Published on: 8 September 2023 | Last updated: 25 April 2024
About ACLED & Background on US Coverage
What information is captured?
ACLED systematically collects the dates, actors, locations, fatalities, and types of all political violence and demonstration events in the US using a methodology consistent with our global coverage of conflict and disorder around the world. ACLED also collects information on specific, contextually-relevant strategic developments that offer further insight into the political environment; these may be precursors to changes in patterns of political violence and demonstrations (see below for further descriptions of these various types of events).
What information is not captured?
ACLED does not record all instances of violence. For example, criminal violence, defined as violence that is motivated by personal or purely criminal motives, is excluded from the ACLED dataset. Some mass shooting events fall into this category (see below for more on mass shootings). Violence that takes place in the private sphere, such as domestic or interpersonal violence, is also not recorded in the ACLED dataset, even when these events could have wider repercussions among the public. Similarly, events that are considered to be standard police enforcement are excluded from ACLED’s coverage. These typically include incidents where law enforcement agencies appear to have used violence within the bounds of the legal constraints on their activity, either in reaction to an attempt on the life of a police officer or otherwise in the presence of a threat (see below for more on police brutality). Finally, ACLED only captures events that are reported to have actually occurred. As such, ACLED Researchers do not record threats of violence or intimidation. Non-physical violence, such as online or cyber-violence, is also outside of the scope of ACLED’s data capture and mandate.
Please note: ACLED does not collect what is commonly referred to as ‘big data,’ such as phone records or internet browsing habits, in our data collection. ACLED does not include personally identifiable information about participants in any events, except for rare cases involving public figures (e.g. politicians, leaders of movements, etc.). For more information about our do no harm and conflict sensitivity policies, please contact us at [email protected].
Who is coding these data?
All ACLED data are coded by a team of ACLED researchers. ACLED’s team of over 150 researchers is based all around the world. They review thousands of sources in over 75 different languages on a weekly basis. Each researcher collects data for a specific country or set of countries. They are recruited for their specific language and contextual knowledge. If you are interested in joining the ACLED team, check out our Careers Page.
Has ACLED covered the US before?
Yes, ACLED covered the US originally as part of a three-month pilot project during the summer of 2019; read more about preliminary data and pilot findings in the capstone report. That pilot served as a foundation for ACLED’s current US data collection. The data produced during the course of the pilot are not publicly available at this time. More recently, coverage of 2020 began as part of the US Crisis Monitor initiative – a joint project between ACLED and the Bridging Divides Initiative (BDI) at Princeton University. The project sought to provide the public with real-time data and analysis on political violence and demonstrations around the country, establishing an evidence base from which to identify risks, hotspots, and available resources to empower local communities in times of crisis (to learn more about the former project, see the now-archived project page).
Sources of Information
Where does ACLED’s information for the US come from?
The ACLED team reviews over 2,800 sources to collect information on events in the US. In addition, ACLED constantly reviews new sources to determine their ability to provide information on distinct events; therefore, it is continuously expanding the scope of coverage. When a new source is identified and is determined to yield many events, the source’s events will first be supplemented in the older data before being released. This limits artificial spikes in the data (for more on ACLED sourcing, see this methodology primer).
Does ACLED use crowdsourcing?
No, ACLED does not crowdsource information – such as through broad scraping Twitter for tweets – given sources of such information cannot be verified. ACLED does use social media in its coverage; however, this is limited to only trusted accounts, such as those of journalists.
How does ACLED verify its information?
In the age of ‘fake news,’ this question is both ubiquitous and critical. ACLED researchers assess thousands of sources in their data collection process. Sourcing for each country is distinct and based on the particular country’s media landscape and conflict dynamics. This deliberate sourcing methodology prioritizes an assessment of each source’s reliability prior to inclusion into the source list. As noted above, ACLED does not use crowdsourcing (for more on ACLED sourcing, see this methodology primer). During weekly coding, ACLED researchers find that many events are reported several times by multiple sources. The details of each report may differ, but ACLED researchers only extract factual information about the event: who was involved, where did it happen, when did it occur, what occurred, and in what sequence (see below for more on why certain event details are not systematically coded by ACLED). It is rare to find events that fall within ACLED’s catchment in the US context that are reported by only a single source, without any references to triangulate the information. In those cases, additional research occurs to verify – where possible – the details of the occurrence. As such, ACLED’s information verification process centers on a triangulation system of (1) source review, (2) coding and sourcing review, and (3) event reporting review.
Does ACLED have any US data partnerships?
Yes, ACLED has established a number of partnerships, particularly in response to a noticeable deficit in coverage of militia activity outside of violent events and demonstrations.
MilitiaWatch is a research project and blog that tracks, documents, and analyzes contemporary US militia movements and provides reports connecting long-term militia trends to broader political events. MilitiaWatch gathers data from open source and semi-open source content created by and for militia members, allowing for an analysis of militia activity from the perspectives of the militia members themselves. MilitiaWatch partner data typically provide ‘Strategic development’ events, such as recruitment drives, training exercises, the creation of new groups or splinter groups, or important announcements, such as a militia’s support for a political group. Furthermore, MilitiaWatch may provide more specific details as to the identity of militias active during demonstrations, allowing for the improvement of already published data.
