merging data according to geographic level.

Hi Joe,

Thanks, I got your reply.

Then I think I may adjust my analysis on more aggregate level. I am considering doing analysis based on geographic reagion.Can you provide me some information on how to merge data on geographic reagion? Say, county level, city level, MSA level, or urban area level? My target variable is demographic characteristics of unemployment population and change of income inequalities.

Thanks much!



How to aggregate information up to a geographic unit is largely dictated by what statistical program you are using, taking into consideration the type of summaries you would want on a variable to variable basis (e.g. mean vs. median vs sum vs counts). However, I would actually recommend checking out the NHGIS project (also housed here at the Minnesota Population Center). NHGIS offers aggregate data for many geographic levels, using Census Bureau published summary tables. This way you can get complete coverage using mostly consistent geographic units, such as counties (the public use microdata does not allow for the identification of all counties due to confidentiality restrictions), and consistent aggregate measurements for modern samples. The NHGIS project is primarily intended to facilitate mapping data, but the CSV-format tables provided can be used independently of any GIS software.

I hope this helps.