An Interview with Paul Romer on Economic Growth
By Paul Romer
In the summer of 2007, I interviewed Paul Romer of Stanford University
for EconTalk. Paul has been the driving force behind the “new growth theory” writing a series of influential papers that put the role of ideas at the center of growth theory. Our conversation focused on the importance of ideas, the role of institutions and the legal environment for creating incentives for new ideas and how ideas spread beyond national borders helping people around the world. Here are edited excerpts from that conversation.
The podcast can be found at Romer on Growth.
Paul Romer: This is a classic application of the power of compounding, that if you have a slightly higher growth rate, as growth rates compound over many years, it leads to dramatically higher levels of income.
Paul Romer: For example, at 2.1 percent rate of growth per year, income per capita in a nation can increase by a factor of about 8 over 100 years. So if income per capita is $30,000 per person in the United States, just round numbers, in 100 years, it could be $240,000 per person. Now imagine you had a slightly faster growth rate. Suppose it was 2.6 percent instead of 2.1 percent. Then it could increase by a factor of 13 instead of 8, so you’d end up with $390,000 per person instead of $240,000—
Russ Roberts: Huge difference. Almost twice as much.
Paul Romer: Yeah, so small differences, half a percent per year, accumulate into very large differences in standards of living and the extra income will make life better for everyone in a nation—no matter what it wants to do with those extra resources—spend them on education, more vacations, more quality time with each other, whatever.
Russ Roberts: How much richer are we in the United States than we were 100 years ago? How have we been doing?
Paul Romer: Well, I picked the 2.1 percent because that’s about the average rate of growth that we had over the last century and we did experience about an eight-fold increase over a period of 100 years. It’s a really phenomenal increase and something that’s unprecedented in human history. We’ve never seen the leading nation in the world grow at such an astonishing rate. Now if you listen to the news reports, you’ll know that there’s some places like China which are growing at 10—8 and 10 percent per year. That’s because they’re catching up with the frontier and they’re starting from way behind. When you start from a low base, it’s always easy to have a faster rate of growth. But as they start to catch up with us, their growth rates will slow down.
Russ Roberts: You mean they’re not going to pass us in five or six years per capita?
Paul Romer: They’re certainly not going to pass us in five or six years. No, and they might not pass us ever.
Russ Roberts: I understand.
Paul Romer: We’ve had historical episodes where the United States started out below the UK and it’s also why I picked a half a percentage point change. We grew over the twentieth century by about a half a percentage point more per year than the Brits, and we went surging ahead of them in terms of income per capita.
So sometimes a nation that’s behind can grow faster and can go surging ahead of what was then the leading technology nation in the world. But right now, the United States is still a real powerhouse in terms of developing technology and sustaining growth.
Russ Roberts: You picked China as an example and you said their base was low, so it’s easier for them while they’re catching up to grow at a faster rate, but it’s not just that their base is low. It has something also to do with the reason that their base was low, which was both the institutions and the technology that they were using at that low level. Correct?
Paul Romer: That’s right.
Russ Roberts: Because I’m interested in that American example. Why do you think we’ve grown at a faster rate in America over the last 100 years than England? Technological opportunity in both countries are relatively similar.
We don’t have access to secret technology that they didn’t know to use. What are some of the reasons they might have grown more slowly than we did? And other nations, for that matter, over the same period.
Paul Romer: Let me answer a different question and then come back to that one. The easy question to answer is, “Why is China growing so fast?” If you look back when the United States was at the same level that China is at, we were growing more slowly. How can they grow so fast?
Well, they have the advantage of being able to import—essentially just copy—technology that already exists in places like the U.S. and adopt it very rapidly. So when they opened their economy, they could rapidly take advantage of things that were already known. Now the question about the U.S. versus the UK is a harder question about not copying things that are already known but developing brand-new things. In the United States, we developed a set of institutions. Institutions are the rules—the rules of the game that structure what everybody does in the nation.
We developed a set of institutions which encouraged more rapid discovery. We discovered and implemented things more rapidly than they did in the UK. And the interesting question that historians and economists are still struggling with is, “What were the precise details of our institutions that made them better, just enough better to get an extra half a percentage point per year compared to the British?”
Russ Roberts: Of course, one of the reasons that we’re able to do it is that we import people. You talked about China importing technologies. We imported—because we have relatively open borders, we’ve imported a lot of smart people from overseas that have helped us. But I want to stop before we get into that and ask you to clarify something more basic. You attribute the growth in the United States—the growth rate in the United States being above that of England’s—to our ability to develop new stuff. Yet how important is it for a nation to be the developer rather than the importer? Do we really care if the television was invented in the United States versus Russia? Do we really care where the car was invented or where the next great breakthrough comes from? Does it matter?
Paul Romer: No, we don’t and it’s a really good point to emphasize. People often use these national comparisons as if it’s a race where there’s winners and losers. I often tell my students, “Any time you’re thinking about rivalry between countries, reframe the question as rivalry between states in the United States.”
Would it upset us if we lived in Illinois and Intel was making microprocessors in California? Is it bad for us that Intel develops microprocessors? Of course not. We’re glad they make them and we’re happy to use them. We get the benefits from using them. Does it make people in Illinois any worse off if people in California grow richer or develop new technology? Absolutely not. People in Illinois are better off being able to trade with California and people in California are better off being able to trade with people in Illinois and New York and the rest of the country.
So this notion that there’s a kind of a rivalry with winners and losers when we think about nations, it’s really very misleading about the underlying economics. This, by the way, was one of the advantages the United States had in the early part of the twentieth century. We were already a big free trading block when a lot of the world was still relatively closed.
Russ Roberts: We had a big physical area. What about our level of population and human capital at the beginning of the twentieth century? Any thoughts on that?
Paul Romer: I think that we were already overtaking Britain in population by around the turn of the century and then certainly grew faster as the twentieth century progressed.
Russ Roberts: Let’s turn to the question of how economists have looked at growth and how that view has changed.
Paul Romer: The basic economic analysis—starting with Adam Smith and certainly with Malthus, who came afterwards—was an economics based on physical objects, notions of scarcity. If you have a plot of land or a piece of wood, only one person can use it and there’s a finite number of those pieces of wood and plots of land. Economists recognized that there were other things out there besides physical objects—things like ideas, a formula or a recipe for how to rearrange physical objects and make them more valuable.
