I understood his claim a bit differently, namely as "Lead is an epiphenomenon, and a single variable with a big effect is likely to be an epiphenomenon". To be

I understood his claim a bit differently, namely as "Lead is an epiphenomenon, and a single variable with a big effect is likely to be an epiphenomenon". To be sure, I think the logic goes like this: - Lead is positively correlated with crime - Reduction in lead-levels is associated with multiple factors that are in turn associated with gentrification - Thus, these gentrification factors are the true cause, and lead-levels just happen to correlate with them.