Business applications fell 7.8% in August. The bigger story is what did not collapse.
New U.S. business applications retreated sharply from July, yet they remained well above a year earlier and far above the pre-pandemic benchmark. The gap between a monthly drop and a still-elevated startup pipeline is the number worth understanding.
One weak month does not erase an unusually high level of entrepreneurial activity
The headline number from the Census Bureau’s September 11 Business Formation Statistics release looks abrupt: 531,728 applications for new businesses in August, adjusted for seasonal variation, down 7.8% from July. That is a meaningful month-to-month decline, and it arrived just as financial conditions were becoming a more prominent concern for households and firms. But the monthly change is only one layer of the story.
August 2026 applications were still roughly 11% higher than the 477,409 recorded in August 2025. They were about 82% above August 2019, when the comparable seasonally adjusted count was 292,063. The post-pandemic startup boom has cooled and changed shape, but the filing rate has not returned to the world that existed before 2020. In level terms, Americans are still creating new business entities at a pace that would have looked extraordinary a few years ago.
That distinction matters because a monthly growth rate answers a different question from a level. The first tells us whether momentum improved or deteriorated from the prior month. The second tells us how much activity remains in the system. August says momentum weakened. It does not say the startup pipeline disappeared.
What the Census series measures — and why an application is not the same thing as a new employer
Business Formation Statistics start with applications for Employer Identification Numbers, or EINs. An EIN is a federal tax identifier used by businesses and other entities. Census excludes several categories that are not useful for tracking ordinary business creation, such as estates, trusts and certain financial filings, and then classifies the remaining applications according to characteristics on the underlying form.
Four signals, four different meanings
- Business applications
- The broad pool of qualifying EIN applications. Useful for measuring entrepreneurial intent, not completed startups.
- High-propensity applications
- Applications with characteristics associated with a higher likelihood of becoming an employer business.
- Planned-wage applications
- A narrower group that reports an expected first date for paying wages, one of the clearest signs of hiring intent.
- Projected formations
- A forward-looking Census estimate of how many employer businesses from a month’s applicant cohort will form within four quarters.
This hierarchy is why the August report is more nuanced than the 7.8% drop suggests. The broad pool fell sharply, while the employer-oriented indicators weakened by less. High-propensity applications totaled 145,387 in August, down about 4.1% from July’s 151,606 but still about 1.9% above August 2025. Census projected 28,501 employer business formations within four quarters from the August applicant cohort, down 4.6% from July.
The employer-intent measures are therefore saying “cooler,” not “crash.” They do not guarantee job creation, but they help separate a change in general filing activity from a deeper contraction in the pipeline of businesses more likely to hire.
The July-to-August drop is large, but July itself was a high bar
Seasonal adjustment is designed to remove recurring calendar patterns, so the 7.8% decline should not be dismissed as a simple end-of-summer effect. At the same time, month-to-month percentages can look dramatic when the comparison month is unusually strong. The latest FRED version of the Census series puts July at 576,512 applications and August at 531,728. Even after the retreat, August remained above every month from January through May 2026.
Revisions are another reason to avoid treating any one release as a fixed monument. Historical monthly values can change as seasonal factors and underlying data are updated. The July value visible before the August release was higher than the currently published estimate. That is normal for economic time series, but it reinforces a basic reporting rule: use the newest vintage when comparing adjacent months.
The better question is not “Did applications fall?” They did. The better question is whether employer-oriented demand for starting firms is weakening at the same speed.
So far, the answer is no. High-propensity applications fell less than the broad series, while some industry measures held up or even edged higher. That is precisely the kind of divergence economists watch when they try to distinguish noise, composition shifts and a genuine change in the business cycle.
Two sector clues show why the aggregate can hide very different startup stories
The composition of new applications matters. In August, seasonally adjusted high-propensity applications in retail trade rose to 20,371 from 20,131 in July, a small increase even as the national high-propensity total declined. Applications in the information sector also edged up, to 13,180 from 13,136. Neither move is large enough to define the national picture, but both demonstrate that the broad retreat was not universal.
20,371
High-propensity retail applications in August, slightly above July.
13,180
Total information-sector applications in August, also slightly above July.
41,075
Applications from corporations, down from 44,700 in July.
145,387
The broader likely-employer pool, down from 151,606 in July.
New businesses are entering a more difficult financing conversation
The timing of the August data is important. The Federal Reserve’s target range for the federal funds rate has been 3.50% to 3.75% since the July meeting. The next Federal Open Market Committee meeting is scheduled for September 15–16, with the policy decision due on September 16. Recent inflation data have made the direction of that meeting unusually consequential for small firms that depend on credit cards, bank loans, equipment financing or commercial real estate.
That makes the persistence of high application levels more notable. Prospective entrepreneurs are still filing in large numbers even though financing is not obviously becoming easier. The next question is how many of those intentions survive the expensive middle stage between obtaining an EIN and opening a payroll, signing a lease, buying equipment or building inventory.