Live Universal Awareness Map (Liveuamap) is a leading independent global news and information site dedicated to factual reporting of a variety of important topics, including conflicts, human rights issues, protests, terrorism, weapons deployment, health matters, natural disasters, and weather-related stories, among others, from a vast array of sources.
The Network Contagion Research Institute is a not-for-profit organization that uses big-data approaches to detect and map the false narratives and motivated reasoning that undermine institutions, threaten vulnerable communities, and harm democracy.
Additionally, ACLED’s US coverage began as part of the US Crisis Monitor, a joint project with the Bridging Divides Initiative (BDI), a non-partisan research initiative based at the Princeton School of Public and International Affairs, co-hosted by the Empirical Studies of Conflict and Liechtenstein Institute on Self-Determination. BDI works with a network of over 3,000 organizations to build community resilience and to track and mitigate political violence in the US.
ACLED has also previously worked with Political Research Associates (PRA), a national non-profit which provides research, strategy, and briefings to expose movements, institutions, and ideologies that undermine human rights. PRA has been tracking incidents of far right and paramilitary interference with racial justice and police accountability protests since Minneapolis police killed George Floyd in May 2020 by collecting secondary data through the monitoring of various online forums and media reports. PRA data have been used to enhance ACLED’s coverage and analysis of militia group activity and hostile political activity in the US. PRA data typically take the form of Strategic development events as well as demonstrations featuring both violent actors, such as militias, as well as non-violent political activity, such as voter intimidation and suppression.
ACLED continues to pursue partnerships with other organizations or research projects whose data or expertise may improve the quality of the US dataset. This section will be updated once new partnerships are established.
Political Violence
How does ACLED define political violence?
ACLED defines political violence as the use of force by a group with a political purpose or motivation. ACLED records political violence through its constituent events, the intent of which is to produce a comprehensive overview of all forms of political disorder, expressed through violence and demonstrations, within and across states. A politically violent event is a single altercation where force is used by one or more groups toward a political end. ACLED employs this definition of political violence in every country covered. It is a core component of the established global methodology. (for more on ACLED terminology, see the ACLED Codebook).
What types of political violence does ACLED capture?
In the US context, ACLED’s coverage of political violence includes, but is not limited to: excessive use of force by police; mass shootings; non-state militias operating in both urban and rural contexts (especially those which consider themselves local security providers); violent mobs that seek to take justice outside of state law; violence involving extremist groups (such as white nationalists, far-right groups, and far-left groups); hate crimes against minority and vulnerable populations, such as women, LGBTQ+ groups, religious groups, race or ethnic groups, etc.; and others.
Does ACLED record mass shootings?
Yes. However, only mass shootings (shooting events with three or more victims) that are designed to terrorize a population or specific group (rather than target(s) personally connected to the perpetrator) are included in the dataset. This means that an event such as the shooting on 3 August 2019 in El Paso, Texas, where a white nationalist opened fire inside of a Walmart shopping center, killing 22 civilians including eight Mexican nationals, would be included given its political links and its intent to challenge public security. However, an event such as the shooting in Dayton, Ohio on 4 August 2019 that resulted in the deaths of nine people would not be included as the motives for the attack remain uncertain and likely linked to personal dynamics, given one of the victims was the shooter’s sister.
How does ACLED code ‘lone wolves’?
ACLED has introduced a new actor — ‘Sole Perpetrator’ — to the data in the US context specifically. Those familiar with ACLED know that elsewhere in the data, when political violence is perpetrated by unknown agents, these agents tend to be in a group, and are referred to as an ‘Unidentified Armed Group.’ In the US, however, the most lethal cases of political violence are mass shootings, which are often carried out by a single person described as a ‘lone wolf’ without a clear affiliation to a specific named group. Hence, attributing such shootings to an ‘Unidentified Armed Group’ would not be accurate given the sole perpetrator of such events. Despite being unaffiliated to any group, these sole perpetrators frequently connect to a kind of political subculture or social movement that motivates their actions, making the tracking of their activity important. The introduction of this new actor for the US context underscores how definitions of political violence must be flexible to allow for accurate representations of political violence across varied contexts, and how ACLED works to develop context-specific methodological approaches while still ensuring standardization across contexts to allow for comparability.
Does ACLED capture political violence involving both far-left and far-right groups?
Yes, ACLED codes activity by any group engaging in political violence, regardless of affiliation or ideology. By definition, a group must first engage in a political violence event in order to be added to the dataset.
How does ACLED attribute events to groups?
Reports will often note the stated affiliation of an armed actor according to the perpetrator as noted once the perpetrator is in custody, by a manifesto they may have left behind, and/or through known associations. ACLED uses these associations, regardless of whether the larger movement ‘claims’ an event or not. Many of the actors active in the US context are affiliated with movements with no clear leader or formal membership. As such, ACLED assumes that the stated affiliation of the armed actor is the motivating impetus for an event and codes the actor as such.
