clock menu more-arrow no yes mobile

Filed under:

How Republicans are undermining the 2020 census, explained with a cartoon

Why it’s so hard to count everyone — especially marginalized populations.

In March, the Trump administration added a question about citizenship to the upcoming 2020 Census.

Census Bureau research shows that this question could actually deter people from responding to the survey, which means it might cause an undercount of vulnerable populations.

This is why residents from Maryland and Arizona filed a lawsuit alleging that the US government is violating the Constitution by asking about citizenship, given the current political climate. The Trump administration has tried to get the case dismissed, but three federal courts have said the case can go forward.

Already, lawmakers have grilled Census Bureau stakeholders, including John Gore, who represented Trump’s Department of Justice and was the key person who pushed this citizenship question onto the census. Gore has a history of defending redistricting and voter ID policies that have adverse affects on people of color.

This is among the many reasons census advocates say Republicans are undermining the census. But it’s worth understanding how exactly their actions have hurt our effort to count every person in the country — an effort that undergirds our democracy.

What is the census and why is it hard?

Every 10 years, the US has to count every person living in the country.

This process is called the census.

The last one was in 2010, which means the next one is coming up in 2020.

Animation of a census enumerator going house to house to count people

This count is crucial because it serves as the basis for the apportionment of all 435 seats in the House of Representatives. That means that the number of seats a state has in the House, and, by extension, the Electoral College, is determined by its share of the population as counted by the census.

For example, between the 2000 and 2010 censuses, Florida gained nearly 3 million residents. This gave the state two more seats in the House of Representatives. Meanwhile, New York’s population stayed steady, falling behind many other states that grew, so it lost two seats.

In addition, the census data is used to figure out how the federal government will distribute funding for programs ranging from highway construction to food stamps to Pell Grants.

And this process counts citizens and noncitizens.

Animation of how the above counts affect representation

This is an incredibly hard problem, and the process is very expensive.

In fact, over the past several decades, the cost of the census has skyrocketed:

But even with all this spending, there’s a problem.

After the count, the Census Bureau does follow-up surveys to figure out how accurate it was.

And the bureau always finds that it undercounted vulnerable populations — for example, minorities, young children, people who are poor, and people who don’t speak English.

Chart: White residents were overcounted in 2000 and 2010, while black and Hispanic were still undercounted.

The undercount happens because those people are harder to find. They tend to be more transient, less trustful of government, and less tied to communities.

We can use a statistical method calling “sampling” to get a more accurate count. But Republicans have opposed any kind of manipulation of the raw data because they believe it’s an attempt to hurt them electorally. The Supreme Court has also said sampling can’t be used to apportion seats in the House of Representatives, though it left room for sampling to be used for state redistricting and apportionment of federal funds to states.

So even though we know there are vulnerable people in the US who might go uncounted, the census can’t account for them.

Comic panel: Enumerators saying we missed people. Census worker saying we can’t do anything about it.

All of this brings us to the upcoming 2020 census.

Former Census Director John Thompson recently said, “There’s a lot of concern about the potential for an undercount in the census.”

Experts are worried that we will severely undercount vulnerable people. There are many reasons, but the biggest worries are:

  • Republicans have underfunded the effort.
  • The Census Bureau is using new technology to try to cut costs — but that new technology isn’t being properly tested because of the lack of funds.
  • The Trump administration also added a question about citizenship that could scare immigrants away from responding.
  • Oh, and the Census Bureau has been without a leader since Thompson resigned last year.

The census undergirds our democracy — yet we’re about to conduct it under highly worrisome circumstances.

Animation of enumerator counting houses, but not getting to the most vulnerable places.

The census is now using online surveys. Why does this cause problems?

In the past, the US Census Bureau sent out surveys in the mail. But over time, people have become less responsive.

Chart: The number of people responding to the mail survey has decreased over time.

So for 2020, the Census Bureau will urge people to fill out the survey online.

Most people will get a letter in the mail with a unique security code, and you’ll be able to use that code to answer the 10 questions on the Census Bureau’s website.

Households that don’t respond will get a follow-up paper survey that they can mail back.

The census plans to mail 80% of households a postcard telling them to fill out the survey online.

In addition, the census will allow people to call in and answer the questions. It will use a new language recognition technology.

This use of technology is expected to save the census a lot of money, since it will encourage people to respond on their own rather than requiring census workers to knock on their door.

The new technology will save the Census Bureau $5.5 billion in 2020.

Here’s the problem: This is a completely new procedure with a whole new set of complications.

For one, there are the technical complications that come with building this system.

But there’s also the possibility of hacks. The US Government Accountability Office warned that cybercriminals would likely try to steal personal information from people filling out the survey. In addition, news of digital databases leaking personal information — like the Equifax hacks and the Facebook scandal — could scare people into not responding.

