Terrorism, Islam, and Religious Violence: What the Data Actually Shows

Between 1979 and April 2024, at least 66,872 Islamist terrorist attacks killed at least 249,941 people worldwide [1]. Five groups — the Taliban (71,965 deaths), ISIS (69,641), Boko Haram (26,081), al-Shabaab (21,784), and al-Qaeda (14,856) — were responsible for 81.8% of those fatalities [1]. These numbers are not close to those produced by any other ideological category over the same period. This is a statistical fact, documented in the Fondapol database, the Global Terrorism Database at the University of Maryland, and the Global Terrorism Index published by the Institute for Economics and Peace.

It is also a statistical fact that more than 91% of the victims of Islamist terrorism are themselves Muslims [2]. That 83% of attacks and 90% of deaths occurred in Muslim-majority countries [1]. That in the United States since 2001, excluding the September 11 attacks, right-wing extremists have killed more people than Islamist extremists [3]. And that 99.99% of the world's 1.8 billion Muslims have never engaged in any form of political violence.

Both of these realities are true simultaneously. The failure of public discourse on this subject stems from the insistence that only one of them matters.

The Global Numbers: Scale and Trajectory

The Fondapol database, maintained by the Foundation for Political Innovation in Paris, represents the most comprehensive accounting of Islamist terrorist attacks worldwide. Its data covers three distinct periods that reveal a clear trajectory [1]:

• 1979–2000: 2,194 attacks, 6,817 deaths
• 2001–2012: 8,265 attacks, 38,187 deaths
• 2013–April 2024: 56,413 attacks, 204,937 deaths

The acceleration is stark. The post-9/11 period saw a fivefold increase in annual Islamist attacks compared to the previous two decades. The rise of ISIS after 2013 drove another sevenfold surge. At the peak of the Islamic State's territorial control between 2014 and 2017, Islamist terrorism fatalities reached levels unprecedented in the modern era [4].

Source: Fondapol Islamist Terrorism Database

Data as of Apr 12, 2024CSV

The Global Terrorism Index 2025 shows that this trajectory has shifted geographically. Burkina Faso ranked as the country most affected by terrorism for the second consecutive year, with 1,532 fatalities. Pakistan surged to second place with 1,139 deaths — its highest level since 2013. The Sahel region now accounts for over half of all terrorism-related deaths worldwide [5]. The number of countries experiencing terrorism increased from 58 to 66 [5].

The decline of Islamist terrorism in the Middle East following ISIS's territorial defeat has been partly offset by expansion into sub-Saharan Africa, Mozambique, and parts of Southeast Asia. JNIM (Jama'at Nusrat al-Islam wal-Muslimin) and Islamic State affiliates in the Sahel have made that region the new global epicenter [5].

The Western Picture: A Different Dataset

In the United States and Western Europe, the data tells a different story — and the frame you choose is itself a political act.

The Cato Institute's comprehensive analysis of 51 years of politically motivated killings on U.S. soil (1975–2025) found that Islamist attacks account for 87% of all terrorism deaths, driven overwhelmingly by one event: the September 11 attacks, which killed 2,979 people [3]. Remove 9/11 and the 1995 Oklahoma City bombing, and those two outlier events account for 88% of all terrorism murders over the entire half-century [3].

Excluding those two outliers fundamentally changes the picture. The share of deaths from right-wing attacks rises from 10% to 61% — the highest of any ideological category [3].

Since 2020, the Cato data shows right-wing terrorists accounting for over half of the 81 terrorism murders on U.S. soil, Islamists for 21%, and left-wing extremists for 22% [3].

The Center for Strategic and International Studies tracked a parallel trend: right-wing extremists perpetrated two-thirds of terrorist attacks and plots in the United States in 2019, and over 90% between January and May 2020 [6]. The 73 far-right incidents recorded in CSIS's database represented an all-time annual high going back to 1994 [6].

A 2022 study published in the Proceedings of the National Academy of Sciences (PNAS), authored by Jasko, LaFree, Piazza, and Becker, found that in the United States, "there is no difference between the level of violence perpetrated by right-wing and Islamist extremists." Globally, however, "Islamist extremists engage in deadlier attacks" compared to other ideological categories [7].

The FBI and DHS have repeatedly identified domestic violent extremism — primarily racially motivated violent extremism — as the top terrorism threat to the U.S. homeland [8]. White supremacist extremists were assessed as the "primary threat" among all domestic violent extremists, responsible for more than 50% of the 443 extremist-related deaths recorded by the Anti-Defamation League between 2012 and 2021 [8].

