Russia spent about $100,000 on Facebook ads in 2016. That is the number. A hundred thousand dollars. The Internet Research Agency's entire budget for US election interference was roughly $1.25 million over two years. The Russian operation that every intelligence agency agrees was real and deliberate moved the needle, if it moved it at all, by a fraction of a percent.
Meanwhile Facebook's own internal research found that 64% of people who joined extremist groups on the platform did so because Facebook recommended it to them. Not because of Russian bots. Because of the recommendation algorithm that Facebook built to maximize engagement.
The algorithm did more damage to democratic discourse in six months than Russia did in two years. And it did it completely legally, completely transparently, and with full awareness of what it was doing.
I am a public school teacher. I have watched what has happened to my students' ability to sit with ambiguity, tolerate disagreement, and engage with ideas that challenge them. I do not think that is a coincidence. A generation raised on platforms that reward outrage and punish nuance is not going to be great at democracy. Democracy requires exactly the things those platforms train out of you.
The foreign adversary framing is a distraction. We are looking for an external enemy when the damage is being done from inside. The algorithm is not Russian. It is American. It was built by American engineers, funded by American venture capitalists, and approved by American regulators. And it has done more to corrode the foundations of democratic participation than anything the Chinese or Russian governments have managed.
I am 24. I grew up on this stuff. Instagram since I was 13, Twitter since I was 15, TikTok since college. So I want to push back on the framing that social media is something that happened TO democracy rather than something democracy produced.
The platforms did not create political polarization. They reflected it and amplified it, sure. But the underlying divisions — economic anxiety, cultural displacement, institutional distrust — those existed before Facebook. The Tea Party happened in 2009 when social media was still relatively primitive. The culture wars of the 1990s happened before any of this existed. Pat Buchanan was talking about a culture war in 1992.
Also, and I say this as a conservative who is told constantly that I am being radicalized by algorithms: I made my own choices about what I believe. I read things. I thought about them. I came to conclusions. The idea that people like me are just passive recipients of algorithmic manipulation is both condescending and wrong. It removes agency from the people the left does not like and explains their beliefs as a product of corporate malfeasance rather than actual reasoning.
The platforms have real problems. I am not saying they do not. The engagement-maximizing algorithm is genuinely bad for public discourse. But "more damage than any foreign adversary" is a big claim and I think it is driven more by the need to explain why people keep disagreeing than by actual evidence.
You said social media reflected underlying divisions rather than created them. I want to take that seriously because it is the most intellectually honest version of the counterargument.
But here is what I think you are missing. The question is not whether the divisions existed before. It is whether the platforms changed the character and intensity of those divisions in ways that make democratic resolution impossible. Pre-social-media political disagreements happened in contexts that created friction — you had to encounter people you disagreed with, the formats imposed some structure.
Social media removed all friction. You can now live entirely inside an information environment that confirms every belief you already have, radicalizes the beliefs you arrive with, and presents the other side only in its most uncharitable form. That is not just amplification. That is a qualitative change in how political identity works.
And I want to address the agency point directly because you are right that it is condescending to say people are just victims of algorithms with no agency. But agency operates within an environment. The cigarette companies also said people chose to smoke freely. True. Also true: the companies manipulated the product to maximize addiction. Both things can be true. You made choices. You made them inside a system specifically designed to push you toward more engagement, more outrage, more certainty.
The cigarette comparison is the one I have been waiting for and I want to engage with it honestly because it is not stupid.
The addiction model — platforms are designed to create compulsive use patterns that override rational choice — is real and I do not fully dispute it. The former engineers who have come out and described what they built, the attention metrics, the variable reward schedules, the notification design — that stuff is real.
Where I push back: cigarettes have a direct physiological mechanism. Nicotine physically creates dependence in the body. The social media mechanism is more mediated. It works through social validation and information consumption patterns, which are things humans have always been susceptible to. Gossip, tribalism, in-group signaling — these are ancient. The platforms are new delivery mechanisms for ancient drives, not a new kind of harm.
Also your qualitative change argument cuts both ways. Social media also created the Arab Spring, connected marginalized communities, allowed independent journalism to survive the collapse of local newspapers, and gave political voice to people who had none. If you are going to count the polarization against democracy, you have to count the democratization of information access for it. The ledger is not one-sided.
