Quote:
Originally Posted by galleyfox
I believe the ACS uses the previous census to make some calculations. So the ACS compounds any errors that were in the census.
|
The ACS does not actually adjust for what the actual Census has. I asked this question a handful of years ago to some at the Census and they confirmed it. If the Census shows that a tract was undercounted by 500 people, the ACS doesn't actually adjust to the new information. The next survey is tied to the previous one.
They're complete different studies with different methodologies so they don't mix. That would be bad science.
Census is an actual count whereas ACS is a sample survey from a small percentage of the actual population.
The issue is that you are supposed to only compare ACS to ACS. you aren't supposed to compare ACS to Dicennial Census. All the media does this and they aren't supposed to. The Census will tell you this, but they aren't assertive enough to give everyone shit for erroneously doing it.