The Christchurch mosque shooting in New Zealand, the El Paso Walmart massacre, the Pittsburgh Tree of Life synagogue shooting, the Buffalo supermarket shooting — these were all white supremacist attacks. They killed scores of people. They were not Islamist.

Who Dies: The Victims Are Overwhelmingly Muslim

Between 2001 and 2015, 167,221 people died in terrorism-related attacks worldwide. Of those, 163,532 — 98% — died outside the United States and Western Europe. Muslim-majority countries accounted for 75% of all fatalities [9]. The Fondapol data shows 192,782 deaths from Islamist terrorism in Muslim countries, representing 91.7% of all deaths from Islamist attacks worldwide [2].

This is the single most important statistic in this entire debate, and it is routinely ignored by both sides.

For critics of Islam, it complicates the "clash of civilizations" narrative. If Islam itself were the driver, the victims would more plausibly be non-Muslims. Instead, the violence is overwhelmingly intra-Muslim — Sunni against Shia, extremist factions against civilian populations, insurgents against states. The primary victims of the Taliban are Afghans. The primary victims of Boko Haram are Nigerians. The primary victims of ISIS were Iraqis and Syrians.

For defenders who minimize the problem, it should prompt an honest reckoning with why certain Muslim-majority regions produce catastrophic levels of political violence that other Muslim-majority regions do not.

Source: World Bank Open Data

Data as of Dec 1, 2024CSV

The 100x Variation Problem

If Islam itself — its texts, its theology, its doctrinal structure — were the primary driver of terrorism, we would expect relatively uniform terrorism rates across the Muslim world, adjusted for population. We do not see anything close to that.

The GDP per capita data from the World Bank illustrates the spread. In 2023, the UAE had a GDP per capita of $49,851, Saudi Arabia $36,157, Malaysia $11,386, Kazakhstan $12,879, Turkey $13,375, and Indonesia $4,876. Afghanistan stood at $414, Somalia at $597, Pakistan at $1,360, and Yemen had no reported data due to state collapse [10].

The countries with the highest terrorism rates — Afghanistan, Somalia, Yemen, Syria, Iraq, Burkina Faso, Mali, Niger — share not a uniquely Islamic theological tradition but a cluster of structural conditions: state failure or fragility, active armed conflict, foreign military intervention, extreme poverty, and governance vacuums.

Indonesia — the world's largest Muslim-majority country with 277 million people — has a terrorism death rate orders of magnitude lower than Afghanistan or Somalia. The UAE and Kazakhstan, both Muslim-majority states, experience virtually no terrorism. Bangladesh, despite significant poverty, has relatively low terrorism rates.

This variation does not prove that theology is irrelevant. But it establishes that theology alone cannot explain the distribution. Something else — or several something elses — must account for the difference.

The Theology Argument: The Strongest "Yes" Case

Sam Harris, Ayaan Hirsi Ali, the late Bernard Lewis, and Maajid Nawaz have argued that specific Islamic doctrines create a unique theological infrastructure for violence [11]. The argument, steelmanned, runs as follows:

The concept of jihad as armed struggle — though contested by Muslim scholars who emphasize its meaning as inner spiritual striving — provides a scriptural basis for violence that has no direct equivalent in the current practice of other major religions. Martyrdom theology in Islam promises paradise for those who die fighting for the faith. The historical concept of dar al-harb (the "house of war," meaning non-Muslim territory) and dar al-Islam frames the relationship between Muslim and non-Muslim societies in terms of inherent conflict. Apostasy is punishable by death in several major Islamic legal traditions. And unlike Christianity's institutional separation of church and state (itself a relatively modern development), mainstream Islamic political theology has historically integrated religious and political authority.

Pew Research Center data from Muslim-majority countries lends empirical weight to parts of this argument. Substantial minorities express support for suicide bombing being "sometimes justified": 40% in the Palestinian territories, 39% in Afghanistan, 29% in Egypt, 26% in Bangladesh [12]. Support for sharia as "the official law of the land" ranges from 8% in Azerbaijan to 99% in Afghanistan, with solid majorities in most Middle Eastern, North African, sub-Saharan African, and South Asian countries surveyed [12].

