2020 Senate Forecast

Map created at 270toWin.com

2020 House Forecast

Map created at 270toWin.com

Despite always-robust fundraising numbers by the known-to-be competent Republican National Committee, the 2020 cycle was destined to be a rough one for the Republican Party’s congressional caucus, as evidenced by another six “strategic retirements” by House Republicans over the 26 who retired prior to the 2018 cycle. And that was before the COVID-19 pandemic arrived to strip the party of its best line of defense (Democrats are coming to take away your private health insurance) and shine an even brighter light on its core weakness: unpopular incumbent Republican President Donald Trump

As the Republican Party seeks to defend its last holdings in America’s suburbs in the House — and defend its Senate majority via the slate of swing-state seats up in this season’s more “blue wave-conducive” Senate cycle — some Republicans, including President Trump himself, are hostile to enacting electoral reforms to enable “remote” voting, believing erroneously, that such systems electorally benefit Democrats. Yet, the results of the Wisconsin primary, which had nested inside it a race for a Republican-held state Supreme Court seat the party was keen on defending, suggest Democrats can still win even under low-turnout scenarios. In fact, the GOP’s efforts to force the state to hold the primary on its original date at the height of the pandemic created blowback, both greatly increasing the overall salience of the election in the media and Democrats’ desire to vote in it. Ultimately, what determines electoral victories in the partisan era is the partisan composition of the electorate: the percentage of the electorate that is composed of Democrats and left-leaning independents versus the percentage that is composed of Republicans and right-leaning independents. And in 2020, in places where partisan competition is equalized (places like Wisconsin), it will be the coalition that is angrier and/or more frightened that wins that battle, even when voting infrastructure is manipulated to discourage participation. 

This “coalitional” turnout advantage for the out-of-power party, combines with an advantage among a small, but still very important, voter block: “pure” independents. Pure independents are that 5-12 percent of any given electorate who are real independents, that is to say, they are not what political scientists call “closet partisans.” Independent leaners — those voters who report being independent, but when pressed by pollsters to disclose whether they “lean towards” one party or the other readily admit they do — behave similarly to their confessed partisan peers. They vote and exhibit policy attitudes much like their partisan peers. While it’s not a perfect overlap, independent leaners overwhelmingly vote for the candidate represented by the party they lean toward. 

But “pure” independents are different. Although some are highly engaged and highly informed, many are fairly apolitical. They don’t have strong opinions on political issues or candidates because they don’t have a lot of interest in politics. Often, however, they vote regularly out of a sense of civic duty, or as remnants of political socialization they received as children from parents who voted and instilled the importance of voting in their children. 

They observe America’s increasingly dysfunctional political parties with disdain, often reporting that they dislike both candidates in presidential elections and, my research theorizes, break predictably against the status quo so long as the opposition party effectively frames the party in power as incompetent and corrupt. 

I open my 2020 congressional forecast with “The Parable of the Wisconsin Special Election” because it perfectly encapsulates what the fall of 2020 is likely to look like. Everywhere and anywhere competitive elections run up against Republican control of some or all of a state’s government, it’s probably reasonable to expect some degree of election-administration strategic skullduggery. Over the course of the past decade the Republican Party has hardly been coy about their electoral strategy of “subtraction,” helped along by an American media system keen to cover elections in a “nonpartisan” manner, which allows much of this strategy to be normalized even within the “World’s Greatest Democracy.”

Every election cycle there are stories of voter-roll purges and registration matching, and in recent years, states implementing voter ID laws, sometimes clearly designed to suppress voter access, but these efforts are not the “bread and butter” of voter suppression or voter enfranchisement. 

That’s because voter participation is largely determined by the institutional design of a state’s voting system. The main determinants of whether a state’s system is suppressive or built to enhance enfranchisement (meaning, is the system designed to facilitate voting and maximize voter participation or it is institutionally hostile to voter participation) are the state’s rules and procedures governing the participation in, and the conduct of, its election system. Generally, in states where voting is made easy, more people vote, and in states where it is made difficult, fewer people vote. 

And while using registration and voting rules to influence elections may seem a modern tool, American politicians and parties have always manipulated access to the ballot box as a way to hold onto, or prevent others from gaining, political power. Indeed, the techniques the Republican Party deploys today are tricks of the trade invented by conservative Southern Democrats, who used similar methods to lock black voters out of political power in the South for nearly 100 years after the Constitution was amended to formally grant them the right to vote, and to try to hold onto their declining power as the South realigned to the Republican Party in the wake of the civil rights movement. 

