We use a tax policy example to show how debate on the value of an elasticity parameter translates into a debate about policy choices. To construct this example, suppose that the choice among four particular tax reform options is based on a single measure of efficiency gain. For each reform, we show how the size of this gain depends upon the elasticity of saving with respect to the net rate of return. Moreover, within quite narrow and reasonable bounds for the elasticity parameter, we find regions in which each of three different tax reforms turns out to dominate the others.