Research

Use of lethal weapons to kill protesters by Bangladesh police: A discussion analyzing 253 deaths of July 24 revolution

Download PDF

Authors:

Md Abu Bakkar Siddik

The Center for Social Policy and Justice, Dhaka 1200, Bangladesh

Md Syful Islam

Research & Integrated Thought (RIT), Dhaka, Bangladesh

Sheikh Muzzammil Hussen

The Center for Social Policy and Justice, Dhaka 1200, Bangladesh

Md. Khalid Syfullah

Department of Public Administration, University of Dhaka, Dhaka 1000, Bangladesh

Published Date:

January 8, 2026

Keywords:

Lethal force; Police violence; Protesters; July 24 Revolution; Bangladesh

Abstract

The use of lethal force by law enforcement against protesters is a critical issue in Bangladesh, where police actions during the July 24 Revolution resulted in more than a thousand recorded deaths. This study analyses the circumstances, demographics, and geographic distribution of these fatalities through a quantitative content analysis of national newspaper reports of 253 deaths, verified through a multi-step validation process to reduce bias. The findings reveal that young adults aged 18 to 29 accounted for 58.1% of victims, underscoring the historical role of youth in political mobilization. Lethal bullets were responsible for 78.66% of deaths, with most fatalities occurring on the spot (77.47%) and predominantly at demonstration sites (75.89%). The chest (46.64%) and head (29.25%) were the most frequently targeted areas, suggesting intentional use of lethal force inconsistent with international norms on proportionality and necessity. These patterns illustrate how the securitization of dissent discursively frames protesters as threats, legitimizing violent repression consistent with broader theories of state violence. In doing so, the findings extend global debates on protest policing by showing how securitization discursively legitimized the lethal targeting of youth protesters in a South Asian context. By situating Bangladesh within South Asian and global debates on protest policing, the study contributes to understanding how state violence undermines democratic participation and public health. The analysis aligns with the World Health Organization’s objectives on reducing violence and promoting human rights. While limited by reliance on media reports and cross-sectional data, the findings highlight the need for future research on long-term impacts and for context-specific reforms in accountability, training, and oversight.

 