For a long time, they recognized that those formulas or insights or ideas were extremely important and that we were discovering more of them over time and that discovery was driving growth, but they felt like we didn’t have the economic tools to study that process of discovery or technological change the way we studied the market for corn.
So economists said, “Let’s just set aside the question of where technology comes from,” and they made up this highfalutin jargon to cover up for their ignorance in this area. They said, “Let’s treat technological change as exogenous. It comes from outside the economic system. We don’t know where it comes from and why, but given that there’s technological change, let’s study how an economy transforms scarce resources like iron ore into tractors and forklifts and structures and so forth.” So we studied capital. We studied labor. We didn’t pay much attention to the underlying process of technological change.
Russ Roberts: We didn’t really pay much attention to what you call the recipe, which is a metaphor that you use in your article, “Economic Growth,” and that is really a very powerful way for thinking about it. We had a thing called the production function, which is really the ultimate black box. Until recently, we didn’t think much about what was going on inside that black box. I mean, we realized there were things called factories where the capital and labor were combined in different ways, different materials, different things came out of the box, but the whole idea of creativity and innovation was not really considered.
Paul Romer: I think part of why this question attracted me was because of my background in physics, and to a physicist, the whole notion of a production function sounds wrong. We don’t really produce anything. Everything was already here, so all we can ever do is rearrange things. Think of conservation of mass. We’ve got the same amount of stuff we’ve always had, but the world is a nicer place to live in because we’ve rearranged it. It’s like we fixed up the house and it’s nicer now that we’ve fixed it up. So then you have to think about, “What is ‘rearranging?'”
Rearranging involves connecting things, or modifying them chemically or structurally. Then you realize there are recipes or formulas you’d use to do that rearranging. It’s like cooking. Take the same raw materials in the kitchen. You can create something—a soufflé which is really valuable. It gives us great pleasure when we eat it, or it could be sour milk. Thinking about ideas this way makes you think about how we actually create value in an economy.
It also helps you think about questions of sustainability. It’s not that we’re using up raw materials and we’re going to run out of them at some point. It’s really that there’s a lot of stuff like this huge Tinker Toy set we’ve got to play with, and we can rearrange things and take them from states where they’re not very valuable and rearrange them into configurations that are worth much more to us.
Russ Roberts: When you and I were in graduate school, we were taught that you combined capital and labor and for a while you got a big kick and then after a while, there were diminishing returns and you hit the steady state. To add another metaphor, you could think of it as there being low-lying fruit. There are easy things to think of that are productive about the way you combine things, and then you get to the subtler things.
But that clearly isn’t the way the world works because we wouldn’t grow at 2.1 percent per capita, year-in, year-out, if all the good stuff had been discovered and all that’s left now is mint-flavored floss. Adding mint to the floss? It’s a breakthrough, right? No one combined those things. It’s a new recipe, but it’s not going to add so much to the pile of goods and services. Yet the pile of goods and services grows in a healthy way even though it’s a very large pile now. Why is that?
Paul Romer: The notion of diminishing returns is very important and none of this new work overturns diminishing returns.
Russ Roberts: Maybe you should start by explaining ‘diminishing returns.’ I didn’t say it very well.
Paul Romer: Think of an activity like moving goods around in a distribution center. Goods come in from manufacturers, and then the distribution center gets them on different trucks and sends them out to stores. You could run a distribution center with 100 workers and just one forklift, and the first forklift would be really valuable for moving the heavy things.
Then you could add a second forklift and that would still add real value. You’d get a lot more done in that distribution center. But by the time you’ve added the 30th or the 40th or the 50th forklift, each additional forklift is really not helping you very much. So with fixed recipes for how you arrange things while you’re adding more and more physical capital, you do run into diminishing returns.
Economies which try to grow by just adding more and more forklifts eventually do run into serious trouble. The Soviet Union tried to grow like that for a while with essentially no innovation but very heavy investment in physical capital. And they grew for a bit because they started out short on capital, but they rapidly ran into diminishing returns from accumulating capital.
So you have to keep discovering ideas. Then the interesting question you’re posing is: Is it getting harder and harder to discover new ideas because we found the good ones first? Or is it getting easier and easier?
To maintain a given percentage rate of growth, you’ve got to discover more things every year. So if each discovery is worth X and you want to grow from 6 percent from a level of, say, $10,000 per capita, you’ve got to add $600 worth of value in new things. But if you want to grow at 6 percent when you’re starting from $30,000 per capita, you’ve got to add a lot more new things. What it looks like is, as we learn more it’s getting easier to discover new things, so somehow knowledge is building on itself. Newton had this great evocative phrase that he can see farther because he “stood on the shoulders of giants.”
But it’s not quite getting easier fast enough to maintain a constant rate of growth. It’s getting easier to discover new things. But instead of adding five, you know, 100 years ago and we can add ten now, it’s more like we could add five before and we can add six or seven now, so that wouldn’t be enough to keep growth going. So how have we kept growth growing at the same constant rate? And there what it looks like is we’ve been putting more and more people to work on the discovery process. We’ve been training those people who do discovery more.
They spend more years in education getting really good at using prior knowledge, so they’re a lot more skilled than they used to be. But there are also more people worldwide who are engaged in the discovery process—more people per capita in the United States doing discovery, and more total people worldwide.
We’re getting the combined effects of knowledge building on knowledge that makes it easier to discover, and having more and more people all engaged in the discovery process. And those two things together seem to explain why we’ve had growth rates which are actually getting faster over time, not slowing down.
Russ Roberts: You’re really talking about investment in what economists call human capital. But a simpler way to say it is from the traditional model of investment—you don’t eat all your seeds. You put some aside for fruit tomorrow or seeds tomorrow. And as we get wealthier what do we choose and are we able to afford more folks doing the looking around?
Paul Romer: Think back to 100 years or 150 years ago when the vast majority of the labor force was engaged in agriculture.
Russ Roberts: About 40 percent in 1900.
Paul Romer: Yes. Where are we ever going to find jobs for all of those people if agriculture becomes a much smaller source of employment? Well, what we’ve done is we’ve educated the children and grandchildren of those farmers, and many of those people are now engaged in the discovery of better ways to do things. As productivity has grown, we’ve freed up human resources which really is, in some sense, the scarcest commodity: the power of the human intellect. We’ve freed up more of that power to engage in this discovery activity.
