ACS Data Users Group

Is it possible to get the unemployment rate by census tract level? I could not find it in the census reporter.

Thanks!

Robin

That's a measure you'd have to calculate yourself.
table B23025: Unemployed / table B23025: In Labor Force

Good luck.

Yikes! I see what you mean. I think I will leave it out of my project and stick with what I have. Thanks!

PolicyMap has this pre-baked, if you have access to it: https://plcy.mp/H3XG8Vw

Thanks Bernie. Unfortunately PolicyMap will not let me signup.

Hi Robin. I work at PolicyMap and would be happy to help you sign up and subscribe. Feel free to message me directly!

In the meantime, if you would like to explore the unemployment data by census tract, our latest mapping platform (which we are just in the process of releasing) will let you see the values without needing to sign in: https://www.policymap.com/newmaps#/. To find the unemployment data you can either search in the data search bar or go to the Economy Menu > Employment and Unemployment > Unemployment > Census ACS Civilians Unemployed. The shortcut link that Bernie posted above sends you to our old platform which is still accurate but lacks some of the new features.

Found it! Thank you so much!

Hello Lauren, may I ask where did you get this cbg level unemployment data? I'm trying to find the unemployment data from 2019 to 2020, but seems the newest data in your link is 2019. Thanks!

I'm looking in PolicyMap but can only see down to the city level data. Am I missing something?

Never mind, I was looking at the BLS data instead of the ACS data. My bad.

Just a heads up...
The Unemployment Rate from the ACS will be a five-year average if you're looking at tracts, or any other level smaller than mid-size cities. Even for the larger cities, counties and metros, the data is a couple of years old by the time it's published. For something as time-critical as the unemployment rate, this may be meaningless for your purpose.
For the most recent job stats (incl unemployment rate) use the BLS :LAUS, and it will be only about two months old. The downside is the smallest geography is all counties and many mid-size cities.

Thanks for the info Bert. For the work I'm doing the tract level is more important than the age of the data. We're looking at improving neighborhood outcomes and the five-year average data should be fine to help identify high priority areas.

Sounds good, Marshall. In this case, the spatial is more important than the temporal.

Robin, you can get table B23025 for all US tracts at once on the MCDC site.

https://mcdc.missouri.edu/cgi-bin/uexplore?/data/acs2019/base_tables_5yr

Use the dataset ustracts21_24 for table B23025: https://mcdc.missouri.edu/cgi-bin/broker?_PROGRAM=utils.uex2dex.sas&path=/data/acs2019/base_tables_5yr&dset=ustracts21_24&view=0

Find the variables you need, in your case it's B23025i3 (Civilian labor force) and B23025i5 (unemployed). Save as a CSV, then just use Excel to do the math.

Let me know if you need assistance with this data tool.

hello Glenn, I'm trying to find the census block groups level unemployment data for 2019 and 2020, but through the page you provided I didn't find it, and the staff of LAUS replied that they don't provide data of this level. Would you please help to see what's the problem? Thanks!

The link I provided was for tracts, not block groups. For block groups you would use https://mcdc.missouri.edu/cgi-bin/broker?_PROGRAM=utils.uex2dex.sas&path=/data/acs2019/base_tables_5yr&dset=usbgs21_24&view=0

Sorry but it seems the same link. Should I choose "BG" as the identifier?

It's not the same link. The first one goes to dataset containing tracts (ustracts). The second one goes to a block group dataset (usbgs).

Depending on which census tracts you're interested in, the City Health Dashboard has this available for download as well (the measure is called "Unemployment - annual, neighborhood-level"). www.cityhealthdashboard.com/

Just a heads up...
The Unemployment Rate from the ACS will be a five-year average if you're looking at tracts, or any other level smaller than mid-size cities. Even for the larger cities, counties and metros, the data is a couple of years old by the time it's published. For something as time-critical as the unemployment rate, this may be meaningless for your purpose.
For the most recent job stats (incl unemployment rate) use the BLS :LAUS, and it will be only about two months old. The downside is the smallest geography is all counties and many mid-size cities.

LAUS geography varies by state. In Massachusetts it includes all 351 cities and towns, as well as counties and other geographies.

Bert, do you know if/how this affects the specific occupational codes (SOCs) at the tract level? I'm using ACS data now that have been appended to the Add Health study, and the codes they provide for tract-level occupations do not sum to 1 (indeed, they sum to a mean of 2). I found that odd but perhaps it's an artifact of the 5-year nature of the data.

Good question, Adam. You may be on the right track. The inner workings of the 5-year datasets are a mystery to me.
I'd like to believe that the Census Bureau, in its infinite wisdom, does some adjustment so the 5-year average skews towards the present. Otherwise, 5-year stats for a red-hot area like Boise are going to be far from the present-day reality.
But that totally does not address your question. Sorry!