Convergence of resource allocation algorithms is well covered in the literature as convergence to a steady state is important due to stability and performance. However, research is lacking when it comes to the propagation of change that occur in a network due to new nodes arriving or old nodes leaving or updating their allocation. As change can propagate through the network in a manner similar to how domino pieces falls, we call this propagation of change the domino effect. In this paper we investigate how change at one node can affect other nodes for a simple power control algorithm. We provide analytical results from a deterministic network as well as a Poisson distributed network through percolation theory and provide simulation results that highlight some aspects of the domino effect. The difficulty of mitigating this domino effect lies in the fact that to avoid it, one needs to have a margin of tolerance for changes in the network. However, a high margin leads to poor system performance in a steady-state and therefore one has to consider a trade-off between performance and propagation of change.