Science AMA Series: Hi, we're political scientists who study if (and how) people can be persuaded on political topics. Ask us anything!

Abstract

Hi reddit!

In advance of this year's national election, AAAS is bringing together scientists who have studied how people make up their minds about political issues and, once their opinions are set, how people can change their views.

Science Magazine has published a few articles on this topic in 2016. One paper, by Noah Friedkin, explored the question "how do some beliefs within groups persist in the face of social pressure, whereas others change and, by changing, influence a cascade of other beliefs?" Another written by two of us, David Broockman and Joshua Kalla, describes our field experiment that showed that 1 in 10 Miami voters shifted their attitudes toward transgender individuals and maintained those changed positions for 3 months.

We are joined by Drs. Samara Klar and Yanna Krupnikov, authors of "Independent Politics: How American Disdain for Parties Leads to Political Inaction."

In the final weekend before the election, we suspect that many family and friends will be speaking about issues that are important to them. Ask us anything on the science of political persuasion!

Dr. David Broockman is Assistant Professor of Political Economy, Graduate School of Business, Stanford University

Joshua Kalla is a PhD Candidate in Political Science at University of California, Berkeley

Dr. Samara Klar is Assistant Professor of Political Science at University of Arizona.

Dr. Yanna Krupnikov is an Associate Professor in the Department of Political Science at Stony Brook University.

We’ll be back at noon EST (9 am PST, 4 pm UTC) to answer your questions, ask us anything!

Hi everyone, thanks for the AMA!

As I understand, a lot of your work involves in-person interaction, but a lot of interactions these days also occur online (including on sites like this one). When users can customize their own experience, they seem to cluster in self-reinforcing groups on social media some have called echo-chambers. How concerned are you with this echo-chamber effect, where many people are less exposed to ideas that challenge their beliefs? Does it make persuasion harder? If so what changes might bridge the gap?

I realize I'm also pre-supposing that this effect is inherently bad, so related question: Are there benefits to this phenomenon? Any silver linings?

Thanks!

superhelical

Samara Klar (SK): Thanks for your questions!

On one hand, you’re right. Individuals on social media do tend to cluster into like-minded groups. For eg. Conover et al. show a nice example of this on Twitter

But, on the other hand, social media also allow for access to what network scientists refer to as weak links -- casual acquaintances, for example, who connect you to more disparate parts of the network. Scholars at Facebook have shown that weak links actually drive the majority of information exchange on Facebook. Due to this access to weak ties, some scholars are arguing that the “echo chamber” effect on social media has been overstated. For example, Pablo Barbera and his colleagues find evidence of ideologically diverse communication.

In research I’ve done with Yotam Shmargad at the University of Arizona’s Center for Digital Society and Data Studies (forthcoming at the Journal of Politics), we find that the structure of on-line networks can actually facilitate exposure to diverse information and can even result in attitude shifts. Specifically, networks that contain more weak ties who are connected to more distant regions of the network allow individuals to be exposed to under-represented viewpoints and, subsequently, those individuals become more sympathetic to those viewpoints. So one way to combat the “echo chamber effect” that you speak of is for individuals to maintain structurally diverse networks and for social media platform to encourage these types of links.


Which action is more effective at changing the mind of voters on political topics:

  1. Watching the candidates debate directly about the topics

or 2. Reading debates between complete strangers on the internet about the topics?

adamento

DB: My favorite study on this is an oldie-but-goodie by Alan Abramowitz http://www.jstor.org/stable/2110467. The study finds that when people watched a Presidential debate, they just updated their views on the issues to match whatever their favored candidate said, and never actually changed their views on the candidates. So, on the one hand, debates between candidates seem to be effective at changing people’s views on issues -- but not their views on the candidates! This is consistent with some other research not on debates by myself and by Gabe Lenz (see http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.175.8385&rep=rep1&type=pdf and http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.711.8724&rep=rep1&type=pdf) that when people learn what politicians think about issues, they are much more likely to just start parroting whatever the politician says than to change their view on the issue. In a certain way, it all suggests that what people think on issues doesn’t matter as much as we like to think. More generally, there’s actually been a fair amount of research where people are experimentally encouraged to watch debates, or randomly assigned to have access to them. Here are a few of my favorites:

You also asked about whether interactions between strangers might change people’s views. There, I think there is some reason to be optimistic. Here’s my favorite study on this, where people were randomly assigned to tables at a “deliberation” event were more likely to agree with what people at their table said after the event: https://research.hks.harvard.edu/publications/getFile.aspx?Id=986. And our research on door-to-door canvassing (as well as other research on the same) has likewise found that when a stranger knocks on your door, they can change your mind.


Are some people inherently more open to changing their views, and if so, what causes this?

thefizzingwhizzbee

Yanna Krupnikov (YK):Much of the research in political science focuses on the idea that people aren’t likely to change their minds Even though many scholars (see for example Lodge and Taber’s book) find this in many contexts, but not all people are immune to persuasion.

First – and most obviously – people who have weak opinions are most open to new information and changing views. But its more than just a weak opinion. In their book, Hillygus and Shields show that people who have positions that are in some way unusual for their partisan group (so, for example Republicans who hold a liberal position on one issue or Democrats who hold a conservative position on one issue) are also more likely to change their minds.

Your question is about individual characteristics, but there is also the question of context. Some people are more or less open to persuasion depending on the context they are in. In my own research with Eric Groenendyk, I've shown that once you put someone into a very combative, political context, their “shields” go up and they become much more likely to dismiss a lot of information. In contrast, you may find people to be more open to political information in a situation that is less combative and less political.

Similarly, Bashir’s work in psychology suggests that people are resist \ information transmitted by those who call themselves “activists.”

So, what this may suggest people are more likely to resist information from those who have a long history of trying to persuade them or a long history of combative behavior but may be more welcoming of new information from those that they do not necessarily view as having purely political goals.