Russ Roberts: We’re so wealthy. In fact, we can even have people spending their time listening to podcasts figuring out how growth occurs, which is a glorious thing. Talk about what you mean by meta-ideas.
Paul Romer: To recap, we’ve maintained accelerating growth over time, partly because the more we know, the easier it gets to discover. But we’ve also maintained it because we’ve got more and more people pitching in on this discovery activity.
Now how did we get more and more people pitching in on discovery? Some of that has come purely from population growth. There are just more people around. But the most important part of it has come from changes in our institutions.
We have things like universities, and we have things like patent laws, and we have things like research grants which have created incentives for those individuals to engage in more discovery. The institutions—again, the rules of the game—create incentives. And we’ve found ways to create incentives for people to do more discovery. So a meta-idea would be something like the modern research university.
It’s an idea that helps us get better at discovering ideas. When we essentially invented the modern research university with the creation of the land-grant university system in the United States with the Morrill Act in 1870, we created a whole new idea-discovering system with these universities that were focused on very practical problem-solving tasks rather than abstract, ivory-tower examination of the classics. The classical activity at a place like Oxford was to study Greek manuscripts.
The job of the universities in the United States created by the land-grant system was to figure out how to grow more crops. The football team at Purdue is still called the Boilermakers because engineers at Purdue used to work on railroad boilers. They would weld railroad boilers because boilers were very important technology for transportation.
Russ Roberts: We seem to have invented a lot of the cool meta-ideas. A lot of our institutions are really unusual and extremely productive.
Paul Romer: We in the United States did two things that were complementary, that reinforced each other. One of those things was committing to what we were just talking about: education, universal primary education, then universal secondary education, developing the university system, and encouraging research. We committed heavily to institutions of learning and discovery. But we also committed heavily to the market mechanism, to property rights, to free entry, to competition, competition in all its many forms.
The institutions of the market and broadly speaking, the institutions of science—we got both of those right. And it’s the combination of those two which has been so powerful. Many nations of the world have tried to push the institutions of science alone and are learning but have been slower to adopt the full institutions of the market.
When we were kids, when we were going through college, it was still an open question whether market systems would turn out better than centrally planned economies. A lot of intelligent people thought central planning was a better system than the market.
In our lifetimes, everybody has been—virtually everybody’s now been persuaded of the power of the market system. Many institutions around the world still lag behind where we are in the United States in terms of committing to markets and competition.
Russ Roberts: A dramatic example that comes to my mind is, you hear that Cuba has a great educational system. You hear that the Soviet Union in the heyday of communism had great scientific institutes, and yet many of the things they spent their time on were not productive.
Lysenkoism was a dead end in science that didn’t face the market test. A lot of resources were poured into science with no value produced.
Paul Romer: The Soviet Union is an interesting example of how powerful a force competition can be. Where was the Soviet Union closest to the technology frontier? It was actually in military equipment.
A MiG fighter jet was not a bad airplane. Why was it good and their washing machines were terrible? Well, the MiGs had to compete with—at least they were worried about competing and staying up with U.S. fighter technology. Competition clearly stimulates better performance, new discovery, and better productivity.
A lot of the world has yet to really embrace competition in all of its power and discomforts, because it is discomforting sometimes as well as being powerful.
Russ Roberts: I think one of the great advantages we have in the United States is a tolerance for change that Europe, for example, doesn’t appear to have.
That may be masked by political forces that make it hard to figure out what the real preferences are. But on the surface, they appear to have more of a taste for security rather than dynamism and it seems to handicap them tremendously.
Paul Romer: I sometimes tell my students that everybody’s in favor of growth but nobody wants change.
Russ Roberts: It’ll be interesting to see how Europe copes if we’re right, and I think we are, that the cost of their desire for security is going to become increasingly apparent.
Paul Romer: I’m sure many of your listeners are young people, possibly college students, and I think it might help to give a little bit of historical perspective on this. I was talking with somebody my age from France and he was describing the thought processes, what the French were thinking in the ’70s and the ’80s.
Even in the 1970s, their reaction was, “How could you possibly have many different phone companies, and how could people even decide which phone company to use? Wouldn’t that be chaotic? Of course the government has to run the phone company!” And so even in the last 30, 40 years, now the whole world has seen the power of cell phones.
My daughter lived in India last year. She could get cell phone service out in the middle of India because instead of having to wait for the government telephone provider to put a telephone line in, which took decades in some of these countries, private companies, competing private companies, entered and they put cell phone towers all over India.
So much of the world in our lifetimes has gone from thinking competition would be this chaotic, wasteful process to recognizing that this is how we produce a higher quality of life.
Russ Roberts: So to go back to our narrative about the evolution of the way economists think about growth. We started off with this idea that physical things mattered—adding capital, adding machinery to people to make them more productive—and, clearly, that was part of the story.
But in recent years your work and that of others has focused on ideas and the power of innovation of human creativity. What are the policy implications that you think are relevant for both the United States as a country that has a very successful growth record, and for poorer countries that would like to emulate the United States’ rate of growth?
Paul Romer: Let’s spend a minute talking about institutions and the role they have in fostering ideas as opposed to trading objects, and then we’ll get to the policy implications.
Russ Roberts: Okay, sure.
Paul Romer: We recognized that discovery is something where incentives matter, and if we create the right institutions people will do more of it. What was missing was a simple economic theory that could describe that process. And that was where endogenous growth theory, the kind of growth that I worked on, came in. We tried to create models where people had an incentive to go out and discover things like ideas, not to do things like dig up another cubic yard of iron ore.
When we looked at that, we noticed that ideas differ in this very fundamental way from a scarce resource like iron ore, so the optimal institutions that we’ve used for fostering the production and distribution of ideas are somewhat different from the optimal institutions for iron ore.
With iron ore, there’s this just wonderful, miraculous result from Adam Smith that one price can serve two jobs. It can motivate the production of an additional unit of a good, and it can allocate that good to the right person.
Take something like ethanol. If you’re using the price system correctly, it’ll tell you whether or not we should make ethanol and use it as a fuel, and the same price decides who should get that ethanol. As an aside, we don’t use the price system for ethanol. The market for ethanol is heavily distorted with subsidies and we might actually—scientists aren’t sure—but we might actually be destroying energy every time we make a unit of ethanol. If we did use the price system, that wouldn’t happen. We just take it for granted, but it’s remarkable and just a subtle mathematical fact that one price can do two very different jobs for something like a gallon of ethanol.
