Walz to Victory?

What a great time to start a blog about Minnesota politics! Not only is the media doing its quadrennial “maybe Minnesota is actually a swing state this time” speculation but our very own Tim Walz is one of the final four on Kamala Harris’s VP shortlist.

Minnesotans are probably too used to losing once a contest gets down to four to get their hopes up, but what if Walz actually wins this one? I’ll go over the case for Walz and in doing so discuss what political scientists know about the effects of VP selections. I’m drawing heavily on Christopher Devine and Kyle Kopko’s research on the effects of Vice-Presidential candidates (see their take on the VP-stakes here).1

Briefly, Democrats shouldn’t count on Walz to win the Midwest or “white dudes” who are not already for Harris, and while he’d likely help her in Minnesota, that’s of negligible value when if comes to winning the Electoral College.

Where Walz distinguishes himself from the other contenders is his relatively long electoral track record, which reduces the chance that voters end up seeing him as unprepared for the office and the chance that some unexpected weakness turns voters against him. These risks matter less because voters will vote against Walz, and more because they might lead voters to question her judgement.

As you might expect from a Minnesotan, Walz is the safe, reliable pick when it comes to electoral considerations - the hotdish among Harris’ possible VP candidates. Fundamentally, though, voters will reward Harris if she picks someone who they end up liking and thinking would make a good president. If Harris likes Walz and thinks he would make a good president, she should pick him.

Probably, but it wouldn’t matter.

There are very few worlds where Harris loses Minnesota but wins Wisconsin. Beyond Wisconsin, Minnesota is also highly correlated with the other “Blue Wall” states of Michigan and Pennsylvania. If Harris is on track to lose in Minnesota, a lot of other things will have already gone wrong.2

In more general terms, do presidential candidates win more votes in their running mate’s state – the so-called “home state” advantage? The answer is a resounding… maybe, but like many things in political science, both the effect and our datasets are small, making it hard for us to say for sure. The most definitive political science research on the subject concludes that “home state” effect of a VP pick is statistically indistinguishable from zero. Other work suggests that parties win one or two percent more of the vote in the VP candidate’s home state. It is, at any rate, small, but potentially decisive in a very close election.3

How potentially decisive? Let’s peg the home-state effect at one percent. If Harris picks a VP candidate from the state most likely to decide the election … why, hello Josh Shapiro, governor of Pennsylvania … Nate Silver estimates she’d increase her chances of winning the election by about two percent. Not huge! But not nothing.

An extra one percent in Minnesota is, however, nothing. The world where that one percent puts Harris over the top in Minnesota is a world where she’s already lost by the time Minnesota’s polls close.

Here, again, the political science research suggests probably not, and if so, only by a tiny bit. Devine and Kopko find basically no evidence of a home-region effect for presidential candidates, much less for a VP candidates. From a political science perspective, the idea that Walz would help deliver Wisconsin, much less Michigan is… interesting, in the Minnesotan sense.

This makes sense if we think about the factors – media coverage, state pride, having already voted for the guy – that might produce a home-state or home-region advantage. Walz has two decades of Minnesotans accustomed to pulling the lever next to his name, good name recognition from frequent media coverage, Minnesotan’s love for Minnesota … we think all of that will get him one measly percentage point in Minnesota. No one in Wisconsin or Michigan (much less Pennsylvania) has ever voted for Walz, nor has he had years of (mostly positive) press coverage in those states.4

We can see more evidence for the lack of regional effects in the horse-race polling from that brief period when it looked like Biden might be out, but Harris might not automatically be in.5 In mid-July, Gretchen Whitmer was doing averaging 7 points better than Harris in Michigan, but basically the same in Pennsylvania/Wisconsin/Minnesota. Josh Shapiro was averaging nine points better than Harris in Pennsylvania but, again, basically the same in the other three states. In a move no doubt calculated to feed Minnesotans’ inferiority complex, no one polled Walz as a Biden replacement, but I assume the pattern would be similar. Whatever bonus a candidate receives in their home state, it stops at that state’s borders.

Walz’s recent bout of internet fame was mostly fueled by him calling Donald Trump and J.D. Vance “weird,” but his widely covered speech on last week’s massive “White Dudes for Harris” Zoom call helped a lot too. Could Walz deliver the “white dude” vote?

