The decennial census enumerates each person in the United States. The results of the census apportion seats in the House of Representatives, guide legislative redistricting, and form the basis for allocating half-a-trillion dollars in federal assistance annually. The census also provides a social, demographic, and economic profile of the country’s residents, informing decisions by policymakers and businesses across the state. Census results also are used to determine legislative districts for federal, state, and local representatives. The census is the responsibility of the federal government and is conducted by the Census Bureau, a division of the U.S. Department of Commerce. Census Day—for the purposes of the count—is April 1.
Significant Changes to Count Operations. The 2010 Census cost over $12 billion over the life cycle of the enumeration (which includes the preparation for and aftermath of the count). The Census Bureau estimated that completing the 2020 Census in a similar manner as 2010 would cost over $17 billion. To keep costs closer to the costs of the 2010 Census, the Census Bureau is making a number of significant changes to count operations, including:
Internet Response. Historically, individuals primarily responded to the census via a mailed questionnaire. In 2020, only a portion of households will receive a mailed census. Most individuals will be encouraged to respond online. The Census Bureau aims to have over 50 percent of respondents respond via the internet. (Individuals also will have the option of responding by phone through the Census Questionnaire Assistance Center.)
Address List Update. Typically, the Census Bureau relied heavily on field workers to update its national address list. (The national address list is used to mail census forms and follow up with nonrespondents.) For the 2020 Census, the Census Bureau will rely primarily on local government data, satellite imagery, and other administrative records to update its address list.
Reduced Follow Up. The Census Bureau expects to hire 50 percent fewer fieldworkers for nonresponse follow up in 2020. Consequently, field operations will be substantially reduced in 2020. In previous years, census field workers visited households that had not yet responded up to six times to complete their surveys. For 2020, the Census Bureau is committing to visiting nonrespondents at least once. Given the reduced field presence, the Census Bureau will rely heavily on administrative records—like those from the Internal Revenue Service—to complete the count.
Current Funding Below Estimated Costs. Funding for the 2020 Census has been below estimated costs for a number of years. For the current year, the latest federal budget agreement provided $180 million less than was requested by the Census Bureau. These reductions have had a number of impacts to census preparations, including notable impacts to field testing:
2017 Field Test Canceled. The Census Bureau planned to test a number of features of the online response system in 2017, but canceled these tests due to budgetary uncertainty.
Significantly Reduced Census “Dress Rehearsal.” Typically, the Census Bureau does a comprehensive test of census operations—known as an end-to-end test—in three areas of the county. The different locations chosen reflect the differing response challenges throughout the country. This year, this rehearsal will only occur in Providence, Rhode Island.
Changes to Residency Determination. Historically, the Census Bureau counted people where they lived, generally defined by where they sleep most of the time on census day. The 2020 Census will change this practice for some residence situations. Generally, these changes attempt to capture more people where they live day-to-day. One of the largest changes is for deployed civilian and military personnel. For the 2020 Census, individuals on deployment will be counted at their domestic address, typically the installation from which they deployed. (This change, however, will not apply to personnel stationed overseas. Individuals stationed overseas typically live at that location for at least a year, whereas deployments are shorter in duration.) Previously, deployed personnel were counted using administrative records. Often, these administrative records used a different permanent address (typically, their last address before joining the military). These changes will not affect college students or individuals in prison, who will continue to be counted at their college residence or jail/prison facility.
Possible Changes to Census Questions. Census forms historically asked two separate questions about race and ethnicity. Increasingly, however, many respondents have not identified with any particular race and instead selected “Some Other Race.” This raised concerns about the accuracy of the data collected. While a proposal for combining the two questions (and changing the possible categories) was proposed previously, the Census Bureau recently announced that no changes will be made to these questions. In addition, the U.S. Department of Justice requested that a question on citizenship be added to the census. No decision on this issue has been announced. The final census form is due to Congress at the end of March 2018.
California Has a Large “Hard to Count” Population. Historically, the census has undercounted particular groups of people including young children, the elderly, low-income individuals, minorities, renters, foreign-born individuals, and individuals living in crowded households. As a majority-minority state, most residents of California fit into at least one of these categories. In addition to over half of residents being nonwhite, over a quarter of residents are foreign born, close to half live in rental housing, and 14 percent have incomes at or below the poverty line.
