Levente Littvay on Contending with and Measuring Populism

Levente, you have been working on populism both in Europe and in the US. Let’s begin with your research on political psychology. You look at the sometimes conspiratorial nature of populism. Is there a particular bias of conspiratorial thinking when it comes to right-wing populism? Or are the conspiracy theories we’re seeing gain traction these days simply a symptom of the state of right-wing politics?

We are starting with a question that I am not at all a fan of, but you are certainly not the first to ask it. So, let’s unpack it a little bit and I’ll also explain why I am not a fan. Populism, as viewed from the ideational perspective I and most others in my field subscribe to, is a worldview wherein the pure people are in a Manichean struggle with a conspiring and conniving elite who exploit them. Given this, the connection between populism and conspiratorial thinking is almost definitional, but of course there is more to it. On the other hand, it is orthogonal, or unrelated, to ideology.

In fact, in our 2019 book Contemporary US populism in Comparative Perspective (with Kirk Hawkins) we conducted content analysis of candidate speeches from 2016, which clearly show that the most populist politician running for president of the United States was Vermont Senator Bernie Sanders—hardly a right-winger. This may come as a surprise to people, but if you really think about it, the three components that go into these speech analyses are people-centrism, anti-elitism, and the struggle between the two sides. Sanders is very strong on all these components; Donald Trump, not so much. Trump, especially at the time, was quite unable to praise the people (or anyone but himself, really). His anti-elitism was certainly matching that of Sanders, if not greater. But still, given that both components of people-centrism and anti-elitism are important for the presence of populism, Trump scored much lower than Sanders, based on their speeches.

Going beyond the US, not all populists are right-wingers. Barring new entrants like Brazilian President Jair Bolsonaro or former Peruvian President Alberto Fujimori’s daughter Keiko, who leads the Fuerza Popular party, Latin American populists have historically been very leftist. Just think of the late Venezuelan President Hugo Chávez, former Bolivian President Evo Morales, or current Mexican President Andrés Manuel López Obrador, to name but a few of the more recent successful ones. In Europe, especially in the South, Podemos in Spain and Syriza in Greece are clearly both populist and left-wing parties, while in Italy the 5-Star Movement is also populist, albeit quite ambiguous (or at least eclectic) in its ideological leanings.

So, what is the case with conspiracy theories? Sure, they are present among right-wing populists. But I sincerely do not believe that the left is immune to these. I cannot speak to the magnitude of the phenomenon without looking at data, but it is not that difficult to imagine leftist stereotypes (straw persons or Weberian ideal types) who, when you look at them closely, also subscribe to conspiracy theories. I have seen leftist communities that reject Western medicine, that are suspicious of the pharmaceutical industry (just think of Steve Jobs who, early on, tried to cure his cancer with natural practices—to his literal demise), that have refused to vaccinate their children (even before covid-19), that opt for homeopathic placebos (what modern medicine calls “medicine without an active ingredient”), or that vehemently reject genetically-modified foods (the technology that appears to be our best hope of curbing human starvation) even though little to no evidence exists that these would pose any danger.

There is a weird connection between the green left and the traditionalist right on the political spectrum that goes back to nature, to how things were before. And it is also this place where conspiratorial thinkers meet. Why do the leftists get a pass on conspiratorial thinking as we attribute the conspiratorial stereotype to right-wing actors?

In the same article you write about how the conspiratorial mind is a difficult one to change. What policies do you recommend to ensure against populists wielding conspiracy theories being able to damage democracy?

Well, we wrote that it is difficult for a reason. If I had good solutions, you’d see them in print already. I don’t. Federico Vegetti and I had a priming experiment pilot study where, for one group of survey respondents, we tried to be understanding of a conspiratorial perspective, and for another we tried to set people straight by pointing out how that perspective was wrong. Both approaches increased people’s conspiratorial thinking when compared with a control group who received no priming.

Not specific to conspiratorial thinking, Jennifer McCoy and I have been piloting various experimental treatments—priming to see if anything can be done to make people want to curb the rights of their political out-groups any less—and it has been among the most frustrating research endeavors I have engaged in to date. Nothing works. Even if something reaches statistical significance, substantively it still has very little effect. It is easy to rile people up against the out-group. The opposite is near impossible. I was talking to my political psychologist colleague and coauthor Cengiz Erizen once about this, asking him what we should try, what would work. His response was, “What else do you want me to do, cure cancer while I’m at it?” And that is what this research agenda feels like.

