Democrats must flip states like Iowa to win the Senate

They will have to buck partisan trends to defeat Republicans on red-leaning turf

🍺 What’s on tap 🚰

Today’s newsletter features:

Opening Bell: Must-read items about elections and politics, with a note about our polling memo and a fun Substack Live conversation.

Political Analytics Conference: Decision Desk HQ is a co-sponsor of Political Analytics Conference 2025 at Harvard University, which takes place on Sept. 26.

The Frontrunner: Democrats can only win the Senate by flipping seats like Iowa.

Parsing Predictions: Could Zohran Mamdani win a majority in New York City’s mayoral election?

Around the Corner: Upcoming elections we’re tracking at DDHQ.

---

🔔 Opening Bell 🐏

Must-read items about elections and politics.

• This week, we plan to send out the Decision Desk HQ polling memo to Substack subscribers! We meant to begin last week, but better late than never. The weekly memo will be full of key political and electoral polling trends based on DDHQ’s polling averages. You can subscribe to receive the memo in your inbox by clicking on your account settings and turning on the option to receive the Poll Memo (see image below).

  

  Click on the toggle on the right to subscribe to our polling memo!

• I chatted with Chris Cillizza on Substack Live last Friday. Chris runs So What on Substack and was kind enough to invite me on! Our conversation delved into all sorts of things, including redistricting, the 2026 midterm picture, and presidential approval. You can watch or listen here!

• A judge ruled that Utah’s legislature must immediately redraw the state’s congressional district map. The ruling rejected an attempt by Republican lawmakers to delay the implementation of the same judge’s ruling late last month that found the legislature had acted unconstitutionally when it ignored the state’s independent redistricting commission while drawing lines.

• Amid rampant rumors that he might drop out and take a job in the Trump administration, New York City Mayor Eric Adams said last week that he would continue his independent reelection campaign. Calling rival Andrew Cuomo “a snake and a liar,” Adams proclaimed that he is the “the only one” in the field who can defeat Democratic nominee Zohran Mamdani.

---

Political Analytics 2025 is a one-day conference at Harvard University featuring top minds from media, politics, and academics. We are starting an exciting conversation about the growing role of data and analytics in determining the winners and losers in politics. This event will highlight and showcase how emerging forms of analytics are changing the way we understand politics in the 21st century.

The gathering will feature many well-known figures, including NBC News’s Steve Kornacki, The Cook Political Report’s Dave Wasserman, and CNN’s Harry Enten. You can find out more about the conference and buy tickets to attend here.

Personal plug: I will be on the Election Analyst Panel with Dave Wasserman and Jacob Rubashkin of Inside Elections, which is going to be a ton of fun. I’m also moderating a panel on post- and future-mortems of the 2024 election that will include David Shor of Blue Rose Research.

---

📊 The Frontrunner 📈

Democrats can only win the Senate by flipping seats like Iowa

Last week, Republican Sen. Joni Ernst of Iowa announced that she would not seek reelection in 2026. At first blush, this bolsters Democrats’ chances of winning the seat because incumbents may still enjoy a small electoral advantage over non-incumbents in federal elections. While Democrats face a tough midterm Senate map with few clear GOP-held seats to target, Iowa seemingly could give Democrats an opportunity since the state has been highly competitive in recent memory. That matters because Democrats need a net gain of four seats to overcome the Republicans’ 53-47 majority in the U.S. Senate (three would be insufficient because of Vice President JD Vance’s tie-breaking vote).

Yet Democrats will have to buck recent history to win a Senate race in a state like Iowa, much less claim a majority in the Senate. Iowa leaned almost 12 percentage points to the right of the country in the 2024 presidential election, as now-President Donald Trump carried it by 13.2 points while winning the national popular vote by 1.5 points over then-Vice President Kamala Harris. And in recent midterms, the president’s party has successfully defended almost every seat that leaned 10 points or more toward it.

Iowa exemplifies an important wrinkle in midterm math. Although the president’s party tends to lose ground in midterms, this is more true for elections to the U.S. House of Representatives, where the broader political situation influences the outcome in all 435 seats. By contrast, the Senate is affected not just by the electoral environment, but also by which seats happen to be up. Senators are elected to six-year terms and are divided into three groups (called classes), such that only about one-third of the Senate is up every two years. As a result, the electoral landscape and the makeup of the seats on the ballot both help determine how a party performs. Tellingly, while the president’s party lost House seats in 18 of 20 midterms from 1946 to 2022, it lost Senate seats in only 13 of 20.¹

Consider President Donald Trump’s first midterm in 2018, when Republicans made a net gain of two Senate seats while losing 40 House seats. Why the contradictory performances? The GOP only had to defend nine Senate seats, just one of which was in a state that leaned to the left of the country in 2016 (Nevada, with a lean of less than D+1). Conversely, Democrats had to defend 26 Senate seats, 12 in states that had leaned to the right. Democrats flipped two seats — Nevada and Arizona (a seat with about an R+6 lean in 2016) — but Republicans captured four — three with a clear red lean (Indiana, Missouri, and North Dakota all were R+20 or more) and one with a slight red tint (Florida, R+3).

