Nones, No Religious Preference, No Religion and the Misclassification of Latino Religious Identity

Abstract

This article challenges the conventional wisdom about the reported decline of Christianity and Protestantism in the U.S. and the rise of the “nones” among Latinos. It does so by cross-examining the growth of the “nones” (those respondents reportedly having no religion and/or no religious preference) in the U.S. Latino community, which is slated to make up almost 28 percent of the U.S. population by 2060. In 2000, we stumbled, quite by accident, upon a remarkable discovery in the Latino community: that many of the so-called “nones” were, in fact, under cross examination against other religious identity questions, religious, spiritual, and/or believed in God or a higher power. In some cases they were born-again Christians who rejected the label “religion” (and thus reported having “no religion”) as a descriptor of their faith. Many self-identified as and/or attended independent and non-denominational Evangelical and/or Pentecostal/Charismatic churches. To test these initial findings and to try to secure a more accurate reading of the “nones” respondents, our research team added a follow-up question for the “no religion” respondents to the screening questions section so they could explain what they actually meant in the Latino Religions and Politics (LRAP) national surveys in 2012 and 2020. We also cross-analyzed the “no religion” and “no religious preference” respondents against other religious identity questions, such as being born-again and church attendance, and was surprised to find that more than 60% of them reported believing in God or a higher power and/or being Christian, Catholic, or Protestant, religious, spiritual, or something other than having no religion. More surprising and counterintuitively, we found that more than 40% of those reporting “no religious preference” and 17% of those who reported having “no religion” also self-reported being born-again Christians. All of this problematizes the conventional wisdom about the identity and growth of the “nones” in the Latino community and could (though we do not claim or explore this here) problematize our understanding of the “nones” and the changing contours of religion and secularization in American society.

1. Introduction

There has been an explosion of popular and scholarly literature over the past twenty years on the rise of the “nones”, or those Americans that reportedly have abandoned religion and report having no religion on national surveys.1 This work is insightful and has raised important questions about the saliency and creativity (or lack thereof) of religion in contemporary American society.

2. Literature Review

All of this literature has contributed to a largely unchallenged declension narrative, which posits that religion, in general, and Christianity and Protestantism, in particular, are witnessing rapid decline across America. Newsweek sensationalized these developments in its 2009 cover story “The Decline and Fall of Christian America.”2 The struggling periodical seemed to echo the spirit of the “Is God Dead?” TIME magazine cover story from 1966.3 These articles have been joined by a growing cacophony of scholarly books (of mixed quality and partisanship) and clergy in some denominations echoing similar prognostications. Their main claims are that Christianity in America is in a downward spiral and that traditional mainline Protestant and, more recently, Evangelical and Catholic denominations have witnessed steady and in some cases rapid losses over the past twenty years. This decline, they argue, is due in part to the growing politicization of American Christianity by Evangelicals and the rapid secularization and growth of Americans with no religion. This latter theory contends that people do not want to associate any longer with Christianity in general because Evangelical Christianity in particular is writ large for the Christian religion by secular elites and ordinary Americans, and for this reason, moderate and liberal Christians are abandoning the faith in unprecedented numbers. A similar corollary seems to suggest that these defections are due to some Christians being uncomfortable with traditional claims that the Bible is infallible, Jesus is the only way to heaven, marriage should be between one man and one woman, and because of their biblical criticisms of feminism, abortion, and same-sex marriage (Hout and Fisher 2002; Thomson-DeVeaux and Cox 2019).

This is not the first time in American history that scholars and clergy heralded the decline of Christianity. In colonial America, mainline clergy, such as Lyman Beecher, lamented the steep drop off in religious attendance and what they perceived as the cool indifference to (their) religion in general. This lament was picked up by nineteenth-century historians and scholars, such as Robert Baird in 1844 and later by William W. Sweet in 1930 (Sweet 1930), and Sydney Ahlstrom in 1972 (Ahlstrom 1972), who spotlighted declining church attendance and formal religious affiliation within mainline Protestant traditions (e.g., United Church of Christ, Episcopalian Church).

However, this declension theory has since been challenged by a growing number of historians of colonial America, such as Patricia Bonomi, and historical sociologists, such as Roger Finke and Rodney Stark, who, after looking more closely at archival, statistical, church membership, and other records across a broad swath of mainline and non-mainline Protestant and other religious traditions, argued that while there was a real decline in church attendance and affiliation for some denominations, it was not uniform across all denominations and religions. In fact, sectarian Evangelical and other religious traditions were growing rapidly, albeit under the radar of mainstream churches.4

Furthermore, while mainline and more liberal varieties of Christianity were, in fact, declining, the first and second great awakenings in the late 18th and early 19th centuries, along with other new independent and non- and trans-denominational movements, birthed upstart revivalist Protestant Evangelical sects that made converts and drew congregants away from colonial mainline traditions. These new Evangelical revivalists led by George Whitefield, Jonathan Edwards, Gilbert Tennent, Charles Wesley, and later Charles Finney, Francis Asbury, and others created revivalist-based movements that transcended denominational boundaries and were, therefore, hard to track, measure, interpret, and understand since not all required membership or kept records. Unlike mainline traditions, which tended to center their religious practices around rich liturgies, church membership, institutional credal formulations, infant baptism, and an increasingly elastic theological framework, these upstart Evangelicals from both credal and non-credal traditions focused on having a personal born-again conversion experience with Jesus Christ, revivalism, evangelism, youth outreach, evangelistic social work, and living a holy life according to the Bible. Their revivals helped give birth to a revivalist tradition that was carried on through the work of countless other evangelists, such as Dwight L. Moody, Billy Sunday, Billy Graham, and more recently, Luis Palau, T.D. Jakes, Samuel Rodriguez, and Greg Laurie. They brought about massive revivals of religion in general and Christianity in particular that ran roughshod over American religion well into the 21st century.

On top of these new largely Protestant Evangelical traditions, there was an explosion of alternative forms of Christianity in the late eighteenth and nineteenth centuries (e.g., Unitarians, Mormons, Adventists, Jehovah’s Witnesses, Christian Science) that also made major inroads into the U.S. religious marketplace. This led Thomas Jefferson to predict that Unitarianism would one day become the dominant religion of America, something that never came to pass because the Second Great Awakening and other alternative forms of spirituality changed the trajectory of American religiosity (Lambert 2006). Taken together, there was not so much a raw numerical decline in Christianity, but rather a radical shift to upstart and often unruly religious traditions that, in many ways, were theological and ethically conservative but also stylistically and culturally more adaptive, creative, innovative, and entrepreneurial than some previous traditions. Moving beyond just believing, many of these new upstart sects required people to belong, as concretely demonstrated by their level of religious commitment, participation, proselytism, and devotion.

However, these new upstart Evangelical Presbyterian, Methodist, American Baptist, and Disciples of Christ sects created in the wake of the first and second great awakenings, over time became by the mid-20th century mainline Protestants and theologically moderate and liberal (though not always culturally or economically) in the twentieth century. They, like their colonial mainline Protestant predecessors, have struggled since the 1970s to maintain their once dominant status in the American religious marketplace and today have been outpaced (at least in the pews, though not in politics, media, and the arts) by a wide variety of Evangelical and new religious movements. They have been joined not only by new independent and non-denominational churches and experientially-oriented movements, but also by other world religious traditions and a bewildering array of new religious movements that run the scale of human creativity. While these world religions and new other religious movements have contributed richly to religious pluralism, their share of the marketplace has remained relatively modest, with less than six percent of Americans practicing another world religion because the majority of Americans that are religious still self-identify with the Christian tradition (Finke and Stark 2005; PRRI Staff 2021).

