The Changing Age Distribution of the U.S. Population 1950-2050

I owe, or blame, my sudden interest in demographic issues on Edward Hugh of Bonobo Land.  While at the Census web site yesterday, I came across census data on the age distribution of the U.S. population from 1950 through 2000 and projected through 2050 so I took a moment to look at it graphically.  I decided to post it so I would have the data and graph readily available.  If you are interested in the diagram, note that as you age you move to the northwest.  The baby boom shows up clearly in the diagram.  That the age bands are lower on the right-hand side of the diagram than on the left-hand side is indicative of the increasing average age of the population.  One big change is the increasing size of the 80+ band over time and more generally the 70+ range shown by the top three bands.  The compression in the distribution below this range is evident:

This brings up issues I want to think about further.  For example, as I assume others have noticed, with a larger and larger fraction of the population moving into the asset liquidation phase of their life-cycle, how is the saving rate affected?  How much of the change in the saving rate in recent years is attributable solely to changes in the age-distribution of the population?  What other changes in economic behavior might this bring about?

UPDATE:  Some of you asked about population pyramids, though as shown below by the projected 2050 population distribution it no longer looks much like a pyramid.  If you click through to the census site listed above there is a dynamic version of the population pyramids showing the changes from 1950-2050.  Here are static versions spaced 25 years apart available at the Census site: