In this paper, we address the problem of associating mobile stations with base stations towards more energy-efficiency, from the perspective of population game. Using the population game, which allows tractable analysis of many selfish mobiles without growing mathematical complexity, our study provides two practical implications on energy-efficient BS associations: (i) how to control so-called association pricing so that an entire cellular network is operated with the goal of optimizing an socially optimal objective, and (ii) how to develop distributed, energy-efficient association algorithms. To that end, we first define a game, where mobile stations are the players, and their association portion for different base stations are their strategies. Then, from our equilibrium analysis, we prove that a simple power-dependent pricing by operators leads Nash equilibrium to be equal to the optimal solution of a social optimization problem (i.e., zero price-of-anarchy). Next, we study three evolution dynamics of mobile stations, each expressed as a differential equation, and connect each of them to a distributed association control mechanism, where all of those dynamics converge to the Nash equilibrium (which is equal to the socially optimal point).