----------------------------Original message----------------------------
Art L. is correct, there is little correlation between ZIP Codes
and population density.  You can however, develop reasonable
estimates by using polygon overlay with Block Group polygons and
BG population with ZIP Code polygons.  Since a ZIP+4 commonly
represents only one side of one city block, ZIP+4 polygons
aren't feasible.  ZIP+2 polygons are possible, but they will be
oddly shaped and are inappropriate for most applications.  Why
not simply go with Block Groups and be done with it?
 
GO>----------------------------Original message----------------------------
  >>I know that zip code boundary files are not in the public domain, but
  >>they just keep asking...  Has anyone ever seen the area (as in sq. miles)
  >>of zip codes?  Five digit would be fine now, but you know he'll be back for
  >>+4.  This is for a population density question and the user has been advise
  >>that zips are not an appropriate statistical area...
---
 * OLX 2.2 TD *
 ~ GISnet BBS  303-447-0927  Boulder, Colorado