What History Teaches About Pennsylvania's AI Chase

Pennsylvania is not alone in navigating the vast expansion of both private- and public-sector investment in artificial intelligence. States across the country are simultaneously grappling with how to attract more AI-focused investment and with the consequences of that investment. 

Competition for new investment, and the public sector’s ever-growing role in attracting investment, is not new. History can’t tell us what the prospects for AI are in the long run. However, history provides warnings about the pitfalls of narrow industry-targeting strategies and the risks of using public financing to attract new industries.

Pennsylvania's future Volkswagen plant would be the first foreign automaker to operate a production facility in the United States since Rolls-Royce operated a much smaller factory in Massachusetts between 1921 and 1929. The public was told that the plant would immediately create upwards of 5,000 jobs, and spur the creation of an additional 23-25,000 jobs in a new automobile manufacturing cluster. “We could have a mini Detroit,” predicted one Pennsylvania official.

The victory came at a cost to Pennsylvania and significantly altered the role of public money in private-sector investment to this day.

When Volkswagen announced in 1975 that it was looking for a location to build an entirely new automobile assembly plant, the frenzy it generated in state capitals was unprecedented. State competition for new investment was not new, and in many ways dates to the formation of the republic. But the size and scale of the prospective plant, and the changing nature of the public’s role in economic competition, marked something different by the 1970s. In the end, 47 states and Puerto Rico proposed specific locations. 

The process included state delegations and more than a few governors, travelling to Germany to personally sell their locations to corporate executives. What they also included in their sales prospectuses was unprecedented levels of public investment that they would provide to induce the decision on where the new plant would be located.

State investment for new investment was far from new in the mid-1970s. The modern era of state competition arguably extends to the efforts of southern states to induce northern factories to relocate. Led by Mississippi and the passage of the Mississippi Industry Act in the 1930s, states brought public money into what would otherwise be private-sector location decisions. For decades, those incentives included limited backing for industrial revenue bonds and other modest programs. The competition for Volkswagen’s future plant pushed states to dedicate unheard of levels of public money to influence corporate site selection decisions.

Volkswagen first narrowed the potential locations to sites in five states. Eventually, the choice came down to a site in Brook Park, outside Cleveland, Ohio, and New Stanton, where plans for a Chrysler Plant in the 1960s had been cancelled. Near the end, the media began reporting that the Ohio site was the likely winner, only for Pennsylvania to dramatically increase the stakes.

Both states had put together generous packages of public financing to entice Volkswagen. But when faced with losing the competition, Pennsylvania went for broke and increased the bid. The package of public funding Pennsylvania put on the table was valued at over $78 million, the equivalent of nearly half a billion in 2025. It was a bridge too far for Ohio to match, and Volkswagen selected Pennsylvania.

Investment in a new automobile plant seemed a sure thing, until it wasn’t.

What was missed in the calculations were the unavoidable vicissitudes of the free market. What appeared to be a sure source of job creation, not just for the present but for an almost unlimited future, was anything but. Volkswagen had jumped to build production capacity in the United States to fill an emergent market niche. The national energy crises of the 1970s had rapidly ramped up demand for fuel-efficient vehicles, a market that American producers had virtually abandoned. The plant in New Stanton would produce fuel-efficient VW rabbits.

At first, Pennsylvania's investment appeared to justify the immense cost. Operations at the site began in 1978. At its peak, the new plant employed over 6,000 workers, but it was not too last. In the 1980s, energy prices collapsed, and along with it, the demand for the fuel-efficient vehicles being produced in New Stanton. At the same time American producers had expanded options into competing. The new Volkswagen plant lasted one full decade before permanently shutting down in 1988. Had the short operational horizon been known, the calculations for providing public money to the project would have been even harder to justify.

The scale of investment that Pennsylvania risked to bring Volkswagen to New Stanton was unprecedented, but it soon set the example for future site selection competitions into the future. Major corporations now routinely expect and receive public financing packages from both state and local governments.

What lessons does the Volkswagen investment have for the pursuit of AI? Certainly, the peculiar history of a long-closed automobile plant tells us little about the future of a new emerging technology, particularly the impact of AI investment on local communities. Promoting AI is a complex issue involving the concentration of the most specialized workers, the development of research and technology ecosystems, and, as is dominating the news of late, the expansion of electricity-generating capacity. The unchallenged presumption is that demands for AI will only increase in the future, and with that investment will flow jobs. 

Arguably, the unknowns surrounding the future of AI far exceed the seemingly sure bet of increased automobile production in the 1970s. Given the much more rapid rate of change in AI, nobody can predict what role AI will play in local economic development in a decade, the span of time Volkswagen made cars in Pennsylvania.  

None of this argues against either public and private investment in AI, nor against the need for greater electricity generation across the nation. In an ideal world, a coherent national industrial policy would be guiding the expansion of electricity generation capacity across the United States. What we are left with is more haphazard public competition between states that may or may not make sense for all communities.

Promoting AI industries must remain just part of a comprehensive economic development strategy. There is no free lunch, and public investment in AI is likely to come at the expense of many other economic development priorities. Long-term growth will rely on encouraging innovation across many industries, strengthening supply chains, and expanding opportunities for workers.

Christopher Briem is a regional economist at the University Center for Social and Urban Research (UCSUR) at the University of Pittsburgh. His book: Beyond Steel – Pittsburgh and the Economics of Transformation, is being published next month.