Consultants Like to Talk about Messaging
Political scientists, however, are skeptical.
As was noted in the comments earlier this week, a post-mortem on the Democrats’ loss in 2024 was recently released by Welcome PAC: Deciding to Win.* The funny thing is that while people like James Carville (i.e., consultants) love this kind of thing, political science types aren’t as impressed.
Here are three examples (noting that Masket is the Director Center on American Politics at the University of Denver and is an expert on American parties, Bernstein holds a PhD in Political Science and is a former polisci professor turned analyst, and Morris is a well-known data analyst).
- Seth Masket: That Democratic post-mortem.
- Jonathan Bernstein: A Mantra for the Democrats. I would note that Jonathan referenced my post from Monday, and I would answer that my goal was not to suggest that focusing on democratic reform is a winning issue. I am making a governing/long-term need argument.
- G. Elliot Morris: The Strategist’s Fallacy in American politics
All of these folks note that the study is well done and impressive, and I would agree.
But I also agree with Masket’s subtitle: “A report reaches the conclusions it wanted to reach.”
I would have readers note Bernstein’s mantra (in the post) but also state that I thoroughly agree with this assessment of the report.
The truth is that there are two substantial biases at work causing trouble. One is the illusion that campaigns determine outcomes. The losing party almost always believes that they could have done things to change that, and in fact it’s always easy to find things that could have been done better. The truth is, however, that electioneering and messaging just aren’t that important.2 Republicans didn’t suddenly lose their ability to run a decent campaign in 2008; what happened that year is that a massive recession started while a Republican was president, and the party suffered. It’s just a mistake to think that the campaigns and candidates were why Barack Obama defeated John McCain that year. The same with 2024. The same with every election year.
The overlapping bias is that people whose job it is to work with words tend to think that proper use of words is more important than it actually is. Not just words, but ideas, images…all of that stuff.3
Elliot notes, among other things,
The report also spends little time discussing the role of inflation in Kamala Harris’s loss in 2024, and frames the economy as an exclusively Democrat vs Republican problem instead of acknowledging an incumbent vs opposition dynamic at play last year — a big mistake that I have covered before.
And also,
But my biggest problem with the report is with the conclusion itself. The report suffers from something I’m going to call the “Strategist’s Fallacy” in politics — the tendency for campaign consultants and political strategists, especially on the Democratic side (where quantitative analysts are overwhelmingly focused on policy positions and ideological point-positions), to map their mental model of how they make political decisions onto voters. They implicitly assume all voters make choices and select candidates the same way elites do. This is wrong for several reasons, which I will explain.
[…]
The Strategist’s Fallacy, simply put, is the incorrect belief that voters think like strategists — i.e., that they hold concrete issue positions and ideological labels and match them up to candidates, like a customer picking an entree off a dinner menu — so the way to win is for a candidate to tweak their individual positions on policy. This mental model does not hold in reality, leading individuals to provide poor assessments of voters’ decisions and organizations to give candidates poor advice.
This fits, I would note, what I am constantly arguing about “messaging” and dovetails with my post on Monday.
As a general matter, I have a set of core objections about the approach that studies like this undertake.
- The country isn’t one big district wherein the median voter can be targeted. While I wish (and that is the appropriate political science term at the moment) that our electoral institutions produced electoral outcomes that better reflected national political sentiment, none of our electoral institutions do that. The Electoral College is the closest we come, and it filters national preference through 51 individual contests that are weighted to privilege some voters more than others.
- Like Bernstein and Elliot both note above, campaigns are not as important as candidates, consultants, or the public thinks they are. Context matters a lot, and the 2024 election cannot be understood without the context of inflation (and, I would argue, general frustration over the pandemic).
- Along those lines, Elliot is correct to highlight incumbency factors that are ignored in purely left/right analyses.
- Given the way that party leadership develops in the United States, as I noted yesterday, this strikes me as a not particularly useful approach. I mean, remember the 2012 post-mortem the GOP did? Yeah, Trump ignored that and then went on to reshape American politics, doing a lot of things that were the opposite of that report.
At any rate, I recommend all of the above posts.
One last passing comment on the framing, which includes the following.
Republicans also moved toward the center on several issues, including moderating their stances on Medicare and Social Security and dropping pledges to repeal the Affordable Care Act, ban abortion nationwide, and pass a constitutional amendment to prohibit same-sex marriage
It seems to me that it is noteworthy that, despite not trying to repeal the ACA, the last month in particular has shown a real willingness to attack the ACA nonetheless by Congressional Republicans. The entire shutdown is focused on ACA subsidies that the GOP cut, for example, (and many GOP states never expanded Medicaid under the ACA).
