Is there a way to get this level of granularity for census data?

Questions & Help
1

Hi everyone, is there a way to get this level of granularity for census data?

| census_block_group | age   | sex  | race / ethnicity | income             | education              | count (estimate) |
+--------------------+-------+------+------------------+--------------------+------------------------+------------------+
| 10010201001        | 22-24 | Male | White alone      | $45,000 To $49,999 | No Schooling Completed | ###              |
+--------------------+-------+------+------------------+--------------------+------------------------+------------------+
| 10010201001        | 25-29 | Male | White alone      | $45,000 To $49,999 | No Schooling Completed | ###              |
+--------------------+-------+------+------------------+--------------------+------------------------+------------------+
| ...                | ...   | ...  | ...              | ...                | ...                    | ...              |
+--------------------+-------+------+------------------+--------------------+------------------------+------------------+```
2

Hey @Thomas_Young_Econometric_Studios_Utah_Legislature thanks for reaching out! Yea, I am looking for CBG counts broken down by age, sex, race, income, and education. I could try to calculate it from separate parts but then I would have a bunch of NaN values. I just looked at row 8 did not realize that converting to zip codes would be non trivial. Did you have a solution in mind?

3

@Steven_Du_Cato_Institute I don’t - yet. I’ll look into how I might do it and let you know. I’m interested in the same set up.

4

@Thomas_Young_Econometric_Studios_Utah_Legislature I found this on the census website https://www.census.gov/cps/data/cpstablecreator.html

5

and this https://data.census.gov/mdat/#/

6

These tools allow you to do limited custom queries of the census. But you have to manually select all the ‘public use microdata area’, which is the smallest geographic division available (no zip or CBGs). There should be an api that will make this easier which I am looking into right now.

7

@Steven_Du_Cato_Institute Just getting back to looking at this. Did you make any progress? Wondering if there’s anything I can do to help.

8

IPUMS USA provides 5% of random sample of US population (1% for each year between 2014-2018). It provides household and person level details of these samples. However, the smallest spatial unit is PUMA (Public Use Microdata Areas), which is about 100k populations each (in census 2010).