By linearity of expectation, . The variance of a sum and the negative covariance just proved give
Apply Chebyshev's inequality to the empty-box indicators in uniform allocation:
This proves the required concentration of the empty-box fraction. Since , it also yields convergence in probability of to .