Voter suppression is best measured at the 30,000 foot level rather than with microstudies testing the effects of one or two novel actions. As you can see from Mike McDonald’s gold-standard election turnout data, stored at his United States Election Project website, voter turnout rates by state are highly variable, even within a “wave” cycle like 2018. For longtime readers of my research, you’ll recall I was unequivocal that the entire slate of California’s Orange County Republican-held districts would flip, en masse, in the 2018 Blue Wave, starting on day one of my midterm forecasts, released on July 1, 2018. One of the reasons I could be so bold in predicting Democratic victories in districts that had been solid red since the Reagan administration was the institutional design of California’s voting system, which had recently been reformed to enhance, rather than to suppress, voter participation. When the negative partisanship model meets a contest with demographics ready to produce a large turnout surge, and voting is easy, the probability for a Democratic victory is enhanced. Indeed, it’s this same “ease of access” that helps keep these California districts safely out of reach for the GOP in 2020, something the other race handicappers and forecasters will eventually figure out (though Nathan Gonzales of Inside Elections also senses the size of the Democrats’ embedded edge in California). 

I’m hesitant to say the CA-25 special election on May 12 for the seat vacated by Rep. Katie Hill, who was forced to resign partway through her first term in a sex scandal, will be a bellwether, but it’s the best chance Republicans have at reclaiming one of these seats this cycle. Democrats have been hesitant to spend heavily on the race whereas Republicans, increasingly bearish about their fall prospects under Trump, badly need any positive optics they can rustle up, and unlike other 2020 races where fundraising and candidate recruitment have reflected a hostile political atmosphere that still favors Democrats, their recruitment effort in California’s 25th district produced a top-notch nominee. 

Ironically, outside of gender, Democrats struggle to embrace the electoral power of descriptive representation (i.e., identity politics) when it comes to candidate recruitment; however, Republicans have long understood the benefits of recruiting and running minority candidates in diverse districts. It is a strategy that has worked well for them in both Florida and Texas, and one they’re hoping might mitigate a massive demographic advantage for Democrats on the 25th. Their candidate, Mike Garcia, has an impressive background as a Navy fighter pilot, and perhaps more relevant for the GOP’s electoral purposes, Garcia’s Ballotpedia entry provided by his campaign notes he is a “first-generation American citizen whose family came to the United States legally.” Republicans hope he may be able to pull over some voters from the district’s 37-percent Latino population and potentially erode some of the Democrats’ advantage among this core voting bloc. 

As my 2018 voter file analysis of the district reveals, a surge in turnout among Latino voters played a pivotal role in the district’s transition from red to blue. They doubled their 2014 turnout rate of 25 percent to 54 percent despite a massive uptick in registrations. As such, a strategy that offsets Democrats’ advantages among Latinos is critical for Republican efforts to reclaim these, and similarly situated, seats. So, CA-25 may be a beta test for the more expansive strategy the GOP plans to deploy in 2020 in similar districts and states such as Arizona, where the party is poised to lose a second Senate seat and see the longtime Republican stronghold flip to the Democrats in the Electoral College. 

But it’s hard to predict what will happen because, although a special election normally has lower participation, and one that occurs during a pandemic may have even lower-than-normal turnout due to the inability of campaigns to engage in critical get-out-the-vote activities, as the pandemic emerged, the state’s Democratic governor, Gavin Newsom, issued an executive order switching the election to an entirely vote-by-mail system, which could cause a turnout increase. Unlike most of the other California districts from 2018, the 25th has a modest registration advantage for Democrats and every voter in the district received a ballot. For now, I remain bullish on Democrats holding onto this seat. 

As I discussed in my presidential forecast, prior to the COVID-19 pandemic, I was anticipating something approaching — perhaps even reaching — 70 percent turnout in the 2020 general election, which is about 10 points higher than the 2016 turnout and would be history-making. This prediction was driven by the fact that 2018 midterm turnout increased just over 13 points over its 2014 levels (I was expecting about 10 points) and that based on my own calculation, in contests and cycles since Trump’s 2016 election, the average turnout increase has been about 10 points. The 2018 election overperformed my bold estimates because, unlike with Democrats during Obama’s tenure, turnout of Republicans in 2018 did not decline, it increased. 

The image of a disaffected Republican Party, embarrassed by their “chaos” president, so far runs into an irrefutable data-reality that Republican voter turnout, even in the 40 suburban districts Democrats flipped, was robust — and it did not break in favor of Democrats in rates any higher than normal for the polarized era. Instead, the blue wave that washed through America’s suburbs in 2018 was powered by a massive turnout of Democrats and independents, who showed up in droves to toss Republican House incumbents out of office and send a message to Donald Trump. 

