946,590 Workplaces Opened Outside America’s Metros in 28 Years. 947,046 Closed.

Census has counted both sides since 1978. Fewer businesses close out there than in the cities, and have for most of four decades. It is the opening side that never kept up.

A count that has not moved in a generation

In 1995 there were 366,221 workplaces outside America’s metropolitan and micropolitan areas. In 2023 there were 366,033. That is 188 fewer after 28 years, which for a number that size is the same number twice. It did not sit still the whole way: the count climbed through the late nineties, peaked at 387,244 in 2007, fell through the financial crisis and the decade after it, and has been clawing back since 2021. It is still 21,211 workplaces — 5.5% — below where it stood in 2007. Over the same 28 years the metro count went from 5,552,591 to 7,091,407. The gap is not that non-metro America shrank. It is that it did not grow while everywhere else did, which over a generation amounts to the same thing: its share of all American workplaces fell from 6.18% to 4.88%.

It is not that more of them close

The obvious explanation is failure, and the data says no. Census publishes an opening rate and a closing rate for each side of the line, both measured against the same denominator, so the two can be subtracted from one another honestly. In 2023 the closing rate outside the metros was 8.60% against 9.44% inside them. Fewer of them shut, not more — and that is not a one-year fluke: the non-metro closing rate has been the lower of the two in every year since 1987 except 2006. It was the higher one through the farm crisis of the early eighties, and it has not been since. A second, independent measure agrees: 6.76% of non-metro workplaces belonged to a business that shut down entirely in 2023, against 7.18% of metro ones. The opening rate is where the two Americas part. It was 9.26% out there against 10.68% in the metros, a gap of 1.42 points, and it has been the lower number in 45 of the 46 years on record — 1984 the only exception. Subtract one from the other and the net rate is 0.66% against 1.23%. Averaged across the whole 28 years the non-metro net rate is −0.003 points a year. That is what a flat line is made of.

Nothing out there is failing faster. There is just less of it starting.

Which businesses are actually out there

Non-metro America holds 4.88% of all American workplaces, but it is nowhere near that evenly spread. In agriculture, forestry and fishing, 26.8% of every workplace in the country sits outside a metro area. In mining, oil and gas it is 20.2%, and in utilities 15.8%. At the other end, management of companies runs 2.2%, educational services 2.4% and professional and technical services 2.6% — a twelve-fold spread from top to bottom. The industries that hold up out there are the ones tied to land and to infrastructure that cannot be moved. The ones that do not are the ones that can sit anywhere and choose a city, which is also the fastest-growing half of the economy. That is the arithmetic behind a flat count: the non-metro business base is weighted towards industries that are not adding establishments anywhere.

What a flat count means if you are the one running the business

A number that has not moved in 28 years is easy to read as stability, and from the inside it is not. The business that survives out there is genuinely more durable — lower closing rate, every year for a generation. What is missing is everything around it. Fewer new workplaces opening on the same street means fewer suppliers to switch to, fewer people who have just done the thing you are about to do, fewer buyers when you want out, and a smaller pool of anyone who has run a business locally at all. The metro workplace also sits in a bigger crowd: 17.9 employees on average against 12.1 out there. Resilience is not the same as support. The Census numbers show a place full of businesses that are very good at not dying and a place that has stopped producing new ones — and those are the same place.

Four limits, and one that was checked and cleared. First, these are establishments — workplaces — not companies. Census also publishes a firm count on this cut and it does not add up: a company with sites on both sides of the line is counted on both, so the three parts over-sum the total by 21,128 firms in 2023. Every number here is an establishment count, which partitions exactly. Second, the flow pair (946,590 opened, 947,046 closed) and the stock pair (366,221 then 366,033) differ by 268 because openings and closings are measured March to March while the count is a single March snapshot; both say the same thing and neither is the other. Third, the survey carries a third bucket beyond metro and non-metro — “statewide” establishments, 0.50% of all workplaces but 4.32% of all employment, because they average 158 employees each. They are excluded from both sides above rather than folded into either. Fourth, this is a different universe from the workplace counts in our earlier pieces, which came from the Bureau of Labor Statistics; the two agencies count different things and their totals should not be set against each other. The cleared one: metro boundaries get redrawn every decade, which could have moved counties across the line and faked this whole result. It did not. There is no step in the series at either redrawing — the largest single-year move in 46 years is 1984, and the metro side moved just as much that year, so the boundary is applied as one fixed definition throughout.

Published by The Lonely Entrepreneur — the community and coaching platform for entrepreneurs who are building alone. lonelyentrepreneur.com

This article is for educational purposes and is not a substitute for professional financial or legal advice.