How many people are members of any faith in the United States? Which religions are growing and which are in decline? Researchers at the Pew Research Center recently released a sweeping analysis addressing these questions.
“The Christian share of the U.S. population is declining, while the number of U.S. adults who do not identify with any organized religion is growing, according to an extensive new survey,” according to the study, entitled “America’s Changing Religious Landscape.”
Pew estimates that there were 51 million adult Catholics in the U.S. in 2014, making up 20.8 percent of all adults. By comparison, in a similar study in 2007, Pew estimated that there were 54.3 million Catholic adults, representing 23.9 percent of those age 18 and older. After taking margin of error into account, Pew estimates, at a minimum, that the adult membership of the Catholic Church declined by 1 million between 2007 and 2014.
How sure can we be that this shift actually occurred? After reviewing other available data sources, not very sure.
Understanding changes in religious populations has always been challenging in the United States. In 1957, the U.S. Census Bureau asked about religious affiliation in the Current Population Survey (CPS) and estimated there to be 30.1 million Catholics in the nation ages 14 years and older. Since 1976, by law, the Census Bureau cannot ask Americans about their religious affiliation. Therefore, unlike many other countries the U.S. does not have anything like a “religious census,” and 1957 is our last reliable government estimate.
Of course, religions make their own efforts to count their members. Worship attendance counts, registration rolls, baptisms and deaths can be utilized. Yet these data miss those who self-identify with a religion but who do not have a very active faith life within their religious community. The Catholic Church in the United States and the Vatican both keep an annual tally on membership. For example, in 1957 the Church estimated there were 36 million Catholics of all ages in the United States. This is fairly consistent with the census estimate for that year, which did not include those ages 13 and younger.
How many Catholics did the Church think there were in the United States in 2014? The official Church number is 68.1 million adults and children. In 2007, the Church estimated there were 64.1 million of members of the Faith. Yet we know both of these estimates miss many who self-identified as Catholic but who were not parish-affiliated and practicing at that time and thus went “uncounted.” The Church’s estimates also do not concur with Pew’s claims of any decline.
Beyond Church counts (and without a census), the best way to estimate the size of the adult Catholic population is to utilize aggregated polling. Researchers take the average across a number of polls, smoothing out variations related to margins of error and the individual practices of polling groups. The Center for Applied Research in the Apostolate (CARA) recently completed such a meta-analysis of 275 national surveys conducted by the Gallup Organization, the Public Religion Research Institute (PRRI), the Pew Research Center and the General Social Survey (GSS). These groups span academic polling, commercial polling and private research organizations.
Since 2010, on average, 23.2 percent of U.S. adults have self-identified as Catholic in these surveys. The lowest reading in these surveys was 18.3 percent and the highest 27.6 percent. Most observations fall between 21 percent and 26 percent. The annual averages in these polls go from 23.2 percentin 2010 to 23.2 percent in 2011, 22.9 percent in 2012, 23.5 percent in 2013 and 23.1 percent in 2014. This is about as steady a pattern as one could imagine, indicating little if any change in Catholic affiliation among U.S. adults. Only Pew’s surveys have a declining trendline.
Shortly after Pew’s Religious Landscape study completed interviewing in 2014, the General Social Survey also completed its data collection. Unlike Pew, they estimate that 25.4 percent of U.S. adults self-identified as Catholic in that year. More importantly, the General Social Survey series, which began in 1972, does not show any significant prevailing trend up or down in the share of the adult population that is Catholic at any point. A similar stability is evident in Gallup’s data all the way back to 1948.
The census estimates that the adult population of the United States in 2014 was 244,563,362. If Pew is correct, that means there were 50.9 million adult Catholics in 2014. If the GSS estimate is to believed there were 62.1 million adult Catholics. Reality is likely somewhere in between. If one uses CARA’s poll-aggregated estimate of 23.1 percent affiliation, this would mean there were 56.5 million adult Catholics in 2014.
|Rise of the 'Nones'
According to the Pew Research Center’s “America’s Changing Religious Landscape,” more U.S. adults are identifying as atheist or agnostic, or saying their religious affiliation is “nothing in particular.” In 2007, an estimated 16 percent of Americans identified as “nones.” In the 2014 survey, the number rose to 23 percent. Here is the percentage of “nones” for each adult generation:
Younger millennials (born 1990-96) 36%
Older millennials (1981-89) 34%
Generation X (1965-80) 23%
Baby boomers (1946-64) 17%
Silent generation (1928-45) 11%
Predicting the size of this population in any given year is difficult enough with many surveys. Comparing two polls for two years and discerning a decline or increase is nearly impossible. By definition, one needs multiple surveys over time to reliably identify a trend. Pew’s use of only two survey-based data points to estimate change is a bit premature. Even if one interviews 35,000 people, issues related margins of sampling error, confidence intervals and error components unrelated to chance can still lead to estimates that are off the actual population mark.
For example, in Pew’s 2014 study, only 48 percent of Hispanic or Latino respondents self-identified as Catholic. CARA’s meta-analysis of surveys since 2010 indicate this outcome is strongly related to surveys having lower than average estimates of the total size of the Catholic population. On average, surveys estimate a majority of Hispanic or Latino respondents self-identify as Catholic. It is well known that surveys that do not reach representative samples of Hispanics or Latinos, especially those who speak Spanish, can lead to lower observations of aggregate Catholic affiliation. Those who survey in the English language only regularly estimate the adult Catholic affiliation percentage to be 17 or 18 percent.
It is possible that the Catholic population in the United States declined between 2007 and 2014, but this is very unlikely. It is more likely that the Catholic affiliation percentage will begin to fall in the next decade. CARA has noticed that the numbers of infant baptisms in the Church have been falling below 23 percent of all live births for several years. Once these younger cohorts become adults, the Catholic affiliation percentage will be affected. The Catholic Church is also certainly not immune to the rise of the “nones,” those without a religious affiliation who are well-documented in Pew’s study (see sidebar). According to the GSS, 3 in 10 “nones” today were raised Catholic. Absent a religious revival of some sort, or a shift in generational patterns of religious affiliation among millennials, the Catholic population may begin to consistently fall below 20 percent in the next decade.
Yet, even if this occurs, membership in the Catholic Church is expected to continue to grow in the aggregate — just more slowly than the overall population. CARA estimates the Catholic population — adults and children — will likely reach 95 million by 2050. One of the big unknowns in this equation is immigration. Forty-five percent of foreign-born individuals in the United States in 2014 self-identified as Catholic (58 percent of foreign-born adults were raised Catholic).
A really big survey, such as the one Pew conducted, is helpful, but it is still only one poll. In a sea of other studies that disagree, the safer bet is still to go with the aggregate. This implies no recent decline in the Catholic population.
Mark M. Gray is a senior research associate for the Center of Applied Research in the Apostolate.