2020 Census FAQ

What data does the 2020 Census cover?

The 2020 Census covers a surprisingly large number of issues with very few (7) questions:

The data may be combined to provide multiple data points, e.g., the number of non-Hispanic, Black children in a geographic area or the percentage of Native American homeowners in a geographic area.

How does the 2020 Decennial Census differ from the American Community Survey?

The American Community Survey (ACS) is an ongoing effort to assess the demographic, social, economic, health, and education status of the U.S. population. The Census Bureau sends surveys to a random sample of U.S. addresses every month, with each household having a roughly 1-in-480 chance of being selected every five years.

The ACS is generally released in two phases. A single-year sample for the previous year, which includes detailed data for larger geographies (i.e., states, most U.S. counties, metro areas, and places with more than 65,000 people), generally comes out in September. A five-year sample usually is released in December and includes detailed data for smaller geographies, usually down to a tract level.

Due to the COVID-19 pandemic, a single-year sample will not be released in 2021.

How does the 2020 Decennial Census differ from previously conducted Censuses?

The 2020 Census represents an evolution of the decennial headcount in some ways; in others, it’s a devolution. It asks the same questions as the previous Census and adds space for Black and white people to describe their race in more detail (i.e., Caribbean, Irish, and so forth).

From a historical perspective, the 2020 Census continues to evolve. Consider the nation’s first Census in 1790 had the same basic categories, with the exception of a native/foreign-born and slave category. Its racial categories also were limited to white and non-white; the 2020 Census includes white, Black, Asian, Native American, Pacific Islander, other, and multiracial. 

The 2020 Census diverges from its predecessors in its attempt to alter the results with a process known as differential privacy. Although the risks of people breaching Census data are small and limited in their ability to cause personal injury—since most of the decennial results are usually obvious (race) or easily discovered (homeownership)—the Census Bureau is moving people among different geographies, ostensibly to protect their privacy. 

While the efforts to ensure privacy are laudatory, the shifting of populations has created demographic anomalies, such as neighborhoods populated only by children, areas where there are no racial or ethnic differences, or areas where racial differences could affect the redrawing of voting districts that generally occurs every decade.

How is the Census used?

The Census has three primary purposes:

What is included in the data and why?

The basic purpose of the Census is to count the population within each state, so that congressional seats can be proportional. In a representative democracy, states with large populations should have more members in the U.S. House of Representatives than smaller states.

Over time, however, the use of the Census has expanded. It has been used as a tool to make sure that congressional and other voting districts include a representative population under laws, such as the Civil Rights Act of 1964 and the Voting Rights Act of 1965, enacted to ensure that the voting rights of Black Americans were protected. 

The decennial Census also serves as the population base for subsequent American Community Survey releases, which are used to distribute trillions of tax dollars over a decade. Because it’s an actual headcount—as opposed to a survey—the decennial Census creates a foundation for the annual ACS data.

What was not included in the data and why?

Contrary to fever dreams among right-wing politicians, the Census does not ask about the following:

Every question asked by the Census, or the American Community Survey for that matter, is the result of a congressional law. Extremely controversial questions, such as pertaining to political or religious affiliation, would stand little chance of winning congressional approval.

Former President Donald Trump, for example, attempted to dilute minority representation by trying to insert an additional question into the 2020 Census that would have required people to provide their citizenship status. Fortunately, the effort was defeated by a series of lengthy and time-consuming court battles.

The Census is often confused with the ACS because of the relative novelty of the survey. Prior to 2010, 1 in 6 American households would receive a “long form” from the Census, asking questions about their housing, income, family, education, work, and transportation. Businesses and policymakers, however, realized that the long form data became quickly outdated. For instance, many things changed between 2001 and 2009, which led to the push for a more up-to-date product. The decennial Census, however, is still often confused with the American Community Survey.

What geographies are available for the 2020 Census dataset?

The Census is broken down into several dozen geographical units, ranging from the esoteric (Traffic Analysis Zones and Urban Growth Areas) to the standard (States, Counties, Cities). The most important geographies, however, are generally:

Other permutations include almost 40,000 county subdivisions, which largely correlate to towns and villages in the Northeastern U.S.; metropolitan and micropolitan statistical areas; school districts; legislative districts (state and federal); zip code tabulation areas or ZCTAs that approximate the boundaries of a postal delivery area; and lands occupied by Native Americans, Alaskans, or Pacific Islanders. 

What issues may arise from the data?

Problems associated with differential privacy, or the shifting of certain groups of people from one geography to the next in the name of security, stands to create an enormous headache for the redrawing of voting districts. In some areas, it may decrease minority representation; in other areas, it may increase minority representation. Either way, introducing deliberate error into the smallest of Census geographies is a recipe for a nation that fails to provide equal representation for its citizens.

While the differential privacy issue looms largest, it’s well understood that the 2020 Census had one of the lowest response rates in modern history. According to Census data released on Aug. 25, 2021, 5.95 percent of households failed to respond to questions about their date of birth; 5.35 percent refused to answer whether they were Hispanic or not; and 5.77 percent didn’t answer the question about their race.

The numbers increased significantly in comparison to the 2010 rates, when only 3.35 percent refused to specify birth; 3.99 percent didn’t answer the question about Hispanic ethnicity; and 3.31 percent refused to describe their race.

What is unique about the 2020 Census?

The 2020 Census was the first Census that allowed people to provide their data online or by telephone.

It was only the 12th Census to be conducted during a presidential election.

The Census was conducted during one of the worst pandemics in modern history (the 1918 Spanish flu epidemic was winding down by 1920).

It was the first Census to allow white and Black people to elaborate on their race. Previously, only the Asian, Pacific Islander, Native American, and Hispanic communities had been allowed to specify their origins.

Its PL94-171 results, used for redrawing voting boundaries, were made available in one release. Previously, releases were made in groups of five or fewer states over a period of weeks.

What makes Social Explorer's census data unique?

Social Explorer is the only data provider that quickly and easily allows you to compare results of the 2010 Census to the latest version, even down to more than 11 million blocks, the smallest geography available from the decennial headcount.

However, there are two caveats. First, geographic boundaries for places and counties may change over the course of a decade due to political factors, such as annexation, remapping, or renaming. Likewise, the Census Bureau makes population-based adjustments for lower-level geographies, such as blocks, block groups, and tracts. Social Explorer has made adjustments to allocate changes in population based on land area.

Second, the Census Bureau used a Disclosure Avoidance System to ensure that no individual's race, ethnicity, age (older or younger than 18), or housing tenure (own or rent) could be identified. The system meant shifting individuals across Census blocks, resulting in false information being presented. For example, areas being composed entirely of people younger than 18 years old or neighborhoods of exclusively one race that don't exist. Social Explorer recommends extreme caution when making comparisons at very small geographic levels, especially among blocks, block groups, and tracts.