Clarity on Tract Crosswalk Calculations

NHGIS
1

I’m trying to calculate the net difference in population from a 2010 tract to a 2020 tract. I’m aware of all the difficulties with accuracy, but I want to be positive I’m following the methodology of the crosswalks correctly. I’m using a 2010 Blockgroup to 2020 Tract crosswalk file, sample image uploaded.

  1. Iterate through the block-groups in column 1 and multiply their 2010 population data by the weight.
  2. Then for each 2020 census tract in column 3, sum up the 2010 product values made in Step 1 if it has the target 2020 census tract.
  3. This summation from Step 2 represents a weighted estimate of 2010’s population relative to comparable boundary of the 2020 Census tract.
  4. To calculate the change from 2010 to 2020, you simply subtract 2020 Census Tract’s data from the data compiled in step 3.

Is this correct?
Below is a screenshot of the crosswalk data example.

Screen Shot 2023-06-29 at 12.38.27 PM
Screen Shot 2023-06-29 at 12.38.27 PM1004×352 101 KB

2

Yes, what you describe sounds exactly right, with one minor wording issue in your Step 4…

You say “subtract 2020 Census Tract’s data from the data compiled in step 3”. I’d flip-flop these… To calculate change, subtract the 2010 estimates for 2020 tracts (from step 3) “from” 2020 data for 2020 tracts, not the other way around. I.e., change = 2020 data - 2010 estimates. But that may be what you meant anyway!