As the GOP struggles to defend itself from a second cycle powered by Trump backlash, I expect it may resist any efforts to expand voting access even in states where it has the power to do so. That said, although the top-line turnout number is important, at the end of the day (and as Wisconsin shows) suppression can only take you so far in the face of a riled-up opposition. Far more important than the overall turnout is who is voting. As my colleague,  Niskanen Center President Jerry Taylor, eloquently put it after the negative partisanship model delivered him a successful election prediction: Right now, Wisconsin voters “would drink bleach for hours in those lines to kill the GOP,” an oddly prophetic claim given that the president suggested bleach as a potential cure for coronavirus at his press conference on April 23.    

Whatever 2020 turnout is, barring something extraordinary that disrupts the election, if more Democrats and left-leaning independents vote than did so in 2016 and pure independents break against Trump and congressional Republicans, Democrats will not only hold their 2018 House gains — they are poised to expand on their House majority and are competitive to take control of the Senate. In fact, at the writing of this article, the COVID-19 death toll exceeds 50,000 Americans. Yet Republican-led states like Florida and Georgia are reopening their states even in the midst of rising infection and death rates. By the time we enter the fall the death rate from the pandemic may well reach the dire predictions of early models

In addition, the pandemic has leveled both the American and global economies, and America’s weak social welfare infrastructure and Republican control of the Senate have blunted the types of aid being made available to displaced workers. As they say, voters are likely to forgive Trump the crisis itself, but they will judge him on how he, and by extension, his party, handles the response, a record that is already plagued with allegations of mismanagement

We’ve already witnessed something unprecedented in terms of the public’s reaction to the crisis: a near-total lack of what is known as the “rally around the flag effect,” which refers to the public’s tendency to rally around their leader during national crises. To be sure, only two presidents have served during measurable mass polarization, Obama and now Trump, and any president serving under polarized conditions might expect a muted “R.A.T.F.E” as a result. Yet, polarization cannot account for the total lack of any effect for Trump because mass polarization, like its elite polarization cousin, is asymmetrical, tending to be stronger among Republicans than among Democrats. The lack of any “national unity” effect speaks to more than polarization — it suggests that there is something more going on, something specifically lacking in Trump’s handling of the crisis. And that is a bad sign for the president, because when a crisis of this magnitude sweeps in, the public tends to start out forgiving and end it angry. Presidents need that reservoir of goodwill to drain out slowly, over time, as the inevitable inadequacies of crisis response take their toll. Trump will not have a goodwill reservoir to draw from as the extent of the crisis begins to materialize — a scary prospect for his reelection team. 

That said, polarization will likely provide Trump a robust cushion regardless of how severe the crisis gets, a solid floor that presidents from other eras could not have enjoyed. Jimmy Carter, for instance, was also a victim of circumstance when his 1980 reelection campaign faced similarly daunting conditions. Like Trump, Carter entered his reelection year deeply unpopular, even more so, in fact, because in 1980, polarization did not exist, and public opinion was more elastic and less partisan. Indeed, without polarization and the hyperpartisanship that fuels it to inoculate him, Carter found himself in the unenviable position of having to defend himself from a challenge for the Democratic Party’s nomination by Sen. Ted Kennedy. As if that wasn’t enough, the American economy was in a long-term tailspin as the oil shortages of the 1970s fed into an inflationary crisis and high unemployment. And then, the Iranian hostage crisis came along, providing the Reagan campaign the perfect political wedge issue to exploit. Sound familiar? It should, because in the electoral bloodbath that followed, down-ticket Republicans flipped control of the Senate from Democrats for the first time in 25 years. They also used the chaos to remake the American economy, but that is a story for another day. 

Deep in the heart of Texas 

In 2018, under more favorable conditions for Republicans, the Democrats raked in 40 House seats and might have picked up more if the party had leveraged a more modern strategy. As I noted back in August 2019, based on my theory and model, which is interested less in historical performance and compelling candidates (although it is always nice to have one) and more in current demographics of states and districts, there were several realigning suburban districts in 2018 that the Democratic Congressional Campaign Committee did not invest in or recruit candidates for that, based on the two-party vote share the party’s nominee earned that cycle without party support, likely would have been winnable had the party chosen to pursue them. These districts are predominantly located in Texas, in the realigning suburbs of Dallas and Houston and in the heavily gerrymandered districts that tap into portions of Austin. 

But, before I explain why these districts are such attractive targets for Democrats in 2020, I want to make it clear that Democrats can only run deep into these Texas districts via significant strategic adjustments to both their voter-targeting and messaging strategies. These are not like the Orange County districts of 2018 where victories were all but assured no matter the strategy deployed. For a good number of the 2018 Democratic House freshmen, their 2020 reelections will likely play out similarly to those of the 2017 freshman Democrats in the Virginia House of Delegates, who stood for their first reelections in 2019 and overperformed (well, at least other people’s expectations) by winning decisively. This even includes a victory by Lee Carter, a self-described democratic socialist who ousted a five-term Republican incumbent in 2017 and, despite robust attacks from the state’s Republican Party over his ideological extremism and an endorsement by Bernie Sanders, won his 2019 reelection bid by seven points.

