🎆 The American Growth Machine

A Precision Analytica Research Note

We usually tell the American growth story as a parade of inventions. After the long slowdown of the 1970s, one engine followed another: the personal computer, business software, the internet, the smartphone, the cloud, social media, and now artificial intelligence. Each arrived just as the last was settling into middle age. That is the story we can see. The deeper story is harder to see, and it matters more. It is not that America keeps inventing things. It is that America keeps converting them.

Plenty of countries produce brilliant scientists, fine engineers, and clever inventions. The harder thing, the rarer thing, is to turn a possibility into a company, a company into an industry, an industry into jobs and wages and new wealth, and then to do it all again when that wealth grows old. Few economies manage that repeatedly at frontier scale. America has managed it over and over. The American growth machine is not mainly an invention machine. It is a conversion machine. At 250 years, that is the achievement worth understanding.

The car that keeps moving

The economist Charles I. Jones offers a useful way to see the puzzle. Growth, over the long run, cannot come from simply piling more machines onto the same factory floor. Add enough and each new one does less than the one before. A big economy is heavier to move than a small one. To keep its speed, it has to keep finding new ideas, new methods, new ways of working. And ideas, it turns out, get harder to find as you go. We put more people to the search and get less from each of them than we once did.¹

So here is the strange fact to explain. Through recessions, oil shocks, a financial crisis, a pandemic, and no shortage of political noise, the American economy has kept moving forward over the long haul. The road got harder. The car kept its speed. Why?

The answer is not one thing. It is a system. Let me describe it with a simple picture.

Engine, fuel, road, and driver

Technology is the engine. It is the part we point to, the part that makes the headlines. But an engine does not start itself, and many fine engines never leave the garage.

Capital is the fuel. In a market economy, capital is restless. It is forever looking for a better return than sitting still, and that restlessness is what carries a new idea from a sketch to a startup to a public company. Most of the bets fail. A few pay for all the rest. America is unusually good at making that wager at scale, at turning an uncertain future into something people will fund today.

Institutions are the road, the traffic lights, and the patrol.² Property you get to keep. Contracts a stranger will honor. Courts that settle the disputes. A bankruptcy law that lets you fail without being finished for life. These rules sound dull. They are the surface the whole race is run on. Too many rules and nothing moves. Too few and everything crashes. The American knack, when the system is healthy, has been to leave enough freedom to start and enough order to trust.

And then there are the people. Schools and a loud, argumentative civil society produce the drivers, and not only the famous founders. The driver pool is engineers and managers and salespeople and lawyers, and the critics who say it will never work. A growth economy needs people willing to believe something before the spreadsheet agrees, willing to leave a safe job, willing to look foolish. America has long produced people like that, partly because it lets them argue.

Immigration supplies the builders and the mechanics. We like to tell this as a story of celebrated founders born somewhere else, and some of it is. But the fuller truth is that the frontier rests on a base. Immigrants are disproportionately visible among high-growth founders and inventors, and foreign-born workers also make up a large share of the ordinary labor force that keeps the economy operating.⁴ The engineer needs a city to live in. The hospital needs nurses. The warehouse needs hands. A country can have the finest laboratories on earth and still seize up if no one will build the housing, staff the care, or move the goods. Immigration keeps the whole machine running, top and bottom.

Markets run the race, and failure clears the field.³ No committee decided in advance that the personal computer, or the internet, or artificial intelligence would be the next engine. The system did not need to know. It only needed to let many people try, fund the attempts, and test the results in the open.

The deeper code

Here is the part I find most worth saying on an anniversary. Underneath the road there is an operating system, and it was written at the founding.⁵

The founders did not build a perfect country. The early republic left out most of the people living in it. But they did something that turned out to matter enormously. They built a system that could be corrected. They made it updateable.

The Declaration was the kernel. It did not lay out an economic plan. It said something deeper, that legitimate power rests on equality and consent, and that a people may change institutions that fail them. That gave the country a standard it could be held to, in its own words, by every generation that came after. Reform became part of the design rather than a betrayal of it.

The Constitution was the architecture. It assumed people would disagree, and it built channels to handle the disagreement, so that conflict could be argued through rather than only fought out.

