Is the “God Gap” about to Strike Back?

Two years ago we highlighted the importance of the “new Catholic vote” and the “God Gap” in presidential elections. Americans without a religious affiliation, often referred to as “Nones,” are as important to the success of the Democrats in presidential elections now as Catholics were to Democrats in the early 1960s. With the Nones growing in number this will likely become a stronger advantage for Democrats in the years ahead. 

However, before anyone starts running for president again there is another religious wrinkle ahead in the 2014 midterm elections. In many cases, the Democrats are contesting the 2014 Senate elections as a visitor on the Republican Party’s more religious home field and this may lead to a loss of their majority. The cohort of Senate seats that are up for election in 2014 includes many states where residents say “religion is an important part” of their daily life (“somewhat” or “very” combined). When more than 70% of a state expresses this, more often than not, they vote majority Republican.

The scatterplot below shows the 2012 percentages of adults saying religion is an important part of their daily life (“somewhat” or “very” combined) for each state on the horizontal X-axis (source: Gallup) and the percentage of a state’s voters casting ballots for Republican presidential candidate Mitt Romney in 2012 on the vertical Y-axis.

The data fall into four sections. On the lower left is the strength of the Democrat’s geographic coalition—states where less than 70% of residents say religion is an important part of daily life. This group includes Northeastern states such as Vermont (43% say religion is important), New Hampshire (50%), Maine (50%), Massachusetts (54%), Rhode Island (56%), Connecticut (60%), New York (62%), Pennsylvania (68%), New Jersey (66%), Delaware (66%), Maryland (69%), and the District of Colombia (59%). It also includes Western states such as Washington (56%), Oregon (56%), Hawaii (59%), California (64%), Nevada (59%), and Colorado (61%). A few Midwestern states are in this block as well including Wisconsin (66%), Minnesota (66%), Michigan (67%), Illinois (68%), Iowa (69%), and Ohio (69%). Collectively, these more “secular” states represent 285 Electoral College votes. Only 270 are needed to win the presidency.

The upper right portion of the scatterplot represents the strength of the Republican’s geographic coalition—states where 70% or more of residents say religion is an important part of daily life. This group includes Southern states like Mississippi (88% say religion is important), Alabama (86%), Louisiana (83%), South Carolina (82%), Arkansas (81%), Tennessee (81%), Georgia (80%), North Carolina (80%), Texas (79%), Oklahoma (78%), Kentucky (75%), West Virginia (74%), and Florida (70%; voted majority Democrat in 2012). It also includes some states in the Midwest and Western states like Kansas (74%), South Dakota (73%), Utah (73%), Indiana (72%), Missouri (72%), Nebraska (71%), North Dakota (71%), and New Mexico (70%; voted majority Democrat in 2012). Collectively, these more religious states represent 216 of the 270 Electoral College votes needed to win the presidency.

The two blocks of states described above, are shown in the maps below:

The upper left “outlier” portion of the scatterplot includes a set of Republican states where less than 70% of the population say religion is an important part of your daily life and where a majority voted Republican. These states include more rural locations such as Wyoming (60% say religion is important), Alaska (60%), Montana (53%), Idaho (66%), and Arizona (67%). These collectively represent 24 Electoral College votes.

There is only one lonely outlier in the lower right segment of the scatterplot—Virginia. Here, 73% of the population say religion is an important part of daily life but only 48% voted Republican in 2012. Virginia is currently what I would call “pocketbook blue.” Voters here in many ways resemble the portrait of other Republican states (rural and suburban with higher religiosity). However, Northern Virginia is a rather unique place. Many residents either work for or contract with the Federal Government. Those who don’t, often work for businesses that cater to and/or need the patronage of those working for the government. As long as Republicans have austerity and shrinking the size of government as one of their top priorities they will not win Northern Virginia, which makes it difficult to win all of Virginia.   

Virginia as well as other “borderline” religiosity states like Ohio (69%), Florida (70%), and New Mexico (70%) are the “battlegrounds” sitting at the fulcrum point of the “God Gap” and American electoral politics and this may have shifted a bit since 2012 (Minnesota and Iowa are now at or above the 70% line). Gallup’s religiosity data for 2013 are shown in the figure below in the states where Senate seats are up for election in 2014. The color of each bar represents the current incumbent party. There are nine seats currently held by Democrats in states where 70% or more say religion is important in their daily lives (“somewhat” or “very” combined). One of these is Virginia, which will remain blue (current polling data suggests a very strong Democratic lead). The other eight seats may be in play. Republicans have only two seats to defend below the 70% line and one of these, Wyoming is a safe outlier. The math and the model seem to indicate an opportunity for Republicans to gain a slim majority.

Best When Mixed: Polling and Demography

A recent infographic has stirred up a new discussion about who is “leaving” the Catholic Church in the United States and where are they going. The graphic shows declining numbers of white Christians across generations while showing an increase in young adults without a religious affiliation or “Nones.” Mark Silk, Professor of Religion in Public Life at Trinity College, writing for Religion News Service, noted in a response to the graphic:

“White Christians turning into Nones. That’s pretty much the message of this week’s graphic from the Public Religion Research Institute (PRRI). The news is that white evangelicals, mainliners, and Catholics go from 69 percent of those 65 and older to 25 percent of those 18 to 29. And the “unaffiliated” — i.e. those who say they have no religion — go from 11 percent of the over-65′s to 31 percent of the 18-29′s.”

