Every time the U.S. Census is conducted, New York City makes for an especially tough place to count. Its diverse demographic groups typically respond at lower rates than most of the country. The large immigrant population often requires language assistance. And just getting up-to-date addresses for the city’s many transient residents is a problem in itself.
So as the 2020 Census approaches, city staff are already canvassing neighborhoods, identifying potentially problematic blocks and doing other work to ensure these residents are counted. “We’re starting earlier than in the past because of threats to enumeration,” says Joseph Salvo, the city’s chief demographer.
For a number of reasons, concerns over missing hard-to-count residents are perhaps greater than ever. The Census Bureau’s current budget hasn’t gotten anything like the funding boosts it’s received prior to previous counts. If limited funding persists, it could hamper outreach budgets and in-person enumeration, both key to accurate counts. Online responses are also being used for the first time, but many urban poor still lack access to the internet. And more nontraditional living arrangements will make it harder to count certain households.
All of this explains why government officials and local advocacy groups are looking to get a head start. Ditas Katague, California’s Census coordinator, says several of the state’s larger jurisdictions have formed teams and started preparing earlier than usual. “I’m encouraged that localities are taking responsibility for what’s typically more of a federal operation,” she says. Officials are well aware that the Census count will carry far-reaching implications for federal spending and apportionment of political representation.
One of the best measures of participation is the mail return rate. Census statisticians compare the number of completed Census survey forms to the number of occupied housing units that received them. The national return rate in 2010 was 79.3 percent, but it varied significantly by state and region. Less than three-quarters of households responded in Alaska, Louisiana and New Mexico, while nearly 86 percent did so in Minnesota. In certain counties, more than a third of households failed to return their forms, and some parts of cities responded at even lower rates.
In order to respond, people have to receive the forms in the first place. Renters are typically undercounted because the forms don’t reach many of them, while homeowners are overcounted. High-rise apartments with mail delivered to a single central location also pose problems.
Census participation is largely tied to a jurisdiction’s demographics. The last Census is estimated to have undercounted the nation’s black population by 2 percent, and Hispanics by only slightly less. Whites, particularly older white women, were overcounted. One of the toughest groups to count is children under the age of five, many of whom live in low-responding neighborhoods. According to calculations from the Leadership Conference Education Fund, the great majority of young children in cities such as Cleveland, Memphis and New York reside in hard-to-count Census tracts.
Immigrants could be especially hard to reach this time, given heightened fears of deportation. New York’s Salvo says it’s not just the city’s undocumented residents who may not respond, but naturalized citizens as well. “The environment of hostility toward immigrants is a huge threat to the Census,” he says. Also reluctant to respond are those in any demographic who have a cynical view of government or don’t trust the Census. A report prior to the 2010 count identified 19 percent of the U.S. population holding such views.
Hard-to-count demographics don’t always spell reporting problems. Residents of Washington Heights in New York City, a predominantly Hispanic neighborhood with large numbers of undocumented immigrants, responded at high rates in the last Census. Community leaders and neighborhood groups worked assiduously to ensure people were counted. But Washington Heights was an exception to the typical pattern of poor response in low-income urban neighborhoods.
Some rural areas of the country have also been historically difficult to reach. One region that responded at low rates in 2010 is southeastern Oklahoma, which includes many Native Americans and high-poverty communities. Melanie Poulter, who coordinates counting efforts for the Tulsa-based Community Service Council, says they’re exploring partnerships with convenience stores to advertise the Census.
The single biggest source of anxiety for officials has been funding. To save money, the Census Bureau plans to rely more on administrative records to accumulate its list of mailing addresses, spending less on canvassing operations. Julie Dowling, a University of Illinois associate professor who convened a working group on the issue, warns that transient individuals, immigrants living with other families and other difficult-to-reach groups are frequently missed in such records. “The fact that administrative records will be used will leave certain people out,” she says.
In 2020, Americans will be able to respond for the first time to the Census online. But this, too, could undercount poorer households without internet access. An analysis by the City University of New York’s Center for Urban Research found that in the past, congressional districts with lower mail return rates were correlated with higher percentages of households who don’t have internet access or are burdened by poor connections.
To better reach these neighborhoods, the Census Bureau has pushed partnerships with local organizations. New York’s Salvo is asking philanthropists to help local community groups raise money for hiring staff and Census outreach. In the past, California has contracted with supplemental nutrition program counselors to talk about the Census with their low-income clients.
Reaching the hard-to-count, says California’s Katague, requires trusted messengers in communities. These can include pastors and faith groups who emphasize completing the Census. It’s also crucial that messages are tailored to individual communities. Katague explains that, depending on cultural attitudes, some communities might not trust any web address that ends with “.gov.”
One point that no one disputes: For communities that are determined to get an accurate count, the time to begin work is now. “States and localities need to be doing things in 2018,” Katague says. “You need to be engaging and finding out who these leaders are in the community.”
2010 Census Mail Return Rates
This map shows final mail return rates by county. Open interactive map.
< 65% >= 65% >= 70% >= 75% >= 80% Not Applicable Source: U.S. Census Bureau