Consider the Jackknife variance estimator for a population mean $\bar{y}$. Under what sampling design conditions does the Jackknife estimator $v_{jack} = \frac{n-1}{n} \sum_{i=1}^n (\bar{y}_{(i)} - \bar{y})^2$ become exactly equal to the standard variance estimator $s^2/n$?