Are We Headed to a Ticket-Splitting Renaissance...or Are the Polls Just Wrong?

Rep-Elect Michael Rulli (R-OH)

This week, we saw yet another special election.  In Ohio's 6th congressional district, stretching must of southeastern Ohio voters went to elect a candidate to succeed Rep. Bill Johnson, who retired to become Youngstown State University's president.  Johnson, a Republican, will be succeeded by a Republican-State Sen. Martin Rulli, a formidable candidate who defeated a former congressman (John Boccieri) to win his state senate seat.  But Rulli did not get this victory easily, or at least he did not get it impressively.  Rulli won the seat by 9-points, in a district that Donald Trump won by 46-points.

That sort of insane gap is something we've seen a lot of in the past few years.  Since the Dobbs Supreme Court decision, the Democratic Party has consistently over-performed in special elections.  Congressional races, state legislative, even judicial contests...the Democrats continue to over-perform the margins.  This has been chalked up to college-educated voters flocking to the Democrats, but OH-6 is one of the least educated districts in the country, and yet a candidate with no money or name recognition got over 30-points advantage on the GOP this week.

This happened as Joe Biden's polling position continues to sour.  His approval ratings are now in the high 30's, basically a death knell for a campaign.  His position in states like Maine, Virginia, & Minnesota in polls frequently come back with tied results even though he won these states handily four years ago, and swing states like Georgia, Arizona, North Carolina, & Nevada are a nightmare for him in polling.  These polls, though, concurrently show Democrats in a uniquely strong position down-ballot.  Democratic senators like Bob Casey, Tammy Baldwin, & Jacky Rosen look fine for reelection; Sherrod Brown in Ohio leads in most polling even as Biden loses by nearly 10-points in the state.  This is a conundrum, because it flies in the face of conventional wisdom.

Since the 2000 election, we have consistently seen a decline in ticket-splitting, even for districts that historically were kind to down-ballot parties that couldn't win presidential elections.  Republicans cleaned up in 2004, 2010, 2014, & 2018 in the Senate almost completely because of beating moderate Democrats in red/pink states/districts.  Conversely, Democrats course-corrected moderate Republicans in places like Oregon, Rhode Island, & Minnesota in Bush's second term.  Currently, there are only five US Senators who represent states that their party didn't take in a presidential election, and only one of them (Susan Collins of Maine) was able to do so in a recent presidential election.

This presents a conundrum-do you believe polling, or do you believe elections?  Polling would have you believe that Joe Biden is a uniquely unpopular Democrat, and that voters are willing to throw him out in favor of Donald Trump (a man they tossed in favor of Biden four years ago)...even as they will be willing to elect Democrats to stop Trump down-ballot.  This is technically borne out in the Democratic over-performance in special elections the past two years...but only because we don't have any presidential election results to back it up.  

The reality is that a complete reversal in ticket-splitting on this magnitude is not possible in one full sweep.  It is far more likely that (rather than the polling scenario being accurate) one of two things is actually true.  First, Trump is leading, and the Republicans are in a stronger position than we think.  The Trump strength at the top of the ticket will start to leak down as voters get to know the Republican candidates, and we'll start to see people like Sam Brown, Bernie Moreno, & Mike Rogers gain leads in states where Trump is ahead in the polls, and he'll get his victory.  There might be an outlier or two (Kari Lake feels uniquely unpopular in Arizona, enough so that Ruben Gallego can make up the gap), but by-and-large Trump's support is solid, and the Democrats' down-ballot are standing on sand.  This is totally plausible, and if you believe this theory, you do so with sound logic as it's happened before.

But the special election consistency has me starting to wonder if the opposite is true-what if Americans don't like Joe Biden and clearly wish that they got to vote for someone like Gretchen Whitmer or Raphael Warnock...but they still vote for him anyway?  The polling scenario would take a while to catch this.  It's early (most people aren't paying attention)...anecdotally, I spoke to someone just last week who still hoped that one or both of the candidates would drop out (I suspect this is driven because the media seems obsessed with hoping one of the candidates drops out)-this could be driving people to be "undecided" who actually aren't, or who are mad they're voting for Biden, but would vote for him anyway.  This isn't a good way to win an election, but it happens (Nevada's 2010 Senate race comes to mind), and it has to be pointed out that virtually every other metric favors Biden.  The economy is good, he had a great midterms, his party dominates in special elections, his party is dominating in fundraising, he has a better GOTV model than Trump (Trump's legal problems, and mismanagement at the RNC has meant that he's months behind Biden in opening field offices in key states).  He has some trouble on his left flank due to Gaza, but I'll be honest-it's hard to tell how much of that is just online noise from voters/bots that won't actually vote for him.  He also polls unusually well with senior citizens, the most important voting bloc in terms of consistency other than college-educated or suburban voters...which he is winning handily.  Special elections like the one recently in NY-3 do not show any deterioration amongst voters of color for Democratic support (neither did the midterms) that is largely a polling crosstabs phenomenon to date.  In fact, EVERYTHING other than polling is in Biden's favor right now, and even that it's ONLY Biden that is in a tough position-downballot Democrats have every reason to assume they will do well in November if you just look at polls.  Hell, this past week we saw a poll where Democrats were in striking distance of taking the Florida Senate seat, which would open up an entirely new avenue to win the Senate if they pulled it off.

We will not know whether or not this polling is right or not until November, and it presents a clear problem for Biden as long as it's an issue.  I've said this many times on this blog, but the maxim "run like your ten points behind" is terrible advice.  No one wants to vote for a loser, and if Biden looks like a loser, he will lose a lot of momentum this fall-he has to at least make this look like a race that is close, and it needs to look like a race that is close enough that the media will bill it as such.  He needs to find a way to lead at least some polls in places like Georgia, Arizona, Nevada, North Carolina, or Florida, enough to keep passion in these states alive amongst his base.  But at some point, after seeing the results in Ohio and countless other special elections, you have to beg the obvious question-why is it that Republicans only win in the polls, and never when it comes to actually voting?  Is it because the polls (gasp!) are currently too tilted to the conservatives?  I think that's a far likelier explanation than us getting a sudden about-face when it comes to trends against ticket-splitting.