The Bill of Rights protected the space where argument lives. Speech, the press, assembly, worship, and limits on arbitrary power. A society that cannot tolerate dissent will struggle to produce the people who challenge what everyone already believes. And challenging what everyone already believes is where new things come from.

On top of that founding code, each generation has installed its own software. The research university. Public science. Venture capital and the startup. Civil rights. The open protocols of the internet. The point was never that each update was wise. Many were wasteful, or unfair, or half-finished. The point is that the system could take a new part without tearing down the whole machine. It could turn public research into private industry, dissent into enterprise, newcomers into citizens, and failure into experience worth having.

The newest test

Artificial intelligence is the latest engine bolted onto this machine, and so far it looks the way every new engine looks at the start: a great deal of building, a great deal of money, a great deal of excitement, and not yet a clear, broad gain in how much an ordinary worker can produce. That gain may come. It is too early to be sure.

But AI is not only a test of technology. It is a test of the operating system. The American system is very good at scaling a new engine. It is less obviously good at protecting people while the engine roars. A new technology can lift output and hollow out the first rungs of the career ladder at the same time. It can create fortunes and concentrate them. It can make us faster while quietly letting some of our skills go soft.

So the questions for this engine are the old questions, asked again. Can we fund real experiments and not just speculation? Can we spread the gains and not only the wealth? Can we keep teaching people to judge and to build, even as the machines do more of the routine work? Can we still produce drivers, and not only passengers? These are not side questions. They are the growth questions.

What the anniversary is for

A growth machine can run beautifully and still strain the society around it. That is the honest part of the story. Every new engine reshapes who has work, where opportunity lives, and how the rewards are shared. People will put up with that churn only as long as they believe the road is still open to them and their children, that the rules are roughly fair, and that the next generation can still get behind the wheel.

So a 250th anniversary is not really a victory lap. The founding made renewal possible. It never made it automatic. The road has to be repaved. The rules have to be kept fair. The people who build and repair the machine have to be treated as part of it, not as parts to be used up.

That, in the end, was the founders' deepest promise. Not that the country would be perfect. They knew it would not be. The promise was that it could be argued with, and changed, and made better than it was. The growth we celebrate has always rested on that quieter foundation.

Technology is the frontier. Capital is the fuel. Institutions are the converter. People drive and build and repair the whole thing. And the founding made it a machine that can be rebuilt while it runs. Whether it stays that way, in the age of intelligent machines, is the American question for the next 250 years.

Notes

1. Charles I. Jones, an economist at Stanford, studies why economies grow over the long run. The finding that we now devote far more effort to research while getting fewer new ideas per researcher is documented in Nicholas Bloom, Charles I. Jones, John Van Reenen, and Michael Webb, "Are Ideas Getting Harder to Find?" American Economic Review 110, no. 4 (2020). His broader case that lasting growth depends on a steady supply of new ideas appears in "Sources of U.S. Economic Growth in a World of Ideas," American Economic Review 92, no. 1 (2002).
2. That prosperity rests not only on resources but on the rules that govern how people own, trade, and cooperate is the central theme of the economic historian Douglass North, who won the Nobel Prize for this work. A short and readable statement is his essay "Institutions," Journal of Economic Perspectives 5, no. 1 (1991).
3. The notion that a market economy renews itself by constantly replacing old products, firms, and methods with new ones belongs to the economist Joseph Schumpeter, who called it "creative destruction" in Capitalism, Socialism and Democracy (New York: Harper & Brothers, 1942).
4. The role of immigrants in the American growth machine is visible both at the frontier and in the operating base. The National Foundation for American Policy reported in 2026 that immigrants founded or cofounded 59 percent of privately held U.S. billion-dollar startups. Shai Bernstein, Rebecca Diamond, Timothy McQuade, and Beatriz Pousada find in an NBER working paper that immigrants are 16 percent of U.S. inventors but produce 23 percent of patent output. The Bureau of Labor Statistics reports that foreign-born workers accounted for 19.1 percent of the U.S. civilian labor force in 2025.
5. The three founding documents referred to here are the Declaration of Independence (1776), the Constitution of the United States (1787), and the Bill of Rights (1791). Official transcriptions are published by the U.S. National Archives.