Bi-Weekly Discussion - Can We Predict Geopolitical Conflict?

Details

We're currently hosting our discussions at Café Walnut. The cafe is near the corner of 7th & Walnut, just off Washington Square Park in Olde City. The cafe's entrance is below street level down some stairs, which can be confusing if it's your first time. Our group meets in the large room upstairs.

Since we're using the cafe's space, they ask that each person attending the meetup at least purchase a drink or snack. Please don't bring any food or drinks from outside. If you're hungry enough to eat a meal, they have more substantial fare such as salads, soups & sandwiches which are pretty good and their prices are reasonable.

-----------------------------------------------

CAN WE PREDICT GEOPOLITICAL CONFLICTS? AN INTRO TO THE EMERGING FIELD OF GEOPOLITICAL FORECASTING

INTRODUCTION:

This discussion will look at the emerging field of geopolitical forecasting, with a particular focus on current efforts to predict various forms of conflict both between states & within states. Current developments in geopolitical forecasting in the U.S. stem mostly from a geopolitical forecasting tournament, the "Aggregative Contingent Estimation" (ACE) program, which was run by the U.S. intelligence community's "Intelligence Advanced Research Projects Activity" (IARPA) from 2011 to 2015. The ACE program engaged tens of thousands of forecasters and drew over one million forecasts across roughly 500 questions of geopolitics relevant to U.S. national security. Typical questions included things like:

"What is the chance that a member will withdraw from the European
Union by [target date]?"
"What is the likelihood of naval clashes claiming over 10 lives in the
East China Sea by [target date]?"
"How likely is the head of state of Venezuela to resign by [target
date]?"
The tournament challenged several groups at academic institutions to come up with innovative methods of recruiting gifted forecasters, methods of training them in basic principles of probabilistic reasoning, methods of forming teams that are more than the sum of their individuals, and methods of aggregating predictions that most effectively distill the wisdom of the crowd.

The group that won this tournament was the Good Judgement Project (GJP), organized by UPenn political scientist Philip Tetlock. Using personality-trait tests, the researchers at GJP were able to select forecasting participants with less cognitive bias than the average person & train them in probabilistic reasoning. As the forecasting contest continued, the researchers were able to find so-called "superforecasters" who were in the top 2% for prediction accuracy. Amazingly, the superforecasters in GJP who only had access to publicly available information were reportedly 30% better than intelligence analysts with access to classified information.

The GJP's findings on what traits & methods allow successful geopolitical forecasting were laid out in his 2015 book, "Superforecasting: The Art and Science of Prediction". The GJP surprisingly found that good forecasting does not require a PhD in political science or statistics, powerful computers or access to classified intelligence, and amateurs working part-time could potentially beat intelligence analysts provided they followed the best practices developed by Tetlock. These best practices were summarized at the end of Tetlock's book in his "Ten Commandments for Aspiring Superforecasters". For our purposes here, the most pertinent rules are:
(2) Break seemingly intractable problems into tractable sub-problems, using Fermi estimates.
(3) Strike the right balance between "inside" (event-specific) and "outside" (reference class) views.
(4) Strike the right balance between under- and overreacting to new evidence, by using Bayesian-style updating.
(5) Look for the clashing causal forces at work in each problem, and consider multiple theories & models.

For the entire list & a more in-depth explanation of these rules, check out this article:
https://fs.blog/2015/12/ten-commandments-for-superforecasters/

Unfortunately, Tetlock's book does NOT provide any advice specifically for predicting geopolitical conflict (maybe due to IARPA secrecy), but it does suggest that by studying the theoretical models of geopolitical conflict and sifting through publicly available information on the latest developments in international relations we could predict conflicts in the short term (roughly the next 1-2 years) at a rate significantly better than chance. This discussion aims to familiarize members with some of the theoretical models of geopolitical conflict & current methods of predicting conflict.

