Data for: The transition between energy efficient and energy inefficient states in Cameroon
Abstract of associated article: I use a two-state (energy efficient/inefficient) Markov-switching dynamic model to study energy efficiency in Cameroon in a novel manner, employing yearly data covering 1971 to 2012. I find that the duration of an energy inefficient state is about twice as long as an energy efficient state, mainly due to fuel subsidies, low income, high corruption, regulatory inefficiencies, poorly developed infrastructure and undeveloped markets. To escape from an energy inefficient state a broad policy overhaul is needed. Trade liberalization and related growth policies together with the removal of fuel subsidies are useful, but insufficient policy measures; the results suggest that they should be combined with structural policies, aiming at institutional structure and investment in infrastructure.
Steps to reproduce
I used the OXMetrics software to estimate the two-state Markov-Switching model. The following steps were followed. 1. checked data for nonstationarity 2. nonstationary series were differenced. 3. i determine the appropriate autoregressive lag structure by using different model diagnostic statistics 3. For the model estimation, I first estimated a univariate autoregressive model with state-dependent effects 4.I then control for the effects of other variables. At this stage i do not allow the effects of these variables to be state dependent. 5. I relax the assumption of state-independent effects of the regressors. 6. Codes for each stage is included in the Batch code file.