do 5-yr ACS freq wgts for a migplac1 tab(le) yield avg ann. or cumulative totals?

#1

Dataset: 2012-2016 ACS

Variable: migplac1

Issue: estimatse using freq weights yield … avg annual counts??

I’ve generated descriptive tabs for migplac1 from the 5-yr 2012-2016 ACS file

I want to be certain that the freq weighted estimates are average annual frequencies across the 5 years, not a cumulative total (I can’t think of how they couldn’t be averages, but don’t want to xxxx these counts up).

code: tab migplac1 if statefip==6 & migplac1!=6 & sex==1 & incearn>0 & educd<101 & migrate1d>10 & migrate1d!=23 & migrate1d!=20 & migrate1d!=25 & labforce==2 & school==1 [fw=perwt]

Some output:

| California In-Migrants | |
| All male workers without a 4-year College Degree |
| State or country of residence 1 year ago | Freq. | Percent |
| Total | 107,566 | |
| Texas | 6,627 | 6.16 |
| Arizona | 6,581 | 6.12 |
| Washington | 5,405 | 5.02 |
| Nevada | 4,495 | 4.18 |
| Florida | 3,772 | 3.51 |
| Oregon | 3,273 | 3.04 |

0 Likes

#2

The sampling weights available in IPUMS USA for the 5-year ACS files will calculate a representative statistic for the given five year period. So, in your case when you are calculating weighted averages with the 2012-2016 ACS data these averages will represent the average of your variable of interest for 2012-2016. If you are looking for year-specific estimates, you’ll need to use the 1-year ACS files.

0 Likes