Professor Silk was certainly not alone in drawing this conclusion and I am not alone in noticing the graphic's potential for creating distortions. PRRI’s graphic simply highlights growing diversity of the U.S. population (i.e., "America the Diverse"). PRRI does polling on religion and politics and a common practice in that endeavor is to break out religious groups by race and ethnicity as there are often distinct political differences between people of the same religion who have different racial and ethnic identities. What the figure lacks is a break out of race and ethnicity for the “Nones” (...which is also understandable given that this sub-group tends to have very similar political attitudes regardless of race and ethnicity).

Are Nones disproportionately white (i.e., the children of Catholics and Protestants)? Not quite. Generational differences in a snapshot poll may not always really reveal a trend one seems to “see” in the data. First of all, a smaller proportion of people self-identify as non-Hispanic white among younger generations, as shown in the General Social Survey (GSS) data below. Much of the population growth in the U.S. in recent years has come among those self-identifying as Hispanic or Latino. This has primarily occurred through immigration and historically higher fertility rates.

Despite a decline in percentages across generations, the U.S. non-Hispanic white population is growing in absolute numbers, just at a slower rate than the Hispanic population. With smaller proportions self-identifying as non-Hispanic white there will most definitely be smaller proportions self-identifying as “white Catholics,” “white Protestants,” etc. in younger generations. This has little to do with shifting religious identities and is not “White Christians turning into Nones” (...which reminds me of another instance when some commentators and reporters misunderstood percentages/graphics and concluded that Christians were "migrating from Europe to Africa, Asia and the Americas").

As the population becomes more racially and ethnically diverse, Nones are actually becoming less likely to self-identify as non-Hispanic white as well (82% in 2002 and 69% in 2012). Currently, one in four Nones self-identifies as either African American of Hispanic/Latino.

The GSS also indicates that Americans who self-identify as Non-Hispanic white aren’t becoming less Catholic across generations either, as shown in the figure below. Between one in four and one in five non-Hispanic white Americans of each age segment self-identify as Catholic (similar proportions, given margin of error). Differences by age group among non-Hispanic white Protestants are more evident...

Even though Non-Hispanic white Catholics make up a smaller proportion of all 18 to 29-year-olds than they do among all of those ages 65 and older in the 2012 GSS (14.9% compared to 15.3%) it is also important to point out there are actually more of these young Catholics in absolute numbers than the older Catholics. Using the GSS data and Census estimates there were approximately 7.3 million non-Hispanic white Catholics ages 18 to 29 in 2012 compared to 6.0 million non-Hispanic white Catholics ages 65 and older. Such is the nature of demography. People die (i.e., "exit the sample") and they tend to be older. The size of the pie (or bars in PRRI's graphic) for those ages 65 and older is simply smaller than the pie for those ages 18 to 29. 

"Conventional wisdom," many religion reporters, and even some researchers repeatedly claim that the Catholic Church is “hemorrhaging” members and we just can't see it because of immigration. In response, we’ve repeatedly pointed out that this hypothetical equation simply does not balance with immigration alone and that there has to be more to that story that goes ignored (1, 2, 3). Yet few ever seem interested in running the population numbers. What percentage of the adult Catholic population in the U.S. is foreign-born? (Pew: 22%, GSS: 26%). What percentage of former adult Catholics in the U.S., those who have left the faith, are foreign-born? (Pew: 18%, GSS 17%). Plug in these numbers (and more) and see how everything adds up. Make comparisons over time and see how different things are now from decades ago. Sometimes it takes a bit of demography and polling to see the clearest portrait of reality. Right on time to test this point again is Gallup’s annual aggregated estimates of religious affiliation in the United States.

How many years of the Church “falling apart” will it take for the Catholic affiliation percentage to decline (beyond margin of error)?... Reality can be so stubborn!

Update 2/27: One final graph below shows the U.S. adult population by religious affiliation and age group...

Mixing image provided by RLHyde.

Ordination All-Stars

The research note is from the Winter edition of The CARA Report. CARA has been releasing research in this quarterly print format for nearly 20 years, since 1995, and has won a number of awards for "general excellence" from the Catholic Press Association during this time. You can get your copy here. 

Following the procedure originally devised by the late Fr. John Klein of the Archdiocese of Chicago, and using data from The Official Catholic Directory, from time to time CARA has reviewed the proportion of ordinations to the diocesan priesthood relative to the latest Catholic population figures for each diocese. Only Latin-rite dioceses that have had at least five ordinations during the reporting period are included.

The tables here present the results of the latest review, for the three ordination years 2010–12 and using year-end Catholic population figures for 2012. The first table lists the top 20 dioceses according to the priest-to-parishioner ratio; the second the top 21 dioceses (because of a tie) based on the largest absolute number of new priests ordained.

CARA has presented the results of the five top-20 comparisons for ordination years 1993-1995, 1997-1999, 2000-2002, 2003-2006, 2007-2009, and now 2010-2012. An analysis of the results of all six data points indicates that of the 57 dioceses that placed in the top-20 for Catholics per ordinand ratio at least once:

• Only the Diocese of Lincoln was in the top-20 all six times
• Three dioceses were listed five times: Bismarck, Fargo, and Wichita
• Three more were on the list four times: Peoria, Sioux Falls, and Tyler
• And 13 were listed three times: Alexandria, Atlanta, Birmingham, Duluth, Knoxville, Mobile, Omaha, Owensboro, Pensacola-Tallahassee, Savannah, Springfield-Cape Girardeau, Tulsa, and Yakima

At the other extreme, for the years 2010-2012, six dioceses with a total of almost 500,000 Catholics had no ordinations, and another 16 dioceses with almost 1,700,000 Catholics had only one new priest.