The Perfect Enemy | Coding Religious Repression and Disorder: Outcomes and Critical Reflections from the ACLED-Religion Pilot Project - World - ReliefWeb
May 27, 2022

Coding Religious Repression and Disorder: Outcomes and Critical Reflections from the ACLED-Religion Pilot Project – World – ReliefWeb

Coding Religious Repression and Disorder: Outcomes and Critical Reflections from the ACLED-Religion Pilot Project – World  ReliefWeb

Read Time:15 Minute

Between January 2021 and March 2022, the ACLED-Religion pilot project collected and published real-time data on religious repression and disorder in seven countries in the Middle East and North Africa: Iraq, Egypt, Israel, Palestine, Yemen, Bahrain, and Iran. ACLED-Religion also published historical back-coding data for 2020. Overall, ACLED-Religion recorded nearly 14,600 religious repression and disorder events, including over 6,100 harassment events, over 5,200 political violence events, and over 1,500 demonstration events, as well as over 2,000 strategic developments.

This final report explores the main trends in repression and disorder observed in each country covered by ACLED-Religion. Additionally, it presents critical reflections on the methodological challenges that emerged during the pilot project, outlining the coding decisions and sourcing strategies that ACLED adopted to capture religious repression and disorder.

Religious Repression and Disorder Trends

Between January 2020 and March 2022, ACLED-Religion records over 6,800 religious disorder events, mostly concentrated in four countries: Iraq (3,228 events, 47%), Palestine (1,363 events, 20%), Yemen (1,134 events, 17%), and Egypt (737 events, 11%) (see graph below). In each of these countries, the acts of a few religion-based actors drove the most significant religious disorder trends: in Iraq and Egypt, the vast majority of disorder events involved interactions between government forces and the Islamic State; in Palestine, most events involved Israeli settlers and Palestinian rioters; and, in Yemen, most disorder events involved Islah and Houthi forces targeting civilians.

ACLED-Religion also records nearly 7,800 religious repression events, 53% of the total number of events included in the dataset. Among the countries covered by the project, the highest levels of religious repression were observed in Palestine—1,798 recorded events (over 23% of total events) — followed closely by Egypt— 1,764 recorded events (nearly 23% of total events) (see graph below). Religious repression in Palestine mainly focused on religious practice, belief, and political expression.1 In particular, the suppression of Muslim religious rituals — mostly at the hands of Israeli settlers and state forces — remained consistently high between January 2020 and March 2022 (in dark blue on graph below), mainly concentrated around Islamic holy sites such as the complex of Al Aqsa in Old Jerusalem. This form of repression surged notably during Ramadan 2021 in connection with the enforcement of COVID-19 restrictions by Israeli security actors. Similarly, most repression of belief involved the desecration of Islamic holy sites. Meanwhile, widespread Islamic protest prayer gatherings and the repression of religious leaders for their political opinions account for the majority of political expression events.

Violent incidents that occurred in Palestine during Ramadan 2021 also had an impact in Israel, where ACLED-Religion records 870 repression events (11% of total events), mostly connected with the suppression of religious belief and practice (in dark orange and dark blue, respectively, in graph above). Demonstrations against perceived offenses to Islamic beliefs and the sacredness of Al Aqsa also surged in Israel between April and May 2021. Other belief and practice events were mainly associated with the repression of Ultra-Orthodox Jews and ensuing riots amid the enforcement of COVID-19 restrictions.

In Egypt, the majority of religious repression events targeted religious practice and political expression (in dark blue and light orange, respectively, in graph above), with high levels of belief and imposition events also recorded. The suppression of practice remained at similar levels between 2020 and 2021. State enforcement of COVID-19 restrictions mainly involved the Muslim population, whereas Christian religious authorities self-regulated access to places of worship. The vast majority — 83% — of political expression and belief events were connected with state repression of the Muslim Brotherhood, recorded by ACLED-Religion since December 2020. These events have included judicial harassment, discrimination against religious leaders, and repression of activists on trumped-up accusations of Muslim Brotherhood membership. Meanwhile, imposition events — concentrated primarily between June 2020 and February 2021 — were connected with state enforcement of moral values and suppression of non-normative sexualities, underpinned by systematic monitoring of cyberspace.