Of course, this is extremely difficult when everyone’s gut reaction is to find what’s pathological in the other side. One of the reasons the first question triggered such a negative reaction from me was exactly this. My mentor in college, John Hibbing, worked on the biology of political attitudes and there were times when he received media inquiries daily analogous to, “So, what does your research tell us about what’s wrong with conservatives? Finally, someone is looking at the physiology and we will figure out why they are so pathological.” I am exaggerating, but no matter how sophisticated one asks the question, it boils down to this: a healthy dose of othering, pathologizing those who think differently.

And this was over a decade before Trump. We need to do better. I am a big fan of the works of Theda Skocpol on the Tea Party, or Arlie Hochschild or Kathy Cramer. (Notice, all women, but maybe I could cite men as well like Justin Gest.) They went into communities they certainly disagreed with on most things and worked on developing some understanding, some empathy. We need more work along these lines and their experiences in how to connect with people they disagree with—something they don’t write enough about—which could be invaluable in answering this extremely important question you ask.

I mentioned the experiments with Jennifer McCoy. In one of these we asked people to imagine a coworker who is vocal about their politics, and with whom the survey respondent very much disagrees with. Then imagine that the coworker’s kid becomes very sick and may die. We tried to elicit some empathy, some common humanity. But when we asked an open question on what people think, some wrote that it’s good, they should suffer, their kid deserves to die. In another US study, we have some evidence that encouraging people to take the perspective of someone they disagree with actually increased people’s support for curbing that person’s rights. It was just a pilot. It was not replicated. Hopefully it was just an accidental result.

But this is where we are at as a community. It’s depressing. I recently told Jennifer that working with her has been the most depressing thing I have done professionally. (She is amazing to work with, but what we are finding along the way is enough to make someone lose all hope for humanity.) If we want to find ways to improve things, we have our work cut out for us.

Let’s move now to some methodological questions. In a 2020 article with Bruno Castanho Silva, Sebastian Jungkunz, and Marc Helbling, you explore ways of measuring populist attitudes. Can you talk about the advantages and disadvantages of the different measurements?

Measuring things is difficult. And political scientists do not think about the issues involved with measurement as much as they should. The first contemporary attempts to measure populism date back to Kirk Hawkins’ efforts to place a few populist attitude questions on the Cooperative Congressional Election Study in 2008. (There were some other ones in the ’60s and ’70s, but what they measured had little to do with populism as we know it today.) The items Kirk produced were later adopted for the Netherlands and became known as the Akkerman et al. items. Kirk’s paper (with Scott Riding and Cas Mudde) was never published in a peer-reviewed journal; instead, it lives on the web as a working paper.

When I looked at the items with my survey methods/psychometrics mindset, I thought, “These are not very good.” And Kirk would be the first to admit he was not a survey researcher coming up with these. He got some help from his American Public Opinion colleagues at Brigham Young University, but that was it. So, these items became the thing everyone used—mostly macro-comparativists who wanted to venture into populist attitudes research. And I thought we could do better. Kirk also encouraged this line of inquiry. So, Bruno and I set out to take a psychometrics approach to produce something we thought would be better, and the results of these efforts are in our edited book The Ideational Approach to Populism (co-edited with Kirk Hawkins, Ryan Carlin, and Cristóbal Rovira Kaltwasser.) We also posted a short memo on it on the Team Populism website.

I would say one of the main advantages of our scale was that it disentangled sub-components of populist attitudes. It gave the researchers some flexibility. But our scale was not the only one to do this. Anne Schulz and coauthors also worked on a scale that uses sub-components of populism. We approached scale development a little differently, though. Bruno and I did it fully inductively. We catalogued all our concepts related to populism by surveying colleagues, took all questions we could find that measured those concepts, and wrote a bunch more. Then we checked the underlying dimensionality of the data after we asked around 150 questions from our respondents. As it turned out, populist attitudes have four dimensions, the three that were theoretically sound—people-centrism, anti–elitism, and a Manichean worldview—plus authoritarianism, which was an independently-established and well-studied construct, so we discarded it. Anne and her team went another direction. They identified people-centrism, anti-elitism, and popular sovereignty (a construct that did not emerge in our inductive analysis) as theoretically relevant for populist attitudes and wrote questions to tap them.