Break-away varieties of these new Evangelical traditions within historic mainline Protestant traditions (e.g., Free Methodists, Conservative Baptists, Lutheran, Missouri Synod, the Evangelical Presbyterian and Presbyterian Church in America), along with those that stayed true to their earlier Evangelical heritage (e.g., Southern Baptists) for the most part quietly grew throughout the twentieth century. They were joined by an influx of Pentecostal and Holiness denominations (e.g., Church of God in Christ, Assemblies of God, United Pentecostal Church) and a plethora of newer independent and non-denominational Charismatic churches (e.g., Calvary Chapel, Vineyard), which of which were racial-ethnic and indigenous the Black and Latino communities (e.g., COGIC, PAW, AG, Apostolic). They came out of the shadows in the 1940s, led by Billy Graham and the National Association of Evangelicals (NAE), and by the 1970s, had become a highly decentralized and unruly and yet a somewhat culturally and morally cohesive social and political force to reckon with. Initially led by symbolic leaders, such as Billy Graham from the 1940s to the 1980s, they were then joined by a cacophony of other leaders and organizations with varying degrees of political involvement, all of which helped propel Jimmy Carter to the presidency in 1976, which led Newsweek and other periodicals to call it “The Year of the Evangelical”.

However, Evangelicals come in a wide variety of theological, cultural, racial, ethical, and political traditions. After Jimmy Carter’s progressive policies at home and abroad failed to live up to the hopes, dreams, and political expectations of his grassroots Evangelical base, especially in the South where they historically voted Democrat, other Evangelical leaders rose in Republican ranks to fill the gap and leverage their frustration in American politics. White Evangelicals began to switch parties in large numbers across the nation and helped usher Ronald Reagan into the White House, all of which seemed the stop the cultural hemorrhaging of traditional values unleashed by the countercultural movement of the 1960s and 1970s. This shift contributed to a string of Republican victories from Reagan to George W. Bush to Donald Trump. Black and Latino Evangelicals remained true to the Democratic Party, and also contributed to their victories from Bill Clinton to Barack Obama to Joe Biden. Jerry Falwell’s Moral Majority and Pat Robertson’s Christian Coalition led the vanguard. They were later joined by more mainstream and centrist (on racial and social justice issues, though not on moral issues) Evangelical leaders such as James Dobson’s Focus on the Family, Pastor Rick Warren, T.D. Jakes, Charles E. Blake, Jesse Miranda, Samuel Rodriguez, Paula White, and others in national politics. The resurgence of Evangelical and new religious movements challenged the earlier speculation] about the “Death of God” in America. This led to TIME running cover stories on 26 December 1969, “Is God Coming Back to Life” and on 21 June 1971, “The Jesus Revolution”. Resurgent Evangelicalism in all of its bewildering varieties and racial-ethnicities seemed to demonstrate that God was alive—and all too well—for many stunned scholars and cultural elites.

Despite these past misfires about the decline of Christianity and the Death of God, the new declension theory claims are serious because researchers have better tools to survey and measure religious identity and practice across the nation and traditions. This enables them to ascertain a more accurate reading of the status and health of religious traditions and identity. This research is augmented by other social and cultural indicators across academic disciplines and spheres of society, all of which make the claims about decline more compelling and seemingly incontrovertible. In its wake, a new and emerging narrative has emerged in some circles about the decline and fall of Christian America. This new narrative has taken on a kind of doctrine of inevitability, which is welcome news to some atheists, strict secularists, and people who loathe Evangelical Christianity’s defiant cultural intervention and political influence.

However compelling this new declension theory is at first sight, it has been challenged by scholars who argue that not all religions and forms of Christianity are declining equally. Some are more or less stable, and a few are even growing in their share of the religious marketplace. They point out that church attendance and religious beliefs and practices have, in many circles, remained surprisingly stable and constant over the past decade or two, as have people’s views of the Bible and salvation in Christ alone. In their rush to herald these new seemingly incontrovertible discoveries, the print media and even the scholarly community have at times pushed forward often simple, unrefined, and unqualified narratives and theories about religious declension in America.5

Despite this fact, over the past ten years, the evidence, scholarly basis, and theoretical sophistication of their arguments have grown and seem to have all but put the nail in the coffin of Christian America. In many respects, they are correct. There has been a real decline in religious affiliation across many (though not all) traditions. There also has been a corresponding increase in those individuals (many of which formerly identified as Christian—largely mainline Protestant or liberal or disaffected Catholic) who now say they have no religion, or perhaps more accurately—have no religious preference. However, whether this is also true for all Christians and people with a religious or spiritual identity is a more complicated story.

While there are many factors that have gone into the calculus of this new declension theory, perhaps the single most important is the rise and growth of the “nones”—or those Americans with no religion. Almost every social science indicator that tracks religious identity and affiliation over the past twenty years points to the decline of Christian religious affiliation and a corresponding rise of the “nones”.

However, a growing number of scholars are beginning to problematize, refine, and even challenge some of these findings about the purported massive decline of Christianity and Protestantism and the growth of the “nones”. They start by pointing to serious methodological and theoretical problems with how scholars and journalists measure religious and non-religious identity. This is consistent with our own research and writing on this, which we first noted in 2000 and published in 2006. Some of our findings problematized this declension theory, at least in the Latino community (Espinosa 2006; Johnson and Levin 2022; Levin et al. 2022).

First, until relatively recently, in some surveys, the “nones” were a composite category made up of smaller religious and non-religious identity survey response options and collectively called the “nones”, by which they meant people that did not identify with any specific denomination or religion. Often included in this group were people who reported having “no religion”, “no religious preference”, other religion, something else, independent, non-denominational, and did not know/refused to answer the question. In some cases, “other Christian”, “just Christian”, and “other Protestants” were also included. However, we found that just because a person reports having no religious preference, this does not necessarily mean they do not have a religion. They might see themselves, for example, as broadly Protestant and, therefore, have no problem hopping from a Presbyterian to an Episcopalian church, or among Evangelicals, from a Baptist to an Evangelical Free church, or from a Pentecostal to an independent church. In short, while they might not have one particular denomination they have remained true to their entire life, this does not necessarily mean that they do not attend church or identify with a larger or broader movement (mainline, Evangelical, or Pentecostal/Charismatic, born-again, other, etc.).

Second, until recently many scholars did not cross-analyze these respondents by other religious identity, belief, and practice questions to test if those identifying as a “none” did not, in fact, have any personal religious beliefs under closer inspection and cross analysis. Neither do most surveys ask follow-up questions to these respondents.

