The Islamic State's Strategic Trajectory in Africa: Key Takeaways from its Attack Claims.

• Author: Rolbiecki, Tomasz

In March 2019, the Islamic State was declared defeated after it was routed by the coalition-backed Syrian Democratic Forces (SDF) in the town of Baghouz in eastern Syria. (1) In the time that has elapsed since, the fallacy of this declaration of defeat was rendered apparent countless times_whether by mass-casualty terrorism or the Islamic State's expansion into new lands. (2) Indeed, by the summer of 2020, it had become resolutely clear that the Islamic State was a changed organization, but by no means a beaten one.

This has been most apparent on the continent of Africa. The Islamic State has had an active presence in north, west, and east Africa for years, but in 2019, the military potential of its affiliates there--especially in west and central Africa and the Sinai Peninsula--surpassed that of its residual core in the Levant. This is most starkly the case in northeastern Nigeria, where its supporters have been engaging in attacks that have exceeded the scale and complexity of those being deployed by their counterparts in Syria and Iraq for at least a year now.

Exploring the group's insurgent prospects in Africa, this article makes an operational assessment of the Islamic State's provincial and non-provincial affiliates on the continent based on two streams of data--the first, an 83-week aggregation of attack statistics published in the Islamic State newspaper Al Naba between December 28, 2018, and July 31, 2020; the second, an exhaustive collection of Africa-focused attack reports prepared and distributed by the Islamic State's Central Media Diwan in 2019 via an outlet called the Nashir News Agency. Through the lens of these data streams, the authors evaluate the geographic, tactical, and targeting characteristics of the Islamic State's presence across Africa, identifying its key hotspots, emergent strongholds, and potential future trajectory.

The article proceeds as follows. After a brief discussion of the data collection and analysis methodology, the authors disaggregate the data by wilaya (province), focusing first on West Africa and the Sahel, before moving on to the Sinai Peninsula, Somalia, Central Africa, and Libya, Tunisia, and Algeria (which are considered collectively due to the comparatively low amount of Islamic State-reported activities in them). (a) For each location, the authors describe contemporary attack trends (based on the December 28, 2018, to July 31, 2020, Al Naba dataset), contextualize the recent history of the affiliate in question, and identify key operational dynamics (based on attack reports from 2019).

The conclusion considers the implications of this data in aggregate, holding that the Islamic State's forays into Africa are no longer a sideshow to its operational core in Syria and Iraq. Rather, its brand as a globalized insurgency is dependent now more than ever on the military activities of its affiliates there.

Methodology

This assessment is based on two complementary datasets, both of which were drawn from the Islamic State's official propaganda output. The first dataset, which is used for high-level quantitative analysis of Islamic State attack trends in Africa between 2019 and 2020, is drawn from the aggregated weekly statistics the Islamic State publishes in its newspaper Al Naba. In accordance with the Al Naba publishing cycle, its start date is December 28, 2018. Because the authors use this dataset to track trends through both 2019 and 2020, its end data is July 31, 2020. This dataset is henceforth referred to as "the Al Naba dataset."

The second dataset comprises every operation claim published via Nashir, the group's official media distribution channel, in relation to Africa in 2019. Its start date is January 1, 2019, and end date December 31, 2019. For sake of clarity, this dataset is henceforth referred to as "the Nashir dataset." This dataset was used to quantify attacks in 2019 and for the qualitative analysis of Islamic State attacks (location, type, scale, and target, etc.) deployed in/by the different wilayat. (b)

The Al Naba Dataset

The authors use the Al Naba dataset to ascertain Islamic State attack trends in Africa during the 83-week period between December 28, 2018, and July 31, 2020. This dataset relates how many attacks, according to the Islamic State, occurred on a weekly basis in a given wilaya in Africa. Besides total kill and casualty counts, it does not provide specific details as to what exactly those attacks looked like. That is because this dataset was drawn from already-aggregated statistics prepared and distributed by the Islamic State in its weekly newsletter, Al Naba. Specifically, these numbers originally appeared in the "Harvest of Soldiers" infographic series, which has been running on a weekly basis since July 2018. (3)

The Al Naba dataset's start date is December 28, 2018, because each issue of Al Naba provides aggregated statistics for the seven days preceding its publication. Issue 163, which appeared on January 3, was the first issue of Al Naba to be published in 2019. Hence, that issue covers the period between December 28, 2018, and January 3, 2019.

