The 10 States at Highest Risk of an Undercount

Which states will be hurt the most by a 2020 Census that does not perform up to par? A study by the Urban Institute examines three performance scenarios and projects the undercount by state for each one.

• In the low-risk scenario, the 2020 Census performs as well as the 2010 Census, with only demographic change influencing the outcome. The result is a net undercount of 0.27 percent versus the net overcount of 0.01 percent in 2010. So, even if the 2020 Census performs as well as the 2010 Census, "demographic changes alone would create a net undercount," explains the Urban Institute. 
• In the medium-risk scenario, the 2020 Census performs according to its operational plan (60.5 percent of households self-respond within six weeks), with no surprises. The result is a net undercount of 0.84 percent. 
• In the high-risk scenario, the 2020 Census performs below expectations (55.5 percent of households self-respond within six weeks, which is the Census Bureau's predicted lower bound of response) and participation is suppressed by all the hoopla surrounding immigration and the citizenship question. The result is a net undercount of 1.22 percent—more than 4 million people will not be counted. 

Under the high-risk scenario, the District of Columbia will experience the largest undercount (2.68 percent). After D.C., the 10 states that will experience the largest undercount include both blue states such as California and New York and red states such as Texas and Georgia....

10 states with largest projected undercount in 2020 Census (high-risk scenario)
1.98% in California
1.96% in Texas
1.76% in New Mexico
1.73% in Nevada
1.65% in Georgia
1.58% in New York
1.48% in Florida
1.47% in Maryland
1.40% in Arizona
1.33% in Louisiana

Source: Urban Institute, Assessing Miscounts in the 2020 Census