Third, until relatively recently, on many past surveys, people who responded as having no religion were treated and/or analyzed as literally having no religion and/or were often lumped together with atheists and agnostics. More refined analyses by Pew Research Center in 2018 and others have called into question this practice of lumping them all together into a “nones” category. Pew Research Center reported that upwards of 44 percent of the “nones” actually reported that religion is important in their daily living and 72 percent of “nones” believe in God or a higher power, which is relatively consistent with the findings discussed later in this study among Latinos (Espinosa 2006, 2014; Lipka 2015; Fahmy 2018). Despite this fact, it has not stopped journalists and even some scholars from continuing to classify respondents who say religion is important in their lives as religious “nones”. This is surprising and begs the question: why? There are many possible reasons for this decision. The most straightforward are that they did not cross-analyze the respondents against other religious identity, belief, and practice questions, they have little training in U.S. religious taxonomy and terminology especially among non-mainline traditions, and/or because it would problematize and punch a hole in the current declension theory, which has become almost an uncontested doctrine in some circles.

Fourth, there are classification and taxonomical problems in past social science research on religious identity in the U.S. and U.S. Latino communities. The main problems are that past scholars were not trained in American religious history, Christianity, theology, terminology, and the Protestant family traditions (mainline, Evangelical, Pentecostal, Independent) and denominationalism (more than 2000 in the U.S.) and the bewildering varieties of racial-ethnic and U.S. religious traditions (over 3000). Religious diversity is also present in the U.S. Latino community, which beyond Catholicism is heavily Pentecostal, Charismatic, and independent and non-denominational. This lack of attention to history, classification, terminology, and denominational and non-denominational nuance has led, in some cases, to the misclassification of survey respondents. The survey vendors who are hired by sociologists to field the surveys and call and classify the respondents in the past were not trained in religious studies and religious classifications, and if they did, they tended to work from a largely dated mainline Protestant classification system in labeling, coding, and analyzing respondents from smaller religious traditions. Even if they allowed respondents to state their own religion, tradition, or church, the list of traditions provided by the sociologists often reflected a simplistic or mainline Protestant framework that did not take into account the post-1970 explosion of Pentecostal and other Evangelical and Charismatic independent and non-denominational churches (especially racial-ethnic) and religious identity and language. For example, many Evangelicals and Pentecostals/Charismatics described themselves—especially if they attended independent and non-denominational churches—as “just Christians”, “believers”, “Spirit-filled” (shorthand for Chrismatic), and “people of faith”, while people from mainline Protestant churches are much more likely to list their actual denomination. Most Evangelicals do not use the label “Evangelical” or “Protestant” to describe themselves, and thus, when this is used on surveys, they do not always respond affirmatively, especially if they are new converts or if their church is non-denominational or Pentecostal/Charismatic. For this reason, many end up getting misclassified. This lack of understanding has led vendors to place respondents that did not specify large, recognizable Protestant traditions into composite categories, which often misclassified individually small, but combined collectively a notable number of respondents. In one study, social scientists classified people from the Church of the Nazarenes as Pentecostals, when in fact, their tradition explicitly repudiates Pentecostalism, even dropping the word “Pentecostal” from their original name, the Pentecostal Church of the Nazarenes. Similarly, some classify all Methodists and Presbyterians with the mainline Protestant tradition, when it is largely the Evangelically oriented Free Methodists and PCA and EPC Presbyterian churches that are growing. In some cases, this has led to some studies arguing that liberal mainline Protestant traditions are growing. Still others listed Mormons and/or Jehovah’s Witnesses with Protestants in their data analyses, neither of which identify as Protestant (Hunt 1999, 2000; Espinosa 2006; PRRI Staff 2021).

Fifth, there are nomenclature and survey language problems with the survey instruments and questions themselves. The most common national survey screening question asks respondents if they practice a “religion”. However, many Evangelicals throughout American history and especially since the explosion of independent and non-denominational Evangelical, Pentecostal, and Charismatic Protestantism in the latter half of the 20th century rejected the label of “religion” as a descriptor of their faith and instead focus on having a personal born-again experience with Jesus Christ. And most others simply say they are “just Christians”, “believers”, or “Spirit-filled”, or “people of faith”. This is true in racial-ethnic and Latino communities, where many are recent converts and know little about the spiritual roots of their pastor and/or church. This was driven home after a former African American doctoral student once told us in a conversation that she did not consider herself “religious”, but rather was a born-again, Spirit-filled woman of faith. Similarly, in churches across the U.S., people say they are not religious and do not believe in “man-made” religions, but rather in having a personal relationship with Jesus Christ. This is why scholars are often puzzled by the relatively high percentage (33–42%) of Americans who say they are born-again, which is much larger than any single denomination.

Sixth, there were also language problems. Many surveys used to analyze religious identity were only conducted in English. This resulted in undercounting the nation’s estimated 45 million legal and 10 million undocumented immigrants. This is a problem because sociologist R. Stephen Warner found in his own research that the majority of religious immigrants arriving in the U.S. are disproportionately Christian per their home country and have higher rates of Christian identity than their U.S.-born Euro-American counterparts. In fact, many arrive predominantly traditional Catholic, Evangelical, and/or Pentecostal/Charismatic, among others. Similar studies by the Pew Research Center and others found that over 83 percent of undocumented immigrants are Christian, which is much higher than the Christian identity national U.S. population (Pew Research Center 2013). Since immigrants generally have higher birthrates than the general U.S. population and because some (e.g., Latin America) do not often assimilate as quickly as European immigrants did in the past because the border is so close, it is possible that they may actually contribute to the growth of more traditional forms of Christianity and other religions in the future, albeit in racial-ethnic forms. Since the racial-ethnic population is projected to skyrocket from 40 percent of the U.S. today to 56 plus percent of the U.S. by 2060, there is good reason to believe that they will contribute to the growth of Christianity in particular and religion in general in the future. The white non-Hispanic population is projected to decline from 60 percent today to 44 percent by 2060 (Vespa et al. 2020). While assimilation may mitigate this immigrant religious growth among some populations, it is unlikely to completely stop or significantly retard it overall, especially among Latinos with the U.S.-Mexico border so close and Latin America constantly replentishing those that assimilate secular values. The rapid racialization of American society and religiosity may result in what Robert P. Jones calls The End of White Christian America (2017), wherein a majority of Christians in the future will be racial-ethnic. This shift in the complexion of Christianity was driven home in the 15 April 2013, TIME cover story: “The Latino Reformation: Inside the new Hispanic Churches Transforming Religion in America”. It spotlighted socially progressive Latino Pentecostal and Baptist leaders leading the reform and transforming the face of American Christianity.

Seventh, the methodological approaches to survey research on religious identity in America have changed significantly over the past twenty years, going from almost exclusively random sample dialing digit telephone calls to surveying people via a mixed medium of telephones and computer-based surveys. The latter, despite claims to the contrary, run the risk of biasing the sample in favor of respondents from higher educational and income brackets, both of which past scholarship has demonstrated are less likely to be religious and undercount racial-ethnic Americans, immigrants, the poor, and rural Americans—all of which tend to be more religious than the general population. Despite controls put in place to counteract these biases, some of these controls are based on the faulty sampling of these groups that affect the control figures themselves. All of these factors run the risk of biasing the results slightly in favor of people who have higher educational and income brackets and who are, therefore, less religious. This is even a greater risk among Latinos and racial-ethnic persons, since people with computers would be more educated and likely to assimilate larger U.S. values, including secular attitudes. While this might accurately capture the views of more highly educated and assimilated Latinos and racial-ethnic persons, it may not accurately capture their general populations and provide an accurate picture nationally. In short, the growth of the “nones” may also be due in part to the change to mixed-medium sampling methods.

