With images of violent unrest, Donald Trump is trying to recreate Richard Nixon’s 1968 law and order campaign. What were the effects of racial protest and riots on elections back then? And will they be the same today? Omar Wasow finds that non-violent protests helped Democrats win nearby elections as the media emphasized civil rights. But looting, burning, or rioting associated with later protests played up by the media likely cost Hubert Humphrey the 1968 election.Those effects might still linger. Elizabeth Coggins finds that liberalism in the abstract has remained unpopular since the late 1960s, when it became increasingly associated with race, crime, and militancy, in both the media and public views. Even though the public favors liberal policies, liberal ideology still brings to mind unpopular and racialized symbols. The historical question matters for how minority communities gain political voice and how the public understands liberalism—with the effects still felt today.
Matt Grossmann: This week on The Science of Politics, racial protest, violence, and the backlash against liberalism. For the Niskanen Center, I’m Matt Grossmann.
With images of violent unrest, Donald Trump is trying to recreate Richard Nixon’s 1968 Law and Order Campaign. What were the effects of racial protests and riots on elections back then? And will they be the same today? It’s not just a historical question or one about campaign strategy. It matters for how minority communities gain political voice and how the public understands liberalism with effects that linger today.
This week I talked to Omar Wasow of Princeton University about his American Political Science Review article Agenda Seading: How 1960s Black Protests Moved Elites, Public Opinion and Voting. He finds that nonviolent protests helped Democrats win nearby elections as the media emphasized civil rights, but looting, burning or rioting associated with later protests played up by the media likely cost Hubert Humphrey the 1968 election. Those effects might still be with us.
I also talked to Elizabeth Coggins of Colorado College about her ongoing work on symbolic liberalism. She finds that liberalism in the abstract has remained unpopular since the late 1960s, when it became increasingly associated with race, crime and militancy in both the media and public views. Even though the public favors liberal policies, liberal ideology still brings to mind unpopular and racialized symbols. The 1960s were both the period of civil rights success and backlash. Wasow says protests helped to explain both.
Omar Wasow: One real puzzle in the period I look at is that between 1950 and 1980, if you look at public opinion on what’s the most important problem, there are these two bumps, one around civil rights in 1963 and another around crime and riots in the late 1960s. And if your theory of politics is elites dominate well, it’s hard to explain the rise of civil rights, but if you look at the role of protest activity, it really is consistent with the March on Washington and other kinds of protests that are shaping not only newspaper coverage, but in turn public opinion.
In the later part of the 1960s, we see this concern in public opinion around crime and riots, and there I find protest tactics are really helpful. And so if your story is that while tactics don’t matter, then it’s really hard to explain why as violent protest activity is going up public opinion on concern about crime and riots is becoming the most important problem. And then again, we’re seeing kind of the rise of law and order. And so by focusing on that puzzle of this, these variation in public opinion, we can really get at how two prominent theories of how protests work, maybe they don’t or maybe tactics don’t matter are insufficient. And that’s a lot of where my paper is trying to intervene.
Matt Grossmann: Protest tactics matter, nonviolent protests help, but violent protests cause a backlash.
Omar Wasow: There are two main questions that my paper gets at. One is do protests matter and another is do protest tactics matter. In political science there’s a large literature that says elites dominate politics. And I find contrary to that, there is a kind of what I call punctuated pluralism. There is a way in which marginal groups are for at least brief periods of time, able to punch through and make their voices heard through things like disruptive tactics and protest.
And secondly, that protest tactics matter. That the kinds of strategies organizers are using on the ground influence media coverage and that has a profound effect on the kinds of outcomes that happen for a movement. And so in particular, I find that in the 1960s, not only were protests able to get those issues in the national conversation and the issues of civil rights, but that when protest tactics were largely peaceful, particularly when they were targets of violence by state and vigilante actors, that media coverage tended to really elevate rights as a concern in the public consciousness. In a later period of the 1960s, as protesters used more aggressive resistance to state violence, more aggressive resistance to white supremacy, when the protest included protestor initiated violence, that those events tended to be covered by the media as headlines focused on crime and riots, public concern focused on crime and riots and campaigns focused on law and order were successful, particularly Richard Nixon’s campaign for the presidency in 1968 was helped by the kind of law and order campaigning he was doing in response to what was happening on the ground in violent protests in the sixties.
Matt Grossmann: Wasow finds that protests make headlines emphasizing either violence or civil rights.
Omar Wasow: So, I collected 275,000 front page headlines from eight newspapers as a way of trying to gauge does an event today predict a headline tomorrow that mentions either a rights related topic like civil rights or voting rights, or a kind of crime and riots related topic, particularly the word riot or Attica. There was a prison uprising in Attica earlier in the 1970s. And what I find is that a protest today that uses nonviolent tactics predicts a headline tomorrow that mentions rights. And not just a headline, a front page headline with a high degree of, it’s probabilistic, but it’s a high degree of predictive power. And that protest today that includes protestor initiated violence predicts a front page headline tomorrow with the word riot.
