THE JOURNAL OF ENERGY AND DEVELOPMENT

Mehdi Sadeghi,

Analysis of Distorted Energy Prices in Iran,

Volume 37, Number 1

Copyright 2012

ANALYSIS OF DISTORTED ENERGY PRICES IN IRAN

Mehdi Sadeghi*

Introduction

nergy pricing is one of the most fundamental tools in the integrated national energy planning (INEP) procedure in developing countries. The scope and objectives of pricing in this integrated framework would include the following. First, social welfare considerations cannot be attained without rational pricing mechanisms. If there are distortions in energy prices they easily result in misallocation of resources. Second, energy conservation cannot be realized by improper energy pricing. For instance, in order for the effective implementation of demand-side management (DSM), rational energy prices are needed. Third, optimal investment decisions are based on demand fulfillments and this will not be realized with distorted energy prices. Fourth, to reach partial equilibrium in the energy sector, prices should be based on marginal opportunity cost of the energy resources and, therefore, it is argued that energy subsidization will inhibit optimal resource allocation. Finally, energy is regarded as an essential input in the production function; thus, distortions in energy prices will cause higher production costs.

*Mehdi Sadeghi, Associate Professor at the Imam Sadiq University in Tehran, Iran, holds a Ph.D. in economics from Tehran University and masters and bachelors degrees in economics from Imam Sadiq University. The authors areas of academic expertise are economic planning, econometrics, energy economics, and energy demand and supply modeling. Dr. Sadeghi previously held positions with the Iranian Ministry of Energy and is currently President of the School of Economic Sciences. The author, who has served as Scientific Secretary of the Eighth International Energy Conference held in Iran, has participated in projects dealing with energy economics and presented papers at international conferences. The Journal of Energy and Development, Vol. 37, Nos. 1 and 2 Copyright 2012 by the International Research Center for Energy and Economic Development (ICEED). All rights reserved.

45

46

THE JOURNAL OF ENERGY AND DEVELOPMENT

The energy sector in Iran is facing a critical situation at the present time (2012). A reduction of oil revenues, coupled with increasing domestic energy consumption, has forced the government to implement policies to reduce energy subsidies in order to manage public expenditures in this sector. Quantitative analysis of the distributional effect of energy subsidies in Iran indicates that they result in the welfare transfer to higher-income households. This contradictory situation enhances the social welfare losses due to energy subsidies, which are the sum of losses due to the price distortion and unjustified redistribution of economic resources to the benefit of groups with higher incomes.1 In this paper, first the theoretical background for energy pricing and distortion (subsidization) is introduced. Next, the econometric model and the empirical evidence of energy price distortions (subsidization) is analyzed using cointegration analysis. The policy implications of price distortions and the main conclusions comprise the fifth and sixth sections, respectively.

Theoretical Background Two basic mechanisms can be considered in order to decrease energy consumption growth rates. The first is the removal of price distortions (subsidies), which is concerned with pricing mechanisms, and the second is autonomous energy efficiency improvement (AEEI), which represents a non-pricing mechanism. Regarding the removal policy for price distortions, the important issue is the estimation of the shadow price and its difference with domestic market prices. In figure 1, efficient energy pricing with shadow prices has been shown. The (subsidized) domestic market-priced demand curve for the energy carrier is given by the curve D(Q), which is the consumers willingness to pay. The dashed curve b3D(Q) shows the shadow-priced demand curve. Here b is a conversion factor that shows the relative distortion or ratio of the shadow price to the domestic market price of the energy carrier (e.g., if the energy domestic market price is one half of its shadow price, then b = 2). MOC(Q) and MOC/b represent the shadowpriced and the domestic market-priced supply curves, respectively. Assume a small increment of consumption DQ at the market price P. The traditional optimal pricing approach attempts to compare the incremental benefit of consumption due to DQ, (the area between the demand curve and the horizontal axis), with the corresponding supply cost (the area between the supply curve and the horizontal axis). Since MOC(Q) is shadow-priced, D(Q) also must be transformed into a shadow-priced curve to make the comparison valid. The shadow cost of expenditure is b(P3DQ). Thus, at the price P, incremental benefits EGJL exceed incremental costs EFKL. The optimal consumption level is Qopt, where the MOC(Q) and b3D(Q) curves cross or, equivalently, where market-priced supply curve MOC(Q)/b and market demand curve D(Q) intersect. The efficient price to

IRAN: ANALYSIS OF DISTORTED ENERGY PRICES Figure 1

EFFICIENT PRICING WITH SHADOW PRICES

47

Source: M. Munasinghe, Energy Analysis and Policy (Guildford, United Kingdom: Butterworth & Co. Ltd., 1990).

