Skip to Main Content
Speaking of the Economy
Person looking through window at opportunities
Speaking of the Economy
May 17, 2023

Creative Destruction: The Birth and Death of Firms

Audiences: General Public, Economists

Hugo Hopenhayn discusses his research on the factors that affect the entry and exit of firms in a market, as well as the reallocation of labor and other resources that results from this "creative destruction." Hopenhayn is a professor of economics at UCLA and a long-term consultant at the Federal Reserve Bank of Richmond.



Tim Sablik: Hello, I'm Tim Sablik, a senior economics writer at the Richmond Fed. My guest today is Hugo Hopenhayn, an economist at UCLA and long-term consultant with the Richmond Fed's Research department. Hugo, welcome to the show.

Hugo Hopenhayn: Thank you. It's a pleasure to be here.

Sablik: We're recording this episode in March 2023, and I'm excited to talk to you about some of your research in the areas of innovation and firm dynamics, as well as about a conference you're organizing here at Richmond in May on innovation.

Most people who have had some exposure to economics have probably heard of "creative destruction," a term typically attributed to Austrian-born economist Joseph Schumpeter. It refers to the process by which old, inefficient businesses fail and are replaced by newer, more productive firms. This process is considered a cornerstone of capitalism, but I suspect that most people don't give it a lot of daily thought.

You've spent much of your career as an economist studying firm dynamics — this process of firms entering and exiting the market. Why are you so interested in it?

Hopenhayn: As social scientists, we see facts and we want to think of what might be explaining these facts. Firms are born. Some of them die very young. Some of them survive, grow over time [and] eventually die. There is this process in which there's survival and death.

The process of firm creation and destruction, and the process of firm growth and shrinkage, involves the reallocation of resources — a very large reallocation of resources. Labor is reallocated from shrinking firms to growing firms, from exiting firms to firms that are newly created. Capital is relocated, too — the order of magnitude of this capital relocation is perhaps as large as the creation of new capital.

So, is this reallocation good? Does it contribute to the well functioning of the economy? Well, that's part of what the Schumpeterians had in mind — that behind this process of creative destruction is somewhat of a process of natural selection, where this job reallocation will reallocate talent and resources from less to more valuable activities.

Sablik: Mm-hmm.

Hopenhayn: It's this process of reallocation that makes the economy good and function well and be able to progress. So, that's the economic reason for thinking about this.

In terms of the creation and destruction mechanisms, you can think, well, why are some firms not able to survive? Why are some firms not able to grow? There might be many reasons behind this.

One reason might be that firms are not able to keep up with progress. As the world is changing, as new technologies appear [and] as new needs appear, firms might not be able to adapt. Firms that are not able to adapt to the new environment are likely to not succeed and exit.

The other reason might be the displacement by new firms. This is something I think was central to the idea of creative destruction. New firms, in some ways, are displacing existing firms.

The specific way in which this creative destruction takes place is important. If you're creating a completely new activity, that might not displace directly any resources from other firms. But at the end of the day, you need resources. If you're creating a new activity, you're going to need workers, you're going to need equipment and so on. Indirectly that is going to be competing for resources with other firms. At the other extreme, a firm may come and basically displace completely one other firm that was the leading firm in that particular activity. By doing that, it's adding some new value. But at the same time, there's another force which is taking away value from the firm that exits.

While these resource reallocations might be good from the efficiency point of view, not everything is perfect. There might be reasons why creative destruction might generate some sources of inefficiencies. Even though it might contribute to this reallocation, it might do it [while] burning some value that was created by other firms. The anticipation of that burning of value might affect those firms in decisions to innovate to begin with.

Sablik: Right. As you mentioned, this reallocation process can be a bit messy sometimes. It can be particularly painful for workers at firms that fail and lose their jobs. A number of countries have adopted policies to help with that transition process, to help protect workers who lose their jobs through no fault of their own.

A 1993 paper that you wrote with Richard Rogerson in the Journal of Political Economy explored this issue of how interfering with the job destruction process can have an impact on the overall efficiency of the labor market. Can you talk a bit more about those findings of that paper and the legacy of that work today?

Hopenhayn: Sure. Just to focus on that particular kind of policy, this is like a tax on firing. A tax on firing one might think protects the workers that are currently in been in the firm. And that is right in some ways. But one has to understand that a tax on firing has implications for hiring. In anticipation of the fact that if you were to downsize in the future you're going to have to pay these taxes, firms are going to hire less.

