Not looking for hate crimes. Looking for percentage of crimes per capita broken down by religion of the perpetrator. (Example: 1M Protestants in USA, 10k crimes by Protestants, rate of 0.01% per capita.) Expecting to see a bar graph. Prefer a reputable source, not KKK. Might consider them if multiple opposing sources like an atheism organization show similar stats. Have only found an infographic. I know FBI has crime rates broken down by skin color, but unsure if they have them broken down by religion. Serious crimes (violent) vs. non-serious (victimless or non-violent) is a plus.

  • 6
    I've never been arrested, so can't say for sure, but I don't believe "your religion" is part of the arrest record.
    – user1530
    Commented Dec 14, 2015 at 21:57
  • @blip That's why it might be a hard question to answer, but I'm sure there are studies that survey prison populations.
    – Chloe
    Commented Dec 14, 2015 at 21:58
  • 1
    There are, but I think there's a challenge there in that many people 'find religion' in prison...be it a sincere conversion or one to appease fellow inmates or the parole board...so stats may be skewed.
    – user1530
    Commented Dec 14, 2015 at 22:00
  • 7
    That information would be meaningless without other variables. I mean, perhaps you find that Cthulu worshippers are the most likely people to commit crime, but maybe it is because the average Cthulu worshipper is uneducated and the real relationship is not between religion and crime, but between education and crime. The only way you could make any research using religion data alone would be if you assume that each religion has the same % of believers in every social class/education level/geographical zone & others, which is a rather questionable proposition.
    – SJuan76
    Commented Dec 15, 2015 at 12:30
  • 1
    @Trilarion: Which comment was a perfectly good answer - saying that a meaningful answer is not possible - that someone arbitrarily decided to delete :-(
    – jamesqf
    Commented Jul 17, 2019 at 19:28

1 Answer 1


While official stats on actual affiliation of inmates are not measured, Pew did a study in 2012 that notes the following:

Presence of Various Faith Groups

The relative size of each faith group within the prison population is difficult to gauge. The Pew Forum survey asked chaplains to estimate the approximate percentage of inmates in the prisons where they work who identify with each of 12 religious groups. It should be noted, however, that these findings cannot be used to reliably estimate the religious affiliation of the U.S. prison population. They provide only an impressionistic portrait of the religious environment in which chaplains work.

On average, the chaplains surveyed say that Christians as a whole make up about two-thirds of the inmate population in the facilities where they work. Protestants are seen, on average, as comprising 51% of the inmate population, Catholics 15% and other Christian groups less than 2%. The median estimate of the share of Protestants is 50%, meaning that half of the chaplains estimate that Protestants comprise more than 50% of the inmate population where they work, and half of the chaplains estimate the figure to be below that.

The chaplains’ responses also suggest that many other faith groups are represented in the prison population. On average, the chaplains surveyed say that Muslims make up 9% of the inmates in the prisons where they work, with half of the chaplains saying that Muslims comprise 5% or less of the inmate population and half saying that Muslim inmates make up more than 5% of the inmates where they work. On average, other non-Christian groups are perceived as considerably smaller in size.

Chaplains’ perspectives on the religious makeup of inmates may reflect a number of different influences, including their degree of exposure to various groups in the course of their work. But even if the chaplains interviewed had perfect information about the relative distribution of religious groups among inmates in the prisons where they work, the findings would not be weighted in proportion to the size of the overall U.S. prison population. As a result, they would not provide an accurate count of religious affiliation in the U.S. prison population.

I would caution that this certainly does not measure affiliation against things like severity of crime, so is relatively meaningless for statistical correlation against concepts like violent crime, but only may have some relevance (with a very large grain of salt) against overall population figures on affiliation.

The closest thing to "official" numbers I could find was the following, but take note of the list of problems with the dataset that follows the chart, especially noting that the chart is based on a response of only 75,000 prisoners in federal prisons - which is peanuts compared to the total prison population.

  • 1
    Hmm, that's not very useful. It's not per capita. Of course the majority religion will have the majority of inmates. I thought it was obvious, but I'll add it to the question anyways.
    – Chloe
    Commented Dec 15, 2015 at 22:16
  • 1
    75,000 can be more than enough for a survey. A sample size of 1,000 is usually considered to be sufficient, if the sample size is selected representative. But the other problems mentioned about the dataset are far more severe.
    – Philipp
    Commented Dec 16, 2015 at 12:10
  • 2
    Philip, I'd agree that 75K could indeed be a statistically relevant sample of a 2.2 million population under some circumstances, but that assumes careful selection of your representative sample. Assuming that a survey of Federal prisoners would be relevant to a population predominantly made up of State prisoners seems a stretch unless one is to assume that demographic propensity for federal crimes = propensity for all crime, especially when considering that white-collar crime is almost entirely prosecuted in federal court, so demographics are skewed by income between state and federal. Commented Dec 16, 2015 at 13:39
  • 1
    Prisoner data may also encode religion-conditional arrest probabilty, conviction probability and sentencing severity probability. You also exclude all crimes that don't lead to a custodial sentence.
    – James
    Commented Jul 17, 2019 at 15:06
  • A difficulty of the Pew sample is also the considerable variation in religious affiliation regionally. A Utah or Idaho sample will suggest that Mormons are very criminal. A Rhode Island one would suggest Catholic criminality. An Alabama survey will point to Evangelical Protestants. A New York City survey will make non-Christians look far more criminal. You need lots and lots of data to get a representative survey and it also matters if you are controlling for income and race which aren't distributed equally by religious affiliation.
    – ohwilleke
    Commented Jul 18, 2019 at 0:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .