Missing tract-level MOE estimates for ACS 2023 5Y

[jvtcl]

Observation

https://data.census.gov/table/ACSDT5Y2023.B01001?q=b01001&g=1400000US20081463100_g=010XX00US

• Seems to affect just the MOE estimate. Tracts have valid count estimate values.
• Seems to affect total population (B01001001) and Hispanic/Latino origin (B03003xxx). Not clear on whether this affects other variables as well.
• Block groups do have estimates in this case [1, 2, 3, 4], but it's not clear whether the problem can be present at this scale as well.

Possible solutions

• In acs.puma() using block group level aggregate standard errors in place of those derived from ACS (MOEs --> SEs)
• In acs.puma() using Variance Replicate Estimates to approximate MOEs/SEs
• Substituting in a placeholder value