Islam: The Numbers Don't Lie

Nine years after Ben Affleck accused Bill Maher and Sam Harris of being "gross" and "racist" for criticizing Islam, Sam Harris engaged in a frustrating 90-minute debate with Rory Stewart on the same topic. Rory, armed with a posh British accent and constant appeals to authority, managed to sound reasonable while insisting that basic statistical inference would be an epistemic mistake when discussing Islam.

Rory fell back on his "lived experience" in Muslim communities, suggesting that Muslims are diverse, heterogeneous, and mostly just normal people trying to live their lives – all of which isn't disputed by most reasonable people. In fact, as Sam Harris tries to point out, it's people who are keen to exonerate Islam that might be forced to attribute pernicious Islamic laws and norms to something "inherent" about Muslims as people.

Their conversation, though frustrating, provides a framework for identifying key points of contention. Here, I intend to tackle three questions: Is Islam a meaningful variable when analyzing geopolitical risks, terrorism, and threats to freedom and human rights? How seriously should we take the theocratic aspirations of Islamist movements, and what are the implications for free societies? And finally, how can we foster an environment that encourages reform within Islam and empowers liberal Muslims to speak out against extremism?

The wrong question: Who gets to define Islam?

Rory Stewart: I concede that there are versions of Islam that are problematic. Groups like ISIS, Hamas etc. are mostly horrible. And you can’t understand them without understanding the strong religious motivations that underpin their ostensibly crazy actions. However, I don’t think there’s just one Islam. There are many different Islams - why do you let ISIS or the Ayatollahs in Iran define what Islam is rather than Muslim clerics who disavow violence and advocate peace?

Rory, and people in his camp, often respond to descriptive claims about Islam and Muslims with normative arguments about whom they would like to empower to define Islam as a religion. This line of thinking is not too different from the argument of some liberal women that mainstream feminism is really stupid these days, but real feminism, which they have the authority to define, is really good actually. But these discussions elevate the theoretical above the actual. If most people calling themselves feminists and making careers as feminists have bad ideas, at some point we need to be able to criticize feminism (à la Bryan Caplan). For a more extreme illustration, imagine you’re in the late 1930s, trying to make the case that German nationalism is problematic, citing the rise of antisemitism, subversion of political norms and an ideological obsession with reclaiming national pride that is likely going to lead to bad outcomes. Your debate opponent responds with the following:

“I agree that German nationalism can be problematic and has problematic elements, and I’m no fan of the Nazi leadership. But there are plenty of decent people who love their country and want to see it rise and prosper, and support peaceful means to get there. Why not empower the good actors to define what it means to be a German Nationalist?”

Your beliefs about who should get to call themselves the real German nationalists doesn’t change facts about the current momentum and character of an ideology, the power structures that control the dissemination of that ideology or the relative popularity of good and bad ideas within it. Now, there may be times, perhaps like at the end of World War II, when the Allies probably did want to empower liberal Germans to redefine German national identity, after having completely dismantled the previous system.

But if you’re having a debate in the 1930s about the source of all the worrying developments you see around you, pointing a finger at a particular ideology obsessed with righting historical wrongs and ensuring German revival would be a pretty good place to start. Similarly, arguing about what constitutes “real Islam” is futile, since the point of this exercise is not to judge Islam in some abstract way. It’s to identify the common variable that’s driving the cluster of bad outcomes/behaviors that we’re concerned about in a hope to point us to the right solutions.

Is Islam the right variable to focus on?

Now one could reasonably ask - Is Islam even the right variable to hone in on? One can find examples of reprehensible behavior by Muslims that is not attributable to Islam, for example the horrendous practice of female genital mutilation. Female genital mutilation (FGM) is prevalent in many Muslim countries. Yet, one can find little justification for the practice in the Quran. More importantly, the countries with the highest prevalence rates are heavily concentrated in North Africa, and many non-Muslim tribes and countries in this region also practice it. If you aimed to identify the causal drivers of this practice impartially, your focus would shift from Islam to the cultures and traditions of African tribes in the region.

Unfortunately, for a whole basket of other problems, Islam, the religion, seems to be a reasonable variable to drill down on. One such problem is rates of political violence and terrorism. Rory’s instinctive response to this is to claim that most of the violence in the world doesn’t have much to do with Islam; and that he doesn’t see Muslim regions of the world as that much more violent than non-Muslim regions. This is either vague enough to be meaningless, plainly false or attacking a straw man. If he means that Islam as a variable doesn’t explain most (>50%) of the variance in violence and war globally, that’s obviously true. But no one variable can explain most of almost any complex phenomenon.

I ran a quick regression to check how levels of violence within countries correlate with the presence and influence of Islam within a country. I used the Global Peace Index developed by the Institute for Economics and Peace, which scores countries on levels of violence, as the dependent variable, and proportion of the population that identifies as Muslim as the independent variable. We can explain ~16% of the variance in levels of violence within countries using the % of Muslims in the population.