So there is a very clear statistical relationship between what’s happening on the ground and what’s happening the next day in the news. And so that relationship to me suggest that what protestors are doing matters. And I should also add that a protest in which the tactics were primarily non-violent does not predict a headline about riots. And conversely, a protest in which there are violent tactics used by protestors does not predict a headline around rights. This is not just that there’s something kind of arbitrary or kind of random here. There’s a very clear relationship between non-violent protest and a kind of focus on civil rights and as I mentioned, the violent tactics and a headline on crime and riot.
So that’s the 1960s. And what we also see is that public opinion follows that pattern. And it’s very hard to explain some of the core patterns in public opinion in the 1960s without differentiating tactics. And so I think one core thing that is sort of obvious once you think about it, but wasn’t front and center in much of the literature, is that almost nobody directly observes a protest, right? So almost the entire way of protest might be having an effect on politics is through the media. That’s not all protest. Sometimes you have a boycott or there are other kinds of tactics, but these kinds of protests are primarily reaching the public through the media. And what we see is that public opinion is covarying really closely with protest tactics, headlines, and then shifts in what people say is the most important problem in America.
Matt Grossmann: And he connects that to election results. Violent protests hurt Democrats, but nonviolent protests help.
Omar Wasow: One other thing I do in the paper is look at the effect of protests on voting. I look at three presidential elections in particular: 1964, 1968 in 1972. And the way the statistical analysis is set up is I imagine that a county is treated by a protest if it’s within a hundred miles of a protest, either non-violent or violent and within about two years of the election. And then I take certain measures of protest activities. So number of people who participated or a number of people arrested, and if it meets this criteria, it’s within a hundred miles, it’s within two years, there are at least 10 participants or 10 arrests, I call that county treated. And if there’s not an event within a hundred miles and within two years and so on, then it’s a control county. And so we can kind of compare counties that got a, “Protest treatment,” and counties that didn’t get a protest treatment.
What I find is that the, “Treated counties,” the ones exposed to nonviolent protest activity vote slightly more liberally between 1964 and 1972. And counties exposed to violent protest activity, vote more conservatively. And that is an important confirmation of work we see cross nationally that there may be differential effects of protest tactics on things like voting behavior.
Matt Grossmann: And there is some evidence from the Martin Luther King assassination that suggests that this relationship is causal.
Omar Wasow: One question in all statistical analyses is can we not just show a correlation between say, non-violent protest activity and more liberal voting in nearby counties and a more conservative voting when there are tactics that escalate to violence. But can we show a causal relationship? Can we show that violent protest is causing voters to shift from the Democratic coalition to the Republican coalition?
And one way to get at that is to try and find something that approximates random assignment so that we can say, well, we’ve got this thing that looks almost like a medical trial. Some people get the treatment and some people get the control by a coin toss. And building on work by two economists, Margot and Collins, I use the assassination of Martin Luther King Jr. in April of 1968 to approximate a kind of random assignment. And so the way this works is King was assassinated, a terrible moment for the country. There are 137 protests that follow that escalate to violence. And that’s an unusual moment. It’s in April, not the summer. And it’s a very specific kind of window in time. And that’s helpful because I can then use rainfall in April. And what we find is that rainfall is very predictive across a wide variety of event types, that if there’s more rainfall, people don’t gather outside, right? It’s sort of pretty intuitive.
And so I can use rainfall to approximate random assignment of protests. Where there’s more rainfall, there’s less likely to be a violent protest. And so using rainfall to predict protest activity, I’m able to create a kind of natural experiment. And what I find is that there is very good evidence that what’s happening on the ground in April is influencing, is causing changes in voting behavior in November, right?
It’s possible that rainfall could be influencing voting in November through a variety of channels. So I also look at rainfall just before Martin Luther King was assassinated, April 1st to third. I also look later in April. And what I find is that it’s not something like region in all likelihood because the only time rainfall is predictive of voting in November is in the week following King’s assassination when most of the 137 violent protests occur. And so that means that unless you’ve got a better story about why rainfall in April is correlated with voting in November, it almost certainly is coming through the channel of violent protest. And that suggests that what’s going on with those violent protests is causing about a six to eight percentage point decrease in support for the Democratic candidate in the 1968 presidential election.
Matt Grossmann: Coggins finds that these effects might have been enduring. She’s trying to answer the question of why liberalism is less popular than conservatism.