Does ACLED capture police brutality? And if so, how?
Yes, ACLED includes events of police brutality where the details conform to ACLED’s global standards of political violence. Such events are coded with the ‘Violence against civilians’ event type and ‘Attack’ sub-event type, with police forces and civilians coded as the two primary actors. Violent policing is often particularly excessive towards select groups in the US, but much of this policing is within the bounds of the law — and these bounds can vary, with different levels of protections extended to law enforcement by state legislatures, in addition to the federal government. As such, only police engagements that are explicitly outside of established legal parameters are included, however unfair the existing constraints on police behavior may be. In practice, this means that the following types of events are the only ones categorized as such: (1) events in which an individual was not engaging in a crime, yet was seriously hurt or killed, e.g. the case of Breonna Taylor, who was fatally shot by police on 13 March 2020 in Louisville, Kentucky after entering her apartment on a no-knock search warrant; (2) events in which an individual was (assumed to be) engaging in a crime though was visibly unarmed, yet was seriously hurt or killed, e.g. the case of George Floyd, who died of asphyxiation on 25 May 2020 in Minneapolis, Minnesota after a police officer knelt on his neck for nearly eight minutes, despite Floyd’s pleas that he could not breathe, in response to Floyd’s arrest after his use of a counterfeit bill; or (3) events in which an individual was in pursuit of a crime and was seriously hurt or killed, but without evidence that the suspect in question posed a risk based on subsequent reports, e.g. the case of Sean Monterrosa, who was fatally shot by police on 2 June 2020 in Vallejo, California on the assumption that he had a gun, despite the fact that Monterrosa only had a hammer in his pocket and was kneeling with his hands above his waist at the moment he was shot.
Does ACLED capture hate crimes?
ACLED concentrates on collecting group activity, not the activity of individuals. As such, ACLED captures hate crimes when they are perpetrated by groups, or single members of groups, and only codes an event perpetrated by a “lone wolf” without group affiliations in a public “mass” attack (for more on the coding of mass shootings, see above). Hate crimes can be found in the data coded with ‘Event type’ Violence against civilians — with some rare events coded with ‘Sub-event type’ Mob violence if the perpetrator of the attack is a spontaneous, violent mob.
Does ACLED capture the presence of armed demonstrators?
All demonstration events in which firearms are present, whether those armed are participants in the demonstration (hence coded as an actor in the demonstration event) or are present yet do not engage in the demonstration (hence not coded as an actor in the demonstration event, though noted in the Notes), will include a tag “armed” in the ‘Tags’ column to denote the presence of firearms at the demonstration. This tag is applied regardless of whether the firearms are used or simply carried. For more information, see this page on Methodology and coding decisions in the United States.
Demonstrations
Which events in the ACLED dataset are ‘demonstrations’?
Events coded with ‘Event Type’ Protests include three specific sub-types, which appear as ‘Sub-event Types’ in the data: Peaceful protest, Protest with intervention, and Excessive force against protesters. Events coded with ‘Event Type’ Riots include sub-event type Violent demonstration. For more on definitions around ACLED terminology, see the ACLED Codebook.
How does ACLED define and code demonstrations?
ACLED codes all physical congregations of three or more people (single-person demonstrations are not coded) as a demonstration when they are directed against a political entity, government institution, policy, group or individual, tradition or event, businesses, or other private institutions. This includes demonstrations affiliated with an organization (e.g. NAACP), a movement (e.g. Black Lives Matter), or a political party (e.g. Republicans), as well as those affiliated with identity groups (e.g. LGBTQ+, women, Native Americans). Whenever such salient identities exist, they will be coded as an ‘Associated Actor’ to the respective primary actor (for more on coding decisions, see the ACLED Codebook). In addition, ACLED also codes demonstrations around a certain topic, even if not associated with a specific identity group or organization (e.g. against climate change, anti-vaxxers, COVID-19 restrictions, etc.). For more on what does not constitute a demonstration per ACLED methodology, see below.
What ‘demonstrations’ does ACLED not cover?
Given the above-outlined definition, political or party rallies, town hall meetings, and caucuses are not coded as ‘demonstrations’ by ACLED, as they reflect regular political activity by members of political organizations, civil society, and the general public. ACLED covers the occurrence of events, not the absence of action; this means that physical congregations of people are coded, while a labor strike where workers stay at home is not (though significantly large strikes would be captured by ACLED, though as ‘Strategic developments’, explored further below). Symbolic public acts are also not coded as ‘demonstrations’ — such as displays of flags, putting up a sign in one’s yard, written chalk messages on sidewalks, a congressional walkout, etc. Additionally, vigils that are not intended to manifest any protest message also do not fulfill ACLED’s requirements for inclusion.
What does a Violent demonstration entail?
