Congressman Panetta Leads House Democrats in Calling on the Commerce Department to Explain Reduced Number of Census Offices

January 16, 2019
Press Release

WASHINGTON, DC – Today, Congressman Jimmy Panetta (CA-20) led 54 of his colleagues in writing Commerce Secretary Wilbur Ross expressing concern that the administration’s decision to decrease the number of Area Census Offices could lead to a potential undercount of rural, immigrant, low-income, communities of color, and traditionally hard to count communities during the 2020 Census. 

"As Members of Congress concerned by a potential undercount of rural communities, communities of color, immigrant, low-income and traditionally hard to count communities, it is critically important to our districts, and the country, that the Census secure a faithful accounting.  Failure to do so will be enormously damaging and prove profoundly consequential for decades to come," the Members wrote.

"The Bureau halved the number of Area Census Offices, reducing local presence and increasing reliance on automation, which could cause an inaccurate count of population.  An undercount of these communities would have disproportionate impact on the federal share of resources rural and minority communities receive to pay for critical services such as education, healthcare, nutrition assistance, hospitals, and child care," said Congressman Panetta about leading this letter.  

The Members of Congress are requesting clarification on the process by which the Census Bureau determined the number and locations of the 248 planned Area Census Offices.  In September 2017, the Bureau slashed the number of Regional and Area Census Offices from 12 and 435 in 2010 to 6 and 248 planned for 2020, leaving several congressional districts without a physical Census Bureau presence.  This decision will place thousands of critical enumerators far from many of the areas they are responsible to count, seriously jeopardizing the accuracy and completeness of the decennial census.  

To compensate for the fewer number of census offices, the Bureau will instead rely heavily on technology and internet-based enumeration.  This strategy lies in stark contrast to the Census Bureau’s own estimates, which predict lower self-response rates compared to 2010 and widespread mistrust and/or cybersecurity concerns.   A lower self-response rate will have a disproportionate effect on rural and minority communities, and potentially result in a historic undercount.  The consequences of an incomplete decennial census will be enormously damaging and prove profoundly consequential for a decade, if not longer.

The full text of the letter can be found below.


January 16, 2018

The Honorable Wilbur L. Ross, Jr.
U.S. Department of Commerce
1401 Constitution Avenue NW
Washington, D.C. 20230

Dear Secretary Ross:

We write to express our significant concern regarding the number of Area Census Offices (ACOs) currently planned to administer the 2020 Decennial Census and the resources available to ensure a complete and accurate count. As Members of Congress concerned by a potential undercount of rural, communities of color, immigrant, low-income and traditionally hard to count communities, it is critically important to our districts, and the country, that the Census secure a faithful accounting. Failure to do so will be enormously damaging and prove profoundly consequential for decades to come. 

The Decennial Census is a cornerstone of American democracy. The data influences congressional representation; Electoral College vote distribution; state, local, and federal district map delineation; and billions of dollars in resource allocation.

During the 2010 Decennial Census, the Census Bureau (Bureau) established 12 Regional Census Offices (RCOs) and over 500 ACOs throughout the country, ensuring enumerator proximity to community residents. This broad distribution of RCOs, ACOs, and enumerators provided the necessary program staff, local communication efforts, and field operations necessary to develop sufficient trust in the Bureau and maximize response rates.

However, in September 2017, the Bureau halved the number of RCOs and ACOs to six and 248, respectively. To compensate, the Bureau suggests its field operations will instead “use technology to efficiently and effectively manage the 2020 Census fieldwork, and as a result, reduce the staffing, infrastructure, and brick and mortar footprint required for the 2020 Census.”[1]

According to the October 2017 Department of Commerce Lifecycle Cost Estimate for the 2020 Decennial Census Program, moving away from a robust local presence will pose two major risks. First, the report stipulates that “[the] self-response rates decline below assumed and modeled levels, increasing the Non-Response Follow Up (NRFU) caseload”, which the Bureau estimates will decrease the self-response rates from 60.5% to 58.5% and 55.5%.” Second, the report highlights “widespread mistrust and/or cybersecurity concerns [will] decrease the public’s willingness to respond online.” [2]

We worry that a heavy reliance on internet automation, compounded by the decision to dramatically reduce the Bureau’s field presence, will disproportionately underrepresent rural, low-income, and minority communities. According to the 2015 American Community Survey, 21% of rural residents lack internet service at home, compared to only 13% of urban residents. Moreover, the same study found that 40% of low-income minority communities in the rural South and Southwest lack internet access at home, and for many other minority communities, internet access is completely unavailable. [3]

A Bureau study of the 2010 Census found a net undercount of 2.1% of African Americans and 1.5% of Hispanic Americans, totaling 1.5 million residents. This contrasts starkly with the overcount of 0.6% of homeowners and 0.8% of non-Hispanic whites.[4] In addition, 5.3% of Asian Americans, totaling almost 900,000 residents, who should have been counted were not, as opposed to 3.8% for non-Hispanic Whites.[5] Further to the detriment of minority communities, the Bureau has reversed previous policy and will no longer hire non-citizen enumerators to ensure optimal access to non-English-speaking households. This is particularly concerning because, matched with the reduction of ACOs, it denotes that there will not be adequate on-the-ground resources to fully count communities who already struggle to be counted.

While the Bureau does not produce undercount rates for rural areas, its calculation of 2010 Census mail return rates show a disproportionately low mail return rates in rural areas. In the 2010 Decennial Census, of the 316 counties with mail-in returns of less than 72.7%, 251 were in rural areas.[6]

Given lower projected self-response rates, reduced local presence, and increased reliance on automation, we ask that you respond to the following questions for the purpose of clarifying the originations, intentions, and justifications related to the number and location of ACOs:

  1. What formula did the Bureau rely upon to determine ACO locations?
  2. Did the Bureau reassess ACO allocation after the cost estimate projected a self-response rate decrease?
  3. Does the Bureau plan to provide additional field resources, through additional ACOs or questionnaire assistance centers, to increase response rates in rural areas, communities with decreased internet access and use, and traditionally hard to count communities?
  4. What is the plan to increase response rates in rural, minority, and low-income communities that have disproportionately lower rates of internet access and use?

We ask that you provide responses to these questions within 30 days of receipt of this correspondence. Rural and minority communities receive a disproportionate share of federal resources to pay for education, healthcare, nutrition assistance, hospitals, and child care. Communities currently suffering economic anxiety cannot afford to lose federal resourcing due to a lack of Census accuracy.