Ma, Sicong2023-02-152023-02-152023-02-15http://hdl.handle.net/10393/44625http://dx.doi.org/10.20381/ruor-28831Referring to the results of most empirical studies, an increase in expenditures on public capital leads to an appreciation of the real exchange rate. This result contradicts the findings in theoretical papers. In the latter studies, the real exchange rate depreciates, following a rise in public capital spending. The first chapter examines the puzzle related to the impact on the real exchange rate of government spending on physical capital. We have developed a theoretical model to sort out the puzzle, showing that the real exchange rate can move in both directions following an increase in public spending on capital. In contrast to some existing theoretical studies, we depart from an executive focus on the impact of government spending on capital on the demand side of the economy. Considering the impact of public capital on the supply side of the economy, we developed a small-open New Keynesian Dynamic Stochastic General Equilibrium model with tradable and non-tradable goods, real and nominal rigidities, and productive government spending. The latter affects the productivity of private factors aside from impacting the demand side of the economy. We explored the role played by the output elasticity of public capital on the dynamics of the real exchange rate. By being able to focus on both the supply and demand sides of government spending on public capital, we can shed light on its net effects on the dynamic response of the real exchange rate. Specifically, we assessed the responses of the real exchange rate and other real variables following innovations in government spending on public capital, and ran some sensitivity analyses on the magnitude of the output elasticity of public capital and of other key parameters of the model. The second chapter explores the dynamic effects of government spending on private consumption and other macroeconomic variables. Responding to a rise in government spending, both neoclassical and New-Keynesian models predict that private consumption will decrease. However, empirical analyses find a rise, no change, or a fall in private consumption. To address the results' inconsistencies, this paper distinguishes public consumption from public investment because public capital is productive. Since the traditional structural vector autoregressive (SVAR) models neglect the forward-looking behaviours of economic agents, we built SVAR models with rational expectations motivated by a New-Keynesian model. We found a rise in public investment in response to a shock to public consumption, and identified 'deep' structural parameters (not subject to the Lucas critique) via full information maximum likelihood estimation. Our findings suggest that an increase in public consumption associated with a positive response to public investment can result in an increase in private consumption. Blocking the response of public investment to a rise in public consumption leads to a decrease in private consumption. Some studies have suggested that when the composition of government expenditures (consumption vs. investment) is suboptimal, a shift from public investment to consumption, or from public consumption to investment, may boost growth. The third chapter uses autoregressive distributed lag (ARDL) models to investigate, for the United States, the impact of the composition of government expenditures on real GDP growth. It is apparent that increasing the ratio of total public consumption to total public investment has a positive effect on GDP growth. In a disaggregated analysis including different government levels, we find that increasing the public consumption-to-investment ratio at the federal level does not affect GDP growth, indicating that the federal public spending composition is optimal. In contrast, an increase in the ratio at the state and local level has a positive impact on GDP growth. This suggests that the ratio of public consumption to investment is less than optimal at the state and local levels. It follows that state and local governments can reduce public investment to finance public consumption in order to promote GDP growth, while keeping government expenditures constant. The ARDL results still hold up in a simple linear model with Ordinary Least Squares (OLS) estimation. Our findings offer an alternative explanation for the observation in some studies that the government prioritizes public consumption over public investment, despite the high productivity of public investment found in several studies.enAttribution-NonCommercial-NoDerivatives 4.0 Internationalhttp://creativecommons.org/licenses/by-nc-nd/4.0/Government spendingReal exchange ratePrivate consumptionEconomic growthThesis