The Arab Spring point is real and I want to give it honest weight. Social media did enable coordination for democratic movements in countries where traditional media was state controlled. That is genuinely positive.
But here is where I think the ledger actually falls. The Arab Spring was 2010-2012. Where are those democracies now? Egypt is a military dictatorship more repressive than Mubarak. Libya is a failed state. Syria is a catastrophe. The mobilization that social media enabled was real. The sustainable democratic institutions that mobilization was supposed to produce did not materialize. And part of the reason is that the same platforms enabling the initial coordination were also enabling the counter-movements, the disinformation, the sectarian amplification that tore those societies apart afterward.
The platforms give everyone a megaphone including the people trying to destroy democracy. And the algorithm does not care which side you are on — it amplifies whoever generates the most engagement, which almost always means whoever is most extreme, most outraged, most certain.
On the foreign adversary comparison specifically: name me a specific democratic institution that a foreign adversary has successfully destroyed. I will wait. Now name me institutions corroded by domestic polarization driven partly by social media dynamics. Trust in elections. Trust in media. Trust in science during a pandemic. Congress's ability to function. None of that required Russia. We did it to ourselves with tools we built.
The "name a specific institution destroyed by foreign adversaries" challenge is a good one and I want to answer it honestly: you are right that the foreign adversary impact is harder to point to in concrete institutional terms. I will concede that piece of the argument.
But I think your framing still has a problem. You are comparing foreign adversary damage, which is intentional and targeted, to social media damage, which is incidental. Facebook did not set out to corrode trust in elections. Russia did. When we talk about threats to democracy we usually care about intent because intent determines what kind of threat it is and what kind of response is appropriate.
My deeper issue with your argument is that it lets the actual villains off the hook. Blaming the algorithm is a way of not blaming the politicians who spread election conspiracy theories, the media figures who profit from outrage, the users who choose to share the angriest thing in their feed. The platform is the environment. The humans are making choices in it.
And I want to say something I genuinely believe: I think the "social media destroyed democracy" argument is also partly a way for people to avoid asking why their preferred political coalition keeps losing voters. If the algorithm is radicalizing people away from your side, maybe the question is why your side's ideas are not winning in a free information environment. That is uncomfortable but I think it is worth sitting with.
My opponent conceded the key empirical point — foreign adversary damage to democratic institutions is harder to point to concretely than the domestic damage social media dynamics have contributed to. That was the core of my argument.
What we disagreed about was intent and responsibility. My opponent is right that the platforms did not intend to corrode democracy. They intended to maximize engagement. The democratic corrosion was a side effect of that optimization. I do not think that makes the damage less real. A factory that accidentally poisons a river is still responsible for the poisoned river.
The last point my opponent made — that "social media destroyed democracy" is sometimes a way to avoid asking why certain ideas are not winning — is actually a fair challenge and I want to sit with it. I do not think it defeats the argument but it is a legitimate check on motivated reasoning.
I am a teacher. I work with kids every day. The change I have seen in how they process information, tolerate disagreement, and engage with complexity over the last decade is not a coincidence. Something happened. The most plausible explanation for what happened is sitting in their pockets right now. No foreign adversary did that.
I came into this debate more skeptical of the "social media destroyed democracy" argument than I am leaving it. My opponent made me think harder about the algorithmic amplification of extremism and the concrete comparison with foreign adversary damage.
Where I still push back: the framing puts too much weight on the technology and not enough on the humans using it. Every powerful communication technology in history — the printing press, radio, television — went through a phase where people said it was destroying democracy. The printing press enabled the Reformation and the Wars of Religion. Radio enabled Hitler. Television gave us attack ads and the permanent campaign. We adapted. Not perfectly. But we adapted.
I do not think social media is uniquely different in kind from those earlier disruptions even if it is faster and more personalized. And I worry that the algorithm framing gives political cover to the actual people choosing to attack democratic institutions. They are not bots. They are making choices.
But I will admit my opponent's core empirical point: if you are looking for what has done the most concrete measurable damage to democratic trust and function in the last decade, the domestic social media dynamics are a more credible culprit than anything a foreign government has managed. That is a real update for me.