This question is basically moot because the other VP short-listers are also (not coincidentally) white dudes. But even if they weren’t, there’s little evidence that picking a white dude would help win the votes of other white dudes. Devine and Kopko find that the two previous female vice-presidential picks had no effect how women voted. They also look at religion and find no effect of evangelical Christian picks (Palin, Pence) on evangelical Christian voting, the lone Jewish VP pick (Lieberman) on Jewish voting or, post 1960s, any effect of Catholic VP picks (numerous) on Catholic voting. In short, to the extent we can say anything about the effects of VP demographics, it is that they don’t matter.

This is the kind of conclusion we should treat with caution, because there’s very little historical variation to learn from; almost all VP picks have been white dudes. Harris herself is, in terms of race and gender, unprecedented – and the only other non-white and female nominees (Obama and Clinton) both picked white dudes. Would Obama and Clinton have done worse if they hadn’t picked white dudes to give demographic balance to their tickets? Its hard to say, but the evidence we have says no.

All of the preceding discussed what Devine and Kopko call “targeted effects.” With these, we can (try to) draw a direct line between some characteristic of the VP pick (Minnesotan, white dude), and how that group of people vote. But what about the more general effects of Walz on how voters in general vote? Here, we do think that VP picks matter – a little.

First, the Vice-Presidential candidate’s popularity has a (small) effect on how people vote, even when we take into account the popularity of the Presidential candidate. To give a specific example, Devine and Kopko’s estimates suggest that the gap in favorability between John Edwards and Dick Cheney, the 2004 VP candidates, increased the average voter’s probability of voting for the Democratic ticket by a little more than two percent. Edwards and Cheney, however, are outliers, with the biggest gap in favorability in the last 30 years.6 The average gap over those seven presidential elections corresponds to a roughly 0.7 percent increase in the probability of voting for the favored VP candidate’s ticket. Not huge. But, like the home state effect, not nothing.

So… pick a running mate who will turn out to be popular!7 The problem is that most VP candidates aren’t well known outside of their home state, so its hard to forecast how popular they’ll be once they’ve campaigned, participated in debates, been mocked on SNL, etc. Maybe Walz’s good electoral track record in Minnesota and relative popularity in the state suggest that he’d be relatively popular nationally. On the other hand, Walz is a less-popular-than-average governor, and Minnesota has no shortage of locally popular politicians who face-planted on the national stage.

The key insight from Kopko and Devine’s work is that the impact of a VP pick goes beyond what people think of the VP pick him or herself – it also influences what people think about the presidential candidate doing the picking. In other words, pick a VP who people like and people will like you more; pick one they view as a bad choice, and people will question your decision-making.

The canonical “don’t do this” example is Sarah Palin. John McCain reportedly picked Palin to reinforce his reputation as a “maverick.” Instead, voters saw Palin as a very conservative pick whose youth and relative inexperience (and possibly gender) made her unready to be president. They, in turn, viewed McCain as more conservative than they previously had and worried about judgement. Interestingly, Palin wasn’t really disliked to an unusual degree – people liked her more than Dick Cheney four years before or Paul Ryan four years later – so she probably didn’t hurt McCain’s favorability ratings. But you can like someone, even enjoy their subsequent reality TV show, and still wonder why the hell someone picked them to be second in line for the Presidency.

I think this is where Walz really shines, in that he is in a lot of ways the anti-Palin. Not young (though, at 60, not too old), experienced, and male, which probably translates into being perceived as ready for the job which, in turn, increases perceptions that Harris has good judgement.

Predicting how voters will perceive his ideology and general likability is a bit harder. In Minnesota, Walz was generally perceived as a moderate until the last legislative session, when he helped deliver a lot of progressive wins with a one-vote majority in the state senate . His 12 years of electoral success in a red-tinged congressional district suggest that he could be perceived as a moderate. That said, he is being discussed nationally as the most progressive of the possibilities (and now has the Bernie seal of approval). My guess is that he’d be able to project a moderate image on the national stage, but there’s uncertainty here.

Similarly, predicting Walz’s national favorability ratings once he jumps from Minnesota to the big leagues is hard. He is moderately popular state-wide, though less popular than the average state governor. The “weird” thing plays well with the base, but perhaps not so well with disengaged voters who think that politicians bicker too much. Still, I think his national favorability would probably end up about where it is currently in Minnesota – fine, not great.