The Senate map for Trump’s second midterm differs in many respects from 2018, but once again the GOP holds very few seats on unfavorable turf. Republicans must defend 22 seats in 2026, far more than eight years ago, but only one that leaned to the left in 2024 (Maine). Overall, the GOP is defending 19 seats that voted at least 10 points to the right of the nation, easily the most favorable collection of seats that a president’s party has had to defend in midterms dating back to 1994.

Over the past three decades, despite usually enjoying a midterm boost, the opposition party has rarely flipped Senate seats that leaned 10 or more points toward the president’s party. In 41 races in which the president’s party defended seats that leaned 10 to 20 points in its direction, only twice did the opposition win. In 2006, Democrats flipped Montana, which had voted 18 points to the right of the country in 2004, while in 2010, Republicans flipped Illinois, which had leaned about 18 points to the left in 2008. During this time, the president’s party successfully defended every seat it held in a state that leaned 20 or more points in its favor.

Conversely, the White House party has struggled to retain seats in states that leaned against it in the last presidential race. From 1994 to 2022, the president’s party lost 61% of its seats in states that leaned 5 or more points toward the opposition, while it lost slightly less than half of its seats (44%) in states that leaned between 0 and 5 points toward the other party. Helped by a typical midterm swing in its direction, the opposition party has also picked off some seats that leaned somewhat toward the president’s party (e.g. Democrats capturing Arizona in 2018). From 1994 to 2022, the White House party lost 29% of the seats it defended in states that leaned 0 to 5 points in its direction, and 22% of those in states that leaned 5 to 10 points toward it.

Now, what’s past is absolutely not prologue, but the fact that the president’s party usually defends seats that clearly lean toward it gives the GOP a clear edge in 2026. Maine is the only Republican-held seat that leaned toward Democrats in 2024; at D+8, its lean is in the category where the opposition party has found the most success. But the GOP only has to defend two other seats, in North Carolina and Ohio, that leaned fewer than 10 points to the right of the country.

And even if Democrats could win all three of those seats, that would still leave them short of the four they need for a majority. In other words, Democrats have to flip those three and at least one seat in a state like Iowa — all while not losing any of the seats it’s defending. This will be a tall order, one that may require especially favorable electoral conditions. That would involve a big backlash against the White House party, signaled by polling that shows low presidential job approval, dissatisfaction with the economy, and frustration with the state of the country as a whole. At the moment, though, no major race ratings outlet gives Democrats an edge in any 2026 Senate contest in a Republican-held seat that was not more than R+20 in 2024. Meanwhile, the Kalshi prediction market only makes Democrats the favorites in one such race — North Carolina.

We can say that Maine and North Carolina are must-wins for Democrats. Thanks in part to her moderate reputation, Republican Sen. Susan Collins has repeatedly won in light-blue Maine. However, she has never had to run in a midterm environment with a Republican president whose approval rating sits below 50%, which will likely be Trump’s situation come November 2026.² In North Carolina, where Republican Sen. Thom Tillis is retiring, Democrats probably have their strongest possible standard-bearer in former Gov. Roy Cooper. Still, he is far from a shoo-in to defeat the likely GOP nominee, former Republican National Committee chairman Michael Whatley, who has Trump’s endorsement.

Turning to more clearly red-leaning states with 2026 Senate contests, Ohio is the only other state that was fewer than 10 points to the right of the country in 2024. This special election is on the ballot because Vance resigned to become Vice President, creating a vacancy that Republican Mike DeWine filled by appointing Sen. Jon Husted. The winner here will serve out the seat’s current term and then face another election in 2028 for the seat’s next six-year term.

Yet with an R+9.7 lean, Ohio barely fell shy of the 10 points or more lean category that has proved so hard for an opposition party to flip in a midterm. Nonetheless, Democrats are hoping that former Sen. Sherrod Brown can mount a serious challenge to Husted, who has Trump’s endorsement. In 2024, Brown lost reelection by fewer than 4 points to his GOP opponent, now-Sen. Bernie Moreno. But Brown won about 3% more of the vote than Harris garnered in Ohio (46.5% vs. 43.7%), a sign that he could outperform the average Democratic contender. Husted still defaults to favorite status in a state this red, but not because he’s technically an incumbent — seats with appointed incumbents have historically functioned more like open-seat races, which tend to be more vulnerable to capture by the opposition party.

