In 1988, a political consultant named Roger Ailes helped produce an ad about a convicted murderer named Willie Horton. The ad ran in the final weeks of the presidential campaign and is widely credited with helping sink Michael Dukakis's candidacy. It was not the first political attack ad. But it was among the most carefully studied, and what the political consulting industry took from it wasn't just a tactical lesson. It was a data point in an emerging science.
Fear works.
Over the next 35 years, that data point was confirmed, refined, and operationalized at industrial scale. Today, the American political advertising industry is not primarily a communications industry. It is a fear-delivery system of extraordinary precision, funded by $11 billion per election cycle and operated by a professional class that has spent decades learning exactly which anxieties, in which combinations, delivered through which channels, produce the highest return.
The A/B Test That Built an Industry
The first thing to understand about modern political advertising is that almost nothing in it is accidental.
Every major campaign runs extensive message testing before a dollar goes to broadcast. Focus groups identify which attack lines land and which don’t. Online surveys test which framings produce the strongest emotional responses. Digital campaigns A/B test ad creative in real time — running two versions of an ad simultaneously, measuring which generates more clicks or donations, and automatically shifting spend toward the winner.
What this optimization process has consistently found, across thousands of campaigns over three decades, is that negative emotional framing — fear, anger, disgust — outperforms positive emotional framing by significant margins.
The result is an advertising environment calibrated to produce maximum negative emotional activation in the viewer. Not because anyone sat down and decided to make American political culture angrier and more anxious. But because the optimization process, run at scale, converges on the same destination: whatever makes you feel worst about the other side.
Who Actually Profits from Toxic Politics
The popular narrative about political polarization focuses on the politicians. But the financial beneficiaries are mostly not politicians. Most politicians would, in fact, prefer a less polarized environment — governing is genuinely harder when the other party views you as an existential threat rather than a political opponent.
The people who profit directly from political toxicity are several layers removed from the candidates.
Television stations and their parent companies. Political advertising operates under must-carry rules in the U.S.: broadcast stations must accept political ads at the lowest unit rate during the 45 days before a primary and 60 days before a general election. This makes political advertising a reliable, protected revenue stream for local broadcasters. In the final weeks of a competitive cycle, political ads can represent 30-40% of local TV revenue. The more contentious the environment, the higher the demand, the higher the rates for non-political advertisers who get priced out.
Digital platforms. Meta, Google, and the connected TV ecosystem collectively collected roughly $4 billion in political advertising revenue in 2024. Their platforms are designed to maximize engagement, and political content — particularly outrage-generating political content — is among the highest-engagement content they host. Political advertising and political organic content reinforce each other on these platforms in ways that benefit the platforms financially regardless of what it does to the public sphere.
The consulting and vendor ecosystem. Media buyers, creative agencies, data vendors, polling firms, and opposition research shops collectively capture an estimated 10-15% of total cycle spending. Their revenue scales directly with the total size of the market. A less polarized environment with less competitive races would mean a smaller market. Their structural incentive is for the market to grow.
The Geography of Fear
Political advertising is not distributed evenly. It concentrates in competitive markets with a precision that reveals exactly how the industry understands its own purpose.
In the final week of October 2024, Atlanta aired over 1,000 political ads per day — roughly one every 86 seconds on local broadcast stations. Philadelphia saw 862. Phoenix saw 729. In many of these markets, political advertising accounted for more than 60% of all advertising during prime time.
Meanwhile, voters in non-competitive states saw virtually no political advertising at all. A voter in deep-red Wyoming or deep-blue Massachusetts was essentially invisible to the political advertising industry — not worth spending money on because their vote, in the aggregate, was already determined.
The political advertising industry is not primarily in the business of persuading voters. It is in the business of activating already-committed voters in already-competitive places by making them afraid of what will happen if they don’t turn out.
This is not a small distinction. Persuasion-focused advertising would look very different from fear-activation advertising. It would be more informational, more comparative on substance, less emotionally extreme. The fact that modern political advertising is overwhelmingly fear-based is a direct result of what the optimization process discovered works. Persuading a genuinely undecided voter is hard and expensive. Terrifying a committed partisan into showing up costs less.
The Hidden Cost
The $11 billion figure captures what campaigns and outside groups spent. It doesn’t capture what the rest of us paid.
The ambient effects of a fear-based political advertising environment — on civic trust, on social cohesion, on the simple capacity to disagree with someone without viewing them as a threat — are real but unquantified. The research on this is not subtle: sustained exposure to threat-based political messaging increases in-group/out-group thinking, decreases willingness to engage across political lines, and produces measurable increases in political anxiety.
Put plainly: the industry is externalizing costs onto everyone who lives in the political environment it creates. The consulting firm that earned $7.5 million placing Senate ads in Pennsylvania in 2024 did not pay for the degradation of civic life in Pennsylvania. The voters did. The neighbors who can no longer have a Thanksgiving dinner did.
This is the definition of a market externality — a cost generated by economic activity that falls on parties who didn’t participate in the transaction. The political advertising industry is, in this sense, one of the most effective generators of negative externalities in the American economy. Nobody has figured out how to bill them for it.
This is the third installment of "The Incentive Machine," a series examining the structural forces that shape American political spending. The series continues with an analysis of power in Washington in "Why Congress Feels Broken Even When Smart People Are Elected."