Aghion’s message, pitched before Monday’s headlines, is timely for bankers, payments executives and FinTech builders now navigating a world where governments are writing big checks for chips, clean tech and health innovation.
In “Competition-Friendly Industrial Policy,” Aghion, Mathias Dewatripont and Patrick Legros said industrial policy doesn’t have to mean picking national champions or protecting laggards. When designed to work with competition policy, targeted, sector-specific support can speed innovation and raise productivity. The authors blended theory with evidence from China, and then pointed to the U.S. Advanced Research Project Agencies (ARPAs) as a template, as its focused public funding invites multiple teams to compete, with milestones and accountability.
Europe’s productivity slump versus the U.S. partly reflects the absence of a true European ARPA-style engine for “frontier” innovation, they said. For financial services readers, the significance is that smarter public incentives can catalyze the next wave of digital infrastructure, such as artificial intelligence, cybersecurity and payments rails, without dulling market rivalry.
Among the key takeaways:
- Competition and industrial policy are complements, not opposites. The authors’ model showed firms invest more in breakthrough ideas when they face real rivals, not cozy monopolies. Policy should push activity toward the markets where innovation pays off most, and support should be spread across firms rather than concentrated in a single “national champion.” That dispersion, coupled with strong antitrust guardrails that make collusion unlikely, produces more innovation and faster diffusion.
- Design details change outcomes. China’s experience proves it. Studying Chinese firms from 1998 to 2007, the authors found that some tools, like direct subsidies and tax holidays, are associated with higher productivity (total factor productivity), while others, like cheap loans and tariffs, aren’t. Results are stronger when support is dispersed and when smaller firms are eligible. The policy miss is that authorities didn’t steer more aid to the most competitive markets, leaving gains on the table.
- The ARPA playbook shows how to “pick winners” competitively. BARDA’s COVID-19 program concentrated large, milestone-based funding across six vaccine teams and three technologies, including mRNA, viral vector and protein subunit, inviting global entrants to compete. That included multibillion-dollar bets such as approximately $2.48 billion for Moderna and about $1.95 billion for BioNTech/Pfizer. All six projects ultimately secured authorization in the U.S. and/or the European Union. This is industrial policy that keeps rivalry alive while solving coordination bottlenecks.
Other Notable Findings
The paper highlighted European innovations in procurement. Joint EU vaccine purchases improved access and created buyer leverage, an approach the authors said could help tame prices for high-cost rare-disease drugs by reducing “divide-and-conquer” pricing across countries.
Advertisement: Scroll to Continue
They also flagged a governance fix. “Me too” products can be curbed by tightening approval standards so public money and private R&D chase genuine breakthroughs, not incremental tweaks.
Finally, they underscored a political economy point. Channeling aid to more competitive, less concentrated sectors can reduce rent-seeking pressure and lobbying distortions. The larger European takeaway is that today’s programs like the European Innovation Council serve useful purposes, but the region still lacks a DARPA/BARDA-style engine for disruptive bets in strategic sectors.
For the full argument, data and policy blueprint, including the China evidence, the page-by-page design checklist, and the BARDA case study, read “Competition-Friendly Industrial Policy” by Philippe Aghion, Mathias Dewatripont and Patrick Legros in the May CPI Antitrust Chronicle.