In Yemen, ACLED-Religion records over 1,100 repression events (14% of total events) between January 2020 and March 2022, 90% of which occurred in Houthi-controlled areas. Houthi repression mainly involved imposition events and the suppression of religious belief and practice for sectarian reasons. Houthi forces suppressed Sunni-leaning rituals, particularly during Ramadan 2021. Later that year, they targeted Salafi mosques and proselytizing movements. Imposition events included moral policing, imposition of Houthi ideology on the religious and educational sectors, and exaction of religious taxes. Non-Muslim religious minorities, which make up less than 1% of the population in Yemen, were disproportionately targeted by the Houthis (*for more on this topic, see this ACLED-Religion *infographic). In Hadi and Southern Transitional Council (STC)-controlled areas, ACLED-Religion captures pandemic-related suppression of religious practice in 2020, and continuous suppression of alcohol consumption and production throughout the whole period of the pilot project.

In Iran, ACLED-Religion records over 930 events (12% of total events), with belief and imposition being the main drivers of religious repression (in dark orange and light blue, respectively, in graph above). Imposition events mainly involved state actors engaging in moral policing under the ‘forbidding wrong’ pretext (*for more on this topic, see this ACLED-Religion *report). This included the imposition of an Islamic dress code and the suppression of gender-mixing, music, and alcohol consumption, as well as anything generally considered by the state to be a ‘Western influence.’ Similar to Egypt, the monitoring of cyberspace also played a crucial role in morality policing in Iran. While imposition events started trending downwards in the second half of 2021, belief trends remained at relatively similar levels and focused on the suppression of religious minorities. State repression of Muslim minorities fell in 2021 compared to the year prior, whereas violations against non-Muslim groups rose by 19% (*for more on this topic, see this ACLED-Religion *infographic).

Iraq is the only country covered by the project where religious disorder events – 3,228 – outnumbered repression events (759 events, 10% of total events). The Islamic State was involved in the vast majority of disorder events, and yet it played a negligible role in perpetrating religious repression. Contrastingly, Shiite militia activity took the form of both religious repression and disorder, with the latter trending upwards in 2021. The main drivers of religious repression in the country were religious belief and imposition events (in dark orange and light blue, respectively, in graph above). State repression of Sunni believers accounted for 95% of belief events, which followed an upward trend in 2021 and the first quarter of 2022. Imposition trends mainly captured two activities: Shiite militias suppressing leisure activities, accounting for almost 40% of imposition events between November 2020 and March 2021; and the government’s reinvigorated imposition of Islamic morality, which accounted for more than 50% of imposition events between late 2021 and the first quarter of 2022.

Although the Bahraini population accounts for less than 1% of the more than 260 million inhabitants in the countries covered by ACLED-Religion, Bahrain was home to 7% — 538 events — of the total number of repression events recorded by the pilot project. The dominant forms of religious repression in Bahrain included suppression of religious practice and expression (in dark blue and gray, respectively, in graph above). Bahrain joins Egypt and Israel as the only countries covered by ACLED-Religion that continuously enforced COVID-19 restrictions on religious rituals throughout 2020 and 2021. In doing so, the regime selectively applied restrictions to the Shiite population, leading to spikes in repression events during Islamic commemorations, particularly during the days leading up to the Ashura season (*for more on this topic, see this ACLED-Religion *report). Similarly, religious expression — in particular the display of Shiite symbols related to Ashura commemorations — was systematically censored by state security actors.

Methodological Challenges

ACLED-Religion builds off the main ACLED dataset of political violence and demonstrations by adding new events, event types, and information about religious dynamics. The pilot project is organized around a core distinction between religious disorder and repression, which prompted specific methodological reflections.

Religious disorder events included in the ACLED-Religion dataset meet two main requirements: the involvement of at least one religion-based actor; and the absence of an explicitly religious dimension to the event. This approach implied a systematic and iterative assessment of the actors’ religious agenda and ideology. Nonetheless, it created clear-cut criteria on which disorder events could be included in the dataset and assigned a religious affiliation to religion-based actors. On the other hand, coding religious repression presented methodological challenges, required ad hoc coding decisions, and new sourcing strategies. These challenges are discussed below.


According to ACLED-Religion methodology, religious repression involves the suppression or imposition of religious activity and belief. This approach demands a systematic assessment of each event’s religious dimension based on two premises: first, no event is purely religious in nature; and second, religious motivations are often omitted by secondary sources. To overcome these challenges, ACLED-Religion codes events based on the outcome of repressive practice, minimizing assumptions regarding the actors’ motivations.