And how does that compare to the other three white dudes on Harris’s short list, Josh Shapiro, Governor of Pennsylvania, Andy Beshear, Governor of Kentucky, and Mark Kelly, Senator from Arizona?

1. Walz would probably have a small home-state effect in Minnesota, which is pretty useless. He’s unlikely to have any particular effect on the Midwest as a whole, or with “White Dudes” who are not already for Harris. Here, the advantage obviously goes to Josh Shapiro and, to a lesser extent, Mark Kelly.

2. Walz would probably be viewed moderately favorably – this would have a small positive effect partially because people like Walz and a small positive effect because people who end up liking Walz would like Harris because she picked Walz.

   This is the point with the greatest uncertainty. Shapiro is more popular in Pennsylvania than Walz is in Minnesota, and you have the kind of vague journalistic paeans to Shapiro’s charisma that you don’t see about Walz. Which, fair enough, charisma isn’t the first word that comes to mind when I think about Walz either.

   But Walz has one big thing going for him: His long electoral track record suggests less risk that something could go wrong – say, a tendency towards gaffes, an inability to debate or talk to the media, former friends willing to publish embarrassing personal correspondence, or embarrassing op-eds he wrote in college. Contested elections, especially state-wide elections, are a great opportunity for these things to come to light, and the fact that Walz has won two suggests an absence of skeletons in his closet.

   Here, Shapiro is the riskier choice. He’s been through one gubernatorial campaign against an extremely flawed candidate and two Attorney General campaigns which, yes, are statewide but receive much less publicity. This is not to say that there is some huge hidden flaw that hasn’t yet been exposed in a previous campaign, but rather that there’s a higher probability of one than with Walz. The recent hubbub around his college op-ed is an example of this. While I don’t think this revelation is in itself going to be all that damaging, it is exactly the kind of thing that running in a hotly contested statewide campaigns (in Walz’s case, hotly contested primary and general election campaigns) is likely to uncover. Particularly given the short vetting period, Harris should probably worry that there are more as-yet-uncovered revelations waiting about Shapiro.

   Andy Beshear and Mark Kelly both have two high-profile statewide campaigns under their belts, but less experience than Walz preceding that. So, if Shapiro’s experience is closer to Palin-level than you might like, Beshear and Kelly are closer to Walz.

3. Walz would probably be perceived as ready for the job, ideologically moderate, and, again, moderately popular. Picking Walz would lead voters to view Harris as having slightly better judgement, being slightly more moderate, and generally feel slightly more positive towards here.

   All of this is mostly true for the other contenders too, however. Here, again, I think Walz’s relatively longer track-record in office puts him slightly ahead of Shapiro, and possibly also Beshear and Kelly. He is just slightly less of a wild card, in a way that makes it slightly less likely that voters will end up thinking “what was she thinking?”

In sum, Walz would be a good pick. To the extent that any of this matters, Walz would probably help – a little – for Harris’s chances of winning. On average, I’d guess Shapiro would help more. But Shapiro is closer to Palin’s level of experience and previous electoral vetting – there’s a greater chance that something comes out about Shapiro that makes them question Harris’s aptitude than there is for Walz. Kelly and Beshear are somewhere in-between.

Not much. None of the four names reportedly on Harris’s short list is a real wildcard and all are comparable in their home-state popularity, ideological moderation, and white-dude-ness (slight edge to Shaprio on the first two of those). Walz’s relatively long track-record gives him a (small) advantage – something that I think is being overlooked in a lot of the national commentary. Shapiro’s (and, to a lesser degree, Kelly’s) swing-state residence is also a (small) advantage. None of these picks is going to shake up the race, but there’s also no clear best pick among them.

All of which is to say, Harris should probably pick the one that she thinks would make the best President. Even in a narrow electoral sense, whoever she picks is likely to be the Democratic frontrunner in 2028 or 2032. Here, the candidate’s ages come into play; Beshear (46) and Shapiro (51) have an edge over Walz and Kelly (both 60).

But fundamentally, Harris needs to make a choice based on her judgement of each of these dudes’ ability to sit in the Oval Office and do a good job. The nice thing is that Harris’s private judgement about which of them is more ready to be President is probably a pretty good predictor of voters’ eventual judgement.

And, of course, picking the best potential president is also the best thing for the country (what a happy coincidence!). So, the preceding 3,000 words are mostly about tiebreakers. If Harris thinks Walz would make the best President of the bunch, she should pick Walz.