Religious Outliers Nonsense (or “Atheists Are Richer Than Religious People If You Take All Poor Atheists Out Of Your Sample”)

Charles Blow’s most recent New York Times op-ed is something of a boon for visualization enthusiasts. He replaces almost his entire article with a visualization. This illustrates that he recognizes power of visual communication to make and reinforce a point in a way that is self-obvious and can stick with the reader better than words.

Unfortunately, he has decided to use data that misleads his audience to such an extent that I can only conclude that he is unconcerned with the truth insofar as it undermines his desired objective.

Blow’s main point is that the US is an outlier in the world because we’re religious but also rich while “religiosity was highly correlated to poverty”.

I’ve reproduced the chart in question below. (Click to enlarge)

image

Now, keep in mind that this is not charting religion as it is listed in the CIA World Factbook, but according to the specific question: “Is religion an important part of your daily life?” That will be important in a little bit.

This chart seems to prove his point. Until you realize what isn’t on the map.

Here is a list of the countries that didn’t manage to make their way onto the map due to the fact that Gallup didn’t poll them:

China – 1.33 billion people, heavily non-religious, poor

North Korea – 22 million people, heavily non-religious, unbelievably poor

Cuba – 11 million people, presumed non-religious, poor

Taiwan – 23 million people, 93% Buddist*, rich (comparable to Japan)

Problem number one – Charles Blow has a duty to inform his audience of these omissions. The countries without data represent nearly 25% of the world population and skew heavily toward non-religious. They are too large and too important to the data set and visual reference to simply ignore. Yet Mr. Blow doesn’t seem interested in mentioning them.

Problem number two – Mr. Blow heavily implies that there is a causal relationship between religiosity and wealth. But (as we all know) correlation doesn’t imply causation. Western European countries (and countries filled with people from Western Europe) are richer, as are developed Asian countries. Eastern European and South American countries are less rich. Middle eastern, and African countries tend to be much poorer. There’s a correlation in geo-political histories here that is stronger than religion.

Of course Mr. Blow could always go to rural India and inform them that their poverty is related to their devotion to Hindu and has nothing to do with British imperialism. Or perhaps to the deep south where he can proclaim to the +90% Christian black population that their economic woes are related to their religious tendencies.

Problem number 3 – But the final problem is the worst one because it involves an outright lie:

Singapore is more religious and richer than the United States. And Mr. Blow didn’t map it. At all.

It’s possible that Mr. Blow is actually so numerically illiterate that he didn’t know he was supposed to tell people about key missing data points. But taking out data that doesn’t align with his point is disgusting manipulation. The end result of his deception (conscious or otherwise) is “If you take out all the poor atheists and take out all the rich religious people, then this pattern emerges…”

Mr. Blow should put Singapore back in to the data set and add a correction to his article that announces how his data set has enormous gaping holes. And he should probably never be allowed to touch charting software again.

* The CIA Factbook has Taiwan listed at 93% Buddhist, but I’m not sure how they would answer the specific question that Gallup asked. I’ve heard some atheists claim Buddhism as an “atheistic religion” (no personal god) so it could be that the citizens of Taiwan wouldn’t say that religion plays a big role. I simply don’t know.

14 comments

  1. Kuvetli says:

    Great article! I agree with your corrections, but when you click to enlarge the picture of your reproduction of his visualization it doesn’t enlarge… I can’t tell what’s on his version

    Thanks!

  2. FriendIndeed says:

    I hope you’re planning to send this in to the NYT as a letter to the editor. I don’t imagine that China allows western pollsters to interview its citizens, so Blow would probably rebut that there’s no data for that country. However, it’s still a glaring omission, and reasonable estimates can be made given the size of China’s population and GDP. You may want to include the correlation coefficient and the p-value, given that even the data presented aren’t super. You can use a utility called “engauge digitizer” to get estimated numbers from the image, and you can probably even develop a weighted metric based on the population size.

  3. razib says:

    minor note: hindi is a language. hindu(ism) is the religion.

  4. politicalmath says:

    razib,

    How complete horrifying that I made that mistake. I’ve fixed it. Thank you for pointing it out.

