Skip to Main Content
Speaking of the Economy
ribbon cutting ceremony
Speaking of the Economy
Jan. 18, 2023

Start-Up Surge

Audience: General Public

Chen Yeh discusses the factors that have contributed to an increase in new business formation during the COVID-19 pandemic and the economic implications of this entrepreneurship surge. Yeh is an economist 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 Chen Yeh, an economist in the Richmond Fed's Research Department. His research focuses on understanding the differences between firms, and how those differences affect the broader economy.

Chen, thanks for joining me.

Chen Yeh: Hi, Tim. Thanks for having me.

Sablik: Our main topic of conversation today is an interesting and, perhaps, surprising story to emerge from the pandemic. There's been a surge in applications to start new businesses, as measured by the Census Bureau's business formation statistics. I wrote about this for our Econ Focus magazine back in 2021. We'll include a link to that story in the transcript.

But a quick summary for listeners. New business formation has been declining for decades. After the Great Recession of 2007-2009, start-up activity fell and remained depressed for years. This time around, the business formation statistics declined at the very start of the pandemic but then surged to record highs. Last I checked, that measure has remained elevated above pre-pandemic levels.

I should note that these statistics measure applications for an Employer Identification Number [EIN] with the IRS, which is often required for many types of business activity. But they don't correspond one to one with actual start-ups. Still, they do provide some signal of intent to start a business.

Chen, you've been looking into this topic as well. What seems to be behind this surge?

Yeh: You mentioned the Census Bureau's business formation statistics. There are two salient facts that you can read off the data, especially when you look at the industry breakdown.

For example, when you look at the industry [of] non-store retailers, a big chunk of the surge was coming from that industry alone. About a third of the spike between 2019 and 2021 came from that industry. That industry, for example, also includes online retailers. Think about people who want to earn extra money by having a side gig or people who were laid off during the pandemic and needed to make ends meet. They think, "Maybe I can create something myself and start a business." Think about someone crafty who posts something on Etsy.

Another thing that you see popping up very clearly from the data are industries that provide a supporting role for another industry. Think about professional, scientific, and technical services, so interior designers and computer system designers. You can imagine during a pandemic, we all wanted to stay at home. We wanted to spruce up our living room or decorate our home office, so you hire an interior designer. Sometimes your employer requires you to work from home, so you would need this advanced computer system.

You also see a surge in administrative and support services. Think about businesses that had a hard time retaining people on payroll. They were uncertain about their demand. So, what did they do? They go to a leasing agency to have the flexibility of hiring a certain [number of] people.

Another industry that people like Ryan Decker and John Haltiwanger pointed out had a large surge in [EIN] applications is truck transportation. We stopped going to restaurants, for example, and we just started buying a lot of stuff on Amazon. Someone needs to deliver those goods, so you see a large surge in that.

Sablik: Got it.

Taking a step back and thinking about start-ups in general, how do new businesses differ from established firms? Another way to ask that question is why should we care about how many start-ups there are month to month or year to year?

Yeh: New businesses or start-ups … come in all shapes and sizes, right? The U.S. economy consists of about five million firms and about eight to 10 percent of them are start-ups. Many start-ups and young firms display what the literature calls an "up or out" dynamic. Either you take off and you grow very fast, or you simply don't survive, right? This is the narrative that you hear, for example, in restaurants. You need to survive the first two years. That applies to basically a lot of businesses in the U.S.

Just to give you a statistic, of all firms that are born in a certain year, more than half of them won't survive the first five years. But those that do survive — a small chunk of them, actually — they grow really, really fast in terms of employment and revenues. And those are the firms that we tend to care about from the aggregate point of view. These are the set of firms that the literature also refers to as gazelles.