DOI: https://doi.org/10.1016/j.dialog.2026.100277

Full Article

1. Introduction

The Bangladesh Police, established during British colonial rule in 1861, has played a crucial role in maintaining law and order throughout the country’s turbulent history. From the fight for independence in 1971 to the management of internal security challenges in subsequent decades, the police force has been integral to Bangladesh’s state apparatus [1]. Over the years, the Bangladesh Police have been involved in various operations, from counterterrorism to crime prevention [2]. Despite this legacy, the force has often been criticized for its methods, with accusations of corruption and excessive use of force tarnishing its reputation. This historical context sets the stage for understanding the complex relationship between law enforcement and the public in Bangladesh, where the legacy of the police is increasingly marred by allegations of human rights abuses [3]. This historical overview motivates our focus on protest policing, in which questions of legitimacy, proportionality, and accountability are most salient.
The reputation of the Bangladesh Police is further complicated by widespread reports of human rights violations, including extrajudicial killings, torture, enforced disappearances, and the mistreatment of vulnerable groups, including children [2,4]. These abuses are not isolated incidents but rather part of a broader pattern of law enforcement practices that have drawn significant domestic and international criticism. The systemic nature of these violations suggests deep-rooted issues within the policing system, where excessive force and lack of accountability are prevalent [3]. This backdrop of persistent human rights abuses is crucial for understanding the police’s role in the violent suppression of public dissent, as it reflects a troubling trend in the use of state power to stifle opposition and control the population. We therefore read these events through two lenses used in research on contentious politics: securitization of dissent and state violence, which together help explain how exceptional force becomes normalized. Securitization discursively justified the lethal targeting of youth protesters, normalizing exceptional force and embedding repression within the logic of national security.
The July 24 Revolution in Bangladesh began in response to the government’s civil service quota system, which reserved a large share of jobs for the families of 1971 war veterans, which was regarded as unfair and discriminatory among government job aspirants and students, as it limited opportunities for others who were qualified [5]. The protests began with student groups, particularly the “Students Against Discrimination” movement, and soon gained support from broader segments of society. Although economic problems such as inflation, unemployment, and corruption contributed to public frustration, the primary cause of the protests was the demand for equal treatment in government hiring [6]. The government described the protests as a threat to national security and reacted with force. It imposed a curfew, shut down internet access, and used the police, military, and other forces to control the demonstrations [7]. These curfews, communications shutdowns, and deployment patterns are characteristic of a securitizing move that reframes protest as a threat to national integrity and public order. This strong response reflects a pattern in Bangladesh’s history, where peaceful protests are often met with state violence. The events show an ongoing conflict between government authority and the people’s right to express dissent.
The way Bangladesh’s police respond to protests today is shaped by laws and practices that go back many decades. The country still relies on rules from the British colonial period, such as the Police Act of 1861, which focused on controlling the population rather than protecting rights [8]. After independence, new laws like the Police Act of 1973 and the Torture and Custodial Death (Prevention) Act of 2013 were passed to improve accountability and reduce abuse [9]. However, these efforts have often failed in practice. For example, enforcement of the 2013 law remains weak, and there have been efforts to reduce its effectiveness. In recent years, policing in Bangladesh has become more militarized, especially during political unrest. There are still no clear national rules governing the permissible use of force during protests. As a result, violent crackdowns continue, and the deadly response in the July 24 Revolution reflects a broader pattern, not just an isolated event [10]. Against this backdrop of weak guidance on the use of force and episodic militarization, it is not surprising that our data later reveal clustering of on-the-spot fatalities at demonstration sites and concentration in the capital, consistent with escalatory protest policing.
The police response escalated rapidly, with hundreds of deaths reported during July and August 2024 and verified accounts of shootings at close range, including inside residential areas [11,12]. Independent assessments documented widespread violations, including lethal force against unarmed protesters and child casualties [13], [14]. These findings are consistent with those documented by the Office of the United Nations High Commissioner for Human Rights (OHCHR), whose 2025 Fact-Finding Report identified widespread human rights violations, including the use of lethal force against unarmed protesters, child casualties, and killings inside residential areas, during the July and August 2024 protests [10].
Despite the alarming nature of these events, there remains a significant gap in scholarly research that fully examines the correlation between the historical role of the Bangladesh Police, ongoing human rights abuses, and the lethal force used during the July 24 Revolution. While media reports provide detailed accounts of individual incidents, there is a lack of comprehensive analysis that connects these incidents to broader systemic issues within Bangladesh’s law enforcement [15,16]. This article makes three contributions. First, it systematically documents 253 police linked deaths with demographic, anatomical, and spatial detail. Second, it links those patterns to the twin lenses of securitization of dissent and state violence [17]. Third, it situates Bangladesh comparatively within South Asian and global protest-policing debates to assess transferability. We use parallels and contrasts from Myanmar, Hong Kong, and the United States to contextualize Bangladesh, while emphasizing South Asian policing traditions in India, Pakistan, and Sri Lanka as the closest regional comparators [18].
This study aims to: (1) describe patterns of police use of lethal force during the July 24 Revolution; (2) analyze demographic, anatomical, and spatial distributions of fatalities; and (3 interpret these patterns through securitization of dissent and state violence, frameworks that explain how protest was reframed as a security threat and how lethal targeting became normalized as a tool of repression [19,20]. Our analysis aligns with World Health Organization objectives on violence and injury prevention by strengthening evidence on violence as a determinant of health and by informing governance and accountability measures relevant to prevention. We next detail our data and coding strategy, then present results before returning to the conceptual and comparative implications and policy recommendations.

2. Methodology

2.1. Study design

This study employs a quantitative research design, focusing on a content analysis of reports from national newspapers in Bangladesh. The research aims to explore and provide a descriptive overview of the use of lethal force by the Bangladesh Police during the July 24 Revolution and to analyze these events within the broader context of human rights abuses in the country. The descriptive and exploratory study seeks to provide a detailed account of the incidents and identify patterns of police behavior during the protests. The study describes and analyzes the deaths caused by lethal weapons used by the Bangladesh Police, which were reported in Bangladeshi newspapers and might not be documented elsewhere. All figures and tables are original visualizations produced from the authors’ dataset of 253 verified cases.