Elizabeth Coggins: Conservative as a label is more popular, right? And has been since we’ve been collecting data on how people choose to identify on the liberal to conservative scale. So when we think about it long-term, conservative has held at 10 to 20 point lead really over liberal for the past three decades, since we’ve really had solid data. The most recent numbers from places like Gallup and Pew show that about 24% of Americans choose to identify as liberal with about 37% identifying as conservative, and then about 35% identifying as moderate. And that’s been pretty stable that the gap between conservatives and liberals has shrunk a bit. And that I think we can talk about this more too, but I think has come mostly from moderates moving to the liberal ranks. But over time it’s not the case. And I think this is a really common misconception when thinking about ideological identification, that there was a period of time where the liberal identity was dominant. That’s just not true. It never has been. The gap has been anywhere from 10 to 20 points, but a gap, nonetheless.
Matt Grossmann: Liberal policies are popular, but liberal symbols are not.
Elizabeth Coggins: How can it be that the majority of Americans do and have for decades preferred liberal policies, wanting their government to spend more and be a more present part of their everyday lives and yet prefer to call themselves conservatives? How can this be? And just to sort of add to that, and it may even be more perplexing if we look at those numbers today. Our policy liberalism or policy mood, Stimson’s mood, is at an all time high in the United States. And yet the conservative label still wins out. And we have what we can call for sure a conservative president.
The gains that liberal has made come from moderates mostly over time. But so how do we explain this operational symbolic disconnect? That’s what we usually choose to call it in the literature. And I think trying to explain it requires trying to understand a few moving parts. In my research I find that folks are really drawn to that connotation.
In my research, I find that folks are really drawn to the connotations of the word conservative, and that plays a pretty big role in why this is true. Conservative means things like family and prudence and practicality, and it conjures images of the American flag in our minds. And it’s been rich with meaning for decades. While liberal, on the other hand, holds very different connotations for people, broadly speaking. It carries a sense of defiance, of sort of God and country, and really critically carries these images of race and protest for many people in the US public. And importantly, liberal in the 1960s was more… One way to think about it, I think, is it’s more of an empty cup when it came to meaning. The term liberal itself became popularized by FDR and his fireside chats. He began giving it meaning to ordinary citizens, linking it to the common man. But when compared to conservative, it was a lot less rich in meaning, it had less of a history.
But the 1960s changed that and changed that dramatically. In that decade, that liberal cup got filled. It was filled with a lot of things, but mostly by the end of the decade in the minds of Americans, it kind of conjured these ideas and these images of riots and really racialized images. In brief, liberal became racialized in the 1960s. So conservative became tied with law and order, and liberal became tied with race rights in the 1960s. We see that in our correlational data, which I can talk more about, but it has never as a term really shed that racialization or even really shared it with moderate or conservative. Instead, that link has really just been perpetuated by the media, most especially by politicians, too. As we like to say, JFK may not have been the last liberal president, but he was certainly the last to call himself one. In fact, the most common source of the use of the word liberal in the political sphere has been by conservative politicians. To be called a liberal, we understand, is to be called out as one, politically speaking.
Matt Grossmann: And that may be traceable to the same media coverage that Wassow looks at in the late 1960s. Coggins looked at what else is mentioned when the media is discussing liberal and conservative ideology over a long time period.
Elizabeth Coggins: Our denominator is every article on the front page of the New York Times that mentions the word liberal and or conservative. Then we can kind of start working on that numerator.
And those other important words in groups came from the American National Election Study. And what we did is we pulled the groups that they ask people about over time in the feeling thermometer section, right? And we did that on purpose so we could do some analyses. There are lots of other linkages we’d like to know about, but this would help us gain some traction on how these things shift in connection over time. So we matched up those groups and those people from the ANES, and the feeling thermometer scores, what they do with this is they say to people, we want to get a sense of how you feel on a scale of zero to 100 about this particular group or this particular person, really commonly use scale and political science, and it helps us get a sense of how folks feel about societal groups or certain individuals like political leaders. We generally assume that lower numbers mean you feel coldly toward this group of person or don’t like them, higher numbers mean you feel warmly toward them or like them.
Then with these data, we could see how often those words and groups from the American National Election Study appeared alongside terms like liberal and conservative in the news. And we find that over time, these ideological labels become more and more linked to race, while labor becomes much less prominent of a link to the ideological terms over time in the news media.
Matt Grossmann: And she found that public opinion associations matched those in the news coverage. People’s views of liberalism came to match their views of black militants and civil rights leaders.
Elizabeth Coggins: We find also that feelings toward liberalism became more strongly correlated toward views of civil rights leaders and black militants. So this is where those two analyses link up. We again consider those feeling thermometer scores from the ANES, and these are great data for us because most of them are longitudinal, asked every few years of respondents, so we can get an overtime sense of how feelings change about certain groups and how those correlations between ideological terms and those groups change over time.
