Let’s talk US religion and politics … with charts!

Let’s talk US religion and politics … with charts! November 8, 2015

Stereotypes aren't the way to understand religion and politics in America
Stereotypes aren’t the way to understand religion and politics in America—it’s too complicated for that!

How should we respond to political and religious diversity? A great first step is actually to understand it before we start making judgments about groups other than our own. One of the best ways to do that is through data, especially through careful interpretation of visual presentations of data.

True confessions: I’m a chartgeek, and I hope you are a bit of one, too. That’s one purpose for this Charting Church Leadership blog, to highlight insights into the church as an in-the-world-but-not-of-it community that has implications for politics, the economy, and much more of everyday life. If you think charts aren’t your thing, now’s the time to change your mind!

With thanks to Dr. Tobin Grant for the chart concept, here’s a brand new Tableau Public visualization of the 2007 Pew Religious Landscape Survey, using two forced-choice survey questions to indicate economic and social ideology. Click each box in the gray bar at the top to read through a story about the survey data. You can investigate further at any time by hovering over the colored circles or using the menus at right.


The Pew Religious Landscape Survey

In May 2015, the Pew Research Center on Religion and Public Life released results from their second Religious Landscape Study, based on telephone interviews with over 35,000 Americans. The RLS is one of the most comprehensive social-scientific research studies on the planet. Why? American religion is incredibly diverse, so it’s not enough to survey a few thousand people. If you want to say something about the difference between Missionary Baptists and American Baptists or between Jews and Jehovah’s Witnesses, you need to interview many dozens of members of each small group.

The results are important to America’s self-understanding as a nation and to the American church’s self-understanding as a part of the Kingdom of God.

For years, the hottest topic of conversation has been the increase in “religious nones”, people who say they have no religious affiliation, who rose from 16.1 percent of the US population in 2007 to 22.8 percent in 2014. There’s been less attention to the implications of these shifts for American politics and the future of the church. Is the “rise of the nones” really a substantive force for secularization? Or is it just an adjustment in survey responses from demographic segments who have always been basically irreligious in practice?

Just last week, I enjoyed new, insightful chart work on political differences between religious groups on the Corner of Church and State blog by Dr. Tobin Grant at Southern Illinois University, a colleague in political science and a fellow protege of a great scholar of religion and politics, Lyman “Bud” Kellstedt. Take a look, and don’t miss out on the enlarged view of the chart.

The horizontal dimension is the traditional left-right dimension representing preferences about the size of government and availability of public services. The vertical dimension reflects support for gay marriage, as an indicator of social ideology. The chart updates a previous version based on the 2007 data (enlarged image), where the vertical dimension is a less specific question about government protection of morality. These charts are not for republication, so you’ll have to visit Dr. Grant’s blog to see them.

We’re used to expecting a correlation that puts most people either at lower left (liberal on both dimensions) or upper right (conservative on both dimensions). Political leaders often try to enforce this correlation by demanding consistency. However, consistency between ideological dimensions is not a law of nature, and exceptions abound.

Most religious groups are internally diverse

Another skill in reading these graphs is to remember that focusing on averages (“central tendencies”) too easily tempts us to indulge our love of broad generalizations and stereotypes. For example, one of the most visible features of Dr. Grant’s charts is a large, central circle representing Roman Catholics. In 2014, the average Catholic was only slightly to the left of the average American on both government size and gay marriage. In 2007, the average Catholic was average on government size and only slightly more traditional than average about government protection of morality. But the apparent breadth of the large Catholic circle reflects the large numbers of Catholics, not the actual variation among Catholics. I was left wondering just how much variation there is within Catholicism and the other traditions. How accurate would it be to predict that an individual American Catholic would be a political centrist just based on their religious tradition?

Not very accurate. Lo and behold, Dr. Grant has already done that work on the 2007 data, showing that Catholicism includes subgroups that are distributed all over political space. The new visualization above permits slicing on church attendance, age, race, and more (with more to come through updates over coming days).

We can show the similar internal variation for clergy. For example, the 2009 Clergy and Public Affairs survey conducted by Dr. Corwin Smidt and his partners showed that each of ten Protestant denominations includes pastors whose political and theological views differ widely from their denomination’s average. We can view that variation as heresy—”these pastors don’t believe what they should!”—or we can view it as normalcy: every group has internal differences that have to be acknowledged and managed.

New interactive visualization

To produce the interactive visualization, we need individual-level survey data. Pew hasn’t released the 2014 data yet, so I turned to downloading the 2007 data (it’s free, but completing a form is required).

In the interactive charts for 2007, the horizontal dimension is still the classic left-right debate about the size of government. The vertical dimension is social issues, represented by opinion about government and morality. Religious groups at lower left favor lots of public services and “worry the government is too involved in the issue of morality.” Those at upper right prefer a small government and say “the government should do more to protect morality in society.”

The charts are particularly helpful in illustrating just how unusual Historically Black Protestants are. Historically Black religious denominations are almost alone in the upper left quadrant of these charts, economically liberal but socially conservative. In 2007, they shared this space with Hindus, Muslims, and Jehovah’s Witnesses and, surprisingly, Mormons 25 and under (see the fifth story page, titled “Old and young divide left and right in most traditions”).

Meanwhile, the mainline (some would say “oldline”) denominations live in the lower right quadrant, economically conservative but socially liberal.

Keep in mind that these graphs are just average answers to forced-choice survey questions—there isn’t actually an average person in the middle, only a percentage who went one way or the other (though I have coded the few respondents who said “both at the same time” or “it depends” as “0.5” on the forced choice, so they pull the averages toward the middle a little bit). In a way, every dot that’s not on the outer edge represents potential internal disagreement.

In the 2007 chart, the Assemblies of God appear right of center on the size of government and toward the top (“right”) on government protection of morality. This might seem to mean Assemblies affiliates are “extremely conservative” on government protection of morality. Many are, of course. But the chart itself only means is that most Assemblies survey respondents say “The government should do more to protect morality in society,” while very few “worry that the government is getting too involved in the issue of morality,” quoting the actual survey question.

In other words, the position of a dot on the chart doesn’t necessarily mean that these respondents hold extreme positions favoring specific, strict legislation. It just means that the A of G had little internal variation on that 2007 survey question.

What do you see in the data?

When the full 2014 dataset is released, I’ll update this visualization with wave-to-wave comparisons. Please post your comments and questions below, or comment on the Charting Church Leadership Facebook page.

As promised, the next post discusses a “growth mindset” versus a “fixed mindset” and the implications of mindsets for business, education, and the church.

Browse Our Archives