The competitiveness of the U.S. economy depends on technological progress, but recent data suggests that Innovation is getting harder and the pace of growth is slowing down. A major challenge in business and policy spheres is to understand the environments that are most conducive to innovation. One way to do that is to look to history. In our research we focused on the golden age of invention: the late 19th and early 20th centuries, when America became the world’s preeminent industrial nation.
The golden age is associated with some of America’s leading technology pioneers, such as Thomas Edison and Nikola Tesla in electrical illumination and Alexander Graham Bell and Elisha Gray in telephony. Our analysis goes beyond these well-known individuals. We built a systematic data set that contains millions of patented inventions and millions of individuals in Federal Censuses from 1880 to 1940. We also linked patent data to state- and country-level information. By analyzing this data, we were able to shed light on why the U.S. was so innovative.
The context for technological development was very different a century ago. For instance, in 1880 most inventive activity was the result of inventors operating outside the boundaries of firms. Research laboratories, such as the famous one opened, in 1876, by Thomas Edison in Menlo Park, New Jersey, were rare. From the middle of the 20th century, however, the modern corporation started to dominate patenting. By 2000 almost 80% of patents were assigned to inventors associated with firms.
Nevertheless, the impact of innovation on economic growth was typically large. The chart below illustrates a strong relationship between patenting activity and GDP per capita at the state level. It predicts that an innovative state like Massachusetts, which from 1900 to 2000 had four times as many patents as a less innovative state, like Wyoming, would become 30% richer in terms of GDP per capita by 2000.
Innovation was more prevalent in some areas than others. The map below shows regions that today are declining, such as the Rust Belt, used to be innovation hotspots during the golden age. Our research finds that innovation flourished in densely populated areas where people could interact with one another, where capital markets to finance innovation were strong, and where inventors had access to well-connected markets. States with a legacy of slavery were considerably less innovative, and religion had a negative effect, too, though to a lesser degree. Places that were economically and socially open to disruptive new ideas tended to be more innovative, and they subsequently grew faster.
Inventors in the golden age were overwhelmingly white and male. They were less likely to marry and they had fewer children, perhaps because of the time commitments associated with making technological discoveries. Inventors in U.S. history have tended to be highly educated, in contrast to the common portrait of the uneducated amateur. They typically invented in pursuit of profit, and the financial returns to innovation were large. The innovation sector was highly competitive. The best inventors survived. The worst exited quickly.
The family backgrounds of inventors were distinctive. Having a father who was an inventor increased the likelihood of becoming one, perhaps because fathers passed along their aspirations, or perhaps because it facilitated access to the right types of social networks. Fathers’ incomes were positively correlated with the probability of becoming an inventor. This means that talented individuals from low-income families were more likely to be excluded. (This remains the case today.) Much of the link between family income and invention appears to have been due to education. High-income families invested in the education of their children, and, in turn, educated inventors were more productive.
Our study also examined the relationship between innovation and income Inequality. New innovation is a disruptive force, which may reduce inequality or perpetuate it.
We found that the relationship between innovation and inequality depends on the type of inequality we’re talking about. Innovation was negatively correlated to the Gini coefficient, a broad measure of inequality. On the other hand, innovation was higher in places where the share of income held by the top 1% was larger, including in states like New Jersey, Massachusetts, and Connecticut, where patenting activity was extensive.
Our findings are consistent with two different approaches to thinking about inequality. If innovation is associated with financial rewards from patents and the associated monopoly rights, then we should see a positive association between innovation and inequality. But if innovation permits new entrants or small business owners to catch up with incumbent leaders, then innovation should lead to lower income inequality.
Our study is predicated on the idea that what made the United States an innovation powerhouse during the golden age is relevant to the way technological development progresses in the modern era. History matters because innovation and growth are largely about long-run changes. Creating an innovation sector that is both dynamic and inclusive was as challenging a century ago as it is today.