> Using 2010 census population data, we select the urban areas from which data on prices are collected and choose the housing units within each area that are eligible for use in the shelter component of the CPI.

> Participation in the Consumer Expenditure Survey is voluntary. However, when you participate, you are representing thousands of other households like yours. In order to be able to create quality statistics, we rely on the participation of those who are randomly selected.

i.e. Volunteer census data from self-reporters.

It is the census, but one step away from outright unreliable information. This doesn't engender confidence.

▲

amanaplanacanal 6 days ago | parent [-]

Do you have a proposal for a better way to do it?

	▲	Supermancho 6 days ago \| parent [-]
		Being as objective as possible and having worked for Experian, I will say yes. Americans, in aggregate, are one of the most monitored populations in the world^1. Financial institutions are painfully aware of the spending habits of Americans. I believe credit agency data, which is paired with monitoring bank accounts via social security tie-ins (that already exist), over census data. Integrate credit companies with social programs (like food cards) and you'll have a complete picture of the lowest brackets. ^1 China does better for specific demographics, but the rural populations are not tracked very well. I could also stand to be corrected for conditions in some very small countries.