Changes

Consider a two-dimensional case. any Any policy <math>[z_{i}] <math> is characterized by coordinates (x_i, y_i)

In first stage we vote on x_i <math>x_{i}<math> and obtain policy equal to median voters bliss point x_m. In second stage we vote on y_i and obtain policy equal to median voters bliss point y_m, so we obtain unique outcome z=(x_m, y_m)T is always distributed equally among n districts so <math>t_{i}=T/n</math>.