The U.S. Census Bureau released statistics on business formation on September 11, showing that in August, there were 531,728 business applications after seasonal adjustment, which is a decrease of 7.8% compared to the revised figure of 576,512 in July. This decline is notable, but since it follows a 8.1% month-over-month increase in July, it seems more like a fluctuation at a high level rather than a signal that entrepreneurial activity has entered a recession on its own.
Enterprises apply for statistical data from the employer identification number, which is EIN. This process is fast and covers a wide range, allowing for an early observation of potential entrepreneurial activities, and therefore it is often used as a leading indicator of economic vitality. However, there is still a long way to go between "applying for a number" and "establishing a business that continuously employs staff." Some people may cancel their plans, some entities use it solely for asset holding or individual operations, and it can take several months for some applications to generate payroll tax records.
530,000 applications represent intended traffic, not that 530,000 new employers have already opened for business.
The Business Formation Statistics of the Census Bureau includes four types of application sequences and eight types of formation sequences. The total number of applications excludes items such as tax liens, estates, trusts, certain financial filings, agriculture, and public administration; high-probability applications are further filtered based on industry, legal form, and whether wages are planned to be paid, to identify those that are more likely to become employer enterprises. The questions asked vary depending on the perspective, so one cannot simply take the largest total number of applications as the number of newly established companies.
Seasonal adjustment is also crucial. The start of annual tax filing schedules, changes in academic semesters, holidays, and policy milestones can all affect EIN applications. Statistical departments use historical patterns to account for expected seasonal fluctuations. A 7.8% decline month-over-month in August refers to the adjusted change and does not mean that every state or industry experiences a simultaneous decline. To determine a trend, one should observe at least three months of average data, applications with a high tendency, applications for planned salary payments, as well as forecasts for the next four or eight quarters.
The data for July was revised to 576,512, representing a month-on-month increase of 8.1%. Looking at the two months together, the decline in August offset the surge from the previous month. Monthly figures are inherently susceptible to the impact of concentrated filings and calendar effects. If the media only focuses on one negative figure, it is easy to misinterpret a "cooling down in enthusiasm" as a "collapse of entrepreneurship." A more accurate description would be that the application volume has fallen from its peak in July, and whether this trend will continue downward remains to be seen with more months of data.
Statistical definitions also changed in 2026. Starting from January, the Census Bureau began using the 2022 version of industry classifications to restate time series data, and applications related to internet sales were excluded from certain high-priority and corporate application categories. Annual updates also revise industry codes and actual business formations based on more complete administrative records. When comparing long-term periods, it is essential to use the same version of the time series; otherwise, changes in methodology should not be misinterpreted as real economic turning points.
To judge the quality of entrepreneurship, one must continue to look at employment, financing, and survival rates.
The number of applications reflects the willingness to enter the market, but what truly affects the economy is whether this willingness can be translated into actual business operations, investments, and job creation. When interest rates are high, the cost of registering a company may not increase significantly, however, rent, inventory, equipment, and working capital become more expensive. As a result, applicants may be more cautious before embarking on actual business activities. Digital platforms have further reduced the barriers to entry, allowing many people to apply first and then test the market, which has widened the gap between the number of applications and the number of traditional physical businesses.
The industry structure also affects the implications. There are significant differences in capital requirements, hiring speeds, and failure rates among professional services, construction, catering, transportation, and e-commerce. For a given total number of applications, if more come from entities with a higher propensity to pay wages, it usually indicates a stronger signal for future employment; whereas if they are concentrated among those with low capital and no employee activities, it is more indicative of changes in self-employment and project-based income.
Regional data cannot be obscured by the national average either. States with rapid population growth, active housing construction, or concentrated industrial investment may maintain a strong demand for applications; regions that rely on a single industry may be more vulnerable. When comparing inter-state variations, it is also necessary to use the population or the working-age population as the denominator, otherwise larger states naturally have more applications. Disasters, tax deadlines, and state-level registration policies can also create short-term peaks and troughs.
From the perspective of business owners, applying for a reduction may mean a decrease in expected returns, or it could simply be that previously pent-up demand has been released ahead of time. After a significant increase in July, some applications that were originally scheduled to be submitted in August may have already been completed. Without data on the distribution of application dates and subsequent developments, it is not possible to distinguish between these two explanations. Therefore, while the news headline can describe a decline, it should not directly give entrepreneurs confidence or label the economic cycle as certain.
Investors should also consider this set of data in conjunction with small business confidence, bank lending standards, job vacancies, and local consumption. An increase in applications but tightening of lending may indicate an increase in creativity yet difficulty in securing financing; a decrease in applications along with improved sales for existing businesses does not necessarily mean a weakening economy. The advantage of BFS is its early detection, but the trade-off is that it is far from providing a definitive conclusion, making it suitable for offering directional clues rather than for independently drawing growth conclusions.
The Census Bureau clearly defines the projected establishment formation as a forward-looking estimate: it estimates the number of businesses that will incur payroll tax obligations over the next four quarters, rather than the total number of newly established companies in any given month. News reporting should maintain this distinction. The 531,728 applications in August and the 7.8% month-over-month decline are actual statistics, but they only indicate changes in the flow of new entrants. Whether the wave of entrepreneurship will continue will ultimately be determined by the number of applications, the emergence of actual employers, and the survival of these businesses.
