Drug Convictions as a Statistic for Other Crimes
Suppose we catch somebody dealing drugs. Does that tell us that that person has probably committed other, non-drug crimes that are harder to detect? (drug dealing is relatively easy to catch because [a] the seller is at risk from the buyer, and [b] the seller at risk from his supplier, and [c] the seller has to find customers). If so, then even if we didn’t care anything about drug dealing directly, it might be a good idea to lock up drug dealers.
This could be looked at with numbers. It would be a good research topic.