In a traditional Turing test, a programmer tries to write a computer program which is indistinguishable from a real human in conversation. In an ideological Turing test, {discussants try to present each other’s views in a way indistinguishable from how their opponent would present them}.
Coined by {Bryan Caplan} in The Ideological Turing Test - Econlib, in response to a self-congratulatory passage from Paul Krugman:
If you ask a liberal or a saltwater economist, “What would somebody on the other side of this divide say here? What would their version of it be?” A liberal can do that. A liberal can talk coherently about what the conservative view is because people like me actually do listen. We don’t think it’s right, but we pay enough attention to see what the other person is trying to get at. The reverse is not true. … We have views that are different, but they’re arrived at through paying attention. The other side has dogmatic views.
Pragmatically, Caplan’s suggestion is that each party responds to a survey on their views, then attempts to guess the responses of their counterparty to the same survey. The implication is that if one side can’t guess their opponent’s responses, they don’t understand the arguments very well.
Related: Steelman argument