I know I’m fighting a losing battle with this post. It won’t go viral. It probably won’t change many minds.

But I’ll give it a shot anyway.

No matter how many times people make the claim, it is simply wrong to say that 81 percent of white evangelicals in the United States voted for Donald Trump to become president.

First (and I know this is quibbling), the number that people are meaning to cite is actually 80 percent.

(Media originally reported 81 percent, but that was based on initial reports of the exit poll before the tabulations were complete.)

Second, the statistic was not purporting to measure the total percentage of all white self-identified evangelicals.

Rather, the number is supposed to indicate the number of white voters who self-identify as born-again or evangelicals and voted for Trump.

That sounds like mere semantics, but it actually represents a significant difference. Evangelical historian Thomas Kidd uses recent statistical analysis to estimate that 40 percent of white evangelicals didn’t vote in this election (see, e.g., this).

If we then grant the 80 percent figure for the remaining 60 percent who did vote ended up casting their ballot for Trump, then it would be the case that less than half (48 percent) of white self-identified evangelicals voted for Donald Trump.

Third, we know almost nothing about the 80 percent beyond a religious label they affirm or an experience they claim.

Do they go to church? Are they Protestant? Unless we are willing to say that “an evangelical is anyone who says he or she is an evangelical or says he or she has been ‘born again,'” then we have to admit that we are talking more about a label of self-designation than an actual movement or network, much less a reflection of theological belief or religious practice.

For example, an array of theological traditions outside of the traditional evangelical movement have adherents who say they are “evangelical” or have been “born again,” including:

  • mainline Protestants (27 percent)
  • Roman Catholics (22 percent)
  • Orthodox (18 percent)
  • Mormons (23 percent)
  • Jehovah’s Witness (24 percent)
  • spiritualist Christians (24 percent)

[As an aside: some people rightly point out that 35 percent of those who self-identify as evangelical are minorities:

  • 19 percent black
  • 10 percent Hispanic,
  • 6 percent combination of Asian, mixed race, or other ethnicities

But the 80 percent number was specifically limited to white voters who self-identify as evangelical/born again, not to all who identify with these religious markers.]

Finally, it is reasonable to estimate that less than half of the 80 percent actually hold to traditional evangelical beliefs.

As researchers have discovered serious limitations in how to identify and measure evangelicals beyond the exit-poll data that merely asks this as a yes-or-no self-identification question, more sophisticated polling data has recently been used, revealing that fewer than half (45 percent of respondents) who self-identify as evangelical strongly hold to evangelical beliefs as articulated by the Bebbington Quadrilaterial (with its emphases on conversionism, biblicism, activism, and crucicentrism).

This is not a defense of evangelicals and their relationship to Donald Trump. (I have been a critic regarding that relationship, especially with regard to what historian John Fea calls the Court Evangelical leaders who serve as his acolytes.) This says nothing about what his current level of support may or may not be. It’s simply a little plea for more accuracy about how we describe the role of evangelicals in the 2016 election. Criticize them all you want—but let’s first define our terms and strive for accuracy.