Manufacturing Output vs. Jobs: Why U.S. Factories Make More With Fewer Workers

Two numbers get quoted in almost every argument about American manufacturing, and they seem to tell opposite stories. One says factories are thriving; the other says they're in decline. Both are right, because they measure different things. Output is how much U.S. factories make. Jobs is how many people they employ. Here's what the Federal Reserve and Bureau of Labor Statistics data show through September 2026.

Is America making more or less than it used to?

More, a lot more, than a generation ago.

Seasonally adjusted. September 2026 is preliminary.

Source: Federal Reserve G.17 via FRED (IPMAN)

View the data

Year (Sept.)	Index (2017 = 100)
1972	37.9
1973	41.2
1974	41.1
1975	37.2
1976	40.3
1977	43.9
1978	46.6
1979	47.1
1980	44.5
1981	46.2
1982	43.1
1983	46.9
1984	50.2
1985	50.8
1986	52
1987	55.5
1988	58.1
1989	58.2
1990	59.4
1991	58.7
1992	60.6
1993	62.4
1994	67.1
1995	70.7
1996	74.9
1997	81.4
1998	85.9
1999	89.7
2000	93.7
2001	88.9
2002	91.5
2003	92.6
2004	95.6
2005	98.4
2006	103
2007	106
2008	96.9
2009	87.8
2010	93.7
2011	96.2
2012	98.1
2013	99.6
2014	101
2015	100
2016	99.4
2017	99.7
2018	102
2019	98.9
2020	94.6
2021	97.1
2022	98
2023	97.3
2024	96.1
2025	98
2026	99.4

The Federal Reserve's industrial production index for manufacturing (FRED series IPMAN, where the 2017 average equals 100) measures the real, inflation-free volume of what U.S. factories produce. In September of each year:

• 1972: 37.9
• 1979: 47.1
• 1990: 59.4
• 2000: 93.7
• 2007: 106.0, the all-time high for a September reading
• 2009: 87.8, after the Great Recession
• 2026: 99.4

So U.S. factories now produce about 2.1 times as much as they did in 1979. The catch is the last 20 years: output is still about 6% below its 2007 peak and has moved sideways, mostly between 94 and 102, since about 2011.

The most recent months, from the Fed's October 16, 2026 G.17 release, show a small upturn: 98.98 in May, 99.27 in June, 99.43 in July, 99.18 in August and 99.44 in September. September 2026 was about 1.4% above September 2025 (98.05).

How many manufacturing jobs has the U.S. lost?

Factory employment tells the other story. The BLS payroll count (MANEMP) peaked at 19,553,000 in June 1979. In September 2026 it was 12,652,000.

The decline came in two very different phases:

1. 1979–2000: a slow slide. Jobs fell from 19.44 million (September 1979) to 17.23 million (September 2000), about 11%, while output nearly doubled.
2. 2000–2010: a collapse. Jobs fell from 17.23 million to 11.55 million in September 2010, a drop of about a third in a decade that spanned two recessions (2001 and 2007–2009).

Since 2010, jobs have recovered only partly, to about 12.65 million, roughly 10% above the 2010 low.

For the full 1939–2025 history, see U.S. manufacturing jobs since 1939.

How big is the gap between output and jobs?

Putting both series on the same scale makes the gap easy to see. We set each to 100 in September 1979, the year factory employment peaked:

OutputJobs

Our calculation: each September value of IPMAN and MANEMP divided by its September 1979 value (IPMAN 47.0605; MANEMP 19,442 thousand), times 100.

Source: Federal Reserve (IPMAN) and BLS (MANEMP) via FRED

View the data

Year (Sept.)	Output (1979 = 100)	Jobs (1979 = 100)
1979	100.0	100.0
1990	126.1	90.6
2000	199.2	88.6
2007	225.3	70.9
2010	199.2	59.4
2019	210.2	65.8
2026	211.3	65.1

By 2007, output was at 225 while jobs were at 71. In 2026, output stands at 211 and jobs at 65. In other words, U.S. factories produce a bit more than twice what they did in 1979 with about two-thirds as many workers. Divide one by the other and output per job is roughly 3.2 times its 1979 level (our rough calculation from the two series; it ignores changes in hours worked).

Why do factories make more with fewer workers?

The main reason is productivity: more output per hour of work. Several forces drive it:

• Automation and better machinery. Robots, CNC machine tools and computer-controlled processes let one worker oversee far more production.
• A shift toward more complex goods. Plants that make sophisticated products tend to need more machinery and fewer hands per unit of output.
• Offshoring of labor-intensive work. Industries that need many hands per unit, such as apparel, moved largely overseas. Our look at American-made socks and the U.S. textile industry shows how steep that fall was.

Productivity has been roughly flat recently. BLS's manufacturing labor productivity index (OPHMFG, 2017 = 100) was 100.38 in the second quarter of 2026, compared with 99.31 a year earlier, a gain of about 1.1%. That's in line with output growing a bit faster than jobs over the past year.

Does fewer factory jobs mean less is made in America?

No, and this is the most common misunderstanding. A "Made in USA" label today usually reflects a more automated, more productive factory than the same product made in 1979. Job counts are an important measure of how many families factories support, but they're not a measure of how much America makes.

That said, the 2007 output peak is a fair benchmark for skeptics. Output hasn't returned to it, and that stall since the Great Recession, not a long-run collapse, is the real story of the past two decades.

What does this mean for shoppers?

Buying American-made products supports both halves of the story: the plants that keep U.S. output high and the 12.65 million people who work in them. Hand tools are a good example of goods still made in American factories:

Lodge casts its skillets in South Pittsburg, Tennessee; Channellock makes pliers in Meadville, Pennsylvania; and Estwing makes hammers in Rockford, Illinois. Browse more in tools and home and kitchen, or see which states have the most manufacturing jobs.