studyfinds.org /americas-political-divide-surged-2008-2020/

America's Political Divide Surged 64% Between 2008 And 2020, But Only One Side Grew More Extreme

StudyFinds Analysis 11-14 minutes 2/5/2026
DOI: 10.1098/rsos.251428, Show Details

Partisan politics of the democrats and republicans are creating a lack of bipartisan consensus. In American politics US parties are represented by either the democrat donkey or republican elephant

Both sides of the political aisle can't help but see the world their way. (© Victor Moussa - stock.adobe.com)

Study shows liberals ‘shifted substantially’ more to the left, while conservatives held steady.

In A Nutshell

  • The polarization surge was concentrated: American political polarization stayed relatively flat from 1988 to 2008, then surged 64% between 2008 and 2020: all the growth happened in just 12 years.
  • One side moved more than the other: The left-leaning group shifted substantially further left while the right-leaning group barely moved, contradicting the “both sides getting extreme” narrative.
  • Every issue became tribal: Abortion, healthcare, racial justice, and family values all became tightly bundled, knowing someone’s position on one issue now predicts their stance on everything else.
  • The divide may be reversible: Because polarization surged during a specific historical period rather than gradually over decades, understanding what drove it could help address it.

Arguments at Thanksgiving dinner feel more heated than ever. Your uncle who voted one way can’t fathom your cousin who voted another. They’re not just disagreeing anymore, they’re living in completely different political realities. Now, interesting research reveals this great divide didn’t creep up gradually. It surged rapidly during a concentrated 12-year period.

MY LATEST VIDEOS

Researchers at the University of Cambridge and the London School of Economics analyzed 36 years of American voter data to make this discovery. From 1988 to 2008, political polarization in America (at least measured by the gap between opposing political viewpoints) barely budged. Then between 2008 and 2020, it rose sharply in a sustained climb, growing more in those 12 years than in the previous two decades combined. By 2020, the gap between America’s two political camps had widened by roughly 64 percent compared to where it stood in 1988.

This research, published in Royal Society Open Science, doesn’t measure whether Democrats and Republicans simply dislike each other more, though that’s also happened. It measures something deeper: whether Americans genuinely hold more divergent beliefs about the issues that matter. The answer is yes. And the timing raises a pressing question. What happened during those specific years that pulled America’s political fabric apart?

The Flat Years Nobody Remembers

Using machine learning to analyze thousands of survey responses from every presidential election since 1988, researchers sorted Americans into two groups based purely on their political opinions without using party identification at all. They looked at 14 major issues spanning abortion, healthcare, income equality, racial justice, and traditional values.

For each election year, they measured how far apart these opinion groups had drifted. In 1988, there was disagreement, sure, but both groups occupied somewhat overlapping political territory. That distance stayed essentially flat through the 1990s and into the 2000s. Americans argued about politics, but their actual beliefs weren’t pulling dramatically apart.

Then 2008 hit, and the pattern began to change.

What Changed During the Obama-Trump Years

By 2012, the gap between the two groups had noticeably widened. By 2016, it gaped. By 2020, Americans inhabited two distinct political worlds. The 2024 data shows a slight narrowing, but the divide remains far wider than it was in 2008.

The study itself doesn’t assign causes, but the timing overlaps with seismic shifts in American life: the Great Recession and its lingering economic anxiety, the Tea Party uprising, Facebook and Twitter becoming political battlegrounds, the Affordable Care Act fights, Black Lives Matter protests forcing conversations about systemic racism, and Donald Trump’s norm-shattering presidency.

Any one of these forces might explain part of the shift. In all likelihood, they combined in ways researchers are still untangling. Crucially, this timeline challenges the idea that polarization simply rises automatically over time. Instead, it concentrated in a specific 12-year burst. Something about that era specifically pushed Americans into opposing camps.

Liberal vs Conservative road sign
The amount of Americans whose opinions matched their chosen party increased significantly between 2008 and 2020. (Image by Jane0606 on Shutterstock)

Abortion, Family Values, Healthcare: Everything Became Tribal

The divide grew across every single issue researchers studied, but some issues transformed more dramatically. Take abortion. In 1988, plenty of Americans with generally conservative views on most issues held moderate positions on abortion. It was possible to be a fiscal conservative who supported abortion rights, or a Democrat who opposed them.

By 2024, that cross-cutting complexity had declined sharply. Abortion had become tightly bundled with every other political position. Know someone’s view on abortion, and you can probably predict where they stand on healthcare, immigration, and climate change.

“Traditional family ties” followed the same pattern. So did views on government healthcare and racial justice. Issues that once allowed for nuanced, independent thinking became loyalty tests. Americans didn’t just sort themselves into political tribes during this period, they adopted comprehensive ideological packages.

