We’re often told to “follow the science”—a comforting phrase that suggests clarity, objectivity, and consensus. But in today’s hyperpolarized world, even science itself has become a political Rorschach test. A new study in Science reveals that Democrats and Republicans cite science differently and effectively operate from separate scientific realities.“I want to pause here and talk about this notion of consensus and the rise of what has been called consensus science. … Historically, the claim of consensus has been the first refuge of scoundrels; it is a way to avoid debate by claiming that the matter is already settled. …In science consensus is irrelevant. What is relevant is reproducible results. The greatest scientists in history are great precisely because they broke with the consensus.”
– Michael Crichton
Crichton’s critique challenges a popular narrative: that science speaks with a unified voice. While his stance may oversimplify science’s collaborative and cumulative nature, he raises a valid concern—consensus can be misused as rhetorical armor to shut down debate. In practice, however, scientific consensus is often a signal of accumulated evidence, not ideological conformity.
Partisan Science
Following the science seems to be a homily best honored in word rather than deed. “scientific” evidence or reports are often flashed around to support regulation. According to a Pew survey, there is a clear divide between how Democrats and Republicans view science. As noted by Vox in describing Pew’s findings:
Of course, how citizens feel and Congress acts may differ, and we get a fascinating picture of what following the science means to partisans from a new study in Science.
How Science Splits Along Party Lines
Is science used differently by policymakers in different parties? Spoiler Alert: Yes, and consistently across the years.
As we might expect, the researchers found
“systematic differences in the amount, content, and character of science cited in policy by partisan factions in the United States. These differences are strikingly persistent across fields of research, policy issues, time, and institutional contexts.”
In other words, how science informs policy is not just a matter of quantity but of ideology. It’s not simply that one side cites more science—it’s that both parties consistently select different kinds of science to support their policy positions, reinforcing partisan narratives rather than converging around shared evidence.
Researchers analyzed a massive dataset to explore how science is used in American policymaking: over 49,000 congressional reports and hearings since the mid-1990s, plus more than 190,000 policy documents from U.S. think tanks dating back to 1999. They zeroed in on the 424,000 scientific papers cited in these documents and traced them to a global database of over 122 million academic publications.
By linking these two worlds—science and policy—they could see not just what science was being cited but also the nature of the science used: whether it was peer-reviewed or a preprint, its influence within its field, and how often it had been cited by other researchers. This comprehensive mapping allowed the researchers to reveal how deeply political ideology shapes the selection of scientific evidence in policymaking, not just the presence.
A Tale of Two Approaches
Let’s begin with the good news. Policy has increasingly cited science in the last 25 years. It has grown from 20% to 35% of policy statements, especially in documents generated by think tanks. This suggests that, at least on the surface, science is playing a larger role in shaping policy conversations.
But dig deeper, and a partisan pattern emerges. Congressional committees, central engines of policy development, show a clear divide. Committees led by Democrats are nearly twice as likely to cite scientific research as those led by Republicans. This partisan disparity becomes especially evident when committee control switches from Republican to Democrat. The shift is immediate and substantial, spanning nearly every domain—“20 of the 23 scientific fields and 15 of the 17 issue areas.”
“You are entitled to your opinion. But you are not entitled to your own facts.”
– Senator Daniel Patrick Moynihan
Yet, in today’s policy landscape, facts are increasingly factionalized. While both Democrats and Republicans draw on scientific research, they are rarely drawing from the same sources. Only 5 to 6% of scientific citations are shared across party lines. The authors note that “partisans [are] consistently less likely to cite the same science.” In fact, the actual level of bipartisan overlap is about half of what would be expected by chance. This gap has held firm over time and across institutional settings—suggesting it’s not a passing trend but a persistent feature of modern policymaking.
This divergence is often driven by political agendas. Take the House Energy and Commerce Committee: under Democratic leadership, the scientific topics cited differ sharply from those cited when Republicans are in charge. The committee’s focus shifts—not just in priorities but in the kinds of studies deemed relevant.
