One of the biggest upside surprises of the pandemic years was an explosion in the number of new businesses being created by Americans. That “startup surge,” as the Economic Innovation Group (EIG) labels it, has persisted past the formal end of the pandemic. As shown in the chart above—from the Census Bureau’s Business Formation Statistics (BFS)—the initial spikes upward and downward in the second half of 2020 have since moderated. Yet business creation in the aggregate remains well above pre-Covid levels, a “new, significantly higher baseline,” according to EIG.
Before proceeding, some definitions:
- “Business formation” or “business creation” is defined by the Census Bureau as an application for an employer identification number (EIN) with the Internal Revenue Service (IRS). These “business applications” are sliced and diced in a number of ways.
- “High propensity” business applications are EIN filings that, based on various Census criteria, are deemed to have a strong likelihood of hiring employees. All the other business applications could turn into employer firms; most of them, however, are likely to remain nonemployers for some period of time, if not permanently.
- “Projected business formation” is a projection by the Census Bureau of how many employer businesses will “originate” from business applications within four and eight quarters of the application.
Let’s start with the basic numbers.
The Facts on Business Creation
From 2005 through 2016, the average annual number of total business applications was 2.6 million. In no year during that period did the annual total surpass 3 million. Toward the end of that timeframe, as can be seen in the Census BFS chart above, total business applications began to rise. From 2017 through 2019, the average annual number of business applications rose to 3.4 million, a 23% increase. Then, the surge.
In the three-year period from 2020 through 2022—even including a drop in the early pandemic months—the average annual number rose to 4.9 million. That represents an 89 percent increase compared to the 2005-2016 period.
On its website, the U.S. Chamber of Commerce has a slick interactive map using BFS data that includes useful comparison of business applications—and, importantly, projected business formation—by sector. Even though the Construction sector, for example, has had about half the total number of business applications as Retail Trade, it has a slightly higher level of projected business formation.
The reason that Census tracks business formation and projects future business formation is that not every business application will become an actual business, let alone one that has paid employees. The number of “high propensity” or “likely employer” business applications has also risen well above pre-pandemic levels, albeit at smaller scale.
The monthly average of high-propensity business applications between June 2020 and January 2023 was 36% higher than between July 2004 and May 2020. That’s a lot more employers: 1.2 million more, to be exact, relative to the pre-Covid trend.
The 2022 Business Openings Report from Yelp corroborates the overall trends tracked by Census and offers, based on data from its platform, additional insight into the micro-dynamics. According to Yelp, new business openings “reached an all-time high” in 2022, “largely driven by new home and local services businesses.” Other types, such as new restaurants, were still lower than pre-pandemic levels. New business openings, per Yelp, were 12% higher in 2022 than in 2019. That’s a more modest increase than shown in Census data, where high-propensity business applications were 28% higher in 2022 than in 2019. But it’s a useful reminder of what the Census data show: applications versus actual business openings on Yelp.
Why Is It Happening?
That’s the immediate question presented by the entrepreneurship data. And, what explains not merely the spike in 2020 and 2021 but the persistence of the surge through 2021 and 2022?
Various explanations have been put forth. It’s possible that the owners of the millions of small businesses that closed in the early months of the pandemic in 2020 started totally new businesses later that year or in 2021. This would basically be “replacement” entrepreneurship. Layoffs in the spring of 2020 may have shoved many toward entrepreneurship. As the Chamber puts it: “Many individuals laid off as a result of pandemic shutdowns turned their ideas and hobbies into a business that could be run from home.”
More broadly, the Chamber posits: “Entrepreneurs solve problems, and when America experienced huge problems in a concentrated time frame during the COVID-19 pandemic, entrepreneurs rose to the occasion. New economic needs and changing consumer preferences created more circumstances for new businesses to start.” This seems true regarding changing consumer preferences, as non-store retailers (read: e-commerce) have dominated the business application increase.
The Yelp data point toward more prosaic but no less insightful explanations. In its data, sectors such as Hotels & Travel, Automotive Services, and Event Services saw lots of new business openings. E-commerce businesses may have driven business creation in 2020 and 2021, but new business creation in 2022, at least according to Yelp, was driven by everyone’s desire to get back to those things we missed during the pandemic.
We also know that venture capital investments into startups hit all-time highs in 2021 and early 2022. While the Census BFS data don’t allow us to qualitatively distinguish VC-backed startups from other types of businesses, it does break out business applications from corporations. Those also spiked in 2020 and have remained at an elevated level, though they’ve fallen in recent months back toward pre-pandemic trend. Some researchers have used the fact of a business being a corporation as a mark of quality and economic impact, so an increase in corporation formations could be positive. Interestingly, however, the spike in high-propensity business applications has been mostly driven by “other” applications, not those from corporations.
(Here’s another question for contemplation, by the way: why were high-propensity business applications (including those by corporations) so high between 2005 and 2007? They were a much larger share of overall business applications than in recent years. One answer is the housing bubble, but that wouldn’t necessarily solve the higher share puzzle.)
