The ROCKWOOL Foundation Berlin finds that German regions with heavy steam engine use in the late 19th century now earn roughly 4.3% more and have higher shares of technically trained and university-educated workers. These areas also host more productive firms and generate more patents. Researchers say the pattern suggests that fast, early adoption of AI could similarly shape which regions reap long-term benefits — though outcomes will depend on policy, training and investment.
Steam Engine’s 150-Year Legacy: German Regions Still Earn More — Early AI Adoption Could Repeat The Advantage
The skyline of the banking district is seen during sunset in Frankfurt, Germany, April 21, 2024. REUTERS/Kai Pfaffenbach
A new study from the ROCKWOOL Foundation Berlin (RFBerlin) finds that regions in Germany which embraced the steam engine by the late 19th century continue to show measurable economic advantages today — a pattern the researchers say offers lessons for the likely long-term effects of early artificial intelligence adoption.
According to the study, areas where steam engines were particularly widespread at the end of the 19th century now have average wages about 4.3% higher than other regions. Those same areas also have a larger share of workers with technical training and university degrees, host more productive firms and register greater numbers of patent filings — roughly 150 years after the steam engine first spread across the country.
"The steam engine didn't just power factories; it also changed technology and the development of skills across entire regions over generations," said Christian Dustmann, director of RFBerlin and a professor at University College London. The study, shared with Reuters ahead of publication, traces how early technological adoption can reshape local labor markets and innovation ecosystems over decades.
Early AI Adoption Can Have Long-Term Positive Effects
Sascha Becker, the RFBerlin project lead, said the steam engine set off a self-reinforcing cycle: technological change encouraged skills acquisition, which supported further innovation and growth. He argued that this historical pattern suggests — while cautioning that it is still early to judge AI’s full impact — that rapid adoption of artificial intelligence could again help determine which regions and countries capture sustained long-term gains.
"These historical lessons suggest that rapid adoption of AI could once again help determine which regions and countries will benefit in the long term from the advantages of change," Becker added.
The authors stress that historical parallels are indicative rather than definitive: the exact outcomes will depend on policy choices, investment in training and infrastructure, and how widely and responsibly AI technologies are deployed.
Reporting by Rene Wagner; written by Maria Martinez; editing by Aidan Lewis. Research by ROCKWOOL Foundation Berlin (RFBerlin).
Help us improve.
