Environmental Preferences and Technological Choices: Is Market Competition Clean or Dirty?

GAICIL Insights

Green consumers are often dismissed as virtue signalers whose choices change nothing.¹ This paper finds their signals have a powerful amplifier: competition. Environmental Preferences and Technological Choices: Is Market Competition Clean or Dirty?, by Philippe Aghion, Roland Bénabou, Ralf Martin and Alexandra Roulet, asks whether consumers' stated environmental concerns can shape what firms develop and sell, and how they interact with market competition in driving clean innovation.

In the model, agents care about both consumption and environmental footprint, and firms pursue greener products to soften price competition. Acting as complements, these forces determine research and development, pollution and welfare. The theory is tested with panel data on patents by 7,060 automobile sector firms in 25 countries, combined with measures of environmental willingness to pay and competition. Going green, in this model, is a way to escape a price war.

The predictions hold. Exposure to prosocial attitudes fosters clean innovation, all the more where competition is strong. Plausible increases in both together can spur clean innovation as much as a large fuel price increase. Results are robust to a broad set of indicators for environmental values and competition. They support models in which intrinsically and reputationally motivated individuals incur costs to act responsibly despite negligible impact on aggregate outcomes. Leveraged by strong competition, such motivations move markets, even at the upstream stage of research and development. The individual purchase is small; the competitive response is not.

The policy implication is notable. Although the analysis is primarily positive, the findings suggest that educational policies raising consumers' environmental awareness, together with more active competition policy, can be quite effective, especially when direct instruments such as fuel or carbon taxes face strong popular opposition. The combination offers an alternative or complement to price-based environmental policy. Where a carbon tax is politically dead, competition policy may still be alive.

The amplifier works for artificial intelligence too. Consumers increasingly care about the ethical dimensions of AI products, such as privacy, fairness and environmental footprint. In competitive markets, firms respond to such preferences by differentiating along these dimensions, investing in research to meet demand. Competition channels consumer values into innovation. Privacy can become a product feature firms race to build.

That is a market-based path to responsible AI. If consumers value privacy-preserving or energy-efficient AI, and markets are competitive, firms have incentives to innovate in those directions. Policies informing consumers about AI products' characteristics and sustaining competition could encourage responsible innovation without heavy-handed mandates. Disclosure plus rivalry may achieve what decrees struggle to.

AI's own footprint is the obvious test. Training and running large models consume substantial energy. Consumer and business demand for cleaner computing, combined with competition among AI providers, could drive innovation in energy-efficient AI, a welcome application of the mechanism the authors document.

The complementarity cuts both ways. Where competition is weak, consumer preferences have less effect on innovation, and concentrated markets may be less responsive to consumers' values. Sustaining competition matters for efficiency and for ensuring firms respond to what consumers care about. A monopolist can afford to ignore its customers' conscience.

The fuel-price comparison adds perspective. Price instruments are powerful and often face political resistance. Preferences and competition together can produce comparable effects on clean innovation. Policymakers have multiple levers, and combining them may achieve goals more acceptably.

Scale strengthens the evidence. Patent data from thousands of firms across many countries let the authors relate innovation to variation in attitudes and competition systematically.

Something for those who want greener to sit with: more private AI: the fastest route may run through antitrust and transparency, with fewer new mandates. Concentrated AI markets can afford to ignore what users value; competitive ones cannot.