Of course, some people are closed to persuasion no matter what, but for others context/persuader may play a key role.


Thanks for doing an AMA! Is there any research on whether people are more or less likely to change their beliefs when confronted with a person outside their "group." (For example, someone against gay marriage having a discussion with someone who is gay).

yrfavtrash

DB: There is a large research literature in this, mostly in psychology, on this. The hypothesis is called the “contact hypothesis.” There’s a great review of this and other research on prejudice reduction by Paluck and Green here: http://www.cscc.edu/about/faculty/pd/files/Paluck%20-%20Prejudice%20Reduction.pdf. My view, which is similar to theirs, is that there is much less evidence on this question than many assume. It’s difficult to just look out into the world and see whether people who have, e.g., gay friends, are more supportive of gay marriage. That might just be because people who are okay with gay people are more likely to become their friends in the first place! There’s hundreds of psychology lab experiments on the topic, where undergraduates have a brief experience in the basement of a psychology department and claim their mind has changed a few minutes or days later. But research in the field - where people actually meet and get to know members of an outgroup - is a lot more rare.

There are a few studies, though, that I think are particularly interesting on this:

There’s also a couple studies that I think should make one less confident in the contact hypothesis:

  • Ryan Enos did an ingenious experiment (http://www.pnas.org/content/111/10/3699.full) where he randomly assigned some people riding trains into Boston to have some Latinx riders join them for the ride, as if they had moved to their neighborhood. He finds that people on these trains became more anti-immigration. This suggests contact might even have the opposite effect, producing threat and feelings of hostility, even when there isn’t any kind of “competition” going on. For a real-world example of this effect in action, check out this by the WSJ: http://www.wsj.com/articles/places-most-unsettled-by-rapid-demographic-change-go-for-donald-trump-1478010940.
  • In our study in Science we found that non-transgender canvassers could be effective. So even if contact with outgroups works sometimes, I am pretty sure it is not the only tactic that can work.

As you can tell, I think how to reduce prejudice is a fascinating question - and one where we don’t have nearly enough clear answers and what works, why, and when.


I recently had an online discussion where I was told "I don't believe 'facts'. I know what I know and won't ever change." On what basis do people form ideas and opinions if they are consciously anti-fact? What are other types of input and motivators that are being employed and how can they be used to sway an opinion?

copopeJ

SK: The truth is  that most Americans, no matter how “sophisticated” or interested in politics you might, simply do not have the time or the resources to gather all the facts. While we might have opinions on a particular legislative policy, for example, we probably have not read the hundred-page policy that circulated Congress.

Instead, we rely to a large degree on heuristics or short-cuts. And, in fact, scholars have shown that short-cuts can be extremely helpful in reaching the same decisions that we would have  reached had we actually read all of the hundreds of pages of pertinent information. Michael Bang Petersen suggests that heuristics are  “evolved, biological adaptations” that help us to navigate the world around us.

This, of course, does not mean that individuals can remain completely oblivious to the facts and still expect to make informed decisions, but rather than shortcuts and cues in our day-to-day life often fill in the gaps we need to make a reasoned choice. For example, Lupia and McCubbins write in their book “The Democratic Dilemma” that “reasoned choice does not require full information or unlimited attention. Instead, reasoned choice requires information that generates accurate predictions about the choice.” And some scholars, for example David Redlawk and Rick Lau, show that heuristics might benefit more informed voters.

Examples of heuristics include party endorsements and expert endorsements, emotions, and candidate traits (for example, gender which Monika McDermott shows can have adverse effects).


I recently had an online discussion where I was told "I don't believe 'facts'. I know what I know and won't ever change." On what basis do people form ideas and opinions if they are consciously anti-fact? What are other types of input and motivators that are being employed and how can they be used to sway an opinion?

copopeJ

JK: Brendan Nyhan, a political scientist at Dartmouth, has conducted a number of studies on the role of conspiracy theories and misinformation in American politics. You can check out his research at (http://www.dartmouth.edu/~nyhan/). Nyhan has found that fact checking can reduce the likelihood that state legislators will make untrue claims (http://www.dartmouth.edu/~nyhan/fact-checking-elites.pdf) but that among voters, sometimes correcting facts is effective, while other times it may backfire.


I recently had an online discussion where I was told "I don't believe 'facts'. I know what I know and won't ever change." On what basis do people form ideas and opinions if they are consciously anti-fact? What are other types of input and motivators that are being employed and how can they be used to sway an opinion?

copopeJ

SK: The truth is  that most Americans, no matter how “sophisticated” or interested in politics you might, simply do not have the time or the resources to gather all the facts. While we might have opinions on a particular legislative policy, for example, we probably have not read the hundred-page policy that circulated Congress.

Instead, we rely to a large degree on heuristics or short-cuts. And, in fact, scholars have shown that short-cuts can be extremely helpful in reaching the same decisions that we would have  reached had we actually read all of the hundreds of pages of pertinent information. Michael Bang Petersen suggests that heuristics are  “evolved, biological adaptations” that help us to navigate the world around us.

This, of course, does not mean that individuals can remain completely oblivious to the facts and still expect to make informed decisions, but rather than shortcuts and cues in our day-to-day life often fill in the gaps we need to make a reasoned choice. For example, Lupia and McCubbins write in their book “The Democratic Dilemma” that “reasoned choice does not require full information or unlimited attention. Instead, reasoned choice requires information that generates accurate predictions about the choice.” And some scholars, for example David Redlawk and Rick Lau, show that heuristics might benefit more informed voters.

Examples of heuristics include party endorsements and expert endorsements, emotions, and candidate traits (for example, gender which Monika McDermott shows can have adverse effects).


