While Facebook was buying Instagram on Monday, a couple of my favorite wonky policy journalists were hailing the White House’s plans to revive American manufacturing. Specifically, Ed Luce from the Financial Times and Ezra Klein of the Washington Post were both very taken with National Economic Council director Gene Sperling’s recent speech (PDF) that attempted to put some analytic meat on the bones of President Obama’s manufacturing-heavy State of the Union address.
I don’t share their enthusiasm. The extra effort that went into Sperling’s speech raises the troubling possibility that these ideas will actually guide policy in a second term rather than simply serve as props in a re-election campaign. It is sensible for public policy to pay attention to the creation of great firms, to strength in specific sectors, and to the quality of the jobs generated by different economic models. But it should be obvious that the path forward for America is to focus on our strengths in information technology and media, and not compete with the Chinese for manufacturing supremacy.
Klein’s gloss on Sperling’s argument is that there's “a market failure” regarding manufacturing in which an open marketplace “is failing to appropriately price the benefits of manufacturing firms.” So America should push for more manufacturing jobs because such jobs are worth more than they appear to be worth. Such things can happen. An unregulated market leads to overproduction of air pollution, so you need to tax it, and underinvestment in children’s education, so you need to subsidize it. If an unregulated marketplace underproduces manufacturing firms or establishments, then we should subsidize factories just like schools.
Sperling’s case rests on two main legs—externalities associated with research and development and externalities associated with what’s known as “agglomeration.” The basic idea in both cases is that spillovers from manufacturing benefit the rest of the economy.
On the R and D front, it’s clear that companies that come up with great ideas can’t capture them all. New inventions prompt imitators and new processes have a way of leaking out as workers switch jobs. When flat panel high-definition televisions came onto the market, it wasn’t just one company that knew how to make them. Suddenly everyone was making them. That’s R and D spillover. Agglomeration spillover helps explain why the two most successful TV makers are both in South Korea. It’s not just a weird coincidence, it’s something you see all the time. Sperling notes that researchers have found that “spillover benefits decline with distance, indeed by over half when they are more than 700 miles away” and tend not to cross national boundaries. That’s because spillovers are fundamentally made of people—gossip over dinner, workers and managers drifting from one firm to another, casual inspection of the other guy’s setup, etc.—and people don’t move around that much. Long story short, if the awesomest, most innovative widget-making factory in the world is in your country, that gives you a huge leg up on your odds of becoming home to a disproportionate share of the next 10 awesome widget factories.
The problem is that none of this has much to do with manufacturing. If you want to subsidize R and D, then subsidize R and D—there’s no need for the backdoor of an across-the-board subsidy to factory owners regardless of how much R and D they actually do.
On agglomeration, the irrelevance of manufacturing per se is even clearer. It’s not a coincidence that Twitter, Apple, Google, and Facebook are all located on a narrow corridor between San Jose, Calif. and San Francisco, that all the movie studios are in Los Angeles, or that nonlocal journalism happens overwhelmingly in New York and Washington, D.C. Industry clusters happen in all sectors. But if you look at America’s metropolitan areas, it’s clear that manufacturing-oriented places are relatively poor. The wealthy clusters in the United States are built around things like software, biotechnology and medical devices, higher education, finance, and business services. Places like California, Minneapolis, Seattle, and the Northeast corridor are far richer than the factory-oriented Rust Belt and Southeast.
Sperling notes that “if an auto plant opens up, a Wal-Mart can be expected to follow,” but that opening a Wal-Mart doesn’t bring an auto plant. But, again, this highlights the need for communities to have firms that are competitive in global markets, not manufacturers per se. A global firm—be it Ford or Amazon or Paramount—calls into being a local economy of shops and barbers. But the road to a more prosperous America is to learn from the most prosperous parts of the country, not to imitate Chinese clusters that are even poorer than America’s industrial hubs. The potential of America’s most productive places is tragically limited by restrictive zoning policies that keep the cost of living high and population growth low. The number of American students getting degrees in computer science and other technical fields is actually falling even as the number of people going to college grows. Short-sighted politicians are underinvesting in the transportation infrastructure even as people need to access our most vibrant labor markets. These kinds of issues don’t do as good a job of addressing the anxieties of Midwestern swing state voters as visits to lock-making factories, but creating new billion-dollar software startups has a lot more to do with the future of American prosperity.
The scary thing about the factory-driven view of the American future is that it’s not totally implausible. The “insourcing” trend where firms move production back to North America is real enough. The drivers are rising Chinese wages and falling “unit labor costs” in the United States. But that’s just a way of saying that America can regain factory parity with China by eliminating the prosperity gap between our two countries—a very strange policy aspiration. Most likely there’s nothing we can do to prevent some narrowing of the gap, which will have the consequence of bringing some jobs back. But we should measure our success by the extent to which this doesn’t happen, and we instead build and expand new industries that push living standards up and keep factory owners searching abroad for cheap labor.