Photo of a paper copy of the 2020 US Census

Forsyth County, North Carolina   |   December, 2021

The impacts of the COVID-19 pandemic on our community and our broader society have been widespread, including specific federal sources of local data on which our community relies, like the US Census and the American Community Survey. This article takes a closer look at the nature of these disruptions, how they might affect Forsyth County, and provides some helpful background information to help us better understand these critical sources of local data.

Key Takeaways

  • Much of the data that Forsyth Futures’ produces is sourced from the American Community Survey; we are awaiting guidance from the Census Bureau on the status of the 2020 ACS 5-year estimates. It will likely be necessary to assess what data can be leveraged for 2020 informational requests and projects on a case-by-case basis.
  • The Census Bureau has decided not to release the 2020 ACS 1-year estimates because the survey gathered approximately two-thirds of its normal response volume due to the pandemic-related challenges (the lowest response rate in recorded history).
  • Discrepancies in response rates are mostly due to the structure of the instruments themselves coupled with pandemic-related challenges around data collection.

What is the American Community Survey?

The American Community Survey (ACS) helps local officials, community leaders, and businesses understand the changes taking place in their communities.[1] Many nonprofit and for-profit organizations repackage the ACS’s data for detailed population and housing information about our nation.

The ACS is a demographics survey program conducted by the U.S. Census Bureau. It regularly gathers information previously contained only in the long form of the decennial census, such as ancestry, citizenship, educational attainment, income, language proficiency, migration, disability, employment, and housing characteristics.[2] These data are used by many public-sector, private-sector, and not-for-profit stakeholders to allocate funding, track shifting demographics, plan for emergencies, and learn about local communities.[3] Sent to approximately 290,000 addresses monthly, it is the largest household survey that the Census Bureau administers.[4] The ACS gathers information annually in the 50 U.S. states, the District of Columbia, and Puerto Rico.[5]

1-year Data

1-year data reflects estimates for a given calendar year. For example, 2019 ACS estimates cover the period from January to December, 2019. The main advantages of 1-year ACS data estimates:

  • They reflect the most current data available.[6]
  • They provide estimates for specific years, which is especially useful for any significant local or national changes that could affect those estimates (e.g., the Great Recession 2007-2009).
5-year Data

The 5-year data reflects estimates for a period of 5 calendar years. For example, the 2015-2019 ACS estimates cover the period from January 2015 to December 2019.

The main advantage of using 5-year estimates is that the samples are larger; therefore, they increase the statistical reliability of the estimates produced, especially when looking at subgroups in the data.[7] In addition, 5-year estimates are the only sample that provides local Census tract-level data, and for some rural counties they are the only county-level data available.

Interpreting 5-year Data Estimates

5-year estimates are calculations (more or less averages) that are designed to show cumulative (or overall) changes over a 5-year period. If the unemployment rate is 10% from 2007-2011, then that could result from:

  • a constant percentage of 10% across those 5 years (i.e., 10% unemployment in 2007, 2008, 2009, 2010, and 2011);
  • a steady increase from, say, 7% to 13% or a corresponding decrease; or
  • any other increase and decrease across all the years.

Comparing 5-year estimates to 1-year estimates can help contextualize the patterns in the 5-year data to better understand if the estimate was relatively constant or if it fluctuated over time.

Differences Between the Census and American Community Survey Estimates

There is often confusion whether the ACS is the same as the Census, and vice versa. In short, they are not the same! So, what are the differences? Fundamentally, the Census is a total count of all individuals who live within the United States (US) along with a short selection of general questions that occurs every ten years.[8] The ACS is a national survey that relies on data provided by a sample of 290,000 households monthly within the US on a monthly basis to represent the entire population.[9] The data collected by the ACS encompasses a variety of question topics such as social, economic, housing, and demographic characteristics.[10] After receiving information from each month of an individual year, the Census Bureau combines the monthly results into a set of 1-year estimates at the national and state level, as well as for all communities with a population of 65,000 or more.[11]

Table: Comparison of Differences [12]

American Community Survey (ACS)US Census
Purposesample estimatesofficial counts
Collectsdetailed social, economic, housing, and demographic characteristicsbasic demographics
Producespopulation and housing characteristicspopulation and housing totals
New Data Everyyear10 years
Data Reflectperiod of timepoint in time

Why were the 2020 Census and ACS 1-Year Estimates Impacted Differently?

In July 2021, the Census Bureau announced that they would not be releasing the standard 2020 ACS 1-year estimates because of its low response rate.[13] However, the Census Bureau is more confident in the 2020 Census.[14] How could this be?

