Cities, with their dense mixtures of people and economic activity, have long been fonts of innovation. To start, density spurs innovation by pushing people and ideas together, enabling them to combine and recombine in new ways. And advances in transportation—from railroads and subways to automobiles, planes, and high-speed rail—increase the circulation not only of goods and people, but of ideas as well.
Most of the evidence connecting urbanization and innovation focuses on recent times. Now a study by Elisabeth Ruth Perlman, a Boston University economist and doctoral candidate, documents the ways that urbanization and transportation spurred American innovation in the late 18th and 19th centuries.
The study examines the connection between innovation, urbanization, and transportation over the period stretching from 1790 to 1900. According to economists and economic historians such as David Landes and Joel Mokyr, this was perhaps the most inventive period in American history, beginning just after the American Revolution and running through the Civil War and the Industrial Revolution. It was also the century when America began its shift from a rural to an urban nation. Just 5 percent of Americans lived in urban areas in 1790. By 1900, urbanites were nearly 40 percent of the population, according to the U.S. Census. New York City alone ballooned from 33,000 people in 1790 to over 3.4 million in 1900—an increase of over 10,000 percent.
To get at the connection between cities, transportation, and innovation, Perlman’s study combines several unique data sets. She begins with detailed data on more than 700,000 patents—which she uses as a proxy for innovation—for all U.S. counties from 1790 to 1900. (Patents are the most accessible written records of innovative activity: Even though they are imperfect, economists and social scientists have long used them to study the dynamics of technological change and innovation.) Perlman then compares this to data on the accessibility of canals, railroads, and other forms of transportation; distance to major ports; and the key terms used in patent descriptions, which enable her to examine the effects of local transportation on the speed of innovative activity.
The first chart below shows the connection between innovation, population, and transportation. From 1780 to 1900, both the number of patents and total population saw an overall trend upward, with a sharp spike around the 1850s due to an increase in railroad access.
Urban historians have long identified the role of new transportation systems like canals and railroads in spurring urban development. But Perlman’s study finds a connection between these transportation improvements and innovation. She observes that the arrival of a canal in a county is in fact positively associated with more patenting activity. And, in the 20 years following the arrival of a railroad, counties saw a twofold increase in their number of patents per capita. In fact, the study finds that 8 percent of the increase in 19th century patents was the result of the spread of transportation.
The two maps below show the connection between innovation and transportation systems by plotting the rate of patenting per 10,000 people by county, while highlighting major railroad lines. Notice how innovation frequently clusters along these major railroads.
First, take a look at the map of 1840: Innovations are highly clustered in the Northeast and New England, and then begin to spread out along railroad lines into the Great Lakes states and even farther-flung parts of the South and Midwest.
Now look at the map for 1860, just before the Civil War. This same basic pattern continues, although innovation hotspots are pronounced along canal lines (shown in blue) as well as railroad lines (shown in red). Twenty years year, patents were increasingly concentrated in the same Northeast and New England regions, while also expanding farther into the Midwest. Western states like California, Arizona, and New Mexico and Southern states like Texas, Florida, and Louisiana also saw increased innovation.
Interestingly, Perlman’s research shows that the geographic concentration of patents declined steadily throughout the 19th century. This seems counterintuitive to the idea of innovative clustering. A big impact of advances in transportation was that it enabled innovation to spread from its initial sources in New England and the Northeast to generate new clusters in more rapidly growing parts of the country.
Overall, Perlman finds transportation to be a key determinant of innovation in rapidly industrializing and urbanizing America. But the mechanism by which this occurs is even more interesting—and more important—to our understanding of innovation. Her analysis shows that, though both matter, innovation is not just spurred by population, nor is it merely a function of access to larger markets brought on by transportation. Instead, it has subtler effects. As Perlman explains in an email, “If transportation simply runs through a place, it doesn't do any good unless it also forms a nexus around which agglomerations occur. This is shown in the different effects estimated in the North and the South.”
Perlman than examines the connection between new transportation and the flow of information. To get at this, she tracks the spread of key words and phrases in more than 700,000 patents to determine how quickly a new idea or technology spread into an area, and to what degree new forms of transportation contributed to this expansion. Surprisingly, she finds that transportation doesn’t seem to be encouraging more references to new technologies. If anything, areas that are newly connected to transportation incorporate technology at a slower rate than they did before.
Still, advances in transportation help to create the the kind of urban environment that can speed the arrival and pick-up of new innovation. “Because transportation encourages urbanization,” Perlman writes, “it encourages patenting in many ways. Urban areas allow for greater specialization, which might encourage patenting by giving people in those areas better access to the bureaucracy of patenting (e.g. lawyers, machinists, draftspeople), or by encouraging innovation directly.”
What matters most is not the ability to trade with a large area, but the way in which urban areas and advances in transportation accelerate the movement of goods, people, and especially ideas. What’s more, these advances are not solely the product of history. While the car facilitated the spread of innovation to suburban nerdistans in the 1970s, ‘80s, and ‘90s, the past decade or so has seen innovation and high-tech startup companies move back to urban centers near transit lines where people and ideas can connect more efficiently. Perlman’s detailed research shows the key role of urban areas, transportation, and connectivity in spurring American innovation throughout the nation’s history.