Now it turns out that doesn’t work with an idea. Take some really valuable idea. My favorite really valuable idea is something called oral rehydration therapy which is this formula, this insight, for how to save the lives of kids who get diarrhea. Many of them will die of dehydration if you don’t give them fluids. If you just give them water or even give them just water plus a little salt in it, they’ll actually get an electrolyte imbalance and die. But it turns out if you mix in a little bit of glucose, a little bit of sugar, along with the salt and the water, you can save millions of lives—from figuring out just a simple formula to mix some sugar in with the salt and a few minerals with the water.
Now what’s the right price for a discovery like that? Well, society should be willing to pay a huge amount to have somebody go discover something like that because it can save so many lives. But then what’s the right price for deciding who gets to use it?
Once that idea is discovered, the efficient price for that idea is zero because there’s no cost associated with using the idea. Another way to say it is there’s no congestion.
If we had a field, a pasture, and we let everybody use it for free, we know what happens. You get the tragedy of the common pasture. It gets overused. You get congestion. You get waste. But there’s no tragedy of the intellectual commons. There’s no overuse or congestion from having everybody use an idea once it’s discovered.
Russ Roberts: And let’s emphasize we’re talking about using the idea, not the glucose, not the salt, not the water, just the knowledge.
Paul Romer: Any good recipe will have to have some ingredients, and the ingredients are just standard old economic goods and standard analysis applies. But the idea itself has this different characteristic. So what we created were models that didn’t rely on the classical institutions of perfect competition.
We brought in models that had to achieve some kind of a compromise. Some kinds of ideas, we might want to treat like oral rehydration therapy and make them public goods. So you might have the government pay for a research project to discover an idea like this and then once it’s discovered, we give it away for free. That’s the kind of difference between an idea and a standard kind of economic good.
We talked about the institutions of market and the institutions of science. It’s interesting how different they are. The market—the most important idea in the market is the notion of a property right. You give somebody an ironclad right over something like a piece of land. They get to decide—and that right lasts forever.
They get to decide what to do with that land, but with an idea in science, it’s the opposite of a property right. You say, “We’ll reward you for publishing and giving away and renouncing any property rights, any control over an idea, if you come up with something really, really valuable.” So we have these two extremes of the institutions of the market and the institutions of science.
What’s really interesting at the policy level, finally getting back to your policy question, is where do most of the interesting ideas lie? Should we treat them more like the market or should we treat them more like science, with strong property rights like the market and ownership or weak property rights like science? I think where we’ve come out in the United States is a kind of a healthy middle point where we assign some kinds of property rights and control over most discoveries.
For most new things you discover, you can go out and potentially become a wealthy person if you come up with a better idea, so we give some property rights. It might be a patent. It might be a copyright. Maybe it’s just secrecy. You can keep people from copying you, but it’s not a perfect property right.
It doesn’t last forever like your ownership of land. Somebody can copy you. They can start to compete with you. The ideas leak out. So where we’ve settled for most ideas is in the middle between pure systems of the market and the pure system of science.
Russ Roberts: We’re talking about two things here, the ideal and the fundamental nature of these things. The last part of your paper “Endogenous Technological Change” is only accessible to a graduate student or someone with a serious mathematics background. But the first part is fascinating to anybody and there are a lot of interesting insights there. In the first part of that paper you talk about how physical goods are “rivalrous”—only one person can use the iron ore at one time. But ideas aren’t rivalrous—an infinite number of people can use the oral rehydration recipe.
It’s hard to make a living selling it once everybody knows what it is. It’s impossible to make a living selling it. So on the surface, it appears the economic incentives to discover these type of ideas are zero but they’re not zero because of secrecy. So, for example, take the oral rehydration example. Suppose I can produce a bottled solution of this stuff and let’s assume there’s no institution requiring me to reveal the ingredients.
Suppose there’s no labeling requirement and I can make an immense amount of money selling it. And if I can’t make money because the people I’m selling it to are poor, other people transfer income to these folks to allow them to purchase my expensive, valuable solution. We could be talking about pharmaceutical products here, not just this example of oral rehydration therapy.
So the fact that the idea can be shared by many doesn’t mean that it has to be shared by many. We understand that the world would be a good place if it is shared by many. That’s where the tension is. But there are many different ways to solve that problem. Correct?
Paul Romer: Yes. Secrecy, as you suggested—in a market system, secrecy alone can create important incentives. Go back to the story I was telling about the warehouse where you had workers and forklifts. People at Wal-Mart actually developed an interesting idea about how to build distribution centers.
It’s what they call cross-docking. Instead of just having one set of docks that some trucks would come up to and unload and then another set of trucks would come up to and then load back up to go out to the stores, they put a building with docks on both sides of the building, on the east side and the west side.
So one set of trucks would come up to the east side and they’d unload things from the manufacturers, and then they’d pick them up off those trucks and then drive them over to the other side of the building and put them on trucks going directly to the stores. And the efficiency advantage there was that you only picked things up once. You didn’t pick them up once, set them down, and then pick them up again a second time.
That simple idea is part of how Wal-Mart can deliver goods to people at lower price than anybody could before. Target and K-Mart and Sears could see what Wal-Mart was doing. They tried to copy it. Wal-Mart knew a whole lot of details about how to get that system right, so the big picture idea could be copied.
All the details were harder to copy and Wal-Mart’s made a lot of money for the Wal-Mart family and then the shareholders of Wal-Mart by discovering little ideas like that—seemingly little but very important economically, keeping them secret, running faster than the rest of the competition, and earning a real profit from it.
But, as I said, that kind of information leaks out over time, so eventually Target catches up. They compete with Wal-Mart, so what does Wal-Mart have to do? Well, they have to go out and discover the next thing and move a little bit further ahead in productivity. That kind of racing to stay ahead and discover new things is part of what drives the innovation machine here in the United States.
Russ Roberts: And secrecy is like a patent with a finite life. Obviously, it’s very hard. Coke had a formula for Coke that was supposedly a secret. Pepsi tastes similar to it, not exactly the same, and there are fans on both sides but these examples of innovations, the return to them, they do get copied. Secrecy can’t be sustained for a long time. Ideas get copied.
