The Washington PostDemocracy Dies in Darkness

McConnell’s flub on Black voters misstated the facts, too

The Senate Republican leader got in trouble for his comments, but we’re more interested in McConnell’s factual dexterity than his verbal dexterity

On Jan. 19, Senate Minority Leader Mitch McConnell (R-Ky.) said that African American voter turnout was “just as high” as for "Americans." (Video: C-Span)
Placeholder while article actions load

“The concern is misplaced, because if you look at the statistics, African American voters are voting in just as high a percentage as Americans.”

— Senate Minority Leader Mitch McConnell (R-Ky.), in response to a question from Pablo Manríquez of Latino Rebels, Jan. 19

This article has been updated

McConnell stepped in it the other day, making remarks during the debate over new state voting laws that were roundly criticized for suggesting that Black voters were not Americans. He later clarified that he had inadvertently failed to include the word “all” before “Americans.”

We’re more interested in McConnell’s factual dexterity than his verbal dexterity. His statement plays sleight of hand with the facts.

The Facts

For months, Democrats have pressed for a final vote on two proposed laws. The Freedom to Vote Act would override many state laws and set national standards such as a minimum of 15 days for early voting, mail-in ballots and universal rules for voter identification. Another bill, the John Lewis Voting Rights Advancement Act, would strengthen and restore parts of the 1965 Voting Rights Act and make it easier for voters to challenge state rules.

Republicans have said the legislation is unnecessary and tilted to favor Democrats. McConnell’s statement was intended to show that Black turnout has not suffered in recent years.

On the face of it, his comment would appear only slightly overstated. A McConnell spokesman directed us to census data showing that Black turnout was relatively close to overall turnout in recent elections. That’s not “just as high” as McConnell claimed, but maybe it’s close enough for government work?

  • 2020: Total turnout 66.8 percent, Black turnout 62.6 percent
  • 2018: Total turnout 53.4 percent, Black turnout 51.5 percent
  • 2016: Total turnout 61.4 percent, Black turnout 59.4 percent
  • 2014: Total turnout 41.9 percent, Black turnout 39.7 percent

In the most recent presidential election, the gap was about four percentage points, but it was closer in the three previous election cycles.

But here’s the trick McConnell is playing. He’s comparing Black turnout to overall turnout. These are not comparable data sets.

Michael McDonald, a University of Florida political scientist who analyses census data on voter turnout, said it is more appropriate to measure Black turnout against White turnout.

“This statement is false, according to the Census Bureau’s Current Population Survey Voting and Registration Supplement,” McDonald wrote in an email. “Non-Hispanic African Americans currently have lower turnout rates than non-Hispanic Whites in the 2020 and 2016 elections.”

In 2020, for instance, 72.6 percent of non-Hispanic White Americans voted, compared with 65.6 percent of non-Hispanic Black Americans — a gap of seven percentage points. In 2016, the gap was 4.8 percentage points.

That trend is consistent dating to 1986, according to McDonald’s analysis, with only two exceptions — the presidential election years of 2008 and 2012. Those are the years that Barack Obama, the United States’ first Black president, was on the ballot.

McConnell’s spokesman responded with a curveball. “This whole debate is in the context of the Voting Rights Act, looking at the turnout in some of the former preclearance states is instructive,” he said, referring to a section of the law that had prevented Southern states from making voting changes without preapproval from the attorney general or a court ruling. Recent Supreme Court rulings, such as Shelby County v. Holder in 2013, have gutted that section of the law.

The spokesman said McConnell’s staff had analyzed census data from the Current Population Survey and come up with interesting findings. He claimed that for the states affected by the Voting Rights Act, the gap between Black and White turnout was much smaller than the national average. (Note that some of the numbers are averages — we will explain that below.)

  • Alabama: Black turnout was above White turnout in 60 percent of the past five elections, and averaged just 0.64 percent less than White turnout.
  • Georgia: Black turnout below White turnout by 1.15 percent.
  • Mississippi: Black turnout has been above White turnout for five consecutive elections and on average is 5.9 percent higher.
  • Louisiana: Black turnout on average is 1.2 percent below White turnout.
  • South Carolina: Black turnout on average is 0.18 percent below White turnout.

