Ensaios em modelos DSGE novo-keynesianos aplicados à economia brasileira

Abstract

This thesis is based on two essays with applications of the DSGE-TANK model to Brazil. In the first essay, a DSGE-TANK model is developed for Brazil, with Ricardian households and non-Ricardian households, and the Galí, López-Salido and Vallés (2007) model is expanded with distortive taxation and introduction of the public capital stock in the production function. The bjective is to discover the effects of government spending shocks and government investment shocks on aggregate private consumption and on Ricardian and non-Ricardian decentralized private consumption through impulse response functions (IRF's) and verify the presence of crowding-in effect or crowding-out effect. The results found from the calibration and Bayesian estimation point to the crowding-in effect of both shocks on aggregate private consumption and non-Ricardian private consumption and the crowding-out effect on Ricardian private consumption in Brazil. Both shocks lead to an increase in output, a reduction in private investment, an increase in inflation, an increase in the nominal interest rate and, subsequently, an increase in taxation on consumption, capital, and labor so that the government can finance its expenditures and investments. Therefore, in Brazil, an expansionist policy of government spending and/or investment has a positive effect on general private consumption and on the consumption of the poorest portion of the population. However, it has a negative effect on the consumption of the portion of the population that has access to the credit market, that is, it reduces the consumption of the richest portion of the population. In the second essay, the objective is to estimate the impacts of shocks from the COVID-19 pandemic in Brazil through a DSGE-TANK model from Galí, López-Salido and Vallés (2007) expanded with distorting taxation, pandemic demand shock (negative shock on consumption preferences), according to Faria-e-Castro (2021), and pandemic supply shock (positive shock on the disutility of work), following Mihailov (2020). The calibration of the model parameters and the IRF's of the variables for the two pandemic shocks were used. The IRF's results showed that both shocks resulted in a drop in aggregate output, aggregate consumption, aggregate working hours, Ricardian and non-Ricardian consumption, and Ricardian working hours. Both shocks had more perverse effects on the consumption of the poorest households, that is, on non-Ricardian households, which do not have access to credit. While the negative shock of consumption preferences presents a rapid recovery in aggregate product and a deflationary effect, the positive shock of labor disutility provides a slow recovery in aggregate product, a stagflationary effect and an increase in public debt.