My friend from South Korea explained that to me. It's wild.
If researchers are just using government stats, they might be finding a completely meaningless 1.5 year difference in life expectancy.