If Democrats Get a 2006-Style Senate Result, It May be Because Trump Has a 2006-Style Approval Rating

KEY POINTS FROM THIS ARTICLE

— While the 2018 midterm was generally a “blue wave,” Republicans still held the Senate. If Democrats want to flip the Senate, they’ll need a year that’s more like 2006.

— The Senate map was in some ways completely different heading into the 2006 election as opposed to the Senate map we have now, and it was less tied to presidential partisanship.

— However, there are some commonalities between the top Democratic Senate targets in 2006 and in 2026, although the comparisons require some stretching.

— But more important is President Trump’s approval rating, which probably needs to be more like George W. Bush’s in 2006 than Trump’s own approval in 2018 in order for Democrats to flip the Senate.

Comparing 2006 to 2026

In many respects, Democrats will have done well if 2026 ends up being similar to 2018. That year, Democrats flipped the House and also gained considerable ground in gubernatorial and state legislative elections.

But the 2018 cycle also featured Republicans netting a seat in the Senate.

So when it comes to the Senate, Democrats don’t want a repeat of 2018. They instead want a repeat of 2006.

Two decades ago, and in what at the start of the cycle would have seemed like quite a surprise, Democrats flipped six Senate seats, eking out a 51-49 majority just two years after President George W. Bush’s reelection had the GOP dreaming of a “permanent Republican majority.” Senate Minority Leader Chuck Schumer (D-NY), who chaired the Democratic Senatorial Campaign Committee in the 2006 cycle, has compared this year to 2006.

While readers probably have a good sense of what happened in 2018, they may have a hazier memory (or no memory at all) of 2006. So we thought we’d take a look back at the Senate playing field from two decades ago and how comparable it may be to 2026.

First of all, let’s set the table for what the Senate looked like leading into the 2006 election. Map 1 shows the partisan makeup of the Senate following the 2004 election.

Map 1: Party control of Senate by state following 2004 election

Note: Independent Jim Jeffords of Vermont is categorized as a Democrat on this map because he caucused with Democrats after leaving the Republican Party in mid-2001

One thing that immediately jumps out is that there were many more split delegations: 13, not including Vermont (which then and now had an independent who caucused with Democrats: Jim Jeffords prior to the 2006 election and Bernie Sanders after it). There are only three split delegations now: Maine, Pennsylvania, and Wisconsin (for the purposes of this discussion, we’re counting Sanders as well as independent Sen. Angus King of Maine as Democrats).

In 2004, Bush had won the national popular vote by 2.5 points, while Donald Trump won it by 1.5 points in 2024. But both won 31 of 50 states, indicating then (and now) a generic Republican lean in the Senate based on presidential performance.

Just as is the case today, Democrats were in the Senate minority despite holding more “crossover” seats, Senate seats in states that had voted for the other party in the most recent presidential election. But there were many more of them back then. Democrats held 16 seats in states that had voted for Bush, while Republicans held 9 in states that had voted for John Kerry. Today, the Democrats hold 10 seats in states that voted for Donald Trump, while Republicans hold just one in states that voted for Kamala Harris. While that is lopsided in favor of Democrats, all of the current “crossover” Democratic Senate seats come in states that voted for Joe Biden in 2020, meaning that they are all in somewhat marginal states. Democrats held many seats in redder states 20 years ago. As Michael Barone noted in his introductory essay to the 2006 Almanac of American Politics, 11 Democratic senators held seats in states that had given Kerry less than 47% of the vote. Of the current crop of Democrats, that applies only to the two Democratic senators from Arizona, and Harris’s share of the vote there (46.54%) actually rounds up to 47%.

But one can see from the map that Democrats in 2004 still held both seats in states that were or were becoming deeply Republican at the presidential level, such as Arkansas, North Dakota, and West Virginia. Democrats do not hold any such states now. Republicans, meanwhile, only held three Senate seats from states where Bush got less than 47%: both seats in Maine and one seat in Rhode Island (the latter of which they would lose in 2006).

Let’s now turn to the seats that the Democrats flipped in order to win the majority.

