The 2017-18 Budget

Local Update of Census Address Program Overview

Decennial Census. Since 1790, the United States has conducted a census every ten years. Census results are used for many purposes, including reapportioning seats in the House of Representatives, realigning congressional and state legislative districts, and distributing roughly $400 billion in federal funds to the states. The next U.S. Census will be conducted in 2020.

Information Collected in the Decennial Census. The U.S. Census counts every person who lives in the United States (including territories like Puerto Rico) whether he or she is a citizen, legal resident, or undocumented immigrant. Traditionally, each household receives a questionnaire in the mail that asks a few basic questions including the number of people in the household, their names, their ages, their race/ethnicities, and whether or not the occupants of the household own or rent their home. If households do not return their census forms, the Census Bureau follows up with door-to-door household visits. In the 2010 Census, 72 percent of households returned their questionnaires by mail. Other information collected by the Census Bureau—like socioeconomic and housing characteristics—used to be collected as part of each census, but now are collected through the Census Bureau’s American Community Survey exclusively.

Local Update of Census Address Program (LUCA). For the purposes of mailing census questionnaires, the Census Bureau maintains a national residential address list. The Census Bureau takes many steps to keep this list up to date, including incorporating new information from the U.S. Postal Service and using satellite imagery and other geographic information systems to identify areas of development. The Census Bureau also works with local governments to verify residential addresses. Through LUCA, local governments can review Census Bureau information on residential addresses for accuracy. Participating in LUCA is not required, but the federal government encourages local governments to participate.

2010 Census Count Determined Accurate. Since 1980, the Census Bureau has attempted to measure the accuracy of the decennial census. The process currently is referred to as the Census Coverage Measurement. To reach this measurement, the Census Bureau estimates what proportion of the count was enumerated correctly, what proportion was enumerated incorrectly (double counted, for instance), and the extent to which individuals were omitted from the count. In 2012, the Census Bureau estimated that the 2010 Census undercounted the overall population by only 0.01 percent. For California, the undercount was just over one-quarter of 1 percent (0.26 percent). This net undercount percentage reflects the offsetting effects of both over counts and omissions.

Probable Changes to 2020 Census. The costs of conducting the decennial census have increased substantially over the decades with the 2010 Census costing over $12 billion. Notably, this cost reflects the life cycle of the census, which not only covers the count itself, but also preparations by the Census Bureau for the count and analyzing the data. The Census Bureau plans to make a number of changes to the 2020 Census to keep costs roughly similar to the 2010 Census. In particular, the Census Bureau plans to try to reduce the need for door-to-door household visits by improving outreach and providing more modes for response. To that end, the Census Bureau plans to send mailed questionnaires to only a portion of households. All other households will receive invitations—by mail and perhaps electronically—to fill out the census questionnaire online. The Census Bureau hopes 55 percent of households will respond to the 2020 Census via the Internet.

Accurate Census Count Clear State Interest. Not only do U.S. Census results affect the number of seats in the House of Representatives apportioned to each state, but the results also affect how much federal funds states receive. That is, many federal funding formulas use the census counts to determine what share of funds each state will receive. Thus, historically, the larger a state’s census count, the more federal funding the state likely will receive. If there is an undercount in any particular state, there is no mechanism to correct that error until the next U.S. Census. Consequently, California has a significant interest in achieving an accurate census count. (Importantly, the Department of Finance [DOF] also uses the U.S. Census to rebench its population estimates, which affect state and local spending limits, infrastructure plans, and state funding distributions.)

Previous State Efforts to Ensure an Accurate Count. As part of the 2007‑08 Budget Act, the state provided $3 million for incentive grants to local governments to participate in LUCA. Grants ranged from $1,000 to $75,000 per local government. The amount of funding provided to each local government was based on the amount of new construction in each jurisdiction. Overall, 42 percent of cities and counties participated and $1.8 million in grants were distributed. The state also provided $2 million to the Complete Count Committee for the 2010 Census, which encouraged full participation in the census. (The Complete Count Committee also received $10 million in private funding.)

Governor Proposes $7 Million for Incentive Grants for Local Governments to Participate in LUCA. The Governor’s budget proposes a substantial increase in the funding available for local governments to participate in LUCA. Under the Governor’s proposal, grants would range from $7,500 to $125,000 per local government based on housing activity in the jurisdiction. DOF defines housing activity as the amount of new construction, demolitions, conversions (to housing units), and housing unit annexations since the 2010 Census. The administration cites the low take-up rate of the 2007‑08 grants as an indicator that the grants need to be increased in order to achieve better participation from cities and counties. Moreover, the administration believes city and county participation is particularly important because the Census Bureau will no longer be validating 100 percent of the national residential address list through fieldwork. In prior years, the Census Bureau validated each addressed with in-field canvassing. For the 2020 Census, the Census Bureau will only be doing in-field canvassing where addresses cannot be verified through other means.

LAO Comments. Given the impact the U.S. Census results can have on California, taking steps to ensure an accurate count is meritorious. The Legislature may want to consider, however, whether the LUCA funds could be better targeted. In particular, county assessors could be better positioned, in some cases, to verify the Census Bureau addresses. County assessors are charged with maintaining the county rolls, which contain the addresses of all legal properties in the counties. (Properties built without permits may not be contained within the county rolls, however, city and county governments likely would not have these addresses either.) Rather than distributing the funds across all cities and counties, encouraging LUCA participation by county assessors (on behalf of the county and its cities) could increase the funding available to each county. Moreover, many cities may not have the resources available to review the federal address data. Targeting the work to the county assessors could result in a larger portion of California addresses being reviewed and verified.

Given the proposed changes to the 2020 Census, the Legislature will want to consider what types of outreach should be done in the years and months leading up to the count. While some residents may be very comfortable with filling out online census forms, some may not want to use that format. Though the Census Bureau will be undertaking various efforts to encourage participation, the Legislature may wish to consider in the coming few years how the state can ensure an accurate count given these changes and potential challenges.