Scott Alexander questions the plausibility of multifactorial explanations for trends, using the US crime rate decline as an example, and seeks literature on comparing single-factor vs multi-factor explanations.
Longer summary
Scott Alexander discusses the plausibility of multifactorial trends, using the decline in US crime rates as an example. He presents two perspectives: one arguing that complex phenomena like crime are likely caused by many small factors, and another suggesting that simultaneous changes in multiple factors is improbable. Scott leans towards the second perspective, questioning the likelihood of ten different factors each accounting for about 10% of the crime decline. He asks if there are ways to calculate the relative likelihood of single-factor versus multi-factor explanations, noting that he feels he's missing an existing body of literature on this topic.
Shorter summary