Some people argue we need a lot more incentives for ideas—they say we need to extend patents. We need to create a big incentive for research and development. Let’s make the monopoly rights bigger.
Other people argue that’s a mistake. Let’s get rid of all the monopoly rights. In this other view, what we should do is get rid of patents—get rid of them—and we’ll let the market evolve other methods for protecting intellectual property and create these incentives. Where do you come out on that?
Paul Romer: We know an awful lot about how to structure perfect competition in the world of physical objects. There are things we don’t know about how best to get these institutions to do this tradeoff between incentives and distribution.
Now I think one thing to recognize is that the right solution has got to be both a lot of stuff which is somewhere in the middle of the extremes, and then also some real tolerance for letting a hundred flowers bloom, to try lots of different systems.
Take software, for example. Microsoft is still the dominant provider of software under the kind of property rights model. They have strong copyrights. They’ve got strong monopoly power. They’ve got a lot of incentives to develop their software system and a lot of market power. They can charge a high price for their software compared to the cost of selling it to one person, so we’ve got a property rights solution for developing software.
At the same time, we have the open source movement which is actually developing software under the kind of no property rights system, and it’s good that we’ve got both of those. They’re competing and it’s not 100 percent clear who’s going to win. There are different niches.
Open source has decisively won the day in a few niches like Web server software. The Apache Web server is still the most widely used Web server that serves up Web material throughout the world. And that was developed purely by open source—no property rights, just the institutions of science, just “you get credit for improving the Web—the Apache system.” On the other hand, the property rights solution from people like Microsoft seems to have been a better system for developing systems that are easy for lots of people to use.
You know, geeks wrote systems that were easy for geeks to use, but the interfaces just weren’t as clever. Take Steve Jobs. Steve Jobs used Linux as the backend for the new operating system for Apple, but the front end was developed with some property rights. And, you know, they’ve come up with very wonderful front ends for us to use— music players and now phones and operating systems. So it’s great to have a system that allows competition between different systems of innovation.
But there’s not going to be a one-size-fits-all answer. There will be this middle ground, even with something like Microsoft or Apple or pharmaceuticals.
If someone comes up with a new drug or a new interface, you might want to let them have control over that for many years. But we wouldn’t want to give somebody blocking power so that no subsequent innovation could ever take place without the permission of the initial property holder. That could slow things down.
Imagine if we’d given an infinite life property right to the person who came up with A minor, and nobody could play an A minor chord for the rest of human history without negotiating a contract with the great-great-grandchildren of whoever it was who came up with A minor.
It sounds absurd, but if we took extreme property rights to the limit, that’s what could happen. You’d have to negotiate with the owner of every single different chord when you’re trying to compose a song. So even though infinitely strong property rights on land are a great idea, infinitely strong property rights on ideas could really hamstring future innovators.
And so we’ll ultimately have to be in a position which is somewhere the middle— where you can control something for a while and you can control it for sale to consumers, but you can’t stop somebody from coming along and coming up with an even better version down the road.
Russ Roberts: So let me give you my worry. When we put this in the political process you have a situation where Congress perpetually extends Disney’s exclusive right to Mickey Mouse.
Mickey Mouse has a long life because Disney makes a lot of money off of Mickey Mouse. The fact that creative people can’t use Mickey Mouse without Disney’s permission isn’t really a big loss to human welfare.
But there are a lot of things that are a big loss. The holders of these property rights use the political process to protect them. I have very little confidence that the political process will pick the right mix.
Take another example. I’m a big fan of universities, as clearly you are. We make a large part of our living in them and they’re glorious and there’s wonderful things that happen here. We’ve subsidized them tremendously in the United States and there’s a strong intellectual case for subsidizing them along the lines that you’ve talked about.
On the other hand, there’s an enormous amount of stuff that takes place here that has nothing to do with growth, nothing to do with these ideals that we’re discussing. They’re just due to the fact that people can use the political process, be it at the state house to get grants for their state university or at the federal level to favor their own particular flavor of stuff.
Does that push you at all, as it does me, toward lower property rights for intellectual ideas because of the worries about public choice concerns?
Paul Romer: Yeah. I think that when you set a variable somewhere in the middle—you try and set the dial in the middle—and you have people who have a huge stake in moving the dial one direction or the other, you’re going to create large lobbying efforts and political dynamics that will move the dial in the direction of the people who have a very large concentrated stake. They may not represent the interests of the nation as a whole.
So as we think about policies, it’s very important to think about the political dynamic that you unleash when you create a policy framework. I think patents and copyrights have been an area where people had a lot at stake and they’ve really pushed in the direction of strengthening property rights, potentially going too far.
What’s interesting right now is we’re actually going through a correction in the area of patents. And there’s even a little bit of discussion in the legislative area, although we haven’t seen legislation yet. But the Congress has been talking about it.
The courts have been responding. We’re actually moving a little bit back towards the middle without making patents quite as strong as they were. There are some restoring forces even in the political dynamic, so I don’t worry quite as much. And copyrights are not as damaging as patents if they get out of control, so I’m not quite so worried about it. And I’m encouraged by the movements towards slightly narrower patent rights. But this leads to a very interesting question which is that if you have weak property rights on discovery, we won’t get as much discovery—
Russ Roberts: What do you mean by that?
Paul Romer: If we let people copy ideas or even if patents expire after a certain period of time, or we move more towards the open source system for discovery, in general, the value to society for somebody to develop something new would be bigger than the value to that individual. The innovator just won’t capture all the value. So there’s a good reason why we might want to encourage more of that discovery. We’d like to create institutions. I mean meta-ideas here again. Let’s come up with a meta-idea that could help encourage discovery using our collective resources—let’s use the government to try and encourage discovery.
Now you could do that in ways which unleash strong lobby groups, or I think you can come up with ways to do it which are much less likely to create those kind of lobbying forces. I think we should really focus on that issue when we decide what policies to adopt.
For example, I’m very big on government incentives which could reward students, young people, for going on and getting additional training in, say, science and engineering. Science and engineering training is great for the discovery process.
I think the government should give out more grants to students to just pay for their graduate education—like vouchers—or even maybe their undergraduate education in science and engineering.
But then once they’ve graduated, they just go out in the market and try and discover new things. You don’t know what they’ll do, but they’ll do something clever.