Meanwhile, Americans became dramatically better at aligning their party identification with their actual political views. The proportion of people whose opinions matched their chosen party increased substantially over this period.

One Side Moved, One Side Held Steady

As America polarized, it also became more liberal overall. The left-leaning group shifted substantially further left between 1988 and 2024. The right-leaning group moved slightly rightward, but the shift was statistically small, barely noticeable compared to the left’s movement.

This contradicts the common story that “both sides are becoming extreme.” At least on the issues studied, one group moved significantly while the other held relatively steady. The polarization came from that growing gap, not from two groups racing toward opposite extremes.

What Happens Now?

The 2024 data hints that the divide narrowed slightly from its 2020 peak, though it remains far wider than before 2008. Whether that signals the end of the polarization surge or just a temporary pause won’t be clear for years.

What researchers can say with confidence about past patterns: American political polarization isn’t some inevitable force of nature. It was relatively stable for 20 years, surged for 12, and may now be shifting again. The forces that drove Americans apart during the Obama-Trump years were specific to that moment, not permanent features of democracy.

The real challenge is figuring out what specifically about 2008-2020 created this divide. Was it economic anxiety from the Great Recession? The rise of social media echo chambers? Increasingly partisan media? Political leaders who benefited from division? Probably all of the above, in ways that fed off each other.

The findings suggest polarization is not fixed in stone, but understanding what drove it is a necessary first step toward addressing it. Americans didn’t gradually drift apart over generations, they broke apart sharply during a specific window of time. If those forces can be identified and understood, the divide that feels so permanent today might eventually narrow. The pattern doesn’t appear inevitable. It reflects a historical moment that researchers are only now beginning to fully map.


Paper Notes

Study Limitations

The researchers acknowledge several important limitations in their analysis. The regression analyses examining predictors of polarization across countries could not establish causation due to insufficient longitudinal data—the associations found might be explained by unmeasured variables or even reverse causation. Additionally, there was some imprecision in matching economic and institutional predictor data to the exact years when opinion surveys were conducted, as researchers used data from within 5 years when exact matches weren’t available.

The k-means clustering approach forces all individuals into one cluster or another, creating a qualitative difference from traditional approaches that might exclude moderates or independents. While the study examined 14 issues in each dataset, results are necessarily tied to those specific issues and may not generalize to all political topics. The researchers also note that if systematic sampling errors exist in the American National Election Studies data, the patterns might not perfectly represent the entire US population, though the ANES uses sophisticated random-probability sampling methods.

Finally, the study could only analyze data available in large-scale surveys like the ANES and World Values Survey. Some potentially relevant political issues weren’t included because they weren’t consistently measured across all survey waves, meaning the analysis might miss divisions occurring on unmeasured topics.

Funding and Disclosures

This research was supported by a grant from the Templeton World Charity Foundation (TWCF number: 31453) awarded to Jens Koed Madsen and Lee de-Wit. Andreas Kapounek received support from a Leverhulme Trust International Professorship grant to the University of Cambridge. David Jack Young, James Ackland, and Lara Jane Greening were supported by Economic and Social Research Council Doctoral Training Partnership Studentships. The authors declared no competing interests.

Publication Details

Authors: David Jack Young, James Ackland, Andreas Kapounek (University of Cambridge, Department of Psychology); Jens Koed Madsen (London School of Economics and Political Science, Department of Psychological and Behavioural Science); Lara Jane Greening, Lee de-Wit (University of Cambridge, Department of Psychology) | Journal: Royal Society Open Science, Volume 13, Article 251428 | Publication Date: 2026 | DOI: 10.1098/rsos.251428 | Received: July 29, 2025 | Accepted: October 22, 2025 | Data Sources: American National Election Studies (ANES) Time Series Cumulative Data File (1988-2020) and 2024 Time Series Study Preliminary Release; Integrated Values Survey combining World Values Survey and European Values Survey data

Called "brilliant," "fantastic," and "spot on" by scientists and researchers, our acclaimed StudyFinds Analysis articles are created using an exclusive AI-based model with complete human oversight by the StudyFinds Editorial Team. For these articles, we use an unparalleled LLM process across multiple systems to analyze entire journal papers, extract data, and create accurate, accessible content. Our writing and editing team proofreads and polishes each and every article before publishing. With recent studies showing that artificial intelligence can interpret scientific research as well as (or even better) than field experts and specialists, StudyFinds was among the earliest to adopt and test this technology before approving its widespread use on our site. We stand by our practice and continuously update our processes to ensure the very highest level of accuracy. Read our AI Policy (link below) for more information.

Our Editorial Team

Steve Fink

Editor-in-Chief

John Anderer

Associate Editor