I want to calculate the percent change for median household income in Los Angeles at the block group level from 2000 to 2010. Someone told me I have to reaggregate 2000 block data to 2010 block group geography, do a spatial join based on centroid, and that I have to download "crosswalk" or relationship data from the U.S. Census comparing 2000 to 2010. I downloaded the relationship .txt file with 2000 California Census tabulation data and the 2010 tabulation data. Right now I have the unjoined data files. I don't know what to do next. What do I join together?

1) Can someone give me step-by-step instructions on how to calculate the percent change at the block group level? :/

2) If calculating it at the block group level is too complicated, I may just change my level of analysis to census tract level. How do I calculate the percent change at the census tract level?

1 Answer 1


If you want to use Census tracts the good people at Brown University have already done the hard work for you:

Brown University Longitudinal Tract Database

This resource contains tract-level variables from 1970-2000 interpolated to 2010 boundaries, facilitating longitudinal analysis.

  • This is great! How do I use the crosswalk data?
    – KTA
    Commented May 15, 2012 at 19:32
  • 1
    From what it sounds like you don't even need to use the crosswalk files. The researchers at the US2010 project have tabulated a whole host of variables for you from the various Censuses, including median household income (also stratified by race/ethnicity). Just go in and download the files and everything will be there for you; they can then be easily joined to your 2010 tract feature class ('trtid10' or 'tractid' corresponds to 'GEOID10' in the Census TIGER/Line files).
    – qgeog
    Commented May 16, 2012 at 13:02
  • The crosswalk files are intended for users who have other historical tract-level variables, not available in the pre-made Brown files, that need to be interpolated to 2010 boundaries. They provide instructions to do this in either Access or Stata on their site.
    – qgeog
    Commented May 16, 2012 at 13:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.