The U.S. Holocaust Memorial Museum recently published a report evaluating U.S. engagement in the Syrian civil war. The report examined whether counterfactual U.S. policy decisions at five critical junctures would have reduced atrocities against civilians, and it was immediately met with outrage over its summary finding: Compared with President Barack Obama’s actions, alternative policy choices were not “realistic and clearly more effective in mitigating atrocities.” The museum has since retracted the report.
Most criticism came from intervention advocates, who branded the report a partisan whitewashing of Obama’s failed Syria policies. But critics mostly neglected its substance. Employing a variety of methodological approaches, the report is a serious examination of the Obama administration’s response to a profound humanitarian tragedy. Setting the political furor aside, how would political science assess its counterfactual analysis?
What the report actually says
The report — including a survey and interviews with Syria experts and former officials, game-theoretic and computational models, and a historical review — offered more criticism of Obama’s policies than either the report summary or its detractors suggest. While contributors did not universally agree, they at various points critiqued the following:
Obama’s August 2011 call for Syrian President Bashar al-Assad to step aside, which foreclosed diplomacy and locked the United States into a false promise of regime change.
Obama’s repeated rejection of no-fly zones, which could have reduced atrocities.
Obama’s failure to uphold his chemical weapons “red line” in 2013 with airstrikes, which could have deterred atrocities and rejuvenated the peace process.
The report also pans the Obama administration for misjudging the inevitability of Assad’s demise, the commitment of his regional allies and the United States’ ability to prevent conflict spillover.
Critics especially derided the report’s agent-based computational model (ABM). In brief, ABM is a computer program designed to approximate “real Syria” and simulate how Syrians might interact with one another to produce violence.
While critics deemed Virtual Syria “inherently absurd,” their sweeping condemnation of mathematical modeling is misguided. Traditional game-theoretic models have produced major insights in war and conflict studies, including civil wars. Unlike those models, ABMs are often too complex to solve by hand. Yet complexity is a feature, not a bug; computational processing permits ABMs to incorporate more real-world nuances into their simulations.
Some criticism is warranted. Good models are relatively transparent, basing the design on expert advice and widely accepted theories of human behavior. Unfortunately, the Holocaust Museum report does not explain how agents’ preferences are defined or how agents interact. It is therefore difficult to trace how the model explains violence in Syria. Some results are difficult to rationalize. For example, the authors find little support for sectarianism as a driver of conflict, but most other accounts emphasize sectarian violence as a defining aspect of the Syrian conflict.
ABM techniques will never be fully transparent to lay audiences, but they should be accessible enough for informed readers to see how their intuitions align with the model’s design. Here, a more detailed description of the background mechanics may have allayed understandable concerns that “Virtual Syria” is divorced from reality.
Inevitable counterfactual uncertainty
What these contributors do not say is that counterfactual options would have clearly reduced atrocities. The authors emphasize that the effects of U.S. actions were uncertain and contingent on other countervailing forces, chief among them possible counter-escalation by Russia, Iran or Hezbollah.
Perhaps a more muscular foreign policy would have deterred these parties from defending Assad; if so, then Obama’s timidity was a grave error. But if Assad’s allies were highly committed, then supporting the rebels would only elicit retaliatory escalation, worsening the killing.
The likelihood of counter-escalation
While the report flags “relentless counter-escalation” as a potential trigger for increased atrocities, it leaves this question unanswered: How likely were Assad’s allies to counter-escalate in the event of U.S. intervention?
To start, states often fail to consider their adversaries’ core interests. Even if U.S. intentions were purely humanitarian, Russia would probably interpret intervention against its loyal Syrian ally as an effort to cripple its regional influence. For Iran, that interpretation is accurate; intervention hawks boast that U.S. engagement would weaken Iran in its cold war with Saudi Arabia.
As a result, observers overestimate the United States’ ability to deter counter-escalation. For example, archival studies of Mao Zedong’s Korean War telegrams show that neither U.S. bombing of Chinese cities nor a demilitarized buffer zone would have averted China’s surprise counter-invasion; China’s security imperative made escalation inevitable, even at great cost.
Let’s assume that these security motives were not overriding, such that the United States could deter counter-escalation. Scholarship suggests that two conditions are vital to successful coercion; opponents would have had to believe that the United States was highly committed (asymmetry of resolve) and that counter-escalation would be forcefully prevented (escalation denial).
In hindsight, it would have been difficult to achieve either condition in Syria. Escalation denial would require targeting Russian or Iranian troops, but the risk of accidental escalation has historically led major powers to avoid direct conflict, instead settling for prolonged proxy wars. Moreover, it is difficult to signal resolve ex ante; opponents would likely doubt the United States’ willingness to bear the long-term costs of intervention. Along these lines, limited military intervention may actually signal weakness by revealing the upper limit of American involvement. These difficulties explain why U.S. coercion fails far more often than it succeeds.
Moral imperatives, limited intervention
The report’s harshest critics believe that inaction in Syria is inexcusable. Their conviction reveals the heart of the problem: The United States writ large was not so resolute. For this reason, the report only considers limited (“plausible”) uses of force. As Andrew Kydd, author of one of the report’s components, explains: “After 15 years of war in Afghanistan and Iraq … even the harshest critics of the Obama administration ruled out a large commitment of ground troops.”
These domestic constraints were real. Yet the report does not demonstrate that the United States could do nothing to stop the killing; rather, it suggests that this goal may have required a more serious American commitment. While the United States wanted to buy democracy in Syria on the cheap, sometimes limited engagement will not suffice. Sometimes the “responsibility to protect” comes at a price.