Violent demonstrations denote demonstration events in which the demonstrators themselves engage in violently disruptive and/or destructive acts targeting other individuals, property, businesses, other rioting groups, or armed actors. Such demonstrations can involve engagement in violence (e.g. fighting back against police), vandalism (e.g. property destruction), looting, road-blocking using barricades, burning tires or other materials, amongst others. Demonstrators active in Violent demonstrations are denoted as Rioters as the primary actor (for more on ACLED coding, see the ACLED Codebook). They may begin as peaceful protesters, only turning to violence after facing violence from others (e.g. police) or after agents provocateurs instigate violence (more below); or they may be intent on engaging in spontaneous and disorganized violence from the beginning of their actions. Contrary to organized armed groups, Rioters typically do not use sophisticated weapons such as guns, knives, or swords (with some exceptions on an individual level). A large number of “armed protesters” therefore either have their affiliation with an organized armed group noted in the corresponding ‘Associated Actor’ column (if peaceful), or are instead coded as militias directly if engaging in armed violence (see “How does ACLED code ‘armed protesters’?” section below).
What does Excessive force against protesters entail?
ACLED codes events as ‘Excessive force against protesters’ only when demonstrators are not engaging in any sort of destructive or disruptive activity, yet are met with excessive force. Excessive force is categorized only as violence that is lethal, has the propensity to be lethal, or results in serious injuries/hospitalization. This means that the use of force such as tear-gassing or the use of rubber bullets are not coded as ‘Excessive force against protestors’, and are rather coded as ‘Protest with intervention’, unless they result in fatalities or serious injuries, e.g. someone being blinded by rubber bullets (for more on this coding decision, see below).
Why doesn’t ACLED code all use of force (including tear-gassing, use of rubber bullets, etc.) as ‘excessive force’?
As outlined above, only the use of force that is lethal (e.g. live fire), has the propensity to be lethal (e.g. car ramming), or that which results in serious injuries/hospitalization (e.g. being blinded by a rubber bullet) is coded as Excessive force against protesters. This is not meant to be a normative evaluation of the justified use of force. Rather, this is to stay in line with ACLED’s coding of the use of force in demonstrations around the world. The use of tactics and equipment like tear gas are commonplace around the world. ACLED strives to use a conservative coding of ‘excessive force’ in order to be able to best capture the ‘outliers’ when it comes to the suppression of demonstrations. In this light, the coding of Excessive force against protesters in the US context abides by the same methodological rules in order to ensure cross-country comparability within ACLED data. Events in which actors (police or others) physically intervene with peaceful protesters are coded with ‘Sub-event type’ Protest with intervention; such events include intervention ranging from the use of tear gas or rubber bullets (without reports of serious injuries/hospitalization) to arrests, with such details about the event included in the Notes.
Three different protests happened in my city yesterday; why do I only see one protest event in the data?
ACLED is an event-based dataset, and therefore only records demonstration events. Events are disaggregated by date (when it happened); actors (who it involved); location (where it happened); and event type (what happened). In practical terms, this means that only events taking place on different days, involving different types of violence, with different types of actors, or in disparate locations are considered separate events. Events that share the same date, actors, a proximate location, and event type will be aggregated when it cannot be clearly discerned that the events were discrete and independent. The number of demonstrations is reliant largely on reporting and the terminology used in doing so. For example, five separate demonstrations happening in Atlanta around a single topic within a few blocks of each other may be reported in a newspaper as “demonstrations happened in Atlanta” or “five demonstrations happened in Atlanta.” Both are ‘correct’ in their terminology, but if they are coded differently as a result (1 vs. 5), this would introduce a bias. ACLED codes an event based on an engagement in a specific location, such as at a city level (e.g. Atlanta) on a specific day in order to avoid such biases. The number of ‘Demonstration’ events recorded by ACLED may differ from the number of ‘demonstrations’ recorded via other methodologies by other datasets.
How do you capture demonstrations that last multiple days?
ACLED is an event-based dataset, as outlined above. Each day of a multi-day demonstration is therefore treated as a distinct event given the difference in date. This means, for example, that a demonstration about the same topic involving the same protesters in the same city, lasting an entire week, would be coded as seven distinct events, with the same actor and location information. The difference is the distinct days reflecting each day the protesters congregated. ACLED would only code a week-long demonstration as such if reports indicate that demonstrators indeed met on every day that week — taking a conservative approach so as not to inflate event counts.
Why does ACLED call some demonstrators Rioters and others Protesters?
This is not a normative distinction, but rather a methodological one. Protesters refers to demonstrators that are peaceful, and not engaging in any destructive or disruptive behavior (e.g. violence, vandalism, looting, etc.) Rioters, on the other hand, refer to demonstrators engaging in violence, vandalism, looting, etc. If the character of an event changes from non-violent to violent, then the event will be coded as how it ultimately ended (i.e. violently). However, if two events occur concurrently in the same or a similar location, and it is clear that they are distinct, independent, and differ with respect to peaceful protest or rioting, two events are coded. If reports note the presence of any violence, vandalism, looting, etc. by demonstrators, then the event is coded with Rioters rather than Protesters. It is important to note, however, that an event noting Rioters does not indicate that every single demonstrator engaged in violence.
