The early 2000s recession was notable for a "jobless recovery," i.e. economic output bounced back fairly quickly but employment was slower to follow. It was also the start of a steady and unprecedented decrease in workforce participation, reversing decades of growth as women joined the workforce, primarily driven by men without college degrees leaving the workforce. https://www.epi.org/publication/issuebriefs_ib186/ https://fred.stlouisfed.org/series/CIVPART https://www.journals.uchicago.edu/doi/full/10.1086/700896# Note that U.S. population was growing steadily over the entire period of the graph you link to. In fact, it more than doubled between that peak in WWII industrial employment around 1943 and 2000. The fact that manufacturing jobs were essentially flat over that period already reflects a massive fall in the fraction of Americans that were working in manufacturing, and the fact that there WERE new manufacturing jobs/factories built anywhere in the U.S. means that a lot of jobs already were lost in the Rust Belt. But, why did such a sharp loss happen when it did, given the preceding decades of industrial decline and plenty of other recessions? Some argue that improved productivity was making many jobs redundant. Some attribute it to rising competition from China, India, and other nations. There was a housing bubble at the same time, so the construction industry was soaring, which if nothing else less obvious that manufacturing jobs were drying up. https://www.industryweek.com/the-economy/article/22006840/why-the-2000s-were-a-lost-decade-for-american-manufacturing https://www.aeaweb.org/research/did-the-us-housing-boom-obscure-manufacturing-bust https://fred.stlouisfed.org/series/USCONS