So for this particular analysis, we considered the correlation between feeling thermometers for liberals and a host of other groups. And we were primarily interested again in that overtime change. So if we find that correlations between feelings about liberals and feelings about black militants, for example, increase over time, then we can say that these things are linked in the minds of Americans. And that’s just what we found when it came to two really important groups. Over time, correlations are sort of connections between liberals and black militants and liberals and civil rights leaders rose, which means, to us, that they became linked, liberals and black militants and liberals and civil rights leaders in the minds of Americans. This rising correlation also happened with liberal and “black people” as the ANES ask and writing as well. And this is all sort of another cog in that story of how and when that liberal label became racialized.
Matt Grossmann: Wasow and Coggins agree that this should change our analysis of the era in policy and political terms. Wasow says it should move analysis away from policy-led stories.
Omar Wasow: It’s really important to think not about Nixon and Humphrey as just candidates getting elected, but as people who are holding together coalitions. And Nixon is running on law and order, and this is the first time law and order has won at the national level, right? Goldwater runs on law and order and loses. So law and order as an issue comes to the fore and becomes part of the governing strategy of the Republican party for the next half century. So I think it’s important to think about this, not just as Nixon versus Humphrey, but as a kind of durable governing coalition that captures the most powerful institution in the country, and then sees that those methods work for maintaining power. Conversely, Humphrey was the lead author of the Civil Rights Act, and so we could imagine a counter factual where Humphrey wins and law and order doesn’t carry the day, the war on drugs isn’t started by Nixon, and maybe law and order doesn’t become a durable part of the GOP governing strategy. So that’s one thing, is we should think about not individuals, but coalitions and the particular issues those coalitions advance.
A second concern is, well, look at something like the 1968 Fair Housing Act, Civil Rights Act of ’68 that passes following King’s assassination and the violence that follows and how can you sort of dismiss the importance of violence? And what I would say there is two things. One is, first, it is important to realize that the ’64 Civil Rights Act and the ’65 Voting Rights Act come after periods of significant non-violent direct action, right? So we have the March on Washington in ’63, and within a year the Civil Rights Act of ’64 is passed. And that’s a buildup after years of sit-ins and other kinds of protests where, again, protestors are putting their bodies on the line in the front of dogs and fire hoses and making themselves targets of state violence.
Similarly in ’65, famously the Bloody Sunday march at the Edmund Pettus Bridge. That’s the one where John Lewis has his head bashed, and that gets covered nationally. Remarkably, ABC News was broadcasting a documentary about the Nuremberg trials at that moment. They cut in to this war crimes footage to show state and vigilante actors beating peaceful protestors in Selma. And so that shocks the conscience of the nation, and within five months we get the Voting Rights Act. So I think it’s really important to appreciate that there is a period of enormous legislative success on behalf of civil rights, dismantling Jim Crow without significant protest or initiated violence.
Then on the 1968 Fair Housing Act, I think it is entirely plausible that the violence that comes in those protests helps create a sense of urgency, but we also can’t really distinguish there’s sympathy for King and his assassination. Into that ’68 Fair Housing Act, there’s also anti-riot legislation proposed by Strom Thurmond. So I think just a much more complicated case in ’68, it’s hard to say specifically what caused it, and it’s much less clearly a victory because of some of the, again, kind of tough on crime policies that get enacted as part of it.
So to my mind, the main takeaway is that activists were able to achieve really landmark victories without having to use this more aggressive resistance to white supremacy and state violence. So you can get a lot of what you want without the potential blow back that followed in the later period. And that’s the kind of the really important lesson. It’s not to say that violence doesn’t help create a sense of urgency or help attract attention, but it also seems to help mobilize opposition to civil rights.
Matt Grossmann: And Coggins says policy views did become more conservative in response to liberal policy, but they moved back to the left. That’s unlike the symbolic meanings, which stuck.
Elizabeth Coggins: Do we look at public policy mood, which is a measure of policy preferences, right? So that’s more of the operational side of things. Do we see a shift toward conservatism? We do, yes, but it’s after a sort of big sweeping liberal agenda happened. And so we see that happen often in public policy mood, so that’s a pretty predictable response in many ways. But what I think is really interesting is that we don’t see that sort of switch-back when it comes to liberal identification, right? So sometimes things come off of their equilibrium, but they returned back, right? That’s the point of equilibrium. This didn’t happen with the ideological labels, and that’s why this is such a fascinating story, right, is that yes we see the sort of the policy content of liberal ideas survives and thrives really in this country, but that isn’t true of the label itself.
And so those are really two different stories, I think. Are they connected? Yes, but I don’t think that we can talk about them in the same way as what are the things that affect each one? What are the things that affect people’s policy preferences? Concrete policy preferences are so distinct from what affects their sort of gut level attachment to a team. And that is how I see the difference in policy preferences and ideological team preferences.