Populist attitudes have four dimensions, the three that were theoretically sound—people-centrism, anti–elitism, and a Manichean worldview—plus authoritarianism.

When we did our final data collection to validate the items we selected from the exploratory analysis, we were a bit strategic and decided to ask every other scale used for populist attitudes at the time (except the ones we missed, and there were several). This allowed for a comparison. And it was fun do. It is what led to that paper. Along the way, we found out that Marc Helbling and Seb Jungkunz were working on the exact same thing trying to compare scales. I suggested that we should just share data and write one good paper. And the team was great to work with.

In the end, the paper concluded that I was wrong all along. The Akkerman items were great for measuring populist attitudes. But our scale did not fare too badly either, and had the additional advantage of being able to separate the three core components of populism: people-centrism, anti-elitism, and a Manichean worldview. To be perfectly honest, I grew quite disillusioned with our scale afterward (with all the scales we evaluated, really, including Akkerman et al.’s).

Where all these measurement instruments fail is the study of populists in power: Hungary, Poland, Venezuela, even Greece was there at one point. If you look very closely at the article, the Greek problem is visible, though definitely not adequately discussed. (There could be other things going on there as well since, unlike in most of the countries assessed in the article, Greek populists were mostly leftist.) There is a footnote in the paper saying we had to discard the Hungarian data because it simply did not work as expected. At the time, we did not fully comprehend all these complexities we now see. What’s going on was that in these two countries populists were in power and none of the scales could deal with this political reality effectively.

Since the paper was written, there has been quite a conversation about the nature of populism’s components and this was made possible directly by scales that allowed for the measurement of sub-components of populism. Alex Wuttke asked us if he could use our data from the populist scale comparison paper. Today, his work is in the American Political Science Review (the discipline’s flagship journal) and is the article I use in my measurement workshops to highlight the necessity of deep thinking on the nature of concepts and their components.

In that article, you explain the correlation between populist sentiments and the propensity for populist parties to have electoral success. What happens when there isn’t a party that maps neatly onto populist sentiment?

This finding is not exactly attributable to us. We looked at it with various scales, but I would cite Van Hauwaert and Van Kessel’s article “Beyond Protest and Discontent.” They were the ones who broadly established the connection between populist attitudes and populist party support.

Of course, if there is no populist supply, if there are no populist parties, it doesn’t matter if there is demand from the side of the public. When we analyze public opinion data in countries where there are populist parties, we see that those with higher populist attitudes are more likely to support populist parties, though we do not think of this as a perfect relationship. These are just general tendencies, or average results. But this result does not cleanly aggregate to the macro level. In our book on the US with Kirk, we have a figure that places the US in comparative perspective with regards to the average level of populist attitudes in other countries. The result is quite striking. There is barely any difference between the countries on their level of populist attitudes: there is huge variation within the countries, but country averages barely differ. When we do take a close look at the minute differences between countries in average populist attitudes, we see that Switzerland (a country with quite a bit of populist supply) is the lowest. One of the highest was Chile, a country that, at the time of data collection in 2016, had no populist parties visible on the political landscape.

We never explicitly wrote it down, but if you read between the lines of our work, it says that populist attitudes are very much present in all societies and populist political strategies would probably be successful everywhere, but especially in countries with a lot of populist attitudes. I have been keeping an eye on countries where there had been little to no populist supply in terms of elected officials, as I had an idea for a survey experiment to run in one such country. But I never fully developed the idea to conduct the study. Since I started thinking about this, Chile (which was a low populist supply country prior to 2020) has fallen to strong populist players. Portugal has fallen. If Ireland also falls, maybe I should scrap the idea altogether.

Of course, in Hungary, we have the opposite problem. I recently collected some populist attitude data in my attempts to develop the scale I mentioned above. My first intuition was to check quickly if people with populist attitudes vote for populist parties. Then I looked at the list of parties and seriously wondered, do we even have one that is not populist? Fidesz, of course, is quite populist. Gyurcsány’s Democratic Coalition is clearly trying to position itself as a left-wing populist party. The Momentum Movement party has been picking up the populist rhetoric very quickly. It is a misperception that the Jobbik party has historically been populist. Scholars tend to lump all of the European radical right, Jobbik included, together with populists, but in fact Jobbik was always a very elitist party. Power to the people? Maybe not. In Parliament they have been impressively technocratic. They have never shied away from working with experts. But their new leader, Péter Jakab, has been quite populist in his rhetoric, whether in support of the “good people” of Hungary just trying to make ends meet or in his critiques of the government, or their cronies, who deserve nothing less than long prison sentences. We have no non-populist parties left in Hungary.

