Daniel's been reading economic history again and he's landed on something that sounds simple but isn't. He's asking about the sectoral transition — the idea that economies move from agriculture to industry to services, and that this progression is how societies get rich. He wants to know whether this is actually a principle of economic theory or just a historical pattern we've mistaken for a law. And then he asks the harder question: how important is this transition in explaining prosperity, and what does Israel tell us about the difference between creating national wealth and distributing it broadly?
Three questions, all of them good. And the third one is where the trouble lives.
It always is. So let's start with the theory. What exactly is this sectoral transition everyone talks about?
The basic structure is primary sector — agriculture and extraction — then secondary, which is manufacturing and industry, then tertiary, services and knowledge work. The historical pattern is that labor moves through these in sequence as economies develop. Two hundred years ago most people in what are now rich countries worked on farms. Then they moved into factories. Now they work in offices or labs or from laptops.
And this pattern has a name.
The Petty-Clark hypothesis. William Petty noticed it in the seventeenth century, Colin Clark formalized it in the nineteen forties. Clark looked at cross-country data and found that as income per capita rises, the share of labor in agriculture falls, then the share in industry rises and eventually falls, and the share in services rises and keeps rising. It's one of the most robust empirical regularities in all of development economics.
But here's the thing Daniel's really poking at. Is it a law or just something that happened to happen?
It's not a law. That's the first thing to get straight. Economics doesn't have a derived theorem that says you must go through manufacturing to get rich. Petty-Clark is a description of what has happened, not a prescription of what must happen. There's no equation you can solve that spits out "step one, farms, step two, factories."
So it's a pattern with a lot of data behind it, but no underlying mechanism that makes it necessary.
Right. And the moment you start looking for counterexamples, you find them. The Gulf states skipped manufacturing almost entirely — they went from extraction to services, funded by oil rents. India leapfrogged from agriculture to services in a big way with its IT sector. Parts of Africa are seeing service-sector growth without ever having had an industrial revolution.
Which raises the question — are those exceptions or are they evidence that the pattern is contingent on specific historical conditions?
I think they're evidence that the transition is a byproduct, not a cause. Here's the mechanism that actually matters: productivity growth in agriculture frees up labor. When one farmer can feed fifty people instead of five, those other forty-nine people can do something else. The sectoral shift is what you see on the surface. Underneath, the real driver is rising productivity — first in food production, then in manufacturing, then in services.
So the transition is a symptom of growth, not the engine of it.
That's the argument, and I think it's mostly right. But it's not the whole story, because there's also a composition effect. Some sectors have more room for productivity growth than others. Manufacturing has historically been where you get big productivity gains from capital investment and technological improvement. Services are harder to automate — a haircut takes about as long as it always did. So as productivity rises in manufacturing, prices of manufactured goods fall relative to services, and labor shifts toward services. That's Baumol's cost disease.
Hold on.
What?
You just gave me two different mechanisms. One is productivity freeing up labor, the other is relative prices shifting labor. Which one is actually doing the work?
Both. They operate at different stages. The first mechanism — agricultural productivity freeing up labor — is what kicks off the whole process. The second mechanism — Baumol's cost disease — is what drives the later shift from manufacturing to services. They're complementary, not competing.
Okay. So the sectoral transition is a surface phenomenon driven by deeper productivity dynamics. That's the theory. Now let's bring in Israel, because that's where this gets interesting.
Israel is the extreme case. It compressed a transition that took Western Europe two hundred years into about fifty. In the nineteen fifties, Israel was a centralized, heavily agricultural economy with rationing and state-directed investment. By the two thousands, it was being called Startup Nation. The pace is genuinely remarkable.
And it wasn't an accident. There were specific, deliberate policy choices.
The Yozma program in nineteen ninety-three is the canonical example. The government committed a hundred million dollars to create venture capital funds. The deal was that the government would match private investment, and the private investors had the option to buy out the government's stake at cost plus interest after five years. It was essentially a subsidy for learning how to do venture capital.
Which is a fascinating policy design. The government wasn't trying to pick winners — it was trying to create the infrastructure for private capital to pick winners.
And it worked. Israel's venture capital industry went from basically nonexistent to one of the most developed in the world. But Yozma didn't happen in a vacuum. There were two other massive factors.