RELEVANT INFO FROM PAST MEETUPS:

The Philly Political Agnostics had a meetup on "The Tactics & Patterns of Political Violence" back on May 28, 2017 which addressed some issues relevant to the 3rd & 4th sections of our meetup. In terms of tactics, we looked at Herbert H. Haines' "radical flank effect" & Eric Chenoweth's "Rule of 3.5%", which both have to do with mobilizing protesters & winning public support for political change. In terms of patterns of conflict, we looked at "Strauss-Howe generational theory" & why it's often dismissed, as well as the more reputable "structural-demographic theory" of Peter Turchin which posits 3 intersecting social cycles that drive conflict (i.e. popular immiseration, intra-elite competition, state fragility). You can review that material here:
https://www.meetup.com/Philadelphia-Political-Agnostics/events/239786314/

Since we're consider the causes of revolts & coups, it may also be worth looking at the first half of the outline from a meetup we had on February 5, 2017. We looked at Bruce Bueno de Mesquita's selectorate theory from "The Dictator's Handbook" and Daron Acemoglu & James Robinson's institutional theory from their book "Why Nations Fail", and both dealt the way rulers must gauge the portion of the populace they have to satisfy in order to stay in power. In a "rentier state" whose economy is based primarily on resource extraction, the ruler's power base is small, whereas in a modern industrialized nation with an educated populace, the power base is typically much larger. You can review that material here:
https://www.meetup.com/Philadelphia-Political-Agnostics/events/jlzgxlywdbgb/

NOTE: If you're interested, this discussion will be immediately followed by a Skeptics meetup on "Intelligence Failures & Conspiracy Theories" that is somewhat complimentary to this discussion. It will address competing explanations of the 9/11 attacks, the 2003 Iraq invasion, and U.S. intervention in the civil wars in Libya & Syria. To check out the discussion outline & RSVP, go here: https://www.meetup.com/Philly-Skeptics/events/254780511/

----------------------------------------------------------------------------

DIRECTIONS ON HOW TO PREPARE FOR OUR DISCUSSION:

The videos you see linked below are intended to give you a basic overview of geopolitical forecasting & current efforts to predict conflict. The easiest way to prepare for our discussion is to just read the numbered videos linked under each section - this should take about 55 minutes total to get through. The articles marked with asterisks are just there to supply additional details. You can browse and look at whichever ones you want, but don't worry - we'll cover the stuff you missed in our discussion.

In terms of the discussion format, my general idea is that we'll address the 4 topics in the order presented here. I figure we'll spend about 30 minutes on each section. I've listed questions under each section which we'll do our best to address.

----------------------------------------------------------------------------

I. CAN WE PREDICT THE ONSET OF WARS WITH THE BARGAINING MODEL & RATIONAL, UNITARY ACTOR ASSUMPTION? CAN RESOURCE CONFLICTS & COMMODITY PRICES PREDICTS WARS? HOW CAN NEWS REPORTS ALLOW US TO GAUGE A STATE'S INTENTIONS & WAR READINESS?

William Spaniel, "Understanding War" (video - 6:51 min.)
https://www.youtube.com/watch?v=bq1YHbFTAuE

1b) Michael Klare, "The coming resource wars" (video - 5:29 min.)
https://www.youtube.com/watch?v=ETEOdtIm_ac

Emily Meireding, "Oil Wars: Why Nations Aren't Battling Over Resources"
https://www.washingtonpost.com/news/monkey-cage/wp/2016/05/19/oil-wars-why-nations-arent-battling-over-petroleum-resources/
Patrick Tucker, "How the Internet Could Have Predicted the Invasion of Ukraine"
https://www.defenseone.com/technology/2014/04/how-internet-could-have-predicted-invasion-ukraine/82480/
Richard Laurent, "Using the internet to predict the future of the South China Sea dispute"
https://medium.com/@predata/using-the-internet-to-predict-the-future-of-the-south-china-sea-dispute-7ff9fc6cee38

II. CAN WE PREDICT WHETHER OR NOT A COUNTRY WILL DEVELOP WMDs USING SELECTORATE THEORY? CAN SPY SATELLITES & WEAPONS INSPECTIONS PREDICT WMD CAPABILITIES? IS IT DANGEROUS TO ASSUME ALL LEADERS WITH WMDs ARE RATIONAL & DETERRABLE?