You have also worked on US populism from a comparative perspective. Could you discuss US populism specifically, as compared to in Europe or Latin America? Where do you see similar transnational trends?

We have been dancing around this question a little already. In terms of attitudes, the US was in the middle tier of the countries we compared, not that it mattered as per the conversation above. Looking at the leaders, Trump was quite populist though still far from Hugo Chávez or Evo Morales. He was close to Hungarian Prime Minister Viktor Orbán or Turkish President Recep Tayyip Erdoğan in our speech codings (though Orbán has been upping his populist rhetoric lately; the data we had was a few electoral cycles back). Needless to say, former US Presidents Barack Obama, Bill Clinton, and President Joe Biden do not register much on the populism scale. Obama was not at zero, but he was very close. Sanders, on the other hand, scored higher than Trump, approaching but not really reaching the Latin American populist archetypes. Texas Senator Ted Cruz was also comparable to Trump in his populist rhetoric, but he could not benefit from this in the crowded primary-election space with competition from Trump. People with populist attitudes were more attracted to Trump than Cruz.

The US, in fact, is an extremely important case study, as both leftist and right-wing populist forces are present in the political system. This allows us to disentangle the impact of populism from ideology, something that cannot really be done in most polities. There aren’t too many countries with both left- and right-wing populist present. Of late, Spain comes to mind, with Podemos and Vox. Austria had a budding left-wing populist party, the Green Party splinter Pilz list, but they fell out of Parliament completely after the last election and did not regroup. I mentioned Hungary, but here the problem is the lack of potential for comparison with non-populists, since they don’t exist.

In your latest article on “Democratic Hypocrisy,” you uncover a fundamental pattern, which seems to confirm that US citizens are more ready to support norm-eroding policies when their party is in power—a finding that challenges the common idea that leaders are often the main ones responsible for democratic norm erosion. Can you tell us more about your findings?

The source of this idea came from a time when Jennifer McCoy and I started working on the connection between populist attitudes and liberal-democracy-eroding attitudes for the aforementioned edited volume on the consequences of populism. This data was collected in the first or second year of the Trump administration and, like many others who were looking at the attitudes towards democratic values, we also started noticing the partisan differences.

As also clearly seen in the “Democratic Hypocrisy” article you asked about, in this earlier data collection, Republicans were more okay with the erosion of liberal-democratic protections. And one would be justified to simply believe this finding, especially given how little respect the Republican elites have for institutions that protect the minority party. But I was suspicious. Elites are not the masses. Also, the Democrats do not get a pass here either. The filibuster, the institution that protects the rights of the political minority in congress, was most recently weakened by Democrats under Obama. So the question became: is it really just the Republicans, or is it whoever happens to be in power at the time, who are more forgiving when liberal democracy is eroded by the country’s rulers?

This, of course, is not an easy question to test. Despite conspiracy theories floating around about Central European University where I work and its founder George Soros, I really don’t have a way to experimentally manipulate who is in and out of power. So, we decided to turn to the next best thing. We sketched a scenario for the next election in which, for some, the Republicans won control of both houses of Congress and the presidency, and for others, the Democrats did. Comparing the two groups will give us a way to measure the gap between the support for policies eroding liberal-democratic protections when people are in and out of power: what is the magnitude of democratic hypocrisy? Of course, the first question everyone asks us is if this is bigger for Republicans than Democrats, and I wanted to be very careful in interpreting the partisan differences. In such a scenario as an experimental treatment, it is harder for study participants to imagine the people who are out of power at the time of the survey being in power for the purposes of the experiment. So, in this 2020 data collection under the Trump administration, it would have been harder for Democrats to comply with the experimental treatment of being in power, and for Republicans to comply with the treatment of being out of power. Simply speaking, it just would have been harder for them to imagine. So partisan differences may be biased even if, overall, I am inclined to believe the results. Of course, the ideal situation would be to do the study with the Republicans in power and then another study now that the Democrats have taken over. Stay tuned.

