It strikes me as you're forgetting that percentages must total to 100. When you reduce the other causes of death, as we have very successfully done with almost

It strikes me as you're forgetting that percentages must total to 100. When you reduce the other causes of death, as we have very successfully done with almost every single one of the causes of death in e.g., 1900 (see http://www.cdc.gov/nchs/data/dvs/lead1900_98.pdf http://www.cdc.gov/nchs/data/dvs/lead1900_98.pdf), then it follows from the fact that 100% of people die, that the few we haven't done as well against must be higher. What happens if we manage to half the deaths from heart disease, for example? Well, the % of deaths from cancer will increase substantially.