3. The propensity score has several properties, the first of which is its balancing property. Recall the propensity score is the conditional probability of exposure to treatment (T=1) rather than control (T=0) given observed covariates, Z. If I sample at random treated subjects, those with T=1, and control subjects, those with T=0, with the same value of the propensity score, then what does that tell me about the distribution of Z between the treated and control subjects?