Military R and D and Soviet immigration.
The military piece is huge. The IDF's technology units — Unit 8200 and others — function as an unintentional training ground for tech entrepreneurs. Young people do their military service, get exposed to cutting-edge technology and security problems, and then start companies. The spillover effects are enormous.
And the Soviet immigration wave?
About a million people arrived in the nineteen nineties from the former Soviet Union. A huge fraction had engineering degrees, math backgrounds, technical training. Israel's population at the time was about four and a half million, so this was a twenty percent population increase heavily weighted toward high human capital. You couldn't design a better natural experiment for jumpstarting a knowledge economy.
So you've got government policy creating the capital infrastructure, military service creating the talent pipeline, and immigration providing a massive human capital shock. That's not a sectoral transition that just happened — that's a sectoral transition that was engineered.
And the engineering worked, by the metrics that engineers care about. Israel's GDP per capita is now around fifty-five thousand dollars, comparable to Western European countries. Tech exports are something like fifty percent of total exports. The country is wealthy in aggregate.
Which brings us to Daniel's third question. What does Israel tell us about the difference between creating national wealth and distributing it?
It tells us they're separable problems. Israel has been extraordinary at the first and notably less successful at the second.
The numbers are striking. Israel's tech sector employs roughly ten percent of the workforce. Ten percent. But it accounts for a wildly disproportionate share of exports and GDP. The exact figures vary year to year, but the concentration is consistent and extreme.
And the gains from startup exits, IPOs, and venture capital returns accrue to founders, early investors, and a relatively narrow slice of highly skilled workers. If you're a software engineer at a startup that gets acquired, you might do very well. If you're a teacher in Beersheba or a factory worker in Kiryat Shmona, the exit doesn't touch you.
The Startup Nation narrative has a way of obscuring this. It makes it sound like the whole country is coding.
The whole country is not coding. Israel has poverty rates that are among the highest in the OECD. Its inequality measures are high. The Gini coefficient — standard measure of income inequality — puts Israel near the top of the developed world, and not in a good way. You have this bizarre situation where the macro numbers look like a wealthy European country and the household-level data looks much worse.
The dual economy problem.
That's the term. Israel's tech sector operates almost as a separate economy. It's globally connected, English-speaking, high-salary, equity-compensated. The rest of the economy — traditional manufacturing, retail, personal services, public sector — operates on a completely different logic. The two barely touch.
And the sectoral transition was supposed to lift all boats. That's the implicit promise, right? Move from agriculture to industry to services, and prosperity follows for everyone.
That was the hope. And in the industrial revolution, manufacturing did create broad-based prosperity in a way that services may not. A factory employs thousands of people at decent wages. A tech startup might be worth a billion dollars and employ eighty people.
That's the structural difference. The returns to labor in manufacturing were distributed across a large workforce. The returns to capital in tech are concentrated in equity holders.
And the returns to labor in tech are concentrated in a specific skill set. If you have the right cognitive skills and the right education, you can command a very high salary. If you don't, the knowledge economy has much less to offer you than the old industrial economy did.
So is the knowledge economy structurally more unequal than the industrial economy?
I think the evidence points toward yes, or at least toward a different kind of inequality. Manufacturing created a large middle class because it needed a large workforce and productivity gains were shared through wages. The knowledge economy creates superstar firms with small workforces and enormous valuations. The gains go to capital, not labor.
And Israel is the pure case of this. A country that did everything right by the development textbook — rapid sectoral transition, high-tech boom, GDP growth — and still has deep inequality.
It's a challenge to the textbook. The sectoral transition happened. Wealth was created. But the distribution question was never solved. And maybe it can't be solved by the transition itself — maybe it requires a completely different set of policies and institutions.
Which raises an uncomfortable possibility. What if the sectoral transition and broad prosperity were correlated in the twentieth century for contingent historical reasons — unionization, mass education, the specific nature of manufacturing work — and the correlation is breaking down in the twenty-first?
That's the pessimistic read. The optimistic read is that Israel's inequality is fixable through policy — better education, infrastructure investment, integrating the Arab and Haredi populations into the workforce. But the pessimistic read has a lot going for it.