How does ACLED code demonstrations that have been ‘infiltrated’?
Reports of infiltrations in demonstrations are very difficult to verify, as demonstrators may themselves be violent for various reasons. ACLED relies on its sources to investigate and report on instances of suspected infiltrations, and only codes such events when the evidence is sufficient. If an infiltration is reported within a demonstration, and the infiltrators are the only individuals engaging in violence, then the event is coded as a separate event (a ‘Violent demonstration’) from the ‘Peaceful protest’. Depending on information provided by the source on the identity of the individual(s), the infiltrating actor is coded as either a Sole Perpetrator (for lone-wolf style attacks), an Unidentified Communal Militia (for reports of local “militia” groups), an Unidentified Armed Group (for reports of unidentified armed persons, or suspected members of radical right-wing or left-wing groups), or the specific group name if known (e.g. Boogaloo Boys). Often, in cases where infiltrators are reported, they serve as instigators of violence, and an initially peaceful protest turns violent. Such cases will be coded as a single ‘Violent demonstration’ — even if the violence was instigated by an infiltrator, given that (some) demonstrators themselves engaged in violence, vandalism, looting, etc. as well. Events in which reports suggest such instigators were present are denoted with a ‘Tag’, “suggested agents provocateurs” in the Tags’ column of the event. The “suggested agents provocateurs” tag is applied to events from 1 January 2020 to 31 December 2020 in the data.
I saw details about a protest being organized; why don’t I see it in the data?
ACLED only codes reports of a demonstration actually occurring, not announcements of planned demonstrations. However, ACLED Researchers use reports of planned protests to identify reports of whether demonstrations actually occurred through supplemental research. If there are reports that it did indeed occur, then the event will be recorded.
Does ACLED capture journalists targeted during demonstrations?
Yes. Journalist targeting often occurs within the context of demonstrations that they may have been trying to cover. As such, this targeting will appear within the context of the demonstration event; it will not be coded as a separate and distinct event as that would lead to double-counting of the same event. To identify such events, look for demonstration events (for more on how to identify those, see above) which list both Journalists and Civilians as an ‘Associated Actor’. Events listing Journalists alone as an ‘Associated Actor’ to Protesters or Rioters refer to events in which journalists themselves may be demonstrating.
How can I find which demonstrations involve car rammings?
Car rammings against peaceful protesters are coded as Excessive force against protesters, given the propensity for such attacks to be lethal. If a single person drives their car into demonstrators, the primary actor is coded as Sole Perpetrator (United States) against Protesters, unless the actor belongs to a specific mentioned group (for example, the Ku Klux Klan). Such events include a tag “car ramming” in the ‘Tags’ column. ACLED only codes events as ‘Excessive force against protesters’ resulting from car ramming when a car was driven into protesters on purpose, or is under investigation due to suspicion of purposeful intent (with the event updated if/when new information comes to light). This means that events in which a car accidentally hits protesters (e.g. a car is surrounded at an intersection, slowly rolling forward) are not coded as ‘Excessive force against protesters’; because such events are not intentional, the driver is not driving with enough force to cause serious injury, and there are therefore no serious casualties reported (in the event of serious casualties, the event would indeed be coded as ‘Excessive force against protesters’). If a car is driven into demonstrators who are engaging in violence or destructive activity (e.g. Rioters), the event is coded as ‘Violent demonstration’, rather than ‘Excessive force against protesters’, and the same metrics mentioned above (i.e. the driver’s intent and/or serious casualties) are used to determine whether the event should qualify for a second actor (as above, ‘Sole Perpetrator’ or the specific group) and the “car ramming” tag.
How are counter-demonstrations coded?
If two demonstrations occur at the same time, in the same place, and they are essentially ‘counter’ to one another — for example, one is pro-vaccination, and the other is anti-vaccination — the events are coded as a single event instead of as two separate events. Depending on whether each side engages in destructive or disruptive behavior, each side is coded as Rioters or as Protesters. As such, the ‘Interaction’ code for such events can be 55, 56, or 66 (for more on ‘Interaction’ codes, see the ACLED Codebook). If both sides engage in violence, the event is coded as a ‘Violent demonstration’; if only one side engages in violence against the other, the event is coded as a ‘Protest with intervention’ or as ‘Excessive force against protesters’, depending on the degree of violence; and if both sides are peaceful (i.e. they do not engage in violence against one another), the event is coded as a ‘Peaceful protest’, respectively. Sometimes, other parties, such as police, may get involved in such a demonstration — on behalf of one side or the other, or to ‘keep the peace.’ In such cases, police may be coded as one of the primary actors. All events that capture a counter-demonstration include a tag “counter-demonstration” in the Tags column, which can be used to identify all counter-demonstrations.
Can I analyze the broad ‘Stop the Steal’ and ‘Count Every Vote’ movements post-Election Day 2020?