But I still think that regardless of all of that that protests and then riots, I think, were sort of a really key feature of what caused policy backlash. But we see that backlash turn around again and go back liberal again if we think about that, if we’re calling backlash policy mood, if we’re tracking backlash with policy mood. But we don’t see that with ideology, right? So I think also Wassau’s research demonstrates the framing around that was really, really critical and his research shows that really strongly. And then for me the question is, “And then what happens?” Right? Well, we know from public health stats research that racial policy preferences are directed in many ways by the media, and so we just see this cycle sort of start over again. But the one thing that stays pretty predictable and pretty stark in the minds of Americans is that racialized nature of the ideological term, even if their policy preferences shift back and forth.
Matt Grossmann: Wassau finds public concern about social control did follow riots, not just campaigns.
Omar Wasow: A core insight, for me, of this work just looking at raw data was when I was first looking at public opinion data on concern about crime and riots what, in the paper, I refer to as social control is there’s this enormous movement on concern about crime and riots. And at first I just thought, “Well, it’s public opinion data. People are being asked, ‘What’s the most important problem?’ And it’s moving around because public opinion data is noisy.” And it wasn’t until I overlaid that public opinion data on violent protest data number of arrests per month, that suddenly what seemed like noisy public opinion data had a structure. It was going up in the summer and down in the winter, up in the summer, down in the winter. It was not moving particularly in response to elections. It was spiking massively when there were big protest events that escalated to violence in places like Newark and Detroit. It also was working kind of cumulatively as there were more of these violent protests across ’65, ’66, ’67. We see concern about social control go up in the summer even more and then down again in the winter.
And so that suggests that if it’s not moving in response to presidential campaigns, that maybe the elite ability to move public opinion on this is not that great. And it also suggests that if your story was, well, this is simply a function of white racism, which certainly is an important factor, it’s hard to explain why it goes down in the winter, right? We don’t think of racism as being seasonal. And so that leaves this sort of large amount of unexplained variation, this enormous up and down where it goes to become the most important problem in America around ’67 and then again in the early ’70s-
Problem in America around ’67 and then again in the early ’70s, you’ve got to have some other story about what’s driving this enormous movement. And given the way in which protest activity on the ground is co-variant with public opinion, that suggested to me that that it really is doing a lot of the work for what’s happening. And again, it’s not just what’s happening on the ground because of course if you’re 100 miles away, you’re not observing that directly, you’re getting it through the media.
Matt Grossmann: So what about today? Coggins project is mostly about some enduring changes. Liberalism and conservatism still maintain their associations with big government, but that’s not the part that’s increasing.
Elizabeth Coggins: There are some components of what are linked to the ideological terms in the minds of Americans that are quite stable over time. We find that big government, big business is sort of over time, that’s a pretty important dimension in how people link up their own feelings about the ideological terms and the things that fill up that cup as well.
Matt Grossmann: People who identify as liberals are liberal across policy issue areas, but conservatives hold mixed policy views and are instead united by symbols.
Elizabeth Coggins: So we ask both groups for their policy preferences across a broad swath of areas. We talk about economic issues. We ask them even about cultural things. So we ask them these 10 questions. And what we find is that 90% or so of people who call themselves liberal also have liberal policy preferences in almost every area. But self-identified conservatives, a full quarter of them, more than 25% have majority liberal policy preferences across both economic and cultural issues. And I think that quarter of people, that’s 25%, let’s say, roughly of the voting electorate is probably the most interesting when we think about that in like an applied sense, how do they vote.
And this, we like to call like the movable middle. These are the folks who really move things in elections. Those thermostatic shifts we see in policy mood, all of that, what happens in outcomes. I think that those are the people doing that shifting
Matt Grossmann: It’s become more liberal, so current protests might not have the same impact.
Elizabeth Coggins: The landscape is quite different. And so, a law and order message lands at least somewhat differently than it did in the 1960s. So I think that people have, the country as a whole, and this is some of the work I’ve been doing with Stimpson and Bumgarner and Mary Layton Atkinson, the country as a whole on cultural types of issues has shifted in a more liberal direction. And so over time, we’ve seen even from the 1960s, that even if we, even if we consider that there are still very racialized messages out there, that liberal is still very racialized.
The fact of the matter is, is that over time on multiple cultural dimensions, the country itself is just more liberal, racial liberalism, gay rights liberalism, women’s rights liberalism. And so the fact of the matter is it’s like, again, that the playing field is quite different, I think, than it was in the 1960s. And the denominator is quite different than it was in the 1960s. But I don’t think that means that we can say for sure that when we think about how that plays out in an electoral sense, that that means Biden will definitely win.
I don’t think that that is an obvious conclusion from what we’ve seen. I just think it means that sort of the center point is more liberal, but there’s still a lot of movement around that center point, and still a lot of opinion that doesn’t line up with that center point as well.