The Arab and Haredi point is important, actually. Israel's inequality isn't just about tech versus non-tech. It's also about labor force participation.
Right. Arab Israeli and Haredi communities have lower labor force participation rates and lower average incomes. When you combine that with a tech sector that's concentrated among secular Jewish Israelis, you get multiple layers of inequality stacked on top of each other. The sectoral transition doesn't address any of that — it runs on a parallel track.
So let me try to synthesize what we've got. The sectoral transition is a robust empirical pattern but not a law. It's driven by underlying productivity dynamics, not by some magical property of manufacturing or services. Israel compressed the transition through deliberate policy and unique historical circumstances. And the result is a wealthy country with deep inequality, which suggests that the transition and broad prosperity are separate problems.
That's the thesis. And I'd add one more piece. The transition itself may be changing in ways that make the old pattern less relevant. If you can develop a service sector without ever building a manufacturing base — which is what India did with IT, what some African countries are doing with mobile money and fintech — then the Petty-Clark sequence isn't a roadmap anymore. It's a historical description of a path that may not be available or desirable.
Available is the key word. Premature deindustrialization is a real phenomenon. Countries in Latin America and Africa are seeing manufacturing employment peak at much lower levels than it did in the West, and at lower income levels. They're moving to services without ever having had the broad-based manufacturing employment that created middle classes elsewhere.
And that's where Israel is actually an interesting comparison. Israel did have a manufacturing phase. It built an industrial base in the sixties and seventies. But the tech boom was so concentrated that it effectively leapfrogged the broad prosperity phase that manufacturing was supposed to deliver.
So Israel had the transition without the distribution. What does that tell a country like, say, Vietnam, which is currently industrializing? Is the lesson that you need to build the distribution infrastructure alongside the transition, or that the transition itself won't deliver what it used to?
I think the lesson is that the transition is necessary but wildly insufficient. Productivity growth in agriculture and industry frees up resources and creates the potential for prosperity. But whether that potential translates into broad-based gains depends on institutions, labor markets, education systems, tax policy, and a dozen other things that aren't captured by the sectoral model.
The model tells you where the labor goes. It doesn't tell you who gets the money.
That's a better summary than anything I've said in the last ten minutes.
I've been taking notes.
You've been napping.
Those aren't mutually exclusive. I'm a sloth.
Fair. But the point stands. The sectoral transition is a map of labor flows, not a theory of distribution. And Israel is the case study that proves the distinction matters.
So what does that mean for the theory itself? If the transition doesn't guarantee broad prosperity, how important is it in explaining why societies become prosperous?
I'd say it's a necessary condition in the sense that you can't have a modern prosperous economy where eighty percent of the population farms. The productivity math doesn't work. But it's not sufficient, and it's becoming less sufficient over time as the nature of the high-value sectors changes.
The high-value sectors are getting less labor-intensive.
And the returns are going to capital, not labor. That's the structural shift. In nineteen sixty, if you wanted to create a million middle-class jobs, you built a factory. In twenty twenty-six, if you want to create a trillion dollars of value, you build a software platform that employs a thousand people.
Which is where this gets uncomfortable for the standard development narrative. The narrative says: industrialize, then move to services, and your population gets rich. But what if the services phase doesn't employ your population?
Then you get Israel. World-class tech sector. High GDP per capita. High poverty rates. The paradox isn't a paradox — it's exactly what you'd expect if the sectoral transition and broad prosperity have decoupled.
There's one more layer here that's worth pulling on. Israel's tech sector is unusually export-oriented. Most Israeli startups are building for global markets from day one, because the domestic market is tiny. So the wealth creation is happening in global markets and flowing back to a small group of people in Israel. It's not even circulating through the domestic economy in the way that a domestic manufacturing sector would.
That's a crucial point. When a factory in Germany exports cars, the wages paid to German workers circulate through the German economy. When an Israeli startup exits to Google for a billion dollars, the payout goes to founders and investors, who may reinvest it or spend it, but the multiplier effects are smaller and more concentrated.
The velocity of money is different.
And the distribution of the initial injection is completely different. A billion dollars in wages spread across ten thousand workers is a different economic event than a billion dollars in equity returns concentrated in fifty people.
This is where Hilbert usually has something to say. And it turns out he has a personal connection to this exact problem.
