I’m impressed: Cato’s Michael Cannon bet Karen Davenport, and let her adjudicate the outcome:

I bet Davenport $20 that I could convince her that the following two claims are true:

cannon.jpg

We agreed on three rules.  First, there would be no splitting the
difference or agreeing to disagree — I would either succeed or I would
fail.  Second, Davenport would be the ultimate arbiter of whether I
succeeded or failed.  Third, if I failed, Davenport would have to
explain why she was not convinced.

The outcome:

After my presentation and a subsequent exchange (mostly about the
second claim), Davenport was unconvinced and she took the $40.  She
made two arguments for why I failed:

  1. A crucial part of my argument — the claim by Helen Levy and David Meltzer that there is “no evidence
    that expanding health insurance is a cost-effective way of improving
    health — is not necessarily true and is, in fact, controversial among
    health economists, and
  2. The lives that would be lost by adopting universal coverage (rather
    than a more cost-effective strategy for improving health), would be
    less than the lives lost during the time it would take to conduct
    experiments to determine which strategy is most cost-effective.

You can watch Cannon’s case here.

P.S. Last year Robin refused me analogous terms.