state median income

In the Build Back Better federal legislation there appears to be a shift to program/service eligibility toward percentages of State Median Income (SMI), versus the currently more common federal poverty level (FPL). This makes some sense because FPL is a broad brush for all states, while SMI is more calibrated by the economic conditions for each state. For example, a family could be eligible for child care assistance if their income is less than 250% of the state median income.

Am I correct that there is no clear way to compile SMI by age and SMI threshold at the county level using ACS tables and IPUMS extracts. If so, will the ACS be creating these breakdowns since these kind of measures will be potentially used everywhere in the U.S. for a wide array of programs?

By way of example, we'd like to create a table that shows:
number of children under 6 yrs of age in families <75% SMI for Ford County, IL
(etc, for all counties, and additional thresholds of SMI)

We found a way to do this via IPUMS using FAMSIZE, AGE, and FTOINC variables, but the geographic level is PUMA, which of course does not line up with counties. ACS tables S1903 and S2503 could be helpful, but do not provide income and age breakdown in a way that would be useful.

(This is all assuming that the BBB legislation passes in anything close to its current form. But I wonder if in general there will be a shift toward SMI away from FPL anyway.)

  • I am looking for similar information. I am trying to find data on residents within a specific county who are elderly and have an income less than 60% AMI. Did you receive any assistance?

  • The CHAS data supplied by HUD might be of use to you:

    This is a special recompilation of ACS data by proportion of AMI.

  • I never made progress on my question above.

    If the county you are interested in consists of 1 or more PUMAs, then you can do this via IPUMS. The CHAS tool gets you part way there, but does not indicate any age ranges. I would find the specific AMI value you need and then create an IPUMS extract for the PUMA(s) you need with the variables PUMA / age / total family income. Then sort and filter by income and age to find what you need.

    Keep in mind there could be low sample size creating accuracy problems.

    Again, someone else should correct me if I'm wrong here! There may simply be a sampling limitation in ACS data to creating age ranges for state median income (and similar measures) at the county geography.