Hmm. Not sure here. As the average life expectancy increases, this should also delay the demise of those who are below the mean. Unless all the increase is taken up by those who are already healthy but live longer due to advances in medicine, the same advances must also affect the bottom end of the distribution curve. The average life expectancy over the last generation has increased by at least half that number of years, so a 5 year increase in the retirement age shouldn't result in more people dying in employment. Given the acknowledged difference between the life expectancies of those at opposite ends of the social spectrum, it is more likely that those dying younger will be unemployed in the years leading up to the age of 70 - so probably irrelevant to insurance companies paying out for those dying whilst in companies' employment.