Further, attempts at prohibiting same-sex marriage via an amendment (which would fail) may be off the table, but there are legitimate hopes in some circles of the GOP that SCOTUS will take care of that problem for them.
*It is no doubt a comment on my age and the way my brain works that “Welcome PAC” made me think of the theme song to Welcome Back, Kotter. To my amusement, the first two stanzas (and arguably the whole song) could make for a bit of an ode to the common refrain that the Dems just need to move to the middle.
Welcome PAC
Your dreams were your ticket out
Welcome PAC
To that same old place that you laughed about
Well, the names have all changed
Since you hung around
But those dreams have remained
And they’ve turned around
[…]
Yeah, we tease him a lot
‘Cause we got him on the spot
Welcome PAC
Welcome PAC, welcome PAC, welcome PAC, Welcome PAC, welcome PAC
BTW: I hope this sidebar doesn’t sound too flippant or seem overly denigrating. My brain really did go here immediately, and the lyrics really do fit to me in an amusing way. Let me stress that the report has some quality information and analysis, even if I find the conclusions to be questionable.
********************@
It may be that messaging and campaigns do matter in some primaries, as a higher percentage of engaged voters take part.
One point that has been used against Democrats since last year’s election actually undercuts the idea that the problem was one of messaging. It’s the fact that Trump did better than usual for a Republican in many blue states. In fact, Trump improved his margin in all 50 states relative to 2020. Yet his improvements in the core battleground states were comparatively small. What this suggests is that in the very places the campaigns focused their attention on–the places where voters were being inundated the most by canvassing, voter registration drives, ads, rallies, and so on–Harris overperformed, and Trump underperformed, relative to the national average. This, in my view, cuts a serious dent into the narrative we’ve been hearing that Harris’s defeat was at bottom a campaign problem. The fact that so few pundits and politicians have made this fairly obvious inference is, to me, proof of the degree of confirmation bias that goes into most post-election narratives.
Let’s flip this around. Is there any way a really poorly-crafted message might damage a candidate? For example, Obama endorsing Sharia law in majority Muslim neighborhoods? Or Texas governor Abbott calling for tax money to be used to finance a gun buy-back program?
Or how about a real world example. In 1994 California governor Pete Wilson endorsed a crackdown on undocumented immigrants. Prop 187 is credited with helping to turn California from purple to blue. Wilson tried out a nativist message, Californians turned into Democrats. (Yeah, obviously that’s oversimplifying.)
If the wrong message can be fatal then messaging matters.
The country as a whole is split 50/50. A tiny number of votes in a handful of states can determine the outcome. For example, Arab voters in Michigan angry about Gaza may be the reason why Michigan went for Trump – 49.7% vs. 48.3%. That’s 15 electoral votes.
The wrong messaging also affects candidate selection, which in turn affects the election, whether local, state or national. Messaging also defines the parties. See: the American South starting in the 1960’s. Messaging via media, let’s say Fox News, can motivate turnout, or depress it.
Make America Great Again is a message which helped to solidify a cult of personality. If Trump’s message had been America’s Already Great, Let’s Just Make A Few Improvements*, he’d have lost. MAGA is messaging that shaped the electoral battlefield.
Academics love them some theories, especially theories that dismiss or diminish the effects created by unpredictable humans.
*Hard to fit on a hat.
@Michael Reynolds:
To be blunt: I am not saying, nor have I ever said, the messaging doesn’t matter at all.
But what I, and a lot of people who study this stuff for a living, note that it simply does not matter the way you, and others, argue that it does.
On the one hand, I know you know, as a writer, how far you had to reach for that example.
On the other hand, it isn’t like someone campaigned on being a dictator or anything like that! I mean, can you imagine?
@Michael Reynolds:
Yes, let’s take the RFK, Jr. route and use our guts to understand the world!!
@Kylopod: This is an excellent point that I think one of the posts I linked also made.
And as Bernstein specifically notes, both parties engage in messaging and campaigning.
Winning means: message good!
Losing means: message bad!
But this ignores any number of other very important variables.
It also tends to end up being cherry-picked as to which message really mattered.
You know when I bring up commercial aviation as an example? Sometimes it fits well, especially when considering accident investigations, things like safety measures (even for unlikely possibilities), and other related matters.
Other times it’s “When what you know is aviation….”
@Steven L. Taylor:
It doesn’t have to be likely in order to prove the point that if messaging can kill a campaign, then messaging matters.