Matt Grossmann: Walso sees important lessons for today about human sympathy and the importance of protests.
Omar Wasow: Broadly, I want to be humble about what a lesson from 50 years ago tells us about today. Some of the dynamics are clearly different. There is dramatically different kind of media environment in technology. There’s the voting public has changed, but to kind of key in on three broad things. One, I think there’s an element of this that is really telling us something about human nature. That when people see somebody who is ostensibly a peaceful protester being beat up or kind of contemporary case, there’s enormous sympathy for George Floyd who is crying for his mother as he’s has the full weight of Derek Shelvin, Officer Shelvin on his neck.
That has a very clear kind of good guy, bad guy arc in the public mind. And we see police forces, conservatives condemn that. And that to me is a very clear echo of the beating of Rodney King in 1992, the lynching of Emmett Till in 1955, that these images of state violence against, or vigilante violence against African-Americans do a lot of work to shape public discourse and public imagination about is our rights important. So I think that is very clearly a kind of consistent theme. To tie it to my work, a moment like Selma is a moment where we see this huge spike in national concern for civil rights.
And it’s short-lived, but it jumps to the top of the public’s concern is the most important problem. And again, within five months we get the Voting Rights Act. So, I think that is very clearly an echo of these sort of almost, as John Lewis said, we’re trying to dramatize injustice, and these images do a lot to dramatize and justice and that’s a consistent thing. Second pattern that I think is similar is that protests clearly, going back to the question of do protests shape the agenda. The act of one young woman, Darnella Frazier, 17-year-old documenting, bearing witness to the killing of Floyd has had this profound effect of mobilizing by some estimates over 10,000 protests in the United States and around the world.
And that has pushed issues of rights and police reform to the forefront of our politics, so that clearly is … There may be academic debate about that, but it just seems pretty clear that protests have driven the national agenda for the last few months and not exclusively, but significantly. And then I think on the question of tactics and whether violence or non-violence matters. I think there’s a little bit for everybody. By that, I mean, kind of in the academic community, I think there’s some evidence that is entirely consistent with some of my work and Dan Gillian’s work.
Violence clearly helps to create a sense of urgency, draw attention, and I do have findings that show violence increases media coverage. So that’s not something I disagree with. And then where it gets a little complicated is by one estimate, 93% of the protests have been non-violent. Media coverage has focused way disproportionately on the events that have had protest or initiated violence. There have also been these incidents of vigilante violence. And so how we parse that out into is this hurting the movement, is it helping the movement, is hard to say.
And generally, my read has been that the press coverage and the kind of the public impression has been that this is overwhelmingly a peaceful movement. Again, 93% of the events were non-violent and that the public has been generally sympathetic. We’ve seen in recent polling, 60% of the public favors police reform over more punitive responses, but it’s also highly partisan. About 90% of those supporting Biden favor police reform and only about one in five people supporting Trump favor police reform.
Matt Grossmann: Coggins says there are some reasons to expect the same political playbook to work, but some differences.
Elizabeth Coggins: We know after the 1960s, sort of the late 1960s, that Nixon really capitalized on this image of law and order and saw that they could link, maybe less in your face, liberal and race and Democrat and race. But, we did see that this cue, that we could change the way we talked about it and use this cue, this heuristic of tough on crime and law in order to sort of do really similar work. And then we saw Reagan do that. Will it work the same way in 2020? I think we have a lot of reasons to think that it will, and then maybe some reason to think it might not land as solidly as it did then.
So, I’ve been thinking about this question a little bit, but I do think that if we think about Trump and his sort of playbook right now, somebody’s reading the playbook from Nixon as well. We saw the John Ehrlichman, the expose come out about the playbook that they were using around race and appealing to a silent majority. I do think that Trump is banking on there being that silent majority out there but because law and order still lands in a really similar way. But I think the real question is, does it land with the same number of people in the same way?
I think it’s a Pew study that says almost three quarters of Biden voters say it’s a lot more difficult to be a black person in this country than a white person versus what like 10% I think of Trump voters. This number again, which I think this is a really critical component, is up from even 2016. We asked that same question in 2016 of potential Hillary voters or people we know voted for Hillary, and it was just a little over half of the people then who actually cast a ballot for Hillary Clinton, who then said, yes, it’s harder to be a black person in this country than a white person.
Now we have like 76% of people. So, we have reason to believe those denominators will be similar, that the same bulk of people will be voting for Biden who voted for Hillary, and maybe even more. So the question really comes down, I think, to that movable middle, those people who hold those conflicting symbolic and operational views when it comes to politics. So how are those people thinking? Where did those people fall?
Matt Grossmann: And Walso says Trump might not be able to replicate Nixon’s message because he sounds more like Wallace.