Why the September Fed meeting matters for startups
- Working capital: short-term borrowing becomes more or less expensive as benchmark rates move.
- Equipment and property: rate expectations influence financing costs for vehicles, machinery and commercial space.
- Demand: tighter monetary policy can cool household and business spending, changing the revenue outlook for a new firm.
The post-pandemic startup boom was real — but the hiring payoff has been fading toward trend
The long-run context is one of the most important checks on the August number. Research from the Federal Reserve Bank of Richmond found that the surge in EIN applications after 2020 did translate into a substantial rise in establishment openings and entrant job creation. Employer startups were not just a statistical mirage produced by people registering side projects.
But the same research also cautioned against assuming the pandemic-era jump would permanently raise the pace of employer creation. By 2025, establishment openings and job creation by new entrants were cooling toward their historical trends even while application levels remained elevated. In other words, the conversion rate from “I filed to start something” to “I opened an employer business and hired workers” is the economic bridge that deserves the most attention.
Why elevated applications still matter even when many never become employers
It can be tempting to discount the broad application series because many applicants will remain self-employed or never launch. That would be a mistake. High application volume still tells us that a large number of people are reorganizing economic activity around new entities. Some are replacing wage income with independent work. Others are creating firms that may hire later. Some will fail quickly. A smaller group will become fast-growing employers.
This is why the high-propensity series deserves more attention than it usually gets in headlines. At 145,387 in August, it is not rising explosively, but it remains above the same month a year earlier. That resilience suggests the employer-oriented side of the pipeline is holding up better than the broad monthly decline might imply.
The next data will show whether August was a pause or the beginning of a broader cooling
The most useful way to read a noisy monthly series is to set up a test for the next release. If September applications rebound while high-propensity filings remain near current levels, August will look more like a correction from a very strong July. If both broad and likely-employer applications decline again, the case for a genuine downshift will strengthen.
The next monthly Business Formation Statistics release is scheduled for October 14. Before then, other data can show whether the environment facing young firms is improving or worsening. Census will release August retail sales and July business inventories on September 16, followed by August housing starts and building permits on September 17. Those reports are not direct measures of startup creation, but they help answer whether demand, inventory management and construction are supporting or pressuring new businesses.
Rebound
September broad applications rise and employer-intent measures stay firm. August looks like a one-month reset after a hot July.
Slowdown
Broad and high-propensity applications fall together. That would signal weakening intent is spreading toward likely employers.
Split signal
Total filings remain soft while high-propensity applications stabilize. The pipeline may be getting smaller but more employer-focused.
For Main Street, the real test is conversion: filing, financing, opening, hiring
For that reason, the strongest conclusion is neither bullish nor bearish. It is that America’s startup pipeline remains unusually large while its near-term momentum has softened. The economy still contains far more business-creation intent than before the pandemic, but the burden of proof is shifting toward conversion: can those filings become durable firms under tighter and more volatile conditions?
Four numbers worth tracking together
- Total applications for the broadest signal of entrepreneurial intent.
- High-propensity applications for the likely-employer pipeline.
- Projected formations for Census’s estimate of employer conversion within four quarters.
- Actual establishment openings and entrant job creation for the eventual real-economy payoff.
What the August report says about the U.S. economy right now
The risk is that a large filing pipeline produces less hiring than its size suggests. Richmond Fed research has already documented some reversion in employer openings and entrant job creation toward historical trends. If that divergence widens — applications staying high while actual employer creation fades — the startup boom will look more like a durable rise in self-employment and small-scale experimentation than a new era of job-creating business dynamism.
Questions readers may have
Did U.S. business formation fall 7.8% in August?
The broad count of business applications fell 7.8% from July after seasonal adjustment. That is not the same as saying the number of operating businesses or employer startups fell 7.8%.
Are business applications still high by historical standards?
Yes. The August 2026 level was roughly 11% above August 2025 and about 82% above August 2019, based on the current seasonally adjusted Census series distributed through FRED.
What is a high-propensity business application?
It is an application with characteristics that Census associates with a higher likelihood of becoming an employer business, such as corporate organization, indications of hiring or a planned first wage-payment date, or activity in certain industries.
When is the next Business Formation Statistics release?
The Census Bureau’s current schedule lists the September 2026 report for October 14, 2026.
August delivered a sharp monthly pullback in business applications, but not a collapse in entrepreneurial activity. The U.S. is still generating far more new-business filings than before the pandemic, and the employer-oriented pipeline has held up better than the headline decline. The decisive signal will be conversion into operating, hiring firms.
Sources and methodology
- U.S. Census Bureau — Business Formation Statistics, August 2026
- Federal Reserve Bank of St. Louis FRED — Total Business Applications
- FRED — High-Propensity Business Applications
- Federal Reserve Bank of Richmond — Will the Pandemic Surge in Employer Business Formation Last?
- Federal Reserve Board — FOMC meeting calendar
- U.S. Census Bureau — Economic Indicator Release Schedule
Comments
Post a Comment