© PhotoXpress.com

Mitigation costs to reach ambitious stabilization targets are estimated below 5% of GDP (IPCC, 2007). Most models used represent a first based world (optimizing agents with perfect foresight, market clearing equilibrium in all markets, and full utilization of production factors). With such assumptions, the optimal policy is a carbon tax, or a cap and trade system to equalize marginal costs. Any deviation from this is supposed to lead to additional costs and to veer from the optimum. By nature, synergies between development and climate policies and no-regret potential are excluded. The world is on the contrary, full of imperfections that could lead to higher mitigation costs however, they can also provide opportunities for implementing synergies between climate and development. In this article, we analyze the potential for implementing these synergies in a case study focusing on the power sector sub-optimalities in India. To do this, we rely on the modeling framework, Imaclim-R, which is well suited to represent a second best world and to embark sub-optimalities, disequilibrium, and inertia in technical systems.

Institutional and Market Failures of the Indian Power Sector

The Indian power sector is characterised by many institutional and market failures. The most critical symptom of these failures is a restrained access to energy services for both households and productive sectors: while electrification covers only 60% of Indian households (the remaining 40% that are not connected to the grid rely mainly on traditional biomass or on diesel generators), capacity shortage in 2007 reached 9.6% of total demand from the grid reinforcing the use of diesel generators by households and industries.

These power cuts and capacity shortage are caused by structural underinvestment in the power sector. The opening of the sector to the private sector in 1991, failed in absorbing the shortage and in compensating for the constraints on public funding. Less than half of the additional power capacity that had been programmed in the 10th Plan has been built. This underinvestment is largely due to the critically low profitability of investments in this sector induced by a very expensive cross subsidy system.

In 2006, the average level of administrated price covered only 77% of the average production cost: tariffs for households and farmers covered respectively 56% and 12% of the generation costs while industries and the commercial sector were partly compensated because they paid respectively 108% and 122% of production costs. Between 1993 and 2000, subsidies to households and to agriculture more than tripled and reached respectively 0.4 and 1.1% of GDP. In 2000, agriculture uses represented 1/3 of electricity sales in volume but less than 5% of total revenues.

These subsidies are justified by positive externalities on development (access to cheap energy for irrigation to promote food production). Nevertheless, they have significant side effects:

• the combination of critically low prices and of frequent but unpredictable power cuts create a strong incentive to a continuous use of electric pumps for irrigation and induces over consumption (estimated at 30% at least of consumption in agriculture), which increases the magnitude of capacity shortage;
• low revenues from electricity sales induce maintenance under financing and increasing inefficiencies in transmission and distribution, which represented at least 30% of the production in 2005.

Moreover, power sector inefficiencies constrain economic growth: electricity shortages hamper productivity and competitiveness. This reduces government tax revenues, and therefore capital availability to invest in additional power capacity.

Source: Annual report 2008-2009. Ministry of Power, India Click to enlarge

These sub-optimalities namely, power generation shortage, under-investment, tariffs not reflecting costs and T&D losses¹ are implemented within the Imaclim-R² hybrid recursive modelling framework. Imaclim-R has been developed to account for sub-optimalities and inertia in technical systems. Underinvestment in the power sector is represented by a gap between the estimated need of investments and realised investments in the power sector. The induced power generation shortage is represented by over utilisation of generation capacities, as Imaclim-R allows for endogenous disequilibrium due to inflexible characteristics of equipment vintages available at each period (putty-clay technologies). In case of underinvestment, in the short term, the main available flexibility lies in the rate of utilization of capacities, which may induce shortage of production factors, entail extra generation costs³, and raise the electricity usage cost.
Whether these sub-optimalities will persist or not is a determining factor for defining climate mitigation strategies.

Deadlock of a Carbon Price

Even if the demand decrease in the power sector for a 450ppm CO2 stabilization objective is large enough to absorb the power shortage and to remove over costs of generation due to over-utilization of capacities, persisting energy inefficiency in power production, transmission, and distribution coupled with over-consumption habits for farmers will introduce a substantial delay in the potential pace of decarbonization in the power sector⁴. A carbon pricing policy will then impact for a longer time the usage cost of electricity for households and industry.

As a result, a carbon price only scenario will induce prohibitive macroeconomic costs for India (-20% of the baseline GDP in 2030) even if GDP losses decrease after 2030 because of a faster technical change induced by the carbon price and of less vulnerability of the Indian economy to oil prices rising. The increased usage cost of electricity has a very important eviction effect for households on other goods consumptions. It induces a decreased competitiveness of Indian production due to the higher energy share in industrial production costs compared to other regions.

Financial transfers from a cap and trade system could reduce these losses. Indeed, with a contraction and convergence rule, financial transfers amount to 1.5% of GDP in 2020, and more than 2% of GDP between 2030 and 2040. On the long term, GDP is 15% higher than in the baseline due to the income effect of financial transfers and to the induced technical change. But on the short term, transition costs persist and India still loses 6% of its GDP in 2015 compared to the baseline. Organizing higher financial transfers to alleviate transition costs have little chance to be accepted by industrialized countries.

Looking for Synergies Between Climate Policies and Development

An alternative approach is investigated to enlarge the spectrum of climate policies to domestic policies targeting sub-optimalities in the power sector:

• the progressive reduction of subsidies to electricity consumption for the agricultural sector, accompanied by demand side management (improvement of irrigation pumps, evolution of consumption habits⁵);
• the implementation of a program to reduce electricity T&D losses to 15% over a 20-year period;
• the revenues following the partial removal of electricity consumption subsidies in the agricultural sector are used to finance all of these components. Remaining additional revenues are rebated to households through lump-sum transfers.

These measures are implemented within the same contraction and convergence allocation and result in GDP losses limited to 2% of the reference scenario during the first three years after the implementation of policies. The main reason for this is the decrease in the electricity usage cost induced by the reduction of T&D (transmission & distribution) losses and by demand side management. These measures lead to a reduction of energy budget share for households and less crowding-out of other goods consumption; in industry these measures lead to a substitution from fossil fuels to electricity and a reduced share of the energy budget.

Therefore Indian products lose relatively less competitiveness than in other simulations, and salaries are less squeezed to restore commercial balance: wage levels are 9% higher in 2020 and 2030. A positive income effect stems from higher wage levels. More wages increase due to more overall activity, and transfers to households coming from the remaining revenues generated by subsidies decrease. They are not used to finance the policies. Climate policies focusing on the alleviation of sub-optimalities constitute a win-win solution for climate and development.

Conclusion

From a methodological point of view, modeling frameworks are able to represent second best world characteristics in order to embark specificities of each national economy and to exhibit no regret potential. Possible synergies between climate policies and development policies should be used for mitigation cost assessment.

From a political point of view, the future international climate agreement should support the implementation of specific policies such as no-regret potential and energy efficiency measures to account for the heterogeneity of the real world instead of only dictating the implementation of a cap and trade system.

1. By default, tariffs not reflecting costs and T&D losses are expected to persist until 2050 [↩]
2. For a detailed description of Imaclim-R and of the implementation of sub-optimalities within Imaclim -R, please refer to Sassi et al. (2007) and to Mathy and Guivarch – under review [↩]
3. Mean generation costs increase when capacity is overused due to the existence of static decreasing returns due to higher labour costs and because less efficient units are switched on at the last aggregate level. By default, in our model, the increasing factor is attached to wages. [↩]
4. In India, 69% of electricity is produced with coal. [↩]
5. If the service is improved, the number of unpredictable power cuts will be lower, and farmers will not have to leave their pumps switched on all day long. [↩]