Yes. Because both movements represent important trends in post-election unrest, ACLED has enabled users to easily analyze them through the use of tags in the ‘Tags’ column. Tags are used instead of ‘Associated Actors’ because they are meant to capture the broader movements which lack a unifying group term or specific actor, and are often organized under different names (e.g. similar demonstrations are organized under the ‘Count Every Vote,’ ‘Respect the Results,’ ‘Count All Votes – Defend Democracy,’ etc. movement names). Broadly, the “count every vote” tag captures all demonstrations whose goal it is to call for all votes to be counted, including valid mail-in/absentee ballots counted after Election Day 2020. In opposition, the “stop the steal” tag captures all demonstrations which call for action against alleged Democratic voter fraud and ‘illegal’ ballot counting, as claimed by President Donald Trump and other Republican officials. Since both movements are essentially counter to one another, many demonstrations featuring one or both of these tags will also feature the “counter-demonstration” tag if opposing groups demonstrate at the same location (see previous section).
How does ACLED code ‘armed protesters’?
Demonstrators are typically, per ACLED definition, spontaneously organized and unarmed (or, often crudely armed if engaging in riotous behavior). However, on occasion, demonstrators can be armed with more sophisticated weapons in certain contexts. In the US in particular, demonstrators at times appear with firearms; under these circumstances, their actions dictate how they are coded. If ‘armed protesters’ engage in a demonstration peacefully, and do not appear to be part of any organized group, they are coded as Protesters. If ‘armed protesters’ are seemingly part of an organized militia group (named or otherwise) yet they remain peaceful, the ‘Associated Actor’ to Protesters is coded as either Unidentified Communal Militia, if the specific group name is not named, or as the specific group name if known (e.g. Proud Boys). On rare occasions, demonstrators who are members of a militia may engage in low-scale violence (e.g. pushing, punching, spitting, etc.) yet still remain ‘demonstrators’ within that context. However, if militia members engage violently with demonstrators using weapons, or in another large-scale and organized fashion, they are coded as a militia (by name if known, or as an Unidentified Armed Group or Unidentified Communal Militia, depending on their goals, if name is unknown) rather than as a Protesters (i.e. with an ‘Inter code’ of 3 or 4). In other cases, armed individuals may be present at a demonstration, yet are not themselves involved in the demonstration as a demonstrator, and do not engage physically with demonstrators. Often these reports cite the armed presence as locals present to ‘defend their community’ from rioting, or to intimidate demonstrators. Such cases are coded with the ‘Strategic developments’ event type and ‘Other’ sub-event type. The ‘Notes’ for these events will start with “Non-violent activity:”. For more on ‘Strategic developments’, see below. The demonstration event at which these armed individuals are present will also include a tag “armed presence” in the ‘Tags’ column. Some users may choose to include such events alongside counter-demonstrations (described above).
How can I see when police have used force against demonstrators?
Police can engage with demonstrators in a number of ways. Police, as with all state forces, are coded with an ‘Inter’ code 1; as such, all demonstrations involving police will have an ‘Interaction’ code of 15 or 16 (for more on ACLED coding rules, see the ACLED Codebook). Demonstration events coded with an ‘Interaction’ code of 15 are ‘Violent demonstrations’, meaning that while police may or may not use force against demonstrators, demonstrators were engaging in destructive or disruptive behavior (e.g. violence, vandalism, looting, etc.). Demonstration events coded with an ‘Interaction’ code of 16, meanwhile, may be one of two types of events. These events are either a ‘Protest with intervention’ — meaning that police may have used some force though not resulting in deaths or serious injuries/hospitalizations (e.g. arrests, tear-gassing without injuries, etc.) — or are a case of ‘Excessive force against protesters’ — meaning that police used lethal force and/or force that resulted in serious injuries/hospitalizations. If police are merely present at a demonstration yet do not physically interact with demonstrators in any way, they are not coded as a primary actor as they are not viewed as active actors within the event.
Does ACLED capture detentions of demonstrators and civilians by state forces in unmarked vehicles?
Yes, both during and outside of demonstrations. Events in which demonstrators, or other civilians, are taken and detained by state forces in unmarked vehicles outside of demonstrations are coded with ‘Strategic developments’ event type and ‘Arrests’ sub-event type and are denoted with a tag “detentions” in the ‘Tags’ column of the event. If the detention occurs during a demonstration, the tag will be added to the specific demonstration event during which the detention occurred. These events are not limited to federal forces, but rather any state security force which meets the above parameters. The “detentions” tag is applied to events from 1 January 2020 to 31 December 2020 in the data.
Does ACLED capture political violence or demonstrations related to the coronavirus pandemic?