Visualized in Figures 6 and 7, this dataset shows high-level attack trends in Africa in 2019 and 2020. (c)

The Nashir Dataset

The authors use the Nashir dataset--which, disaggregated by wilaya and relating detailed information about individual attacks claimed by the Islamic State, is rich with tactical, geospatial, and targeting data--to develop an understanding of the full spectrum of Islamic State operations in Africa in 2019. This spectrum includes anything from its affiliates' sporadic raids in southern Libya to their more consistent, higher impact operations in northeastern Nigeria. The authors are able to do this because this dataset comprises a complete sample of detailed Islamic State operation claims relating to Africa in 2019, all of them prepared by its Central Media Diwan (i.e., not the provincial affiliates themselves, which, as officially designated affiliates of the Islamic State, do not have their own media presence). (4)

These claims were collected by the third author exclusively from Telegram, a privacy-maximizing social media platform favored by violent extremists for propaganda distribution (among other things). (d) In 2019, two outlets were charged with distributing all official Islamic State communications: the Nashir network, which was tasked with disseminating materials produced by central and provincial media units; and the Amaq News Agency, which essentially acted as its newswire service. Operating alongside these was a separate, supporter-run dissemination hub called the Nashir News Agency (note: despite the name, this entity is distinct from the Nashir network, which is internal to the Islamic State). (5) The Nashir News Agency aggregated all posts from both Nashir and the Amaq News Agency on a minute-by-minute basis. It was from this hub, the Nashir News Agency, that the Nashir dataset was compiled.

Prior to the analysis, the authors filtered the Nashir dataset so that it only contained operation claims published in 2019 in relation to Africa. This involved removing all photo, audio, and video files. This was done to help avoid duplication. Photo, audio, and video content only cover a small number of attacks and only provide supplementary coverage of attacks claimed in statements. Moreover, until June 2020 when the Amaq News Agency stopped publishing news bulletins altogether, the Islamic State almost always issued duplicate claims for its attacks--one prepared by the Central Media Diwan and one by the Amaq News Agency. To avoid accounting for duplicates, all Amaq News Agency claims were also removed from the dataset. In total, this cleaning process resulted in the exclusion of some 5,248 pieces of content from the corpus, leaving 453 Central Media Diwan- prepared and Nashir News Agency-disseminated operation bulletins relating to Islamic State activities in Africa in 2019. Each of these was then manually processed by the authors to make sure that no duplicates reports found their way into the dataset.

After this, each bulletin was coded according to several criteria, among them:

* Week and date of the attack;

* Longitude and latitude of the attack location (as visualized in Figure 2);

* Wilaya, country, and region to which the report relates (as visualized in Figures 2, 3, and 4);

* Weapons used in the attack;

* Attack type (i.e., ambush, assault, assassination, bombing, etc.);

* Target (i.e., Nigerian Army, Egyptian Federal Police, etc.);

* Target type (i.e., military, intelligence, civilian, government, etc.) (as visualized in Figure 5); and

* When mentioned, number of kills reported.

This data from the posts on the Nashir News Agency is visualized in Figures 1 to 5.

Correlating the Datasets

While the two datasets are structurally distinct and of a moderately different size (453 attacks are recorded in the Nashir dataset versus 595 in the Al Naba dataset), they correlate closely, as indicated in Figure 1. The Islamic State's explanation for the numerical discrepancy is that not all attacks are separately claimed through Central Media Diwan bulletins for reasons of operational security. (6) Furthermore, some attack claims in the Nashir dataset mention two or more different locations. In the Al Naba dataset, they are listed separately. (7)

Notwithstanding this variance, the correlation between the two datasets is clear from Figure 1. Based on it, the authors are confident that the two datasets are...