posted on 2025-05-07, 19:02authored bySamson S. J. M'boueke
This dissertation consists of three chapters on the application of local projection (LP) methods in international macroeconomics. In the first chapter, I revisit the empirical role of terms-of-trade shocks as a source of business-cycle fluctuations in emerging and developing economies. Specifically, I test the robustness of the empirical results in Schmitt Grohe and Uribe (2018) to: (1) using a commodity price-based measure of terms of trade to identify exogenous terms-of-trade shocks versus the aggregate indices of export and import unit values commonly used in the literature, and (2) using the local projections method introduced by Jorda (2005) versus the SVAR model.
In the second chapter, I borrow from the Global Vector Autoregressive (GVAR) literature to construct country-specific "foreign" terms-of-trade shocks, and study how these shocks spill over to countries via the trade channel in a LP setting.
In the third chapter, I extend the pseudo-panel local projections method developed by Berg et al. (2023) to shrink impulse response standard errors. Specifically, I introduce a method to increase the chances of obtaining pseudo panels with enough similar units to leverage cross-sectional information in the data and increase the precision of impulse response point estimates. I then develop the "PsPLP" R package to mechanically implement this method.