Eighth, there is also a foreign language problem. Today, there are over 45 million immigrants in the U.S., half of which are from Latin America, and half of these are from Mexico and the Caribbean. There are literally millions of Americans that do not speak English as their primary language. On top of this, there are another 10 million undocumented immigrants (primarily from Mexico & Latin America), some of whom do not have computers and, in some cases, regular phones. While a growing number of surveys are now bilingual, they often report Spanish response rates lower than previous Latino-framed bilingual surveys, wherein 34–38 percent of their respondents answered in Spanish—approximately the same percentage of the population that is immigrant. Although social scientists try to take this into account when weighting the samples, this also may skew the sample in favor of non-immigrants, which past research has shown are less likely to be religious than immigrants from Latin America, Africa, Korea, the Philippines, and other sending countries.

In addition to these concerns, there are still other concerns, such as fielding the survey on Sundays, which may also affect the response rates of people who attend religious services, which the scholarship has again shown tend to be more a little more religious than the U.S. population. Many data sets used to analyze religious identity are also sometimes fielded for other purposes, such as political and social attitudes, and thus do not pay careful enough attention to the wording on religious identity questions and/or to the day(s) when the survey is fielded, which if performed on a Sunday will affect who is able to respond.

Still, despite these problems, there is still evidence that there is a decline in religious affiliation in general and Christianity and mainline Protestantism in particular, though not all religions and Evangelical and non-denominational/independent forms of Christianity. Rather than fixate on the religious decline or growth and treat it as a zero-sum gain/loss, we will turn our attention to one of the key lynchpins in this new declension theory—the rise of the “nones” and those who report having “no religion” and “no religious preference” among Latinos—in order to see if the “nones” and those who claim to have “no religion” actually mean what they say. Then we will speculate about what this might mean for the future of American religions in general and Christianity and Protestantism in particular. This is something other scholars would have to explore to draw any larger conclusions (and some are beginning to do this), since this study will focus exclusively on Latinos (Levin et al. 2022).

3. Thesis

While we tend to agree with some aspects of the declension theory about declining religious, Christian, and Protestant affiliation for some denominations, we not so sure that this automatically means there is a radical decline in Christian identity and/or belief in God in general. Our social science research over the past 20 years has given us reason to question some of the survey sampling methodologies and larger conclusions about the rise and growth of the “nones” and those who reportedly have “no religion” and “no religious preference”, especially among Latinos. In 2000, while analyzing the data from the $1.3 million The Pew Charitable Trusts-funded Hispanic Churches in American Public Life national survey (n = 2060), we stumbled—quite by accident—upon a surprising discovery in the sociology of U.S. Latino religions: many Latinos in the composite “nones” category (other religion, something else, independent, etc.) and those claiming to have “no religion” and “no religious preference” actually reported in follow-up questions that they believed in God/universal force and/or self-identified as Christian—Catholic or Protestant and/or as born-again Christian. To test these findings and to secure a more accurate read of these respondents, in our follow-up Latino Religions and Politics (LRAP) national surveys in 2012 (n = 1075) and 2020 (n = 1292), we added a follow-up question for the “no religion” respondents so they could explain what they meant. We also cross-analyzed the “no religious preference” respondents against other religious identity questions, such as being born-again and church attendance. We were surprised by the findings.

While this study does not make claims for the religious identity of the general U.S. population, it does argue that many of the so-called “nones” among Latinos are, under close inspection and cross analysis, not only religious and/or spiritual, but in some cases, Christian, Protestant, and Evangelical/Pentecostal/Charismatic and/or born-again Christian. If what we found among Latinos holds true for the larger U.S. population, it should and beginning to cause scholars to pause and possibly refine and revise their forecast about the inevitable decline of religion in general and Christianity and Protestantism in particular in America (Levin et al. 2022).

In short, the LRAP surveys in 2012 and 2020 found that more than two-thirds of Latinos that reported that they had “no religion”, actually did not mean they had no religion, but rather something else when given the opportunity to clarify what they meant. In 2012 and 2020, more than 60 percent of the “no religion” respondents reported believing in God or a higher power and/or being Christian, Catholic, or Protestant, and more than 40 percent of those who said they had “no religious preference” also reported being born-again Christians, or people who had a personal conversion experience with Jesus Christ. This latter finding points to the fluidity, transdenominational, and experientially- rather than denominationally-oriented nature of born-again Christianity. Furthermore, in cross analyses against other survey questions on religious beliefs and practices (e.g., being born-again, Jesus is the only way to heaven, praying daily, attending church almost every week or more), “no religion” and especially “no religious preference” respondents reported suhigh rates of religious belief and practice. While these findings may or may not hold true for the general U.S. population, they should be investigated nonetheless to secure a more accurate read of the religious identity and the growth of the “nones” in American society.

4. Methodology and Findings from Latino National Surveys in 2000, 2012 and 2020

The findings in this study will draw on three national surveys of Latino religions and politics: (1) the Hispanic Churches in American Public Life National Survey 2000 (n = 2060 U.S. sample), (2) the Latino Religions and Politics National Survey 2012 (n = 1075), and (3) the Latino Religions and Politics National Survey 2020 (n = 1292). The limitation of this study and these surveys are that they are only focused on Latino adults 18 years of age and older and who had telephones (2000, 2012, 2020) and/or computers (2020) to complete the survey. The strengths of this study and these findings are that these surveys were specialized Latino-framed surveys created by experts in U.S. Latino religions and society, were bilingual (English and Spanish—with 35–38% answering in Spanish), included indigenous Latino religious denominations, churches, and traditions in the religious identity questions, and used language and terminology specific to the Latino Catholic and Protestant Evangelical and/or Pentecostal/Charismatic communities, instead of mainstream Euro-American mainline Protestant churches and society.

4.1. Latino “Nones” and “No Religion” from 2000 to 2020: Sources, Methods, Analyses

4.1.1. Latino “Nones” in 2000

By 2000, a major debate had erupted over Latino religious affiliation in the United States. Depending upon whom you believed, Latino Catholic religious affiliation ran from as high as 77 percent to as low as 50 percent. These contradictory findings were due not only to the inherent methodological limitations (e.g., English only surveys) of national data sets at the time, such as the General Social Survey (GSS), the National Alcohol Survey (NAS), the National Survey of Religious Identification (NSRI), the American Religious Identification Survey (ARIS), and the Gallup Poll, but also to unique problems that social scientists faced when conducting quantitative research in the U.S. Latino religious community. Clearly, there was a need to rethink how scholars should go about conducting social science research among Latinos.6

To address this gap in the literature and track changing trends in American and U.S. Latino religions and politics, The Pew Charitable Trusts funded the $1.3 million Hispanic Churches in American Public Life (HCAPL) research project. I was hired to serve as project manager and director of research. This three-year (1999–2002) project was housed in the Department of Religion at UC Santa Barbara and was the largest bilingual study in U.S. history at the time on U.S. Latino religions and politics. It included three HCAPL surveys national random sample of all Latinos in the U.S. (n = 2060) and Puerto Rico (n = 250), U.S. Latino religious leaders survey (n = 205), U.S. Latino civic and community leaders survey (n = 229), 17 commissioned scholarly research-based studies on the history of Latino religions and civic activism, 266 community profile interviews in 40 congregations (3 leaders, 3 lay people in each congregation) from 25 religious traditions (e.g., Catholic, Protestant, Evangelical, Pentecostal, Independent, Alternative Christian, Metaphysical) in eight urban and rural Latino communities in the U.S. and Puerto Rico.7 Our research team collaborated with the Tomás Rivera Policy Institute (TRPI) and a nationally recognized team of scholars in religious studies, political science, sociology, and related areas to create the survey instruments, field the surveys, and direct the community profiles and research.8 One of our major goals was to rethink the best approach to capture, identify, and interpret Latino religious identity and its impact on their political and civic views.

