Harmonizing Discretionary Policy Choice with Fiscal Rule Making

It is widely recognized that under the hypothesis of rational expectations economic agents' intertemporal decisions over consumption and investment are affected by the discretionary use of fiscal policy (Kydland and Prescott, 1977). Consequently, it is not surprising that modern macroeconomics is seeking to develop a "coherent framework for the design of economic policy that takes into consideration three relevant aspects: (i) a model to predict how people will behave under alternative policies, (ii) an economic criterion to rank the outcomes of alternative policies, and (iii) a description of how policies will be set in the future" (Chari and Kehoe, 2006).

This research work incorporates these elements into a quantitative policy research approach that studies the long-term effectiveness and vulnerability of a diverse set of fiscal policies, which propose lump-sum transfers and different consumption tax exemptions. The Robust Decision Making (RDM) (Lempert, Popper, Bankes, 2003) framework is used for characterizing the deep uncertainty conditions of the discretionary use of fiscal instruments and translate results into effective robust alternatives that can set the "rules of the game" for improving economic activity and reducing income inequality across Mexican families.

For this RDM study, I developed a Dynamic General Equilibrium (DGE) model for the Mexican economy that incorporates heterogeneity on productive sectors and households, as well as an input-output approach that reveals the structural interdependency between final demand and domestic production, to assess how major macroeconomic aggregates are affected by deterministic shocks on the tax structure. The model is used as in a computational experiment that creates a large ensemble of future scenarios exploring an ample uncertainty space on consumption taxes and labor income levies on the ninth and tenth deciles of total income. This large database of future scenarios is then analyzed using statistical clustering algorithms to identify the factors that increase or reduce the vulnerability of policy alternatives. Finally, the vulnerability of fiscal options is analyzed in light of two economic targets: the percentage variation of real GDP and GINI index with respect to the initial steady-state conditions.