Beyond that, there is the impact that these policies have on overall productivity. Presumably, if the reallocation of resources is healthy, it's going to be a reallocation that goes from less to more productive or less or more valuable activities. So, taxing reallocation is somehow curtailing the efficiency of this movement of resources. As a consequence, it's going to slow down progress. It's going to reduce the overall productivity in the economy.

Sablik: Mm-hmm.

Hopenhayn: Now, the question of protecting workers is a serious one, especially in the event of large displacements that could happen because there are big reforms — lowering tariff barriers, privatizations and so on — which are very common in less developed countries. This could lead to very large employment losses. The question is what is the best mechanism to cope with that. In this process of creative destruction, there [are] winners and losers and there are people that can lose a lot. We would like to be able to provide some kind of protection and insurance.

Unemployment insurance is an institution that is fairly common in more developed economies, but it's less prevalent in less developed economies. I'm an advocate of those kinds of policies, properly designed. It's understood that unemployment insurance could also generate some distortions.

Sablik: In recent years, there's also been some growing concern about an overall slowdown in business dynamics. Firm entry and exit rates have been falling since the 1970s. This is something that you've also been looking into and researching, particularly looking at how [business dynamics] connects to demographics. What did you find in that research?

Hopenhayn: These demographic changes in the U.S. economy have been tremendous. Around the mid-'70s, the labor force was growing at a rate of 2.5 percent per year. Nowadays, it's growing at 0.5 percent. It's hard to understand what are the implications of these changes. There's lots of research that is concerned with this.

When we think about firm creation, when a firm creates, it's going to need to take its resources from somewhere. Part of where it takes them is from existing firms. But also new resources are added constantly to the economy in terms of labor, which is one of the most important resources, by the introduction [of workers] or the growth in the labor force — basically, new people entering the labor force.

If there's a slowdown in the growth of the labor force, there's a slowdown in these resources that are made available for the creation of new firms and for the expansion of existing firms. What we see in the data is that the incentives for firm creation are affected negatively by this decrease in the rate of growth of [the] labor force. So these demographic forces have a first impact — less people are born, less firms are born.

Another side of the story is that firms are going to be created and firms are going to expand if demand grows. Part of the growth in demand is the increase in population. If population increase is slowing down, then that contribution of that additional source of demand growth is diminished.

In the same way that we think about populations having demographics and populations aging, firms also have demographics. So, in any population where births are going down, that population will age. In the same way that the population of people is aging, the population of firms is also aging. What implications does this have [for] questions such as the extent of reallocation, perhaps innovation? What implications does this have on concentration?

What our paper tried to do is try to account [for] how much of the rise in average firm size, how much of the decrease in entry rates, how much of the decrease in exit rates — the turnover of firms — [and] how much of the decrease in reallocation between shrinking and expanding firms is affected by these demographic forces. We find that it's a very large fraction.

In terms of the implications for the economy, it's unclear. If you think that [firm] entry is contributing to growth by bringing in new ideas and so on, that might be worrisome. Perhaps part of the slowdown in productivity has to do with this. But all of these trends are not showing that the economy is not working well. No, the economy is responding in a natural way to the changes in demographics.

Sablik: Yeah. As you mentioned, the creation and destruction of firms matters a lot for the overall productivity of the economy, and the productivity of individual firms matters for the aggregate. You've explored how frictions like the misallocation of resources also influences productivity. Can you talk about your findings and how they help improve our understanding of the drivers of economic growth?

Hopenhayn: These issues that you're talking about are of the most concern with less developed economies. There is a very important question in economics, which is trying to understand the big disparities that we see in income per capita across countries in the world — much larger than the disparities that we see within developed countries.

One of the leading explanations of this is the aggregate level of productivity of less developed countries is the principal driver. Obviously, there are differences in their human capital and also their physical capital. Their resources are not the same. But it's not only the resources. [It's] how these resources are put together that make these economies much less productive. And so, creative destruction and the reallocation of resources to their most productive uses, in some ways, will determine the productivity of an economy beyond the resources it has.