The HCAPL national survey was unique because, for the first time to our knowledge, the religious identity questions included Spanish-speaking Latino indigenous denominations and religions on the list of response options given to survey vendors for classification purposes. Many were hitherto overlooked, excluded, or not properly listed, identified, and/or recorded in previous surveys. The HCAPL survey included denominations, traditions, questions, expressions, and language that reflected the terminology used by Spanish-speaking and other Latino Catholics, Protestants, Evangelicals, Pentecostals, Alternative Christians (e.g., Jehovah’s Witnesses, Mormons, Adventists), and metaphysical and healing traditions (e.g., Spiritism, Spiritualism, Santeria, Curanderismo), and not simply translations of white Euro-American mainline Protestant denominations, concepts, terms, and words. Another key distinctive was that we asked respondents in English (62%) and Spanish (38%) on the religious identity question who reported “other religion”, “something else”, “independent”, “non-denominational”, “just Christian”, and similar responses to actually “specify” to the interviewer what church, denomination, religion, or other group or thing they meant. The survey interviewers hand-wrote down the specific churches, denominations, religions, traditions, and/or groups they specified (this information was not in the general data set but added later manually for our data analyses), and we found that about 95 percent of them were identifiable and could be traced to and classified with a specific denomination or religious tradition, almost entirely Christian and largely Protestant and Evangelical/Pentecostal/Charismatic.

We were thus able to reassign these respondents to larger existing denominations, traditions, and/or religions (e.g., Free Methodists with Evangelicals; all denominations of Protestants with Christians; Disciples of Christ with mainline Protestants and Christians; Santeria with Metaphysical; Muslims with Islam and world religions; the Vineyard and Victory Outreach with Evangelical and Pentecostal/Charismatic; Jehovah’s Witness and Mormons with Alternative Christian, non-denominational born-again with Evangelical Protestant, etc.). This enabled us to ascertain the religious identities of a very large number of Latino respondents (20+%) that hitherto would have been classified in a composite category called “nones” or “others”, because they did not identify with a recognizable denomination, religion, tradition, or group until they were asked to specify what they meant (e.g., other Christian—specify; other religion—specify). Because we had such a large base sample (n = 1560) and Protestant oversample (n = 500), we were able to secure a much more precise and accurate reading of the religious identity of respondents then in previous surveys that normally would have been classified as “nones”, especially among Protestants. This not only enabled us to determine how many respondents may have actually had “no religion” or were really “nones”, but it also invariably enabled us to secure a more detailed profile of the Latino Protestant and non-Catholic religious community. The vast majority of these other respondents were clearly Protestant and mostly Evangelical, Pentecostal/Charismatic. Many were also independent and non-denominational, with the vast majority of the latter being Evangelical, Pentecostal or Charismatic in beliefs and practices.9

Another key breakthrough was that many of the respondents in the smaller response options that in the past were collectively called the “nones” (something else, other religion, independent, non-denominational, etc.) when cross-analyzed by other religious identity questions ended up self-identifying not only as religious, but also born-again Christians and attending church at the highest response option on the survey—once per week or more. These findings challenge and problematize the notion that the growth of the “nones”, “no religion”, and “no religious preference” in America accurately reflects the religious identities of respondents and automatically signals a decline in religious, Christian, and Protestant identity. As the graph below illustrates, a very high percentage of HCAPL survey respondents (after we recoded the “specify” answers with an identifiable tradition, if they had one) in cross analyses reported also being born-again Christians. These included “independent/non-denominational” (75 percent), “other Christian” (77 percent), “something else” (52 percent), “other religious tradition or denomination” (52 percent), “other religion” (50 percent), and “do not know/unspecified” (48 percent). They also reported high levels of church attendance (almost every week or more): “independent/non-denominational” (67 percent), “other Christian” (71 percent), “something else” (52 percent), “other religious tradition or denomination” (60 percent), “other religion” (40 percent), and “do not know/unspecified” (48 percent).

Given that being born-again is one of the defining marks of Evangelical/Pentecostal/Charismatic Christianity, and that high church attendance is more often associated with people who are religious and Evangelical Christianity as well, these data may suggest that many Latino Evangelicals and Pentecostals and Charismatics may have been misclassified in other national surveys at the time. This is because they did not cross analyze these respondents against the born-again and other religious identity and practice questions to see if they, in fact, had no religion and thus were actually a religious “none”. These born-again figures are not likely to be exaggerated because self-identifying as a “born-again Christian” has a negative social stigma in U.S. society and especially in the heavily Catholic Latino community, where Protestants, Evangelicals, and born-again Christians are sometimes still derisively called “sectas” (or religious sects or cults). This kind of misclassification may also be a problem for analyses of other national surveys as well. Furthermore, other religious groups, such as Jehovah’s Witnesses, Adventists, and Mormons, may have also been underreported because as we noted earlier they have often been inaccurately classified and lumped together with conservative Protestants, despite the fact that none of them identify with the historic Protestant tradition.10

Latino unwillingness to self-identify with any one denomination, tradition, or religion may in some cases be due more to high rates of mobility, employment transfers, and the search for affordable housing and higher-paying jobs than to secularization, a lack of interest in spirituality, and/or a movement away from religion, spirituality, and/or Christianity. Based on our field research and interactions with practitioners, the most obvious reason why some people self-identified as “independent/non-denominational”, “other Christian”, “something else”, “other religious tradition or denomination”, or “other religion”, and then later specified being born-again and/or attending a Protestant church on Sunday is because Evangelicals stress not being religious, but rather being “saved”, “born-again”, “believers”, “Spirit-filled”, and “just Christians” and because many (especially new converts) are simply unaware of the historical and taxonomical distinctions between a religion, denomination, and church and the religious genealogies of their churches. Furthermore, many freely move back and forth across orthodox denominational lines and see themselves simply as Christians (soy Cristiano—which almost always means Evangelical/Pentecostal/Charismatic Protestant in the Latino community and in Latin America). In fact, these findings and others reveal a high level of religious mobility and switching within the Latino Protestant subculture (Espinosa 2004a, 2004b). Finally, Latino independent and non-denominational churches have been one of the fastest growing segments of the Latino religious marketplace for more than twenty years, and people in these traditions—many of whom are first-generation converts to the tradition—would have very little familiarity with their church’s genealogy and tradition. Still, others may, in fact, practice two or more different religious traditions, although this is unlikely given the tendency of born-again Christians to criticize non-Christian religions and Alternative Christian traditions (e.g., Mormon, Jehovah’s Witness) as unbiblical.

The findings in Figure 1 on religious identification of other religious groupings from the HCAPL national survey challenge the notion that in 2000, movement away from denominational affiliation necessarily signaled an automatic movement away from religion in general and Christianity and Protestantism in particular. In fact, many Latino respondents were not only found to be religious in cross-analyses with other religious identity and practice questions, but also born-again Christians: people often associated with among the highest rates of religious identity, belief, and practice in the United States.