Russ Roberts: We wouldn’t tell them to go discover a hydrogen car. We wouldn’t subsidize that.
Paul Romer: Yes, but you could think about another system which is where the government gives grants to firms for specific kinds of discoveries, and that raises two problems. One, can the government really pick the right things? And, two, even if they can pick, it’s going to create a lot of lobbying pressure from those firms who get those grants, so—
Russ Roberts: You don’t worry about lobbying pressure from the music department and the French department and the philosophy department for those science and engineering grant vouchers?
Paul Romer: Well, that’s why I’d give them to the students, because the students get to vote with their feet. I would put a fence around this which would say these students can go get a degree in science and engineering but not in music or history. But, you know, that would be—
Russ Roberts: You start doing the science of music. We have a lot of neuroscience in music going on here at Stanford, actually.
Paul Romer: Somebody will have to draw a line that says “this is science and this is not,” but I think if you look back in our history, we really benefited as a nation from subsidizing discovery and innovation and, as you said, universities were a good way to do that. Universities have developed, I think, some wasteful features and I would like to shake things up a little bit.
Part of the problem in universities right now is we give all the research subsidies to the professors and the students, the bright young students, you know, the Isaac Newton, the 24-year-old Isaac Newton of our day, can’t get subsidies, can’t get research dollars. They’ve got to go cater to some old professor like me or you, and we tell them to go work on some dumb thing we’re interested in. With innovation it’s great to free up the young people, give them the resources and turn them loose.
So I wish we devoted more of our support for innovation not to universities, not to firms, but to young people who have lots of freedom.
Russ Roberts: It sounds like we need a new type of university. That sounds like a good project for the next 20 years of your life, Paul.
Paul Romer: Well, if you gave students these portable fellowships, you might actually see new universities. If somebody came up with a better university, those students would actually go there.
Russ Roberts: I think one of the reasons that we’re so well paid relative to the average is the difficulty of entry both into our profession but especially into the university and it’s hard to start universities.
And one of the reasons is because they do things unrelated to education that are much more difficult to copy. They create identity and lifestyle experiences that go well beyond the educational process. Interesting challenge.
Paul Romer: Well, the three ways universities get money—research grants, alumni giving, and then tuition—the research grants and the alumni giving set up very serious barriers to entry. Graduate education is funded almost not at all by tuition. If there were students out there with tuition dollars—funded by government vouchers—clambering for graduate education, we might actually see some entry into the creation of new fields of training.
Russ Roberts: Graduate education would look more like MBA training in the sense of its desire to cater to the consumer, which is missing from a lot of graduate education, as you point out.
Paul Romer: Yes and worrying if those people could actually get a job.
Just as an aside, part of why I shifted from an economics department to a business school is I think business schools are somehow the model of how the world should look— that we as professors should live in an environment where tuition is an important part of the incentives that guide what we do. And so I wanted to not just talk the talk from an economics department where I could ignore students and tuition. I wanted to come live and work in a business school and see if that really is the way more of higher education should look.
Russ Roberts: One of the great meta-ideas of the last 20 years was the ranking of business schools. Business Week, was, I think, the first organization to publicly rank business schools in a way that was widely shared. Their methodology has been criticized. It’s silly. It’s imperfect and, of course, it is. But, boy, did it shake up the world of graduate education in business, mainly for the best in my opinion.
Paul Romer: Yes.
Russ Roberts: I want to come back to something you mentioned in passing which intrigued me. You said in the case of these ideas where there’s weak property rights, the inventors won’t capture all the value and, therefore, we don’t get enough of it. So we want to make sure to keep it going, subsidize it perhaps.
But in the physical world, the less “intellectual property” scope of things—nobody captures all the value. Nobody comes close to capturing the value of their discoveries and yet we get a ton of it.
There are many other motivations besides money. There’s pride and glory and fame and all the things that go with that that are often financial, so there are some things in place that keep those discoveries coming even in the absence of capturing the right amount.
Paul Romer: This was a point you mentioned earlier I meant to comment on—even if we had no legal protection for ideas whatsoever, there are still lots of incentives for discovery. Some would come from secrecy, as you mentioned, you know, like cross-docking at Wal-Mart.
Some comes from just reputation and curiosity, so there are lots of things that will keep discovery going. What my claim is, is that if we can keep it going at a pretty exciting rate right now with reduced incentives, imagine how fast it could go if we just turned up the dial a little bit.
Russ Roberts: Fair enough.
Paul Romer: But the challenge, the difficulty is in turning up the dial—to really mix my metaphors—you could also strangle the golden goose. You know, if you provided the subsidies in a way which bureaucratized the whole discovery process, you’d be better off with no subsidies at all than creating this bureaucratic structure which sucks talent away from, you know, exciting things and puts them onto terrible things.
So when I say that we could subsidize it and get more discovery, you have to approach that with a fair degree of caution and think about the political dynamics. But when we provided, say, universal primary school education back early in our history or universal secondary school education and then subsidized university education, part of what we did was we gave lots of people just the basic problem-solving analytical skills to go out and discover more effectively.
And those kinds of investments really paid a huge return for us as a nation. So those, I think, are pretty safe. What’s more problematic are these more targeted ones to particular firms, to particular areas, and I think we’d do well to stay away from those.
Russ Roberts: Well, you also raised the point earlier and you emphasize it in the “Endogenous Technological Change” article, the role of the price system in steering innovation. We don’t just want innovation. We want innovation that changes people’s lives and makes those lives better. We want innovation that delights and inspires and is more than just a new flavor of dental floss. If you rely on the non-market incentives—glory, fame and curiosity—you’d still get innovation. But you won’t get the same type.
Paul Romer: Well, one example that I use comes from colleagues at the University of Chicago. I think it was Bob Lucas or perhaps his wife, Nancy Stokey. They describe why research grants to universities are not the best way to develop all different types of ideas. They said, “Well, imagine that all music that we could listen to was produced by academic departments of music on college campuses.” Have you ever listened to what music people write when they do research on music? It’s often pretty unlistenable stuff.
So the pure university research grant path isn’t necessarily the way I want to get the music I listen to or the books I read. But on the other hand, if you had things like open source, there is a kind of a democratic element where people in open source have to cater to not just a narrow group of peers but to a wider audience. There’s an incentive for people to create things that are valuable for large numbers rather than for small elites.