I'm curious about the value/purpose of the standard debate in politics. I've always wondered why we force candidates into a situation where they end up having to aggressively talk over each other in order to save face in front of the nation and nothing is ultimately gained. Wouldn't it make more sense to interview them individually and perhaps ask a series of relevant skill testing questions, their answers to which the nation could use to judge their merit? Is it because that wouldn't be interesting enough to draw attention?

oskiwiiwii

YK: The value of the debate for many political scientists is questionable. Given that people (as a number of questions in this AMA have rightly suggested!) are skilled at dismissing and ignoring the information that they disagree with, it is difficult to imagine that people are going to be persuaded by anything that happens at the debate. Some scholars have seen some post-debate shifts among certain groups of voters (for example, Hillygus and Jackman’s 2003 piece on decision-making in the 2000) election, but arguably responses to the debate may be based on more than just the facts discussed (Jamie Druckman shows this in his paper on debate winners): There is also some research that suggests that the actual debate – candidates arguing and fighting – may lead people to either mistrust or retreat from politics. In some of her work Diana Mutz shows that political debates – especially if they are “uncivil” – can have negative effects on people.

In our book, Samara Klar and I suggest that seeing argumentative politics makes people want to hide their partisanship and pretend to be independents.

So, I think some of this points to the idea that debates aren’t actually all that helpful.

On the other hand, its possible that debates are “exciting” – which makes more people tune in just to see the spectacle, which helps people learn something about politics. So, this may be a silver lining of the combative debate.


Hey, thanks for this AMA, it's a very interesting topic.

I am wondering if you have found anything in your research about what is most effective when trying to persuade someone to your side of the political argument? I am the type to always use facts and reasoning that can be cited and proven but I have heard that emotional reasoning can be more effective. Have you found anything to support either tactic?

coleslaw74

JK: A brand-new book by psychologist Robert Cialdini (https://www.amazon.com/Pre-Suasion-Revolutionary-Way-Influence-Persuade-ebook/dp/B01C36E2YS) looks to general principles of persuasion - from business, psychology, politics, and other fields. Cialdini argues that effective persuasion is less about “altering a listener’s attitudes, beliefs, or experiences” than changing a listener’s “state of mind”.

There is certainly a lot of research in political science that demonstrates the effectiveness of emotional appeals. For example, see Ted Brader’s work (http://press.uchicago.edu/ucp/books/book/chicago/C/bo3640346.html).

Enthusiasm, for example, tends to reinforce individuals’ pre-existing opinions while anxiety has somewhat of the opposite effect, moving individuals to reevaluate their options. In Samara’s work, she finds that threat can drive individuals to support policies that actually counteract their own party, when those preferences line up with the individual’s other identity groups (http://faculty.wcas.northwestern.edu/~jnd260/pub/Klar%202013.pdf). Emotions can be an effective way to persuade.


Thanks for coming!

In my opinion the polarization of people in our bipartisan system has lead to a lot of voters who are unable to objectively consider the information being directed at them by both parties. The "us vs them" mentality of social identity theory and confirmation bias lead to beliefs in objectively absurd things even passing consideration and research could prove or disprove. Would this be a statement you agree with and if so what other factors contribute to this issue and do you see a way forward with a more productive dialog? If you disagree with it, what do you think the causes of the animosity are and what could be done to lessen the party animosity to increase intelligent voting and discussion in the future?

shadus

SK: Thanks for the question! Americans are deeply committed to their party identification and many equate party ID to a type of social identity. (Examples here and here just to name a couple of many.) As the question asker, states, social identity theory tells us that individuals view their in-group as a reflection of themselves and are thus motivated to evaluate their in-group positively and out-groups negatively. This in-group vs. out-group in politics can indeed lead Americans to unduly derogate the other party.

There are instances, though, where partisan adversaries can become less, for lack of a better word, adversarial. The Common In-Group Identity Model suggests that cueing a superordinate identity can unite rival groups. For example, Dr. Matthew Levendusky at the University of Pennsylvania finds that priming Democrats and Republicans to think of themselves all as Americans can unite the two groups. In ongoing research with Yanna Krupnikov and John Barry Ryan, I am finding that Americans are more tolerant of members of the opposing party when those individuals are not politically active. In other work I've done, I have found that bringing Democrats and Republicans together in face-to-face groups to allow for inter-personal discussion actually leads to more moderate and bipartisan policy preferences. So there are certainly moderating circumstance that can temper group rivalry between the two parties, at least within the mass public.


In your study about reducing transphobia, your field experiment method involved going door-to-door and having 10 minute conversations. This had people take the perspective of someone else to great effect - you saw changes in attitudes towards trans* people that lasted 3 months. Specifically (for those who don't have access):

Canvassers first asked each voter to talk about a time when they themselves were judged negatively for being different. The canvassers then encouraged voters to see how their own experience offered a window into transgender people’s experiences, hoping to facilitate voters’ ability to take transgender people’s perspectives. The intervention ended with another attempt to encourage active processing by asking voters to describe if and how the exercise changed their mind.

And the impact was impressive:

Before the intervention, the treatment and placebo groups scored similarly on this index (see tables S13 to S17). After the intervention, the treatment group was considerably more accepting of transgender people than the placebo group (t = 4.03; P < 0.001). These effects are substantively large: These brief conversations increased positivity toward transgender people, as measured with a survey tool called a “feeling thermometer” (20, 21), by ~10 points, an amount larger than the average increase in positive affect toward gay men and lesbians among Americans between 1998 and 2012.

In some ways, this seems like a fairly simple method and one that could potentially be taught. So I have two questions:

1) Can this be adopted for personal use? If so, what tips would you give people who are trying to have difficult conversations with loved ones and co-workers about sensitive issues? In other words, how can we employ this method in our own lives - what scripts, approaches, and cautions would you give for getting conversants to engage in active processing and stepping into someone else's shoes?