4.1.2. Religious Identity, No Religion, and No Religious Preference in 2012 and 2020

The 2000 HCAPL national survey findings raise a number of important questions about how social scientists identify and measure religious identity in general and the “nones” in particular. Why are we creating composite categories called the “nones” and continuing to place people in this category when in follow-up questions, a sizeable number of them also report believing in God or universal force, being born-again Christians, and/or self-identifying as Christian, Catholic, Protestant, and/or with some other identifiable religious group? Is the term “nones” the best way to describe people in this composite group? If a closer analysis of the “nones” revealed that many were not only religious, but born-again Christians, is it also possible that people who said they have “no religion” or “no religious preference” could also be more religious than scholars have hitherto realized? Is it possible that many people in these categories are actually born-again Evangelical and/or Pentecostal/Charismatic Christians and other religious people and have been misclassified in past and current studies? If this is true among Latinos, could it also be true for the larger U.S. population? If so, even if to a lesser degree, how might this affect to the new declension theory about the decline of religion in general and Christianity and Protestantism in particular among Latinos and in America more broadly?

4.1.3. Latino Religious Identity, Nones, and No Religion in 2012

While it was impossible to answer all of these questions, in our follow-up Latino Religions and Politics (LRAP) survey in 2012, we were curious and wanted to know if “no religion” and “no religious preference” respondents were also possibly being misclassified and were in fact religious. Before we go into an analysis of this problem, it might be good to briefly sketch the methodology and key religious identity findings.

The 2012 LRAP survey was a nationally representative bilingual telephone survey of 1075 Latino Christian likely voters and was fielded 4–10 October 2012. The margin of error for 1075 respondents is ±3.0% at a 95% confidence level. More than 94.2 percent of all Latino likely voters are Christian, and Latino Catholics and Protestants made up 98 percent of Christian likely voters in this sample, with 2 percent self-identifying as Jehovah’s Witness and Mormon (Espinosa 2012, 2014). The screening questions included a religious identity question and a follow-up question for those that reported having no religion, and thus included Christian and non-religious respondents.

Regarding the religious identity of Latino Christian likely voters in 2012, the LRAP survey found that Catholics made up 67 percent, Protestants and Other Christians (almost all Protestant) 32 percent, and others 1 percent. Born-again Christians made up 39 percent, and born-again and Evangelical Christians made up 88 percent of all Latino Christian Protestant likely voters. With respect to religious beliefs and practices, 82 percent of Latino Christian likely voters reported that Jesus is the only way to heaven, 78 percent reported praying every day, and 58 percent reported attending church or religious services almost every week or more. Religion was important, with 75 percent reporting that religion provides a great deal (53%) or quite a bit (22%) of guidance in their day-to-day living, and 66 percent reported that a political candidate’s faith and morals were very relevant (42%) and somewhat relevant (24%) to their decision to vote for him or her.

There could be no way to know for sure about the “no religion” and “no religious preference” respondents without asking them a follow-up question and running them against other religious identity, belief, and practice variables. For this reason, in 2012, we asked a follow-up question to the “no religion” and “no religious preference” respondents in the screening question section of the survey about what they meant by these response options. We also cross-analyzed them against other religious identity, belief, and practice variables on the survey, such as being born-again, believing Jesus is the only way to heaven, praying almost once per week or more, and attending church almost once per week or more. What we found was surprising and counterintuitive.

In the 2012 LRAP survey, we asked the no religion and no religious preference respondents the following question below. We provided seven response options and even biased the response in favor of “no religion” by listing it first since respondents tend to select early response options. This was also used for the same question in our 2020 LRAP survey.

Question: When you say that you have “No Religious Preference” or “No Religion”, which of the following statements reflect what you really mean?

I literally have no personal religious belief or faith in God or in organized religion.

TERMINATE 1

I am a Catholic, a Protestant, or other Christian and personally believe in God but do not practice my religion at this time due to job transfers, moving, or other reasons.

TERMINATE 2

I am a Christian, a Catholic, or a Protestant and personally believe in God but I do not have any one particular religious tradition or church that I prefer over another at this time.

TERMINATE 3

I believe in God, but I have no desire to know God personally or practice any religion at this time.

TERMINATE 4

I am religious or spiritual and believe in God or a higher power but I do not have any one particular religious tradition that I prefer over another at this time.

TERMINATE 5

I meant something else.

TERMINATE 6

Don’t Know/Refused to Answer (DK/RF).

TERMINATE …. 9

The 2012 Latino Religions and Politics (LRAP) national survey sampled 1075 Latino likely voters, which tend to have higher rates of assimilation, education, income, and lower rates of religious and Christian identity than the general U.S. Latino population. Not knowing what to expect, we were shocked to discover that only 24 percent of those Latinos that reported “no religion” or “no religious preference” said, “I literally have no personal religious belief or faith in God or in organized religion”. Despite Latino likely voters being less religious than the general Latino population, as shown in Figure 2 over 70 percent reported they believe in God, a higher power, and/or were spiritual, Christian, Catholic, or Protestant. Six percent reported they did not know or refused to answer the question. The fact that 70 percent of the “no religion” and “no religious preference” respondents reported some kind of religious identity was surprising. If this is true for Latinos, what about Blacks, Asians, and even the larger white population? Given that Latino registered likely voters are more likely than the general U.S. Latino population to be assimilated and have lower rates of religious affiliation, we speculate that the actual percentage of Latinos that would report having “no religion” or “no religious preference” would be closer to 21–23 percent. If three-fourths of the bedrock of the “nones” sample report believing in a God, a higher power, and/or being spiritual and religious, what are the larger theoretical and interpretive implications for the rise of the “nones” and purported decline of religion and Christianity among Latinos in particular and in America in general? When “no religion”, “no religious preference”, atheist and agnostic are all combined, 36 percent of this combined group reported that Jesus is the only way to heaven, and 18 percent reported attending church more than weekly. Still, others also attended church, just not weekly.

How might we explain why a person could believe in God and/or self-identify as Christian, but not practice or prefer a particular tradition or spirituality? There are many reasons, including some just do not want to practice or prefer one tradition over another, others see no problem attending many different churches within a particular tradition (e.g., mainline, Evangelical, Pentecostal/Charismatic traditions), some are too busy to attend church and practice their faith because they have to hold down multiple jobs, many are forced to often move in search of employment and, therefore, find it hard to plant roots in just one church or denomination, and because many attend churches of their spouses or partners. There is also a long-standing tradition of practicing popular religiosity outside the confines of the institutional churches, especially given their Catholic backgrounds where home altars and other forms of popular spirituality and devotion to the Our Lady of Guadalupe/Virgin Mary, saints, and popular religiosity take the place of attending institutional churches—which in some parts of Latin America are hard to reach and/or do not have resident priests.

