Income elasticity was increasing gradually until 1999. Moreover, like any other panel studies, it cannot capture specific features of the Azerbaijani electricity system.Īs can be seen from Figure 2, income elasticity demonstrates considerable variation across the time period. Second, the study does not analyze the significant changes in electricity system since it ends in 2009.
TEST FOR COINTEGRATION XLSTAT DRIVER
First, the study only considered income (GDP) as a driver of the electricity consumption but other determinants such as price, demographic indicators, and temperature were missed. The study suffered from some shortcomings. They also found that speed of adjustment coefficient is −0.41. They estimate the income elasticities of the electricity consumption to be 0.52 and 0.77 from the FMOLS and ARDL methods, respectively. However, the nature of causality is bidirectional in the long run. The Granger causality test showed that causality runs from GDP to electricity consumption in the short run. They applied Pedroni co-integration tests and Panel Fully Modified Ordinary Least Squares (FMOLS) and Auto-Regressive Distributed Lag (ARDL) methods to the panel data over the period 1990–2009. Third, as Azerbaijan is an example of a resource-rich developing country in transition, we provide a benchmark of analysis, which can be useful for similar countries, which are going through different economic stages.īildirici and Kayikci explored economic growth and electricity consumption nexus in the Commonwealth of Independent States, where Azerbaijan is a member. Second, considering that there are limited studies on electricity consumption in Azerbaijan, we provide additional evidence of the consumption pattern of this country. Put differently, if the true model is a cointegrated regression with TVC but it is estimated using a regression with fixed coefficients, then the regression outcomes would be spurious and totally nonsensical. Otherwise, the regression becomes spurious. It is very crucial to model TVC properly in a cointegrated regression. However, this is not true in a cointegrated regression.
The point is that if you have TVC and ignore it, then you would still expect to have its average value possibly with some bias in a stationary regression. For example, in cointegrated regressions, to allow for TVC is much more important compared to the usual stationary regressions. There are some key advantages of the TVC method in cointegration framework. The main contribution is that we first use the Time-Varying Coefficient cointegration model to explore electricity demand in Azerbaijan and we find and argue that income elasticity is time varying. This study contributes to the existing literature in three ways. Additionally, Azerbaijan has ranked first in the world in terms of real GDP growth in 2005, 20, being about 26%, 35% and 25%, respectively, according to the World Bank statistics. The real GDP per capita slowed down in 20 with the growth rates of −1.23% and 0.85%, respectively, and recovered since then. It faced very rapid economic growth over 2004–2010, and per annum growth rate of the real GDP per capita was 15.4%, which is very huge. Briefly note that the country was in a very deep recession during 1991–1995 with the declining of real GDP per capita by 2.4% per annum, while it witnessed stable socio-economic development in 1996–2003 with the growth rate of real GDP per capita by 8% annually. The main reason for this suspect is that the Azerbaijani economy had different and contrasting development stages since its independence in 1991 (see inter alia). However, we argue in this paper that one can get suspicious about constancy of the elasticities estimated by the earlier studies.
All the earlier studies for Azerbaijan employ conventional econometric methods, where elasticities of electricity consumption with respect to explanatory variables, for example income and price, are assumed to be invariant over time. comprehensively reviewed these studies with their advantages and disadvantages. There are very limited studies investigating electricity demand in Azerbaijan.
0 kommentar(er)
