Accessing Census Bureau Data in 2019: Transition from American FactFinder to

After continuously developing a new way of accessing data over the past few years, we are now a few months away from the retirement of American FactFinder and official launch of, the Census Bureau’s new dissemination platform.  With no new releases in American FactFinder after June 2019, this platform will be the primary way to access data from the 2018 American Community Survey, 2017 Economic Census, 2020 Census, and more!

The vision for the new platform on is based on overwhelming feedback to streamline the way you get data and digital content from the Census Bureau. Since 2016, we have been developing the site based on customer feedback while continuously releasing new improvements every few months. These updates will continue over the next few years as we are committed to giving you the functionality and capabilities you want in a dissemination system. 

Upcoming Webinar: Join us on April 9, 2019 at 1pm for a live demo of the latest release.

To prepare for this transition we encourage you to learn more about through the release notes and frequently asked questions.

  • So far this looks like a huge step backwards for data users. Tables are displayed with geographies in random order; different geography levels (e.g. state + county) aren't shown in the same table, even if they're both selected. We frequently need to obtain base table data for the nation, our state, county, city and school district; looks like this is no longer possible in one take.

    At least it's still possible to deep link into this application for specific individual geographies, if not for lists.

  • In reply to Glenn Rice:

    agree 100%, Glenn. I am not a fan of this change. I can see the benefits to casual users, but for data folks, this is clunky and difficult. I tried to simply enter a table number into the search hoping to at least get the table then narrow down by geography, to no avail.

    while we mostly use the summary file data stored on our SQL server, I often jump onto FactFinder to quickly double check numbers to ensure my query is correct at a state level before I start pulling tract and block group level geography.
  • Even the CSV downloads are in apparent random order. I mean, look at this!


    How hard would it be to sort by geoid by default?

    One change I do like is that the column names in the CSVs now match the var names in the table shells files, e.g. B05007_001E and B05007_001M instead of the current, incomprehensible, names HD01_VD01 and HD02_VD01. 

  • In reply to Glenn Rice:

    I would think the sort would be an easy fix. and yes, I did notice the variable names now match - so one good thing about it, I suppose.