The three most important reasons are more easily overlooked by social scientists unless they carry out ethnographic, interviews, and community profile research. First, for over twenty years, Christians across traditions are increasingly rejecting the label of “religion” and instead are reporting that they are spiritual and/or have a personal born-again conversion experience and relationship with Jesus Christ, and not what they perceive as inflexible humanmade denominations and traditions. Second, born-again, Pentecostal/Charismatic, and independent and non-denominational Christianity facilitates easy movement across Evangelical/Pentecostal/Charismatic traditions. Third, over this same time period, the American religious marketplace has witnessed the rapid growth of independent and non-denominational Evangelical, Pentecostal, and Charismatic forms of Christianity, with little ties to traditional denominations and institutional structures. Many report being “just Christian” (soy Cristiano). This lack of knowledge about their Protestant and denominational genealogical identity is further reinforced by the fact that many are first-generation converts from Catholicism to Protestantism, especially in independent and non-denominational churches. For example, confirming similar findings in the HCAPL 2000 national survey about “just Christian” and “other religious tradition” respondents, the 2012 LRAP survey found that 77 percent of “just Christian” and 36 percent of “other religious tradition/denomination” respondents reported being born-again Christian, 73 and 36 percent reported being Evangelical, and 58 and 26 percent reported being Pentecostal or Charismatic, respectively. All combined, 52 percent of “just Christians and other religious tradition/denomination” reported praying every day, and 61 percent believe Jesus is the only way to heaven. These response rates are similar to large Latino Evangelical denominations, such as the Assemblies of God, where 68 percent reported being Protestant, 22 percent as “just Christian”, and 10 percent “other religious tradition”.11

4.1.4. Latino Religious Identity, Nones, and No Religion in 2020

Various studies from 2012 to 2020 noticed major shifts in religious affiliation across the nation and the continuing rise of the “nones” and those that have “no religion” and “no religious preference”. In order to see if this held true among Latinos, we analyzed those that said they had “no religion” and/or “no religious preference” in the Latino Religions and Politics National Survey in 2020.

The Latino Religions and Politics National Survey was conducted from 8 to 22 September 2020. The results are based on a bilingual Spanish/English 44-question survey instrument administered nationwide to 1292 U.S. Latino Christian registered voters who are either Catholic or Protestant/Other Christian. These adults were all 18 years of age or older, Latino, registered to vote, and declared that their religious preference was either Catholic or Protestant/Other Christian, which together make up 85 percent of the U.S. Latino electorate and 95.5% of all religious Latino registered voters. The results in this study are based on interviews conducted via landline, cell phone, or online. The Latino national registered voter sample (source L2 Voter Mapping) was pulled at random using an nth selected probability methodology to properly represent the Latino registered voter population across the country and within states. For 1292 completed interviews, the margin of sampling error is plus or minus 2.73 percentage points at the 95% confidence level, and the 1528 interviews for the first seven questions (including the no religion and no religious preference questions and follow-up question) had an MoE of plus or minus 2.51% at a 95% confidence level, for the results based on all Latino registered voters nationwide. The margin of error for the subset of 823 Catholic voters is plus or minus 3.42 percentage points at the 95% confidence level. The margin of error for the subset of 453 Protestant and other Christian voters is plus or minus 4.60 percentage points at the 95% confidence level. The margin of sampling error is higher and varies for results based on demographic sub-groups.

The HCAPL and LRAP surveys have an excellent track record over the past 20 years (2000, 2008, 2012, 2020) to be an accurate reflection of the current voter preference among Latino voters at the time the poll was conducted. The LRAP survey is a study of self-described Latino Catholics and Protestants/Other Christians, which make up 85 percent of all U.S. Latino voters. Every four years, the LRAP survey has always been very close (normally within 2 points) to the final Presidential election result among Latino voters. Internal testing of non-religious Latino voters during the September 2020 fielding confirmed, as in previous election years, that there were not any significant statistical differences between sampling self-described Latino Christian voters and all U.S. Latino voters in the election results and the difference is 0–2 percentage points (Espinosa 2020).

There were eight screening questions that were asked to all respondents (Christian or non-Christian) about age, language preference, voter registration, Hispanic/Latino identity, country of origin, religious identity, no religion and no religious preference follow-up question, church attendance, and born-again identity. The findings below discuss the no religion and no religious preference respondent results.

Regarding the religious identity of Latino Christian registered voters, the LRAP survey found that Catholics made up 64 percent, Protestants and Other Christians (almost all Protestant) 34 percent, and Mormons and Jehovah’s Witnesses almost 2 percent. The screening sample revealed that nationwide Catholics made up 54 percent of the U.S. Latino population, Protestants and Other Christians (almost all Protestant) 31 percent, no religion 6 percent, no religious preference 3 percent, agnostic 3 percent, atheist 2 percent, and other world religion 1 percent.

Born-again Christians made up 35 percent of the U.S. Latino population, 42 percent of Latino Christian registered voters, 64 percent of all Protestant/Other Christian registered voters, 72 percent of independent/non-denominational Protestant registered voters, 42 percent of mainline Protestant registered voters, 70 percent of Other Christian/Other Religion registered voters, and 43 percent of do not know/refused to answer registered voters. Nationwide, 74 percent of Latino Christian registered voters reported believing Jesus is the only way to heaven, 66 percent pray daily, 50 percent attend church or religious services almost every week or more, 37 percent read the Bible at least once a week or more, and 60 percent pray for divine healing at least once a week or more. Religion is clearly important, with 87 percent reporting that religion provides “quite a bit” or “some” guidance in their day-to-day living, and 66 percent reporting that a political candidate’s personal faith and morals are important in their decision to vote for him or her.

With respect to church engagement in public life and church–state issues, 38 percent agree that religious leaders should try to influence public affairs, 67 percent support prayer in school or a moment of silence, 79 percent support keeping “under God” in the Pledge of Allegiance and “In God We Trust” on currency, 68 percent report that America’s Judeo-Christian history, values, and heritage are worth preserving, and 79 percent support the government providing legal protection for religious liberty and freedom of conscience. Religious identity was strongly correlated with being undecided, independent, and female voters. Born-again Christians made up 64 percent of all Latino Protestant registered voters, 46 percent of Latino Christian undecided voters, and 41 percent of independent voters.

Like in 2000 and again in 2012, in 2020 Latino “just Christian” and “other religious tradition” respondents are overwhelmingly born-again Christians. A full 66 percent of “just Christian respondents” report being born-again, 64 percent attend church or religious services every week or more, 94 percent believe Jesus is the only way to heaven, 91 percent pray once a week or more, 77 percent believe religion provides a great deal of guidance in their day-to-day living, 84 percent favor legal protections for religious liberty and freedom of conscience, and 73 percent believe America’s Judeo-Christian values are worth preserving.

In 2020, in the LRAP national survey of 1292 Latino registered Christian voters, 6 percent said they had “no religion” and 3 percent said they had “no religious preference”. Those that reported having “no religion” and “no religious preference” were asked the same follow-up question respondents were asked in 2012: “When you say that you have “no religious”, “preference” or “no religion”, which of the following statements reflect what you really mean?” As shown in Figure 3, we were surprised to find that only 32 percent of the “no religion” respondents affirmed the first response option: “I literally have no personal religious belief or faith in God or in organized religion”. The rest indicated one of the other six response options. No one said they meant something else. In short, 62 percent of the no religion respondents self-identified as religious, spiritual, and/or believing in God when given a chance to clarify what they meant. When analyzed by religious identity and practice questions, 13 percent of all no religion respondents reported attending church once or twice a month, and 17 percent of those that said they had “no religion” considered themselves born-again Christians.