2.2. Data collection procedures

Data for this study were gathered from various national newspapers (e.g., The Daily Star, Prothom Alo, Daily Manab Zamin, The Naya Diganta) in Bangladesh, selected for their extensive coverage, credibility, and consistent reporting on political and human rights issues, particularly during the July 2024 protests. The data collection process followed a systematic approach, beginning with the identification of 1103 news stories published between July 01, 2024, and August 25, 2024, to capture the immediate aftermath of the protests. Of these 1103 stories, 459 included descriptions of deaths, further examined for relevance. The next step involved filtering these stories to eliminate duplicate entries and those lacking critical variables, yielding 261 news stories with usable content. Finally, after thorough review and analysis, the study documented 253 deaths in the final dataset, focusing on key details such as the number of fatalities, the circumstances of each death, police involvement, and the use of lethal weapons. Fig. 1 illustrates the data collection procedures.
Fig. 1

  1. Download: Download high-res image (69KB)
  2. Download: Download full-size image

Fig. 1. Data collection and verification flow for protest-related deaths, July 1 to August 25, 2024.

Notes: initial story identification, relevance screening, duplicate and incomplete record removal, verification across outlets, and final dataset of 253 deaths. All visuals are original and derived from the authors’ dataset of 253 verified cases.

2.3. Coding protocol and variables

Each death was coded using a predefined codebook covering age group; sex; cause of death (lethal bullet; pellets or rubber bullets; assault; chased or drowning); place of shooting (demonstration site; inside or in front of home; other); anatomical targeting (head; chest; full body pellets; other); death location (on the spot; after hospitalization); and administrative division. The codebook specified operational definitions and decision rules. Where accounts differed, we prioritized the description with the most incredible detail or the dominant formulation across outlets. Ambiguities were flagged for team review and resolved by consensus.

2.4. Data verification

To increase accuracy and reduce duplication, we cross-checked names, locations, dates, and incident descriptors across at least two outlets whenever possible. We removed duplicates using exact and fuzzy matching on name, age, location, and incident date, followed by manual review. Records lacking sufficient information to classify core variables were excluded from the final dataset.

2.5. Biases of media reporting and residual limitations

Reliance on news reports introduces several sources of bias. Selection bias may favour observable events and urban centres, leading to an overrepresentation of the capital relative to rural areas. Censorship and self-censorship risks may suppress or delay coverage during periods of political pressure, creating time-lag bias and potential under-ascertainment of fatalities. Source heterogeneity can produce narrative bias, whereby incidents associated with specific institutions, locations, or victim profiles receive greater coverage. Misclassification is possible in early breaking news, for example, distinguishing pellet from bullet injuries or identifying whether a shooting occurred at a demonstration site or near a residence. We mitigated these risks by cross-validating core variables across outlets, excluding cases with unresolved contradictions, and conducting simple sensitivity checks by (a) re-running descriptive statistics after dropping single-source cases and (b) comparing Dhaka-only with non-Dhaka subsets. Nevertheless, possible underreporting of rural deaths, misclassification of anatomical targeting, and incomplete contextual detail remain limitations, which we acknowledge in interpreting results.

2.5.1. Inclusion and exclusion criteria

Inclusion: (i) articles published between July 1, 2024 and August 31, 2024 that address the July 24 Revolution; (ii) reports that explicitly mention Bangladesh Police involvement; (iii) articles providing sufficient detail to code number of deaths, context, and actions by law enforcement; (iv) newspapers with a reputation for credible reporting on political and human rights issues.
Exclusion: (i) opinion pieces or editorials lacking factual incident reporting; (ii) articles that do not address police use of lethal force or lack sufficient incident detail; (iii) stories that lacked intended variables (age, gender, type of wound, death location, division).

2.6. Inter-coder reliability

A formal inter-coder reliability assessment was not conducted during the acute data collection period. Coding followed a structured codebook, and ambiguities were adjudicated through team discussion, but reproducibility cannot be fully quantified without double-coding. We will implement a post-publication audit in which a second coder independently codes a 10% random subset of cases stratified by division. Agreement will be evaluated with Cohen’s kappa for cause of death, place of shooting, anatomical targeting, and death location, and any boundary definitions will be refined considering discrepancies.