Iowa is the Democrats’ next-best target because, with Ernst’s retirement, it is an out-and-out open seat. However, at nearly R+12, Iowa’s red lean is formidable. Republicans are rallying around Rep. Ashley Hinson, who has Trump’s backing. Democrats have a number of contenders, though no one of Hinson’s stature. But the Democrats’ strong showing in special elections for the state legislature in 2025 could signal a friendlier environment for Democrats next year. In four special elections, Democrats have outperformed their 2024 presidential margin by an average of 24 points (!), per data from The Downballot, including two wins in GOP-held state Senate seats that Trump carried with at least 55% of the vote.

Texas could eventually become an open-seat race, too. Republican Sen. John Cornyn is seeking reelection, but he could lose renomination to state Attorney General Ken Paxton. In Decision Desk HQ’s polling average, Paxton leads Cornyn by about 12 points, 47% to 35%, putting Cornyn in a pretty terrible position for an incumbent. But limited general election polling suggests that Cornyn might perform better than Paxton in November. Paxton’s controversial profile, including his legal troubles and his wife’s very public divorce filing earlier this year, concerns Republicans who would like to avoid a close race. The Democrats’ own candidate field looks set to expand further this week, as much-ballyhooed state Rep. James Talarico is expected to announce a bid on Tuesday.

A trio of other seats merit a quick mention. In Alaska, Republican Sen. Dan Sullivan is certainly favored to win reelection, but Democrats could have a high-profile candidate in former Rep. Mary Peltola, who has not indicated if she will run for governor or Senate. In Florida, appointed GOP Sen. Ashley Moody does not yet have a notable Democratic opponent, but the Republicans’ potential mid-decade redistricting effort might draw a Democrat or two out of their House seats and prompt them to consider a Senate bid instead. Lastly, independent Dan Osborn ran a competitive Senate race in Nebraska in 2024, and now he’s back to run against Republican Sen. Pete Ricketts. The incumbent starts as a clear favorite, but Osborn — who has said he would not caucus with either party — could make things interesting again.

***

Republicans have the upper hand in the fight to control the Senate in 2026. Tellingly, prediction markets give them about a 7 in 10 chance of holding onto power. A key factor bolstering the GOP’s position is that the party has few vulnerable seats to defend, which leaves Democrats with a difficult path to gain the four seats they need to win a majority. That’s not to even mention that Republicans can target Democratic-held seats in Georgia and Michigan, two states that Trump narrowly carried in 2024. Democrats must hold onto those seats while also winning in some states that, like newly open Iowa, lean at least 10 points to the right of the country as a whole — a task the opposition party has struggled to accomplish in recent midterms.

✅ Parsing Predictions 🎯

Sizing up a fascinating political prediction market.

As of Sunday evening, Polymarket gave Democrat Zohran Mamdani a 57% chance of winning a majority of the vote in New York City’s mayoral election. This is actually down slightly from the end of last week, when his chances briefly topped 60% when it looked like Mayor Eric Adams might drop out (Adams then announced he would continue his independent bid.)

Mamdani could win a majority, though the crowded mayoral field works against the possibility. We haven’t had any new polling in a while, but Mamdani is hovering around 40% in DDHQ’s average, while independent Andrew Cuomo is at about 25%, Republican Curtis Sliwa is at 15%, and Adams attracts a tad over 10%. Dating back to 1905 mayoral election (the first for a four-year term), New York City has had 14 races for mayor that had three or more candidates win at least 10% of the vote. In nine of those, the winner failed to achieve a majority.

Now, the field could further consolidate because of withdrawals or because voters become more strategic with their votes by deciding that their preferred also-ran cannot win. Those developments would make a majority winner more likely. That would most likely be Mamdani, although limited polling suggests that Mamdani would not be certain to win a head-to-head race against Cuomo.

📆 Around the Corner 📌

• September 9, 2025

  • VA-11 Special Election

  September 23, 2025

  • AZ-07 Special Election

  October 7, 2025

  • TN-07 Special Election Primary

1. This includes two midterms (1982 and 1998) when the president’s party had no net change in Senate seats.

2. Collins was first elected in 1996. In 2002, she ran for reelection during President George W. Bush’s first midterm, but Bush had an approval rating above 60% in November 2002 amid the aftermath of the Sept. 11 attacks and the lead-up to the Iraq War. Collins’s other midterm campaign came in 2014 while President Barack Obama was in office.