Superstitions, street traffic, and subjective well-being

• Congestion costs are central to urban and transportation economics.
• We explore a measure of congestion costs based on self-reported happiness.
• We estimate these costs with random variation in traffic from driving restrictions.
• We find implied congestion costs several times larger than conventional estimates.

Congestion plays a central role in urban and transportation economics. Existing estimates of congestion costs rely on stated or revealed preferences studies. We explore a complementary measure of congestion costs based on self-reported happiness. Exploiting quasi-random variation in daily congestion in Beijing that arises because of superstitions about the number four, we estimate a strong effect of daily congestion on self-reported happiness. When benchmarking this effect against the relationship between income and self-reported happiness we compute implied congestion costs that are several times larger than conventional estimates. Several factors, including the value of reliability and externalities on non-travelers, can reconcile our alternative estimates with the existing literature.