How much better does per capita GDP do at predicting violence - lower levels of which most people would intuit correlate with higher levels of violence. Slightly better, but not by much - it explains about 19% of the variance in global violence. The proportion of Muslims within a country only explains 5% of the variance in per-capita GDP between countries. So it’s not just that Muslim countries tend to be poor and poor countries tend to be violent. If you’re unconvinced, try to think of other variables that might yield a stronger correlation. I doubt % of Christians would do it, or any other religion for that matter. Perhaps some measure of democracy might do better (but the potential for reverse causality is high there, since countries that are violent are unlikely to develop democratic institutions)

The claim that gets thrown around then is that radical Islam is to blame, not the religion per se. But what exactly does the term “radical Islam” refer to? It’s not a special sect within Islam or some category that can be neatly separated from the rest of the religion. It’s just what we use to refer to members of Islamic groups who do bad things in the name of Islam. It’s true that some revivalist strains of Islam, like Salafi Islam, which originated in Saudi Arabia and advocates for returning to the 7th century life of the Prophet Mohammed, have radicalized previously moderate Muslim populations, in places like Malaysia or Kerala, India. But these moderate Muslim populations were able to be radicalized because the tenets of Salafi Islam have ample support within the text of the Quaran, which they already accepted as unassailable.

But the problem is not limited to the Saudi export of Salafi Islam. A majority of the religious fighters that formed the Taliban were drawn from Deobandi seminaries, a strain of Islam which has its roots in 19th century South Asia, and the greatest geopolitical threat that Islam currently poses is in the form of the Islamic Republic of Iran, which has not only birthed several hardline Shia militant groups like Hezbollah, but has also formed coalitions with Sunni groups like Hamas in an attempt to unite against the infidels.

A common counterargument is that Western foreign policy, rather than Islamic ideology, is the primary driver of Muslim violence worldwide. Notice that conceding this argument doesn't make much of a difference. No one is arguing that Islam doesn't interact with other forces in the world, or that it is monocausally responsible for all these ills. The claim is that given contact with the outside world and perceived grievances like oppression, invasion, or humiliation by non-Muslim powers, Islamic doctrines makes it much more likely that these grievances are channeled into violence, with worse expected outcomes.

More importantly, if the motivations were purely political, we wouldn't see second and third generation Muslim immigrants in the West, who enjoy political freedoms and economic opportunities, being drawn to jihad. The 7/7 London bombers and the perpetrators of the Paris Bataclan attack, for instance, were citizens of democratic countries, not people under direct foreign occupation.

We also wouldn't see the cavalier willingness on behalf of Islamist groups to murder Muslim civilians en masse, particularly Shia Muslims and other religious minorities, who had nothing to do with Western foreign policy. The prevalence of sectarian violence and the targeting of "heretical" or "impure" Muslims points to the significant role of religious ideology in justifying and directing terrorism, beyond mere political grievances with non-Muslim powers.

It thus seems clear that Islam, as a religion and ideology, is a significant variable that cannot be ignored when analyzing the drivers of political violence, terrorism, and sectarian conflict in the modern world.

Disagreement #2: How much of a geopolitical threat does Islam and Islamism pose?

While it’s true that Islamic terrorism isn’t the most likely source of catastrophic geopolitical risks, it’s important to remember that the world we live in now is a function of our efforts since 9/11. We decimated Al-Qaeda’s leadership and infrastructure, and stopped ISIS in its tracks, precisely because we took this threat seriously. As Richard Hanania argues in his piece on the American Empire, we can’t use statistics from today’s world order to argue for a live and let live approach with Jihadists, which would very likely produce different outcomes.

There are at least two good reasons to be worried about political Islam on the geopolitical stage. The first is its sincere imperialist aspirations for a global caliphate. This aspiration for political conquest and expansion is not a modern aberration within Islam, it's the continuation of a trend that started with the life of Mohammed and gained much more traction after his death, with Islam expanding and conquering large swathes of the old world. Even if Jihadi terrorism and suicide bombings are modern tactics with controversial theological support, the desire to spread Islam by the sword is deeply rooted in Islamic history and doctrine.

Unlike the Spanish Inquisition or other historical examples of religious violence that have no modern ideological flag-bearers, the fervor for a global Islamic caliphate has been passed on with high fidelity through the centuries. It survived the fall of the Ottoman Empire and reemerged in the form of revivalist movements like the Muslim Brotherhood. Today, groups like ISIS, Al-Qaeda, and the Islamic Republic of Iran openly declare their aim to establish Islamic rule worldwide, often citing religious justifications. As Joshua Hoffman puts it in his piece about Islamophobia:

What many people don’t realize about Islam is that it is not just a religion. Islam is also a political ideology. One aspect of this ideology is that Islam “doesn’t recognize a distinction between mosque and state, the way that Christianity recognizes a distinction between church and state,” according to Niall Ferguson.

