[B]oth extreme heat and extreme cold result in immediate increases in mortality. However, the increase in mortality following extreme heat appears entirely driven by temporal displacement, while the increase in mortality following extreme cold is long lasting.
In other words, deaths from heat reflect what coroners call a "harvesting effect"; heat kills people who didn't have long to live anyway. The same doesn't hold for deaths from cold.
The life expectancy benefit of heat is large, too:
These longevity gains associated with long term trends in geographical mobility account for 8%-15% of the total gains in life expectancy experienced by the US population over the past 30 years.
Does this mean that global warming will, on net, be a good thing? No, but it's striking how few people even consider the possibility.
P.S. The paper also indirectly confirms my hypothesis that people prefer warmer climates, on average: People have been moving to warmer climates over the past thirty years.