Estimation of short-run inflation dynamics is important to policy makers for the effective implementation of monetary policy. My dissertation studies the estimation of inflation dynamics at the aggregate level by considering more disaggregated level data. The first chapter employs a regional framework to study the potential improvement of estimating the inflation dynamics in the US with regional variation and instrument selection. The second chapter switches the angle of disaggregate data by focusing on estimating sector inflation dynamics and implied aggregate inflation dynamics. The third chapter studies the inflation dynamics in the euro area by estimating country-specific inflation dynamics of each member state. All three chapters study the inflation dynamics in a hybrid New Keynesian Phillips curve and the many instruments issue is solved by instrument selection using a kernel-weighted Lasso method introduced in the first chapter.The first chapter examines the inflation dynamics in a hybrid New Keynesian Phillips Curve (NKPC) that encounters the many instruments problem. Monte-Carlo simulations demonstrate that the NKPC can be better estimated in finite samples if the instruments are selected by a kernel-weighted Lasso method. In addition, I consider a disaggregated regional NKPC model and show that this can further help with the efficient estimation of national inflation dynamics. I apply these methods to US data and find that theinflation process is more forward-looking than typically found in other studies.I also find a statistically significant trade-off between national inflation and unemployment in the short run, that is only evident when using disaggregated data.The second chapter examines inflation dynamics for US sectors with emphasis on the various pricing behavior across sectors. It estimates inflation dynamics for the aggregate US economy and for each separate sector. Sectors are assumed to have different pricing behavior and this feature is incorporated in the sector specific New Keynesian Phillips curve. In the model part, I derive the sector inflation dynamics in the NKPC model by assuming asymmetric behavior of the firms, and the result suggests that sector inflation should be examined in separate sector regressions. In the empirical part, I apply US quarterly sector data to estimate the disaggregate and aggregate New Keynesian Phillips curves. I discuss the sector specific results and explain the possible reasons for the heterogeneous behavior across sectors through international competition. More importantly, I find that disaggregate sector inflation dynamics can help uncover a significant relationship between inflation and unemployment at the aggregate level. The third chapter empirically investigates inflation dynamics for the euro area in the presence of heterogeneous economic conditions across member states. It reviews the inflation dynamics since 2000 for the euro area as a whole and for individual euro area countries. Cross-country heterogeneity is considered and incorporated in separate national New Keynesian Phillips curves. Moreover, this paper highlights the improvement of the estimates of aggregate inflation dynamics through national estimates aggregation and instrument selection. In the empirical part, I apply monthly national and euro area data to estimate the national and euro-wide New Keynesian Phillips curves. I discuss country-specific inflation dynamics, but more importantly, I find that disaggregate national estimates can help uncover a significant relationship between inflation and unemployment in the euro area by reducing the standard errors of the implied parameters. Although the Phillips curve is not a sufficient tool to gauge inflation dynamics as already discussed in the literature, a more precise estimate of the relation still helps with monetary policy formation in the euro area for ECB and national central banks.
PDF
download
878987797
028-85220240
