Time for Us to Stop Undercounting Ourselves – Really, Stop It

Gregory Paul

Back in the late 1970s, the religious Right was embarking on social and political activism in a big way that would make it a major force in American life. To promote the effort, in 1979 Jerry Falwell, Paul Weyrich, Richard Viguerie, and others did something as audacious as it was an unethical lie. They founded an organization they titled the Moral Majority. That was a lie because even back then, when the United States was considerably more religious than other Western countries, theological conservatives (“theocons”) were nowhere close to constituting the majority of a population that had already in the main adopted the sexual revolution plus reproductive and women’s rights. It was the swinging 1970s, for heaven’s sake. In issuing such a big lie, the radical “theoright” was shamelessly taking a propaganda lesson from the Russian atheistic communists, a.k.a. the Bolsheviks, who in 1917 declared themselves the majority when they were in fact a small extremist minority. That cynical tactic worked enough to help the Reds seize power, with overall bad results. Likewise, the tactic did not hurt the self-labeled Moral Majority, which went on to help elect Ronald Reagan president and place a series of deep theocons on the U.S. Supreme Court.

Based as they are on as much objective, rational information and analysis as is humanly practical—and seriousness about being ethical—the atheist and humanist movements are not nearly so shamelessly propagandistic as the religious Right has been. Despite all the big gains in American atheism and other expressions of nonreligion over recent decades, none on the nontheist side would imagine calling ourselves the Atheist Majority. That moniker may well be true in Scandinavia and perhaps Holland and France—and could come true in the American future. But instead, if anything, seculars are tending to do the opposite, as is the mainstream news media.

We are chronically prone to undercounting ourselves.

Not so long ago, the United States seemed destined to remain in perpetuity the most religious nation in the developed West. Even back in the early 1800s, Alexis de Tocqueville noted that Yankees were more theistic than Europeans. In time, it became a common premise that America was and always would be pious, one mark of American Exceptionalism that distinguished that nation from the other advanced democracies, a component of the God-blessed “American Way.” By and large, even seculars bought into this premise. Then, starting in the 1980s and into the 1990s, several polling organizations started reporting a distinct rise in what is often called the religiously unaffiliated Nones. Even the number of self-proclaimed atheists ticked up a tad. In the summer of 2002, yours truly wrote an FI article noting that while the United States was behind the times when it came to being a rational country, it showed signs that this might change in the new century1. By the time I did the entry on the subject in the 2007 New Encyclopedia of Unbelief, enough statistics were on hand to know that assorted forms of untheism were definitely on the rise2, a reality Paul Kurtz acknowledged at a conference in Washington, D.C.

Nonbelief—or at Least Nonaffiliation—on the Rise.

Remember how, in the wake of 9/11, a set of books by Sam Harris, Richard Dawkins, Christopher Hitchens, and Daniel Dennett became the first explicitly pro-atheist best sellers? Many in the movement were delighted. But others fretted that being so publicly aggressive against religion would backfire against the seemingly demographically delicate atheist cause.

That did not happen. The big news that American irreligion was on the way up first gained really wide traction in the 2012 Pew report, “‘Nones’ on the Rise,” in which the nation’s premier analytical survey organization stated that the nonaffiliated had risen by a demographically startling 5 percent of the national population in just five years, to comprise a fifth of all Americans. That total is competitive with Catholics and evangelicals in terms of absolute numbers. In 2015 and most recently 20193, Pew found that the same rapid pace had continued, with Christians plummeting by about a tenth of the population while Nones soared the same amount in a mere decade. According to the Pew results, most Nones are theists. With 9 percent of Americans saying they are atheists or agnostics, the atheists are less than half of that.

And this is where things get statistically dicey for nontheism. In part because Pew is the most respected polling institution, its statistics tend to be the go-to data set cited endlessly by reporters—and most everyone else—when it comes to what Americans do and do not believe these days on matters theistic. That includes atheists, humanists, and seculars, including those writing for FI and Skeptical Inquirer.

For a whole bunch of statistical reasons, this is not the best idea.