2) Door-to-door might be effective but it is time consuming and not possible for people to have a 10 minute conversation with everyone in America about every controversial issue. Do you see social media or online games being an effective substitute? Does face-to-face matter? This is something I'm particularly interested in because I study new media engagements.

firedrops

Josh Kalla (JK)

1) None of our research can directly speak to the applicability our the conversations in our field experiment (delivered by strangers in Miami) to a setting involving loved ones or co-workers. Nevertheless, there might be some more general practices, such as not being judgemental during the conversation and working hard to get the other part to actively talk and participate. All of the scripts used in the experiment are available at (https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/WKR39N). There is also a sample video of a real conversation available at (https://www.youtube.com/watch?v=_tdjtFRdbAo). Hope this helps!

2) I’m not aware of any experiments looking at whether social media or online games can be effective at changing minds on controversial topics. In a review of the research on prejudice reduction (http://www.cscc.edu/about/faculty/pd/files/Paluck%20-%20Prejudice%20Reduction.pdf), Betsy Levy Paluck and Don Green found that “Few have gauged the impact of media on large audiences or the impact of large-scale media campaigns (which span long periods of time or multiple theatres, cable networks, or airwaves).”


What is the biggest factor that makes people stick by a political belief in the face of overwhelming evidence that that belief may be misguided or even completely untrue?

kt-bug17

YK: There are number of factors that lead people to stick by their political beliefs even in the face of facts. People often rely on confirmatory biases, which means that they are really good at dismissing and discrediting facts that are contrary to their views. Even when the information seems totally factual, people still can find a way to ignore it. But possibly the biggest factor may be when the incorrect fact is pivotal to someone’s worldview. So, if their opinion is based on misinformation, the misinformation becomes a key component of an opinion. As Brendan Nyhan and Jason Reifler show once the misinformation is part of an opinion, it is very hard to correct it.


Hey! Thanks for taking the time!

I'm greatly concerned by confirmation bias and associated psychological traits—how people tend too manipulate their perception of information to fit their pre-existing conclusions, rather than be willing to modify their conclusions in light of new evidence, and how providing contradictory evidence can actually cause a person to double down on their initial position. It strikes me as disturbing, even dangerous, for our democracy for large segments of our population to make up their minds about an issue and then be essentially impervious to evidence that they are wrong. Persistent disbelief in anthropogenic climate change is perhaps the best example, but you also see it in anti-vaxxers, and in many, many other areas.

From what I've read, this seems like an insurmountable psychological roadblock, but I have to believe there is a way to reach people. Is there any way to defeat confirmation bias and get a person to consider and accept evidence that challenges pre-existing deeply held beliefs?

Bird_nostrils

SK: Thanks for the question. Confirmation bias can indeed seem like a normatively troubling psychological phenomenon: individuals appear to avoid information with which they disagree and unfairly evaluate it when they are forced to confront it. But, in fact, Robert Huckfeldt, Mendez, and Osborn show considerable variance in the degree to which our discussion networks are ideologically homogeneous or diverse. And, moreover, recent research suggests that we might not have as much control over our information environments as we like to think we do.. For example, Kathy Cramer as UW-Madison finds in her book Talking About Politics that political conversations tend to arise in casual settings with informal acquaintances. Wojcieszak and Mutz find that politics will come up in online message boards that were formed to discuss non-political things.

When individuals actually do confront political disagreeable information, there is evidence that it can, in fact, be persuasive. I find that when Democrats and Republicans come together in small discussion groups, their opinions actually moderate  and move toward more bipartisan preferences. Working with social network data, Pablo Barbera and his colleagues show that even social media networks are less ideologically homogeneous than previous work has suggested.

This is not to say that Americans do not prefer to discuss politics with like-minded thinkers -- indeed they do. But it turns out that we’re all forced to confront opposing perspective more often than we realize.


Is there any way, or any known method that can be utilised, that would encourage politicians to take a more positive approach to campaigning. Looking at some of the most recent campaigns (Brexit and the American Presidential campaign), a common theme seems to be 'Project Fear' and marginalising the opposition rather than promoting your own policies/agenda. This also seems to be a situation that is worsening as time goes by, though this may be anecdotal.

It feels like this is a recent phenomenon, however is this new or has this actually always been the case to this extent?

ParanoidQ

YK: There is some evidence -- at least in American politics -- to suggest that political campaigns have become more negative over time (for example John Geer’s book). But, there is some reason to believe that the increase in actual negative campaigning is magnified due to news coverage of negative campaigning -- negative ads are the ones that are most likely to be covered by the news (so they are most likely to have a life beyond the original ad play, as Geer shows. The fact that negative ads get the most attention is possibly the reason why politicians are more likely to rely on these ads. But another reason is that negativity is a way that candidates who are challenging long-term incumbents can actually get some attention during an election. It is very hard to unseat an incumbent in Congress -- and a negative campaign is really one of the only strategies that a challenger has.

Which is to say, the incentives to rely on positive campaigns are often quite low.


I would love to hear if there is any research on why politicians lie. I'm not talking about denial of a scandal, or twisting of facts, but a statement that can, and probably will be, proven objectively false. Does research support this being an effective strategy?

chancho21

DB: There actually is some research on whether this is an effective strategy! From this paper by Todd Rogers:

listeners can fail to detect dodges when speakers answer similar – but objectively incorrect – questions (the “artful dodge”), a detection failure that went hand-in-hand with a failure to rate dodgers more negatively. We propose that dodges go undetected because listeners’ attention is not usually directed at a dodge detection goal (Is this person answering the question?) but rather towards a social evaluation goal (Do I like this person?)


Isn't this more or less political psychology? And isn't your answer going to be obvious? People can be persuaded about a variety of subjects, my question is: What makes political opinions different in terms of persuasion than other subjects?