“The successes in former preclearance states post-Shelby County is weighed down with states like Oregon, a deep blue state which mails every voter a ballot and has Black turnout 28.5% below white turnout; Colorado, a blue state which has a 18.2% delta with the same voting systems as Oregon; Washington, with a 21.3% average delta between white turnout and black turnout; and Massachusetts, which had the worst Black voter turnout in the country,” McConnell’s spokesman said in an email.

It turns out that McConnell, in various venues, has cited such statistics.

“Voter turnout last November was the highest in decades. African-American turnout was twice as high in Mississippi as it was in Massachusetts,” he said on the Senate floor in May.

And, in a January opinion article for the Louisville Courier-Journal, he wrote: “In many of the Republican-led states where Democrats claim ‘Jim Crow 2.0’ is out in force — like Mississippi, Tennessee and Missouri — Black voter turnout routinely outpaces white voter turnout. These are signs of a healthy democracy, not one on its deathbed.”

Hmm. We went to the census website and looked at the tables, and these numbers were confirmed for states such as Massachusetts and Colorado. A group called the Honest Elections Project, run by a former Heritage Foundation staff member, publishes similar data in easier form to digest.

But there’s a catch. Apparently, Black voters in the South are more likely to respond to the census voting questions than those outside the South. That results in a lot of missing data — and the Census Bureau has chosen to count the missing data as “did not vote.”

“When you remove the missing data, the comparison turnout between the South and non-South changes, and between Whites and non-Whites in Southern states,” said McDonald, from the University of Florida. “For some reason, Massachusetts has one of the largest African-American nonresponse rates.”

The estimates are also unreliable in states with small percentages of Black voters. McDonald said the census counts the missing data as non-votes because otherwise the Current Population Survey (CPS) turnout would be higher than the aggregate statistics indicate.

A Census Bureau spokesperson acknowledged the problem.

“If a respondent does not respond to our question of whether they voted, then we do not consider them to have voted. In other words, if they have no response, we do not consider them to have voted. We calculate voters as those who report voting, and nonvoters as those who report not voting,” the spokesperson said, adding that the raw data at the state level needs to be carefully handled.

“Data users should adjust their analysis to account for the fact that we do not exclude nonrespondents from the voting-age population and that we do not know whether nonrespondents voted or did not vote.”

As for the averages of the preclearance states listed by McConnell’s staff, those numbers include the years when Obama ran — and so mask an increase in the White-Black turnout gap in many of those states since the Supreme Court issued its ruling in the Shelby County case.

In South Carolina, for instance, the turnout for Black voters in 2012 was 69.3 percent, more than the 63.5 percent for White voters. But in 2020, the Black voter turnout was just 53.9 percent, compared with 69 percent for White voters — meaning the gap between the two had shifted by almost 21 percentage points. In Georgia, Black voter turnout was three percentage points higher than White voter turnout in 2012 and more than six percentage points lower in 2020.

The Pinocchio Test

We went into an interesting rabbit hole that taught us something new about the state-level census data. McConnell should not cite this data making the case against the laws pressed by Democrats, given the Census Bureau’s warning about how the nonresponse rate messes up the numbers.

Let’s return to the statement that first sparked controversy. McConnell not only overstated the situation a bit, but comparing the turnout of Black voters to the entire voting population is misleading because these numbers are apples and oranges — the entire population (in which the turnout rate is dragged down by other ethnic groups) versus just one ethnic group. The more appropriate comparison is between ethnic groups, such as White Americans and Black Americans. That comparison shows there has been a persistent gap — and it increased in 2020.

McConnell earns Two Pinocchios.

Two Pinocchios

(About our rating scale)

Send us facts to check by filling out this form

Sign up for The Fact Checker weekly newsletter

The Fact Checker is a verified signatory to the International Fact-Checking Network code of principles

Loading...