Omar Wasow: Another issue that might be seen as kind of a limitation or kind of a boundary condition of my work is thinking about Nixon and Trump as comparables. And I looked at American national election study data from 1968 and I found that on a question of how do you view the candidates on a scale from solve poverty and address unemployment to respond to unrest with full force of the state, that people perceived whites and blacks perceived Humphrey to be sort of moderate to liberal more on the soft poverty end of how we should respond to unrest.
They perceive Nixon, interestingly, as moderate as the modal response from whites is that he is right in the middle of that scale. Most whites viewed him as a four on a one to seven scale and the person who was highly rated on law and order was George Wallace. There were three candidates in 1968. And so Nixon was perceived as the moderate candidate on how to respond to unrest between liberal Humphrey and segregationist George Wallace. And I think Trump’s potential miss kind of playing of his hand is that law and order could be an issue that he could campaign on like Nixon, but he uses so much violent rhetoric and has engaged in so much kind of, as those people described it in the ’60s, kind of spectacles of violence.
Like what happened in Lafayette Square, where peaceful protesters are cracked down on that instead of looking like Nixon, the moderate on how to respond to unrest, he looks George Wallace. And that actually is something that most of the public does not support.
Matt Grossmann: Coggins says Democrats still are not defending liberal symbols. So she doesn’t think ideological change is inevitable.
Elizabeth Coggins: Is Biden better able to define liberalism? When I think about that, then those in the past, I mean the real short answer to that is no. And the reason is really simple. I think he simply doesn’t attempt to. He doesn’t call himself a liberal. He doesn’t link his policy preferences to being liberal. Whereas Trump, on the other hand, he calls himself a conservative proudly, constantly. And this is in US politics as strong as a cue as there is, I think. These attachments that we have to these ideological teams I argue are really, are really gut level. They are these prototypes that we see ourselves in and a lot of how we see ourselves in them is controlled by the news media.
And so if we have our leadership using one term and defining it and making it rich with meaning, and then having what is considered to be the other side of that coin, even though there’s a lot of evidence it’s different currencies, liberals and conservatives. But if we have the other side of that being left vacant, like Biden is very much doing, I don’t think that we have a lot of movement because of leadership from the elite.
Matt Grossmann: Walso says videos might be making a difference this time producing more durable shifts in attitudes.
Omar Wasow: One really big difference between the 1960s and now is that activists really needed the NBC, ABC, CBSs of the world, the New York Times to show up and cover a protest for it to have any kind of national resonance. And now, because everybody has a video camera in their pocket, it’s possible for there to be a moment like George Floyd’s killing to be documented by one person and then transform our national conversation. And the steady stream of those videos, whether it’s Jacob Blake or Ahmaud Arbery or the recent Daniel Prude videos, these videos that are in the language of the 1960s shock the conscience, allow whites who might otherwise have not really appreciated the degree to which there is a sustained and kind of brutal level of state violence against, in particular, black men but also black women and also not exclusively black people, that there’s this militarized policing, the unaccountability of policing in America becomes visible to a white audience that might be liberal, but not particularly conscious or aware of this high level, this over-policing.
And so my intuition is that the kind of the dynamic that protests are capturing, which is to kind of take an issue and try to get it into the national attention is actually very similar to what these videos are doing and that they kind of work together. And so that means that this should likely be a more durable shift because these videos are not going to stop coming. And those videos do in conjunction, particularly with protests, force people to look to in some ways, have a level of empathy to kind of sit with what it is to be black in America. And that in turn is causing people to update their understanding of these issues and the kinds of policies they’re willing to support.
Matt Grossmann: And he sees more nationalized effects this time, but some that remain local.
Omar Wasow: I suspect that there will still be local effects of protest activity, that there is a kind of visceral dimension to your community seeing a building go up in flames. I think about this in Princeton, if there were protests where a lot of buildings were set on fire and that’s eight miles. Sorry, in Trenton there were building set on fire and that’s eight miles from Princeton. And I’m imagining in the late 1960s, would that have influenced voters in Princeton in a kind of visceral way, and I suspect that it did and that’s what my data show. At the same time, I think you’re right that the degree to which an event like Kenosha is on heavy rotation on Fox News, I did some conservative talk radio and there was a sense of there’s a crime wave that we’re in the midst of.
And that story is going to mean that even though in the aggregate they’re actually relatively few injuries and deaths at these events, there were some events in the 1960s where dozens of people are shot by the National Guard. There’s just much more significant state repression. Again, hundreds of incidents of arson, much more protest or initiative violence, that because a single event that’s captured on cell phone camera becomes fodder for media, I suspect there will also be significant national effects of some of these events as well. But what we are also seeing is that it’s just playing out in a very partisan way. And so it may be that it is national, but for a subgroup for whom that is a confirmatory story and conversely, the story about peaceful protest and a claim for rights is confirmatory for the Biden supporters and that’s the national narrative for them. It seems like that coalition right now is the larger one and is the one carrying the day. But I agree that there’s almost certainly a national dimension to how these local events are being mobilized in politics.