If we care about aggregate employment or productivity or innovation, it is exactly this set of firms or these gazelles that we care about. Just to give you an idea, a small fraction of firms that are start-ups are responsible for 15 percent or 17 percent of average job creation alone. It is for this reason that sometimes start-ups are referred to as the engine of economic growth. A similar logic can be applied to aggregate productivity or innovation. A large chunk of the growth in these aggregate variables is due to start-ups alone. There are some researchers that have found that up to a quarter of economic growth is due to start-ups.

Sablik: So, start-ups are important for overall economic growth. But, as you mentioned, not everyone who starts a business necessarily wants to go on to hire a lot of employees or grow really fast. What have researchers learned about the particular start-ups that matter most for economic growth?

Yeh: There's one distinction, one dichotomy that the literature has pointed out very clearly, and that's the distinction between what we call subsistence and transformational entrepreneurs. The first group is those individuals that start a business as a way of putting bread on the table, like a restaurant [or] a mom-and-pop store. The second group, transformational entrepreneurs, are the true innovators of the economy. They have the intent to grow large. These are the type of businesses that really want to sell a new product or service, rather than continuing or slightly improving on existing ones. These are the type of entrepreneurs that tend to create these gazelles that I just talked about, and they tend to matter the most for aggregate outcomes.

Sablik: Gotcha.

Getting back to the post-pandemic boom in the business formation statistics, do we know yet anything about what type of start-ups these are?

Yeh: It's difficult to say. We don't have the data yet to verify that. Typically, you verify this with confidential micro data from the Census Bureau, and that data tends to be released two or three years after the actual time episodes. So, I would say we'd have to wait at least another year.

But what's encouraging when you look, for example, at the business formation statistics that you mentioned, Tim, there's a distinction made between EIN applications where you're likely to become an employer and where you're not likely to become an employer. Basically, what the IRS does [is] when they receive applications, they ask these EIN filers, "Are you likely to hire people or not?" What you've seen in the past is there was almost no way of distinguishing these two time series. But especially during the pandemic, you really saw a split between these two — the fraction of EIN filers that said I want to hire employees really took off compared to that other group.

There are some signs that we're going to have start-ups that are going to grow a lot. The question is, are they going to grow by a lot for the aggregate? Are they going to be gazelles? That's always going to be a question that can only be answered in time.

Sablik: Yeah, so as you mentioned, it might be some time before we know for sure what type of firms these are. When I wrote my article a few years ago, there was a lot that still wasn't clear on what this would mean for the overall economy. Do we have any more clarity on that aspect now, of what this boom might mean for aggregate economic activity?

Yeh: I would say we have some more clarity on what hypothesis we might have, but I don't have a smoking gun.

We do see a rise in EIN applications, and what's encouraging is that data from the BLS seems to indicate that it's actually accompanied by a rise in employer start-ups. When you file for an EIN, it doesn't necessarily mean you're going to become an employer, right? Sometimes entrepreneurs just file for an EIN just to obtain business credit, but they don't have the actual intention of hiring employees.

If that's the case, a classical view of why a rise of start-ups is good is something that we mentioned before — start-ups tend to grow by a lot. That's good for aggregate employment, aggregate productivity, and aggregate innovation. You'll see a lot of start-ups creating new products. You have a lot of start-ups that will engage in experimentation that will lead to that innovation. That's typically good for aggregate productivity and, hence, economic growth.

We do see some patterns in the data that seem to imply a restructuring to accommodate the work-from-home lifestyle. What I mean by that is typically when you start a business, you want to locate your business in a thriving city center — I like to think of a Manhattan. But obviously, in the data — something that Ryan Decker and John Haltiwanger also pointed out — is that a lot of these establishment openings are currently in the periphery of the city, the outskirts. That's something we refer to as the "doughnut effect" in the literature.

What that indicates is these are establishments that are opened, as we mentioned before, as a support activity. Think about companies that allow their employees to work from home. You need much more heavy IT support, so you might open an establishment to help with that. But you don't need to put that in the middle of a thriving business district where rents tend to be higher.

So, if that's the case, then this is a mere restructuring as a response to the pandemic and it's temporary. It's not necessarily leading to that innovation that will lead to economic growth.