PoliSciNerd24

What makes political opinions different in terms of persuasion than other subjects?

DB: I think this is a great question. I'll take a quick speculative stab at this from two points of view.

First, when we disagree about political issues we are often engaged in a defense of our identities -- either identities as partisans (e.g., "I'm a Democrat!") or as some group affected by an issue. In these circumstances, people are especially likely to get defensive, embrace evidence that confirms their point of view, dismiss the evidence that doesn't, etc. It's sort of like arguing about which sports team is better with a fan of another team. I really like the work Dan Kahan has been doing on this, which is summarized here: http://www.vox.com/2014/4/6/5556462/brain-dead-how-politics-makes-us-stupid. One reason I mention it is that he explicitly compares how people make similar decisions when they think the decision is about politics versus not, and finds that as soon as you introduce a political aspect to a similar decision task, people start behaving much less... rationally.

Second, and much more speculatively, I think part of what is interesting about political persuasion is that individuals have very little incentive to get the "right" answer on political questions. When we have to form opinions about things like whether a restaurant will give us food poisoning or which repair shop is going to give us the best deal on repairing our car, we think much more carefully because there are personal consequences to us to getting it wrong. But when it comes to political issues, each of us individually doesn't pay a price when we get the answer wrong. Although, of course, we do pay the price as a society. This "collective action problem" where no one has an incentive to think carefully but how everyone thinks adds up to something that matters is, I think, one of the things that makes studying this topic most interesting to me. It's also one of the reasons I've tended to be most interested in studying relatively "high touch" political activities like door-to-door canvassing, where another person can force you to think a little harder about something than you typically would by asking you questions, etc.


In discussing the political spectrum, we usually talk about far left, far right, and moderate people in the middle. We assume that only the moderates can be persuaded one way or the other. Is it reasonable to expect for people at the far ends of the political spectrum to change their minds or vote for the other party's candidate? Would you say that the far left and far right have more in common with each other or their respective moderates?

helpppppppppppp

DB: The word "moderate" has a lot of different meanings, depending on who you ask. In some of my work I've argued that a lot of people we think of as moderate are actually more extreme than the people we typically think of on the far left and the far right.

The idea is this. Quite often when we try to figure out how liberal or conservative someone is in general, we "average over" their views on many different topics. By this measure, you see, no surprise, that Republicans are pretty conservative (they have conservative opinions on most things) and Democrats are pretty liberal (they have liberal opinions on most things). However, most people have a mix of liberal and conservative opinions -- e.g., maybe they oppose gay marriage but favor taxing the rich (or support gay marriage but oppose taxing the rich). These are the "moderates." What I show in my work is that these people are actually less likely to have moderate opinions on the individual issues. So, when it comes to taxes, for example, "moderates" who have some liberal views and some conservative views are actually more likely to favor extreme ideas like a maximum national income or no taxes whatsoever.

The intuitive reason for this is that hard-core Democrats adopt the Democratic position on most issues and hard-core Republicans adopt the Republican position on most issues. By definition, these positions are within the mainstream because many people share them. But moderates are generally less engaged with politics and less likely to adopt the parties' positions as they own -- which leaves them both with a mix of liberal and conservative positions and with positions that aren't within the mainstream of American politics.


In discussing the political spectrum, we usually talk about far left, far right, and moderate people in the middle. We assume that only the moderates can be persuaded one way or the other. Is it reasonable to expect for people at the far ends of the political spectrum to change their minds or vote for the other party's candidate? Would you say that the far left and far right have more in common with each other or their respective moderates?

helpppppppppppp

YK: I am going to start with the far left and far right, and then one of my colleagues may chime on on the moderates. I think that your suggestion that the far left and far right may have something in common is an important one. If we do not consider political attitudes and beliefs and just focus on the extent to which someone can be persuaded, then I think the two extreme ends of the ideological spectrum are equally resistant to persuasion. The far left and far right may view the world very differently, but they are similar in that their political beliefs are very strong and enable them to dismiss new information. If resistance to new information is about being able to dismiss and counter-argue the ideas of the other side, then being far left and far right gives a person a strong foundation to do so.

Now the second point here is whether the moderates are different. This is in some ways a more difficult question to answer, in part because it is possible that the people who report they are moderate may not actually be moderate. In our work, Samara Klar and I have studied political independents -- a different group than moderates, but also a group that tries to take itself out of either “side” of politics -- and we find that many independents aren’t at all independent, but are actually quite partisan. It is possible that the same works for moderates -- at least some moderates may actually have left/right preferences, but prefer to call themselves moderates. This, of course, would affect the extent to which they are persuadable.


It seems that political candidates lately have been trying to win voters over by playing on their emotions rather than facts (some blatant examples of this being the Brexit Leave campaign's "people are sick of experts" and Newt Gingrich's insistence that people don't feel safer). Do you find that generally this actually is the most effective way to get people to change their minds on something? Is there anything people who tend to argue purely from the facts can do to win over people susceptible to this?

senshisentou

Samara Klar (SK): There is certainly a lot of research in political science that demonstrates the effectiveness of emotional appeals, as well the distinct effects of different emotions. For example, see Ted Brader’s work. Enthusiasm, for example, tends to reinforce individuals’ pre-existing opinions while anxiety has somewhat of the opposite effect, moving individuals to reevaluate their options. In my own work, I find that threat can drive individuals to support policies that actually counteract their own party, when those preferences line up with the individual’s other identity groups.

Shana Gadarian and Bethany Albertson have done really interesting work on anxiety, in their new book Anxious Politics, showing that anxiety has pronounced effects on the public’s views on government and the manner with which they engage with politics.

So, in sum, yes, emotions are absolutely an effective way to persuade. Facts that incorporate emotional cues need not be manipulative or dishonest -- but the emotional cues can help to make a persuasive case.