Bruce Bueno de Mesquita, "A Prediction for the Future of Iran" (video - 19:05 min.)
https://www.ted.com/talks/bruce_bueno_de_mesquita_predicts_iran_s_future

Ariana Rowberry & Jane Vaynman, "Keeping Inspections in Perspective"
https://www.brookings.edu/blog/up-front/2013/11/26/keeping-inspections-in-perspective/
Nicholas L. Miller & Vipin Narang, "North Korea Defied the Theoretical Odds: What Can We Learn from its Successful Nuclearization?"
https://tnsr.org/2018/02/north-korea-defied-theoretical-odds-can-learn-successful-nuclearization/
Joshua Keating w/ Alastair Smith, "When Chemical Weapons Are Smart Politics: Why did Assad use chemical weapons?"
http://www.slate.com/blogs/the_world_/2013/09/03/why_did_assad_use_chemical_weapons_the_author_of_the_dictator_s_handbook.html

III. WHAT UNDERLYING FACTORS CAN HELP PREDICT CIVIL UNREST & WHEN IT'S LIKELY TO LEAD TO A MILITARY COUP OR POPULAR REVOLT? HOW DO THE CAUSES OF A COUP VS A REVOLUTION DIFFER & CAN SELECTORATE THEORY HELP EXPLAIN THIS? CAN WE PREDICT THE LIKELY RESULT - I.E. DEMOCRACY, AUTOCRACY, OR STATE FAILURE?

3a) Vice Motherboard w/ Yaneer Bar Yam, "The Math That Predicted the Arab Spring" (video - 2:47 min.)
https://www.youtube.com/watch?v=KpUSNxuuwfA

3b) The Economist, "Predicting unrest in the Arab world [in 2011]" (video - 2:21 min.)
https://www.youtube.com/watch?v=P30s4-94C1A

Andreas Beger & Michael D. Ward, "Where are coups most likely to occur in 2017?"
https://www.washingtonpost.com/news/monkey-cage/wp/2017/01/31/where-are-coups-most-likely-to-occur-in-2017/
Michael Gordon, "Forecasting Instability: The Case of the Arab Spring and the Limitations of Socioeconomic Data "
https://www.wilsoncenter.org/article/forecasting-instability-the-case-the-arab-spring-and-the-limitations-socioeconomic-data
Philip Ball, "News mining might have predicted Arab Spring. Signs of impending social and political change may lie hidden in a sea of news reports."
https://www.nature.com/news/2011/110913/full/news.2011.532.html
Matthew Hardcastle, "The predicted failure of the 'Arab Spring'"
https://phys.org/news/2017-02-failure-arab.html

IV. WHAT UNDERLYING FACTORS CAN HELP US PREDICT WHEN ETHNIC, RELIGIOUS OR POLITICAL TENSIONS WITHIN A SOCIETY MIGHT LEAD TO CIVIL WAR OR MASS ATROCITIES? IS THE "FRAGILE STATE INDEX" A USEFUL WAY OF MEASURING THESE TENSIONS? CAN WE MONITOR DEHUMANIZING SPEECH TO HELP PREDICT THE BUILD-UP TO GENOCIDE?

4a) NowThis World, "Which Countries Are About To Collapse?" (video - 4:28 min.)
https://www.youtube.com/watch?v=GgVmn66oK_A

4b) Jay Ulfelder, "Forecasting Conflict" (video - 17:38 min, start at 2:04)
https://www.youtube.com/watch?v=HmsPnunzgJ0#t=2m04s