Matt Grossmann: Protest is also having organizational and mobilization effects outside of presidential politics.
Omar Wasow: Another dimension of how protests can have long run effects outside of electoral politics is just producing some kind of civic capacity. There are organizations that arise. There are networks that form. There are essentially both informal and formal structures that allow people to mobilize more quickly or to kind of advocate down the road around certain issues. What we’ve seen in some of the earlier BLM protests from 2013 to 2017 is that even after the protests kind of stop being quite as active, there were campaigns for district attorney in places like Chicago and Philadelphia, where reform minded district attorney is one and that became part of a larger kind of criminal justice reform effort. And so I think it’s really important not just to focus on presidential politics and it’s really important not just to focus on outcomes as electoral. Though clearly I think those matter because there are these sort of more local kinds of effects, both on the individual and in a community that can be downstream consequences of protest mobilization.
Matt Grossmann: So where do we go from here? Coggins will be telling the story of how we got to Trump based on voters symbolic attachments.
Elizabeth Coggins: I have so many notes all over my office of many ideas that relate to this particular project. And in a sense I’m glad I didn’t finish this book right after I wrote my dissertation, because even though I do think the longitudinal sense is really important, I do believe that what we have seen over the past six years since I’ve been at Colorado College is critical to telling this whole story. Because when you look back, you can tell it was about to happen. And what it means for these attachments to ideological labels, it has done nothing but sort of make more whole and kind of enrich what I understand these attachments mean to people now. I come from the deep South where conservatism is a way of life. It is a gut level attachment that people have at least in my hometown.
And I have spent a lot of time thinking and a lot of time talking to people from there, including many members of my own family, trying to get a better grasp on not only sort of over time change and over time feelings, but really the past four years. And I think that this research is really enriched by that understanding. But that being said, we’re working there’s some really interesting shifts that happened since we wrote this paper. I think we’ll see a lot of the same, but I think we also have the chance now to do some new things with media analysis especially and strengthen that story of how much the media matters in this story of ideological identifications. But yes, so much of what you have asked has made me write another sticky note and put another sticky note here and there.
Matt Grossmann: And Wasow will be telling the story of how backlash is not inevitable from policy, but from the decisions of activists.
Omar Wasow: In a book project I’m working on now, I’m trying to think about some of the longer term effects of this and some of the kind of how does this maybe spill over into other issues. And there are a couple of things that become clear. One is that it’s easy to forget that in 1964 the Civil Rights Act has passed before Johnson is running. And so there’s an opportunity if you are opposed to civil rights for African-Americans to defect to Goldwater. And we see that happens particularly in the South, but there’s not a national groundswell in opposition to civil rights. And there are other things going on there. Kennedy’s been assassinated. So I’m not trying to say that’s the only factor, but both in the just kind of raw data and in looking at American National Election Study data, what predicts people’s switching parties, I don’t find civil rights is actually a very good predictor. Attitudes about civil rights is a good predictor of whether you switch parties.
What is a good predictor and actually importantly another critique that people often offer of my work is to say, “Well, what about the Vietnam war?” And I find that the Vietnam war is not a very good predictor of switching parties. But attitudes about unrest are very good predictors of whether you’re going to switch parties. And so I think there is a process which other scholars that you’ve had on this show have pointed out that there is this kind of long realignment that’s happening, but in the work I’m looking at, this really does accelerate the degree to which there are certain set of voters who are actually not opposed to civil rights. It’s not a priority for them, but it’s also not something that they stay in the democratic coalition in ’64 and then they defect in ’68. And they defect largely in significant part as a function of attitudes about unrest, about order, and not the Vietnam war, not civil rights.
And so that’s I think one really important lesson is that this is helping to kind of structure the party system for what becomes as other work by Harrington and Weiler point out later, that the parties begin to sort on authoritarianism. And I think this is sort of where that starts to get set in motion, that one party becomes the order party. One is the more egalitarian party. But that this proves to be a kind of critical juncture. And that in a moment where we are seeing nationwide protests against excesses of policing, against the kind of tough-on-crime strategies that militarized policing and other kinds of over-policing, that cleavage and kind of consolidation of a more authoritarian party has these really deleterious effects for African-Americans. And that’s exactly contrary to the kinds of things that more militant protesters wanted.
Matt Grossmann: There’s a lot more to learn. The Science of Politics is available biweekly from the Niskanen center. I’m your host, Matt Grossmann. Please check out our recent episodes at niskanencenter.org or anywhere you find your podcasts. Thanks to Omar Wasow and Elizabeth Coggins for joining me. Please check out agenda seating and then listen in next time.
Photo Credit: The White House from Washington, DC / Public domain