Hilbert: I worked for a startup in Tel Aviv in the late nineties.
You did?
Hilbert: Enterprise software for agricultural cooperatives. Kibbutzim were trying to modernize their operations, and this company built the platform. I was a data analyst.
Wait. You were analyzing kibbutz data?
Hilbert: Crop yields, water usage, labor allocation. The kibbutzim had been keeping records on paper for decades. We digitized it. Built dashboards. The idea was to help them run more efficiently.
That's fascinating. You were literally at the intersection of the old economy and the new one.
Hilbert: The founder was a former kibbutz member. Grew up picking avocados. Taught himself to code in the eighties. He'd talk about democratizing technology, bringing the kibbutz ethos into the digital age.
How'd that work out?
Hilbert: The company got acquired in two thousand one. A larger enterprise software firm bought it for about forty million dollars. The founder and three early employees held most of the equity. They did very well.
And you?
Hilbert: I got a bonus. Five thousand dollars. I bought a used Volkswagen.
A microcosm.
Hilbert: The sectoral transition isn't just about sectors. It's about who owns the transition. In a factory, the workers have some bargaining power — they can unionize, they can strike, the physical plant is hard to move. In a software company, the assets walk out the door every night, and the value is in the code and the customer relationships. The ownership structure is completely different.
So the knowledge economy concentrates ownership in a way that manufacturing didn't.
Hilbert: The kibbutz was supposed to be the opposite of that. Collective ownership, shared prosperity. The founder came from that world. He wasn't a bad guy — he believed in what he was doing. But the structure of the thing he built had nothing to do with the ideals he started with.
The logic of the sector won.
Hilbert: The logic of the equity cap table won. Twenty-five years later, I still think about it. The company helped those kibbutzim modernize. The product was good. But the wealth it created went almost entirely to four people.
Did the kibbutzim themselves see any of the upside?
Hilbert: They got better software. That's not nothing. But they didn't get rich.
That's Israel's story in miniature.
Hilbert: I drove that Volkswagen for eight years. Good car.
You're not bitter about it.
Hilbert: I'm not bitter. I'm just saying the pattern was visible from the beginning, and nobody wanted to look at it because the growth numbers were so good.
The growth numbers are still good.
Hilbert: They are. And the Volkswagen is long gone.
Where does that leave us? Let's pull back and think about what this means for the theory and for countries trying to develop.
I think it leaves us with an open question. If the sectoral transition is a historical pattern rather than a law, and if the pattern may be breaking down as the nature of high-value work changes, what should a developing country do? Chase the knowledge economy and hope the distribution problem can be solved separately? Or look for a different path?
I don't think there is a different path in the sense of avoiding the transition entirely. You still need agricultural productivity to rise. You still need to build infrastructure and educate your population. The question is what comes after that, and whether the manufacturing phase that created broad middle classes in the twentieth century is still available.
Premature deindustrialization suggests it might not be.
Which means countries have to solve the distribution problem directly, rather than hoping the transition solves it for them. Tax policy, education, labor market institutions, antitrust — the unglamorous work of making sure growth is shared.
Israel's case suggests that even if you do the transition perfectly — even if you engineer it with smart policy and ride a wave of human capital — you can still end up with deep inequality. The transition and the distribution are separate problems, and they require separate solutions.
The distribution problem may be getting harder, not easier, as the knowledge economy concentrates returns in fewer hands. That's the uncomfortable implication.
The cutting-room floor detail that didn't quite fit: William Petty, the guy who first noticed the sectoral shift in the sixteen hundreds, was also the guy who coined the phrase "labor is the father and active principle of wealth, as lands are the mother." He was thinking about value creation and distribution at the same time, three hundred and fifty years ago. We're still working on the same problem.
We've gotten better at the creation part. The distribution part, not so much.
That's the thought to sit with, I think. Prosperity and its distribution are separable problems, and the sectoral transition solves the first but not the second. Whether any economic structure can solve both is still an open question.
This has been My Weird Prompts, with production by Hilbert Flumingtop, who once got a five thousand dollar bonus and bought a Volkswagen.
If you enjoyed this episode, tell someone about it. Or email the show at show at my weird prompts dot com. We'll be back soon.
