Why We Don’t Build Anymore

By Daniel Tenreiro

Monday, April 27, 2020

Marc Andreessen wants America to build again. In a recent essay, the technology entrepreneur attributes the U.S.’s lackluster coronavirus policy response to insufficient domestic manufacturing capacity. Shortages of cotton swabs, test tubes, and personal protective equipment, Andreessen argues, reflect a general inability to create tangible goods.

Andreessen’s diagnosis fits into the broader thesis of technological stagnation put forward by Tyler Cowen, Peter Thiel, and Robert Gordon, among others. On that view, sometime in the second half of the 20th century, innovation slowed dramatically, and productivity growth along with it. While the late 20th and early 21st centuries saw a boom in information technology, the Internet and software have done less to spur productivity than innovations such as the airplane that altered our physical world. As Thiel often points out, while the world of bits has transformed over the past 50 years, the world of atoms looks more or less the same.

Slower progress in hardware — or in “building,” as Andreessen puts it — has almost certainly decreased economic growth. Across a variety of metrics, including researcher productivity and patent approval, the rate of scientific progress has slowed, leading to worse economic outcomes, as Cowen and Ben Southwood found in a 2019 paper. Andreessen believes this stagnation is a choice: “The problem is desire. We need to *want* these things. The problem is inertia. We need to want these things more than we want to prevent these things.”

Why don’t we want these things? We used to. After World War II, U.S. corporations were in the business of building. Producing wealth meant producing tangible goods. General Motors and Ford consistently ranked among the world’s most valuable companies, and General Electric transformed American life by making cutting-edge industrial and consumer products — from jet engines to microwave ovens. Thanks in large part to the manufacturing prowess of U.S. corporations, real GDP doubled between 1947 and 1973.

But in the 1970s, businesses turned their eyes from innovation to efficiency. As the post-war boom came to an end and American firms faced increased foreign competition (particularly from Japan), they resorted to cost-cutting and financial engineering in order to stay competitive. Though innovation and efficiency are not mutually exclusive, research and development — especially breakthrough, early-stage research — decreases profit margins in the short term. Firms accordingly deemphasized new product development in favor of streamlining existing businesses.

The rise and fall of GE illustrate this phenomenon. In 1949, GE created the most popular jet engine in history, the J-47. The company went on to produce a slew of appliances that made onerous household chores a thing of the past, at great benefit to consumers as well as GE shareholders. But when Jack Welch took over the company in 1981, process efficiency dislodged product development as the company’s raison d’être.

Welch obsessively squeezed performance out of the business, firing the bottom 10 percent of employees every year and cutting inventories to the maximum extent possible. The company’s theretofore sleepy financing arm — GE Capital — became a profit center in the ’80s and ’90s. GE Capital spurred topline growth by underwriting the company’s sales and lending to a wide array of customers outside GE’s traditional businesses. At the company’s zenith, the financing business generated more than half of its profits. Welch’s process, mimicked by CEOs across the U.S., generated massive returns for shareholders. (Welch himself produced more value than any CEO in history.) Later on, during the 2008 financial crisis, the foray into finance would bring GE to the brink of bankruptcy. In effect, General Electric became a bank that just so happened to produce jet engines, leaving it extremely vulnerable to a financial downturn. The financialization of U.S. corporations also came at the expense of innovation, as financial engineering replaced kinetic engineering.

The production of wealth through streamlining — eliminating redundancies and offshoring — was powered in large part by globalization. In the economy of yore, physical and political constraints limited the extent to which businesses could grow through cutting costs and expanding their customer base. With the elimination of those constraints after World War II came access to labor and consumers around the world, providing a surefire way to grow revenue. (World Bank economist Antoine van Agtmael coined the term “emerging market” the same year that Welch began his tenure at GE.) Foreign competition increased the need to cut costs just as liberalized trade introduced an attractive alternative to toilsome R&D.

High-skilled workers flocked to the sectors that benefited from globalization. From 1980 to 1992, the financial sector’s share of corporate profits tripled, from 10 percent to 30 percent. Physicists who might have built aircraft in the pre-global economy went to Wall Street to build trading algorithms instead. In a self-reinforcing cycle, the international financial system underwrote an expansion of trade, which fed more money into banks. It helped that the end of the gold standard in 1971 allowed previously unsustainable debt levels to paper over declining technological progress: Household and corporate balance sheets ballooned from the 1970s to 2008, allowing GDP growth to outpace productivity.

Where innovation did take place, it was in areas that benefited from globalization. Internet businesses could access the world’s billions of customers for almost zero marginal cost, so tech talent flocked to the IT sector. As Andreessen put it himself in his widely read essay “Why Software Is Eating The World” (2011), “all of the technology required to transform industries through software finally works and can be widely delivered at global scale” (emphasis mine). The latter point — the rapid scalability of software — explains why a globally integrated economy benefits digital business the most. Larger markets can boost revenues in hardware, but they can expand margins only in asset-light businesses. Engineers have followed the money — to the detriment of productivity growth.

Supersonic aircraft, flying cars, and delivery drones seem like no-brainers. The economic incentives are in place: Consumers would pay for great new products. In the long run, growth depends on technology. Businesses that boost productivity by, say, increasing travel speeds could command astronomical valuations, but structural changes to the economy have reduced the relative returns of physical innovation.

Of course, the intrepid entrepreneur can still build: Elon Musk is a case in point. But I’d bet that Musk could have made a handsome fortune in consumer software without exerting (or embarrassing) himself nearly as much. For understandable reasons, his less obsessive counterparts build apps instead of rockets.

The twin forces of financialization and globalization have allowed us to put off the difficult work of building infrastructure and technology in favor of spreading “best practices” and shrewd accounting methods. Not only does this reduce long-term growth, but it leaves us more vulnerable to economic shocks. As the pandemic disrupts supply chains and financial flows, the wobbly foundation of the U.S. economy emerges. It isn’t pretty.