This is not the case in most of the “Texas Eight,” the districts in play in the Lone Star State. If Democrats want to win them, they will need a modernized, strategic approach that recognizes that although Republican refugees are a part of the story of the suburban transformation, it is driven by demographic and generational changes. America’s suburban renaissance is a story about Gen Z, millennial, and even tail-end Gen X college-educated voters who were ambivalent about voting prior to 2016 but have been galvanized to vote since America gave them Donald Trump. And in their defense, these voters have largely been ignored by the Democratic Party’s campaign apparatus, which until recently focused its energy nearly exclusively on so-called persuadable voters. Campaigns are now branching out their strategies, including GOTV efforts, and focusing on techniques long deployed by the GOP to win elections, such as early voting and absentee voting.

It helps, of course, that millennials have reached “a certain age”: that age when one (pre-pandemic anyway) might choose to stay home on a Friday night and find oneself pondering crazy ideas like procreation. Like generations before them, as millennials have matured, they have also matured into voting. They and their Gen Z counterparts may not make much of an impression in the overall numbers of the electorate, at least if the media narrative about the Democratic primary is to be believed. But in 2018, Gen Z and millennials increased their influence significantly in the electorates of the competitive House races. For example, while voters 18-24 made up just 2 percent of the electorate in Virginia’s 10th district in 2014, they made up 7 percent of the electorate there in 2018, while millennials increased their share from 7 percent to 12 percent. All told, the under-35 portion of the electorate in VA-10 increased from 9 percent of the 2014 electorate to 19 percent of the 2018 electorate, and those increases naturally decreased the portion of older voters in the electorate, who tend to favor Republicans.

In the Texas suburbs, voters are also increasingly of Latino origin. They are second- and even first-generation Americans who don’t see a place for themselves in the party of Trump, and present massive electoral growth opportunities for Democrats. But that takes investment, money, and perhaps most importantly, strategic vision. The Democrats’ success in Arizona demonstrates the viability of Latino mobilization. Arizona’s metamorphosis from Republican stronghold to future swing state was set in motion a decade ago. The same nativism that Trump would later ride to national prominence led the state’s Republican Party to enact the so-called “show me your papers” law and politically galvanized the state’s Latino population. 

How much Democrats recognize the viability of the Latino mobilization strategy (which has also proven its worth in California) will have a massive impact on their efforts in Texas and other contests with robust Latino populations. The party’s tangential effort to flip the Texas state house heading into the state’s critical 2020 reapportionment and redistricting cycle undoubtedly encourages DCCC investment in a wider swath of these federal House races than might otherwise land on the DCCC’s radar. Given these coordinated efforts, and what I expect to be at least some investment in the state’s U.S. Senate race, Democrats should be able to gain at least a few of these Texas districts, which means they could potentially make significant gains here. In terms of these initial ratings of “Lean Republican,” I’ll be looking to see which races the DCCC makes a serious investment in — but given these college education rates, it would behoove the party to consider them all. Better to cast a wide net in these realigning Texas districts than to sink $6 or $7 million each into trying to reclaim dealigning districts in the Northeast, as they did in 2018 only to come up short. 

     	 
    
‘Texas 8’ House Seats
 
Hover District for Details

Unless the pandemic has a massive disruptive effect on the election itself, my expectation is the 2020 congressional cycle will closely resemble Virginia’s 2019 cycle. After the 2017 blue wave in Virginia, seats were left on the table because Democrats didn’t appreciate just how deep their newfound advantages ran in America’s suburbs. As in Virginia’s 2019 cycle, Democrats benefit from Republican strategic retirements that all but hand them a couple of easy pick-ups. As in 2019 Virginia, Democrats will also benefit from a couple “gimme” seats from a court-ordered redistricting, picking up two in North Carolina (NC-2 and NC-6) after federal courts ruled the state’s map unconstitutional due to racial gerrymandering. And just as in Virginia in 2019, the Republicans are exposing a couple of very vulnerable swing Senate seats for the first time to the post-Trump electorate, putting a narrow majority in the upper chamber in contention. In Virginia, Democrats barely eked out a trifecta after underperforming in two key state Senate races. What happens in the United States Senate will come down to how much Democrats learned from those underperformances (or not) and have improved their electioneering strategies since the 2018 and 2019 cycles. 

