Trends of Armed Conflict in Kenya from 1997 to 2021: An Exploratory Data Analysis

Abstract

Armed conflict patterns have drastically changed since the post-cold war period. In Sub-Saharan Africa, armed conflict continues to be persistent and on the rise. Kenya has not experienced civil war, but has experienced intra-state conflicts which display themselves as political, natural resources, ethnicity, land, and environmental conflicts. This study aimed to identify patterns and trends of armed conflict in Kenya. Secondary data from Armed Conflict and Location Events Data (ACLED) for the period 15 th January 1997 to 25 th February 2021 was used. Exploratory data analysis and generalized additive model were used to identify patterns and trends. For the period studied, 7,437-armed conflict events and 11,071 fatalities were recorded. There was a non-linear trend and a general increase in the number of armed conflict cases in Kenya. The peaks in the non-linear trend were observed during the years 2002, 2007, 2013 and 2017. On the contrary, the number of fatalities from armed conflict decreased over time and had a non-linear trend, with peaks in the years 1998, 2001, 2007, 2013 and, 2017. Similarly, there was a reduction in the number of fatalities per armed conflict over time with 149 fatalities per 100-armed conflict events recorded in the study period. Linear and non-linear trend of armed conflict events was observed at the county levels, with counties like Nairobi and Nakuru having a non-linear trend similar to the overall trend. The number of events of armed conflict for riots and protests event type had a non-linear trend while the rest had a linear trend with a positive slope. Violence Against Civilians (VAC) event type had the highest number of events followed by Riots and Protests. Additionally, VAC had the highest number of fatalities followed by Battles and Riots. In terms of fatalities per armed conflict, Explosions/Remote violence event type had the highest fatality rate followed by Battles and VAC. The peaks in the number of armed conflict cases and fatalities were observed in the years in which general elections were conducted in Kenya.

Predicting Conflict Zones in Kenya Using a Point Process Model

“…From the exploratory results, armed conflict recorded in the years or months just before the elections have been shown to increase and in particular regions such as Nairobi, Nakuru among others. These results are quite similar to the results presented by Kimani, Mugo and Athiany in an exploratory analysis carried out in 2020 [5]. As per ACLED data [15], the number of armed conflict incidents in Kenya has a non-linear trend, with distinct peaks observed in the years that the general elections were held [5].…”

“…These results are quite similar to the results presented by Kimani, Mugo and Athiany in an exploratory analysis carried out in 2020 [5]. As per ACLED data [15], the number of armed conflict incidents in Kenya has a non-linear trend, with distinct peaks observed in the years that the general elections were held [5].…”

“…ACLED collects information on specific conflict event types such as battles, riots, remote violence, protests, and violence against civilians and fatalities therein. A summary of the classification of these events have been documented by Kimani, Mugo and Athiany [5]. Moreover, it also considers the unique variables of the countries, that is, the geography of the regions, population sizes, and the types of violence in the specified country/ region, providing written reports describing the events after the collected data has been cross validated by multiple coders.…”

“…According to ACLED, armed conflict is the use of force by a group with a political purpose or motivation [4,5]. Violence being part of this, remains a predominant and persistent issue that needs addressing, varying only in area and time of occurrence.…”

“…Studies have shown how conflict can impact the day-to-day activities of the affected areas [13]. Different regions in Kenya have experienced conflicts of varying intensities and forms, such as political conflict during the 2007-2008 post-election violence witnessed in Kenya [5,13] or communal conflict, especially in arid and semi-arid lands (ASAL), which occurs without notice with incidents such as cattle rustling. These events have ripple effects, and the end results include loss of property and lives, displacement of families, and a decline in the country's economy [13,14].…”