2.7. Data analysis

We used descriptive statistics (frequencies and counts) to summarize the distribution of deaths by demographic, anatomical, spatial, and incident characteristics. Analyses were conducted in SPSS and Microsoft Excel, and the results are presented as bar and pie charts. In interpreting patterns, we consider whether distributions by site, victim age, and anatomical targeting align more closely with crowd-dispersal objectives or with a securitized posture consistent with state violence. This analysis provides an overview of lethal force patterns during the July 2024 Revolution and situates them in comparative perspective using the existing literature on protest policing.

3. Results

Table 1 summarizes the distribution of deaths attributed to police actions during the July and August 2024 protests in Bangladesh. Most victims were young adults, with 58.1% aged 18 to 29 years, followed by 18.97% in the 30 to 39 age group. Children and adolescents accounted for 16.6% of fatalities, while those over 40 years represented 6.72%. This concentration is consistent with the central role of student and youth organizations in Bangladesh’s protest repertoires and helps explain the overrepresentation of young adults among the deceased.

Table 1. Distribution of deaths linked to police action during the July 2024 Revolution in Bangladesh.

Variables Category Count Percentage (%)
Age Child and adolescents 42 16.6
18–29 years 147 58.1
30 to 39 years 48 18.97
Above 40 years 17 6.72
Police involvement Police involvement 253 100
Others 0 0
Death place Spot dead 196 77.47
Dead after being hospitalized 57 22.53
Cause of death Lethal Bullet 199 78.66
Pellets or Rubber Bullets 34 13.44
Assault 8 3.16
Chased/drown 12 4.74
Place of being shot Shot at demonstration site 192 75.89
Inside home or in front of home 44 17.39
Others 17 6.72
Shooting point Chest 118 46.64
Head 74 29.25
Pillet full body 38 15.02
Others 23 9.09
Division Dhaka 92 36.36
Chittagong 43 17
Rajshahi 29 11.46
Khulna 25 9.88
Barishal 30 11.86
Sylhet 9 3.56
Rangpur 17 6.72
Mymensingh 9 3.56
Notes: age group, cause of death, death location (on the spot or after hospitalization), place of shooting (demonstration site; inside or in front of home; other), and administrative division. All visuals are original and derived from the authors’ dataset of 253 verified cases.
The timing and circumstances of death further describe the severity of confrontations. A total of 77.47% of victims died on the spot, while 22.53% succumbed to injuries after hospitalization. Lethal bullets were the predominant cause of death, responsible for 78.66% of cases, followed by pellets or rubber bullets (13.44%), physical assault (3.16%), and drowning or being chased (4.74%) (Fig. 2). These distributions are not consistent with crowd dispersal goals.
Fig. 2

  1. Download: Download high-res image (122KB)
  2. Download: Download full-size image

Fig. 2. Cause of death among 253 protest-related fatalities during the July 2024 protests.

Categories: lethal bullets; pellets or rubber bullets; assault; chased or drowning. All visuals are original and derived from the authors’ dataset of 253 verified cases.
The place of shooting also clarifies where incidents occurred. Most victims (75.89%) were shot at demonstration sites, 17.39% were shot inside or in front of their homes, and 6.72% were shot in other locations. The extension of shootings into residential areas indicates that coercive action was not confined to protest venues and increased risks for bystanders and family members. Anatomical targeting patterns show that the chest (46.64%) and head (29.25%) were the most frequently struck areas, followed by full body pellets (15.02%) and other anatomical sites (9.09%) (Fig. 3). The concentration on chest and head is difficult to reconcile with crowd dispersal objectives and is inconsistent with international principles that require necessity, proportionality, and preservation of life.
Fig. 3

  1. Download: Download high-res image (66KB)
  2. Download: Download full-size image

Fig. 3. Anatomical targeting among 253 protest-related fatalities during the July 2024 protests.

Categories: chest; head; full body pellets; other anatomical sites. All visuals are original and derived from the authors’ dataset of 253 verified cases.
Geographically, fatalities were concentrated in Dhaka (36.36%), followed by Chittagong (17%), Barishal (11.86%), Rajshahi (11.46%), and Khulna (9.88%). The remaining divisions, Rangpur, Sylhet, and Mymensingh, accounted for 3.56 to 6.72% of deaths. This distribution highlights the nationwide scope of lethal incidents. The prominence of Dhaka likely reflects both event intensity and higher media visibility in the capital. Taken together, the age profile, the site of shootings, and the focus on vital anatomical areas indicate a securitized posture toward protest rather than a public order approach centered on graded and minimal force.

