This study investigates the causality between energy consumption and GDP in Tunisia for the 1980 to 2009 period at both aggregated and disaggregated levels as oil, natural gas, and electricity. To determine the Granger causality in the presence of cointegration between variables, a vector error correction model (VECM) is used instead of an autoregressive model (VAR). In the short–run, the neutrality hypothesis is supported between total energy consumption and GDP. This is also true between GDP and oil consumption in one hand and the gas consumption in other hand. Whereas a unidirectional is detected from electricity to the GDP is found (growth hypothesis). In the long–run, total energy consumption, in aggregate and disaggregated forms (gas and electricity) causes GDP (growth hypothesis). For against GDP causes oil consumption (conservation hypothesis). Consequently, the policy makers in Tunisia should place priority an increased commitment to aggregated and disaggregated energy consumption will stabilise the country's inefficient spending and allow it to have a stable income stream in the short–term to raise capital for its long–term investments.
Keywords: causality, economic growth, energy consumption, vector error correction modelling, Tunisia, gross domestic product, GDP, Granger causality