Booms, not Busts, Drives Innovation, Especially in Mid Size American Cities
Mid-sized American cities outperform major metros at turning economic growth into patents — challenging where we think innovation happens.
New research provides ammunition for spreading federal R&D dollars beyond Silicon Valley. Economists Federica Coelli (EBRD) and Paul Pelzl (NHH Norwegian School of Economics) studied 2.5 million patents across 759 U.S. communities over 40+ years. Their finding: smaller urban areas innovate effectively when economies improve.
Current reality: Just 5% of U.S. communities produce 75% of all patents.
By the numbers:
The boom effect:
8.3% increase in overall patents when oil/gas employment doubles
8.5% DECREASE in oil & gas patents during their own boom (the paradox)
2x stronger innovation response in non-metro vs. metro areas
1 additional patent per 100,000 residents during booms
Economic impacts:
3.7% employment increase, 2.2% wage growth during booms
4.6% GDP jump, 6.2% local government revenue surge
Who actually innovates:
58% of patents from incumbent inventors (not newcomers)
5% from in-movers, 37% from first-time inventors
57% company patents, 32% individual, 2% university
Study scope:
2.5 million patents analyzed over 40+ years (1969-2012)
759 U.S. commuting zones studied
5% of zones currently produce 75% of all patents
The research design:
Researchers used oil and gas booms as natural experiments through a sophisticated shift-share methodology, not simple correlations. These booms create substantial economic shocks across communities: population increases 1.9%, employment rises 3.7%, wages grow 2.2%, personal income climbs 1.8%, GDP jumps 4.6%, and local government revenue surges 6.2%.
The study's precision comes from tracking patents by filing date (when innovation actually occurs) rather than grant date, and using fractional counting when multiple inventors or locations are involved. The results remain robust whether researchers use oil prices instead of employment or control for coal booms.
Surprising finding: Oil companies innovate less during booms
The paradox: While overall patents surge 8.3%, oil and gas companies cut their own patenting by 8.5% during boom times — even as they dramatically increase extraction activity.
The numbers tell the story: When oil prices and national oil & gas employment rise together, sector profits soar. Companies respond by drilling more wells and pumping more oil, not developing new technology. This direct trade-off between immediate profits and future innovation provides rare empirical evidence for economic theory.
Why this matters: Economists have theorized that firms innovate during downturns when the opportunity cost is lower (Dave’s note: Economists really REALLY need to read more Operations Research stuff like Deming or Beer who would quickly throw cold water on this nonsense, why would anyone spend money on investments when they are having a hard time paying bills or corporate executive bonuses?) — you develop new technology when you can't sell as much product. But most studies find the opposite: innovation rises during good times. This oil sector finding proves the opportunity cost theory actually works (Dave’s note: Why? It’s going to be an interesting research project in itself considering OPEC and greater oil and gas consolidation is a thing) — it's just usually overwhelmed by other factors like better financing and agglomeration effects in other industries.
What it reveals: The oil paradox helps solve a decades-old puzzle about innovation timing. Industries closest to the boom, facing the highest opportunity costs, reduce innovation. Industries further from the boom's center benefit from prosperity without the same profit pressure, so they increase innovation. This explains why studies find conflicting results — it depends which industries and regions you examine.
Three major cycles studied:
Late 1970s-early 1980s oil boom
Extended decline through late 1990s
2000s fracking revolution
Natural experiment strength: 35% of oil/gas reserves were "undiscovered" in 1960 — pure geology, not exploration efforts. Shale reserves existed underground throughout but became economically viable only with 2001 fracking technology. This mirrors how innovation capacity might remain dormant in regions until economic conditions activate it.
Prosperity drives innovation, not hardship
Conventional wisdom: Constraints force creativity. Scarcity sparks breakthroughs. Economic pressure drives invention.
The data shows otherwise. When booms increase wages 2.2%, personal income 1.8%, and GDP 4.6%, innovation rises proportionally. Patent quality remains stable during good times, measured by forward citations. The 1984-2000 downturn saw innovation decline symmetrically — no silver lining of constraint-driven creativity emerged. Even as local government revenue rises 6.2% during booms, private innovation continues to flourish rather than being crowded out. The stark reality: 97% of patents come from companies with capital to invest, not struggling inventors in garages.
How it works: Agglomeration effects
Booming areas see increases in both college graduates and creative workers (up 1.8%), creating a richer environment for innovation. But the gains don't come from attracting star inventors from elsewhere. Instead, existing local inventors become more productive, driving higher patents per capita alongside total patent growth.
The strongest effects emerge in urban areas with median populations around 75,000 — large enough for meaningful knowledge spillovers but small enough to avoid the congestion costs of major metros.
Winners and losers by distance:
Winners:
The booming community
Neighbors within 100 miles
Losers:
Communities 300-400 miles away lose patents
Too far for spillover benefits, close enough to lose talent
Political reality: 8 of 10 people moving to boom towns come from the same state.
Industry breakdown:
Innovation winners:
Highly traded goods: watches, x-ray equipment, aircraft parts
These sell nationally, so local demand doesn't affect them
No change:
Local-serving industries: concrete, ice manufacturing
Oil supplier industries (surprisingly — no "rising tide" effect)
Note: Upward wage pressure from booming oil sector could hurt highly traded manufacturers, but evidence shows only weak production worker displacement and no crowding out of innovation workers (since oil & gas patenting actually declines).
Big companies:
File 53% of all patents
Average 20 patents/year
Spread across 3.3 locations
Keep 71% of innovation at headquarters
Other key findings:
Prior patenting experience matters far more than education levels for capturing boom benefits — areas with innovation history see bigger gains regardless of local college presence. The mechanism isn't financial: public and private companies respond identically to booms, ruling out credit constraints as the driver.
The innovation response lags economic upswings by 2-3 years, consistent with typical R&D project timelines. And the effects show perfect symmetry — economic busts reduce patenting by exactly the same magnitude that booms increase it.
The academic debate:
Berkeley economist Enrico Moretti has argued that concentrated tech hubs produce more innovation per dollar invested — an efficiency case for keeping R&D funding in Silicon Valley. Harvard economists Edward Glaeser and Naomi Hausman warned that spreading wealth to struggling regions might actually reduce work incentives and decrease innovation.
The data proves both critiques wrong. Prosperity boosts both productivity and patent output across diverse regions, with mid-sized cities showing the strongest responses.
Bottomline:
America has vast untapped innovation potential — 95% of communities produce just 25% of patents. Mid-sized cities might not match tech hubs city vs city, but on a per dollar basis it’s a different story. Turns out that focusing on local economies helps "jump-start the American growth engine”.
Requirements for success: Existing innovation infrastructure and acceptance that gains in one region mean might be losses elsewhere — particularly within state boundaries.
For policymakers: The question isn't whether to spread R&D funding. It's how to manage geographic trade-offs when booms, not busts or “capital discipline” drives innovation. With 18% of America producing no patents annually, the opportunity — and challenge — are clear.