Suppose that our consumer is a voter. She begins with a strong preference for Hillary Clinton; that is, Hillary Clinton is her favorite brand. Then questions are raised about certain aspects of her favorite brand. There are nevertheless strong and rational reasons for the voter not to switch. Let’s again consider the consumer.

First, the questions that are raised may not go to the qualities of the preferred brand that the consumer cares about. Suppose, for example, that she learns that her favorite brand of tuna isn’t “dolphin safe.” If dolphins do not particularly interest her, she might well decide that the costs of switching are too great. Why invest time and energy in considering alternative brands if her preferred tuna still has the taste and texture she likes, and will therefore make a satisfying meal?

Second, even if the questions raised about her preferred brand do involve qualities of the product that matter to her, the consumer may nevertheless decide not to take the risk of choosing another brand. Her favorite toothpaste or her favorite tuna, whatever its flaws, might still seem preferable to the alternatives. The consumer has to price into her decision not only the costs of search but also the risks of selecting the wrong brand.

This, of course, is where the analogy breaks down. With toothpaste or tuna, the cost of trying a new brand is small. If the feared risks materialize — bad at cleaning teeth, bad taste — the buyer can throw the product in the trash and go back to the store. In politics, the potential cost of choosing wrongly is much greater. Thus it is entirely rational that voters with strong preferences will be reluctant to switch brands.