The Information War: How Partisan Think Tanks Fractured American Democracy (And What We Can Do About It)

America doesn't just have a polarization problem. We have an information problem.

On the AI in Chicago Podcast, we sat down with E.J. Fagan, professor of political science at the University of Illinois Chicago, and author of "The Thinkers: The Rise of Partisan Think Tanks and the Polarization of American Politics," to understand how we arrived at a moment where Republicans and Democrats don't just disagree on solutions, they can't even agree on basic facts.

The Death of Neutral Expertise

Professor Fagan traces the fracture to a specific moment: 1973, when three conservative congressional staffers founded the Heritage Foundation. Their frustration? The American Enterprise Institute had released a report on federal funding for a supersonic jet the day after the appropriation passed.

They wanted something different: not academic rigor on academic timelines, but strategic deployment of conservative ideas at the moment of maximum political impact.

The model worked. Too well.

What began as a desire for "conservative expertise" evolved into something more concerning: the replacement of a single knowledge regime with competing partisan knowledge regimes. Today, when a policymaker needs to understand climate policy, tax impacts, or AI regulation, they don't consult neutral experts—they consult experts who will tell them what they want to hear.

Why Smart People Believe Wrong Things

One of the most striking insights from our conversation: it's actually easier to fool someone who is smart, educated, and highly motivated than someone who isn't.

As Professor Fagan explains, "When people who are smart and educated and have lots of information hold a strong opinion, they will find information to confirm their opinion, regardless of how well it is true."

This isn't a bug in how think tanks operate. It's a feature. Organizations like Heritage select for ideologues, reward reports that confirm their worldview, and create intellectual environments that are "silently dishonest."

Even RFK Jr.—clearly an intelligent person—genuinely believes the claims he makes about vaccines. The data exists to support almost any conclusion if you're motivated enough to find it and smart enough to construct the argument.

The Cost of Ideological Capture

The consequences extend beyond abstract debates. Professor Fagan points to the 2009 stimulus bill as a case study. When the Obama administration needed to deploy $900 billion quickly to address the financial crisis, they turned to their partisan think tanks for guidance.

John Podesta's Center for American Progress had a ready answer: spend $100 billion on green energy. Was it good environmental policy? Probably. Was it the most effective way to create jobs in a 12-24 month window? No.

The result: "It created net zero jobs by 2010." Democrats got good long-term policy but poor short-term politics—and paid for it in the midterms.

A Path Forward

But here's where it gets interesting. Professor Fagan isn't calling for a return to boring centrism. Political parties need political think tanks. Politics won't—and shouldn't—disappear from policymaking.

Instead, he argues we need more organizations like the Center on Budget and Policy Priorities, the Niskanen Center, and the R Street Institute. These organizations are explicitly ideological but maintain epistemic integrity. They work on issues where their side actually has the facts on their side. They don't fabricate data or torture evidence.

The key insight: "Don't try to take the politics out of information, but try to get it right. Try to be political, to fund political causes that are acting using information that is as close to the truth as we can get."

What This Means for AI Governance

As someone working at the intersection of AI policy and institutional trust, this conversation crystallized something critical: we're about to repeat these mistakes at algorithmic speed and scale.

When policymakers need to understand AI safety, algorithmic bias, or the economic impacts of automation, will they turn to neutral expertise? Or will they consult think tanks that tell them what they want to hear?

We're already seeing the foundation being laid. Partisan think tanks are producing AI policy recommendations. The question is whether they'll maintain the epistemic discipline to get it right, or whether we'll see another cycle of ideology trumping evidence.

The Stakes

Professor Fagan makes a bold prediction: the Heritage Foundation will be less powerful in his next book than it is today. Why? Because as organizations become more revolutionary and extreme, they lose touch with reality and become less useful to actual policymakers.

Project 2025 generated headlines and fundraising, but did it actually shape policy? Professor Fagan argues the Trump administration's domestic agenda looks more like white nationalist interest groups than Heritage Foundation reports.

The real power lies with organizations that can bridge the gap between political goals and empirical reality—organizations that help politicians solve the problems voters care about while staying grounded in what actually works.

What We Need Now

As we face questions about AI deployment, climate adaptation, and economic transformation, we need information infrastructure that serves democracy rather than undermines it.

We need funders willing to support evidence-based advocacy, not just neutral research.

We need organizations with the epistemic discipline to admit when their side is wrong on the facts.

And we need a renewed commitment to the idea that while Americans can disagree on values and priorities, we need to operate from a shared understanding of empirical reality.

The alternative is what Professor Fagan describes as the collapse of "the idea that we are one country with one set of ideas." That's a deeper problem than any policy disagreement—and it's the one we need to solve first.

Listen to the full conversation with Professor E.J. Fagan on the AI in Chicago Podcast, below.