Getting Our Terms Straight

To understand why, we need first to look at some issues, starting with definitions. Take the famed Nones: they are nonaffiliated, in the sense that they do not belong to a particular organized sect or denomination. But as Pew itself has pointed out, the Nones are merely a subset of the demographically much-more-important group we should be talking about: the nonreligious “Nons,” a group that you, the FI reader, almost certainly belong to. Where Nones reject denominational affiliation, Nons eschew identification with religion itself. Neither the Nones nor the Nons are all atheists; it is quite possible not to adhere to any religious institutions or beliefs while believing in the existence of a creator or the like. At the same time, atheists can be affiliated with a religion; for example, some attend church for social and familial reasons.

It gets complicated.

All this brings us to the core etymology of atheist and its linguistic accomplice agnostic. I regularly attend the Baltimore Atheists and Skeptics and the Separation of Church and State meetups. At these gatherings all are welcome. Again and again, I’ve heard newbies claim that they are not “atheists”; they are “agnostics.” They are under the impression that to be an atheist one has to be ardent about it—strongly or totally convinced there is not one or more gods and perhaps deeply antireligious. This when a-theism properly understood marks merely the absence of theism. If one has any significant belief in the reality of at least one deity, from barely casual to zealously absolute, one is a theist. If one is not a theist, one is an atheist. It does not matter how much one disbelieves, nor does it have anything to say about what else one thinks. It is possible to be an atheist and believe in ghosts, Bigfoot, or alien visitors; I know such people. As for agnosticism, that is largely a subset of atheism, in that only absolutist atheists are not agnostics. Agnostics, then, compose the majority of nontheists. Based on his famous dissection of his own beliefs in his landmark book The God Delusion, Richard Dawkins is an accurately self-described agnostic atheist while being a hardline anti-theist.

The above discussion is not a mere idle definitional spat. If even folks who do not believe in the existence of gods and who are serious enough about it to show up at atheist group meetings—where they then deny being atheists on a regular basis—then it follows that many, many atheists who’ve been interviewed by Pew, Gallup, or Harris have told the researchers who called them that they are not atheists even though they most certainly are, even if they are marginal about it. In that case, they are denying being an atheist out of definitional ignorance, which is contributing to a serious undercount of the cohort.

Many others who know full well they are atheists deny it. That brings us to a fact that has been as long-known as it is well-known among pollsters, namely that Americans notoriously over-report their theism, while under-reporting their irreligion, when asked about it by trained professionals. The reasons are similar to the known reluctance of survey respondents to admit to not being heterosexual, or to being a feminist even when their positions on women’s rights fit the definition, and so forth. As secular Americans know well, we remain the targets of considerable disrespect and discriminatory attitudes, to the degree that many unbelievers won’t tell family, friends, and coworkers that they dare not believe in items supernatural. Meanwhile, theists are prone to exaggerating their piety. For example, about four in ten regularly tell Gallup that they attend church on a regular basis. But there are not enough pews in America to accommodate so many Christian rear ends. Careful research shows that the real figure is closer to a fifth. Ergo, a lot of self-described Christians are lying to pollsters about it.

Then there is the growing high-tech sampling problem, as increasing numbers abandon the analog landlines on which major survey organizations still largely rely in favor of less statistically useful digital social media. This poses a special problem for surveying opinions on religion among youth. Younger Americans are much more atheistic than their elders, and they have pretty much ditched Mr. Bell’s invention, so as explicitly noted by Pew, this adds further real risk that nonbelievers will be undercounted. (This is, by the way, directly contrary to the obstinate theocon sociologist of religion Rodney Stark, who for years has been making the increasingly bizarre claim that Americans are remaining as God-believing as they always were even if they are dropping organized theism, a bogus proposition I exposed online4. Of course, this is the same Rodney Stark who opined in a later book that the Crusades were actually a good idea, never mind that Christian soldiers literally ate Muslim captives on occasion.)

Atheism, then, is often inaccurately denied due to some combination of:

  1. Definitional ignorance;
  2. People being uncomfortable admitting not being godly to someone they don’t know; and
  3. Technical problems in getting responses from irreligious youth.

