Costly Electricity May Still Be Cheap

This page in:

ImageThe deep power crisis that Bangladesh is currently living through is affecting more people than the 40 percent population who currently have access to electricity. The reasons are simple.

For industries power outages increase production costs and the operating uncertainty that enterprises face. Losses arise from spoilage of goods-in-process and damage to machinery. Often the cuts in power supply cause production losses lasting beyond the duration of the outage. E-commerce and ICT cannot operate without reliable supplies of electricity. Mechanization of businesses is rendered ineffective, affecting productivity. SMEs rely on electricity for a variety of needs—lighting, refrigeration, grain mills, water pumping, food preservation and you name it. More generally, economic growth that creates jobs and enhances incomes requires electricity. Less and unreliable electricity translates into less and unreliable jobs. This is now a well established fact.

The million dollar question is what do we do to energize the economy?

The government has opted for quick rental stations using liquid fuels and capable of producing electricity within 6-12 months as a short-term response until longer term remedies are built. The desirability of this option has been questioned. The wholesale purchase price reportedly range between Tk 8 to Tk 14 per kWh, compared with slightly over Tk 3 per kWh cost of producing power from the conventional plants that take 4-5 years to build.

Keep in mind that there is hardly any alternative to rental power if we want to begin to alleviate the power deficit in the immediate future. The question essentially boils down to the opportunity cost per kWh of electricity not supplied. If this exceeds the purchase price of rental power then the latter is still worth it, even though more expensive than the medium to longer term option of producing power using conventional plants.

How much is this opportunity cost? Unfortunately, there are no studies investigating this question in Bangladesh. To get some sense of the magnitudes involved, our only option is to look at international experience. The problem here is that most of these studies pertain to either industrialized countries or countries that are approaching high level of industrialization. Often, the lack of relevant data makes such studies impossible in poor countries where the incidence of electricity outages is most acute.

Fortunately, there is one study by Roop Jyoti, Aygul Ozbaffi and Glen Jenkins in 2006 which was made possible due to the availability of a rich source of industrial information on power outages that affected the production at a spinning mill, a steel re-rolling mill, and an oxygen factory in Nepal. This data which covers a period of 5 years in the 1990s is accompanied by sets of detailed cost and operating data for each of these enterprises for the same five years that the power outage data is available. The authors were able to measure the direct impact of electricity outages on the level of profits of the enterprises through the effect such outages have on the contribution to profits that is foregone by the loss in production and increased costs.

This study finds that the loss per kWh not supplied ranges from $0.11/kWh to $0.33 kWh with an average of $0.23/kWh for Jyoti Spinning Mills; from US$ 0.13 to US$ 0.32, with an average of US$ 0.24/kWh for Oxygen; and from US$ 0.47 to 1.28/kWh with a average cost of US$ 0.98/kWh for Himal Iron and Steel. The authors also calculate the opportunity cost of power failures (unplanned outages) and load shedding (planned outages) and find that for these three enterprises the values are very similar.

The upshot is no matter how we look at it the opportunity costs are high. The lowest number on opportunity costs we find above is $0.11, which is equivalent to Bangladesh taka (BDT) 7.6 per kWh, and the highest number is $1.28 which is equivalent to BDT 88.3 per kWh.

How relevant are these to Bangladesh?

One way to judge is to figure whether, in the current state of the Bangladesh economy vis-à-vis Nepal, the opportunity costs of not meeting the electricity shortage are likely to be higher or lower than the ones reported above for Nepal. Bangladesh’s per capita income is higher than Nepal as is the size of the economy which is nearly seven times bigger. The share of industry in GDP is 1.7 times higher and the share of agriculture is nearly half lower while agriculture in Bangladesh is more mechanized and the service sector more energy intensive. All these suggest that the opportunity cost of not having electricity is likely to be higher in Bangladesh relative to Nepal.

There are some ball park estimates. Professor Ijaz Hossain of the Bangladesh University of Engineering and Technology has made a simple back of the envelope calculation by dividing the total energy shortage by the energy intensity per dollar of GDP. This yields a GDP loss of $13.3 billion in FY10, which translates into Tk 70 loss per kWh of electricity not produced. A study by the Bangladesh Institute of Development Studies reports that the elasticity of GDP with respect to electricity is 0.55. If we assume the shortage is about 25 percent of demand on average, this also gives a GDP loss of around $13 to 14 billion annually and a loss of Tk 71.6 per kWh not produced. These may appear to be on the high side considering the Tk 32 loss per unit of electricity shortage estimate that some trade bodies recently came up with. Yet, relative to the cost of obtaining power from quick rentals, the conclusion that the opportunity cost of not having power is very high is inescapable.

Bottom line: Whether a commodity or service is costly or cheap depends on the available alternatives. Rental power is costly relative to what can be made available in the medium to long term from large power plants; but it is cheap relative to what the economy loses by not having power now and in the immediate future.


Join the Conversation

The content of this field is kept private and will not be shown publicly
Remaining characters: 1000