A quick aside about the Texas House races. All “Lean Republican” races are not created equal. My model is built or “trained” in part on the Democratic Party’s historical electoral performance, which is used to help predict the Democratic Party’s future electoral performance. In some districts, the long-recognized swing districts, Democrats usually field a candidate and spend considerable resources trying to elect him or her, providing a certain competitive baseline of two-party vote share. Orange County was like that. Democrats sensed they should be winning there long before they figured out how to (or at least, before nature delivered them their long-sought victories). In other districts, ones recognized by my model as realigning, but never captured by the DCCC’s more blunt “competitiveness” indexes, the party has never made a serious bid for the seat, even if the party had a candidate. This is especially true of what I call my second-level suburban realigning districts, which contain highly educated, suburban populations, but have been GOP-held for years, if not decades. And most of the Texas Eight fall squarely into this category.

What makes the Texas Eight so hot to trot? Education, education, education. As Trump himself once proudly proclaimed, “I love the poorly educated!” And polling data shows education level is a statistically significant predictor of Trump support. This is common knowledge now, but in the run-up to Virginia’s 2017 election, when I was first arguing my theory, and even in 2018, the first time I used college education to predict where Democrats would gain big in the midterms, the idea that the 2018 wave was an education wave was fairly novel. While I was building my 2018 model, I retroactively modeled Virginia’s 2017 state legislative elections and saw that a district’s rate of college education was a good indicator of control flip. I realized I’d uncovered a powerful, new bellwether for predicting Democratic electoral strength and in the case of the Texas Eight, future potential. In my research, a good college education rate for a district is in the 40s, and many of the Texas Eight have college education rates in the 50s. Texas 23rd is the notable exception, but the Republican incumbent, Will Hurd, is strategically retiring after barely holding the seat in 2018. The 23rd is one of the most diverse districts in the state and because of the closeness of the 2018 cycle this district is more likely to receive investment, and hence, flip, so I am starting off more sure about its prospects.

If they want the Texas Eight, Democrats will have to come out against a Texas Republican Party willing to fight hard to hold onto their power. Nationally, the Republican Party recognizes that in terms of its statewide value, Texas is, pardon the shameless Mexican food reference here, the “whole enchilada.” To lose Texas would be to lose control of the presidency permanently, so I assume the national party will treat the state like the Alamo. Losing a slew of House races in the heart of the Republican Party’s economic and political powerhouse would be a significant blow. If Democrats want a piece of Texas, they better bring their lunch. 

The Democrats’ path to a Senate majority

If you’re reading this, you only have one burning question — can Democrats flip the Senate? So, I’ll cut to the chase: They have a good chance of it. But just as with the Texas Eight, the last seat they need in order to pull out a vice-presidential-tie-breaking 50/50 majority that relegates Mitch McConnell to Senate minority leader (assuming the filibuster remains intact, this is a role he will relish as much, if not even more, than his current leadership position) is going to require the Democrats to run an ace campaign strategy. 

Starting the cycle with 47 seats (which includes two independents who caucus with them) Democrats appear to need just 3 seats to win control, but as many of you already know, because one of their current seats is up for reelection in deep red Alabama, Democrats likely need to pick up four seats this cycle to offset that potential loss.

My forecast suggests three of these seats are likely flipping to Democratic control. As I discussed above and in my presidential forecast update, it is simply too early to know for sure, but I am deeply skeptical that the dual crises of the pandemic and economic crisis will affect the election in the way historical events have affected other elections, due to polarization and hyperpartisanship. Good evidence of that comes from this graph from The Economist, which compares Trump’s approval ratings through the crisis  to those of other world leaders, revealing just how much an anomaly we (Americans) are right now, as we are the only country in which citizens have been completely incapable of responding, in terms of presidential approval, to a major political stimulus. 

The hardest of these three to win will be the Maine Senate seat because the formula to win there will be different than the surge formulas that will put Democrats over the top in Colorado and Arizona. That is because America’s Northeast is one of the most politically distinct and interesting (and, oddly, overlooked) regions in the country. This is especially true of its upper neck, which is populated with the nation’s highest concentrations of independents. Keep in mind what I’ve taught you about independents. Although there are a great deal of right-leaning independents in these Northeastern states, there are more left-leaning independents, but they may not vote as reliably, and often vote third-party/independent. Case in point: The two independents in the U.S. Senate — Angus King and Bernie Sanders — are ideologically liberal and from the Northeast. 

So enters Susan Collins of Maine who, for decades now, has co-existed with Alaska’s Sen. Lisa Murkowski as one of two of America’s “Maiden Mavericks” — a female John McCain (minus the war hero background). Until recently, Collins has benefited mightily from a reputational perception as a moderate “lone ranger,” willing to buck her party when they overstep. In earlier years of her career, Collins deserved this reputation. She was reliably centrist and on certain issues, particularly women’s reproductive issues, even leaned towards being liberal (this was back when such things as liberal Republicans existed and roamed America’s Northeast with some regularity). 