Yes. ACLED tracks the impact of the coronavirus pandemic on political violence and demonstrations around the world as part of the COVID-19 Disorder Tracker. To identify events that are directly related to the pandemic — such as health workers protesting for access to better PPE, or armed demonstrators demanding movement restrictions be lifted — look for the word ‘coronavirus’ in the ‘Notes’ section of events (the word ‘coronavirus’ specifically will be used). For more information on events that are ‘directly related’ to the pandemic, see this primer. Furthermore, within this subset of coronavirus-related demonstrations, ACLED provides additional categorizations to help isolate two notable trends: firstly, an additional tag “schools” has been created in the ‘Tags’ column to capture demonstrations specifically against the reopening of schools or school activities during the pandemic; secondly, for demonstrations demanding a cancellation or temporary suspension of rent during the pandemic, the ‘Associated Actor’ Cancel the Rents. The “schools” tag is applied to events from 1 January 2020 to 28 February 2022 in the data.
Does ACLED only cover demonstrations around certain topics (e.g. Black Lives Matter)?
No. Whenever salient identities are reported, they will be coded as an ‘Associated Actor’ to the respective primary actor (for more on coding decisions, see the ACLED Codebook).
What does it mean when ACLED codes a demonstration with BLM: Black Lives Matter as an ‘Associated Actor’?
ACLED codes BLM as an ‘Associated Actor’ for both Protesters and Rioters in data across all regions around the world under the following instances: when the demonstration has a local BLM group involved; when the main issue of the demonstration concerns the police killing of a specific Black person (e.g. George Floyd, Breonna Taylor); when the main issue of the demonstration concerns police brutality against Black people in general; or when the demonstration is in solidarity with the movement in the USA against police brutality against Black people. ACLED does not code BLM as an ‘Associated Actor’ when: the demonstration concerns police brutality in general (without a racial element); when the demonstration concerns police brutality against another identity group (e.g. Latin Americans); when the demonstration concerns the police killing of a specific person who is not Black; or when the demonstration concerns other issues related to racial tensions or racial inequality. As such, it is important to remember that being coded as an ‘Associated Actor’ points to an association rather than the implication that ACLED speaks on behalf of the BLM movement.
Why doesn’t ACLED have a ‘demonstration size’ variable?
The size of demonstrations is one of the most biased and estimated aspects of reporting, with numbers varying largely based on whoever is reporting. This is why ACLED does not record this information as a distinct variable. That said, ACLED researchers will note the size of the demonstration, as reported, in the ‘Tags’ column of each event; this is denoted as “crowd size=x” with “x” denoting the exact terminology used in reports. A cursory look at this information in the notes makes it clear how this information is often not reported at all, and when it is reported is quite approximate.
Why does ACLED report a different number of demonstrations than other datasets tracking demonstrations in the US?
Differences in methodology and sources of information are the primary reasons why demonstration counts at ACLED may vary from those of other datasets tracking US demonstrations. ACLED methodology limits what is counted as a ‘Demonstration’ per a specific catchment. ACLED only collects information on physical congregations of groups (three or more). Single-person demonstrations, symbolic demonstrations, and planned demonstrations without any reports of a demonstration actually occurring are not coded (as outlined above). ACLED sourcing limits information to that which comes from traditional media; select, trusted new media; reports from institutions; and information from partnerships with local organizations. ACLED does not use crowdsourcing, meaning that researchers do not scrape social media (e.g. Twitter). This is because ACLED is not able to verify such reports, especially as they can be susceptible to manipulation (e.g. ‘fake news’, bots, etc.). In this way, data collection for the US context is standardized with ACLED data collection across the globe, allowing for cross-country comparability.
Does ACLED code attacks on statues?
Yes, ACLED collects information around the targeting of statues (including, though not limited to, Confederate statues). Such statues are often targeted in the context of a riot; if a statue is specifically targeted during a riot, the event will be coded with ‘Riots’ event type and ‘Violent demonstration’ sub-event type, with the primary actor coded as ‘Rioters’. If a statue is targeted during demonstrations in which rioters also engage with police, ‘Police Forces of the United States’ will be coded as the other primary actor. For ACLED to code such attacks on statues as ‘violent’ events (i.e. ‘Riots’), there must be substantial damage to the statue — rather than simply graffiti or other similar acts of minor vandalism — such as the breaking off of pieces or the toppling of the statue. Such events include a tag “statue” in the ‘Tags’ column. The “statue” tag is applied to events from 1 January 2020 to 31 December 2020 in the data.
Strategic Developments
What are strategic developments?
‘Strategic developments’ are useful for understanding the context of disorder, and as such are a useful tool for ACLED users. This category is designed to capture contextually important events and developments that are not political violence or demonstrations. These events may, however, contribute to the political environment and/or may trigger future political violence and/or demonstrations. Examples of such events include government-enforced curfews meant to restrict demonstrations from occurring.
Can I work with data on strategic developments the same way I work with the data on political violence or demonstrations?
No. The ‘Strategic developments’ event type is unique from other event types in the ACLED dataset in that it captures significant ‘developments.’ Because what types of events may be significant varies by context as well as over time, these events are, by definition, not systematically coded. One action may be significant in one place at a specific time yet a similar action in a different context during a different time period might not have the same significance. This means that ‘Strategic developments’ should not be assumed to be comparable across contexts and time, as other ACLED event types are. Rather, ‘Strategic developments’ ought to be used as a means to better understand the analysis you are conducting as a user. When used correctly, these events can be a useful tool in better understanding the landscape of disorder within a certain context. It may be useful to think of them as a way to annotate a graph: to make better sense of trends you see in the data before and after that strategic development (for more on Strategic developments, see this methodology primer).
