| In a small structural model of the economy with mircofoundations, three equations summarize the optimizing behavior of both household and firms: an IS curve, a Phillips curve, and a money demand function. If utility is non-separable between real balances and consumption, then current and expected future real balances enter into the forward-looking IS curve. In that case, real balances also enter into the forward-looking Phillips curve, and, consequently, real balances affect the dynamics of both inflation and real output. In the first chapter of my thesis, I explore the role of money in this theory. The first chapter estimates a popular specification of the IS curve using both narrow and broad measures of money as well as Divisia indexes. I find that most measures of money are significant long run determinants of U.S. output.;In Chapters 2 and 3, I explore the Phillips curve. The lack of empirical success for the New Keynesian Phillips curve has motivated empiricists to find alternative formulations of expectations and alternative proxies for marginal cost. In Chapter 2, I replace the rationally formulated inflation expectations with survey expectations of inflation in the New Keynesian Phillips curve. I find that survey expectations have more explanatory power in the Phillips curve specifications when compared to rational expectations (derived using instrumental variables) as well as purely backward-looking specifications.;The third Chapter extends one of the findings in Chapter 2. Labor's share of income has been put forth as a superior proxy for real marginal cost than measures of the output gap. Using common empirical test in the current literature I find labor's share of income is insignificant in the both the forward-looking Phillips curve and the hybrid Phillips curve when the Nixon price control years are omitted from the data set. |
