Studying Nonlinear Relationship between Energy Consumption and Inflation in Agricultural Sector

 Today, the food-energy nexus is a vital issue. Energy in the food production chain is an essential feature of agricultural development and a critical factor in achieving food security. Energy use in the agricultural sector has increased to respond to the growing demand of the population, as well the limited supply of cultivated lands, and the desire for high standards of living. Therefore, the agricultural sector is heavily dependent on energy that affects agricultural prices. Agricultural price fluctuations are one of the most critical challenges for policymakers. The rapid rise in food prices has a significant negative impact on social welfare, especially the poor in developing countries, which is an issue that is more critical in developing countries than in developed countries. According to the Food and Agriculture Organization (FAO) report in 2018, the food world price index increased from 89.6 to 229.9 during the period from 2002 to 2011. Our literature review shows a distinct lack of research on modeling and analyzing the linkage between agricultural input price shock, especially energy and agricultural commodity prices in Iran.

The Markov Switching model is a popular non-linear time-series model that involves multiple equations and can characterize the time-series behaviors in different regimes. This model is suitable for describing correlated data that exhibit distinct dynamic patterns during different periods. So, considering the sensitivity of food security and the impact of agricultural input, the main objective of this paper is to develop an econometric model to gain reliable insight into the impact of energy consumption on agricultural inflation, using the Markov Switching approach. To estimate this equation, we will run a MS-AR model, some preliminary tests, such as unit root test and stability test, are employed to ensure the reliability of MS-AR estimation results.

Due to use of time series data, it is necessary to check the stationary status of variables. We performed a common non-linear unit root test (Kapetanios, Shin and Shell (KSS), Zivot and Andrews, Lee and Strazicich). These results reveal that we can significantly reject the null hypothesis of unit root for API, PPI, FPI, and EC, implying that all four variables considered in this study are stationary with structural breaks at levels. The Markov-Switching model has the various types that each of these is a particular component of the regime-dependent equation. Therefore, to choose the best type, the Akaike information criterion was used, and the model with the minimum value was selected as the optimal one. After model estimation and selection, the LR test indicated that the hypothesis of linearity could be rejected in favor of a Markov switching model. According to this model, the period of the Markov switching model estimation is classified into two regimes. Approximately, all the estimated coefficients of the MSIAH (2) - AR (5) model are found to be significant at the conventional level.

The estimation results are consistent with theoretical foundations illustrating the importance of input prices and energy consumption on agricultural commodity prices. As with most experimental studies reviewed, this study has also shown energy consumption has a negative impact on agricultural commodity prices. In other words, it can be contended that during the study period, agricultural input prices have been influential factors on agricultural commodity prices. The findings revealed that the low inflation rate and high inflation rate regimes are stable and that only extreme events can switch regimes. The results of the MS model showed that the effect of input prices on agricultural inflation is different in regimes. In the case of energy, the impact of energy consumption on agricultural commodity prices in the high inflation rate regime is less than the low inflation rate regime because the elimination of energy subsidies policy has been applied in the second regime (high inflation rate). Thus, the results indicate the asymmetric impact of energy consumption shocks on agricultural commodity prices. The effect of agricultural input prices on agricultural commodity prices indicates that Iranian agriculture is significantly affected by changes in input prices. In this study, changes in input prices were caused by various shocks, such as the elimination of energy subsidies and drought. Therefore, it can be concluded that the elimination of energy subsidies and drought were, directly and indirectly, able to affect agricultural inflations through the price of inputs. In conclusion, planners and policymakers must pay attention to this asymmetry in agricultural commodity prices volatility to increase the price stability in agriculture as much as possible by appropriate policy tools.