The second is the embrace of martyrdom and the promise of paradise. October 7 gave us both quantitative and qualitative evidence of this, and people like Rory seem to have missed the former. Hamas was able to mobilize over 1,000 people within Gaza's population of 2 million to embark on a suicide mission aimed at murdering Israeli civilians. The annual homicide rate in Gaza is a mere 0.8 per 100,000 people, and the suicide attempt rate hovers around 20 per 100,000. If you apply these rates to Gaza's population, you'd expect about 16 homicides and 400 suicide attempts in a typical year. Most suicidal people certainly aren’t homicidal and most homicides aren’t committed by people who would kill themselves or random strangers. In the absence of religious justification, we might expect only a handful of citizens that would be capable of suicide AND murder in its most gruesome forms. Yet, Hamas managed to find more than a thousand willing suicidal murderers, and I suspect they could’ve found more.

Yes, they didn’t explicitly commit suicide but the mission was effectively suicidal, as evidenced by the infamous whatsApp call - “What do you mean come back? It’s either death or victory”. Even for the most delusional, it's clear that "victory" was impossible, making death inevitable. This is a truly astonishing feat of mobilization that should give us all pause. It's worth noting that the US, despite its relative prosperity and stability, actually has higher homicide and suicide rates than Gaza. This suggests that oppression and poverty are insufficient explanations, since Gazans aren’t killing each other or themselves at staggeringly high rates. Sure, you could argue that it’s hatred and not religion. But hatred alone doesn’t explain the religious ecstasy that was on display or the total lack of self-preservation as an instinct. The Japanese disregard for life and self-preservation during World War 2 had an explanation rooted in culture and ideology. Similarly, the Gazans left a footprint of their belief in martyrdom and hatred for the infidels all over Southern Israel that day. [Note: If you don’t trust any data from Gaza, you can uncharitably assume Gaza combines the world’s worst suicide rate with the worst homicide rate, that of Lesotho and Jamaica (thanks to Bukele). Assume 2% of homicides are murder suicides (that’s roughly the ratio in America), you still only get 20 murder suicides a year. And we haven’t started accounting for the ability to kill multiple strangers, the rape etc. ). Still nowhere close to explaining the behavior]

This has profound implications for how we think about conflict and deterrence in the face of Islamic extremism. When your adversary's payoff function is heavily skewed towards the rewards of the afterlife, the usual carrots and sticks of diplomacy are far less effective. If militant groups believe that dying while fighting infidels is a win condition, they'll be happy to keep rolling the dice no matter how many setbacks they suffer.

This isn't to say that all Islamist violence is driven solely by religious fervor. Political oppression, economic desperation, and other grievances undoubtedly play a role in fueling conflict. But the events of October 7th should disabuse us of the notion that worldly concerns are the whole story. The ecstatic expressions of the militants in Gaza, caught on video as they carried out their suicide mission, are hard to square with the image of reluctant soldiers laying down their lives for a purely political cause. There's an otherworldly zeal at work here, one that's inextricable from the theology of martyrdom in Islam.

Disagreement #3: Does Islam pose a threat to free society? 

Consider the following names: Theo Van Gogh, Jean Cabut, Stephane Charbonier, Sabeen Mahmud, Naguib Mahfouz, Elsa Kayat, Philippe Honore, Bernard Maris, Lars Hedegaard, Mustapha Ourrad, Bernard Verlhac, Georges Wolinski, Salman Rushdie, Farag Fouda.

What do they have in common? They were all targets of assassination attempts, often successful, for the "crime" of criticizing or satirizing Islam. The Charlie Hebdo attacks in Paris and the fatwa against and stabbing of Salman Rushdie are just the most high-profile examples of a much broader trend of violence aimed at silencing dissent and criticism of Islam.

Most of these assassinations happened in free countries - in America, Netherlands, France, and a few in Pakistan and Egypt, where atheists and free-thinkers face persecution. Nearly all Muslim-majority countries have blasphemy laws on their books. According to the USCIRF, which studied blasphemy law enforcement from 2014 to 2018, out of 84 countries with such laws, 43 had no instances of enforcement.

Where were blasphemy laws enforced then? "In the 41 countries across Asia-Pacific, the Middle East, Europe, and Africa, researchers found that enforcement exists to varying degrees, from 1 reported case in 8 different countries to 184 reported cases in Pakistan during the study period." The distribution is far from uniform, with five countries accounting for 80% of enforcement cases. The common factor is clear. Four countries - Pakistan, Iran, Brunei, and Mauritania - have death penalties for insulting religion. Just a few days before I wrote this, a Pakistani student was sentenced to death for sharing "blasphemous" pictures that offended Muslim sentiments on WhatsApp.

