EconAI

We are a team of economists and data scientists at the forefront of integrating machine learning with economic research.

[conflictforecast @conflictforeca1]

📢 AUGUST UPDATE - Part 2

New UCDP data bring revisions: Yemen jumps from 133 → 291 predicted fatalities and Nigeria rises to 423. Somalia, Mali, DR Congo & Israel decline.

Rwanda, Uganda & Turkey remain the largest relative-risk outliers.

👉 https://t.co/BWxmCPEAJ8

[conflictforecast @conflictforeca1]

📢 SEPTEMBER UPDATE – Part 1

• Predicted Fatalities: Ukraine remains far ahead despite a decline. Iran rises, while Pakistan and DR Congo remain elevated.

• Risk: Rwanda remains the largest outlier, followed by Uganda and Turkey.

👉 https://t.co/BWxmCPE2TA

[conflictforecast @conflictforeca1]

📢 JULY UPDATE – Part 2

• Predicted fatalities: Ukraine remains far ahead. Sudan, Pakistan, and DR Congo rise, while Lebanon drops sharply.
• Relative risk: Rwanda remains the largest outlier, followed by Uganda and Turkey. Jordan enters the top 15.

👉 https://t.co/id9WaLpiyq

[conflictforecast @conflictforeca1]

📢 JULY UPDATE – Part 1

• Predicted fatalities: Ukraine rises further and remains far ahead. Mexico moves up, while Niger and Colombia also increase.

• Relative risk: Rwanda remains the largest outlier, followed by Uganda and Turkey. Eritrea moves above Egypt.

[conflictforecast @conflictforeca1]

📢 JUNE UPDATE – Part 2
• Predicted fatalities: Ukraine remains the highest-risk case. Sudan and Nigeria decline, while Ethiopia and Burkina Faso rise. Niger enters the top 15.
• Relative risk: Rankings remain stable, led by Rwanda, Uganda, and Turkey. Guinea enters the top 15.

[conflictforecast @conflictforeca1]

📢 JUNE UPDATE – Part 1

• Predicted fatalities: Ukraine remains by far the highest-risk case. Rankings are stable, with Lebanon now slightly above Israel.

• Relative risk: Rwanda remains the largest outlier, followed by Uganda and Turkey. Peru re-enters the top 15.

[conflictforecast @conflictforeca1]

📢 MAY UPDATE – Part 2

• Predicted fatalities: Ukraine remains by far the highest-risk case. Sudan rises, while DR Congo moves up.

• Relative risk: Rwanda and Turkey remain the largest outliers. Uganda enters the top three.

👉 https://t.co/id9WaLpiyq

[conflictforecast @conflictforeca1]

📢 MAY UPDATE – Part 1

• Predicted fatalities: Ukraine remains by far the highest-risk case. Lebanon continues to move up the ranking.

• Relative risk: Rwanda, Turkey, and Egypt remain the largest outliers. Peru enters the top 15.

👉 https://t.co/id9WaLpiyq

[conflictforecast @conflictforeca1]

📢 APRIL UPDATE – Part 2 (with UCDP fatalities)

• Predicted fatalities: Ukraine still far ahead (3,741); rankings broadly unchanged.

• Deviation from expected risk: Rwanda & Turkey remain top outliers; little change overall.

👉 https://t.co/id9WaLpiyq

[conflictforecast @conflictforeca1]

📢 APRIL UPDATE – Part 1

• Predicted fatalities: Ukraine rises further, still far above all others. Rankings otherwise remain stable.

• Deviation from expected risk: Rwanda, Turkey, and Egypt remain at the top, with minor reshuffling.

👉 https://t.co/id9WaLpiyq

[conflictforecast @conflictforeca1]

📢 MARCH UPDATE – Part 2 (with UCDP fatalities)

• Predicted fatalities: rankings remain stable, with Ukraine far above others and the same core group dominating

• Deviation from expected risk: Rwanda, Turkey, and Egypt still lead, with some reshuffling (e.g. Guinea rising)

[conflictforecast @conflictforeca1]

📢 MARCH UPDATE – Part 1

Predicted fatalities remain broadly stable, with the same countries at the top and only minor reshuffling.

More movement in “surprise” risk: Madagascar, Sri Lanka, Indonesia, and Guinea-Bissau re-enter, while Jordan and Senegal remain but move down.

[conflictforecast @conflictforeca1]

📢 FEBRUARY UPDATE – Part 2
🔹 Predicted fatalities: broadly stable, but Iran’s expected intensity is now ~half of earlier estimates
🔹 Deviation from expected risk: bigger shifts—Jordan jumps to #2, and new entries include Rep. of the Congo, Saudi Arabia, Uzbekistan, and Morocco

[conflictforecast @conflictforeca1]

📢 FEBRUARY UPDATE – Part 1 (news-based)
• Ukraine remains far ahead in predicted fatalities over the next 12 months, followed by Iran and Sudan.
• Rwanda, Turkey, and Egypt show the largest deviation from expected risk given peace length.
• Overall rankings remain very stable

[conflictforecast @conflictforeca1]

📢 JANUARY UPDATE – Part 2 (with UCDP fatalities)

🔎 The rankings are largely stable — reassuring for the model.

Key shifts:
• Iran enters at #2 in predicted fatalities (newly recorded January events).
• Iraq jumps to the top of the “deviation from expected risk” list

[conflictforecast @conflictforeca1]

JANUARY UPDATE
We’re highlighting conflict risk in three new ways:
🔹 Top 15 countries by predicted fatalities – led by Ukraine & Sudan
🔹 Top 15 by risk relative to peace length – e.g., Rwanda & Turkey
🔹 Plot shows how expected risk evolves with peace length; outliers labeled

[conflictforecast @conflictforeca1]

📢 DECEMBER UPDATE
We’ve completed our annual maintenance, including retraining our LDA text model.
Below are the latest topic bubble maps for:
🇬🇧 UK (low risk) and 🇮🇷 Iran (high risk), illustrating how different news narratives translate into sharply different risk signals.

[conflictforecast @conflictforeca1]

Update: New-Year Model Retraining in Progress 📈

We are currently retraining our LDA text model as part of our scheduled maintenance. This month's conflict forecast update will be slightly delayed.

We’re excited to share the refreshed data with you soon!

[conflictforecast @conflictforeca1]

📢 DATA UPDATE - New LDA topics live!
Our website now uses new LDA text topics for all conflict forecasts.
📌 Early warnings captured in cases like:
🇪🇬 Egypt, Feb 2012: Port Said riot
🇨🇲 Cameroon, May 2014: Boko Haram-related conflict

[IDEA Graduate Programme in Economics @IDEA_UAB]

IDEA alumni Matt Delventhal and @paupujolas together with former IDEA professor JC Conesa published some exciting new research. These sort of projects are born in this program. Apply now: https://t.co/LoMJpSqC1O. Admission on rolling basis #econtwitter