It's probably because life expectancy, like maternal deaths (of which the US has U.S. the worst rate in the developed world), can't be faked and spin-doctored by government sources as easily as can unemployment rates.
And by "it's", I refer to the fact that this is known, not to the reason why this happens. Unemployment rates, declining middle class, bleeding working class, and people working shitty subsistence jobs that count as "full employment" would be my guess for the latter (and all that comes from eat: low self esteem, depression, drinking, drugs, poor health coverage, homelessness, et al).
Actually it's easy. Just take maternal deaths. The US is pretty straightforward in maternal deaths and baby deaths.
If an infant is born alive, and dies, it is an infant death in the US. Even if it is completely nonviable (e.g. no heart). In the Netherlands any baby that dies in < 2 weeks after birth is not an infant death. If a child dies < 5 years it is not counted in life expectancy calculations. I'm not sure what the situation in the US is. But frankly, any time a number sounds extreme (and this one does) your first thought should frankly be "how have governments faked this number". Just like for private companies, if a number sounds too good to be true, it almost always is.
So the numbers aren't as comparable as people claim. There is obviously a difference, but it's not as dramatic as advertised.
And by "it's", I refer to the fact that this is known, not to the reason why this happens. Unemployment rates, declining middle class, bleeding working class, and people working shitty subsistence jobs that count as "full employment" would be my guess for the latter (and all that comes from eat: low self esteem, depression, drinking, drugs, poor health coverage, homelessness, et al).