Data and Methodology
This section will project the parameters of the economic models through its data sources, sample size, and evidently the estimation techniques. This chapter consists of two parts. The first part will introduce the variables together with the reasoning of their selection. The second part will test the variables through hypothesis testing to ensure a fair degree of adequacy for regression analysis.
The study adopts a comparative approach to examine the effectiveness of fiscal policy in stimulating economic growth under periods of high and/or low economic activity in Brazil. This analysis will initiate a Dicky-Fuller test and Augmented Dicky-Fuller test to ensure adequate data before performing the (OLS) regression. Secondary data will be collected from World Bank before being computed on E-views (Econometric software). Hypothesis testing will be validated through the tau test, to assert the significance of the regression coefficients. (Gujurati 2003), F-test for the overall significance of the model (Patterson and Okafor 2007); (R2) provides information about the goodness of fit of a model (Gujurati 2003) and (Adjusted R2) features the overall improvement in significance of the models when new terms are being introduced or removed.
The empirical and theoretical literature review establishes the effective key fiscal and non-fiscal components of economic growth based on economic theories. All seven variables analysed in the literature review are assumed to be the contributors of Brazil’s economic growth. This study will therefore use all these explanatory variables to show their effectiveness to the economy, however in case of strong collinearity, one or two variables will be dropped. Four multiple regressions established from economic growth theories will be run and observed into order to estimate a benchmarking model that would be most effective and efficient for Brazil’s economy. The multiple regression models have been selected as it incorporates multiple explanatory variables and provides information on the effects of all of these variables on the dependant variable. The model will be estimated by the OLS (ordinary least squares) method.
The conclusion will assess the impact of each fiscal and non-fiscal variable on GDP growth. Examining each regression individually will enable us to determine the variable that performs best through the magnitude of their coefficients and their statistical significance. Having compared all four regressions collectively, a conclusion will be made in regards to the fiscal variables that have contributed the most to Brazil’s growth, according to the model, and the variables used in this study.
Data & Time Span
Time series data has been selected, as it is the best technique to assess a nation’s growth and fiscal effects. The sample size of the data will be forty years, representing economic growth in Brazil annually between 1970 and 2010. This period has...