1. Introduction
The relationship between energy consumption and economic
growth has been much studied using the concept of Granger
causality. Granger causality does not imply that ‘X causes Y’ in the
conventional sense. Instead, as Diebold [1] put it, ‘X causes Y’ in the
Granger sense means that ‘X contains useful information for predicting
Y’. To this point, there is a lack of consensus in the literature.
A major reason for the lack of consensus is that many Granger
causality studies suffer from omitted variables bias. Studies which
conduct Granger causality tests in a bivariate framework are likely to
be biased due to the omission of relevant variables affecting energy
consumption and economic growth [2]. For this reason, some
Granger causality studies examining the relationship between
energy consumption and economic growth have started to include
other relevant variables such as capital and/or labour [3], employment
[4], exports [5], pollutant emissions [6], prices [7] or urbanization
[8]. Most studies for Malaysia on this topic suffer from
omitted variables bias because they only consider the energy-GDP
nexus in a bivariate setting (see [9e11]). Tellingly, most existing
Granger causality studies of the energy-GDP nexus for Malaysia,
which employ a bivariate framework, have failed to find evidence of
cointegration and long-run causality, reflecting the use of a bivariate
model