ACS Data Users Group

So....trying to respond to the new world order that has been imposed on to us.In an effort to possibly aid our grant writing staff, I was looking into trying to address what appears to be some new guidance within the Department of Transportation where preference may be given to projects in locations where the birth rate & marriage rates are higher than the national average. I thought ACS table DP-02 might provide all the elements that I might need for this but now I'm not so sure and my Googling has let me down. What I had envisioned doing was having this information at the census tract level and then being able to present this back to users in map form where I have the rates for the census tract, but also for our county, the state, and national average.

Thoughts?..

Steve

Dear Steve,

You probably will not be able to get the data you need from standard (B S D C) ACS tables at the tract level. One way to deal with the situation is to use PUMS data. With the PUMS data you will the age of children in single year increments so you should be able to determine some trends in birth rates for a geography. For marriages here are the ACS questions https://www.census.gov/acs/www/about/why-we-ask-each-question/marital/ There are the following PUMS variables MARHM (married last 12 months) Times married MARHT and MARHYP year last married.

You can get an estimate for births in the past from the age of the children in the household. So the data is on the ACS questionnaire. The problem is that the data is only available for PUMAs (population >100,000). There are also question on migration. your current address and Where did you live 1 year ago see https://www.census.gov/data/developers/data-sets/acs-migration-flows.html for migration questions and tables. PUMS variable MIG and MIGPUMA. Read up on the PEP (Population Estimated Program) so see how births/deaths/migration are factored in to get the current (intercensal) estimates.

If you want estimates for census tracts, I have a program that will do that. Contact me at info@dorerfoundation.org. If you are from a 501(c)(3) or government entity consults are free.

Best,

Dave Dorer

Hi Steve - My read is that the new DOT guidance isn't yet sufficiently clear on what data are required. Official birth rates and marriage rates are typically calculated based on administrative data, not survey. And ACS data on "birth in past year" often does not align with administrative records of where births are known to have occurred.

Since this is getting a bit beyond ACS data, please feel free to reach out to me directly bjarosz (at) prb (dot) org. I'd be happy to chat.

To follow up on Beth's comments. The birth (and death) datasets and codebooks are here https://www.cdc.gov/nchs/data_access/vitalstatsonline.htm The geography is county FIPS.

I agree with Beth that the new guidance on funding proposals should be considered very preliminary and as with many initiatives coming from the Federal government it will need adjustment and clarification.

I am not aware of any publicly available birth data for all US states beyond the National Center for Health Statistics public use files. Unlike Massachusetts, which has open birth, death and marriage records (our Puritan Massachusetts Bay ancestors wrote everything down for all to see) many states have "closed" records, such as New York and NH. To access those records requires special clearance. The only restricted dataset in this area that I have worked with that has address data is the National Death Index where we needed to clear through 3 review boards to get access to data from all states.

The ACS PUMS data, while coming from a survey, has some advantages. It has many covariates not available in administrative public use datasets. This data would be needed in assembling a research plan for a project proposal or grant application. In the case of transportation studies, ACS data including information about commuting is often used in transportation models.

If you look at: https://www2.census.gov/programs-surveys/popest/technical-documentation/methodology/2020-2024/methods-statement-v2024.pdf you can see what administrative data is used for the Population Estimates Program (basically IRS data, birth and death records, medicare and social security data).

The ACS data is adjusted using the PEP estimates so administrative data does play a role in the ACS microdata and tables.

See my earlier post for an email address.

Best,

Dave

Thanks to both of you for the replies. I completely agree that i'm jumping the gun here a bit since the guidance have been leaked internal memos and not external directions to the state agencies which allocate those funds. My direct role is as a GIS Analyst and not the grants specialist applying for the grants so I do not have a handle on what directions they receive. I do know that for some recent grant applications that I assisted them with, we looked at the various census tracts that intersected with the projects we were seeking funding for and not just at the county level for our county.

I will say that I do think at least a crude birth rate may be possible out of DP-02 since there is one item in that table which has the rate per 1,000 women aged 15-50 that have given birth in the last 12 months. That rate would be a 5 year rate if taken from the 5-year ACS data product but that could be divided by 5 to get a per year rate.

Anyways, I'm not completely sold on what I was thinking but I thought I should ask those who would know better than I would since these are variables I haven't had to look at with the Title VI work I have had to do so far.