4. Discussion

Read through the lens of securitization of dissent, the July 2024 protests display a pattern in which the protest was narrated as a security problem, lowering thresholds for exceptional police powers and normalizing rapid escalation. This framing helps explain why fatalities clustered at demonstration sites and why young adults, who are central to mobilization, were disproportionately represented among the deceased. The concentration of shots to the chest and head is difficult to reconcile with crowd-dispersal objectives and is inconsistent with international principles that require necessity, proportionality, and a constant focus on preservation of life [21]. Taken together, these patterns reflect a securitized posture in which lethal outcomes were normalized as an expected feature of protest management.
A regional perspective clarifies transferability. South Asian protest policing shares institutional legacies, centralized command structures, and episodic militarization. Within this family of practices, Bangladesh’s July 2024 protests results align with a lower threshold for escalation, heavy reliance on firearms in urban demonstrations, and limited external accountability. Similar organizational features across India, Pakistan, and Sri Lanka suggest that lessons about proportionality, graded use of force, and external oversight are likely to travel within the region, even if specific legal frameworks and political incentives differ.
Myanmar provides a closer parallel on lethality and impunity, where security forces repeatedly employed live fire against demonstrators with limited accountability [22]. The Bangladesh data show comparable reliance on live ammunition and a high share of on-the-spot deaths. The broader political trajectories are not identical, but both settings illustrate how securitized narratives, and weak external checks can produce rapid and deadly escalation during periods of contention.
Hong Kong and the United States offer a contrast in which strong legal contestation, civil oversight, and after-action review mechanisms constrained fatalities even when forceful crowd control was used [23,24,25,26]. These contrasts underscore that institutional design, particularly independent oversight and legal contestation, can moderate harm even under conditions of mass protest. The comparison does not imply that non-lethal practices were always followed in those contexts, but it highlights the role of accountability architectures in shaping outcomes under stress.
The findings also speak to generalizability and underserved populations. Within Bangladesh, risks appear higher for children and adolescents, low-income groups, and residents of dense urban neighborhoods. The share of shootings inside or in front of homes indicates permeation of coercive action into private spaces, increasing risks for bystanders and families. Rural deaths may be undercounted because of reporting gaps, which means observed urban concentration may partly reflect visibility rather than incidence. These gradients should inform prevention, service provision, and the design of accountability mechanisms.
As a public health concern, the observed patterns align with World Health Organization objectives on violence and injury prevention, including strengthening data systems on violence, protecting children and adolescents, and supporting governance measures that reduce violent harm. They also resonate with Sustainable Development Goal 16, which emphasizes reducing violence, strengthening accountability, and promoting the rule of law. In the next section, we translate these implications into context-specific recommendations on proportionality standards, non-lethal crowd management, anatomical targeting protocols, oversight mechanisms, and post-incident care for both civilians and police.
This study’s findings align with World Health Organization objectives in three respects. First, by assembling a systematically verified dataset and disaggregating patterns by age, site, and anatomical targeting, it strengthens information systems for violence and injury prevention. Second, the recommendations on national use of force standards, external oversight, transparent reporting, and on scene medical coordination support governance and accountability to reduce institutional violence and improve emergency care. Third, the documentation of child and adolescent fatalities and the call for child sensitive operational guidance advance the protection of children and adolescents. These priorities are consistent with Sustainable Development Goal 16 on peace, justice, and strong institutions.

