Solar panel sales are driven by the most skeptical neighbors, while the eco-friendly crowd has almost zero influence on community trends.
April 26, 2026
Original Paper
Who Follows The Leader? Evidence From a Novel Machine Learning Approach to Studying Heterogeneity in PV system adoption
SSRN · 6615038
The Takeaway
Peer influence for green technology works in the opposite way that most marketing experts assume. Social proof is weakest among people who are already predisposed to care about the environment or tech trends. The most powerful shift occurs when older, low-energy users who have no interest in going green see their peers making the switch. These hesitant adopters wait for a signal from someone exactly like them before they feel safe enough to change their behavior. To accelerate the transition to renewable energy, companies should stop targeting the enthusiasts and focus entirely on the people least likely to sign up.
From the abstract
While a growing body of literature explores heterogeneity in peer effects on photovoltaic (PV) system adoption, most studies rely on strong a priori assumptions and struggle to flexibly capture complex interactions between household characteristics. In this paper, we use a novel Causal Forest approach to estimate heterogeneous peer effects, leveraging exogenous variation from public housing PV installations as a natural experiment. By combining machine learning with a residualization strategy to