Faithland
This post may contain affiliate links. As an Amazon Associate, we earn from qualifying purchases.
Religious adherents as a percentage of all residents by U.S. county mapped (2010). Data is taken from the Association of Religious Data Archives. (ARDA). The Association of Religion Data Archives is a free source of online information related to American religion.
Maps were created using IDW (Inverse distance weighted) technique in ArcGIS.

Animated Map: Faithland (1890 – 2010)
Alex Egoshin, www.vividmaps.com
Related posts:
– The population density of religious adherents (1890 – 2010)
– Largest participating religious group (1906 – 2010)
– Second largest participating religious group (1906 – 2010)
– Black Protestant Population
– Evangelical Protestant population (1936 – 2010)
– Mainline Protestant population (1936 – 2010)
– Orthodox Christians as a Percentage of all Residents (1952 – 2010)
– Population penetration, Baha’i groups in the United States
– Population penetration, Hindu groups in the United States
– The percentage of Jews in the U.S. (1980 – 2010)
– The percentage of Muslims in the U.S. (1980 – 2010)
– The percentage of Catholics in the U.S. (1890 – 2010)
– Religious adherence as a percentage of all residents (1890 – 2010)
– Total religious adherents by U.S. county (1890 – 2010)
– Religious adherence as a percentage of all residents (1890 – 2010)
– Number of congregations in U.S. county (1890 – 2010)
– A number of adherents per congregation by U.S. county (1890 – 2010)
Your 96-100 color (white) looks exactly like your 46-49 color (white). Not very helpful.
Fixed.
Doesn’t look fixed when seen against the background color. They both still look “white” compared with their nearest neighbors.
Maybe put the 40-49 color at the top and move the whole stack down one spot?
How are you defining “religious” in this context?
“Data is taken from the Association of Religious Data Archives”
Whatever that means.
It’s spelled “Pittsburgh”
I think your data source is really suspect. I can see a couple places that have verifiably higher percentages than indicated. Particularly so in the abyss around Ohio, Kentucky and West Virginia.
Can you link to the actual source? I’d love to explore these numbers a bit more.
Data: The Association of Religion Data Archives
http://www.thearda.com/
Population density of religious adherents here: https://vividmaps.com/population-density-of-adherents/
I used data from here: http://www.thearda.com/
Strange the most religious areas have the same color as the moderately religious: white.
Not true, I’m from southeast Iowa which is vast majority white and it’s “underwater”. I’ve also lived in Chicago, Austin, TX and overseas through my work… surprised to see Cook county included. I would have to agree with you to an extent, it has to be the conservative white suburbanites there or the blacks who still believe in the god of slavery.
Dude, read his comment again. He’s talking about the color values, not skin color.
your fort wayne island is more like “coldwater mountain and also ohiosucksland” i can tell by some of the pixels and actually knowing where fort wayne is lol.
Surprised by the big blue spot covering West Virginia, East Kentucky, and South Ohio. I always thought that was Christian country.
The Utah Island should be relabeled to what the locals call “Morridor”. As a reference to the Mormon corridor along I-15 interstate and as a joke about “Mordor” the volcanic plain from Lord of the rings where evil resides.
This just proves that the people who CLAIM to be the most religious in America, are actually the least in practice. The Bible Belt, aka the Belt of Hate.
Knock it off.
Um – you accidentally omitted a couple of words in your last sentence – it should have said “aka the Belt I have been taught to hate” You’re welcome!
“Religious adherents as a percentage of all residents, 2010.”
By county? By zip code? Per square mile? Per 10,000 residents?
Can I see your code please? I have been trying to get unique value renderers to work all day! or is this all done on arcgis desktop?
Alex, I love this! Thanks for helping make sense out of our world.
Two notes re: Texas:
1) The region you labeled as “North Texas Ridge” is actually what we call West Texas, and further up it is referred to as the Panhandle. “West Texas Ridge” would be more appropriate.
2) That deep trench over Loving County (where the Panhandle and Far West Texas meet) is a low spot mainly because of low population. I think Loving Co. has 90 people, most of whom work on the one ranch that covers virtually the entire county, and I’m fairly certain the religious activity there is off the grid.
Thanks for all you do.
Interesting that there are two blips of green showing up north of Moundsville, West Virginia and Alachua, Florida: both places where Hare Krishna communities are located.
Interesting subject, and really good visualization!
The “Jeff Davis Lake” looks really, really out of place…! 🙂
Sure glad I live in blue. North California near Sacramento.
I have some questions:
1. Why did you set your symbology the way you did? With the 46-49% class being set to white, and the 50-55% class being set to green, I think your map exaggerates the difference between areas that are very slightly religious and very slightly non-religious. The result is misleading data. Take Central California for instance – at first glance, it seems strangely more religious than the rest of the state, but an examination of the legend reveals that is probably not actually the case.
2. Why did you perform an interpolation? Presumably you had discrete county-level statistics, and I think a normal county-by-county thematic map would have displayed the data better. Spatial interpolation relies on the assumption that the data is spatially correlated. – that Tobler’s Law is true. Since urban areas tend to be less religious than rural areas, I doubt how reliable an IDW trial would be in this case. I would recommend running a test of Moran’s-I to see if coordinate information, as opposed to some other attribute, is the best predictor of the attribute that you are trying to display on your map. If you were committed to performing an interpolation, an areal interpolation that aggregated county-level data into tract-level polygon data would have been more apt than an IDW trial.
3. Why is the top half of Alaska cut off?
No doubt one could overlay a map of political parties (I refuse to assign the color red to anyone but those who lean towards Communism) and see that those who reject their maker vote almost exclusively for democrats.
Montana is one of the Lower populated states so there is a lot of country with very little population. So you would see alot of blue or low %. But I found it interesting that Missoula and Bozeman areas are blue. College towns? More temporary transplants?
Happy You Are More Perfect Than The Rest Of Us Are……
It’s bc the God of that faith itself (Christianity) actively changed how things get done, which is more to say than others.
Conservative Christian from the South, trying to hold onto this country’s original values set forth in the Constitution!
Just like with any group, please don’t think that all Christians are right wing conservatives who seem to be too quick to judge and too slow to love. Some of us Christians are as sickened by that as some of you (understandably) are.
Happy You Are More Perfect Than The Rest Of Us Are……