Harris's argument is not that most Muslims are violent — it is that "the theological tools for justifying violence are more accessible and more mainstream in Islamic tradition than in contemporary Christianity, Buddhism, or Hinduism" [11]. Extremist groups do not need to distort Islamic texts beyond recognition to recruit; they exploit interpretive traditions that, while rejected by the mainstream, are not fabricated from nothing.

The fact that deradicalization programs in Saudi Arabia, Indonesia, and elsewhere prioritize theological counter-narratives — not just economic aid or psychological counseling — implicitly concedes that theology matters [13]. If religious framing were irrelevant to radicalization, there would be no need to counter it with competing religious arguments. Former jihadists consistently cite religious duty as their primary initial motivation for joining extremist groups.

The Structural Argument: The Strongest "No" Case

Robert Pape's research at the University of Chicago, spanning more than 2,100 suicide attacks, found that 95% occurred in response to military occupation by a democracy [14]. His central finding: "the key to understanding suicide attacks is not religion, but that they aim to compel democracies to withdraw military forces from the terrorists' national homeland" [14]. Half the suicide attacks in his original database were secular [14].

The structural argument holds that every major religion produces violence under certain political conditions, and those conditions — not theology — are the operative variable.

The historical evidence is substantial. The European Wars of Religion (1524–1648) killed millions in the name of Christianity: the German Peasants' War (100,000–200,000 dead), the French Wars of Religion (2–4 million), and the Thirty Years' War (3–12 million dead from warfare, disease, and famine) [15]. Some regions of Europe lost more than 30% of their population. The scale of Christian-on-Christian religious violence in this period dwarfs anything in the modern Muslim world in absolute and per capita terms.

Buddhist violence against Rohingya Muslims in Myanmar, led by monks like Ashin Wirathu and organizations such as Ma Ba Tha, produced what the United Nations characterized as genocide [16]. In Sri Lanka, Buddhist nationalism was invoked throughout the 26-year civil war against the Tamil Tigers and has since turned toward violence against Muslim minorities [16]. Hindu nationalist violence in India — including the 2002 Gujarat pogrom — killed over 1,000 people, predominantly Muslims [16].

The CIA blowback argument is historically grounded. The United States armed and trained the mujahideen who became the core of al-Qaeda during the Soviet-Afghan War. Saudi Arabia's Wahhabist expansion — funded with petrodollars and tacitly supported by Western governments during the Cold War — seeded extremist mosques and madrassas across the Muslim world for decades. The 2003 invasion of Iraq destabilized the region and created the security vacuum from which ISIS emerged.

Critics of Pape's methodology, including scholars at Princeton, have noted that he "sampled on the dependent variable" — studying only cases where suicide terrorism occurred, rather than examining why 58 democratic occupations produced suicide terrorism in only 9 cases [14]. This is a legitimate methodological critique, but it does not invalidate the core finding that occupation is a strong predictor.

Regression Analysis: What Predicts Terrorism?

The PNAS study by Jasko et al. (2022) offers the most rigorous comparative analysis available. Using two unique datasets covering the United States and the world, it found that on the global level, Islamist extremists engage in deadlier attacks than right-wing or left-wing extremists [7]. But the study also identified structural variables — active conflict zones, state fragility, economic deprivation — as significant predictors across all ideological categories.

The Global Terrorism Database at START (University of Maryland) and the RAND Database of Worldwide Terrorism Incidents both show that the strongest predictors of terrorism are: active armed conflict, state failure or fragility, foreign military occupation, and economic deprivation [4]. Religious ideology, when isolated from these confounders, has weaker predictive power than political context.

But "weaker" does not mean "zero." The honest answer from the regression data is that religion functions as an accelerant and a recruitment framework in contexts already primed for violence by structural conditions. It is neither the sole cause nor irrelevant. Countries experiencing identical structural conditions but with different religious demographics do show different terrorism patterns — but disentangling religion from the colonial history, Cold War proxy dynamics, and resource-extraction economies that shaped those demographics is methodologically fraught.

The Media Distortion

A 2019 study by Erin Kearns at the University of Alabama, published in Justice Quarterly, analyzed news coverage of all 136 terrorist attacks in the United States between 2006 and 2015. The findings: controlling for target type, fatalities, and whether the perpetrator was arrested, attacks by Muslim perpetrators received 357% more media coverage than attacks by non-Muslims [17].

