I think it is interesting that people are so quick to understand that measuring productivity is difficult when it comes to software engineering metrics or how promotions can be scuttled by things like internal politics but when it comes to macro scale things suddenly we assume that populations are being accurately evaluated for productivity.