ISSUE 005
Eighty-nine percent never hire anyone
An earlier entry in this pillar found that of Missouri's roughly 30 million American counterparts with no employees, most plan to close rather than sell. This is the other end of that: how many businesses ever hire anybody in the first place. The answer depends enormously on who you ask, and working out why was the useful part.
An idea worth thinking about
Two things associated with hiring your first employee: a spouse and a website
Census researchers looked at which characteristics were positively associated with a business crossing from no employees to having employees. Two of the ones they name are joint ownership between spouses, and having a website.
I find that pairing genuinely funny and also clarifying. Neither is visionary. Neither appears in anything written about entrepreneurial mindset. One is a second person with a stake in the outcome; the other is the minimum apparatus for being found by someone who isn't already a customer.
The heavy caveat: these are associations, not causes. Businesses that build a website may be businesses that were already more serious, better capitalised, or operating in industries where customers search online. Putting up a site will not cause you to hire somebody. But if you are wondering what separates the businesses that grow from the ones that don't, the observable answer is less romantic than the literature suggests — it looks like having a partner and being findable.
A tool I disappeared into
Search the Census Bureau's own research by topic
The Census Research Exploration and Analysis Tool lets you browse the Center for Economic Studies working papers by keyword. These are the papers written by Census researchers using the restricted microdata — the linked administrative records covering the universe of US businesses that nobody outside can access directly.
This is where findings appear years before they reach a headline, and where the ones that never reach a headline live permanently. Searching business formation turns up work like the two-stage entry analysis, which tracks the universe of business applications and their transition into employer firms — a view of the early business lifecycle that the standard statistics miss entirely, because most published business data defines a new business as the moment it goes from zero to one employee.
That definitional point is worth carrying around on its own. A great deal of what gets reported as business formation only counts businesses at first hire, which means the 89 percent that never hire anybody are invisible to the headline numbers about American entrepreneurship.
Learn something
What's known about the businesses that do hire
SBA's Office of Advocacy published a free report on the determinants of hiring a first employee, built on Kauffman Firm Survey data following a cohort of startups for seven years. Among businesses in that survey that did hire, the timing is concentrated early: about 38 percent hired within one year of startup and 54 percent within three. After that the annual probability falls steadily, while the rate of going out of business does not.
That shape is the useful part. The window in which a small business adds its first employee is mostly the first three years, and a business that hasn't hired by year five is statistically unlikely to start — while remaining just as likely to close as it ever was. Whatever the first hire requires, it appears to be something present early or not at all.
One finding with an obvious use: among businesses that received entrepreneurial training, 21.5 percent had hired an employee within five years, against 17.3 percent of those that hadn't. That's a real gap and a modest one, and it is once again an association rather than a demonstrated effect — the people who seek out training differ from the people who don't.
What we’re seeing
One federal source says 1.8 percent cross the threshold. Another says 59. Both are right.
Census researchers tracked the 2011 and 2012 startup cohorts for eight years using the Comprehensive Startup Panel. Their finding: about 89 percent of businesses started with no employees and still had none eight years later. Only 1.8 percent started as a nonemployer and transitioned to having employees. Just over 9 percent started with employees from the outset. And 66.4 percent — roughly 5.8 million businesses — started as nonemployers and did not survive the period.
Now the other number. SBA's Office of Advocacy, using Kauffman Firm Survey data over 2004 to 2011, reported that 59 percent of the nonemployers in that survey group hired a first employee, 13 percent never hired, and 28 percent ceased operations.
One point eight percent against fifty-nine. The explanation is the sampling frame and it matters more than either figure. The Census panel is administrative data covering essentially every registered business, including the millions of sole proprietorships that are somebody's side income and were never intended to become anything else. The Kauffman survey followed a recruited cohort of new businesses that agreed to participate in a multi-year study of startups — a group already filtered toward people who consider themselves to be building something. Older Census work using a different dataset landed near 3 percent over a three-year horizon in selected industries, and roughly 1 percent per year has also been cited.
So the honest answer is that both are correct about different populations, and which one applies to you depends on which you are. If you are running a business you intend to grow, the Kauffman range is closer to your world. If you are reading a statistic about American entrepreneurship in general, the Census figure is the one describing the actual population.
One last figure that reframes why any of this matters. Nonemployer startups that do eventually hire are not marginal to job creation: their share of total job creation rose from about 6 percent one year after startup to 14 percent by year seven, and about 20 percent of new employer firms had previously been nonemployers. The businesses that cross the threshold late are a meaningful share of where jobs come from — which is an argument for paying attention to the one-person business that has been going quietly for six years, rather than only to the ones that arrive already looking like companies.
One question
Do you want to hire someone, or do you want to stay one person?
Both are legitimate answers and the data treats them as the same event, which it isn't. Reply with which you're actually aiming at.
More in this pillar as it's written.
— Shana
Sources link to the original publisher, never to a summary.