Small business failure rates are lower than folklore claims. Bureau of Labor Statistics survival tables show that about four in five employer establishments opened in recent years made it through the first year. The same tables show the risk concentrated in years one through five, not spread evenly across the life of a company.
Money habits shape those early odds as much as sector choice does, and the way a company builds its business credit file is one of the signals lenders read. Coverage of firm-level data sits in the business news section of this publication. The numbers below come from the federal tables themselves.
What does the survival data actually show?
The clearest answer comes from one cohort. In the year ended March 2013, 628,840 private sector establishments opened in the United States. The Bureau of Labor Statistics counted 79.6 percent of them still operating in March 2014, 50.6 percent still operating in March 2018, and 34.7 percent still operating in March 2023. Half of those openings were gone within five years, yet more than a third lasted a full decade. For related coverage, see How the Labor Department recovers stolen wages, and how much workers get.
The pattern repeats across decades. The 1994 cohort posted 79.6 percent one-year survival and 49.6 percent at five years. The 2020 cohort, which opened into pandemic disruptions, still posted 80.9 percent one-year survival. Table 7 of the Business Employment Dynamics program holds this record from 1994 forward.
How fast does risk drop after the early years?
Risk falls fastest at the start and then keeps slowing. For the 2013 cohort, 91.5 percent of the establishments alive in year five were still alive in year six. By year ten, the yearly survival rate among survivors had climbed to 92.3 percent. A business that clears its fifth year faces a materially lower chance of closing in the next one.
Survivors also tend to grow. Average employment at surviving establishments from the 2013 cohort rose from 4.4 workers at opening to 8.8 workers by year five and 11.9 workers by year ten, according to the same table. Survival and hiring move together in the record.
Where did the 90 percent failure myth come from?
Popular claims that most businesses collapse within a year do not match the federal tables. Every opening cohort listed in Table 7 since 1994 kept at least 76.8 percent of establishments through year one, and the figure stayed close to 80 percent across the decades. The 2022 cohort, the most recent one with first-year data, retained 818,549 of its 1,065,228 openings, a rate of 76.8 percent.
Part of the gap comes from what gets counted. The federal tables track employer establishments, not every sole operator. Casual sayings rarely name a source, a year, or a definition, which makes them impossible to check against any record.
What do the tables count and what do they miss?
The BLS series tracks establishments, which are individual business locations, drawn from quarterly employment records. A company that shuts one site and keeps another counts as a partial survival, so establishment closures overstate firm failures to some degree. The data also cover employers with paid workers, which leaves out nonemployers that never hire. We covered a connected angle in Pay transparency laws now cover a third of the U.S. labor market.
The Census Bureau offers a second lens. Its Statistics of U.S. Businesses series covers employer firms by enterprise size, with employment during the week of March 12 and annual payroll, and its latest published data year is 2022. Read together, the two programs describe the employer economy with unusual care.
Conclusion: What the survival odds mean going in
The record gives a launch team a calmer starting picture than the folklore does. Roughly four in five new employer locations clear the first year, half clear five, and a third reach ten. The danger window is the stretch from opening through year five, and the odds improve for anyone who gets there.
The small business failure rate, measured against real data, is a curve rather than a coin flip. Front-loaded losses dominate the arithmetic, and each survived year raises the next year odds above 90 percent for established cohorts. Planning capital and credit around the first five years addresses the risk where the data place it.