Comic panel: Man who encounters a phishing scam, and tells everyone that the census online portion is all a scam.

These factors could have a different effect on different populations. This is why the Census Bureau wanted to test these systems.

But because of budget concerns, two of the three “dress rehearsals” have been canceled, and we are going into this process with limited testing.

There will be 200,000 fewer Census workers knocking on doors

A huge portion of people won’t respond to any of the online or mailing efforts, so census workers will have to knock on doors.

They will do this up to three times — down from six times in 2010 — before they starting asking your neighbors (or other reliable sources, like your mail carrier) about your information. And then, finally, they’ll consult state and local records to try to fill out that info.

In 2010, it took hundreds of thousands of census workers 10 weeks to reach everyone.

It took about 10 weeks to complete follow-up visits in 2010.

And this speed was only possible because the census tries to employ temporary workers from within the community. For example, a GAO report found:

... at a Native American village in New Mexico, local enumerators were aware that according to the community’s customs it was considered offensive to launch into business without first engaging in conversation.

... local enumerators in hurricane-affected rural areas of Louisiana were able to successfully locate households based on their knowledge of the geography.

Animation of census enumerator encountering a house that was destroyed in a hurricane, but knowing where the family lives based on knowledge of the community.

This highlights exactly why vulnerable populations are undercounted.

They tend to have less stable housing situations, whether it’s because they are renters, homeless, or victims of natural disasters. In addition, the process requires a certain community cohesion, like a census worker who knows the area or neighbors who are familiar with your household.

Animation of enumerator not seeing a house listed on her map, but knowing where it is based on knowledge of the community.

But for the 2020 census, because of budget constraints, there will be 200,000 fewer people knocking on doors.

The 2010 census has 500,000 enumerators. The 2020 is projected to have 300,000.

And if one of those people the census doesn’t hire is the person who had local knowledge of a hurricane-affected area, then many people who live there may not be counted.

Animation of enumerator fading away.

Here’s why the citizenship question worries demographers

As if these problems weren’t enough, the Trump administration added a last-minute question about citizenship to the 2020 census.

But remember: The Constitution says we are supposed to make our best effort to count citizens and noncitizens and then apportion representation based on those counts.

And there’s evidence that this citizenship question could undermine that effort.

Enumerator asking about citizenship, which scares Latino communities from responding.

In recent field tests, the Census Bureau found that this question spooked some people, especially Latinos.

One person reportedly ditched the census worker in their own home. Others were less likely to give accurate information. And then there’s this report from a census worker:

There was a cluster of mobile homes, all Hispanic. I went to one and I left the information on the door. I could hear them inside. I did two more interviews, and when I came back, they were moving. ... It’s because they were afraid of being deported.

There is a fear that the Trump administration will use this information to deport noncitizens. This is unlikely — the citizenship question in the census doesn’t separate documented immigrants from undocumented, and the Census Bureau can’t legally share individual information with other agencies.

However, there is another outcome the Trump administration is likely hoping for.

If this question causes an undercounts of noncitizens, then representation of those areas — which tend to lean Democratic — also goes down. This is one of the main reasons a recent must-read Mother Jones piece accused the Trump administration of “rigging” the census.

Diagram of Republican-leaning areas benefiting, while Democrat-leaning areas not.

Attorney General Jeff Sessions said the purpose of the citizenship question was to have a better idea where eligible voters of color live, which would help them “better enforce” the Voting Rights Act.

But the census already asks about citizenship in the ongoing American Community Survey, which surveys a portion of Americans every month and then uses statistical tools to extrapolate the data. This data stays up to date throughout the decade, and the margins of error are quite good.

The cost of sabotaging accuracy

If vulnerable groups are undercounted, they won’t get proper representation when it comes time to apportion representation.

In addition, when lawmakers are figuring out how to allocate funding, the places where people were undercounted could get less money than they deserve.

Until this year, a Republican-controlled Congress seriously underfunded the efforts to prepare for the 2020 census that required a lot more upfront testing than previous counts.

This year, the census got a larger-than-expected budget to make up for the underfunding. But census advocates told Science magazine that it’s only enough to provide the “minimum resources needed to prepare for its constitutional mandate.”

The census is the essence of America’s democracy. It’s the tool that tries to give everyone an equal voice, regardless of where they are in society.

For the most part, our leaders have taken this constitutional edict seriously. This time, it’s unclear if that’s true.

Two houses counted. One house, with American flag waving in the background, not counted.

Correction: A previous version of this piece said that enumerators will follow up with nonrespondents up to six times before asking a reliable source like a neighbor or mail carrier. This was true in 2010, but for the 2020 census they will ask a reliable source after three follow-up visits.


Sign up for the newsletter Today, Explained

Understand the world with a daily explainer plus the most compelling stories of the day.