Suppose that five states reduce income taxes in a given year. You are interested in estimating whether the tax cut has increased saving, and you find that the saving rate for residents of these five states increased by 2 percent in the year after it was introduced. Can you reasonably conclude that the tax cut caused the increase in saving? How would you conduct a difference-in- difference analysis to estimate the impact on saving? What assumption must hold for the difference- in- difference analysis to be valid?