Russ Roberts: Speaking of music, do you think the incentives for music creativity right now are pretty healthy or do you think they’ve been damaged by Napster and the parts of the Web that have not been controlled?
It’s all well and good to theorize about the optimal mix of property rights, but sometimes market forces are going to overwhelm the ideal policy no matter what. What are your thoughts on the music world?
Paul Romer: I’m an amateur here. I haven’t studied the economics of this industry very well but I like music. My sense is that I wish there was more competition in the discovery and promotion of new talent. I wish it was easier for new talent to get heard and for people to become aware of that talent.
I think the Web has great promise for encouraging that. We’re moving away from a system where a small number of record labels have such predominant influence. The record labels and the radio stations still seem to have a huge impact in this area.
Russ Roberts: They’re the primary filters. I’m surprised iTunes isn’t a record label. I don’t know why that hasn’t happened. I guess they’re busy. They’ve got a lot of stuff they’re working on, but I think that’ll happen.
Paul Romer: So when I wonder whether stronger property rights in music are good or bad, I always pass it through the filter of, “Will it actually create more competition in this business of discovering new talent and getting them out there?” iTunes succeeded partly because they have some property rights. It’s a better form of property rights than the old CD form, but they have some property rights. Why don’t they just become a label and just discover new talent and become the intermediary?
Russ Roberts: And they do to some extent but not in the traditional way. Maybe it will never happen in the traditional way now.
Paul Romer: But I’m not alarmed at all by the de facto weakening of property rights in the music business because it’s become so much easier to copy the music. On the other hand, I don’t think it’s necessarily a bad thing that Apple figured out a way to create some property rights in music. I hope what happens is that people like Apple end up shaking up that business and we’ll see lots more entry and turnover both in artists and in record labels.
Russ Roberts: One of the unintended results of people copying music is that it’s put a premium on live music. You can still close the doors of the concert hall. So there’s going to be more touring and other benefits.
Let’s shift gears. We’ve interviewed a number of people here on EconTalk about growth in the past and some of them point to different causes of growth. People specialize in looking at different things—the role of religion, the role of cognitive ability, the role of institutions, which we’ve touched on here.
What about those things? In your mind, do they all work through whether they enhance ideas or not? Is that the bottom line? It seems to me it is. I think it’s a good bottom line.
Paul Romer: I think that’s probably right. There’s a whole cluster of institutions that influence the incentives for the production and distribution of new ideas, and so religious systems can either foster or hinder that process. Legal systems can do that. Cultural norms can do that.
Think of Wal-Mart again. Cross-docking might have been invented by some university researchers, but if there wasn’t somebody out there who built the distribution centers and built the stores and got the shirts to people, it wouldn’t have been very valuable. So institutions that encourage the discovery and the distribution of new ideas alongside a market system with lots of competition— that’s what we want. There may be many different paths towards those kinds of systems, but that’s ultimately where we want to end up.
Russ Roberts: What’s the role of trade, in particular, international trade, in this conversation?
Paul Romer: That’s a great question. Now think about a developing country, for example. There are at least two fairly different paths that some nations have used to try and get ahold of ideas that exist in the rest of the world. Some places like China and Singapore rely very heavily on direct foreign investment. Foreign corporations know a lot of great stuff. Bring that stuff in. Use it with our local talent and let’s produce some goods.
Some other nations—Japan, South Korea—tried to use domestic firms and some licenses, but mostly domestic copying and domestic incentives for copying ideas that existed in the rest of the world. Now it turns out that South Korea has actually done pretty well. Singapore did pretty well, too, so there’s not a decisive answer about whether direct foreign investment is the best system or not. I tend to think direct foreign investment is actually a good system but it’s clearly not central—
Russ Roberts: A good system for the nations that attract it.
Paul Romer: Yes. For a place like China, it’s been a phenomenally effective way for them to bring in all kinds of knowledge that might have taken them a long time to develop—rediscover on their own. But still there are some nations that have done pretty well without relying much on direct foreign investment. What matters in those nations is, are they finding ways to take advantage of the use and production of state-of-the-art ideas?
The one thing we know that doesn’t work is a case like India 20 or 30 years ago where they let some car manufacturers from Britain come in. Then they threw up trade barriers and said, “Okay, you manufacturers can continue to produce those cars you brought in, protected from the rest of the world.” Well, the same darn cars were produced for 30 years with no change while all the rest of the world was getting better cars. There was no incentive, no structure, that led to improvement in the ideas and the use of new ideas.
So there are multiple paths for getting all nations of the world using state-of-the-art ideas and discovering new ideas, but interaction with the rest of the world through trade seems to be an extremely powerful one.
Russ Roberts: But the critics of international trade would argue that multinational corporations exploit the nations that host them. What’s your answer to them?
Paul Romer: I think that’s just a misunderstanding. If you look at a worker in a Nike factory in Vietnam, that worker is worse off and has a lower quality of life than a worker in the United States. That feels wrong to many of us, and that’s a reasonable kind of moral or ethical response. It’s sad that there are people who live lives that aren’t as nice as ours, but that’s not the question here.
The question is, did Nike’s coming in make the life of that person better off or worse off? The unambiguous answer is that Nike coming in really helps that person and helps many other people in that country.
So the fact that they’re still worse off than we are is not a sign that Nike’s doing something wrong. It’s just the fact that the workers are starting from a very low level. But if you look at the change through entry of somebody like Nike, it’s unambiguously positive for these people.
Russ Roberts: What’s the mechanism? Does Nike improve the life of that worker out of kindness or does competition force them to?
Paul Romer: Oh, I think it’s overwhelmingly competition. There’s sometimes a little bit of pressure which makes them do what is basically charitable giving. But look at China right now or India right now. Why are foreign firms that are operating in China and India or Vietnam—why are they paying workers more than they used to?
What happened was that it wasn’t just Nike that came in. The government let in a lot of other firms. All of those firms started to compete for the best talent there in the nation, and that process of competition started to drive up wages. You don’t want to use the Indian strategy of saying, “Okay, we’ll let in one big firm and then we’re pulling up the drawbridges and, you know, you can do whatever you want.” What you want to do is open it up and say, “Hey, any firm that wants to come in, go for it. Compete as hard as you can to get our best workers.”
And that’ll reward the workers who have the best skills. It’ll give incentives for those other workers to acquire skills and it’ll give them opportunities to do things with their skills that they couldn’t have otherwise done.