But the sharp rightward lurches the Republican Party took in the 2000s under the G.W. Bush administration, and then again once Trump’s MAGA doctrines became the party’s ruling orthodoxy, have forced Collins to the ideological right. It’s the same calculus that’s been made by dozens of Republican incumbents over the past two decades (and some Democratic ones, too): Get with the program or get primaried. Now keep in mind, Collins is still a centrist, relative to the GOP Senate caucus. But that caucus is much more conservative than it was when she first entered the Senate, and to maintain a party-friendly record (and thus stave off a primary of the type her Alaska counterpart drew in 2010), Collins has to take votes that often put her to the right of the median Maine voter. Until now, Democrats have failed to dent Collins’ centrist reputation, but this cycle looks to be The One where the Democrats are finally appreciating the efficacy of nationalized messaging. 

Their success in Maine (and elsewhere) will be contingent on how closely they marry Collins to Trump. If the Democratic Senatorial Campaign Committee is smart, they’ll model their messaging on the RNC’s 2014 strategy against the Blue Dog senators, all of whom lost, to see how one structures an effective referendum campaign. 

Both Arizona and Colorado are much easier ground than Maine is, because both states offer racial diversity, while Maine is 94 percent white. Speaking of Colorado, if Virginia had a realignment cousin it would be the state of Colorado, and that is very bad news for the state’s incumbent Republican senator, Cory Gardner. I consider Cory Gardner to be the Barbara Comstock of the 2020 cycle — that is to say, he has the unfortunate distinction of being the most endangered Republican incumbent in the country (out of both chambers). At 47.1 percent college-educated, the Centennial State is one of the best educated in the entire republic, and that is a recipe for disaster for Republican incumbents in the Trump era. 

Moving on to Arizona, where a long-term political realignment has entered its “active” stage, producing its first U.S. Senate gain for Democrats in the 2018 cycle and a second likely seat from the 2020 cycle. The effort is helped with an ace-candidate recruitment in retired astronaut Mark Kelly. It’s always good to run an astronaut as your nominee, and even better to run one in a cycle in which the tailwinds are at your party’s back and said astronaut is married to a universally well-liked and highly sympathetic political star who is popular on both sides of the aisle. Usually, having won incumbency through an appointment, as Martha McSally does, detracts from the normal incumbency advantage, but given that McSally was appointed almost immediately after losing a close race for the state’s other Senate seat, she brings to her effort advantages appointed incumbents usually lack.

Although the quantitative model I use doesn’t include candidate quality, the qualitative process I use to make final race determinations does, and the quality of the Democratic nominee in this race is game-changing. That said, there are limits to what even an excellent candidate can achieve. Even a Mark Kelly cannot deliver you a Tennessee Senate seat, for example, because while Arizona is R+5, Tennessee is now R+14. The polarized electorate is highly inelastic and with very few exceptions, party competition is constrained to those states that are evenly divided between the parties. 

As an aside, many people have an eye on the Kentucky Senate race because of the perceived quality of the Democratic nominee, Amy McGrath, who made a failed bid for Kentucky’s sixth House district in 2018. At R+9, and with low college education rates, my model was unequivocal that McGrath could not win that race (I manually overrode it to predict she could win it) and the state, at R+15, is an even heavier lift. Still, the race attracts a lot of national attention, and money, because her opponent is, of course, Senate Majority Leader McConnell, Democrats’ second-least favorite Republican. Negative partisanship can take you a long way, but for this seat, McGrath needs more than the Democrats and independents, she needs actual Republicans to cross over to topple McConnell. When she announced her run, I told people, unequivocally, she could not win. Then the Kentucky gubernatorial race happened and a similarly unpopular Kentucky official, the state’s incumbent Gov. Matt Bevin, lost due to large turnout surges of Democrats and independents and a revolt by some suburban Republicans who voted for GOP candidates for the other statewide offices and the Democratic nominee on the governor’s spot. Although not common, such activity is not unheard of, but it should be noted, it is more common in state-level races than Senate races, which are more nationalized and thus, more polarized. Right now, I don’t think McGrath can win because the data doesn’t think so, and the data was right last time. BUT, if I should see evidence of crossover support in polling in that race, I will reassess. 

