The effects of expansionary fiscal policy on macroeconomic development: A study on tax competition in an emerging market.

This dataset was assembled to examine the macroeconomic effects of expansionary fiscal policy implemented through tax incentives under the state value-added tax on the circulation of goods and services (ICMS) in Santa Catarina, Brazil. The underlying research hypothesis is that tax benefits granted by subnational governments, as part of a strategy of fiscal competition, are capable of stimulating macroeconomic development, particularly economic activity and employment.
The data consists of monthly macroeconomic time series spanning the period from January 1997 to December 2020. The variables capture key dimensions of the state economy, including the real exchange rate, real interest rate, electricity consumption (used as a proxy for economic activity), exports, imports, formal employment, ICMS revenue, and inflation. These series were collected from official Brazilian institutions and state agencies, ensuring consistency, reliability, and comparability over time.
The dataset supports the estimation of a vector autoregression (VAR) model designed to trace the dynamic responses of macroeconomic variables to shocks in ICMS revenue, which reflect changes in both economic activity and fiscal policy, including the granting of tax incentives. The empirical results reported in the associated article indicate that ICMS shocks generate short-term effects on the exchange rate and imports but produce limited or statistically insignificant impacts on economic activity, employment, inflation, and state tax revenue over time. These findings challenge traditional Keynesian expectations of broad-based stimulus effects and are more consistent with neoclassical and public-choice interpretations of tax incentives as potentially inefficient and weakly connected to employment generation.
All series were seasonally adjusted, deflated when appropriate, and transformed into logarithmic or percentage variations to ensure stationarity and suitability for time-series econometric analysis. The repository includes the original dataset used in the estimations, as well as a PDF log documenting the Stata commands and outputs employed to estimate the VAR model. Together, these materials allow users to understand how the data were gathered, processed, and analyzed, and to replicate or extend the empirical analysis in future research on fiscal policy, tax competition, and regional economic development.