One thing I am interested in how parents affect the political beliefs of their offspring I have seen everything from they completely agree to they outright disagree with everything. What are the factors behind this?

jtrook

YK: There is some research on this in political science, using data that tracks families over time. For example, Jennings, Stoker and Bowers use data from different time points to consider how parents influence children. Their results suggests that parents can have an influence -- but this is most likely to happen when parents are vocal about their politics. In other words, to be influential parents have to give their children cues about politics.

The Jennings paper, however, also raises some interesting issues. For example, they suggest that it is possible that parents who are more political lead their children to be more attentive to politics outside the home, which may in turn actually limit the influence of parents. They also raise the possibility that parents may be even more influential if they continue to be similar (for example, in terms of income and education) with their adult children (though this remains an open question in their paper).

But these types of findings about parents and children, I think, speak more broadly about the power of social influence in politics. For example, Lazer and co-authors find that people in an educational setting (a cohor of MA students) eventually adjusted their political positions to be closer to people in their social networks. In other words, children may have similar positions to their parents because their parents are pivotal political speakers in their social networks. Once their social networks broaden, the influence of parents may decline.


One thing I am interested in how parents affect the political beliefs of their offspring I have seen everything from they completely agree to they outright disagree with everything. What are the factors behind this?

jtrook

JK: We're running short on time, but I would suggest checking out this paper (http://www.jakebowers.org/PAPERS/JenStokBow2009.pdf).


Hi, thanks so much for doing this AMA! As a nineteen-year-old, I feel like I haven't really been around during a time when bipartisan political cooperation has been a mainstay of the political landscape. Do you believe things will change for the better? What do you think has caused this shift towards extremely partisan, divided political parties and factions?

smoke_it_all

DB: It's difficult to predict the future, but my guess would be that partisan divisions will not decrease any time soon. My favorite theory of this topic is by Frances Lee, who argues in two books (http://press.uchicago.edu/ucp/books/book/chicago/B/bo8158910.html and http://www.press.uchicago.edu/ucp/books/book/chicago/I/bo24732099.html) that a lot of what has happened over the last few decades is that US elections have grown more competitive. To vastly oversimplify, if you were an elected official in Congress in the 1960s, there wasn't much hope that your party would win control of Congress if you were in the minority or that you'd lose control of Congress if you were in the majority. But today control of Congress is constantly up for grabs, as is control of the Presidency. If you're a Democratic elected official when Bush is President or a Republican elected official when Obama is President, anything you allow the other party to get done and get credit for with voters thus might actually just lose you an election. Therefore, the incentive is to just not cooperate and trash whatever the other party does -- how else can you go to voters and tell them they should put you in charge instead? Basically all countries feature these bitter political divisions between different parties, with parties demonizing the other and criticizing their every decision.

With this said, I don't think that is inherently a bad thing. Competition is usually a force for good and encourages parties to be better. We want elected officials to have an incentive to be looking for mistakes the others make.

At the same time, we are probably seeing the consequences of exactly this in how voters are reacting to politicians -- people dislike the other party more than ever, and this is probably because politicians on both sides have an increasing incentive to blame the other side for everything.

But I think the real unfortunate thing is that in the United States we don't only have disciplined parties that give opposing parties the incentive to block what the other is doing -- we also have political institutions that give them the ability to do it. This leads to gridlock and, I worry, is going to lead to a constitutional crisis in the medium or long term.

But as a new voter, you totally have the ability to change this. Don't vote for politicians that engage in the behavior you're seeing and don't like -- and write to your elected officials telling them you don't like what they are doing. They really do listen, and there have been experiments to prove this. Voting and reaching out to your elected officials is just about one of the most powerful things you can do, other than running for office yourself.


Are political adds on TV effective? How do you quantify this effect?

JoeRmusiceater

DB: This question is extremely difficult to answer because political campaigns tend to spend more on TV when they need to get ahead -- so if you just looked at which campaigns spent more on TV, you might find that they’re doing worse than the rest! If you’re going to win in a landslide you don’t spend time raising money to buy ads with.

I think there have been a few persuasive pieces of research on this. The general finding is that, regardless of whether the “other side” is advertising or not, the persuasive effects of TV ads appear to decay -- you get an ad today and you’ll answer a survey differently today, but by later this week you’ll have forgotten all about it. This “recency bias” or “myopia” is a more general feature of public opinion, where political events cause some small change in opinion but that this change typically disappears as opinion returns to its previous baseline. We have certainly seen that this election cycle.

There is some new research on TV spending that I think is pretty interesting, this paper by Jorg Spenkuch: http://www.kellogg.northwestern.edu/faculty/spenkuch/research/advertising.pdf. They found a clever way to get around the problem that candidates who buy TV are different than the ones who don’t. The basic idea is that campaigns aren’t able to target TV ads on a person-by-person basis; they have to target entire geographies. Sometimes, you see some small pockets of swing states where, in order to buy TV ads, the campaign would have to spend a lot of money on ads in a market that includes lots of voters outside that swing state. Think a city in a swing state that’s just on a border with another state and gets TV designed for that other state. These people get a lot fewer ads. Looking at these people and comparing them with people who are barraged with ads, they find the people who are shown more ads for a candidate are more likely to vote for that candidate.

The really interesting twist is that the effect doesn’t appear to be driven by people changing their mind, but instead by motivating turnout -- e.g., if you’re an Obama supporter not sure if you’re going to bother to vote, seeing a lot of Obama ads makes you more likely to get off your butt and go to the polls. This seems to happen more often than the ads actually changing people’s minds. More generally, I think the possibility that a lot of campaign events are just changing enthusiasm to participate (in elections, and in polls) is one of the most fascinating nascent themes of recent work. Here’s a couple more pieces on this, outside of the scope of TV ads: https://dl.dropboxusercontent.com/u/11481940/Hall_Thompson_Base_Turnout.pdf and https://today.yougov.com/news/2016/11/01/beware-phantom-swings-why-dramatic-swings-in-the-p/.


