I think that it might explain part of the divergence. There are a couple of obvious reasons that the U.S. might trail the OECD -- the greater number of deaths due to higher car accident rates and a higher murder rate, as well as a well known disparity in the life expectancies of Americans of primarily European descent and those of people of primarily African descent.

I'm still trying to figure out how to say this. I'm wondering if the "natural" life expectancy in the U.S. has always trailed the "natural" life expectancy of the rest of the OECD (judged collectively, anyway) but because of some systemic effects in Europe this natural disparity was offset for some decades. The effect might have been something that gradually faded, exposing the natural life expectancy differential.

I don't want to overstate things here -- there are some clear candidates for things that might have gone wrong with health in the United States in the last couple of decades that we want to look at. But I suspect this is not the only thing we're seeing in this data.

This sounds plausible, though there might be many factors playing into this. Most of Eastern Europe has probably experienced a massive increase in wealth and life expectancy after the Cold War, perfectly lining up with the graph diverging starting at 1990. If you travel around Poland or the Baltics these days, the difference is remarkable. On the other hand, Russia has experienced a massive decrease in life expectancy at the same time.

And stats like these here suggest that Western European states that presumably changed less after the wall came down drifted away from the US, too: http://www.indexmundi.com/g/g.aspx?v=30&c=fr&c=gm&c=uk&c=us&...

Couldn't find graph-able stats that'd go back to the 90s or are totaled by Eastern Europe, sadly, but poking around here suggests that several Eastern European states experienced a significant increase in life expectancy between 1990 and 2000, three years for e.g. Slovakia.

I had completely forgotten about Eastern Europe and the end of the Cold War.

The excess murder rate is tiny. The highest murder rate in the U.S. in the last 50 years was 10 per 100,000 (in 1980). Murder rates in Europe tend to be 1 or 2 per 100,000.

The median age in the U.S. is close enough to 40, so every year, murder impacts life expectancy by something like (8/100000) * 40.

(The error from assuming the highest murder rate should pretty much swamp any other inaccuracy in there)

It would explain the emergence of a gap caused by other factors that harm USA but previously canceled out by the factors noted that harmed e.g. Europe.