The sample mean of Cauchy data never concentrates

Suppose $\{X_1,X_2,\dots,X_n\}$ is a random sample from Cauchy $C(\theta,1)$ distribution. A test procedure for testing $H_0:\theta=0$ against the alternative $H_1:\theta\neq0$ is not consistent if the test statistic is based on: