I’ve been investigating this for five years, and I still don’t have a full answer. But part of the picture is starting to come into focus. Here’s my current, incomplete model:
Progress compounds. It builds on itself. Progress begets progress. This is why progress is super-linear: exponential, or indeed, over long periods, even super-exponential.
The form this takes is a number of feedback loops, or self-reinforcing cycles. By the nature of such loops, they act as if they had inertia: they are hard to get started, but hard to stop once going. Hence, a flywheel: the perfect metaphor for a loop or cycle with a lot of inertia.
There are several of these, at multiple levels, overlapping, and all operating simultaneously. Here are some that I can see:
Technology. Some technologies are fundamental, enabling many other technologies. Precision machining allows for the invention of many more types of machines. New engines and energy sources do the same. Information technology speeds up the dissemination of ideas and makes it easier for people to collaborate. Many technologies enable themselves: we use machine tools to make machines, we burn oil to drill for and to transport oil, we design computers on computers.
Wealth. Some level of surplus wealth is needed to fund research and development. When half the workforce had to be farmers just to feed the other half, the surplus simply wasn’t there. In the Renaissance, science was funded by wealthy patrons. As surplus builds up, we have more to invest in experimentation, invention, and new businesses; and progress in these areas raises our productivity, which gives us more surplus.
Science. Science enables advanced technology: electromagnetism enabled both electrical power and electronic communications; applied chemistry created everything from plastic to the the Haber-Bosch process; microbiology gave us sanitation, vaccines, and antibiotics. And technological and economic progress then in turn enable scientific progress, both by creating surplus wealth to fund it (as per the previous point) and by creating new scientific instruments and techniques, from the microscope and the thermometer to the LHC, LIGO, and JWST.
Markets. Transportation and communication technology have globalized markets that used to be narrowly local. Larger markets support more goods and more kinds of goods. Products that require specialized manufacturing need markets large enough to recoup that investment; you can’t fund the factory needed to build a working threshing machine or reaper if you are only selling to the farmers in your local village.
Government. Progress was enabled in part by reforms in law and government, such as the dissolution of the guild system or the development of corporate law. Nations that have better legal support for progress become wealthier and therefore stronger militarily, and thus able to defend themselves, and their example has inspired other nations to reform their own governments and laws (as India and China did over the last few decades).
Population. All else being equal, the more people who are trying to drive progress, the faster it will go. For a long time, progress led to higher population as well. Improvements in agriculture increased the carrying capacity of the land, leading to higher population densities. By the 18th century, sanitation improvements were lowering mortality rates, and more children were surviving to adulthood. But this cycle may have flipped from self-reinforcing to self-reducing (in engineer’s terms, from “positive” to “negative” feedback): By the 20th century, technology, wealth and education had lowered fertility rates as well. Now global fertility rates are falling, world population growth is slowing, and indeed total world population is set to level off or even begin declining this century. This may turn out to be a significant limiting factor on progress.
Philosophy. The more progress we make, the more people believe it is possible and desirable, and the more they put their efforts into it, which creates more progress. It was a few 15th-century examples such as the voyages of discovery, the compass, gunpowder, and the printing press, that inspired Bacon and his contemporaries to call for more efforts in science and technology. In the 19th century, long-term investments in progress such as the founding of MIT or Johns Hopkins were explicitly motivated by a belief in progress; the same explicit belief was the justification for Vannevar Bush’s call for investment in basic research. (However, this feedback cycle, too, may have inverted: society grew more skeptical and distrustful of progress in the late 20th century, and I think it’s not a coincidence that progress has slowed down; conversely, the less that people see their lives and their world improving, the less they can hold a positive vision of the future.)
In general, the form of a variable that grows in proportion to its size is an exponential curve. If an economy invests a constant percent of its resources into growth each year, and those investments earn a constant return over time, then that economy will grow exponentially at a constant rate. When we see exponential growth in GDP over time, I think this is what’s going on.
But occasionally, something happens that changes the exponent, kicking us into a new mode of production with even faster growth. This happened with the Industrial Revolution. I suspect it happened with the Agricultural Revolution some ten thousand years ago. Perhaps it even happened with the beginnings of behaviorally modern humans more than fifty thousand years ago.
Thus, over the very long term, progress is super-exponential. Not only do we kick into a new, higher growth mode periodically, but it takes less time to do so with each fundamental shift. Indeed, the next shift could happen soon as this century: many have suggested AI as the driver; J. Storrs Hall has suggested nanotech.
By at least one analysis, progress is hyperbolic, and will become infinite at some point in the future, when we will reach the singularity. (One imagines that some source of friction will be discovered that will make progress not infinite but merely unfathomably fast.)
In any case, when we ask, “what causes progress?”, at some level, the answer is, “all of the factors above”: technology, and science, and invested wealth, and good legal foundations, and philosophic ideas… etc. Certainly, if you wanted to advance progress, you could reasonably decide to work on any of those in order to do so.
What’s unclear to me is: what are the root causes of progress? Of all the causal factors, which are necessary and sufficient? Joel Mokyr might argue that it was science; Deirdre McCloskey that it was moral values; Steven Pinker (or Ayn Rand) that it was the Enlightenment; Robert Allen that it was coal.
As the breadth of answers, and of answerers, indicates, this is a hard problem—one that entire careers and shelves full of books have been devoted to; in fact it’s perhaps the biggest question in economic history. So I don’t have the answer now. Maybe in another five years.
Thanks to Kris Gulati for commenting on a draft of this essay.
