Understanding Democratic Backsliding: Insights from Leading Researchers

Authored By 
Rick Harrison
March 17, 2025

Andrew Thompson speaks in a classroom in front of a projection screen showing a chart.

What might America look like without democracy? Are the checks and balances threaded throughout the U.S. Constitution still valid and operating? Do Americans care about living in a democracy?

Institution for Social and Policy Studies faculty fellow Milan Svolik hosted a conference on Feb. 28 and March 1 to help answer these and similar questions. Political scientists specializing in American government or comparing political systems in different countries presented their research on the nature, causes, and consequences of the rapidly unfolding challenges to democracy.

“So many of you study and talk about America’s contested democratic creed,” Svolik said to introduce the conference. “Over the past decade, we have begun questioning many long-held assumptions about the stability of democracy around the world. Especially in the United States.”

The event was sponsored by Democratic Innovations, an ISPS program designed to identify and test new ideas for improving the quality of democratic representation and governance. Recent efforts include a conference on government problem-solving, the founding of Yale Youth Poll, and an upcoming Democratic Innovations Literature Review Series that will offer jargon-free, scientifically accurate descriptions of the key subjects relevant for the promotion of high-quality, honest, and efficient democratic governance.

“America’s democracy rests on a complex mixture of laws and norms” said ISPS Director Alan Gerber, Sterling Professor of Political Science. “As social scientists, we have a vital role to play in helping people understand how this system works, the conditions under which our democracy is threated, and how to improve our political institutions.”

Carles Boix, Robert Garrett Professor of Politics and Public Affairs at Princeton University, said that democratic backsliding is often the result of deliberate manipulation by political elites, adding that the process, which can be gradual, is very much rooted in the social and political environment in which they operate.

“We must consider the economic conditions that enable politicians to manipulate democratic processes,” he said.

Boix presented research using data from all democracies around the world during the period from 1900 to 2019 showing that income inequality increases political polarization and, as a result, the likelihood of democratic backsliding and breakdown. However, he pointed out that, higher income levels attenuate the negative impact of inequality.

Adam Przeworksi and Kevin Elliot listen in the audience of a presentation in a classroom with their hands on their chins.

Susan Stokes, Tiffany and Margaret Blake Distinguished Service Professor at the University of Chicago and director of the Chicago Center on Democracy, explained that backsliding governments around the world have been led either by right-wing ethnonationalists or by left-wing populists. Focusing on the ethnonationalists, she indicated that not all such parties in power have undermined their democracies.

For example, she said we can learn much from comparing cases of erosion of democracy under ethnonationalists (such as in the United States), near-misses (such as the United Kingdom under the government of Boris Johnson), and instances in which ethnonationalists did not even attempt such erosion (such as in Sweden, despite the rise of the Sweden Democrats).

“Effective resistance to democratic erosion requires strong institutions, social trust, and political elites committed to upholding democratic norms,” Stokes said.

Justin Grimmer, Morris M. Doyle Centennial Professor of Public Policy at Stanford University, presented his team’s research on how conspiracy theories on voter fraud spread and affect public trust.

“The power of these conspiracy theories lies in their vagueness,” Grimmer said. “They lack specifics, making it difficult to disprove them, and almost anything can be counted as evidence by a sympathetic audience.”

These vague conspiracies spread widely on social media then create cycles of distrust between activists and local election officials because the increased attention uncovers standard election procedures, innocent mistakes, or limitations of official records that get misinterpreted as confirmation of fraud, he said.

“We need to understand how these theories propagate and the impact they have on local election integrity movements,” Grimmer said. “It’s not just about debunking them but also about addressing the underlying dynamics that make them so compelling to certain groups.”

Matt Graham, assistant professor of political science at Temple University,  presented a paper he wrote with Svolik on how citizens respond to democratic backsliding, as seen through tactics such as manipulating electoral rules, undermining judicial independence, and restricting media freedom.

In an experiment using simulated circumstances, they observed that when a candidate violates democratic principles, a significant number of people switch their vote away from that candidate. The effects were largest among voters who barely prefer the candidate for policy reasons.

“Our findings suggest that voter turnout and engagement are crucial for maintaining democratic stability in the face of backsliding,” Graham said. “At the same time, successful elite framing can turn attacks on democracy from a liability into an asset, altering voter behavior.”

Matt Graham speaks in a classroom, as seen from the audience.

Andrew Thompson, assistant professor of political science at the University of Pennsylvania, addressed the racial and partisan dynamics affecting democratic backsliding in the United States, proposing that changes in racial demographics lead to shifts in power and influence, particularly as racial diversity is perceived as a threat by segments of the population.

“The interplay between racial demographics and partisan politics is a ticking time bomb for American democracy,” Thompson said. “As the racial composition of the electorate changes, so too does the landscape of political power and influence.”

In a series of experiments, Thompson found that Republicans showed decreased racial threat, increased trust in elections, and reduced anti-democratic sentiments when they were told that people of color were more likely to identify as Republicans. Democrats’ reactions were generally unaffected by the framing of racial demographic changes.

He added one major caveat, finding that Republicans only decreased their anti-democratic support when learning white Republicans would maintain control of the party, not when people of color were projected to have control. He concluded that white control of the Republican party influences attitudes toward democracy and white Americans’ perception of losing social status drives anti-democratic sentiments.

“Understanding the racial-partisan dynamics of democratic backsliding requires us to confront uncomfortable truths about identity, power, and the future of democracy in America,” Thompson said. “It’s not just about numbers. It’s about the narratives we construct and the fears we address.”

Yphtach Lelkes, associate professor of communication at the University of Pennsylvania, presented a co-written paper finding that students resist penalizing objectionable speech unless it is considered extremely harmful. However, they tend to react more strongly and support harsher punishments when hateful rhetoric is directed at minority groups compared to when similar statements are aimed at white students.

Rocío Titiunik, professor of politics at Princeton University, shared a paper in which she and her co-authors concluded that charismatic leaders may pose a significant threat to democratic stability, particularly in polarized and less stable democracies. They wrote that political parties rely on the popular support attracted by a charismatic leader for electoral success. This “can lower the parties’ willingness and ability to constrain their behavior, weakening an important institutional check on the potential abuse of power,” they said.

Kristian Frederiksen, assistant professor of political science at Aarhus University, closed the conference with a study of why political elites do not oppose Donald Trump despite his many violations of democratic norms. Through survey data from U.S. state legislators, local officials, and congressional challengers during the 2024 primary elections, he and his co-authors found that politicians are unwilling to pay the costs of breaking with Trump, which is seen as much more costly than breaking with either Joe Biden or Kamala Harris.

Elites have internalized a narrative that closely resembles what Trump attempts to demonstrate publicly: prompt retaliation from Trump and an unwavering party base make defection prohibitively costly,” they wrote. “Our findings underscore the need to better understand how undemocratic leaders are able to shape elite incentives by putting their power on display.”

Svolik called for continued rigorous research on which aspects of American democracy are resilient and why. And which aspects are vulnerable, and why.

“One silver lining is that we, as political scientists, have begun having conversations that we were not having before, and that is across the boundaries of American politics and comparative politics,” he said. “The hope for this conference is that we will continue that conversation.”