The US economy is a job creation miracle, adding more than 70 million new jobs since the late 1970s. Still, over the same time we’ve lost some 7 million manufacturing jobs. For every factory job we’ve lost, we gained another ten jobs in other occupations. These new occupations require different skills and are frequently located in different communities. This has been very disruptive to many people and places. 

There isn’t much serious debate about what has caused these factory job losses. While we lost some jobs to foreign competition, the vast majority – more than 8 in 10 lost factory jobs – disappeared simply because factories became more efficient. This technological improvement has a special name; it’s called economic growth. 

Economic growth is necessary path for a better more prosperous future, but the transition is not costless. Many of the jobs lost to automation and trade relied on very specialized skills that could not be easily transferred to another job. Knowing that this was a risk, Congress created several bi-partisan laws aimed at addressing job losses due to foreign trade. Sadly, this Trade Adjustment Assistance (TAA) is far less effective than anyone hoped.

During the 2000s, factory job losses were unusually fast, spurred by both fast-growing foreign competition and productivity gains. Yet even during this period the traditional TAA accounted for only 6 percent of the total benefits displaced workers claimed in the counties most impacted by imports. This program spent less on affected workers than did Social Security disability, new retirements, other assistance or Medicaid/Medicare. 

In hindsight, it appears the TAA rules were too restrictive. Most workers were not eligible for TAA because they lost their jobs not to imports, but to factories that are more productive. Ironically, even if fear of foreign competition accelerated factory automation, workers were not eligible for TAA. Consequently, many workers claimed disability or took early retirement. This was not good for workers, their families, the communities in which they worked nor state and federal budgets. 

The failings of the TAA also meant the economy was deprived of willing workers who could not afford the time or expense to retrain for another skilled job. Instead, many simply ‘skilled down’ into lower paying jobs. Worse still, the problem is almost certain to grow substantially. 

By nearly all accounts, disruption from automation will creep across much of the American landscape. As many as half of all US jobs will face some risk of automation over the next generation. Fortunately, some folks on Capitol Hill have recognized this problem. Last month, Senator Joe Donnelly, along with Michigan Senator Peters and New York Senator Gillibrand introduced “The Automation Adjustment Act” or the AAA. This legislation would extend trade adjustment assistance to workers who lose their jobs due to automation. 

In my nearly 20 years as an economics professor, I’ve written in support of only a handful of state or federal assistance programs. I’m pathologically averse to new government programs, but this just might be the most obviously smart spending bill I’ve encountered. Here’s why.

First, the AAA addresses an obvious and growing problem of automation-related job disruption. Second, it could well pay for itself by reducing reliance on SSA disability. If the AAA reduces the number of workers who go onto disability by even 3 percent, it should pay for itself several times over. Finally, and most importantly, it targets families in some of the most distressed and vulnerable communities in the nation. Senator Donnelly’s Automation Adjustment Act it is a significant policy innovation and should be passed into law this year.