The Promise and Pitfalls of Conflict Prediction: Evidence from Colombia and Indonesia
Saved in:
| Title: | The Promise and Pitfalls of Conflict Prediction: Evidence from Colombia and Indonesia |
|---|---|
| Authors: | Dube, Oeindrila, Blair, Robert, Gudgeon, Matthew, Peck, Richard, Blattman, Christopher, Bazzi, Samuel |
| Source: | The Review of Economics and Statistics. 104:764-779 |
| Publication Status: | Preprint |
| Publisher Information: | MIT Press, 2019. |
| Publication Year: | 2019 |
| Subject Terms: | bepress|Social and Behavioral Sciences|Political Science|Comparative Politics, bepress|Social and Behavioral Sciences|Economics, 05 social sciences, SocArXiv|Social and Behavioral Sciences|Economics, 1. No poverty, SocArXiv|Social and Behavioral Sciences|Political Science, 16. Peace & justice, SocArXiv|Social and Behavioral Sciences|Economics|Growth and Development, 0506 political science, bepress|Social and Behavioral Sciences|Political Science, SocArXiv|Social and Behavioral Sciences|Political Science|Comparative Politics, bepress|Social and Behavioral Sciences, SocArXiv|Social and Behavioral Sciences, bepress|Social and Behavioral Sciences|Economics|Growth and Development |
| Description: | How feasible is violence early-warning prediction? Colombia and Indonesia have unusually fine-grained data. We assemble two decades of local violent events alongside hundreds of annual risk factors. We attempt to predict violence one year ahead with a range of machine learning techniques. Our models reliably identify persistent, high-violence hot spots. Violence is not simply autoregressive, as detailed histories of disaggregated violence perform best, but socioeconomic data substitute well for these histories. Even with unusually rich data, however, our models poorly predict new outbreaks or escalations of violence. These “best-case” scenarios with annual data fall short of workable early-warning systems. |
| Document Type: | Article |
| Language: | English |
| ISSN: | 1530-9142 0034-6535 |
| DOI: | 10.1162/rest_a_01016 |
| DOI: | 10.31235/osf.io/bkrn8 |
| Access URL: | http://www.nber.org/papers/w25980.pdf https://www.nber.org/system/files/working_papers/w25980/w25980.pdf https://ideas.repec.org/p/nbr/nberwo/25980.html http://www.nber.org/papers/w25980 https://www.nber.org/papers/w25980 https://www.mitpressjournals.org/doi/abs/10.1162/rest_a_01016 https://osf.io/preprints/socarxiv/bkrn8 |
| Accession Number: | edsair.doi.dedup.....b37c513d594155ca93289c67a3cc3a58 |
| Database: | OpenAIRE |
Nájsť tento článok vo Web of Science | Abstract: | How feasible is violence early-warning prediction? Colombia and Indonesia have unusually fine-grained data. We assemble two decades of local violent events alongside hundreds of annual risk factors. We attempt to predict violence one year ahead with a range of machine learning techniques. Our models reliably identify persistent, high-violence hot spots. Violence is not simply autoregressive, as detailed histories of disaggregated violence perform best, but socioeconomic data substitute well for these histories. Even with unusually rich data, however, our models poorly predict new outbreaks or escalations of violence. These “best-case” scenarios with annual data fall short of workable early-warning systems. |
|---|---|
| ISSN: | 15309142 00346535 |
| DOI: | 10.1162/rest_a_01016 |