One of the prevailing views of explaining under development is that these distortions might be much more important in some less developed countries. There are many policies that affect this. One I referred to before was these labor firing taxes. But any distortions that are affecting this process of reallocation will definitely have consequences over productivity.

There [are] other policies that are quite common. The cost of creating a business differs a lot across countries. If you have very high costs of creating business, you're taxing the creation of firms, which will also lead to less destruction of firms. But that reduction of the creation and destruction of firms will reduce the reallocation which is, in principle, a healthy reallocation and as a consequence will have effects on aggregate productivity.

There [are] also incentives for innovation. If resources are not well allocated, that might increase the cost of innovation. Firms below a certain size in many countries, including France, are exempt from some labor taxes.

Or, in less developed countries — perhaps because of the high cost of creating firms — there are many [of] what [are] called informal firms, which are firms that don't comply with any of the regulations and are not paying any of these taxes. The more successful firms, which are the formal firms, are paying these taxes [and] the other ones are not.

You might think that some of this should exist in order to promote a fair economy. They do have a cost in the sense that they there are a tax on being successful. If you have a tax on being successful, that's going to also decrease the incentives to be successful. That decrease in the incentives might imply that firms are more reluctant to invest in innovation and in activities that enhance value. As you reduce the value of firms, you're going to reduce the aggregate productivity.

Sablik: Yeah. Thinking about policies that are designed to encourage innovation at firms, I mentioned at the beginning of the conversation you're organizing this conference at Richmond in May about innovation. What are you hoping to learn there? What are the big questions in this space that economists are researching?

Hopenhayn: I would say the conference is really not only about innovation. It's more about the intersection of networks, innovation and productivity growth.

Lots of papers are exploring the role of inter-firm linkages for productivity. This can happen in different ways. For example, trade economists are interested in the fact that lowering barriers to trade, lowering tariffs, allows firms to search among a larger set of potential providers of inputs.

Another form of interconnection, which is not necessarily between firms that are [in] a supplier-buyer relationship, is between firms in the idea space. When a start-up is thinking about a new idea, a lot of the thinking about a new source of value is inspired by opportunities that it sees from existing technologies or advances in technology. Innovation breeds innovation. There [are] these feedbacks and these linkages that we are interested in understanding better.

One of the forces that is behind growth is not just knowledge creation, but also knowledge diffusion. How does this knowledge diffuse? It diffuses by perhaps observing others that are acquiring and changing their practices. This is something that carries through the rest of the economy. Part of the selection of papers at this conference is also trying to give some light on this.

Sablik: Sticking with the theme of innovation, this conference is part of the Richmond Fed's CORE Week model, which is an innovation that the Research department here adopted after the pandemic as a way to bring economists together, originally virtually and now in person, for seminars and conferences.

You've been quite involved and have attended several CORE Weeks. What do you find valuable about that process?

Hopenhayn: Let me first start [by] saying that the process that is going on here is not unique. It's perhaps quite unique compared to other Federal Reserve Banks.

At UCLA, for instance, every area used to have their seminar. You invite somebody, you hang out with them, maybe some students also hang out with them. And it's a whole day — you go to dinner at night — so there's a lot of informal information exchange plus the paper they present.

A lot [of economists] have been saying instead of doing that you do, monthly, a mini conference where you bring four people and you have a whole day. During that day, you have these presentations of the same length of time that you would have had if people would come every week, but you're concentrating them in one day. So, it's similar to what's happening [during] CORE Weeks.

CORE Weeks concentrate both researchers from inside the Richmond Fed [and] researchers that come from outside — visitors, speakers. It's a critical mass, a very large critical mass. Rather than having these activities distributed evenly over time, you have these weeks where you get this large scale effect. Obviously, this is exploiting, in some ways, economies of scale in knowledge, knowledge creation in the future, right, as we were talking about before.

When I go to a seminar, I learn from the speaker. But you learn almost as much from the questions that are being asked from the people that are there. People are highly energized — speakers sometime start laughing because they get so many questions that they think they're not going to be able to finish.

Sablik: Hugo, thank you so much for coming on to talk with me today.

Hopenhayn: I appreciate the opportunity.

Sablik: I'll remind listeners, as always, that you can head over to the show page for links to Hugo's papers that we discussed today. And if you enjoyed this episode, please consider leaving us a rating and review on your podcast app.

Phone Icon Contact Us

Research Department (804) 697-8000