The Trust Solution

• John Vandivier

David Friedman today in effect asked the question, "who will watch the watchdogs?" <a href="http://daviddfriedman.blogspot.com/2015/05/the-trust-problem.html">He called this the trust problem and it resulted from discussing whether Elizabeth Warren is part Native American. This article contains 5 solutions to the trust problem:

  1. Don't trust anyone. Trustless transactions can be expensive but they occur all the time and in fact they are very useful, secure, and needed in certain cases. Of course, conducting unnecessarily expensive transactions all the time is an inefficient way to use money, but it's better than no action at all.
  2. The scientific method does a great job of resolving the trust issue with a 'trust but verify' attitude. We start from the assumption of the null hypothesis, which is the previously established claim, and we attempt to disprove it. Starting from an assumption of trust is a justified and good actualization of the principle of charity.
  3. Efficient naivete may be the best answer of all. Trust generously, but keep in mind that you will end up trusting when it shouldn't be done from time to time. Try to trust people in a way such that you estimate the marginal expected benefit to exceed the marginal expected cost, without needing the marginal expected cost to be 0, as in the case of certainty.
  4. Distributed trust is another answer. Diversify. Don't put all of your eggs in one basket. Acknowledging that everyone is biased, there may be a canceling out of the direction of bias when you look at many sources.
  5. I think the last answer is to trust yourself and your friends, or to build up your own personal relationships over time or through demonstration of a high propensity to make true claims. If some news organization publishes something and you personally know that author well you can maybe make an estimate of its truthfulness even without knowing the organization well. Unless that person is structurally compromised. But you should also be able to know from your relationship with that person whether or not they are the type to change their behavior significantly or insignificantly from one environment to another.