Transformation into anti-manipulation method in voting. Changes in properties

Abstract

Abstract This paper examines the properties of the anti-manipulation method in voting. Such a method can be used by committees and similar bodies to ensure that votes reflect genuine preferences. The anti-manipulation method is based on the Borda Count and discourages strategic voting by excluding scores that deviate excessively from the mean. The method does not eliminate strategic voting but diminishes the motivation to apply it. We compare the properties of the Borda Count and the anti-manipulation method. The properties, which are most often found in the literature, were chosen for comparison. Thus, the following properties are examined: consistency, vulnerability to the no-show paradox, vulnerability to the subset choice condition, homogeneity, monotonicity, and vulnerability to the reversal bias paradox as well as the Condorcet winner and loser paradoxes. The anti-manipulation method fails to satisfy most of these properties. A real data example, the voting of a certain jury, is used as a counterexample in most cases.