What is the best approach to getting people to actually listen to, and consider, the words you say? It seems that so many people never change their minds because they never take in new information or any information that goes against/does not confirm their bias. Have you found a way to approach people that encourages openness and engagement, if not necessarily a change in viewpoint?

pamplemouss

YK: Just like your question suggests, people are often adept at ignoring information that is contrary to their existing political positions. Many scholars have tried exactly how to break through here, and make people more receptive to ideas from the other side.
There is some research to suggest that the key is to give people a different motivation. If people’s immediate tendency is to be motivated by a need to protect their own viewpoint, then -- the logic suggests -- maybe they can be given some other motivation when approaching new information. One type of motivation is the accuracy motivation: if people can be motivated to be accurate, they may be open to other ideas. Matthew Peitryka recently showed that giving people an incentive to be accurate in making decisions helped them seek out more ideas and opinions (although they didn’t exactly make better choices in the end, they were open to more information). Another approach is to more directly suggest to people that they should be more open-minded. Jason Barabas shows that telling people that their motivation -- explicitly -- is to be open minded, does lead people to be more open to other positions. In some of my own work, Eric Groenendyk have also shown that changing people’s motivations -- so taking them out of a setting in which they are motivated solely by defending their political position -- does encourage them to be more receptive to new information (though, it doesn’t lead people to change their minds). They key here, though, is taking people out of deliberately combative, political environments. Across these different studies, it seems that people are most motivated to dismiss new information when they believe that their context is combative and their goals are political.


Are there such things as core political beliefs and how often do you see these change? How many people typically shift from left to right as they get older and what are their reasons for doing so? When writing my dissertation on the growth of neoliberalism in Britain I read a lot about how its proponents sought to influence 'second-hand dealers in ideas' i.e. newspapers, broadcasters, teachers etc.; do you think this strategy of popularising political and economic ideas is still relevant in the digital age?

TheHalfwayHouses

JK: This is a great question! Going back to the seminal work of Philip Converse from 1964 [http://isites.harvard.edu/fs/docs/icb.topic1384513.files/converse.pdf], political scientists have found that many voters tend not to hold coherent, ideological beliefs (e.g., conservative vs. liberal). For most voters, there is no underlying structure to their beliefs but instead a mix of random opinions.

Instead, two overarching factors tend to explain much of what voters believe and how they vote: their partisanship and their group identities. In American politics, someone’s partisanship is often considered to be the “unmoved mover.” It is a fixed identity that influences their preferences on issues and candidates. Party identity tends to form relatively early and remain relatively stable throughout one’s life. An individual’s group identity - their race, religion, class - and how they see the parties’ relationship to those group identities, is another trait that strongly influences both an individual’s partisanship and their views on specific issues and candidates.

Generally, we see that individuals remain relatively stable in their partisanship over their lifetime. Yair Ghitza and Andrew Gelman compiled data on how birth year influences political views (http://www.nytimes.com/interactive/2014/07/08/upshot/how-the-year-you-were-born-influences-your-politics.html). For example, they find that a white born in 1941 is more likely to vote Republican throughout their entire life while a white born in 1952 is more likely to vote Democrat throughout their entire life. Getting older does not seem to cause shifts in how people vote. Instead, the political situation that existed in their formative years (ages 14-24) tend to drive how someone votes throughout their entire life.

When it comes to “dealers in ideas”, Hans Noel has done some fascinating work on the role of political thinkers in the Civil Rights Realignment - the period when the Democratic and Republican parties swapped positions on civil rights. He finds that “the reversal of the Democrats and Republicans in congressional voting is preceded by a similar reversal, several decades earlier, of liberals and conservatives in the intellectual sphere.” (http://faculty.georgetown.edu/hcn4/Downloads/Noel_CM09.pdf). How this applies today is an open and fascinating question.


Hi, thanks for this AMA, such an interesting topic!

Humans are fact based decision makers.

Could you comment on how true this statement might be, as well as how true we believe it to be?

benrules2

JK: We know that most voters lack full information. Regardless of how we measure political information, the mean level of political information in the electorate is low but there is also high variance. Most voters know very little; a few know a lot. Check out the book What Americans Know About Politics.

If voters just don’t know basic facts, then it is hard to argue that they are fact-based decision-makers.


Thanks for the AMA.

Is there a greater resistance to changing views, when the subject is aware of his/her party's stance, as opposed to when they are not?

I'm especially interested in free trade.Many people on both sides seem to be unaware of their party/candidate's stance, and generally uninformed about the topic.

corner-case

JK: I think you are absolutely right that many voters are unaware of where their party/candidate stand on many important policy issues. In Follow the Leader, Gabe Lenz shows that policy positions generally do no drive how individuals vote. Instead, he finds the opposite: individuals first choose which candidate they want to support, THEN adopt that candidate’s policy positions.


I've had the best success changing minds with the Socratic Method of asking questions and letting my debate partner (NOT adversary; do not treat them as such or you will get nowhere) answer the questions and come to new conclusions themselves.

I find this approach is effective because one feels a sense of autonomy in changing his views in this manner, rather than the traditional approach of "You're wrong; here's what's right". Such an approach deprives one of agency in the process of changing views and more often than not causes him to "lock up" in the face of opposition and double down on his initial views.

What is your opinion on the efficacy of using the Socratic Method as a debate tool, and do you have any quantified evidence exploring the efficacy of the Socratic Method as a debate tool?