The 2020 Census measured the entire US population at a point in time (April 1, 2020).[15] To offset any pandemic-related disturbances with regards to the responsiveness of US residents, the Census Bureau extended the Census’ typical data collection period.[16] This extension allowed for non-response follow-ups to be scheduled at a later date so a potential high non-response rate could be mitigated.[17]

On the other hand, the ACS is a perpetual, period-in-time survey that collects data from selected participants on a rolling monthly basis.[18] Selected participants return their monthly results either through the internet, by mail, or via personal visitations throughout three-month phases.[19] The monthly data is then combined to create twelve months of sample data.[20] Due to the structure of the ACS, extending the data collection period, in particular for non-response follow-ups, like the 2020 Census was not an option once the month had passed.[21] The challenges attributable to COVID-19 on the 2020 ACS 1-year estimates prompted the lowest response rate in recorded history.[22]

What Went Wrong with the 2020 ACS 1-Year Estimates?

Response rates to the 2020 ACS 1-year estimates were hampered by the pandemic specifically through the months of March to September, 2020.[23] Widespread government-mandated quarantines restricted ACS employees from carrying out their general occupational duties[24]; the principal duties obstructed by COVID-19 include but are not limited to disruptions in:

  • mailing survey materials;
  • in-person follow-ups to non-responding households; as well as
  • in-person data collection in group quarter environments.[25]

In the end, the 2020 ACS 1-year estimates gathered approximately two-thirds of its normal response volume due to the pandemic-related challenges brought upon the three modes of collection: the internet, mailing, and in-person interviewing.[26] This meant that many segments of the national population were found to be underrepresented compared to previous years.[27] These segments often had differing social, economic, housing, and demographic characteristics than those who did respond, on aggregate, generating a significant non-response bias.[28] Non-responsiveness was regularly associated with individuals with “lower income, lower educational attainment, and those who were less likely to own their home.”[29] Census Bureau analysts attempted to adjust the non-response bias within the 2020 ACS 1-year estimates; however, despite the efforts made, the Census Bureau’s Statistical Data Quality Standards were unable to be fulfilled.[30]

What Other Data Sources Can be Used in Place of the 2020 ACS 1-Year Estimates?

The Census Bureau plans to release experimental estimates at the national and state level based on the data received from the 2020 ACS 1-year estimates on November 30, 2021.[31] The Census Bureau “urges caution in using the experimental estimates as a replacement for standard 2020 ACS 1-year estimates.”[32] In December, the Census Bureau will announce whether the 2020 ACS 5-year estimates will remain scheduled for release in March, 2022, which could be used in place of the 1-year estimates.[33] Additionally, the Current Population Survey and Survey of Income and Program Participation have been recommended by the Census Bureau to provide the data needs that would have generally been satisfied by the 1-year estimates.[34]

Elizabeth Lees, Forsyth Futures’ Director of Data and Research, further explained, “Through our Community Data and Data Request programs, as well as through collaborations with partners like The Forsyth Promise and The Asset Building Coalition of Forsyth County, Forsyth Futures works to keep a broad range of critical community information up-to-date. Much of this data is sourced from the American Community Survey, so we are keeping a close eye on the situation. Currently, we are awaiting statements from the Census Bureau on the status of the 2020 ACS 5-year estimates and plan to follow any guidance they issue. We will likely need to assess what data can be leveraged for 2020 informational requests and projects on a case-by-case basis. And we will be sharing more as we learn more.”

1. Bureau, n.d.-c.
2. Ibid.
3. Ibid.
4. Bureau, n.d.-b.
5. Bureau, n.d.-c.
6. Bureau, n.d.-e.
7. Ibid.
8. Bureau, n.d.-c; Bureau, n.d.-d.
9. Bureau, n.d.-b; Bureau, n.d.-c; Bureau, n.d.-d.
10. Cook, 2021
11. Bureau, n.d.-b.
12. Cook, 2021
13. Bureau, n.d.-d.
14. Ibid.
15. Bureau, n.d.-d; Cook, 2021.
16. Bureau, n.d.-d.
17. Cook, 2021.
18. Ibid.
19. Ibid.
20. Bureau, n.d.-b.
21. Cook, 2021; 20210729-Transcript-Webinar-Acs-1year.Pdf, n.d.
22. Bureau, n.d.-d.
23. Bureau, n.d.-b; 20210729-Transcript-Webinar-Acs-1year.Pdf, n.d.
24. 20210729-Transcript-Webinar-Acs-1year.Pdf, n.d.
25. Ibid.
26. Bureau, n.d.-b.
27. Bureau, n.d.-b; Cook, 2021.
28. Ibid.
29. Bureau, n.d.-b.
30. Bureau, n.d.-b; Bureau, n.d.-d.
31. Bureau, n.d.-a.
32. Bureau, n.d.-b.
33. Bureau, n.d.-a.
34. Ibid.

Do you have a question about this article?

Did you find this information useful?

We would love to hear from you.

Info@ForsythFutures.org

@ForsythFutures