This outcome-based approach was highly effective for coding suppression of religious practice and expression events. For example, the arrest of worshippers during prayer or the interruption of a sermon objectively point to the repression of religious activity. Contrastingly, imposition and belief events proved to be widely dependent on the perpetrator’s motivations, thus requiring further methodological sophistication.2

Moral Imposition Events

The religious dimension of moral imposition is often self-evident or overtly articulated. At times, it is apparent from the context, as in the case of the enforcement of an Islamic dress code in Iran. In other cases, the legislative framework or public statements make the religious dimension more explicit. For example, political authorities in Yemen explicitly prohibited gender-mixing to preserve the country’s ‘religious identity.’ However, in many other instances, state and non-state actors label behaviors as generically ‘immoral.’ The suppression of sex work in Egypt is a case in point: charges and sentences variably refer to the religious notion of zina, or to the secular notion of ‘debauchery.’ This example highlights the challenge of determining if religion is motivating such ‘morality’ events.

To answer this methodological challenge, ACLED-Religion identified trends of moral imposition suspected to be informed by an implicit religious bias for each country covered by the project, and marked them with the [Morality] tag. This decision has significantly expanded the dataset’s inclusivity; in fact, out of nearly 1,700 imposition events coded by ACLED-Religion, around 49% entail the enforcement of moral values without explicit references to religion. However, there are two shortcomings of this approach. First, it requires assumptions around the religious bias of ‘morality’ events. Second, it brings to the fore questions about the limits of civil liberties for which there is no cross-cultural and universal answer.3

Belief Events and Religious Cleavages

In belief events, the perpetrator targets a ‘passive’ victim because of their religious affiliation. At times, the perpetrator’s motivation is outspoken or apparent from the context. As in the case of the deliberate desecration of a place of worship, or the enactment of discriminatory laws specifically targeted against religious minorities. Yet, in other cases, the perpetrator’s motivations are unreported or deliberately disguised. For example, it is widely recognized that the Iranian regime persecutes Baha’is on religious grounds, yet criminal sentences against them often refer to “threats against national security.”

To capture these forms of implicit and ‘structural’ violence against religious communities, ACLED-Religion surveyed potential cleavages among religious groups and identified ‘permanent’4 religious cleavages in Iran, Yemen, and Egypt involving specific minorities. Violence perpetrated against members of these communities is assumed to be religiously motivated and events are automatically included in the dataset. Out of more than 1,900 belief events, 13% are included in the dataset due to the cleavage designation, significantly improving ACLED-Religion’s coverage of minority repression.

As the project unfolded, ACLED-Religion identified additional ‘permanent’ and ‘temporary’ cleavages based on an iterative process of monitoring and assessment.5 For example, ACLED-Religion identified a permanent cleavage in Bahrain to capture state repression of imprisoned Shiite religious leaders. In contrast, in Israel, ACLED-Religion identified a ‘temporary’ cleavage — limited to spring 2021 — to capture state and societal repression of the Ultra-Orthodox Jewish community in connection with their religiously motivated refusal to abide by COVID-19 restrictions. The self-evident shortcoming of religious cleavages is that they imply assumptions about structural violence. Additionally, they pose the risk of including criminal, racial, and political events in the dataset.

Ethno-Religious Groups

Coding violence targeting tribal and communal ethno-religious groups is an unresolved challenge for ACLED-Religion. Due to their multi-layered identities, these groups were neither designated as religion-based, nor as part of a religious cleavage. Rather, the project opted for assessing the religious bonds of the group or the religious dimension of the event on a case-by-case basis. However, disentangling the ethnic, communal, political, and religious dimensions of specific events involving these groups proved to be problematic — at times even impossible — and extremely time-consuming. As a result, religious repression and disorder involving groups such as the Yazidis in Iraq or the Haredi Jews in Israel may be underrepresented in the dataset.

The Paradox of ‘Legal’ Religious Repression

State forces are the main perpetrator of religious repression, involved in approximately 55% of all ACLED-Religion events between January 2020 and March 2022. Although state forces do engage in extrajudicial repressive activities, in most cases, religious repression is exercised through law enforcement. ‘Judicial violence,’ ‘judicial harassment,’ and ‘change to religion law/policy’ events account for approximately 56% of state repression events. Harassment such as ‘discrimination’ and ‘prevention of practice’ is often a consequence of legal actions.

