This means: whenever commitment can potentially lead to a better outcome, it is theoretically impossible to implement. This article explains how the result applies to a variety of business and economic policy situations – and how its negative consequences can be addressed in practice by relaxing the “power consistency” axiom in various ways. Power consistency is the property that plans are ranked according to the same rule that would govern decision-making without a plan.
For instance, the executive team of a firm might wish to conduct a performance review of its members, but each executive is worried that it might later lead to a vote to get them fired. Advance commitment to future actions solves this kind of problem: the team votes on the proposal ‘We will review everyone’s performance, but nobody gets fired.’ However, this kind of commitment cannot work, unless (1) there is no problem to begin with (the right decisions would be made even without commitment), or (2) voting is restricted to a limited set of choices (disallowing the ones that are going to cause disagreement), or (3) the power to vote on commitments is passed off to a different group of people with different goals (or the power balance in the existing group is altered, say by giving someone a veto right). ‘Nobody gets fired’ would lose in majority voting to ‘if the chief financial officer (CFO) performed badly, she gets fired, but nobody else’ (everyone except the CFO would support it). But this proposal would in turn lose to one where one more person is singled out to get fired, etc. No agreement is possible except if certain proposals are taken off the table, or the team essentially appoints an arbitrator.
Examples analyzed in the paper include joint decisions by entrepreneurs and investors, banks and borrowers, multinationals and governments, and rivaling factions in civil wars and policy debates.