Since Trump’s surprising (if narrow) win in 2016, two contrasting electoral strategies to oppose him have been suggested: mobilization (turnout more people who already are against him) or persuasion (convince existing voters who may have voted for him to vote for his opponent).  One argument against persuasion was that Trump’s base is so racist that it’s impossible to appeal to them or dislodge them without being equally racist—a morally unacceptable path even if it were possible. Trump’s surprisingly narrow loss in 2020, though, gives us more information to think about this in a deeper way. 

Multiple pieces leading up to the 2020 election--and since the election--have argued that persuasion was a lost cause because of the Trump base’s racism. Here are two:

In the Washington Post:

If Democrats are to have any hope of defeating President Trump on Election Day, they should not waste time trying to persuade his supporters to turn their backs on him. It’s simply not going to happen.

They are going to be with him down the line, no matter what. 

In the New York Times:

Don’t waste your time reaching out to Trump voters as I did. Instead, invest your time organizing your community, registering new voters and supporting candidates who reflect progressive values that uplift everyone, not just those who wear MAGA hats, in local and state elections

Obviously nobody is obliged to convince anyone else—or even to engage in a conversation with someone they believe is racist. 

But I think there’s another important dynamic here that’s being missed: electoral persuasion isn’t just about personal appeals to others, and it’s not even always about convincing or changing people through heart-to-heart conversations. In elections, persuasion is about changing voting behavior and political support, not about selecting friends or discerning (or changing) someone’s personality and intentions. 

Let’s back up a bit, because I think this discussion about mobilization vs. persuasion has been muddied by an analytically misguided conversation about “economic anxiety” and racism.

Leading up to the 2016 election, there were assertions that Trump supporters were motivated by “economic anxiety.” Critics said that such assertions were a cover for racism, arguing that  the unemployment rate was low or the income data didn’t support claims about economic anxiety. 

This debate is often presented as a dichotomy: Trump supporters must be either racist or economically anxious—as if they were not unemployed at the moment, the reason for their support of Trump must solely or primarily be racism. 

But in reality, these two traits are neither always separate nor even neatly separable by any known analytic method. In fact, historically, they’ve long been fused in various versions in complicated ways that we missed going into 2016, and I believe still were missed going into 2020. The idea that we can simply measure either factor with polls or economic indicators is not correct: economic anxiety and status, for example, are not just about the unemployment rate at the moment but how one perceives the trajectory of their—and their families’—life chances. Unemployment can be quite low and economic anxiety also quite high at the same time. Racism also isn’t something that can be assessed based on what people say in response to a poll. And crucially, in people’s minds, loss of economic status can be tied to perception of loss of racial privilege and supremacy.

To give context to this conversation, I want to include here a tweetstorm of mine right after the GOP convention in July of 2016, a convention which was deemed by much of the press to be  “disastrous.” It was said that Trump’s speech was not working well, with a lot of brouhaha over the potential plagiarism of Melania Trump’s speech.

I heard something different. 

What does this mean in practice? It means that people respond to policy a lot more than we assume, and that cultural dynamics are then fused with this policy.

This column byNew York Times writer Jamelle Bouie, for example, highlights  something that’s often overlooked: the US government’s response to pandemic included a sizable financial boost to poorer households. I’ve seen commentators say the aid was pathetic since the one-time payment was just $1200. But that claim overlooks a key part of the government’s response: an increase of unemployment benefits to $600 per week. That’s a staggering $2400 per month per unemployed person. The support was so substantial that some households dramatically improved their financial outlook during the pandemic, when the financial aid was in effect. 

Bouie writes:

At the risk of committing the same sin as other observers and getting ahead of the data, I want to propose an alternative explanation for the election results, one that accounts for the president’s relative improvement as well as that of the entire Republican Party.

It’s the money, stupid.

At the end of March, President Trump signed the Cares Act, which distributed more than half a trillion dollars in direct aid to more than 150 million Americans, from stimulus checks ($1,200 per adult and $500 per child for households below a certain income threshold) to $600 per week in additional unemployment benefits. These programs were not perfect — the supplement unemployment insurance, in particular, depended on ramshackle state systems, forcing many applicants to wait weeks or even months before they received assistance — but they made an impact regardless. Personal income went up and poverty went down, even as the United States reported its steepest ever quarterly drop in economic output.

As Bouie states, it’s perfectly reasonable to assume this economic aid may have bolstered Trump’s support among key groups—especially since he made sure to have the checks appear as if they were personally signed by him—an unprecedented move that may have even delayed the payments, but may also have had the intended impact.

Here’s another data point to suggest that policy matters: the dramatic shift to the GOP in the Rio Grande Valley, a Tejano region. This shift surprised many analysts who assumed that Trump’s anti-immigrant stance would backfire among Latino voters. As others have noted, there is no single “Latino” vote: it’s a category that includes a range of backgrounds. And Trump did somewhat better than he did 2016 among Latino voters in multiple regions. However, the Rio Grande Valley is especially interesting for this discussion. I lived in Texas for many years. Even within the state, there are many flavors of “Latino.” The Rio Grande Valley is particularly diverse, with some Latino people being far removed from identifying as immigrants because many of the families have lived in the area since before Texas became part of the United States. As the saying goes, they did not cross the border, the border crossed them.

The other important part of the Tejano reality is that there are a lot of jobs both in law enforcement (including border patrol/enforcement) and in the oil and gas industry. As many news reports have noted since, there was a widespread perception that Trump was more likely to protect their jobs. In addition, the cultural affinity with the Republican party—with individuals being social conservative or even anti-abortion—may well have been the reason the region swung in Trump’s favor, sometimes dramatically. For example, Zapata county, which Hillary Clinton won by 33 percent in 2016, voted for Trump over Biden, 52.5% to 47.1%. A similar pattern was repeated elsewhere in the area.

What does this mean in practice? I don’t have easy solutions to such complicated dynamics, but I think that it’s possible to make a few assertions. First, political persuasion isn’t necessarily about having a heart-to-heart conversation where one person has an epiphany. It’s also about political and policy stances or actions that people perceive as benefiting them. We also need not oscillate between a crude economic reductionism, where people only vote for paycheck considerations, and ideas about incurable racism, where people vote no amount to support racist politicians regardless of other incentives or considerations. 

We can acknowledge when racism and economic considerations are intertwined; economic populism often works by promising welfare state/social-benefits to the “people.” The “people” are often defined ethnically, but appeals are rarely that simple, especially in a country like the United States, where racial identity is complex and shifting among many groups. We can even recognize that opposing racism is a moral commitment that supersedes any electoral calculation.There’s no doubt that there are people for whom racism is an overwhelming consideration who won’t vote for a party they see as opposing white supremacy.

But I don’t think that means there is no hope, ever, of appealing to groups to whom others are trying to appeal based on racism, or, more accurately, on ethno-nationalism populism. What we’ve seen in 2020 is that people do respond, in complicated ways certainly, to what they perceive as both paycheck and economic security issues and cultural affinity. Neither demographics nor turnout patterns are destiny. Trying to change people’s hearts may or may not be futile in the personal realm, but there’s no need to determine that conclusively. People respond, sometimes dramatically, to policies as they perceive and experience them. The personal is political but all politics isn’t always that personal.