Wednesday, February 3, 2010

Islam predicts nothing

An email is going around claiming that Muslims conspire to force Islamic rule onto every country they migrate to. The full text can be seen here; I have copied a few highlights:

"As long as the Muslim population remains around or under 2% in any given country, they will be for the most part be regarded as a peace-loving minority, and not as a threat to other citizens. This is the case in:

United States -- Muslim 0.6%
Australia -- Muslim 1.5%
Canada -- Muslim 1.9%...

At 2% to 5%, they begin to proselytize from other ethnic minorities and disaffected groups, often with major recruiting from the jails and among street gangs. This is happening in:

Denmark -- Muslim 2%
Germany -- Muslim 3.7%
United Kingdom -- Muslim 2.7%..."
So we see the way this is going. The author believes that there are simple thresholds every few percent: as the proportion of Muslims increases, so does their aggression towards other faiths and ideologies:

"When Muslims approach 10% of the population, they tend to increase lawlessness as a means of complaint about their conditions.... At 40%, nations experience widespread massacres, chronic terror attacks, and ongoing militia warfare.... After 80%, expect daily intimidation and violent jihad, some State-run ethnic cleansing, and even some genocide, as these nations drive out the infidels."
One implication here is that one should be able to describe a country well simply by knowing what percentage of the population is Muslim. E.g. Bosnia and Herzegovina, 40%: "Let me check... ah, widespread massacres, chronic terror attacks, and ongoing militia warfare!"

Considering this, and other common anti-Muslim rhetoric, I decided to gather data on democracy, press freedom and (since the Muslims are also supposed to be a prolific lot and rapidly reproducing) total fertility rate, and then compare this with the percentage of Muslims in every country in the world.

I put the data into a Google Spreadsheet and appied it to a motion chart, which looks a bit like this:
On the vertical axis we have each country's 2008 Democracy Index score, as calculated by The Economist Intelligence Unit. On the horizontal axis we have the 2009 Reporters Without Borders Press Freedom Index score. Each dot represents a country, the size representing the percentage of total population which is Muslim. Colour represents geographic region. What this graph shows is very obvious:


There is no clear correlation between the proportion of Muslims in a population and democracy or press freedom.

Muslim-majority countries are scattered all over the graph, with some scoring dreadfully and some rather well. Turkmenistan (93% Muslim) has the third worst press freedom and third worst democracy in the world. Mali (also 93% Muslim) has a press freedom score of 8, which is better than Poland (9.5), South Africa (8.5), Spain (11) and France (10.7), and their democracy score is slightly better than Hong Kong.

One observation, though: there are no Muslim majority countries among the very highest countries, those most democratic and with the most press freedom. Perhaps, then, adherence to Islam does have some negative effect on democracy and press freedom, but certainly not in the mathematical way described on the rather overexcited email quoted above.

...But this negative relationship is unclear or weak. When Muslim-majority countries are compared with their neighbouring states, any correlation breaks down further. For both indicators, Muslim Indonesia is better than Christian Philippines, while Muslim Bangladesh scores far higher than Buddhist Burma, Bhutan and China, as well as Hindu Nepal (though lower than Hindu India).

So far, then, knowing the percentage of Muslims in a society tells us nothing about the extent of its democracy or press freedom - we can guess that a high-Muslim country won't be excellent for these indicators, but it might still be quite good.

What about fertility? Quite a few other internet rumours draw attention to alleged fecundity in Muslims (including this absurd propaganda video). The general argument is that Muslims have huge families, non-Muslims have small families and thus Muslims will take over the world.

Back to our data. This time I plot the percentage of Muslims in a country on the vertical axis (for clarity, size of dots still also correspond with percentage of Muslims), and total fertility rate on the horizontal axis:
Result: no correlation again.

Once more Muslim-majority countries are spread all over the graph. In Niger, the average woman has 7.75 children - but in Iran they only have 1.71, around the same as Sweden, Norway, China and Australia.

The same caveat should be made as before: among the countries with the very lowest fertility rates there are no Muslim-majority countries, and the two highest-fertility countries are Muslim-majority.

But Islam is an extremely poor predictor for fertility. Indonesia has a lower total fertility rate than neighbouring non-Muslim East Timor or Papua New Guinea. Egypt (95% Muslim) has a lower TFR than neighbouring Sudan (71% Muslim), which in turn has a much lower fertility rate than Uganda (12% Muslim).

There is a lot more data available to examine on this, but so far the result is quite conclusive: knowing the proportion of a country's population that is Muslim tells you very little about the political and social situation in that country.

In the second graph we can see sub-Saharan African countries clustering around the same high-fertility section, regardless of the proportion of Muslims in the population. Here geography is a better indicator than religion.

For all the frenzied debate over this in recent years, it could be that, compared with geography, wealth and political policy, religion really isn't all that important.


  1. Nice one Shane, was the creator of the email any well-known individual/organisation?

  2. Hi Nilan, probably not - I came across it first in an Orkut discussion forum. The internet is awash with theories along these lines, I think it's no harm to expose them to a little light sometimes.

    Thanks for the comment :)

  3. Excellent work, Shane! And the thought that crossed my mind when I read this line: "After 80%, expect daily intimidation and violent jihad, some State-run ethnic cleansing, and even some genocide, as these nations drive out the infidels."

    I'm like, what country are they talking about? Indonesia has a Muslim population well above 80% and it is probably a country in which democracy has been consolidated. Yes, it has problems with certain Islamist groups etc., but doesn't the US have problems with Christian Right groups as well?

    But yeah, good job dismantling the stereotype. People make claims so easily and then self-select "evidence" to prove their claims.

    Not that some Muslim countries don't have appalling human rights track records. But then so do China, Burma, North Korea, etc., and these are not Muslim countries.

  4. Thanks Shanon! Really I was shooting fish in a barrel by tackling that particular email, but I used it just as one example of the kind of discussion I've seen about Islam online. Both radical Islamists and anti-Muslimists are determined that Muslims are a monolithic mass of people who are destined to fight against democracy and liberalism.

    This clearly isn't so. As I mention here, geography seems a stronger predictor of culture and politics than religion. Interestingly enough Samuel Huntington, creator of the Clash of Civilizations concept, also argued that there was nothing inherent to Islam that made it violent or oppressive (rather he blamed the rise of Islamist violence on the youth bulge caused by rapid population growth, a bulge that is now receding).

    So there isn't a new Cold War, or civilizational war, going on now. Al Qaeda hoped to spark one, but most Muslims ignored the call to arms. The world is more complex than that, thankfully!


Note: Only a member of this blog may post a comment.