The 2018 Nobel Laureate in Economic Sciences, Paul Romer of New York University speaks with New York University Professor Julia Lane. His work on endogenous technological change transformed our understanding of economic growth by recognizing that ideas—particularly scientific ideas—are neither a private nor a public good. Although he originally argued that the search for ideas is best fostered through competition, his thinking about the best structure has changed over time. In this interview he explains why.
Keywords: value of science, endogenous growth
Julia Lane (JL): You make the argument that science and democracy are under a threat that is more serious than any they faced since the Enlightenment. Can you elaborate on this idea?
Paul Romer (PR): I think that market power presents a major threat. Things like code, that are powerful and reusable but not democratized, that are owned by single entities (with growing natural monopolistic power), can be used to influence large numbers of people and systems (e.g., political advertising). Private companies with lots of data and code have significant influence over things outside of their realm, namely the political process, democratic process, and markets in general.
JL: What do you think is the solution—what would you do?
PR: We must promote open and transparent information sharing, including data production, and finance it. We need to change the incentive structure for information from one driven by pure profit to one driven by something more amenable to the academic model (people doing research because they are interested in sharing ideas). We should de-privatize the information industry—even go as far as putting Google out of business so that we can reduce concentrations of economic and political power as a result.
Data are now and will be the mechanism whereby firms keep a proprietary market and proprietary advantage. If you force disclosure of data, the rate of productivity growth (as in fracking) is higher because firms are able to learn from each other. As Toby Smith notes in his paper, “Demonstrating the Value of Government Investment in Science: Developing a Data Framework to Improve Science Policy,” publicly available information is extremely valuable for the public good.
Science can serve as an alternative market to the private sector to make discoveries and facilitate progress. We should think about adjusting the dial so that things we might have thought were important to do through mechanisms of the market we can actually do through the mechanisms of science. Work at Coleridge is feeding into possibilities for adjusting the dial.
JL: What might a scientific market look like?
PR: One issue that hasn't been addressed well at all is how do we keep our system of funding for scientific research and the government funding for scientific research? How do we keep that flexible? So it's responding to new opportunities?
When the system was first created, we did two things. We delegated some money to the researchers themselves through peer review. We gave money to different agencies to do what they call mission-oriented basic research. So, you've got a lot of money that's controlled by the academic community, and so they work on the things that interests them. That may not be the things that are most important to society, or that are the highest valued opportunities right now.
I think we need to change the dial to be more mission oriented. One example has been the Department of Defense’s [Defense Advanced Research Projects Agency] DARPA model, which produced so many breakthroughs like ARPANET, which was the forerunner of the internet.
The challenge is that scientists have been too hesitant about what they think of as outside meddling and unwilling to address the question about what's the best way to reallocate or decide where the big investments should go. They just argue for more money for everything. That’s politically appealing, but intellectual cowardice. We need to figure out there are more important places to invest, and less important places, and we've got to think about mechanisms that encourage that.
JL: What mechanisms do you have in mind?
PR: I am inspired by the pre–World War Two system, which grew out of the 1862 Morrill Act and the establishment of land grant universities, which feature a lot more local control. That was the system that took us to worldwide preeminence as an industrial nation. As Dick Nelson and Nate Rosenberg argued many years ago, science should situate itself in Pasteur’s quadrant: we should work backwards from problems you want to solve. Part of what the land grant system did was it brought local government and local concerns to the forefront as a source of problems to be to be solved. For example, research at the University of Michigan developed a way of converting power into pellets for easier transportation, which transformed mining industries in those areas. Purdue University has the nickname Boilermakers, because boilers were a very important part of the technology of steam transport. So there was a focus on local concerns.
The important lesson learned about research funding is the indirect effect on the training of future scientists. The schools got their funding to a large extent because of tuition, so that if you had a big program in, say, basic chemistry, and people couldn’t get jobs in basic chemistry, the schools would not get revenue. But new schools of chemical engineering were established because they knew there were a lot of jobs for people to go out and scale up chemistry in an industrial context. Thus, both teaching and research were driven by basic interests.
One recommendation is to reintroduce a demand-driven element by giving large number of graduate fellowships to promising young students and so they are explicitly freed up to go wherever they want. Such an approach would be an alternative to the system we've got right now where almost all of the funding for graduate students goes to the professor who says I want to work on high-energy physics, so I'm going to create some postdoc positions or some graduate student positions in high-energy physics and my peers give me the money (Romer 2001).
Another recommendation is to change the current incentive structure for universities. They create research factories where they get a 60 or 65 percent cut. All the evaluation is on publications, not on student placement.
JL: Turning to data more generally—what is the role of government in providing public data?
PR: My endogenous growth theory (editor’s note—for which Romer got the 2018 Nobel Prize in Economic Sciences) was, in a way, a kind of an endorsement of Chicago-style small government. But I now think we need to have a stronger government that can say that some activities, even if very profitable, shouldn’t be allowed because it’s going to destroy the world. The government then has to have a database from which to figure out the harm—but it currently has neither the data nor the authority. There are real problems to be worried about in that kind of power in centralized government, so that leads me back to the Morrill Act. I think we might see a lot more experimentation, a lot more progress than if we try to do it at the national level.
So, let me give you a specific example. If it were up to me, I would abolish the CDC and the FDA, and I would create state-level, you know, analogues of those two organizations. And then I would tell the states, you should be collecting, you know, data about, for example, new pathogens because you want to be ahead of the game the next time we get hit with something like this. We've got a terrible and appallingly bad system for surveillance right now in the United States. If we delegated this to the states, we'd still have an appallingly bad system in a lot of states. We have some states that would do this right and it would be a wake-up call for the other ones when they see the advantages of being able to act with some knowledge about what's going on. If we're going to strengthen government, we should think about doing it at the state level using competition and collaboration between the states. You can certainly allow for kind of coordination sharing—like a regional CDC. But I think the monolithic federal ideal is just going to lead us to paralysis. And it's also going to lead us to these like single points of failure like the CDC and the FDA, which both failed in this pandemic.
The potential is not just abstract. Fracking provides an example of how state-driven data efforts can have a big impact. Several states put in place regulations that forced disclosure of the fracking fluids that companies were using to do fracking, motivated by concerns about environmental damage because there's sometimes some chemicals, but mostly fracking. So, the effect of this disclosure was when one firm tries something and the results have to be publicized, other firms can actually learn and build on that information when they try things locally. So, the upshot is that productivity in fracking was higher because there was sharing of all of these. There wouldn’t have been any other way to get that kind of information, and the forced disclosure really created great value in a fairly fragmented industry.
Paul Romer and Julia Lane have no financial or non-financial disclosures to share for this interview.
Romer, P. M. (2001). Should the government subsidize supply or demand in the market for scientists and engineers? Innovation Policy and the Economy, 1, 221–252. https://doi.org/10.1086/ipe.1.25056146
Romer, P. M. (2018). On the possibility of progress [Prize Lecture]. https://www.nobelprize.org/prizes/economic-sciences/2018/romer/lecture/
Romer, P. M. (2021, June 23). Conceptual framework for the value of science [Conference keynote]. https://www.youtube.com/watch?v=_FLJbQwpjFc&feature=youtu.be
©2022 Paul Romer and Julia Lane. This interview is licensed under a Creative Commons Attribution (CC BY 4.0) International license, except where otherwise indicated with respect to particular material included in the interview.