Speaking of having to reach to make a point:
With genuine respect I have to point out that political science is not medical science. There are very few political theories that can be tested and falsified.
I’m not overstating the importance of messaging, I absolutely agree with your position that structure matters more. But we can’t realistically fix the structure. So your position is a no-win scenario, and I know you, as a fellow geek will recognize the source: I don’t believe in the no-win scenario.
@Steven L. Taylor:
One caveat is that there are data-oriented types other than political scientists. Pollsters, for one. And certain pundits who act like statistics nerds. The problem is that the use of stats to support one’s argument is often a way of hiding weaknesses in how stats are to be interpreted. It’s how we get theories like the so-called “blue wall” idea that was used to argue that Dems had the advantage in 2016 because of all the states they had consistently won the previous six cycles. The other day I looked back at articles from shortly after the 2004 election to see what pundits were saying, and I found an interview with pollster Peter Hart in which he asserted that every winning candidate since 1912 was the candidate who won the most states along the Mississippi River. I didn’t check whether this was accurate; I’ll take his word for it when it comes to elections up to that point. I did, however, notice that it totally didn’t apply to the three Democratic victories after he made that statement: Obama won just four of the ten states–Minnesota, Wisconsin, Iowa, Illinois–and Biden won three of those (losing Iowa).
I’m not disagreeing with your overall point, just saying that vibes aren’t the only way to make a bad argument.
@Michael Reynolds:
Shakira law for everyone else!
@Kylopod:
This reminds me of a Joyner post decrying the types of maps presented to show x, but really just show population centers.
Statistics are a tool. Like any tool they can be misused. A hammer can be used to drive nails. A hammer can also be used for assault. But deliberate misuse of statistics is not the only danger. If used poorly, statistics is more like a hammer that leaves a nail in a state that may or may not be driven, but appears to have fastened properly.
Not to mention that interpreting statistics introduces another opportunity for bias or outright bad faith.
@Michael Reynolds:
I get your point, but medical science is messier than you are portraying here. Political science is messier, sure. But medical science has similar points that risk erroneous conclusions. The difference is that medical science typically has higher standards of rigor.
@Steven L. Taylor: This discussion reminds me of the common rant I saw during the last cycle that picking between Harris and Trump was like a airline host giving you the dinner option of chicken or broken glass in shit and responding, “how is the chicken prepared?” Somehow all of these people wondered how they might fare under middle-of-the-road Harris and then defaulted to bomb-throwing Trump as if they were driven by perceived risk avoidance. If Trump’s message could not break through, what was there for Harris to promise?
Consultants may like to blather about messaging.
On the other hand actual citizens deliver a coherent message :
No Kings, thank you.
Their reasons may be varied, but the final message is clear as 7 million persons can make it
First, Harris didn’t lose because of inflation. Biden would have lost, unfairly, because of inflation. Unfairly because the admin and the Fed handled the situation about as well as they could, inflation being built into post COVID. Harris lost because of inflation, Biden, and having like six months to campaign. All things considered, she did good to do as well as she did.
Also, yes, there is a consultants fallacy much like the more famous pundits fallacy. But in fairness, what do political consultants do? Messaging is all they really have control over. They have no control over the economy and it’s not on them to reform the system. Maybe the party needs a permanent, standing policy analysis operation. What we have is a loose collection of funders and advocacy groups. The GOPs are closer to it. The RNC produced the post-mortem they all ignored. Heritage produced a policy blueprint, Project 2025. Which isn’t focused on winning election, but on how to restructure and fuck over the country if they did win. Where’s our Project 2029? Who would produce it?
@gVOR10: Where’s our Project 2029
Our Project 2029 is to go back to the rule of law. No amount of messaging can overcome half the electorate knowingly voting for a rapist, racist thief. WHO DOESN’T EVEN HIDE IT. If you are a 20 year-old black or Hispanic male and you voted for Trump, there is not any messaging by the Democrats that will work on you. I am convinced the only way Democrats win the presidency with the current (and even worse prospective) gerrymandering is when things are bad. When Obama won things were so bad that even a black man won. Biden won because people knew just how bad Trump was. Four years later they decided Trump was much better than they originally thought.
The Republicans have the secret sauce – they will say anything to win an election and do anything to stay in power. They don’t pretend to care about the future of the country and half the population is okay with that.
@gVOR10:
Democrats are not now, nor have they ever been in recent decades, playing a long-game. And they do not have their own equivalent of a well-funded powerful Heritage Foundation. A Democratic Party Project 2029 would – de facto – have to be an ad-hoc plan to Make America Normal Again.