I'm trying to calculate the average (mean) 2021 income of the state of Minnesota using PUMS to see if I can come close to the data in the ACS table S1901.
From S1901, I find that in 2021, 1YR ACS, there are 2,281,033 households in the state and households have a mean income of $103,033.
From microdata file PSAM_h27, I am calculating mean income (in a Power BI Measure) in two steps. First, I summarize for all records a product calculation for each record of HINCP*ADJINC*WGTP/1000000; and second, I divide that summation by the sum of WGTP of all records. In this case I get a total of 2,516,964 households with a mean income of $93,370. This does not seem close enough to the ACS table results.
However, if I filter out all records with a minimum HINCP of 1, and use the same formula as above, I get a total of 2,262,459 households and a mean income of $104,990. This seems pretty close to the Table results.
My question, is my general approach to getting mean income correct, and should I be filtering for HINCP?
(While the question above is my main one, I would also be interested in knowing how the IPUM household data can be used to calculate the Median Income.)
Hi!!
absolutely agree with your statement
Filtering for HINCP could be a good idea, as that could help you to remove any outliers that might be skewing your results. In general, your approach seems sound, but I would recommend doing some additional research to see if there are any other factors that might be affecting your results.
regards,
grace