Weekly Conflict Index

ACLED’s Weekly Conflict Index provides a singular measure of conflict intensity in every country in the world. It is updated every Wednesday and leverages the conflict indicators identified in ACLED’s annual Conflict Index — deadliness, danger, diffusion, and fragmentation — to create a single score for each country that can be used to track its conflict evolution each week. 

Explore the table below to see how a country’s score has changed this week, what level of conflict it is experiencing, and how the Conflict Alert System forecasts it will change in the next six months. Full results — including a single numeric score for conflict levels — can be downloaded using the link below.

Methodology

The Weekly Conflict Index is built upon four unique indicators: a conflict’s deadliness, danger, diffusion, and fragmentation. “Deadliness” tracks the number of fatalities in a country, “danger” represents the number of events targeting civilians specifically, “diffusion” measures the geographic spread of violence in a country, and “fragmentation” tracks the number of non-state armed groups engaging in violence in the country.

Each of these indicators is measured weekly at the country level. For more details on how each indicator is measured, read about our methodology. The indicators measure very distinct concepts, and the ranges of possible values are similarly distinct. Diffusion, for example, can have a theoretical minimum value of 0 and a maximum value of 100 (representing 100% of areas in the country experiencing high levels of violence), while bounding a measure like deadliness (the number of fatalities) using historical averages discourages anomalies from appearing in the trends. Still, the values of each indicator need to be somewhat normalized to prevent a single metric from dominating the rest. To do this, the square root of each indicator is taken, both to avoid the extreme distortion of the data that can result from logarithmic transformations and to easily account for values of 0 in the data.

Next, each indicator is assigned a weight, in line with its significance on how conflict conditions change on a weekly basis. Deadliness is weighted most heavily (35%), followed by danger (25%), as both indicators have the greatest direct impact on lives. Diffusion and fragmentation are weighted equally at 20% each. After each country’s normalized indicator value is raised to its weight value, the results are summed to result in a single final score.

While the theoretical minimum value of the final score is 0, there is no theoretical maximum value. This is intended to reflect the reality of conflict dynamics: There is no pre-defined “worst case scenario” score. Each country’s weekly score is intended to be a measure both of its conflict levels relative to other countries that week, but also to itself over time. This score construction allows for both of those comparisons.