California Was “Undercounted” in 1990 Census. California was undercounted by 2.7 percent in the 1990 Census—the fourth highest undercount percentage in the country. This cost California roughly $2 billion in federal funds over ten years. The undercount also likely cost California a seat in the House of Representatives. (See our 1999 report for more on the 1990 undercount in California.)
Changes to 2020 Census May Affect California Count. Moving to a primarily internet-based census is a significant change from prior practice. Only a portion of households will receive a paper census; the rest will receive instructions by mail for how to respond online (or by phone). Concerns have been raised about individuals’ willingness to respond via the internet given concerns about information security. These concerns—in combination with the potential for a question about citizenship—raise the possibility of an undercount in California in 2020. (By law, information collected in the census about immigration status cannot be passed on to immigration enforcement agencies.)
Governor Proposes Providing $40 Million Over Three Years for Census Outreach. The administration proposes providing $40 million and 22 limited-term positions to support the Complete Count Committee. The Complete Count Committee was first formed after the 1990 undercount. The committee is appointed by the Governor and coordinates the state’s outreach efforts, focusing on the hard-to-count population. The funding proposed for 2018‑19 would support the activities of the committee through the 2020 Census. Almost three-quarters of the funds would be dedicated to a media campaign ($17 million) and working with local community based organizations ($12.5 million). Community organizations would conduct most of the direct outreach to individuals to encourage them to complete the census.
Funding Previously Provided for 2020 Census Preparation. The 2017‑18 Budget Act provided up to $10 million for initial census preparation activities. Of that amount, $7 million was provided for grants to local governments for participating in the Census Bureau’s Local Update of Census Address (LUCA) program. (As noted previously, the Census Bureau is relying heavily on administrative data to update its national address list.) The Department of Finance also received authority to spend up to $3 million on initial outreach activities for the 2020 Census. These funds are being used to support initial activities of the Complete Count Committee.
Prior Complete Count Committees. The state provided $24.7 million for the 2000 Complete Count Committee in response to the 1990 undercount. For 2010, however, the state only provided $2 million. (The Complete Count Committee raised roughly $10 million in private funding to augment its efforts for the 2010 Census.)
Proposal Reflects Significant Increase in State Support Compared to 2010. The Governor’s proposal would bring total state funding for census-related activities to $50 million across 2017‑18 and 2018‑19. Due to the significant changes to the census, providing state funding to target hard-to-count populations is reasonable. The Legislature will want to ensure community based organizations can effectively reach out to hard-to-count populations given the notable reduction in fieldworkers for nonresponse follow up. In addition, the Legislature will want to monitor the media campaign costs. The 2020 Census will be taking place in a presidential election year when advertising can be particularly expensive. Census day, however, will occur after the California primaries (which move to March in 2020). Consequently, media costs may not be as high in the weeks leading up to the census as they will be earlier in the year.
Undercount Could Have Serious Consequences for California. The decennial census is one of the main factors that underlie how hundreds of billions of dollars of federal assistance are distributed. For instance, the census count is used to determine states’ Federal Medical Assistance Percentage (FMAP) for Medicaid, known in California as Medi-Cal, which is based on per-capita income. A lower per-capita income can result in a higher FMAP and more federal funds per Medi-Cal participant. The census is used to determine each state’s per-capita income. This year, California expects to receive over $60 billion in federal assistance for the Medi-Cal program. Other major federal assistance programs that use census data include highway funding, Section 8 housing vouchers, and special education grants.
An undercount in California could result in losing a seat in the House of Representatives. These seats are apportioned based on states’ relative populations. California’s population growth has been slower than some other states. By not growing as quickly as other states, the relative proportion of people living in California—compared to other states—shrunk. Based on a recent study by Election Data Services (a research company), were California to have a greater-than-average undercount of roughly 76,000 people it could result in California losing a seat. As we outlined in recent blog posts—here and here—part of the reason California’s population growth has been slower is the net loss of some residents to other states.
Recommend Funding Census Outreach Now. California is the first state to set aside funds for census outreach. Given the major changes to the upcoming census—and the potential impacts to state funding—preparing for a significant outreach campaign can be in the state’s fiscal interest.