4.1. Demographic analysis

The analysis revealed that young adults are disproportionately affected, with 58.1% of the victims aged between 18 and 29 years. This finding is consistent with numerous studies that have documented the vulnerability of this age group to police violence and conflict-related fatalities. For instance, in the United States found that young adults, particularly males, are at a higher risk of fatal encounters with law enforcement due to their increased involvement in public protests, higher rates of unemployment, and general societal marginalization [27]. Similarly, research in South Africa has shown that youth are often at the forefront of political demonstrations, making them more susceptible to violent repression [28,29].
The significant representation of children and adolescents (16.6%) in the fatality data is particularly alarming. This mirrors findings from conflict zones, such as in Syria and Yemen, where children are often caught in the crossfire of violence [30,31]. In the case of Bangladesh, the high number of children and adolescents among the dead, some of whom were shot inside their homes, reflects a breakdown in the distinction between combatants and civilians. These patterns suggest that some victims may not have been active participants in the protests but instead bystanders or unintended targets. This raises serious concerns about the precision and intent of police actions, highlights legal and ethical red flags under international human rights law, and indicates a disregard for international norms regarding the protection of minors and civilians in conflict situations. Their dependence on adults for protection exacerbates the vulnerability of children in such contexts, their physical inability to escape dangerous situations, and the indiscriminate nature of crowd control measures like rubber bullets and tear gas, which do not differentiate between adult protesters and minors.

4.2. Police involvement and lethality

Since this investigation encompassed the 253 fatalities attributable to law enforcement, the involvement of the police in each of these instances (100%) was consequently verified found. However, this high number of such fatalities caused by police underscores the critical role of law enforcement in the use of lethal force. This figure is higher than in many other contexts. For instance, in the United States, approximately 1000 people are fatally shot by police each year, representing a smaller proportion of the overall violent deaths, [25]. However, in countries with high levels of political unrest, such as Venezuela or Myanmar, the proportion of deaths attributed to police or military action is comparably high [[32], [33], [34]]. The near-total police involvement in these fatalities suggests a pattern of state-sanctioned violence, possibly linked to efforts to suppress dissent or maintain public order during times of crisis.
Lethal bullets were the primary cause of death, accounting for 78.66% of the fatalities. This significant dependence on firearms by law enforcement aligns with findings from Argentina and Brazil, where police are recognised for their extensive use of lethal force, leading to thousands of fatalities each year. [35,36]. The use of lethal force, particularly in crowd control or protest situations, raises serious ethical and legal concerns, especially when non-lethal alternatives are available. The United Nations Basic Principles on the Use of Force and Firearms by Law Enforcement Officials emphasize that firearms should only be used in situations where there is an imminent threat to life, and all other means have been exhausted [21]. The data from this study suggests that these principles may not be fully adhered to, reflecting a broader global trend of excessive use of force by law enforcement agencies.

4.3. Place and circumstances of death

The study found that 77.47% of the victims died on the spot, while 22.53% succumbed to their injuries after being hospitalized. The high rate of on-the-spot deaths indicates the immediacy and severity of the violence used, with victims likely receiving fatal injuries that precluded the possibility of medical intervention. This finding aligns with research from conflict zones and areas with high police violence, where the rapid escalation of force often leads to immediate fatalities [37].
Most of the shootings occurred at demonstration sites (75.89%), indicating a strong link between public protests and the use of lethal force. This finding is consistent with global trends. Where protests often turn violent, either due to police action or clashes between different groups. In contexts such as Hong Kong or Sudan, mass protests have frequently resulted in high casualty rates, primarily due to the aggressive tactics employed by security forces. 17.39% of victims who were shot inside or in front of their homes highlight the extent to which violence permeates private spaces, suggesting that even those not directly involved in protests can be caught in the crossfire [24,38]. This phenomenon has been observed in countries like Iraq, where security forces have been known to pursue protesters into residential areas [39].

4.4. Anatomical targeting and intent

The study’s findings on the anatomical targeting of victims further underscore the lethal intent behind many of these incidents. The chest (46.64%) and head (29.25%) were the most targeted areas, both of which are highly vulnerable to fatal injuries. Similar patterns have been observed in regions with high levels of police violence, such as in the Philippines during the “war on drugs,” where many victims were shot in the chest or head, indicating an intent to kill rather than incapacitate [10]. The use of such targeting strategies raises critical questions about the proportionality and necessity of the force used, suggesting a possible disregard for the principles of minimal force and preservation of life. Additionally, the anatomical targeting of victims, mainly the chest and head, supports the hypothesis of intent to kill rather than disperse. This reveals a possible deliberate strategy to deter further dissent through fear and intimidation.