And so, there are three seats for Democrats: CO, AZ, and ME. But as we have already noted, the distinguished gentleman from Alabama, Doug Jones, is unlikely to retain his seat unless Republican voter support for Donald Trump, which at the time of this writing remains at about 90/10 approve-disapprove, begins to erode. The way that Jones has conducted himself in office suggests that Democrats are finally turning away from what I call the “Embarrassed Democrat” model, recognizing the futility of voting with Republicans in the hopes of attracting Republican votes. It’s a model that failed them miserably when they lost their entire slate of remaining Blue Dogs in the South in the 2014 cycle.  Though Jones has certainly maintained ideological and temperamental moderation, he clearly committed to winning or losing along a clear demarcation in terms of presidential oversight, voting for both impeachment trial witnesses and for impeachment itself. His victory in the 2017 special election had two features that offset the Republicans’ significant numbers advantage in this R+14 state but won’t be replicable in this 2020 reelection bid — an incredibly flawed opponent and non-presidential-cycle turnout — so it will take a new, unprecedented deviation to hand Jones a second, full term. Not impossible due to the introduction of the pandemic and potentially, depression-level economic damages, mind you. But not likely. 

So that means Democrats need at least one more seat to secure 50. And they have five viable options. By far the most attractive of these is the North Carolina Senate seat currently held by Thom Tillis. Democrats are perfectly capable of winning North Carolina in 2020. Barack Obama won it in 2008 (bringing with him Democratic Sen. Kay Hagan, I might add), and that was with 10 years less of demographic influx into the state’s so-called Research Triangle and realignment of college-educated voters. The combination of poor organization and negative partisanship during the Obama years benefited the GOP and decimated the NC Democrats in the 2010 and 2014 cycles. Only in the past few years has the state party begun to find its footing. That resurrection story begins with the tenure of Republican Gov. Pat McCrory, who was so divisive and underhanded he managed to lose to a Democratic challenger, Roy Cooper, in 2016, while on the same ballot as Donald Trump. 

Yes, before the GOP brought institutional degradation to the federal Congress, they took it for a test drive in the North Carolina legislature. Given the outcome, you’d think the blowback it caused there, in a conservative-friendly state, would have given Senate Leader McConnell some pause this year as he navigated President Trump’s impeachment trial over his conduct in the Ukraine scandal. As with Arizona’s immigration law, the GOP’s overreach in North Carolina activated North Carolina Democrats, and now, a few years later, the trees of that political activism are starting to bear fruit. As in Texas, Democrats have their sights on building on seat gains from the 2018 cycle, working to flip enough state legislative seats in the 2020 cycle to take control of the North Carolina state house. That down-ticket activity will be a major asset to the party’s Senate nominee, Cal Cunningham, in his effort to topple Tillis. 

That said, Obama carried North Carolina in 2008 for three reasons, and he needed all three to do it: a surge in black turnout, a surge in turnout of voters under 30, and winning over pure independents. In 2012, he got the first two, but not the third, and lost the state to Romney. As my theory suggests, turnout surges of ALL parts of the Democratic coalition, minority and young voters included, and a turnout surge of left-leaning independents as well as a preference advantage for Democrats among “pure” independents explain Democrats’ electoral dominance of Republicans in elections since Trump. North Carolina has the ingredients for Cunningham to win this race, BUT, and you’ll notice I use a literally big “but” here, as my  voter-file analysis of the 2018 House elections shows, not all parts of the Democratic coalition are surging equally. White, college-educated voters are surging the most, because these voters (especially of the female variety) are the most offended by Trump. Therefore, even though the Democrats failed to structure their 2018 messaging around a referendum on Trump and tap into the referendum effect the way the GOP did in their 2010 and 2014 cycles, college-educated whites surged anyway, which is why the Democratic primary produced what appeared to be a suburban wave for Biden. Latinos also seem to have a fairly robust, natural “Trump Effect” surge ( I wonder why), but, as I discussed in my 2020 presidential forecast update, due to institutionalized racism, black voters are less sensitive to the natural Trump Effect surge and therefore, likely need two things to really drive turnout in 2020: descriptive representation and an articulated referendum messaging effort from Democratic campaigns to heighten the referendum effect. 

Cunningham’s fortunes in NC, as well as those of many of the Democratic candidates in the down-ticket state legislative races, will hang on the turnout rates of black voters, so if Democrats are serious about winning there, they should be working feverishly to adopt Stacey Abrams’ “playbook” from her failed-but-close 2018 bid for Georgia governor because that effort, in North Carolina, would be a game-changer. Indeed, failure to understand this, and adapt their strategies to account for it, cost Virginia Democrats those two additional state Senate seats and likely, a couple additional seats in the state’s House of Delegates. Although fortunately, for them, they had room for error as they sought control of the General Assembly. 

