Considering sunk costs in one’s decisions is a cognitive limitation that behavioral economists may underestimate. If you, dear voter, have already lost $100,000 in a project that you are sure will continue to bring you net losses, you will just lose more money by putting more into it. Your previous cost is sunk and won’t be reimbursed to you just because you lose more money. As the dictum goes, don’t throw good money after bad money!
The sunk-cost fallacy is at work in political partisanship too. Suppose you have “invested,” if only emotionally, in politician X. You have followed him for years, applauding his good deeds and forgiving his bad ones: “he just said that to be elected,” “he did not have a choice,” “consider how politician W would be worse,” etc.). Perhaps you even consciously swallowed the lies he wanted you to believe as a badge of loyalty? You were sometimes disappointed, perhaps betrayed, but you had faith that it would get better. It doesn’t get better but you continue to follow. You are a victim of the sunk-cost fallacy if you consider your past emotional investment as a reason for continuing to invest faith, hope, and emotions into X.
You may feel that there is an intellectual or moral reputation reason for doing so. Admitting that you were wrong or have been duped may indeed involve a reputational cost. But waiting longer may impose a still larger reputational cost on you. At any rate, your current decision should not take into account your emotional (or intellectual or moral) sunk costs.
The only reason, it seems, for which a (dismal) economist friend could give you for continuing to follow your politician hero revolves around specific human capital. You may have wasted much of your reputational capital (admittedly a sunk cost) and, in its place, accumulated very specific human capital—the ability to lie or blur the truth, for example—that has little value in honest life pursuits. I am not this dismal economist, but Merry Christmas anyway!