Only one of the Democratic flips came in what was clearly a blue state at the presidential level, as now-Sen. Sheldon Whitehouse (D-RI) defeated then-Sen. Lincoln Chafee (R-RI) by a 7-point margin. Rhode Island was (and still is, to a lesser extent) one of the bluest states in the country, but despite that, the moderate Chafee and his father, John, had held this seat for 30 years, examples of the old Yankee Republican tradition that once used to be so prominent in New England. And even after Chafee’s loss, traces remained: In 2006, Sen. Olympia Snowe (R-ME) won her final election more than 50 points, completely unaffected by the Democratic wave. On the 2026 map, the most obvious comparison for Rhode Island is Maine, where Sen. Susan Collins (R-ME) is trying to hold out as the only New England Senate Republican. One thing that complicates the comparison is that Rhode Island was bluer than Maine is now at the presidential level: Kerry had won Rhode Island by 21 points in 2004, while Harris won Maine by just 7.

Another pair of Democratic flips came in what at the time were core presidential battleground states. Pennsylvania at this time was a somewhat blue-leaning presidential state, while Ohio voted almost exactly in line with the nation. Then-Sens. Rick Santorum (R-PA) and Mike DeWine (R-OH) both ended up being swamped by now-former Sens. Bob Casey (D-PA) and Sherrod Brown (D-OH); the latter pair would win three terms apiece before losing in 2024. On the 2026 map, North Carolina—a Republican-held presidential battleground—is most comparable to Ohio and Pennsylvania in 2006. Though we doubt former Gov. Roy Cooper (D) wins by the same double-digit margins as Brown and Casey did in 2006, it is worth noting that Cooper has led every independent poll of the race so far, much like Casey did in 2005 and 2006 (and Brown consistently led starting in the summer of 2006).

The other three states that Democrats flipped are more comparable to the red state targets Democrats have in 2026 (like Alaska, Iowa, Ohio, and Texas), although there are key differences.

Democrats unseated incumbents in Missouri, Montana, and Virginia, all by close margins. Missouri, which at the time still retained some of its bellwether reputation, was moving to the right much the same way that some of the other quintessential “border states” had been moving following Bill Clinton’s presidency. Bush carried Missouri by a little more than 7 points, and it officially broke from its bellwether tradition by backing John McCain by just about a tenth of a point in 2008 despite Barack Obama’s substantial national victory. Virginia, meanwhile, was still a red state at the presidential level, although Bush won it by roughly the same margin (8 points) in both 2000 and 2004 despite doing better nationally in the latter year. So while Missouri and Virginia were in roughly the same place in 2004, they were ships passing in the night, politically, as Missouri has since become solidly red and Virginia has become Democratic-leaning. Montana backed Bush by roughly 20 points in 2004, making it redder than any of the key red-state Democratic Senate targets this year—Trump carried the Alaska-Iowa-Ohio-Texas quartet by roughly 11-14 points apiece—while Missouri and Virginia were more competitive for president in 2004 than those aforementioned states were in 2024. But Montana was much more competitive down the ballot back then, and it had a long history of electing Democrats to the Senate despite its red presidential lean.

The only close election that Republicans won in 2006 was in Tennessee, an open seat where now-former Sen. Bob Corker (R) prevailed by a little under 3 points against now-former Rep. Harold Ford Jr. (D, TN-9): Tennessee had voted for Bush by 14 points in 2004 and was also in the process of zooming right. Overall, only four of 2006’s Senate elections were decided by less than 5 points (Tennessee, as well as Missouri, Montana, and Virginia).

In 2006, Democrats needed to net six seats to win the Senate. This time, Democrats only need four, although the Democratic defensive assignments probably will be more competitive than they were in 2006. This is despite the fact that Democrats were defending Senate seats in the red presidential states of Nebraska, North Dakota, and West Virginia as well as in the competitive presidential states of Florida, Michigan, and Wisconsin, among others. But Democrats won all of those races in blowouts (something that seems very odd to a modern observer for an election that only took place 20 years ago). The closest Democratic defensive state in 2006 actually came in New Jersey, where the now-disgraced former Sen. Robert Menendez (D) won his first full term against now-Rep. Tom Kean Jr. (R, NJ-7), who is in the news this year for his strange disappearance from public view as he apparently deals with some sort of medical issue. Menendez won by 9 points. The Democrats do have one current Toss-up to defend on this year’s map, the open seat in Michigan.

