2016 ACS Poverty

USA
1

I am trying to calculate the national poverty rate using the 2016 1-year sample ACS, but my numbers are not lining up with what is published in the American Fact Finder here: (https://factfinder.census.gov/faces/t…). I understand that the estimates will not line up, due to differences in defining poverty at the household vs. family levels. However, I would expect the weighted Ns to line up between the two estimates. I am calculating a weighted N of 315,754,035 cases with poverty data while Fact Finder has 315,165,470. Am I wrong in thinking that those two numbers should line up?

Here is my code in Stata:

gen povlevel = .

replace povlevel = . if poverty==0

replace povlevel = 1 if poverty < 100 & poverty !=.

else replace povlevel = 2 if poverty <200 & poverty >= 100 & poverty !=.

else replace povlevel = 3 if poverty >= 200 & poverty !=.

svyset cluster [pweight=perwt], strata(strata)

svy: tab povlevel

Thanks in advance!

Renee

2

As you have noted, estimates are expected to differ due to differences in defining poverty at the household vs. family levels. Additionally, statistics reported in American Fact Finder use full-count census data, whereas IPUMS USA provides a sample of the data (unless you are using full-count samples from 1850-1940), therefore, you can expect estimates as well as a weighted N to be slightly different between the two.

3

Please note that I have since realized that this code does not treat missing data properly. Values of 0 are not actually recoded to missing here, as they should be.