ACLED-Religion methodology considers every suppression of religious practice, belief, and expression as a form of repression, regardless of the state’s legal definitions of such acts.6 The rationale of this approach is twofold. Firstly, it emphasizes the victims’ perspective of what is an offense against their religion. And secondly, it allows for systematic mapping of structural violence, showing how repression is often enshrined in legislative frameworks and law enforcement. The value of this approach is well exemplified by two cases: the repressive use of ‘anti-terrorism’ laws; and the enforcement of COVID-19 restrictions.

‘Counter-terrorism’ as a Repressive Practice

After the 11 September 2001 attacks on the United States, several governments in the Middle East and North Africa region adopted the language of the US-led “global war on terrorism” exploiting broad definitions of ‘terrorism’ to justify repressive practices and restrictions on fundamental rights (Hicks, 2021). The coding of repressive practices disguised as ‘counter-terrorism’ activity poses multiple methodological challenges, including reporting biases. Addressing these methodological challenges required an individual configuration of coding decisions and sourcing for each country.

In Iraq, international reports corroborate the reported use of ‘anti-terrorism’ laws to target Sunni citizens — especially in Sunni majority provinces such as Anbar, Salah al-Din, Kirkuk, and Ninewa (USCIRF, 2020; United States Department of State, 2020). However, local news agencies — especially pro-government ones — tend to disguise this trend as a legitimate security campaign against the Islamic State. Triangulating alternative sources – such as the Sunni Association for Muslim Scholars – with pro-government news agencies, ACLED-Religion recorded nearly 190 events targeting Sunni Muslims under trumped-up terrorism charges in eight Iraqi provinces.

In Egypt, the regime has used accusations of Muslim Brotherhood membership to suppress political dissent for years (CFR, 15 August 2019). Since December 2020, ACLED-Religion has considered this trend a form of religious repression,7 recording over 500 events targeting the Muslim Brotherhood. The use of critical sources, including reports from human rights organizations, combined with ACLED-Religion’s ‘stigmatized affiliation’ variable, allowed for an assessment of civil society repression under trumped-up charges of Muslim Brotherhood membership, highlighting 180 such cases.

COVID-19 Restrictions as a Repressive Practice

The COVID-19 pandemic prompted most governments across the world to enforce preventative measures and restrictions to safeguard and benefit public health. Nonetheless, some regimes used COVID-19 protocols to arbitrarily suppress the religious freedoms of select groups (Foreign Policy, 21 July 2020). Moreover, some religious communities refused to abide by COVID-19 restrictions on religious grounds. ACLED-Religion extended its outcome-based approach to pandemic-related events, coding the suppression of religious practice regardless of the perpetrator’s motivations. This approach guaranteed the full coverage of COVID-19 repression and allowed for a meaningful comparison across countries.

Overall, ACLED-Religion recorded over 1,300 pandemic-related repression events. Although authorities decreed public health measures in all countries examined, the data indicate that only Egypt, Bahrain, and Israel significantly enforced COVID-19 restrictions focused on religious rituals (*for more on this topic, see this ACLED-Religion *report). Some regimes instrumentalized the pandemic to intensify religious repression. Bahrain is a paradigmatic case as restrictions were selectively used by the regime to repress the Shiite population. Among the three countries examined, Bahrain records the highest number of pandemic-related ‘judicial harassment’ events. The targeting of Shiite believers accounts for more than 60% of the pandemic-related events in the country.

A major shortcoming of an outcome-based approach applied to COVID-19 events is that restrictions effectively implemented for the benefit of public health were still coded by ACLED-Religion as religious repression. However, this choice was deliberate and aimed at capturing the repressive dimension of the restrictions as perceived by the victims and potential abuses by state forces.


Groups and individuals use different ideologies to coalesce around a cause and, in some cases, engage in violence and demonstrations. Religion is one of these ideologies. However, all events coded in the ACLED-Religion dataset occur within the larger context of the political and social environment of a country and, therefore, no event is purely religious in nature. Acknowledging this limit, ACLED-Religion determined relevant criteria to include religious disorder and repression events in the dataset, abiding by two main methodological principles: reducing assumptions to a minimum; and adhering to an outcome-based approach.

ACLED-Religion was created to test the potential for real-time data collection on religious repression and disorder and determine avenues for future research. The project achieved these objectives: it recorded over 14,600 religious repression and disorder events across the seven pilot countries and it published regular data-driven analysis tracking the latest trends. Despite the end of the pilot project, the full ACLED-Religion dataset for January 2020 to March 2022 will remain available for further research. Lessons learned can inform programming that aims to build on these results going forward.