But the one thing that might really tie 2006 to 2026 is this: presidential approval.

Bush was actually underwater in approval in all of the states that Democrats flipped in 2006, according to the exit polls. That includes even Montana, which, again, had backed Bush by 20 points in 2004. Bush also was underwater in Tennessee, too, but only by a 48%-50% spread, evenly-balanced enough to allow Corker to beat Ford in the one close race Republicans salvaged.

Table 1 shows the exit poll presidential approval for Bush in the states that flipped (all from Republican to Democratic) in 2006, as well as Trump’s approval in the six states that flipped (four from Democratic to Republican, and two from Republican to Democratic) in 2018.

Table 1: Presidential approval in Senate seats that changed hands, 2006 and 2018

Source: Dave Leip’s Atlas of U.S. Presidential Elections for election results, archived CNN exit polls for 2006 and 2018

This underscores the importance, for Republicans, of Trump improving his approval rating.

Ron Brownstein, one of the sharpest observers of modern politics, has routinely flagged the connection between presidential approval rating and election results, and he noted that during Trump’s first term, the only Republican senator to win in a state where Trump’s approval was underwater was Collins in Maine in 2020.

Republicans netted Senate seats in 2018 despite Trump’s approval rating being underwater nationally. However, exit polls showed that in the four states Republicans flipped from Democrats in November 2018—Florida, Indiana, Missouri, and North Dakota—Trump’s approval rating was net positive. Democrats, meanwhile, flipped Arizona and Nevada as Trump’s approval was narrowly positive in the former and narrowly negative in the latter.

So where will Trump’s approval be in the core Democratic Senate targets in November?

Even if Trump improves his approval a lot, he will be underwater in Maine in November, but Maine may also be an outlier for other reasons. Trump appears to be clearly underwater in North Carolina based on recent polling. A recent Fox News poll of Ohio showed Trump’s favorability (not his approval) at only 42% there. A couple of recent Texas polls showing a close Senate race (Quantus Insights and Texas Pulse) had Trump’s approval narrowly underwater.

There really isn’t recent nonpartisan polling showing Trump’s approval in Alaska and Iowa, although recent estimates from Lakshya Jain’s polling at the centrist liberal site The Argument suggests Trump’s approval may be underwater in those states too (along with the others mentioned in the previous paragraph where we do have recent independent polling).

Approval is not going to perfectly correlate to vote choice. The Republicans who lost in 2006 actually ran ahead of Bush’s approval in the exit poll. More recently, Democrats did well in the Senate in the 2022 midterm despite Joe Biden’s poor approval in key states. And the electorate is a lot different now compared to 2006, including in ways that are helpful to Republicans in at least some of the states at the center of the race for the Senate in 2026 (Iowa and Ohio most clearly). It’s also not hard to imagine, in this era, Trump getting back to positive approval in states he won by double digits even if he’s not there now.

But a lesson from the past two midterms under Republican presidents may simply be that Bush’s approval was bad enough to allow Democrats to win the Senate in 2006, but Trump’s wasn’t bad enough to allow the same in 2018.

Trump’s ascension to the top of the Republican Party a decade ago was in some ways a reaction to the perceived failures of the Bush administration. Trump would never want to be compared to Bush, but if Republicans lose the Senate in 2026, it would likely be in no small part because Trump fell into a Bush-like rut and couldn’t get out in time for the election.

Kyle Kondik is a Political Analyst at the Center for Politics at the University of Virginia and the Managing Editor of Sabato's Crystal Ball.

See Other Political Commentary by Kyle Kondik.

See Other Political Commentary.

Views expressed in this column are those of the author, not those of Rasmussen Reports. Comments about this content should be directed to the author or syndicate.