Muslims committed 12.5% of the attacks studied but received more than half of all news coverage [17]. Each additional fatality generated a 46% increase in coverage, while attacks against government targets received 211% more coverage [17].

This asymmetry produces a measurable perception gap. Americans consistently overestimate the Muslim share of domestic terrorism by large margins. The over-coverage is not fabricated — the attacks are real — but its disproportionate volume creates a statistical illusion about relative threat levels.

This does not mean global-scale attacks like 9/11, the Bataclan massacre in Paris, or the Mumbai attacks should be minimized. Those attacks were genuinely unprecedented in their ambition and lethality. The point is that the media ecosystem amplifies Islamist violence relative to right-wing or other forms of terrorism in ways that systematically distort public risk assessment.

Source: GDELT Project

Data as of Mar 28, 2026CSV

The Domestic Angle: Surveillance and Muslim Americans

Muslim Americans are among the most surveilled communities in U.S. history. The NYPD's Demographics Unit conducted blanket surveillance of Muslim neighborhoods, mosques, businesses, and student organizations. The NSA's bulk collection programs disproportionately targeted Muslim communications. FBI informant programs have been criticized as amounting to entrapment — the "Newburgh Sting" cases involved informants who provided weapons, plans, and money to defendants who might never have acted independently [8].

Despite — or alongside — this surveillance apparatus, Muslim Americans commit terrorism at rates far below their population share. Muslims represent approximately 1.1% of the U.S. population. Even including the September 11 attacks (perpetrated by foreign nationals, not Muslim Americans), the per capita rate of terrorism among Muslim Americans is not higher than among other demographic groups when measured over the full 1975–2025 period [3].

Countering Violent Extremism (CVE) programs, established under the Obama administration and subsequently rebranded, have faced sustained criticism for disproportionately targeting Muslim communities while underfunding programs aimed at white supremacist radicalization — despite federal assessments identifying the latter as the more lethal domestic threat [8].

The Deradicalization Evidence

Which interventions actually reduce terrorism? The longitudinal data is uneven but instructive.

Saudi Arabia's theological rehabilitation program (the Mohammed bin Nayef Counseling and Care Center) claims low recidivism rates, but independent verification is limited. Critics question whether the program deradicalizes hardcore operatives or merely processes low-level detainees and sympathizers. The lack of consensus on what constitutes recidivism in terrorism contexts makes comparative assessment difficult [13].

Indonesia's community-based approach combines theological engagement with vocational training and entrepreneurship schemes. Indonesian authorities found that "stand-alone deradicalization programs focusing on ideological transformation are not effective" without economic reintegration components [13]. The program has achieved notable successes with some extremist leaders but faces obstacles including limited prison resources, inter-agency coordination failures, and inadequate legal frameworks [13].

Military drone campaigns in Pakistan and Yemen killed between 8,500 and 12,000 people from 2010 to 2020, including an estimated 1,700 civilians, 400 of whom were children [18]. The Bureau of Investigative Journalism recorded civilian casualty figures three to twelve times higher than official U.S. government estimates [18]. Only 2% of total drone casualties were militant leaders [18]. Research on whether drone strikes reduce or increase terrorism recruitment is contested, with evidence on both sides.

Economic development in the Sahel faces the hardest conditions. The region's terrorism crisis has escalated even as international development aid has increased, suggesting that economic interventions alone are insufficient in active conflict zones [5].

The consistent finding across programs: no single intervention works in isolation. Theological engagement without economic opportunity fails. Military force without governance capacity fails. Economic development without security fails. The most promising approaches combine elements — but rigorous longitudinal evaluation of combined programs remains scarce.

The Reform Question

Is Islam undergoing a reformation? Liberal Muslim scholars — Irshad Manji, Reza Aslan, Khaled Abou El Fadl — argue that Islam's interpretive traditions (ijtihad) support democratic governance, women's rights, and religious pluralism, and that Western media systematically ignores the ongoing internal reform movement [11].

Critics respond that these voices remain marginal in the Muslim-majority world and that the "reformation" analogy misunderstands Islam's decentralized authority structure. There is no Muslim Pope who can issue doctrinal reforms binding on all believers. Reform must emerge from thousands of independent scholars, institutions, and communities — a process that is underway but difficult to measure.

Saudi Arabia's recent modernization under Mohammed bin Salman — women driving, entertainment liberalization, reduction of religious police authority — has been cited as evidence of reform. Whether it represents genuine theological evolution or cosmetic authoritarianism remains contested.