Flywheels of progress
Link post
What causes progress?
I’ve been investigating this for five years, and I still don’t have a full answer. But part of the picture is starting to come into focus. Here’s my current, incomplete model:
Progress compounds. It builds on itself. Progress begets progress. This is why progress is super-linear: exponential, or indeed, over long periods, even super-exponential.
The form this takes is a number of feedback loops, or self-reinforcing cycles. By the nature of such loops, they act as if they had inertia: they are hard to get started, but hard to stop once going. Hence, a flywheel: the perfect metaphor for a loop or cycle with a lot of inertia.
There are several of these, at multiple levels, overlapping, and all operating simultaneously. Here are some that I can see:
Technology. Some technologies are fundamental, enabling many other technologies. Precision machining allows for the invention of many more types of machines. New engines and energy sources do the same. Information technology speeds up the dissemination of ideas and makes it easier for people to collaborate. Many technologies enable themselves: we use machine tools to make machines, we burn oil to drill for and to transport oil, we design computers on computers.
Wealth. Some level of surplus wealth is needed to fund research and development. When half the workforce had to be farmers just to feed the other half, the surplus simply wasn’t there. In the Renaissance, science was funded by wealthy patrons. As surplus builds up, we have more to invest in experimentation, invention, and new businesses; and progress in these areas raises our productivity, which gives us more surplus.
Science. Science enables advanced technology: electromagnetism enabled both electrical power and electronic communications; applied chemistry created everything from plastic to the the Haber-Bosch process; microbiology gave us sanitation, vaccines, and antibiotics. And technological and economic progress then in turn enable scientific progress, both by creating surplus wealth to fund it (as per the previous point) and by creating new scientific instruments and techniques, from the microscope and the thermometer to the LHC, LIGO, and JWST.
Markets. Transportation and communication technology have globalized markets that used to be narrowly local. Larger markets support more goods and more kinds of goods. Products that require specialized manufacturing need markets large enough to recoup that investment; you can’t fund the factory needed to build a working threshing machine or reaper if you are only selling to the farmers in your local village.
Government. Progress was enabled in part by reforms in law and government, such as the dissolution of the guild system or the development of corporate law. Nations that have better legal support for progress become wealthier and therefore stronger militarily, and thus able to defend themselves, and their example has inspired other nations to reform their own governments and laws (as India and China did over the last few decades).
Population. All else being equal, the more people who are trying to drive progress, the faster it will go. For a long time, progress led to higher population as well. Improvements in agriculture increased the carrying capacity of the land, leading to higher population densities. By the 18th century, sanitation improvements were lowering mortality rates, and more children were surviving to adulthood. But this cycle may have flipped from self-reinforcing to self-reducing (in engineer’s terms, from “positive” to “negative” feedback): By the 20th century, technology, wealth and education had lowered fertility rates as well. Now global fertility rates are falling, world population growth is slowing, and indeed total world population is set to level off or even begin declining this century. This may turn out to be a significant limiting factor on progress.
Philosophy. The more progress we make, the more people believe it is possible and desirable, and the more they put their efforts into it, which creates more progress. It was a few 15th-century examples such as the voyages of discovery, the compass, gunpowder, and the printing press, that inspired Bacon and his contemporaries to call for more efforts in science and technology. In the 19th century, long-term investments in progress such as the founding of MIT or Johns Hopkins were explicitly motivated by a belief in progress; the same explicit belief was the justification for Vannevar Bush’s call for investment in basic research. (However, this feedback cycle, too, may have inverted: society grew more skeptical and distrustful of progress in the late 20th century, and I think it’s not a coincidence that progress has slowed down; conversely, the less that people see their lives and their world improving, the less they can hold a positive vision of the future.)
In general, the form of a variable that grows in proportion to its size is an exponential curve. If an economy invests a constant percent of its resources into growth each year, and those investments earn a constant return over time, then that economy will grow exponentially at a constant rate. When we see exponential growth in GDP over time, I think this is what’s going on.
But occasionally, something happens that changes the exponent, kicking us into a new mode of production with even faster growth. This happened with the Industrial Revolution. I suspect it happened with the Agricultural Revolution some ten thousand years ago. Perhaps it even happened with the beginnings of behaviorally modern humans more than fifty thousand years ago.
Thus, over the very long term, progress is super-exponential. Not only do we kick into a new, higher growth mode periodically, but it takes less time to do so with each fundamental shift. Indeed, the next shift could happen soon as this century: many have suggested AI as the driver; J. Storrs Hall has suggested nanotech.
By at least one analysis, progress is hyperbolic, and will become infinite at some point in the future, when we will reach the singularity. (One imagines that some source of friction will be discovered that will make progress not infinite but merely unfathomably fast.)
In any case, when we ask, “what causes progress?”, at some level, the answer is, “all of the factors above”: technology, and science, and invested wealth, and good legal foundations, and philosophic ideas… etc. Certainly, if you wanted to advance progress, you could reasonably decide to work on any of those in order to do so.
What’s unclear to me is: what are the root causes of progress? Of all the causal factors, which are necessary and sufficient? Joel Mokyr might argue that it was science; Deirdre McCloskey that it was moral values; Steven Pinker (or Ayn Rand) that it was the Enlightenment; Robert Allen that it was coal.
As the breadth of answers, and of answerers, indicates, this is a hard problem—one that entire careers and shelves full of books have been devoted to; in fact it’s perhaps the biggest question in economic history. So I don’t have the answer now. Maybe in another five years.
Thanks to Kris Gulati for commenting on a draft of this essay.