4.5. Geographical distribution

Geographically, the data revealed that Dhaka had the highest number of incidents (36.36%), followed by Chittagong (17%) and other divisions. The concentration of fatalities in Dhaka is likely due to its status as the capital and the epicenter of political and social unrest. Similar patterns have been observed in other countries where capital cities, as hubs of political power, often experience the most intense clashes between protesters and security forces (BBC News, 2019). In contrast, the lower percentages in divisions like Sylhet and Mymensingh may reflect less political activity or more effective control measures in these regions.
The regional disparities in fatality rates could also be attributed to differences in local policing practices, the presence of armed groups, and socio-economic conditions. For instance, areas with higher poverty rates or ethnic tensions might experience more frequent and severe police violence, as seen in parts of Nigeria and India [9,8].

4.6. Comparative framing

While global protests often encounter state violence, the methods and scale of such violence differ. For example, in Myanmar, police employed both lethal and non-lethal measures [22]. Bangladesh has experienced similar patterns regarding the use of indiscriminate force, especially in civilian zones, and the absence of transparency in investigations, which suggests a lower threshold for accountability [40]. Unlike Hong Kong, where legal appeals were available, and international and independent investigations were conducted, with an example of Hong Kong police accepting the responsibility for their misconduct [26]. Bangladeshi victims and their families lacked similar recourse previously, but currently, there are efforts to bring the responsible ones to justice. These differences highlight how institutional context affects state responses to civil unrest [40].

4.7. Underlying causes and policy implications

The findings of this study point to several underlying causes that may contribute to the high levels of lethal violence observed. These include systemic issues within law enforcement, such as inadequate training, lack of accountability, and a culture of impunity. In many countries, police forces are often undertrained in non-lethal methods of crowd control and may resort to firearms as a first response rather than a last [41]. Furthermore, the absence of accountability mechanisms allows officers to act with impunity, knowing they are unlikely to face consequences for the excessive use of force. Political factors also play a crucial role. Governments facing political unrest may empower security forces to use whatever means necessary to maintain order, leading to an increase in police violence. This is particularly true in authoritarian or semi-authoritarian regimes, where the government may prioritize stability over human rights [42,43].

4.8. Broader implications for protest rights and civil society

The data indicate that excessive police violence has wider societal effects. Beyond the immediate loss of life, such state violence discourages public participation in protests and fosters a culture of fear. It reduces civic space and erodes trust in democratic institutions. The suppression of youth-led protests particularly threatens future political involvement and strengthens authoritarian control. These trends align with global concerns about the criminalisation of dissent and the erosion of democratic norms in illiberal regimes. The OHCHR also concluded that the use of force during these protests may constitute serious violations of international human rights and humanitarian law, including protections for children and non-combatants. These independent findings highlight the urgent need for accountability and reform [10].

4.9. Recommendations to reduce police-inflicted deaths

The patterns documented in this study suggest a set of reforms that are both evidence-informed and feasible in the Bangladeshi context. Each recommendation is motivated by specific empirical regularities and is framed with realistic sequencing to support implementation.