Now, let’s talk about women. Our World in Data produces a women’s civil liberties index, which “captures the extent to which women are free from forced labor, have property rights and access to the justice system, and enjoy freedom of movement” Scores go from 0 (worst for women) to 1 (best for women). And surprise, surprise… civil liberties for women are strongly negatively correlated with the % muslim population within a country, the latter accounting for over 20% of the variance in women’s civil liberties between countries. Muslim-majority countries are also about three times as rich as African countries (in per-capita GDP terms), and yet fare worse as a group on civil liberties (median score of 0.50 vs. 0.59).

These statistics themselves are appalling, but they belie the actual scale of suffering. When countries persecute religious minorities, we’re quick to be outraged and rightly so. Meanwhile, in many Muslim majority countries, 50% of the population live perpetually as second class citizens. Women are often married off against their will -sometimes before adulthood, treated as baby factories in their prime years and criminally sanctioned for everything from daring to go out without a guardian to extramarital affairs. Worse still, women’s testimonies are systematically undervalued, restricting the potential for legal recourse from their ‘masters’. I’m surprised I haven’t heard any comparisons of this arrangement to “Apartheid”. To me, it seems worse than Apartheid and on a spectrum with institutionalized sex slavery.

Regan Arntz-Gray

captures my view on feminists who fail to call this out in her piece on feminist cultural relativists:

If you’re a feminist you cannot be both philosophically consistent and a cultural relativist. The stable core of feminist ideology is anti patriarchal. What exactly being “anti patriarchal” means and whether you think America is still a patriarchal society will depend on what you believe about, sex, gender and history. Infighting within feminist movement goes back to its birth. But regardless, being feminist explicitly means that you think some cultures, the less patriarchal ones, the ones that treat women better, are better than others.

I’ve talked about laws here. But in the vast majority of countries, laws grossly understate the regressive social norms and practices. With the exception (perhaps) of Iran, the ruling elite in these countries is often more liberal than the population at large. The reason MBS, crown prince of Saudi Arabia, is lauded for his reforms in this area is because he’s pushing these through in the face of staunch opposition from clerics and the ultra-conservative population they represent. More people are killed in vigilante attacks on suspicion of blasphemy than are imprisoned by the state in Pakistan for the same crime. Most countries have laws against child marriage but have disturbingly high levels of child marriage with double digit percentages reported in countries like Jordan and Syria. This Pew poll released in 2013 that surveyed 38,000 Muslims in over 80 languages across every country with a significant Muslim population, is a bit dated but should give you some sense of what public opinion in the Muslim world looks like.

The evidence is clear and overwhelming: Islam, as it is currently practiced and interpreted by millions of people around the world, poses a grave threat to the core values of free society. The systematic suppression of dissent, the brutal subjugation of women, and the often violent enforcement of religious orthodoxy are incompatible with the principles of free and open societies.

Conclusion - so what now?

In conclusion, it is crucial that we incorporate Islam and Islamism into our geopolitical models. Non-religious liberals often fall prey to the similar mind fallacy, and certainly not only on this issue - they just can’t take beliefs that they deem crazy seriously, insisting that the only real motivation anyone ever has for doing anything is economics or physical safety. The answer to “Why are the Houthis doing this? Don’t they know they’ll get bombed?” is staring at you in plain sight.

Even if you're willing to shrug at the plight of women, minorities, atheists, and liberals in Muslim countries or argue that we can't do much, there's no good argument for being sanguine about importing these norms into Western societies. The last few months since October 7 have provided plenty of anecdotal evidence, including a Member of Parliament's resignation in Britain in response to threats from Islamist thugs. But apart from anecdotes, poll data consistently shows that over 50% of Muslims in Britain think homosexuality should be criminalized, over 39% think wives must always obey their husbands, and over a third refuse to condemn violence against those who insult Mohammed. Think about that last statement. In an anonymous poll where one can reasonably rule out fear as a motivation, a third of British Muslims supported, in principle, the right for Muslim vigilantes to enforce Islamic blasphemy law with violence in Western societies.

Another crucial reason to have this conversation is to create an environment that empowers moderate or secular Muslims to speak out against extremism and bigotry. While some may accuse these "moderates" of being silent out of solidarity, the reality is that remaining silent is the rational course of action for most people within these communities, given the risk of ostracization and lack of clarity on where the community en masse stands. It’s a collective action problem.

To incentivize Muslim communities to purge these extreme elements from within, we need to create external pressure that makes it untenable for Muslims to maintain both their religious identity and credibility in public life without actively denouncing extremism. In other words, we need to make it so that the cost of not speaking out against radicalism is higher than the cost of doing so. This can only happen if accusations of extremism or illiberalism within the Muslim community are taken seriously and not simply dismissed as Islamophobia.