Another way of putting the “why” question is through a geographic lens. It’s one thing to look at sectors and sub-sectors; the surge in non-store retailers would strike most people as completely unsurprising. But take Mississippi, which has experienced a huge spike in business applications. In 2019, according to the EIG analysis, the state ranked 22nd in likely employer business applications per capita; in 2022, it ranked 7th. The Chamber highlights Hinds County, the state’s most populous and where the state capital Jackson is located. Hinds had the most business applications (of all types) in Mississippi. So let’s refine our why question: why did so many more people in Hinds County, Mississippi, file new business applications compared to 2019?
(According to the Yelp data, Home Services and Local Services drove new business openings in Mississippi.)
New Entrepreneurial Hotspots?
The state-level data from Census and analyzed by EIG, the Chamber, and others provides a good way to explore some of the nuances of the business application surge. While every state experienced an increase in business applications (total and from likely employers) from 2019 to 2022, the surge has been far from even. In some states, business applications in 2022 were 10-20% higher than in 2019, a respectable increase. In others, the difference was much larger: South Carolina, for example, saw a 51% increase in likely employer business applications. (It should be noted, however, that in Yelp’s report, seven states saw fewer new business openings in 2022 than in 2019.)
If we cross-reference the Census BFS data with other Census data, the state-level picture gets murkier. Take Mississippi, again. According to the Chamber, Mississippi ranked 9th in the country in 2022 in business applications per capita. Iowa, by contrast, ranked 50th. Yet Iowa can boast a higher number of projected business formations than Mississippi because it has a higher rate of business applications becoming employer businesses. Just because a state experienced a surge in business applications doesn’t mean it’s the new entrepreneurial frontier.
A comparison with the Kauffman Indicators of Entrepreneurship, also based on Census data, provides useful nuance into thinking about both the macroeconomic impact of the business application surge and how it may play out differently across the country. Take the top five states in business applications per capita in 2022: Wyoming, Delaware, Florida, Georgia, and the District of Columbia. Let’s remove Wyoming and Delaware because, as EIG observes, they “have long been preferred states for business incorporation” so the high growth rates there may not tell us much about potential economic impact or local context. The next states up are Colorado and Nevada. If we look at these five states’ data on other entrepreneurship indicators, we see some differences. Let’s look at just Florida as an example.
- Top state in Kauffman’s “rate of new entrepreneurs”
- Above average “opportunity share of new entrepreneurs”
- Top state in “startup early job creation”
- Below average “startup early survival rate”
- Low relative rate of business applications becoming employer businesses
- Above average growth in new business openings (Yelp).
Thus, Florida has a high rate of actual business creation and jobs immediately generated therefrom, but many of those business won’t survive (relative to other states) and, compared to others, a lower share will become employer businesses. Similar discrepancies are seen in other states.
Is Business Creation Permanently Higher?
According to some, yes. EIG, for example, says “the durability of the surge suggests that it is capturing a true renaissance in entrepreneurial activity across the United States.” Its analysis helpfully scans research on the utility of business applications as an indicator and what other datasets show about changes in the business landscape. Their conclusion: “the available evidence suggests that these trends [in business applications] are indicative of genuine entrepreneurial activity.”
John Dearie, of the Center for American Entrepreneurship, is more skeptical. In a recent interview, Dearie reminds us that, prior to the pandemic, new business creation in the United States was in “precipitous decline.” The average number of employer firms created each year had stalled and rates of entrepreneurship had dramatically decelerated. Regarding the pandemic surge, Dearie said his “instincts tell me that we haven’t seen enough to declare that the kind of entrepreneurship America needs—disruptive, innovation-driven, productivity- and growth-driving entrepreneurship—has really turned the corner.” In fact, Dearie is “somewhat skeptical about the longer-term significance of the spike” because of the sectoral composition of new businesses. Many e-commerce businesses don’t necessarily fall into the “disruptive, innovation-driven” category.
Some of the data certainly support Dearie’s points. A surge in automotive repair shops may not scream entrepreneurial resurgence to those looking for “disruptive” business creation. The large share of business applications from those that are unlikely to be employers also weighs in his favor. At the same time, however, we know that a large share of employer firms each year transition from nonemployer status. That’s why the Census Bureau projections business formation: some of those non-high propensity business applications will become employers, too.
And what about all that venture funding? One can certainly quibble that the wisdom of VCs shouldn’t necessarily be equated with “productivity- and growth-driving entrepreneurship”—witness the billions in crypto venture funding that has effectively gone into an incinerator—but it should indicate something, right? In the last three years, for example, about $30 billion each year has been invested in biotech companies through VC deals. Surely some of that is “innovation-driven”?
EIG’s perspective that the pandemic surge in business applications marks a “renaissance” of “genuine” entrepreneurship is also persuasive. Certainly the persistence through 2022 of high levels of business applications from likely employers demonstrates that something is happening beyond just the short-term effects of the pandemic. Yet we also know that, prior to Covid-19, some of what contributing to the long-term fall in business creation were factors such as changing demographics. The U.S. population is growing slowly and getting older; that long-term trend has not reversed.
What’s your vote? A new, permanently higher level of entrepreneurship? Or, a temporary blip that will sooner or later subside?