Introduction
Ever thought about how much the average voter actually values democracy? Spoiler alert: it’s complicated.
Recently, someone sent me an article from the Atlantic entitled “Only About 3.5 Percent of Americans Care About Democracy (gift link)” and it referenced a study done by two professors from Yale.
I was a bit surprised by the title and I went in search of the underlying study.
I found it, and I am going to take some exception to the framing by the Atlantic, but let’s explore.
While Americans wave the flag and champion “democracy” as the best form of government (even though we are a constitutional republic), researchers Matthew H. Graham and Milan W. Svolik put this concept to the test.
They set out to see if Americans would stand by democratic principles—even if it meant voting against their own political team.
So, let’s dive into what this study uncovers about the strength (or fragility) of democratic support in the U.S.
Title: Democracy in the Balance: How Polarization Impacts Americans’ Support for Democratic Principles
Link: Graham, M.H., & Svolik, M.W. (2020). Democracy in America? Partisanship, Polarization, and the Robustness of Support for Democracy in the United States.
Peer Review Status: Peer-reviewed
Methodology
Graham and Svolik used two complementary approaches to assess how willing American voters are to prioritize democracy over partisanship:
Survey Experiment with Hypothetical Candidates: The researchers designed a survey with a nationally representative sample of over 1,600 U.S. voters, asking participants to make choices between pairs of hypothetical candidates. To reflect real electoral scenarios, these candidates were randomly assigned specific party affiliations (Republican or Democrat), policy platforms (economic and social positions), and characteristics (like age, profession, and years of experience).
But here’s the twist: some candidates were also given undemocratic stances, such as supporting a reduction in polling places in areas favorable to the opposition, calling for prosecuting journalists, or endorsing gerrymandering. By mixing these attributes, the researchers could observe whether voters were more inclined to reject undemocratic candidates or, alternatively, overlook these behaviors to support their preferred party. Across 16 randomized choice scenarios, each participant revealed just how flexible they were with their commitment to democratic principles.
Natural Experiment in Montana’s 2017 Special Congressional Election: For a real-world angle, the researchers examined Montana’s 2017 special election for the U.S. House, where one candidate notoriously assaulted a journalist shortly before election day. This incident served as a real-world “democracy violation”—a public act against democratic norms. Here’s the clever part: Montana has a large population of absentee voters, many of whom cast their ballots prior to the assault. By comparing absentee votes (pre-assault) with election day votes (post-assault), Graham and Svolik could isolate the impact of this undemocratic incident on voters’ choices. This natural experiment allowed them to test if, given clear evidence of anti-democratic behavior, partisans would still stand by their candidate.
This dual-method approach offered insight into the general principles voters claim to hold in surveys versus the principles they actually apply in real elections. In total, the study captured data from thousands of voter responses, offering a robust look at how flexible (or inflexible) Americans are when democracy itself is on the line.
Results and Findings
The findings might surprise you—or not, if you’re a realist about American politics. Here’s what Graham and Svolik found:
- Party Over Principle: A solid chunk of voters prioritized their party over democratic values, with about 86% of participants sticking with their partisan candidates, even when those candidates opposed democratic principles.
- Centrist Resistance: Political moderates were more likely than extremists to vote against undemocratic candidates, proving a rare pro-democratic force amid strong partisan loyalty.
- Double Standards Across Party Lines: Both Republicans and Democrats showed leniency towards undemocratic behaviors when it came from their own side but were more critical when it came from the opposing party.
- Natural Experiment Results: The Montana election’s results were consistent with the survey experiment: only voters in moderate areas punished the candidate for undemocratic actions, while partisans tended to stay loyal.
Critiques of the Research
While illuminating, this research has some limitations:
Timing: It’s worth noting that this research was conducted in 2018 and published in 2020, before the events surrounding January 6, 2021, and the ensuing media storm, and the 2024 campaign.
When studied, I’m not sure many Americans were seriously pondering threats to our democratic foundations. This experiment was likely veiwed as more as a hypothetical than a real, potential threat.
Here’s the question: if this study were repeated post-January 6—after such a seismic event—would we see stronger effects? Larger partisan effects? My guess is we would.
Respondents: I cannot access the study’s appendix due to the paywall. However, based on common practices in academic research, I assume the participants are likely a convenience sample consisting of college-aged students. If this is the case, it’s reasonable to suppose that these students, as part of an academic exercise, would be inclined to downplay hypothetical threats to democracy. This is especially true when compared to more immediate concerns, such as the potential rejection of a date for that weekend.
Generalizability and Election Context: The findings from Montana’s special election might be unique to that context, especially considering the high-profile nature of the event and not generalizable to the entire country.
Conclusion
Writing about academic research is difficult (trust me!), balancing nuance with concise writing is a high wire act. And the media has the additional need to attract readers meaning typically any nuance gets murdered in the headlines.
This brings us to a key nuance in Graham and Svolik’s study: it’s not that voters are indifferent to democracy, but rather that, in a polarized society, democratic values end up competing with the pull of partisanship and specific policy goals. They take a back-seat especially if the threat is not salient.
So, while democratic norms matter to Americans, only a small number are willing to sacrifice their other priorities to defend those norms—especially when it means opposing their own party.
This doesn’t mean that democratic principles don’t matter. However, when this study was conducted, I doubt many Americans felt democracy itself was truly at risk. The threat was likely some far-off academic hypethtical scenario.
Since the research and publication, focusing events have likely shifted perspectives, and I suspect future research will show that more than 3.5% of voters truly prioritize democracy.
Though this layered conclusion may sound troubling, it points to a possible path forward. Democratic values are still widely understood and can be made more salient and rise in importance – possibly to the center of our priorities.