Among those that reported they had “no religious preference”, 22 percent attended church almost every week and 28 percent once or twice a month, and a surprising 41 percent reported being born-again Christians. Broken down into greater detail, we found that 10 percent of “no religion” and a surprisingly high 41 percent of “no religious preference” respondents reported they were born-again or Evangelical Christians, that is, someone who has had a personal conversion experience with Jesus Christ. Finally, none (0 percent) of the “no religion” and “no religious preference” respondents said they were atheists or agnostic. This finding is important, because it underscores the need to separate out “no religion” and “no religious preference” respondents from atheists and agnostics in “nones” groupings, since it is very misleading and leads to the impression that respondents in these categories are atheists or agnostics especially when these latter two categories lead the name of this composite group, when this may not be the case.

The LRAP “no religion” and “no religious preference” findings do nonetheless indicate that there was decline among respondents reporting Christian identity between 2012 (28%) and 2020 (17%). This provides evidence of a decline in Christian identification among this group, though not to the extent noted in most of the literature wherein all respondents in this category would have been assumed to have no religion. Interestingly enough, the decline in Christian identity shifted by 14 percent from 2012 (20%) to 2020 (34%) for those self-identifying as Christian and to those identifying as “religious or spiritual and believe in God or a higher power but I don’t have any one particular religious tradition that I prefer over another at this time”. All of this seems to reinforce a movement towards a spiritual but not institutionally religious sensibility for these “no religion” and “no religious preference” respondents. Thus rather than provide evidence of a decline in religion and the growth of secularization per se, they instead may provide evidence for a decline in formal institutional and credal religiosity among this group—often seen as the backbone of the “nones”.

5. Future National Survey Methodology Recommendations

To secure a more accurate portrayal of the U.S. and Latino religious communities, after identifying the research study’s problem(s), framework, theories, and board of advisors, we recommend the following steps to all survey directors, staff, and vendors.

First, secure broad training in the history and contemporary trends and developments of U.S. religions, paying special attention to new religious, immigrant, and Protestant, Evangelical, Pentecostal/Charismatic, and independent and non-denominational groups and movements, because these are the most diverse religious groups and the respondents that are most often misclassified. This is especially true in the Latino community. One should pay special attention to the historical sociology of American religions in order to understand its changing contours over the past two centuries (e.g., Finke and Stark 2005; Albanese 2012; Bonomi 2003).

Second, draw up a culturally, racial-ethnically, and religiously sensitive taxonomy and classification systems that take into account specific indigenous and immigrant traditions in the populations surveyed (e.g., Espinosa 2006). They should train all survey vendors in these taxonomies and classifications to avoid misclassification and allow for multiple, combinative, transnational and/or regional identities, especially within family groupings (e.g., born-again & Methodist, Pentecostal/Charismatic & independent).

Third, allow religious identity respondents in more generic identity categories (e.g., “no religion”, “no religious preference”, none, agnostic, just Christian, other religious tradition, other world religion, metaphysical) to specify what they mean via smaller churches or minority traditions in follow-up questions (e.g., Other religious tradition—Specify___). These can be later reclassified in many cases to provide a more accurate set of findings.

Fourth, ask all no religion, no religious preference, none, and agnostic respondents a follow-up question that allows them to clarify what they mean by the term. As we found in this study, over 60 percent of “no religion” and “no religious preference” respondents clarified in a follow-up question they meant something else other than having no religion.

Fifth, make all surveys bilingual or perhaps multilingual depending on the target communities and make sure the translations are based on the religious terminology in these communities (Espinosa 2006).

Sixth, realize that all English and computer only- & majority samples will probably bias the findings (even with controls & weighting—especially if outdated) in favor of more assimilated, upper-middle class, college-educated, and thus less religious respondents.

Seventh, make sure the demographic and religious controls and weights that are used to weigh and interpret the data are themselves based on more sophisticated sampling methodologies so they are not simply replicating biased methods that reinforce faulty or outdated weighting, controls, taxonomies, and classifications.

Eighth, keep in mind that most independent and non-denominational respondents are Evangelical and the vast majority of them are—at least in the U.S. Latino and Latin American communities—Pentecostal/Charismatic in heritage and theological outlook (e.g., openness to practice tongues & spiritual gifts, charismatic worship styles, healing).

Ninth, broaden the language of the initial religious identity screening question to use terminology that people within the largest religious traditions use as self-descriptors of their religious identity (e.g., Regardless of whether or not you are now actively attending religious services, do you consider yourself religious, spiritual, a person of faith, and/or spiritually open—Specify___?).

Tenth, add a Protestant/Other Christian oversample for cross analyses and make sure the survey is not fielded on days and times when people attend religious services. These oversamples can also be created for other groups for cross analyses.

Finally, avoid clustering religious and non-religious “nones” into a single combinative category, since this is misleading, especially if these “nones” also self-identify as highly religious (e.g., born-again Christian, pray or attend church weekly or more, believe Jesus is the only way to heaven). Avoid lumping all “nones” with atheist and agnostic, since it gives the false impression that they aren’t religious when many are.

6. Findings and Conclusions

The above findings offer a number of important insights that challenge and revise past social science research on the Latino “nones” and those who report having “no religion” and “no religious preference”. First, not all Latinos classified as “nones” actually have no religion. Second, many of those with “no religious preference” self-identify as born-again Christians and attend church once a week or more. Third, from 2012 to 2020, the vast majority of those Latinos that reported having “no religion” and “no religious preference” indicated that they did not mean they had no religion in a follow-up question. While there was decline from 2012 to 2020, still almost two-thirds of them affirmed some kind of openness to religion or spirituality. Fourth, most people who said they have “no religion” and “no religious preference” actually said they believed in God or a higher power, were Christian, or were religious or spiritual. This is consistent with past Pew Research Center studies, which found that almost 44 percent of those that said they were religiously nothing in particular also reported that religion was important to them and 72 percent of “nones” believe in God or a higher power (Lipka 2015; Fahmy 2018). Fifth, the distinction between “no religion” and “no religious preference” is statistically significant, and the two should not be automatically lumped together in data analyses. Sixth, if the respondents in the “nones”, “no religion”, and “no religious preference” categories in our study are reassigned and classified to the various Christian traditions based on self-reporting being born-again and/or other Christian religious identity indicators when asked to specify, then the percentage of Latinos that are Christian will increase and the number of those respondents that truly have “no religion” and that are “nones” decreases. Seventh, even if only some of this holds true for the larger U.S. population, then claims about the current and future decline and fall of Christianity and Protestantism in America—like past declension theories—may prove premature. This is because some respondents are simply being misclassified and scholars have not fully taken into adequate account and methodological sophistication Latino and other racial-ethnic minority faiths and immigration growth rates and identities in their analyses. The high immigration rates, birth rates, and high religious conversion rates of Latinos and other immigrants and racial-ethnic minorities may collectively challenge the declension theory and mitigate the seemingly inevitable decline of Christianity among Latinos and in America more generally and may perhaps even foreshadow its religious resurgence, albeit in de-Europeanized forms. Together, they may contribute to what R. Stephen Warner has described as the de-Europeanization of American Christianity and what Robert P. Jones has called the end of white Christian America, but not necessarily the end of Christianity or Protestantism in America, especially in its Evangelical and Pentecostal/Charismatic independent varieties.12