be charged to consumers (since they react along the market-priced demand curve not the shadow-priced demand curve) will be Pe = MOC(Q)/b. At this level of consumption, the shadow costs and benefits of marginal consumption will be equal, MOC(Q) = b3D(Q). Since b depends on specific consumption patterns, different values of the efficient price Pe may be derived for various consumption categories. Therefore, Pe = MOC(Q)/b shows a second-best adjustment. On the other side, autonomous energy efficiency improvement simply represents the potential increase in non-price efficiency improvements and the decrease in the energy system supply costs. This mechanism is based on endogenous growth and learning-by-doing modeling, in which it is argued that energy system costs will exhibit a downward trend. Market barriers such as asymmetric information and transaction costs should be removed; otherwise, this mechanism will not be effective. In terms of Irans present situation, the nations energy market has its own special set of characteristics. It has been state-owned for many years and the government has responsibility for production, transmission, and distribution of all kinds of energy carriers. In the case of excess demand, sufficient energy is imported by the government to fulfill demand. All investments needed for the development of the energy sector are financed and managed by the government. Thus, the supply side of the Iranian energy market is exogenous and it is adjusted to the demand for energy. Accordingly, the energy pricing mechanism is government oriented and supported by on-budget subsidies. It is quite apparent that off-budget subsidies are less transparent and generally more variable over time than on-budget subsidies.2 However, the energy price is determined in a socioeconomic context with severe political considerations. Iran is an energy abundant country and the general view of the consumers is that no higher costs would be needed for the provision of energy. Therefore, all the

48

THE JOURNAL OF ENERGY AND DEVELOPMENT

prices are distorted from their marginal costs and, thus, the demand side of the energy market is influenced by this kind of pricing mechanism. The Econometric Model The main determinants of energy demand are economic activities and real energy price. Economic theory postulates that there is a long-run relationship between energy consumption and energy price. This long-run relationship and the consequences of energy price distortion are investigated in this paper. In order to express the long-run relationship between energy consumption and energy price, an autoregressive distributed lag (ARDL) approach, based on cointegration, is used.3 Since sufficient data were not available, a methodology utilizing a vectorerror correction model (VECM) encountered limitations. Emprical Results Stationary tests of the augmented Dickey-Fuller were performed based on energy consumption, real energy price, and national income data during the period 19672009. According to these tests, the time series with which we were concerned are integrated to the order of one. For testing cointegration, the Johansen and Juselius approach was applied.4 This methodology is used because of its advantages in comparison to the Engle-Granger approach.5 At least one long-run relationship was proven among energy consumption, economic activity, and real energy price. Seven ARDL models were designed, and in each ARDL model the optimum lag length was found to be 1. This lag length was derived according to the Schwarz information criterion, the Hannan-Quin information criterion, and the Akaike information criterion. The results of ARDL approach are reported in table 1. Cointegrated vectors in energy consumption will show the long-run price elasticity and long-run income elasticity for the selected energy carriers (table 2). All long-run price elasticities are significant with the exception of natural gas and electricity. For gasoline demand all short- and long-run price elasticities are less than unity. The adjustment coefficient is estimated at 0.19, indicating that, following the occurrence of a shock, it takes at least five years to be removed. The same adjustment coefficient would be for gas-oil demand. The largest adjustment coefficient is for kerosene demand. The short-run price elasticity of kerosene demand is insignificant, indicating an unimportant role of price in the short run. The short-run price elasticities of natural gas and electricity were not significant. All of the findings based on the cointegration approach indicate that the removal of energy price distortions (in fact, subsidization removal) cannot play an effective role in energy conservation in Iran. For a measurable energy conservation plan to be implemented effectively, a considerable energy price adjustment would be

Table 1
a

THE AUTOREGRESSIVE DISTRIBUTED LAG (ARDL) MODELS RESULTS FOR ENERGY CONSUMPTION IN IRAN

Energy Carriers Variables Gas Oil 0.6 (1.04) 0.51 (3.68) 0.005 (0.06) 0.52 (4.42) 0.21 (6.39) 0.12 (3.56) 0.35 (2.82) 0.07 (1.45) 0.1 (0.64) 0.48 (2.33) 0.34 (3.17) 0.17 (5.7) 0.22 (2.77) 0.38 (3.07) 0.91 (0.98) 0.10 (0.19) 2.46 (2.51) 0.41 (5.5) 0.02 (0.60) 0.06 (2.32) Kerosene Fuel Oil Electricity

Gasoline

Liquefied Petroleum Gas (LPG)

Natural Gas

Intercept

1.02 (3.5)

D LGDP

0.44 (3.12)

D LVAT

0.18 (3.89)

D LVAIM

D LRPGA

0.22 (4.01)

D LRPKE

D LRPGO

0.10 (1.57)

D LRPF

D LRPLPG

D LRPNG

D LRPEL

D 68

Ect (-1)

0.19 (5.82)

0.19 (2.30)

IRAN: ANALYSIS OF DISTORTED ENERGY PRICES

Numbers in the parenthesis indicate t statistic; D LPGDP = first difference of natural logarithm of Gross Domestic Product; D LVAT= first difference of natural logarithm of Value Added of Transportation Sector; D LVAIM = first difference of natural logarithm of Value Added of Industry Sector; D LRPGA = first difference of natural logarithm of Real Price of Gasoline; D LRPKE = first difference of natural logarithm of Real Price of Kerosene; D LRPGO = first difference of natural logarithm of Real Price of Gas Oil; D LRPF = first difference of natural logarithm of Real Price of Fuel Oil; D LRPLPG = first difference of natural logarithm of Real Price of LPG; D LRPNG = first difference of natural logarithm of Real Price of Natural Gas; D LRPEL = first difference of natural logarithm of Real Price of Electricity; D 68 = Dummy Variable for natural gas network expansion; and Ect = error correction term.

