This dissertation presents a series of three essays that examine the functional form of the U. S. federal income tax and its implications. In the first essay we introduce the convex functional form of the income tax which we believe is superior to the standard income-proportional form. We also describe the parameters within this function and their construction over the years from 1913 to 2005. The second essay discusses the characteristics of the time series of these parameters, the relation of these series to other tax series in the literature, the relation of the intertemporal variation in the tax parameters to the sharp reduction in volatility of macroeconomic time series after about 1950, and the interrelation of the tax parameters with other federal fiscal variables. In chapter three, we use a standard dynamic stochastic general equilibrium model and insert our tax function. We explore the implications that different tax policies will have on the macroeconomy by changing parameter values within this tax function. Specifically we compare the steady states values, second moments, and impulse response functions, of the usual variables, generated by these policies.