Friday, August 09, 2013

Tyranny of Consensus

Erik Loomis posts a brilliant attack on the idea of consensus-based decision-making by Gerry Mackie:
In the late 1970s, some new members imported a belief popular on the liberal-left, that democracy requires consensus. Consensus groups could function, but were unstable and usually the first to fall. There are several problems. Those with the least to do elsewhere in life have the greatest power in the interminable consensus process. Trust, ironically, is absent, in that no delegation of decision is permitted. The thought of a meeting then becomes so horrifying that a larger and larger scope of decisions is left to informal leadership and clandestine process, an undemocratic outcome. Consensus is always biased to the status quo, but problems usually originate in the status quo; rapid external change worsens the conservative bias. Further, consensus invades the individual personality and demands conformity; dissenters may acquiesce but in doing so are implicitly judged to have compromised the moral ideal. The healthy legitimacy of openly holding different views becomes suspect. Finally, rational unanimity is impossible for a larger class of goals. Just to illustrate with a trivial example, suppose it is time to decide where the crew works in the Spring. Six people want to work in Montana because they have friends there. Two people want to work in California because they have friends there. Three people don’t care. Under majority rule, the crew goes to Montana, and those in the minority might feel they are owed a little deference in some future decision (know to political science as “logrolling”). Under consensus, the different sides are denied the legitimacy of their individual interests, because there is only one rational goal for the group, which one side or another must adopt, or the group disband. Under majority rule one is subordinate to shifting impersonal majorities, but under consensus one is permanently subordinated to every other member in the group.
This is from Mackie's masters' thesis, which was written circa 1990 and draws on his own experience in these collectives in the 70s. This is not some new-fangled idea, and its failures are quite well documented.


PG said...

While consensus is a terrible decision making method when members of a group reasonably have differing goals, in the one time I've been in an organization that used it (journal of gender & law in law school) I mostly was OK with it. In particular, I thought it was fine for our selection of articles to publish, because it encouraged people to state their concerns who otherwise might just have voted against without saying "Actually, this idea is totally old-hat in sociology journals, so we won't be publishing innovative scholarship if we accept this article." Since every member of the organization might be called upon to work on the article, it would be bad to pick something people hated.

On the other hand, OMG were the board elections way too long because of the consensus goal. Scholarship ought to be thoroughly discussed and dissected; there is no way to do that decently about another human being.

David Schraub said...

Interestingly, we only used consensus for selecting the next board. We voted on articles with marbles (everyone had five marbles per article and could give it 0-5, and it had to pass a certain threshold to be picked. So functionally someone could basically stop an article from being published by tanking it, but it wasn't pure consensus).