1,774
Views
4
CrossRef citations to date
0
Altmetric
Articles

The role of governmental weapons procurements in forecasting monthly fatalities in intrastate conflicts: A semiparametric hierarchical hurdle model

Figures & data

Table 1. Periodisation of data for expanding window evaluation at time point t{Jan.2017,,Dec.2019} and s{2,,7} and forecasting.

Figure 1. Observed and predicted changes in fatalities Δijts with s = 2 in June 2018 in a sub-sample of countries. The predictions are separated along Stage 1 (first row), Stage 2 (second row), and final predicted change to April 2018 (third row).

Figure 1. Observed and predicted changes in fatalities Δijts with s = 2 in June 2018 in a sub-sample of countries. The predictions are separated along Stage 1 (first row), Stage 2 (second row), and final predicted change to April 2018 (third row).

Table 2. Out-of-sample MSE and TADDA scores from the hierarchical hurdle model with and without the MCW-related covariates as well as the benchmark model.

Figure 2. Forecasted log changes Δijts in fatalities to October 2020 (s = 2) and March 2021 (s = 7), focused on the border region between Nigeria, Niger, Cameroon, and Chad.

Figure 2. Forecasted log changes Δijts in fatalities to October 2020 (s = 2) and March 2021 (s = 7), focused on the border region between Nigeria, Niger, Cameroon, and Chad.

Table 3. The subset of the parametric estimates regarding the governmental procurement of weapons with s = 2 and training data from January 1990 to August 2020.

Supplemental material

Supplemental Appendix

Download PDF (1.3 MB)