  • •
    National use of force standard and training on proportionality and necessity: Given the high share of deaths on the spot (77.47%), the concentration of incidents at demonstration sites (75.89%), and the focus on the head and chest (46.64 and 29.25%), a national standard that prioritizes preservation of life, necessity, proportionality, and graded responses is warranted. An interim ministerial circular could be issued while a full directive is developed; scenario-based training and written rules of engagement should accompany adoption, with pilots in two divisions and public reporting on compliance [44].
  • •
    Non-lethal crowd management and de-escalation: Because 78.66% of deaths involved live ammunition, police agencies should expand non-lethal options, define decision thresholds for their use, and conduct practical drills for dispersal and containment. A staged rollout in higher-frequency protest districts, followed by supervisory review of each deployment, would help manage costs and build capacity over time [45].
  • •
    Anatomical targeting protocols and practical drills: The predominance of chest and head injuries indicates a need for explicit guidance that avoids vital areas except when there is an imminent threat to life. Decision-making drills should be integrated into existing firearms qualifications, and any shot to the head or chest should automatically trigger an after-action review [44].
  • •
    Body-worn cameras for accountability and incident reconstruction: Immediacy of fatalities and clustering at protest sites underscore the value of reliable incident records. A pilot in Dhaka and Chittagong with clear activation triggers, retention periods, and secure access logs, audited by an independent body, would test governance and cost parameters before wider adoption [46].
  • •
    Independent civilian review of complaints: The fact that all deaths in the dataset involved police actors points to the need for external oversight. A first phase could establish transparent intake and case tracking with public reporting; a second phase could add investigative authority and referral powers as capacity and trust develop [47].
  • •
    On-scene medical coordination and rapid transfer: Because 77.47% of victims died on the spot, pre-positioning medical teams near anticipated protest routes, designating evacuation corridors, and adopting rapid transfer protocols with local hospitals may reduce mortality. This can be organized through memoranda of understanding between police, health services, and city authorities.
  • •
    Officer wellbeing, empathy training, and psychological screening: Repeated exposure to intense Clashes and lethal outcomes can increase escalation risk. Regular screening, confidential counselling, post-incident debriefing, and empathy-focused training should be embedded in recruitment, promotion, and annual evaluation cycles, ideally with external clinical partners [[48], [49], [50]].
  • •
    Data transparency with child and adolescent safeguards: Children and adolescents comprised 16.6% of fatalities. Routine public reporting should be disaggregated by age, sex, location, and weapon type, and child-sensitive operational guidance should govern actions in the vicinity of schools and residential areas. A monthly, de-identified dashboard would improve transparency while protecting privacy.
  • •
    Alignment with the UN Basic Principles on the Use of Force: National guidance and practice should be harmonized with international principles on necessity, proportionality, and the preservation of life, and accompanied by periodic independent audits and public reporting. This alignment is consistent with the evidence of high on-the-spot mortality and heavy reliance on live ammunition [21].
  • •
    Implementation sequencing: Piloting in Dhaka and one additional division would allow procurement, training, oversight, and medical coordination to be tested under real conditions before national scale-up. Transparent evaluations, including indicators for proportionality and preservation of life, can support iterative improvement and public trust.
These reforms are essential not only for justice but also for rebuilding public trust and fostering a more secure civic environment.

4.10. Limitations and future research

This study relies on contemporaneous newspaper reports, which introduce biases such as selection bias favouring visible urban events, censorship that delays coverage, and misclassification of variables, including injury and location. We mitigated these issues through verification, duplicate removal, and exclusion of contradictory cases; however, rural death undercounting and incomplete context may persist. The cross-sectional design captures immediate outcomes but cannot assess long-term impacts, practice retention, or policy durability, limiting inferences on accountability, trust, or health effects. Reproducibility is constrained; future studies should include re-coding and Cohen’s kappa for key variables. Additional research should triangulate sources, develop geocoded datasets, and include qualitative interviews, especially in South Asia or among children and adolescents. These limitations align with the WHO’s goals on violence prevention, accountability, and the protection of vulnerable groups.

5. Conclusion

In conclusion, the findings of this study highlight the urgent need for comprehensive reforms in policing practices, particularly concerning the use of force. There is a critical need for improved training in non-lethal crowd-control techniques, greater accountability among law enforcement officials, and adherence to international human rights standards. Additionally, addressing the socio-political factors that contribute to unrest, such as inequality, corruption, and lack of political representation, is essential to reducing the incidence of such violence. By addressing these issues, policymakers can work toward a more just and peaceful society in which the use of lethal force is a rare exception rather than the norm. This study’s scholarly contribution lies in its systematic documentation of a large-scale protest crackdown, spatial and demographic analysis of fatalities, and contextualization within global and local frameworks. It lays the groundwork for future accountability studies and encourages new lines of research into the militarization of protest policing in South Asia.

CRediT authorship contribution statement

Md Abu Bakkar Siddik: Writing – review & editing, Writing – original draft, Supervision, Methodology, Formal analysis, Data curation, Conceptualization. Md Syful Islam: Writing – review & editing, Writing – original draft. Sheikh Muzzammil Hussen: Writing – original draft, Visualization. Md. Khalid Syfullah: Writing – review & editing, Writing – original draft.

Declaration of competing interest

The authors declared that they don’t have any conflicts of interest.

Acknowledgment

This work is dedicated to the unsung heroes of Bangladesh’s July 24 Revolution, and to all who sacrificed their lives in the pursuit of justice throughout the nation’s history.

References