  5. Al says:

    Fantastic rebuttal. I absolutely love reading this stuff.

  6. dagamore says:

    I wonder if he washed is data via the East Anglia University?

  7. Dylan says:

    Cuba non-religious? Are you making that assumption based on the fact that Cuba is a communist country (ie communist = atheist)?

    http://en.wikipedia.org/wiki/Religion_in_Cuba

    “The Roman Catholic Church estimates that 60 percent of the (Cuban) population is Catholic.[1]
    Membership in Protestant churches is estimated to be 5 percent”.

    At least 65% of Cuba is religious. This should put Cuba on par with the US in terms of religious belief.

    As for China, it is more religious than you might think. It is at least 40% religious:

    “Nowadays Shenism-Taoism and Buddhism are the largest religions in China with respectively 20-30%[8][9] (of which 160 millions, or 11% of the total population of the country, are Mazuists[10]) and 18-20%[11][12][13] of the population adhering to them, thriving throughout the country as the government is allowing them to spread.[14] Christians are 3-4% of the population according to various detailed surveys,[15][16][17][18] despite American Christian press claims there might be more;[19] Muslims are 1-2%.[20] The remaining section of the population, ranging between 40% and 60%, is mostly agnostic or non-religious; purely atheists are 14-15%.[21][22][23] Various new religious movements, both indigenous and exogenous, are scattered crosswise the country.[24]”

    http://en.wikipedia.org/wiki/Religion_in_China

    Point taken though. Such large data should not have been omitted or, at best, he should have been transparent about the omission.

  8. alex h. says:

    Agreed -at least with most of your points.

    Whatever the results maybe there is a real danger to play with metrics including religion and economics. Education plays a major part in not re-creating patterns that led to horror (think of WWII) and obviously the author missed something there.

    Note however he could have make things worse by adding the estimated IQ into the equation…

    On a similar note, yet much more accurate and positive, I recommend you this TED video: http://www.ted.com/talks/hans_rosling_shows_the_best_stats_you_ve_ever_seen.html

  9. Russian Spy says:

    While I agree with your general points, I think you misunderstood the point of the article. First of all, correlation obviously does not imply causation. The point of the article is that the US was an anomoly when compared to many other wealthy countries. Your use of the Taiwan and Singapore is a clear refuation of this.

    However, I think the chart is still useful as way of visualizing countries along their relationship with religion. The countries can be grouped in 4 catergories:
    Lower-left: Former or (Current) Communist Countries poor and mostly atheist
    Lower-right: Semi-socialist developed countries (Western European model), generally wealthy, atheist
    Upper-left: developing countries, religious and poor
    Upper-right: religious and wealthy, some are very free-market oriented like US, Ireland Taiwan and Singapore, and some are still semi-socialist (Israel, Italy, Greece)

  10. usc says:

    Great rebuttal. I would take your targets attemp one step further…he is making the assumption that all religions are theocracies. Which we know is not the case. Yes, Islam is a theocracy and therefore does control one’s economic future on an individual level, but Christianity is not a theocracy. So his arguments, regardless of the skeweed and left out data is based on a major flawed assumption. It doesn’t matter what religion I am, if my countries economic system isn’t based or controlled by that religion I could worship space rocks and it wouldn’t affect my countries economy.

    usc

  11. nick says:

    Where is Australia or New Zealand in either diagram? If I were to guess I would say lower right quadrant. They both have comparable populations to Singapore. You have pointed out the original data set was incomplete, but yours is as well.

  12. Jay says:

    Let the willful ignorance reign supreme!

  13. jake mckenzie says:

    Even without the points you made I don’t take much from what he was trying to get across. If he had tried to tie data possibly to economic inequality, or punitive sentencing to religiousness of a society he might have gotten some sort of point across. Possibly that religiousness is leading to inefficiencies in the society culturally effecting it economically. You should never use GDP unless you know that the other data point is directly tied to GDP in some way.

  14. It is real happiness to study you blog!

Leave a Reply

Your email address will not be published. Required fields are marked *