WildBilll33t

DB: It's not exactly the same, but psychologists have a term called "self-persuasion" which captures the intuition that people are more likely to be persuaded when they "own" their new view by feeling that they've reached it themselves. Some of the studies showing this are kind of funny -- you ask people to give reasons that statements are true (e.g., "fire-fighters are more effective when they are risk-takers") and then inform them the statements are actually false. But if they've spent time voicing reasons why the statements are true, they're more likely to believe they are true later.

This is something the canvassers we studied in our Science article often did: asking people to elaborate the implications of what they were saying, hoping they'd reach the right conclusion themselves, instead of telling them what to think. I think it goes hand-in-hand with not making people defensive, as you say.


Thanks so much for doing the AMA!

I might be totally off base, but I've often heard the comparison that trying to convince someone away from political beliefs is a task similar to trying to convince someone out of their religion.

It was described to me as being that people often take a political party, associate themselves as a "[insert party label here]" agenda and use it as a prescription for their political positions in the same way someone would take a religion, identify themselves as being someone of that religion, use its tenets as a prescription for their moral positions.

It seems to be true in my interactions, as I've heard people describe someone as a "good [political label]" in the exact same way as I've heard people say it in religious context where it means that they follow the teachings precisely.

Is there any validity to this parallel? Is it common for people to use a political party as a prescription for their beliefs?

melodyze

JK: There is a lot of validity to this parallel. In their book Partisan Hearts andMinds, Green, Palmquist, and Schickler argue that political parties are akin to a deep social identity. The work of Jennings, Stoker, and Bowers (among many others!) find that partisanship is generally adopted early in life and remains constant through one’s adulthood. Just like religion, children are generally likely to adopt their parents’ political views, particularly in more politicized households.

It is certainly common for people to use a political party as a prescription for their beliefs. In Follow the Leader, Gabe Lenz shows that policy positions generally do no drive how individuals vote. Instead, he finds the opposite: individuals first choose which candidate they want to support, THEN adopt that candidate’s policy positions.

One issue that we think many Americans know a lot about is abortion. Thus it can serve as an interesting test for the role of policy positions in how individuals adopt their political parties. In a series of articles and a new book, Achen and Bartels ask: “Do people vote Republican because they are conservative on abortion? Or are they conservative on abortion because they are Republicans?” What is the causal direction between abortion attitudes among voters and their partisanship.

They find that for some set of voters, particularly women, you see a movement toward the party that matches their views on abortion. They find that, “Almost half of 1982 pro-choice non-Catholic Republicans had disappeared from the party by 1997.”

But on the other hand, you see a significant number of voters who remain with their party and adopt their party’s view on abortion. Achen and Bartels also find that “More than half of 1982 male pro-life Democrats had become pro-choice by 1997.” These voters changed their views on abortion to match their partisanship rather than the other way around. Even on the issue of abortion, you see a large number of voters who use their political party to determine their political beliefs. Thus Achen and Bartels conclude, “Most of the time, the voters are merely reaffrming their partisan and group identities at the polls. They do not reason very much or very often. What they do is rationalize.”


Thanks so much for doing the AMA!

I might be totally off base, but I've often heard the comparison that trying to convince someone away from political beliefs is a task similar to trying to convince someone out of their religion.

It was described to me as being that people often take a political party, associate themselves as a "[insert party label here]" agenda and use it as a prescription for their political positions in the same way someone would take a religion, identify themselves as being someone of that religion, use its tenets as a prescription for their moral positions.

It seems to be true in my interactions, as I've heard people describe someone as a "good [political label]" in the exact same way as I've heard people say it in religious context where it means that they follow the teachings precisely.

Is there any validity to this parallel? Is it common for people to use a political party as a prescription for their beliefs?

melodyze

SK: Party identification in the United States has been described for over 50 years as persistent and resistant to change. Indeed it appears that Americans identify with their party much in the same way that might identify with another social group (see, for eg, Green, Palmquist, and Schickler’s book on the topic).

This particular view of PID as a social identity is not a consensus in the field (other scholars, like Alan Abramowitz and Kyle Saunders argue that party ID is based more on political ideology than on social identity) but political scientists do largely agree that party identification is an important predictor of most political views.

This does not mean, though, that individuals are blindly following their party’s lead. Party identification can be a rational and useful heuristic we can use to judge policy -- if you usually tend to agree with one party’s candidates, then it may make sense to support their policies even if you haven’t heard too much about the specifics.


Engaging young people into the political process is a challenging proposition, and certainly in my home of Australia there exists a growing mistrust and cynicism of the political arena by young people. What are the best ways to engage and activate young people into the political process?

snuffleduff

JK: As the resident young person on this AMA, I’m happy to take this question! There have been a number of field experiments explicitly looking at effective tactics in increasing youth voter turnout.

Generally speaking, the tactics that work to engage older voters also tend to be effective at engaging younger voters. Door-to-door canvassing, high-quality phone calls, direct mail, and text messages have all been proven to cost-effectively increase youth voter turnout. On the other hand, tactics that you think might work particularly well for young people, namely email and online advertisements, have generally been totally ineffective. Just because young people may live online does not mean the online world is the best place to engage them in politics. Instead, the old-fashioned tactics, such as door-to-door canvassing, tend to work much better. For a summary of this research, I highly recommend Don Green and Alan Gerber’s book Get Out the Vote.

Another incredibly important piece in engaging young people is registering them to vote. Registering a young person gets them on the list of voters which opens them up to being contacted by campaigns. Voting tends to be habit-forming, so getting someone to register and vote when they turn 18 increases their likelihood to remain voters throughout their entire lives. Check out these two papers if you’re interested: (http://onlinelibrary.wiley.com/doi/10.1111/ajps.12210/abstract) and (https://sites.duke.edu/hillygus/files/2014/07/Preregistration-10.22.14.pdf).


Additional Assets

License

This article and its reviews are distributed under the terms of the Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and redistribution in any medium, provided that the original author and source are credited.