It follows, then, that the number of atheists and other nonreligious is being seriously undercounted in these United States even by respected and seemingly objective survey outfits such as Pew.

Then again, maybe not.

Now for Some Even Better News

In a 2014 look at the millennial generation, buried in an appendix that even I did not know about until someone else directed me to it, Pew5 used several survey parameters to calculate what percentage of each generation is actually nonreligious. Doing so is of importance, because calculating what people actually think and do when it comes to religion/nonreligion is more accurate than directly asking whether or not one is an atheist, for the reasons detailed above. Aside from showing the usual pattern of decreasing age strongly correlating with lower religiosity, the result found that about half of Americans are actually nonreligious. For reasons that are obscure to me, Pew has never publicized this important result, and as a result no one else has picked up on it.

But wait, there’s more. Pew and Gallup are not the only polling organizations out there. Take Harris, which released a major survey6 that has not received nearly as much attention as it deserves. Its study “Americans’ Belief in God, Miracles and Heaven Declines, Belief in Darwin’s Theory Rises” found that those who do not consider themselves somewhat or very religious rose from 31 to 42 percent between 2007 and 2013 Then there is RedC’s7 “Global Index of Religion and Atheism,” which tallies Americans who deem themselves religious as plunging from 73 to 60 percent since 2006. Combined with Pew’s far better-known stats on the Nones, this furnishes compelling evidence that the cohort that is not religious has been growing by about a full tenth of the total population per decade since the new century began. That’s huge; it’s a demographic explosion of irreligion that few if any predicted at the beginning of the millennium. Then there is the American Family Survey8, which is connected to, of all things, the Church of Jesus Christ of Latter-day Saints. Over recent years, it has been measuring a loss of religiosity of 1 percent a year—again, approximately 10 percent per decade—and found that just 43 percent of Americans consider religion very important in their lives. That is not too far from Gallup’s finding9 that only 37 percent claim to be highly religious. Add to all that the latest Gallup results on church membership. Gallup10 has been asking that same question all the way back to the Father Knows Best/Leave It to Beaver era. In those days, almost three quarters of Americans belonged to a Christian religious institution. By 2000, church membership had gradually declined to just two-thirds of the nation. Since those days, the figure has nosedived like a World War II Me-109G fighter trying to escape a P-51B. It’s fallen to about half, another jaw-dropping loss for the once-mighty institution of American theism. Also down is church attendance, by about a fifth. Such rapid change cannot be limited to just the liberals who make up most of the irreligious. Indeed, nonreligion is now making major gains in the white working class, as well as among the long-pious African American cohort.

But, you ask, what about atheists, who do not possess a belief in deities? Fortunately, the Harris survey mentioned above addressed that question by not making the mistake of directly asking respondents if they are atheists. As we’ve seen, that query only discovers how many Americans are willing to label themselves by a social pejorative of which many respondents don’t even know the definition. It does not smoke out how many atheists there actually are. Wisely, Harris asked instead about levels of belief in deities, which tends to reduce (though not eliminate) the underreporting of nontheism. For example, Harris asked respondents if they were absolutely certain there is no god, somewhat certain, or not sure. Among respondents who are not theists, and therefore qualify as at least marginal atheists, almost a third opted for one of those options in 2013, up from a fifth in 2003. Once again, we encounter that seemingly steroid-driven growth rate of a tenth of the population in ten years.

That one-third of this nation falls into the atheist category is supported by a 2017 analytical study in Social Psychology and Personality Science11. The researchers used standard techniques to compensate for the way members of a disrespected cohort tend to underreport their actual opinions to pollsters to produce more realistic tabulations. Sure enough, it looks like somewhere in the vicinity of a third of respondents are atheists, ranging from hardcore to barely so. Returning to the Harris results, we can further determine that one in ten Americans or more are strongly convinced atheists, and up to a fifth are very skeptical about the whole god thing.