Russ Roberts: In what sense are those workers using the knowledge that that multinational has? I love that idea. What do you mean exactly?
Paul Romer: Nike’s discovered a recipe for taking rubber and cloth and a few other things and then creating something that people value in the United States for a price of, say, $100. They can take raw materials worth probably pennies and create something that I might go to the store and pay $100 for.
To create that additional value, they have to go out and find somebody who does the rearranging according to their recipe. If they could get somebody at an extremely low wage to do that rearranging, then they’d pay that low wage. But over time what they find is they’re competing with other employers. They have to pay higher and higher wages to get people to do that rearranging.
Now if there are lots of people like Nike trying to find workers to do high-value rearranging tasks, they’ll be willing to pay quite a bit as they compete with each other. But imagine that Nike only had ideas that could produce things that were worth, say, $10. Nike could never afford to pay—and its competitors could never afford to pay—very high wages to get people to rearrange something to make something worth $10.
But when they’re making something that’s worth $100, they’ll compete and ultimately start to pay higher and higher wages. So the fact that they’ve got an idea, a recipe, that can create quite a bit of value means that they’ll pay quite a bit to have somebody follow that recipe.
There’s lots of people out there with good recipes competing for workers. They’ll bid up those wages and, in a sense, part of the value that Nike creates will in some sense be taken away by those workers, and taken in a way that we feel is good for the world as a whole. It’s good that workers throughout the world will have higher wages in the future than they have now.
Russ Roberts: One last thought on trade and then I want to ask you a final question. At the end of “Endogenous Technological Change” you say that human capital—that is know-how, knowledge—it’s a bunch of subtle stuff along with basic cognitive skills—human capital is more important than population. That as economists we’ve focused a lot on the size of the market in thinking about growth and not enough about the things that come along with the people, the knowledge.
You say something very profound. You point out that India and China are very large. You’d think that they’d get all the benefits of trade from domestic trade. Same with the United States. We’re also very large. So it’s tempting to say that we capture a lot of the advantages that economists attribute to trade—the idea of comparative advantages. There’s enough diversity here that international trade’s not that important for us. But your argument suggests that that misses something profound. Talk about that.
Paul Romer: Thinking about these non-rival goods, these things like oral rehydration therapy that everybody can use if one person discovers it, gives a very different rationale for why free trade—worldwide free trade, global free trade—can be so powerful. It’s not just that other nations have different endowments of physical goods from our endowments—
Russ Roberts: Ricardian stuff.
Paul Romer: Yes. Say we have a sunny climate. You’ve got a rainy climate. Let’s trade goods that grow in sunny climates for goods that grow in rainy climates. But think about this discovery process we talked about earlier. Imagine that the whole world in 50 or 100 years looks a lot like California, that a large fraction of the population worldwide is out busy trying to discover things. Take a particular industry, biotechnology.
If we trade with the entire world and we can take advantage of any new drug discovery anywhere in the world, we’re much more likely to come up with, say, a treatment for Alzheimer’s or some really valuable good than if we have to discover it ourselves. As more and more people are engaged in discovery, the odds of any particular discovery or the odds of valuable discoveries go up.
So we’d really do ourselves a disservice if we cut ourselves off and said, “Okay, California is big enough to trade with. We don’t need to trade with the rest of the world,” or “The United States is big enough to trade with. We don’t need to trade with the rest of the world.” But already important things are being discovered in other parts of the world, and we can take advantage of those if we engage in trade with people who discover those things.
Russ Roberts: So those articles that talk about the threat to our prosperity of 200 million Indian engineers or 200 million Indian software designers are missing the boat.
Paul Romer: Yes, or even worse, the threat to our biotechnology industry if everybody else develops a biotechnology industry. What do we care about?
We don’t care about whether our biotechnology industry makes a profit. What we care about is whether we have a drug that treats Alzheimer’s for somebody who might otherwise have a miserable quality of life. The emphasis on the importance of non-rival goods, such as ideas and discoveries, means that there are gains from scale, from trading with bigger and bigger markets that don’t max out. You keep getting more and more benefits from having more people to trade with. This sounds similar to the usual rationale for trade but it’s really quite different.
And what’s interesting is, is that we also get more benefits if the people we trade with are more like us. It’s not bad for us if they all look like California, with one exception, which is that right now, we pollute in California. We emit a lot of stuff and we don’t get charged for emitting carbon dioxide.
If they start to live like us, they’ll do more of that, too, so there will be more pollution. So there are reasons why governments should try and stop pollution. As long as the governments keep pollution under control, we’ll all be better off if they’re more like us and there are more of them.
Russ Roberts: Well, I don’t know. If they’re more like us, they’re going to pollute a lot less. If China had our standard of living, they’d have a demand for cleaner air plus there are going to be all those meta-ideas that we’re going to come up with on how to make the air cleaner. I just wish there were more engineers in Jupiter, to outsource stuff to, but what can you do?
My last question, I think, is rhetorical but I’ll ask you anyway. Are you optimistic about the future? Are there limits to growth?
Paul Romer: Oh, you know, I’ve been an optimist ever since I got started in economics. It may be just a personality trait but I think it’s been reinforced by the research. I started work during the ’70s back in a time when people talked about the limits to growth and real pessimism about our prospects. People were saying that our standard of living—it wasn’t just that we were going to grow more slowly. Our standard of living, they said, was going to collapse.
There was no way we could sustain it. Those kinds of pessimistic forecasts have been made ever since the time of Malthus. And they’ve always been wrong. The historical pattern has been one of accelerating growth, not just sustained growth but accelerating growth. I think that process can continue throughout our lifetimes and our children’s lifetimes, and the world will just be a much better place because of it. So I am pretty optimistic.
Russ Roberts: I hope we both live to 200 to see it.
Paul Romer: Right. We just need somebody to discover that pill that makes sure that we’re not only alive but we’re actually functional, you know, mentally competent when—
Russ Roberts: And with those artificial hips and knees we’ll have, we’ll be playing tennis at 170.
Paul Romer: Yes.
Russ Roberts: My guest today has been Paul Romer, the STANCO 25 Professor of Economics in the Graduate School of Business at Stanford University. Thank you, Paul.
Paul Romer: Great. Thanks, Russ.
To read or post a follow-up to this essay, go to Romer on Growth.