About
Our Protest Risk Model forecasts the risk of violent right-wing and left-wing protests on a county level across the US. The product was built using a number of AI and machine learning models combined with over 150 data sets (including political, economic, demographic and historical protest data). Additionally, the model also incorporates the qualitative analysis of our North America desk to provide difficult-to-quantify insights into the final risk gradient. Using this methodology, the model forecasts the separate probabilities of left-wing and right-wing protests occurring over the next 30 days on a county level. Furthermore, the model then predicts the probability of such a protest turning violent.
This final risk gradient can be used by clients to optimise the allocation of resources and security planning for both possible outcomes: a Trump or Harris victory. This ultimately allows for a non-biased and holistic approach to domestic unrest forecasting, enabling data-driven and fact-based mitigation planning on a granular level.
How to interpret the scores
What is the Meaning of the County Risk Scores?
The model assesses the probability of right or left wing violent protests in each county. The County Risk Scores underscore structural unrest drivers which are unlikely to change in the short to medium term. While the risk scores can appear to be relatively low, the outcome of the election will act as the primary risk exponent for these fundamental risk drivers, taking the assessed risk scores much higher.
In short, the model provides a County Risk Score for the probabilities of seeing violent protests in each one of the counties for each ideology.
How to Interpret the Scores?
The outcome of the election will increase the risk equally across the country. So, using these risk scores will allow you to understand which areas will face higher risks depending on the outcome of the election.
For example, if the most dangerous scenario takes place, the national unrest risk baseline will increase to X%. So, all counties will have a violent unrest risk score of X + County Risk Score.