Cass Sunstein Is Sleeping on the Couch Tonight

Is Obama's groupthink guru (and Amb. Samantha Power's husband) opening a window on White House dysfunction?

John Moore/Getty Images
John Moore/Getty Images

Cass Sunstein, the former head of the Office of Management and Budget’s Office of Information and Regulatory Affairs, has long advocated the notion that people left to their own devices often make bad choices and that government has a responsibility to "nudge" people toward better outcomes than they would select for themselves. But now he argues that government, left to its own devices, also makes bad choices. (Wait, wasn’t he supposed to fix this problem?) This is a welcome Damascene conversion from someone who advocates expanding the Bill of Rights to include the right to education, a home, health care, and protection against monopolies. Sunstein’s realization is a damning criticism of the Obama administration’s philosophy of government and, incidentally, of its approach to foreign policy, in which Sunstein’s wife, Samantha Power, ambassador to the United Nations, plays a starring role. 

In an article in the Journal of Institutional Economics, Sunstein and his co-author, Reid Hastie, argue that the very process of deliberation serves to amplify mistakes. They argue that the process of deliberation often conveys to individuals disincentives for providing information that would produce better outcomes. Specifically, they highlight the way people self-censor "out of respect for the information publicly announced by others" or to avoid "the disapproval of relevant others." All this is a fancy way of saying that groups tend to reinforce their initial biases through selective information. Sunstein and Hastie conclude — much as the butterfly flapping its wings causes a hurricane — that these "micro mistakes" lead to macro policy failures, even catastrophes. And the Syria policy of Barack Obama’s administration illustrates their arguments rather neatly. 

The authors argue that corroboration by other members of a respected group raises confidence in its own judgments and reduces the variance of opinion, whether or not their taken position is correct, leading to "sharing a view in which they firmly believe, but which turns out to be wrong." Groups actually don’t defer to internal experts; they tend to adopt positions that the majority supports. Sunstein might have drawn from the vast data trove provided by a National Security Council staff that included his wife (one of America’s leading human rights experts) yet consistently avoided values-based policies and prides itself on being "realist."

In evaluating whether groups correct for or compound the errors of the individuals that compose them, Sunstein and Hastie identify four particular types of error: amplifying mistakes, cascade effects, growing extremism, and the dominance of shared information. Let’s take them one by one.

Drinking the Kool-Aid.
In this type of error, the biases individuals bring to the table cause them to give undue weight to corroborative information. The administration believes that the president was elected to end the wars in the Middle East, that military force achieves no political purposes, that intervention cultivates international hostility, and that international cooperation is the only means of solving foreign-policy problems. So it’s easy to see that the administration would not have been open to data suggesting that problems in Syria would worsen absent early intervention, that discreet uses of allied force could affect the military balance between the Syrian government and rebels, and that countries in the region would welcome us dealing a setback to Syrian barbarism and Iranian proxies — not to mention that there are compelling moral and practical interests in acting even when other states will not.

How can it be that the Obama administration has clung so long to a hopelessly failing policy on Syria? "Groups are more likely than individuals to escalate their commitment to a course of action that is failing, and even more so, if members identify strongly with the groups to which they belong," write Sunstein and Hastie. Political appointees generally identify with their administration, especially so if they consider the politician they work for "transformational," which the Obama appointees most certainly do.

Keeping your head down.
These occur when "participants ignore their private knowledge and rely instead on the publicly stated judgments of others." Take the example of then-Secretary of State Hillary Clinton in March 2011, when she described Bashar al-Assad as a reformer. After Syrian security forces were killing peaceful protesters for months, this stated opinion carried administration policy. But experts like Ambassador Robert Ford in Damascus knew that things were going from bad to worse in Syria. Yet no clarion calls of "never again" rang forth from those who had successfully intervened in the Balkan wars or who compellingly chronicled the Rwandan genocide. Clinton had silenced the Cassandras. But the Cassandras were right.

A second type of cascade effect is said by Sunstein and Hastie to occur when people anticipate and seek to avoid censure within a group, an example of which might occur if, say, a prominent human rights activist and chronicler of an earlier administration that failed to act to prevent mass violence was previously denied a diplomatic post because of her impolitic description of she who would become secretary of state. One can see why someone like this might not want to speak up again. Such activity certainly disincentivizes correcting the mistakes of others.

Yet another type of cascade is caused when groups "draw an undue inference from some failure, thinking that similar plans will fail too." Like if the Obama administration concludes that all interventions must be like the Iraq intervention and therefore all outcomes of interventions will be like the outcomes of Iraq. Or that since the intervention in Iraq faced international condemnation, an intervention in Syria would be likewise received.

Sticking to your guns.
Sunstein and Hastie highlight research showing that "members of a deliberating group end up adopting a more extreme version of the position toward which they tended before deliberation began," and they argue that "the problem is especially severe for groups of like-minded people." Put simply, people reinforce each other’s biases, leading to more extreme positions that reflect their "pre-deliberation tendencies" (i.e., their incoming beliefs before they even started talking). This goes some way in explaining the Obama administration’s predilection for announcing policy reviews that serve only to confirm the previous policy. After allowing President Obama’s Syria red line on chemical weapons to be crossed, the White House announced with much fanfare a review of Syria policy, which resulted in reconfirming for the president that he’d been right all along about doing nothing and that no good options are available.

It’s on a need-to-know basis.
Institutions harbor knowledge that should lead them to accurate understandings, but because of their group dynamics, they suppress that knowledge. As Sunstein and Hastie describe it, "members tended to share positive information about the winning candidate and negative information about the losers. They suppressed negative information about the winner and positive information about the losers." So it is with the Obama administration’s characterizations of the parties to the Syrian conflict: They overweight the concern about Islamic extremists that prevents us from giving them weapons or training to negate the advantages of the Assad regime. Assad’s forces are the beneficiaries of Iranian weapons and the actual participation of Iranian soldiers who are also Islamic extremists, but that information is not considered relevant. Moreover, as policy became more opposed to intervention, administration sources increasingly drew attention to concerns about Islamist rebels, never acknowledging that these choices were affecting the pace of radicalization and the strength of the most virulent al Qaeda-linked groups.

Meanwhile, the U.S. ambassador to the United Nations defends the administration’s policy, saying, "President Obama has put in play every single tool in the toolbox, short of military action…. I’d be careful about suggesting we are not taking the atrocities seriously. This is something the president gets briefed on every day. He’s always asking what we can do." She then gives context to the policy by explaining that "there are other interests at play," such as oil prices and the U.S. economy, and castigating the administration’s critics for supporting a "single-issue" policy when it comes to Syria. Sunstein had a fine case study for his theories without even leaving home.

Samantha Power was once a veritable Delacroix Liberty Leading the People, crying "Never again!" She made her career saying that no longer would America stand by while governments committed mass murder. It was this ringing moral clarity that made her so valuable an asset to the Obama campaign and administration — and that is precisely why its absence in the Obama administration’s foreign policy is such a disappointment. 

Cass Sunstein’s work in behavioral economics may shed some insight into why: the dynamic of policymaking in the administration itself. As Sunstein and his co-author say, "a confident, cohesive, but error-prone group, giving effect to the mistakes of individual members, is nothing to celebrate. On the contrary, it might be extremely dangerous, both to itself and to others."

Kori Schake is the director of foreign and defense policy at the American Enterprise Institute, a former U.S. government official in foreign and security policy, and the author of America vs the West: Can the Liberal World Order Be Preserved? Twitter: @KoriSchake