The role of Saudi Arabia's Wahhabist export machine deserves particular scrutiny. For decades, Saudi funding built extremist mosques and madrassas across South Asia, Southeast Asia, and sub-Saharan Africa, promoting a puritanical interpretation of Islam that provided ideological infrastructure for radicalization. Any honest accounting of Islamist terrorism must reckon with the deliberate state sponsorship of extremist theology — and the Western governments that tolerated it because Saudi Arabia was a Cold War ally and oil supplier.

Comparing Historical Violence

How does current Islamist terrorism compare to other periods of religious violence?

The European Wars of Religion (1524–1648) killed between 5 million and 18 million people in a continent with a fraction of today's global population [15]. The Thirty Years' War alone may have killed 8 million, reducing the population of the German states by 25–40%. Adjusted for population, Christian Europe's religious violence in this period exceeded anything in the contemporary Muslim world.

The Sri Lankan civil war (1983–2009), fought between the Buddhist-majority Sinhalese state and the Hindu-majority Tamil Tigers, killed between 80,000 and 100,000 people. The Tamil Tigers pioneered the modern suicide bombing before Islamist groups adopted the tactic. Buddhist-nationalist violence against Muslims in Myanmar after 2012 displaced over 700,000 Rohingya [16].

Secular ideological terrorism in Latin America during the 1960s–1980s — including campaigns by the Shining Path in Peru, FARC in Colombia, and various military juntas — killed hundreds of thousands. The Argentine military dictatorship alone "disappeared" 30,000 people.

The point is not whataboutism. The point is that the data does not support the claim that Islam is uniquely violent among world religions or ideologies when examined across historical time. It does support the claim that Islamist terrorism is the dominant form of religiously motivated terrorism in the current historical moment — a distinction that matters for present-day policy without supporting civilizational generalizations.

What the Data Cannot Resolve

The evidence assembled here points to a genuinely complex interaction between theology, geopolitics, economics, and state capacity. The data supports several conclusions simultaneously:

What the data clearly shows:

• Islamist groups have produced the majority of global terrorism fatalities since 2001, and the numbers are not close.
• The overwhelming majority of victims are Muslims living in Muslim-majority countries.
• In the U.S. and Western Europe, far-right extremism has overtaken Islamist terrorism as the primary domestic threat since approximately 2015.
• Terrorism rates vary enormously across Muslim-majority countries, tracking more closely with state fragility, conflict, and poverty than with religious demographics.
• Media coverage amplifies Islamist terrorism relative to other forms by a factor of roughly 3.5x, distorting public perception.

What the data does not resolve:

• The relative causal weight of theology versus structural conditions. Both matter; their interaction is what produces the outcome, and isolating one from the other is methodologically difficult.
• Whether Islam's doctrinal content makes it more susceptible to violent exploitation than other religions at comparable development stages and under comparable geopolitical pressures. The historical comparisons suggest the answer is no, but the counterfactual is inherently unprovable.
• Whether deradicalization works at scale. The evidence is promising for individual programs but insufficient for population-level conclusions.

The honest answer is that anyone claiming a simple explanation for the relationship between Islam and terrorism — whether "Islam is inherently violent" or "this has nothing to do with Islam" — is ignoring half the evidence. The data demands a more uncomfortable position: that specific theological traditions provide the recruitment framework and moral justification for violence, but only in contexts where structural conditions — state failure, occupation, poverty, authoritarian repression, and great-power proxy warfare — have already created the conditions for armed conflict.

Source: Cato Institute

Data as of Sep 11, 2025CSV

Neither side of the political debate wants to hold both of these facts simultaneously. The data requires it.

Methodology Note

This analysis draws on the Fondapol Islamist Terrorism Database (1979–2024), the Global Terrorism Database at the University of Maryland's START Center, the Global Terrorism Index published by the Institute for Economics and Peace, Cato Institute analysis of U.S. terrorism data (1975–2025), CSIS Transnational Threats Project data, the PNAS comparative study by Jasko et al. (2022), Pew Research Center global surveys, the Bureau of Investigative Journalism drone strike database, World Bank economic indicators, and GDELT media monitoring data. Where datasets conflict — which they do, particularly on casualty counts in active conflict zones — this report has noted the discrepancy and used ranges rather than false precision. The lack of a universally agreed-upon definition of terrorism across these databases is a known limitation of all comparative terrorism research.