Workforce and Labor

America Stopped Bleeding Factory Jobs. Nobody Noticed Because They Watched the Wrong Number.

American factory employment shed millions of jobs across the 1990s and 2000s, then bottomed and started climbing again. That turn is now over a decade old and it is still narrated as an ongoing collapse. The record says otherwise, and the state map says where it happened.

The most persistent story in American manufacturing is a decline that stopped happening more than a decade ago. The record is unambiguous: employment ran at 17,395 thousand when this archive opens in 1990, ground down through the NAFTA and China-WTO decades, and bottomed at 11,465 thousand in 2009. That is roughly 5.9 million jobs gone. Then it turned. Employment today is 12,611 thousands of employees (Jul 2026), about 1146 thousand ABOVE that trough, and it has been climbing for 17 years. The collapse was real. It is also over, and the sector is still discussed as though it were ongoing.

The national number hides the map

A recovering national line is not the same as a recovering country. Underneath it, some states have won back a real share of what they lost while others never stopped shedding, and the aggregate masks that divergence entirely. The state-by-state data is where the turn becomes legible: which regions reversed the decline, which are still losing ground, and where the sector is quietly reconcentrating. Note how concentrated it already is. For anyone making a siting, sourcing, or workforce decision, the national headline is nearly useless and the state map is nearly everything.

Why the reconcentration matters

Manufacturing clusters, suppliers, skilled labor, and specialized services gather around existing plants, so where jobs stabilize and grow is where the ecosystem strengthens, and where they keep falling is where it erodes. A state gaining manufacturing employment is building a deeper labor pool and supplier base that makes the next plant easier to staff and source; a state losing it is thinning the ground under anyone who depends on that workforce. The stabilization at the national level is real, but its benefits are landing unevenly, and the state data is the only way to see where.

America stopped losing factory jobs around 2009. The debate never got the memo, and it is still arguing about a trend that reversed while it was talking.

Reading the map for a decision

For a manufacturer, the practical use is to treat the state employment data as a leading input to siting and workforce strategy rather than a lagging curiosity. A state where manufacturing employment is stable or growing offers a labor pool and supplier network that a declining state cannot match, and that difference compounds over the life of a plant. The national stabilization is the headline; the state divergence is the decision. Anyone choosing where to build, expand, or source who reads only the aggregate is navigating with the one number designed to hide exactly what they need to know.

The decline, the bottom, and the turn

The 37-year record shows factory employment making a full round trip, which is why point-in-time comparisons mislead so badly here. Its high came at the close of 1997 around 17,587, gave way over the following years to 11,465 by the end of 2009, and has climbed since to 12,611. That leaves it 28% below the peak and well off the floor, so whether today looks high or low depends entirely on which year you anchored to.

Explore manufacturing employment by state alongside the national series on the live data pages. See the state map

Published 2026-08-06.