Sablik: One of the other big stories coming out of the pandemic is that employers have had difficulty finding workers. There are several theories as to why that might be. Could an uptick in people starting their own businesses rather than entering the workforce be a contributing factor? Is there any evidence to support that yet?

Yeh: It's difficult to rule it out. It's definitely possible.

I would say it's a bit unlikely. There are a lot of people that used to be on payroll [who] are now starting their own business, so they're not on payroll themselves anymore. But, as we just discussed, a lot of these EIN applications, these business formations, they actually lead to start-ups that want to hire employees themselves. I would say for every individual that used to be on payroll that now is an entrepreneur, they want to at least hire one or more guys. So, in that sense, it's a bit unlikely. But as I said, it's difficult to rule out.

Sablik: Right.

You alluded to this before, but do you expect this change in entrepreneurial activity to last? Or, are there factors that reduce start-up activity in the decades leading up to the pandemic? Are those going to come back into play in the future?

Yeh: I hate to keep saying it's a difficult question, but that's just how it goes, unfortunately. The recent pandemic is literally one of our last data points that we see in some of the micro data that we use.

From a historical perspective, this rise in business formation is not negligible. But it's relatively small relative to the long-run decline in firm entry that we've experienced in the U.S. economy.

Sablik: Mhmm.

Yeh: There's a literature that tries to explain what's been going on, what causes secular decline in business dynamism. These are narratives that have to do with changing demographics due to the baby boom, or the fact that female labor force participation went up, or the fact that industry leaders find it easier to block industry followers from gaining market share.

If you believe those type of narratives and what that data would indicate, we basically don't have signs that the secular decline in business dynamism, including firm entry, would stop. But, as I said, the pandemic is our last data point, so never say never.

Sablik: Can you tell us anything about what you're researching now around this topic of business formation?

Yeh: I just talked about this secular decline in business dynamism. That's a very interesting topic. We're all trying to understand what caused this decline in U.S. entrepreneurialism.

One thing that Claudia Macaluso and I are working on is trying to understand the cause of the secular decline in business formation through the lens of nonemployers. Nonemployers are businesses that generate strictly positive revenue, but they don't have anyone on payroll. Think about mom-and-pop stores that basically live off their profits.

This sounds a bit contradictory because you're trying to understand the decline in employer start-ups and start-ups that hire people by looking at businesses that don't hire people. The intuition behind that is that a lot of these nonemployers, when they thrive [and] when they grow, eventually they will hire employees, right?

An interesting fact that we found in the data is that the [number] of nonemployers that want to hire people has starkly declined over the last two decades. The question is whether that's a bad thing or not. At first sight, you might think this is bad because there are [fewer and fewer] businesses that don't want to hire. But at the same time, what could be happening is that nonemployers are basically hiring more and more from lease agencies or temp agencies.

Think about a hotel that has a cleaning staff. In the past, that cleaning staff will be put on the payroll of the hotel itself. But these days, when you want to hire employees, you need to pay all kinds of taxes, you need to pay taxes quarterly and abide by all kinds of rules, or maybe even set up your own HR department, which is costly. What hotels can also do these days is go to a leasing or temp agency and say, "I need a cleaning staff" and they pay the leasing agency. The thing is, in the data, what will happen is that this cleaning staff will stay on the payroll of the leasing agency. So, the hotel itself will look like a nonemployer and that's not necessarily a bad thing in our economy, right? The core economic activity is taking place. It's just registered in a different way.

Sablik: Gotcha. Yeah, sounds interesting.

Thanks so much for coming on the show to chat about what's been happening in the business formation space.

Yeh: Well, thank you for having me.

Sablik: As always, I'll remind listeners that if you want to keep up with the work that Chen and our other economists are doing, you can head over to our website, And if you enjoyed this conversation, please consider leaving us a rating and review.

Phone Icon Contact Us

Research Department (804) 697-8000