Does ACLED collect information on curfews and the deployment of federal forces?
Yes, both are coded as ‘Strategic developments’ given the impact such developments may have on future patterns of political violence and/or demonstrations. Such events are coded with the ‘Strategic developments’ event type and ‘Change to group/activity’ sub-event type. For curfews, the ‘Notes’ will begin with “Security measures:” and for deployments, the ‘Notes’ will begin with “Movement of forces:”.
Does ACLED collect information on non-violent militia activity?
Yes, ACLED collects information on a number of different types of non-violent militia activity which may indicate a change in broader political trends. These are coded with Strategic developments’ event type with the appropriate ‘sub-event type’ (see below). Beyond event types and sub-event types, ‘Strategic developments’ are further categorized by tags before the text in the ‘Notes’ column, stylized as “Tag:”. The following list includes the types of events captured, and how they are coded:
- Recruitment drives/events (‘Change to group/activity’ sub-event type, ‘Notes’ tag “Recruitment:”)
- The creation of new militia groups or splinter groups (‘Change to group/activity’ sub-event type, ‘Notes’ tag “Change to armed group:”)
- Training exercises (‘Other’ sub-event type, ‘Notes’ tag “Training:”)
- Announcements and other events (‘Other’ sub-event type, ‘Notes’ tag “Other:”)
The majority of such events are provided through our partnership with MilitiaWatch (see “Does ACLED have any US data partnerships?” above).
Accessing and Using the Data
How can I access the data? Will the US data be included in the ACLED API and global dashboard?
Yes. The data are freely accessible through the ACLED export tool, curated files, global dashboard, and API.
Do you have data for previous years/time periods in the US?
Covering previous years and time periods is a priority for ACLED, but back-coding requires additional support to progress. If you are interested in providing support, please see below.
Can I do regional comparisons with other countries covered by ACLED using these data?
Yes, though with certain caveats. An important point to remember regarding comparison is that ACLED has not covered the same time period for all countries. Please refer to ACLED’s coverage list for further information.
How often are the data updated?
Data are updated on a weekly basis on Tuesdays. The data published on Tuesday will contain events from the preceding week (Saturday through Friday). ACLED data are dynamic. Once the data have been published, if new information becomes available, previous events will be updated or corrected where necessary.
How do I cite the data?
If using ACLED data in any way, directly or manipulated, the data must be clearly acknowledged. Acknowledgement should include: (1) a footnote with the full citation which includes a/the link to ACLED’s website; (2) in text citation/acknowledgement, stating where the data you use are from and that ACLED data are publicly available; and (3) clear citation on any and all visuals making use of ACLED data. If generating a data file for public or private use, and presenting those data to another party, the ACLED data included must be directly acknowledged in a source column, including ACLED’s full name and a link: “Armed Conflict Location & Event Data Project (ACLED); https://www.acleddata.com”
To reference the ACLED codebook, please cite as follows (substituting for the correct year): ACLED. (2017). “Armed Conflict Location & Event Data Project (ACLED) Codebook, 2017.”
If using ACLED data in an academic paper or article, please cite as follows: Raleigh, Clionadh, Andrew Linke, Håvard Hegre and Joakim Karlsen. (2010). “Introducing ACLED-Armed Conflict Location and Event Data.” Journal of Peace Research 47(5) 651-660.
Please note all use of ACLED data is subject to the Terms of Use. For further information, see ACLED’s Terms of Use & Attribution Policy.
For any remaining questions, contact [email protected].
How can I help improve ACLED’s coverage?
ACLED invites all users to contact us directly if they would like to share additional information on an event or events which are relevant to the dataset. They can do so by contacting [email protected] or through the portal on our website.
ACLED works with a unique network of international and local partners around the world. Through partnerships with local organizations and conflict observatories, ACLED is able to collect information in hard-to-access contexts and integrate thematic, regional, and country-level expertise into the data collection program. Partners benefit from ACLED training in data management, visualization, and/or analysis — building local monitoring capacity and enhancing ACLED’s global crisis mapping network. Partner contributions are invaluable to ensuring ACLED is able to produce the highest quality real-time data on political violence and protest around the world. The United States is no exception. ACLED seeks US-based local partners to improve our coverage on an ongoing basis. If you are an organization collecting information on political violence and/or demonstrations and are interested in partnering with ACLED, please reach out to us at [email protected].
How can I stay informed about the project and receive the latest data and analysis?
You can sign up to our mailing list for this project here.
How can I help ACLED to continue collecting this information?
ACLED seeks to continue data collection and analysis on the US through the elections in November 2020 and beyond. If you are interested in finding out more about the project or collaboration and partnership opportunities, please contact [email protected].