Of course, the Cunningham campaign might get a major assist from the Biden campaign if the Biden campaign selects a running mate of color and/or one that galvanizes younger voters. We already know Biden’s running mate will be a woman, but a female woman of color, say, Kamala Harris, Stacey Abrams, or Val Demmings, would be a game-changer for this Senate race. 

It would also be a game-changer in the Democrats’ effort to flip one, or perhaps both, of Georgia’s Senate seats. There is no doubt about it, these seats are a much heavier lift than the seat in North Carolina is, but the viability of all three will be increased dramatically if Biden selects a running mate of color because I expect it will increase turnout of black voters. There is precious little quantitative research out there on descriptive representation and turnout, but we do know that turnout of black voters receded in 2016 after hitting historical highs in the 2008 and 2012 cycles. We also know that in many critical swing states, that recession was more than the margins Clinton lost by. 

And please do not conflate the strong preference black voters have for Democrats, Biden included, with their turnout rates. In 2016, black voters strongly aligned with Hillary Clinton, but their turnout rates declined relative to 2012 and 2008 because enthusiasm (and negative partisanship) was lower. The same is true of Bernie Sanders and young voters. Sanders was widely supported by young voters in both the 2016 and 2020 cycles, but when I would point to the fact that in 2020, young voter turnout did not increase for Sen. Sanders as evidence that he (and the revolution) was not particularly galvanizing among his supporters — whose turnout did not increase in 2016 over 2008 — well, let’s just say I would get dismissed. Let me repeat it here, clearly, so everyone can hear it. If Democrats win the White House and flip the Senate in 2020, it will be because many more young voters and voters of color turn out in 2020 than did in 2016, as was the case in 2018. This surge will be predominantly generated by college-educated voters, and not only white ones because the demographics of America’s suburbs today are profoundly different than they were 30 years ago. Thus, if one endeavors to elect Democrats at any level of office, voter participation and turnout must be the focal point of the campaign’s effort. 

Although the Georgia Senate seats are rated Lean R, like the Texas Eight in the House, there are pathways for Democrats to win there, especially in the special election to replace retired Sen. Johnny Isakson. That contest is a general election conducted as a “jungle primary” and the appointed Republican incumbent, Kelly Loeffler, is facing a robust ideological challenge from a Tea Party-type Republican House member and a potential stock sell-off scandal, which threatens to fracture the GOP vote and potentially allow the Democrats to sneak a victory by without the need for a runoff.

If not via North Carolina or Georgia, the Democrats must look to either Kansas, Montana, or Iowa for their final seat. At first glance, these prospects seems dim, and these are certainly not easy pick-ups because all three are predominately white states and lack large metro areas where key aspects of Democratic Party surges may be generated. But like all states, they do have smaller metros and with them, realigning suburbs that are becoming increasingly hostile to the Republican Party. Kansas is only conditionally competitive (that is to say, only competitive if the radioactive Kris Kobach is the nominee) which seems increasingly possible. Should Kobach win the Republican nomination, Kansas will become a very strong prospect for Democrats. As for Montana, as popular as Democratic Governor Steve Bullock is in the state, he is challenging a well-entrenched incumbent in Steve Daines. The pandemic and economic collapse may be determinative here.

I’ll confess, I was more bullish on Iowa until the Democrats’ Iowa caucuses occurred. And while most people focus on the debacle of the reported results, my main takeaway from the caucuses was the massive underperformance in turnout, given expectations. Now, with additional contests under our belts, we know that turnout in the Democratic primary finally did the thing everyone expected it to do, which was to look more like 2008’s than like 2016’s. But not in Iowa. As I explained in my presidential forecast update, Iowa’s big jump from D to R at the presidential level in 2016 hides a massive amount of third-party balloting, which implies more party competition than Trump’s 2016 margin suggests at face value. And Joni Ernst (who rose to national prominence via this ad, one that I show as a top-five-of-all-time ads in my Campaigns and Elections course), in addition to being an excellent candidate who fits her state like a glove (see Jon Tester and Joe Manchin for other A+ “glove fitters”) and oozes charisma, benefited from a wave election in her initial 2014 bid. Although I have less use for polling than other analysts, Iowa is one state in which polling is going to be useful. 

In terms of overall forecast, Democrats have at least a dozen very attractive prospects in the House to add to their already robust House majority and they should be able to win some of these seats under any electoral strategy. But one that focuses resources on Democratic coalition voter turnout and exploits the referendum effect has potential to increase those prospects significantly. In terms of the Senate, Democrats are in superior positions in three of the four swing races they need to win a 50-vote majority and have six prospects from which to glean their fourth seat. That makes for a lot of points of weakness for Republicans to defend in an environment that grows more politically hostile by the day. 

2020 Senate Forecast

2020 House Forecast