It is obviously not possible to claim to offer an exhaustive database of
Islamist attacks committed worldwide between 1979 and April 2024, for a number of reasons
detailed below.
A certain number of attacks have not been listed. Despite our best efforts,
it has not been possible to identify a certain number of Islamist attacks over the entire
period in question.
a/ In some cases, the number of victims caused
by Islamist attacks is insufficiently detailed. It is sometimes roughly estimated
by the authorities of the country who, for political reasons, do not wish to collect and
publish information on the subject. This is particularly the case with the Algerian state
(see our box below).
b/ A number of attacks do not appear in our
“retained estimate” when religious motivation is not clearly predominant in a
combination involving another determination. Islamist terrorism takes place in
singular and complex contexts that sometimes make it difficult to collect reliable data.
This is particularly the case in situations of war, civil or international, independentist
or separatist struggles and territorial conflicts that persist over long periods of time,
where causalities are shifting or inextricable, as in the case of the Palestinian conflict,
while in a completely different context, in Thailand for example, a separatist movement has
led a Muslim minority to get involved with weapons in the name of objectives that can
achieve, beyond political demands, a religious dimension.
c/ Islamist motivation is not always
identified. Available data do not always allow news agencies to attribute the
attack to the Islamist cause, especially if the country affected by the attack is
characterised by weak administrative structures. The absence of claim of responsability can
increase the likelihood that an attack will not even be recorded by agencies or that this
information will not reach the press.
d/ The number of deferred deaths is
practically unknown. However, it is certainly significant. Since victims who
succumb to their injuries after an attack are almost never mentioned in the available
information, it is impossible to know their exact number.
The death of Tahar Mejri, who died of grief on
June 14, 2019,
three years after losing his wife and son in the Nice attack on July 14, 2016, is
another
tragic illustration of what we refer to as ‘delayed deaths’. It is therefore also
impossible
to integrate these deaths into our database in a reliable way. Thus, according to our
database, we record at least 249,941 deaths and 192,598 people wounded, which is less than
the number of deaths. However, if we consider four cases of attacks, each carried out with
different means, in three countries where the quality of information is excellent, we
observe a completely opposite ratio between the number of deaths and the number of people
wounded: in the United States, there were five times as many people wounded (16,493) than
killed (3,001) in the 9/11 attacks; in France, there were three times as many people wounded
(413) than killed (137) in the 13 November 2015 attacks; in Nice, in the 14 July 2016
attack, there were five times as many people wounded (458) than killed (87). In Israel, the
attacks of October 7th, 2023, resulted in 3,400 injured and at least 1,195 dead, hence in
three times more deaths than injuries. It should be noted here that both the Bataclan attack
and the Nice attack led to survivors committing suicide several months or even years after
the events. Our study goes back on this particular point. This information leads us to
believe that the number of people wounded is much higher than that in our database.
Certainly, developing countries, which are the countries where most attacks take place, do
not have the same capacity to identify and care for people injured in an attack. Some of the
injured are probably not even counted, while others die from their injuries after a certain
period of time due to the inadequacy or fragility of relief systems and health institutions.
If we applied the ratios of the four sample attacks to the number of casualties in our
database (249,941), we would have to adjust this figure by multiplying it by three (hence
749,823 wounded) or five (hence 1,249,705). However, among all these injured people, it
appears certain that several of them die from their injuries, and that the poorer and more
lacking in rescue and medical resources the affected country is, the more numerous these
deaths are.
For these four reasons, we propose two types
of quantification of Islamist violence, in the form of a “retained estimate” and a
“possible estimate”. The “retained estimate” results from our database of attacks
during the period 1979-April 2024, that were clearly motivated by an Islamist agenda. The
“possible estimate” results from the identification of attacks that could be qualified as
Islamist, including certain terrorist acts that are also based on separatist, political or
social logic, which make it more difficult to attribute them to an exclusively or mainly
Islamist motivation. In all cases, the number of victims, dead or injured, is significantly
lower than a reality that cannot be more precisely known.
We have recorded 66,872 Islamist terrorist
attacks that killed at least 249,941 people between 1979 and April 2024.
Share this
image
“The possible estimate” should be between
300,000 and 400,000 deaths considering the victims of Islamist terrorism in Algeria,
according to the hypotheses based on the best-known sources and the most reliable
studies.
Our research began in spring 2018. It gave rise to a first publication, in
201923, and a second in 202124. The study we publish here is based on the database we
developed, within the limits previously indicated. The analyses that follow first present
the evolution of Islamist terrorism from 1979 to the present day, before proposing a
presentation and reading of the data according to the regions of the world and the countries
affected by Islamist violence.
The Algerian “black decade”: only an approximation is
possible
The Algerian case illustrates particularly well the difficulty of providing an exhaustive
database. Algerians refer to the period between 1991 and 2002 as the “black decade”, when
various Islamist groups, in particular the Islamic Salvation Front (FIS) and the Armed
Islamic Group (GIA), opposed the Algerian state in a violent civil war whose outcome is
difficult to document1. According to Fouad Ajami, “the precise number of Algerians who died
in the civil war that broke out in 1992 will never be known. The Algerian leaders, who are
not known for their loyalty to the truth and who have so much to hide, acknowledged in 1999
that 100,000 people were killed. More reliable estimates provided by Algerian civic
organisations put the death toll at 200,000”2. Indeed, President Abdelaziz Bouteflika first
estimated the number of victims at 100,000 in 1999. Then, in February 2005, the Algerian
president gave a speech in which he put the figure at 150,000 deaths3. Fouad Ajami considers
that the death toll could reach 200,000, arguing that it was in the interest of the Algerian
government to minimise the losses4. This figure, which has been used in various works on
terrorism in Algeria5, remains debated, particularly because it does not distinguish between
victims who were members of the police or the army, terrorists and civilians. In a study
published in 2008, Roman Hagelstein tried to distinguish between those killed in clashes
between security forces and terrorists and those killed in massacres, bombings and
assassinations. The study also distinguishes the number of missing persons, presumed to have
been murdered in secret, with the author estimating the total number of victims at 44,0006.
Regarding the victims of Algerian Islamist terrorism in the period 1991-2002, the available
estimates that can be considered reliable because of the sources, authors and types of
publication thus vary between 44,000 and 200,000. In any case, for this country and for this
period, there is no data available to provide information on the number of attacks, their
date, the modalities of action, the group responsible or the profile of the victims. For our
database, these major uncertainties have several consequences: firstly, due to the lack of
information, it is not possible to further specify the situation of Islamist terrorism in
Algeria during this decade. In keeping with our method, we therefore only include in the
“retained data” the cases that were reported. This choice leads to an undoubtedly
significant underestimation of the number of attacks and victims in Algeria. On the other
hand, we will include in the “possible estimate” category the indications concerning
Algeria. Finally, as a consequence, this situation of great ignorance with regard to the
Algerian case leads to an underestimation of the number of attacks and victims of Islamist
terrorism in the world, both for the period after 1991 and for the period as a whole.