EMPIRICKÉ OVĚŘENÍ PHILLIPSOVI KŘIVKY, PŘÍPAD USA

This thesis investigates inflation dynamics through the lens of the Hybrid New Keynesian Phillips Curve (Hybrid-NKPC), a prominent macroeconomic framework integrating forward-looking expectations with backward-looking elements. The research revisits the evolution of the Phillips Curve, highlighting its transformation from a simple wageunemployment relationship into a sophisticated model accounting for price rigidity and expectations. Drawing inspiration from seminal contributions, notably Galí and Gertler (1999), this study addresses key hypotheses: whether forward-looking expectations dominate inflation dynamics, the extent to which lagged inflation persists in influencing inflation, and the significance of marginal costs in the Hybrid-NKPC framework. Empirical challenges include addressing endogeneity, selecting appropriate variables, and accounting for inflation persistence. Using quarterly data 2012–2023 from United States, the analysis incorporates variables such as the Consumer Price Index for Urban Consumers (CPI-U), GDP deflator, output gap, oil prices (WTI), and marginal costs (nonfarm labor share). Lagged variables were used to capture inflation inertia, and natural logarithmic transformations were applied for normalization. Two-Stage Least Squares (2SLS) addressed potential endogeneity in forward-looking expectations. The findings reveal that forward-looking expectations, as captured by survey-based forecasts, significantly influence inflation, validating their central role in the HybridNKPC framework. Backward-looking components show short-term significance but diminish over time, challenging the dominance of lagged inflation. Marginal costs demonstrate moderate influence, suggesting theoretical alignment but limited empirical robustness. Supply-side shocks, such as oil prices, exhibit marginal short-term effects. The results emphasize the Hybrid-NKPC's instability, underscoring sensitivity to methodological choices and variable specifications. This thesis contributes to ongoing discussions on inflation modeling, advocating for refined approaches to address the complexities of modern inflation dynamics.