Extra reading
Serving donuts

A second lease on life

I hinted at it in the last post, and some readers also made similar suggestions.  What happens if we plot the U.S. life expectancy data in relative terms (indiced) rather than in absolute terms?

The result is highly revealing, and that is why we should always look at the data many ways.  While in the original chart, the differences in the race/gender segments were essentially obscured by the overall slowly-growing trend, in our new chart, we took out the trend, isolating the growth rates.


Redo_lifeexpect

The reconstructed chart showed that:

  • Between 1970 and roughly 1990, blacks of both genders gained in life expectancy at a rate higher than the national average, while white females lagged behind white males
  • However, almost all of the gain by blacks were attained between 1970 and 1984, and in the 10-15 years following, this excess gain was wiped out so that by 1992 or so, the black male, black female and white male lines again converged.
  • Starting in 1995, black males again achieved significant improvement in life expectancy.  This time, black females did not follow their male counterparts.  Meanwhile, white females continue to lag behind.

Not being a health care specialist, I can't say what happened to the cohorts of the 1970s, the 1980s and 1995.  One thing is for sure: these insights are hard to glean from the original.

Ap_lifeexpect 


Reference: "CDC says life expectancy in the US is up, deaths not", Miami Herald, Aug 19 2009.  CDC Life expectancy data.

Comments

Cris

But looking at data this way can be deceiving also: that the "white female lagged behind" is because they already had a much higher life expectancy, which makes it harder to make improvements. The same is true for the larger increase in the black male cohort: they're just catching up with the average.

As you say, looking at the data in many ways is the trick!

Matthew

I have to say I prefer the original as I think its more the point of the chart to show the absolute life expectancy, not its rate of increase among sub groups.

What happens if you chart the increase not in % but in years from a common base?

anon

It might be worth making the lines in your graph different colors. I find it a little hard to keep track as it is.

Also, I'm not sure % increase is better than difference in actual number of years.

Kaiser

anon and Matthew: I agree that plotting the age is better than plotting the percent change. However, don't expect the conclusion to change one bit because all the life expectancies are in the 60-80 range and the average percent changes is 10% so we are talking about a total change of 6-8 years (it's more like 5 for white females and 9.5 for black males), depending on which segment.

James Pearce

Actually I prefer this chart.

Charts should tell a story. And if they also provoke further routes of questioning, all the better.

The story from the original chart was "life expectancy increases, but differently per demographic"

The story from the new chart is "life expectancy increases, but differently per demographic...

... AND OMG!!! THERE'S SOMETHING REALLY INTERESTING HAPPENING TO BLACK MALES!!!"

Anyway, you haven't learnt anything less with this second representation - but yet an exciting new area of research has opened up.

Mission accomplished.

The comments to this entry are closed.