The Ending of the American Deal: Political Fallout

Last week I wrote about the American deal.

The unwritten American contract — that what you can do, what you know, what you contribute, can change the circumstances of your life. And about Elon Musk’s prediction that we have roughly three years before artificial intelligence closes that deal permanently, leaving behind a world divided between asset owners and everyone else.

The response was significant. And the question that came back most often wasn’t about the technology or the economics.

It was about the politics. What happens to democracy when you have an owner class and a dividend class? What does power look like in that world? Who runs for office, who votes, who wins, and in whose interest does government operate when the population is divided not by party or ideology but by whether they own productive assets or receive a stipend from the people who do?

Let’s start with what we already know. Because the two-class society Musk is describing isn’t entirely theoretical. We have a preview of it. We’ve been living in an early version of it for the past thirty years — a period in which asset ownership has become increasingly concentrated, wages have stagnated relative to productivity, and the political system has reorganized itself around the interests of capital rather than labor.

The result has not been pretty. What we’ve watched happen to American politics in the era of extreme wealth concentration is not an accident. It is the predictable political consequence of extreme economic stratification. When enough people feel that the system no longer works for them — that their effort doesn’t translate into security, that the rules favor people who already have — they don’t quietly accept their circumstances. They get angry. They look for someone to blame. And they become available to anyone who offers a compelling story about who took what from them.

That dynamic has been the engine of American politics for a decade. It isn’t going away. Under the scenario Musk describes, it accelerates dramatically.

Here is the political math of the two-class world. The owner class will be small. Historically, the concentration of productive asset ownership in any society follows a power law — a small number of people hold a disproportionate share. In Musk’s framework, where the relevant assets are real estate, precious metals, and energy resources, that concentration is already advanced. AI and automation will not democratize asset ownership. They will accelerate its concentration. The dividend class will be large. Potentially very large. If automation displaces the broad middle of the labor market — not just manual labor but cognitive work, creative work, professional work — then the population receiving some form of universal income could constitute a substantial majority of adults.

Now ask yourself: what does electoral democracy look like when one class is small and the other is large?

The naive answer is that the large class wins. They have the votes. The honest answer is considerably more complicated.

Money has always purchased political influence in ways that votes alone cannot counter. We learned this the hard way over several decades of watching campaign finance law erode, lobbying expand, and regulatory capture deepen. The mechanisms by which concentrated wealth translates into political power are well developed, well funded, and highly efficient. A dividend class is, by definition, financially dependent on the owner class and on the political structures that distribute the dividend. That dependency is not politically neutral. It creates a structural incentive — not for gratitude exactly, but for a particular kind of political caution. This is not a new dynamic. It is the oldest dynamic in political economy. What is new is the scale at which it could operate. A UBI administered by the state, funded by taxes on AI-generated productivity, distributed to a majority of the population, creates a relationship between the government and the governed that is unlike anything in American political history.

It is a relationship of dependency. And dependency, historically, is the most reliable tool of political control ever devised. There is a word for a political system in which a large dependent population receives material support from a concentrated ownership class in exchange for political acquiescence.

I want to be careful here. I am not predicting that a UBI society becomes an authoritarian one automatically. The relationship between economic structure and political form is complex and contested. Democracies have survived significant inequality before. But I am saying this: a society in which the majority of people are financially dependent on structures they do not control, administered by governments that are themselves controlled by the owner class, is a society whose democratic institutions will face pressure unlike anything they have faced before. The pressure will not come from a coup or a revolution. It will come from the slow erosion of political independence that follows economic dependency. People who cannot afford to dissent — whose stipend is contingent on behavior, however implicitly — will not dissent.

The second political consequence is subtler and in some ways more dangerous. Democratic politics requires a shared sense of stakes. It requires that enough people believe the outcome of an election matters to their lives in concrete ways — that if the other side wins, something real changes for them. That belief is what drives participation, mobilization, and ultimately the accountability that makes democracy function.

In the two-class world, what are the stakes of an election for the dividend class? If your income is guaranteed regardless of who wins — if the UBI continues under any government because the alternative is social collapse — then the instrumental case for political engagement weakens. Not disappears. But weakens.

And weakened political engagement among the large dependent class is, again, not politically neutral. It systematically advantages the small owner class whose stakes in every election are concrete, immediate, and enormous. A tax rate. A regulatory structure. A trade policy. An energy concession. For the owner class these are not abstractions. They are the difference between vast wealth and slightly less vast wealth. They will organize. They will donate. They will lobby. They will show up.

The dividend class, absent a galvanizing issue, may not. That asymmetry, sustained over electoral cycles, produces governments that represent the owner class regardless of which party holds power. We have seen this movie before. We are watching it now. The two-class world simply runs it at a scale that makes the current version look like a rough draft.

So what is the alternative?

I don’t think the answer is to pretend the technology isn’t coming. It is. I don’t think the answer is to ban AI or reverse automation. That is not available. I think the answer begins with being honest about what we are actually choosing — because this is a choice — when we design the political economy of the AI transition. A UBI that maintains dependency is a choice. An asset distribution that broadens ownership is a different choice. A governance structure for AI productivity that gives citizens a stake in the systems displacing them — through sovereign wealth funds, through cooperative ownership models, through mandatory equity distribution — is yet another choice.

These are not easy choices. They require the kind of political will that concentrated wealth is specifically designed to prevent. But they are choices. And the window in which they can be made — before the two-class structure calcifies into something permanent and self-reinforcing — is not unlimited.

Here is what I know from having covered American politics and finance for thirty years, and from having run for office, and from having built businesses in the real economy: Systems do not reform themselves. They are reformed by people who understand the stakes clearly enough, and care about the outcome enough, to do the hard work of changing them.

The stake here is not abstract. It is whether the next generation of Americans lives in a republic or a managed dependency. Whether democracy means something or becomes a ceremony performed over a structure that has already decided.

That question is being answered right now. Not in Washington. Not in Silicon Valley. In the choices being made today about who owns what, who decides what, and who gets to participate in the future being built.

The owner class knows what it wants.

The question is whether everyone else figures it out in time.