49

Table 2
COINTEGRATED VECTORS FOR ENERGY CONSUMPTION Gas Oil (LGo)
1.74 (1.03) 1 0.98 (5.55) 0.27 (4.55) 0.56 (5.74) 1.04 (3.83) 0.49 (3.76) 0.21 (2.34) 1 1 1 1.11 (21.41) 0.29 (0.73) 0.49 (0.18) 7.08 (7.05) 1 1.23 (36.11) 0.34 (0.78)
a

50

Energy Carriers Variables

Gasoline (LGa)

Kerosene (LKE)

Liquefied Petroleum Gas (LLPG)

Fuel Oil (LF)

Natural Gas (LNG)

Electricity (LEL)

Intercept 1

5.51 (3.7)

3.11 (1.26)

LGa

LGo

LKE

LLPG

LF

LNG

LEL

LGDP 0.51 (3.93)

0.95 (3.7)

LVAT

0.97 (5.96)

LVAIM

LRPGA

0.47 (3.05)

LRPKE

LRPGO

LRPF

THE JOURNAL OF ENERGY AND DEVELOPMENT

LRPLPG

LRPNG

LRPEL

All variables are in natural logarithm form.

IRAN: ANALYSIS OF DISTORTED ENERGY PRICES

51

needed; yet, such an extraordinary energy price adjustment would cause tremendous and simultaneous inflationary effects. Energy prices in Iran are distorted from their optimal levels. In the year 2009, total distortion for seven energy carriers was equivalent to U.S. $41.9 billion. Total goods and services subsidies for the year 2009 were approximately U.S. $8 billion. For the same year, the per-capita subsidy payment, including the energy subsidy, amounted to U.S. $710. In order to compute the inflationary impact of an energy price adjustment, an energy input/output model has been used. Based on this model, all energy price distortions for the year 2006 were removed and their inflationary impacts calculated. The results show that this removal will cause an inflation growth rate of 65.55 percent.6 Policy Implications Our results indicate that the energy price short- and long-run elasticities are less than unity (tables 1 and 2). Therefore, to obtain considerable energy conservation targets will necessitate extraordinary energy price adjustments. Removing all energy price distortions in Iran will result in an extremely large inflationary impact. In order to challenge this situation, non-price policies (e.g., autonomous energy efficiency improvement) in the energy sector could provide the foundation for optimal energy pricing policies. To fully realize a successful energy efficiency plan, the energy price distortions removal is a necessary condition, but other market barriers also should be focused upon and taken into consideration. Conclusions Clearly, energy pricing issues, price distortions, and building an effective energy conservation plan are critical for Iran. From our research, we have reached the following main conclusions. First, short- and long-run energy price elasticities in Iran are less than unity. Second, in order to perform a considerable energy conservation plan, a complicated energy price adjustment is needed. Third, energy price distortions (subsidization) removal will cause an inflationary effect resulting in welfare decreasing. Last, non-price energy policies in the energy sector (such as autonomous energy efficiency improvement) could provide an appropriate background for effective energy pricing policies.
NOTES Y. Saboohi, An Evaluation of the Impact of Reducing Energy Subsidies on Living Expenses of Households, Energy Policy, vol. 29, no. 3 (2001), pp. 24552. R. Steenblik, A Note on the Concept of Subsidy, Energy Policy, vol. 23, no. 6 (1995), pp. 48384.
2 1

52

THE JOURNAL OF ENERGY AND DEVELOPMENT

3 R. Fouquet, The Impact of VAT Introduction on U.K. Residential Energy Demand: An Investigation Using the Cointegration Approach, Energy Economics, vol. 17, no. 3 (1995), pp. 237 47, and M. H. Pesaran and S. Yongcheol, An Autoregressive Distributed Lag Modeling Approach to Cointegration Analysis, in Econometrics and Economic Theory in the 20th Century: The Ragnar Frisch Centennial Symposium, ed. S. Strom (Cambridge: Cambridge University Press, 1999), chapter 11.

S. Johansen and K. Juselius, Maximum Likelihood Estimation and Inference on Cointegration With Applications to the Demand for Money, Oxford Bulletin of Economics and Statistics, vol. 52, no. 2 (1990), pp. 169210. R. F. Engle and C. W. J. Granger, Cointegration and Error-Correction: Representation, Estimation and Testing, Econometrica, vol. 55, no. 2 (1987), pp. 25176.
6 The Energy Input-Output Table for Iran in 2006 (Tehran, Iran: Ministry of Energy, 2008) (in Persian). 5