Presidents, parties and my musings on the latest Marquette Poll
Partisanship rules everything around me
The latest Marquette poll revealed what we all sort of knew and no one really wants to come to terms with: Wisconsin, likely to be a decisive state if not the decisive state in November, is basically impossible to read at this point. (Note: I’m not involved with the production of the poll in any way.) It’s much too early for national polls to be predictive, and state-level polling comes with its own set of caveats. Trump and Biden are essentially tied, well within the poll’s 4.8% margin of error. That’s obviously the topline, but the horserace is never the whole story. What’s striking is how predictable everything is, even if we can’t actually predict the most important outcomes.
Last spring, I warned against the “politics of revelation” and suggested that, in fact, new information could make a difference in American politics. A year later, the evidence for this malleability is still mixed: partisanship drives much of our politics, and it’s not clear how sensitive public attitudes are to events and new developments. This most recent round of polling from Wisconsin residents highlights a couple of things I found interesting and possibly telling.
Issue ownership
The poll asked respondents whether they trust Biden or Trump (or both equally, or neither) to do a better job on various issues. What jumped out at me about these numbers is that you could take out the specific 2024 context and present these as issue ownership data from the 1980s or 1990s and probably no one would notice. (The Wisconsin data seem consistent with national opinion on these issues.)
Issue ownership refers to the idea that voters associate some issues more strongly with Republicans and others with Democrats. In 2024, as in the past, Democrats have the edge on social welfare issues, while Republicans are more trusted with the economy and foreign policy. Scholars have different ideas about why the two parties continue to “own” certain issues – where these connections come from and why they persist among voters – but what’s striking here is that they don’t seem terribly responsive to actual candidate qualities and characteristics. Trump enjoys the Republican advantage on foreign policy issues despite the national security establishment being pretty firmly opposed to him, he gets an advantage on immigration even though some of his main policies were very unpopular. He’s viewed as better on the economy even though the major economic policy of his presidency, the 2017 tax cut bill, was unpopular (this may be more understandable since the economy was strong prior to the pandemic). Regardless of Trump’s record, experience, or public statements, he seems to be enjoying the party advantage on its usual issues.
The one different from decades past is that polls showed Republicans having the edge on “moral values,” fueling a long-standing assumption that social issues – gender, LGBTQ rights, etc. – were electoral winners for Republicans. And while the pro-choice position has been more popular throughout this period, the last two years have produced – or revealed – a powerful new electoral politics of abortion rights.
Do candidates matter?
In general, my instinct is to be pretty close to the “no” position on this question. The 2016 election is among the most powerful examples: in an extraordinary year, with unique candidates, the election outcome looked pretty much exactly as the “fundamentals” would predict. But another point of data from the Marquette poll (and Wisconsin politics more generally) complicates and deepens this perspective. While the presidential candidates are running evenly, Senate incumbent Tammy Baldwin has a lead outside the margin of error against her newly minted opponent, businessman Eric Hovde (though this changes somewhat depending on measurement and which respondents you look at). In 2012, Baldwin looked like a very long shot for U.S. Senator, a Madison liberal running against beloved former governor Tommy Thompson. But Baldwin won the race, becoming the first openly LGBT person elected to the U.S. Senate.
Six years later, in 2018, she won reelection easily in a race that was called just moments after the polls closed – in stark contrast to the governor’s race that year, which came down to about 30,000 votes. Some of this could be incumbency effects or a weak 2018 opponent, and anything could happen in November. Nevertheless, the evidence suggests Baldwin has been pretty successful at building a personal political brand that’s linked to populist appeals around manufacturing and trade – issues important to some Wisconsin constituencies but not front and center in national partisan politics.
When we think about the question of whether candidates matter in terms of the nationalization of politics, it’s not surprising that congressional candidates have an easier time separating from local politics and cultivating their own reputation. Presidential candidates, on the other hand, have become nearly synonymous with national party brands. While intuitive, this is striking to anyone who followed presidential politics at the end of the twentieth century: it was presidents – like Ronald Reagan and Bill Clinton – who were perceived to transcend partisan strictures with their personal charisma. But we may be decisively in the era of the partisan presidency now.
Framing the 2024 election
The issue ownership results from this week’s poll also lend some support to one of my grand theories of the 2024 election. I’ve been predicting that this will be a contest between efforts by the Trump campaign to frame the contest retrospectively – to highlight what critics see as shortcomings in Biden’s handling of the economy and the “border crisis” in particular. For the Biden campaign, the challenge will be to frame the election prospectively, in terms of Trump’s unpopular stance on salient issues like abortion. This entails highlighting, as a recent ad does, that “Donald Trump did this,” but also blasting the implicit message that, if elected, he will do even more.
There’s been some recent work to suggest that persuasion in campaigns is still possible. But presidential campaigns are the hardest for campaigns to have an effect – voters know a lot about the candidates, and this has perhaps never been truer than in 2024. Furthermore, the available evidence suggests that this is about more than just the electorate’s familiarity with the candidates. Nationalized party politics have made presidential politics potentially even more leaden and difficult to move. The twist is that because of uncertainty in states like Wisconsin, and because of a system that amplifies their importance, it makes the outcome of the November election that much harder to predict.
By suggesting that the candidates don't matter (much), are you predicting that the Trump wing of the GOP will continue to dominate after Trump is no longer a potential candidate (through death disqualification, or--god forbid) serving two terms)?