The split in think tanks working on the same issue is even more striking. In comparing minimum wage policies, the Urban Institute (left-leaning) and the Employment Policies Institute (right-leaning) cited 62 different studies—only one overlapped. That’s not just ideological framing—it’s near-total epistemic separation. As the researchers put it, “The science cited by the left- and right-of-center think tanks is nearly perfectly separated.”
And the quality of that science differs as well. Democratic-led congressional committees are more likely to cite peer-reviewed research, studies among the top 5% most cited in their field, and slightly older literature—possibly reflecting a preference for established consensus. While both parties cite research of similar average influence, the Democratic pattern suggests a tilt toward more academically validated work.
Among think tanks, the contrast grows sharper. Left-leaning think tanks prefer more recent, peer-reviewed, and widely cited science, often echoing influential research within academic circles. Right-leaning think tanks, by contrast, tend to cite less prominent or less current studies—possibly prioritizing ideological alignment over academic standing.
Why is this divide more pronounced in think tanks than in Congress? One likely reason is accountability. Regardless of party, legislators must ultimately face voters and weigh the practical consequences of lawmaking. This pressure may push both sides to incorporate more credible science, even if selectively. Think tanks, especially the ideologically rigid ones, have no such constraint. They can indulge in confirmation bias with impunity, curating science that flatters their worldview.
Trust Issues
Researchers went straight to the source to understand how science influences policy: surveying the professionals who shape laws and regulations—from judges and corporate leaders to government officials. They also surveyed 1,000 likely voters to see how elite perspectives compare to the general public’s.
The results were striking. Across the board, political elites—regardless of party—trust scientists more than voters. However, partisan differences remain: Democratic elites show much greater trust in science than Republican elites. For example, over 60% of Democratic elites rated the National Academies of Sciences as “very trustworthy,” compared to just 23% of Republicans. This mirrors the broader populist backlash among Republican voters, where skepticism of expert authority has grown in parallel with political polarization.
Politics Over Peer Review
“Science is often seen as holding a relatively privileged position partly owing to its commitment to Mertonian norms like organized skepticism.” [1]
However, science is far from the only voice in the room. While the composition of Agency committees has and continues to be controlled by the Administration, that is not the case for Congressional Committees or think tanks. Science isn’t always the loudest or most relevant voice in Congressional ears.
While Democrats may foreground evidence to justify regulatory action, Republicans may prioritize individual liberty, market autonomy, or regulatory restraint—values not invalidated by science but which shape how evidence is used. The researchers point out that when seeking information, Democrats and Republicans may use “science in roughly equal measure.” However, Republicans using hearings and committee meetings as performance may be more intent on messaging their political base. A base that is increasingly distrustful of scientific expertise.
The lack of commonality in the science each partisan group highlights may indulge their respective voting bases, but it simultaneously undermines scientific credibility and government trust. In turn, mistrust raises questions about bias, not just in how science is used but also in how it is produced. While a Republican Administration is tearing apart the “deep state” aspects of government-funded research, both parties have culpability in creating an environment that fosters research’s Bonfire of the Vanities.
Science is still a powerful force in policymaking—but its credibility, influence, and utility are increasingly filtered through political lenses. When science becomes just another partisan talking point, we lose more than consensus—we lose the foundation for problem-solving in a complex world. How can we agree on a future if we can’t even agree on the facts? In our fragmented “reality,” good science risks being drowned out in partisan noise.
[1] Robert K. Merton’s four norms of science—Universalism, Communality, Disinterestedness, and Organized Skepticism—form the ethical foundation of scientific inquiry by promoting objectivity, open knowledge sharing, impartiality, and critical evaluation. Together, these principles ensure that scientific claims are judged on merit, shared for collective progress, pursued without personal bias, and rigorously scrutinized for validity.
Source: Partisan disparities in the use of science in policy Science DOI: 10.1126/science.adt9895