Time to Recap and Run Some Numbers

Multiple surveys—even one in cahoots with the Latter-day Saints—finds that about a four-in-ten minority of Americans considers religion very important in their lives. A Pew calculation that has gone nearly entirely ignored estimates that half of Americans are nonreligious. This makes sense because Gallup says half of Americans do not belong to a religious institution. Meanwhile, Harris and a technical analysis broadly concur that somewhere in the area of a third are atheists of at least the light, agnostic variety, and a tenth or more have little or no doubt any gods exist.

So let’s convert those figures to absolute numbers. We’re fast approaching 330 million Americans. So well over 30 million should be staunch atheists; around 60 million are atheists who are at least markedly skeptical about deities being real; perhaps 100 million are at least marginally atheistic; over 150 million are nonreligious; and 190 million do not find religion to be of deep importance. (Not being total separate cohorts, there is considerable overlap.)

In comparison, about three million regularly attend those megachurches we all hear about; 60 million go to church on a regular basis; 140 million feel that religion is very important to their lives; 70 million self-identify as Catholic although a lot of them are lapsed to varying degrees; 160 million are at least nominally Protestant, but, again, many also qualify as nonreligious; some 100 million are born-again and/or evangelical; Mormons and Jews are each about six million, Muslims maybe half that.

So even the most atheistic Americans rival in number to Jews and Mormons combined. Atheists overall probably exceed Catholics and broadly match the evangelical born-agains.

How Should We Talk about These Numbers?

Now, we do not want to falsely boost our numbers like the propagandists behind the long-gone Moral Majority. But chronically undercounting ourselves is demographic madness. Numbers count, and leaving the nation-at-large with the impression that atheists remain a scarce breed helps those who oppose us to continue to perpetuate discriminatory tropes and to discount our social and political sway. To normalize American atheism, humanism, and secularism requires presenting our true large numbers. That means presenting the best science-based numbers of nonreligious and atheistic Americans in the most objective manner possible.

To start, as important as the Pew statistics on the unaffiliated Nones have been, they are now obsolescent. Better-quality statistics are on hand, so stop citing that data set ad nauseam! Seriously, don’t do it. Instead, do the following: Cite the later Pew calculation that around half of Americans are nonreligious Nons. Add in the Harris results and the Social Psychology and Personality Science estimates indicating that about one-third of Americans are atheists of at least the agnostic level. When you do this, be prepared to explain what an atheist actually is! If you need to opine on how many are ardent atheists, it would be about one in ten or more. Feel free to compare the absolute numbers and percentages of atheists and other irreligious to those of the major sects, such as how the total number of atheists matches the number of theocons. As for how fast nontheism is growing at the expense of theism, refer to the many survey organizations that are finding that those without religion—atheists included—are ballooning by an extraordinary one in ten Americans every decade or so. Also note that Gallup finds that just half of Americans are now church members, a statistic that is dropping fast. And while the mainline churches have long been dwindling, these days the religious Right is shrinking too, with, for example, the Southern Baptists in serious demographic trouble, and Bible literalism steadily dropping from the Carter years by nearly half to less than a quarter of the population according to Gallup.

It is often good to note that figures on levels of actual belief are inherently imprecise. That’s because surveying opinion levels always involves a major plus-minus factor, so citing exact figures such as “47 percent believe in one thing or another” is technically inaccurate and gives a false impression of certainty. This should be avoided when trying to give the general public an idea of who thinks what. In any case, the numbers are changing so fast regarding matters a/theistic in this nation that any precise figure is out of date before it is published. What can be said is that all indicators look increasingly bright for the future of American rationalism, although when, if ever, the nation may boast an atheist majority is uncertain.



Gregory Paul

Gregory S. Paul is an independent researcher, analyst, and author. His latest book is The Princeton Field Guide to Dinosaurs (Princeton University Press, 2010).

Back in the late 1970s, the religious Right was embarking on social and political activism in a big way that would make it a major force in American life. To promote the effort, in 1979 Jerry Falwell, Paul Weyrich, Richard Viguerie, and others did something as audacious as it was an unethical lie. They founded …

This article is available to subscribers only.
Subscribe now or log in to read this article.