (D-R)||Hispanic Percentage|
|TX HD 121||46%-54%||26%|
|TX HD 138||48%-52%||24%|
|FL HD 115||43%-57%||67%|
|FL HD 105||46%-54%||64%|
|FL SD 39||43%-55%||63%|
|AZ SD 8||41%-59%||23%|
While a number of factors kept us out of the winner’s circle in these districts, persistent Democratic underperformance among Hispanics didn’t help us get closer to overcoming Republicans’ control of these three states.
4. The state legislative realignment continued, dropping unsuspecting incumbents in what seemed like defensible seats.
The Trump surge affected our existing Democratic incumbents in two ways. Even after bad election cycles in 2010 and 2014, Democrats still held on to state legislative seats that leaned Republican. While we worried about these seats early in the cycle, most polling indicated that Democrats would hold them. That polling, like everything else this cycle, was wrong. Entering the cycle, there were just over 200 D-held seats with Republican leans. Many will have been lost by the time all the votes are counted. In the North Carolina House of Representatives, we saw some of the biggest losses of such districts, as shown below.
Table. Incumbent Losses in North Carolina – D-Held, R-Leaning Seats
|District||Lean||2020 Result (D-R)|
|NC HD 37||R+4||47%-50%|
|NC HD 43||R+3||48%-52%|
|NC HD 66||R+15||40%-60%|
|NC HD 93||R+9||47%-53%|
|NC HD 98||R+3||49%-51%|
|NC HD 119||R+8||46%-54%|
In a highly nationalized environment with high turnout, it is perhaps not surprising that such R-leaning seats would end up in the GOP’s hand. But the surge of Trump vote also resulted in losses for Democrats who were not expected to face serious trouble. In the Michigan House of Representatives, for example, we lost a D+3 (HD 96) and a D+6 (HD 48) district on a night when we were supposed to be making plays for the majority.
The trends that seemed to keep the race out of Biden’s reach in some key states and led to a smaller US House majority were felt acutely in the state legislative space. That spelled disaster for Democrats across the country.
Where Do We Go Next?
Democrats were hopeful to enter the 2020 redistricting cycle stronger than the last. Instead, Democrats are on the outside looking in, with the only saving grace that their control of certain governor’s mansions and the passage of independent redistricting commissions will somewhat limit GOP dominance. Democrats are, as they were in 2010 and in the wake of 2016, a largely powerless majority. As the 2022 elections loom ahead, Democrats will need to identify ways to rebuild with Latinos, bring enough non-college whites into the fold, message the value of Democratic local officials to newer members of the presidential coalition, and ensure that state-level campaigns have the tactics and data to win.
Most of all, Democrats will need to better focus in on the geography for power: where do Democrats need to win to take back the states and the Senate? What will it take to do so? Over the coming weeks, we’ll be laying out our plan for how a Democratic Party rooted in a state-level, grassroots revival can create national power at all levels.