I can also tell you that democratic hypocrisy is very much present in other countries as well. We have looked at this in several countries now and it certainly is present everywhere we look. And we were not the only ones who thought asking this question would be worthwhile. They didn’t call it “democratic hypocrisy,” but Mazepus and Toshkov (2021) looked at the same phenomenon both in observational surveys EU-wide and in a Ukrainian survey experiment. Sasmaz, Yagci, and Ziblatt (2022) looked at the same thing in Turkey. We have heard of colleagues finding similar things in Armenia and in the Republic of Georgia, and we are currently working on related questions in Hungary and maybe even elsewhere. Again, stay tuned.

We believe that democratic hypocrisy is an important finding that will change the way we approach the strengthening of democratic institutions in backsliding democracies. In fact, our Journal of Politics piece you cite was a first attempt not only to show the presence of democratic hypocrisy but to see what affects it. Based on our findings, fear of the other party and strength of one’s partisanship certainly increase it. But the real question—something we are working on now—is how can we make the world a little better by mitigating democratic hypocrisy?

We believe that democratic hypocrisy is an important finding that will change the way we approach the strengthening of democratic institutions in backsliding democracies.

In general, also in populist attitude research, we see many studies that show how we can activate populist attitudes in people. The studies I would much rather see out there are ones that work towards getting people to be better neighbors and more civilized political actors by rejecting populist sentiments. I also fear that we live in a world where the demand for such outcomes is ever decreasing. As one of my mentors once said, “You study public opinion long enough, you start to wonder if the people should be given much power at all.” And indeed, in addition to being more and more depressed about the world, the more I study public opinion, the more I come to favor stronger and stronger institutional checks and balances limiting the power of the majority and protecting those who have less political power.

There is a quote attributed to Benjamin Franklin (though he probably never wrote it and the author is anonymous): “Democracy is two wolves and a lamb voting on what to have for lunch.” Despite the general endorsement of the populists of our day, the illiberal democrats, this is not a good place to be. The quote often goes on to defend the sheep’s gun rights, but I doubt that in the age of satellite-controlled guns, those rights will help anyone defend themselves from tyranny. Instead, we must understand the importance of civil rights and the dangers of the majority’s tyranny. In this light, we can quickly realize that democracy is not just about majority rule, and we can start to cherish all democratic institutions. Without freedom of the press, freedom of assembly, rule of law, transparency, and accountability, democracy is doomed no matter what the people whom one charismatic guy claims to speak for, en masse, may want.

---

Levente Littvay is Research Professor at the Centre for Social Sciences of the Hungarian Academy of Sciences Centre of Excellence and Professor of Political Science and Democracy Institute Research Affiliate at Central European University where he teaches graduate courses in research design, applied statistics, electoral politics, voting behavior, political psychology, and American politics. He is the inaugural and only two-time recipient of CEU’s Teaching Award (2015 for methods-, and 2021 for online teaching). Received his MA and PhD in Political Science and an MS in Survey Research and Methodology from the University of Nebraska-Lincoln. Taught numerous research methods workshops globally and online, and was one of the Academic Convenors of the European Consortium for Political Research Methods Schools (2015-21), is the founder of MethodsNET, head of Team Survey in Team Populism where he helped spawn the New Populism series with The Guardian and co-produces both the MethodsNET and Team Populism YouTube channels. He is a member of the European Social Survey’s Round 10 (2020-21) democracy and COVID19 module questionnaire design teams. Secured close to a million EUR in grants to conduct research on survey and quantitative methodology, twin and family studies, and the psychology of radicalism and populism. In 2019-20 he was European University Institute’s Fernand Braudel Senior Research Fellow. Has publications in Social Justice Research for which he received the Morton Deutsch Award for best article in 2017, Political Analysis, The Journal of Politics, Political Research Quarterly, Political Psychology, BMC Medical Research Methodology, and along with other medical journals, in Twin Research and Human Genetics where he is Associate Editor for Social Sciences. Books include Contemporary US Populism in Comparative Perspective with Kirk Hawkins in Cambridge University Press’s Elements Series and Multilevel Structural Equation Modeling with Bruno Castanho Silva and Constantin Manuel Bosancianu in SAGE QASS (little green book) series.