GDP growth during and after shutdown: Viewed through electricity generation

To contain the Covid-19 pandemic, many countries, including Bangladesh, adopted various degrees of shutdown and social distancing measures. Although vital in suppressing the disease, these measures have also precipitated an unprecedented economic crisis. According to a recent estimate by the World Bank, Bangladesh's Gross Domestic Product (GDP) is now expected to grow at about 1.6 percent in the current financial year. 
Covid-19 led to a drastic increase in food insecurity, which appears to be dispersing rapidly into groups that were formerly food-secure. Help from the government or NGOs were very negligible. The government has now relaxed the country-wide shutdown to augment the economic activities, though we are also witnessing an increase in coronavirus positive cases. 

While the lives versus livelihood debate continues, we take a look at the economic activities following the withdrawal of the shutdown. 

Measuring economic activity for such a short period is not easy. There is also a lack of reliable and real-time data to measure them directly, not least for an economy with a large informal sector. We provide an alternative approach to understand GDP estimates based on electricity generation and consumption, especially when there is a lack of real-time data. 
We apply this approach to gain an understanding of the economic activities following the relaxation of the shutdown. We show this in the current context, but it can be applied for any period. We use a method proposed by Kaufmann and Kaliberda (1996) in a World Bank Research Paper. Our approach can be seen as similar to the method of using night-time lights satellite imagery to derive a measure of GDP.

The idea behind our proposed approach is to identify a simple but important factor of production that can be a reasonable proxy for the level of GDP. We consider electricity generation (or consumption) one such factor. As it is difficult to obtain the real-time data of all traditional factors of production (e.g., capital, labour, etc.) which determine the GDP, we use the readily available electricity data provided by the Power Development Board. 
First, we test the validity of our proposed method in the context of the Bangladesh economy. To do so, we used previous data (FY 2007-08 to 2018-19) on GDP and electricity generation and estimate the electricity production for past years. We found a very high degree of correlation (0.99) between our estimates and actual production, which indicates electricity data can be used to predict economic activities and measure economic growth.
Now, to understand the recent trends in economic activities, we observed the average daily highest generation of electricity in the first six months during 2017-2020 (up to June 15 for the current year). The year-to-year generation of electricity shows a predictable pattern, which rises in summer. 

However, we observed a sharp decline in electricity generation from March to April, the first month since the shutdown. The electricity generation started to rise slowly in May when the government eases the restrictions for various sectors. But, the negative gap in the generation between 2019 and 2020 remain stable until the end of the shutdown. 
Even after the shutdown, if one considers the growth in electricity generation in the previous years, one would expect a higher generation in the current year compared to the previous year, which is not the case. 

Thus, we still observe a decline in economic activity compared to the previous year despite allowing economic activity to function. 

Let's move on to the daily figures of total electricity generation throughout the pre-, during- and post-shutdown periods. We have compared these electricity data with the preceding two years, 2019 and 2018. 

As expected, during shutdown the total electricity generation (including import) reduced by 29.5 million kWh (hereinafter MkWh) per day in April and 43.4 MkWh per day in May 2020 in comparison to 2019. On the contrary, during the same period, electricity generation had increased by 27.63 MkWh and 46.09 MkWh in 2019 compared to 2018. 

This downward-trend is continuing even after the ease of restrictions, a negative gap (-4.92 MkWh) is evident in the first 16 days since the shutdown ended on May 31 2020. Whereas, a positive gap of 28.61 MkWh was evident between 2019 and 2018. 

These figures indicate, even after easing the restriction, economic activities, based on the electricity generation and consumption, have not picked up significantly. 

If we examine the demand for electricity during the same period, a similar trend is evident.
Now, we can use the 'loss' of electricity production to estimate the economic loss associated with the Covid-19 pandemic. As per our estimate, during the general holiday for 66 days (March 26- May 30), daily economic loss was Tk26.56 billion in nominal term and Tk11.01 billion in real term. Built upon the earlier GDP estimates by the government, our calculation is based on the gap between current electricity generation and required electricity to achieve that estimate.
If we assume the economy is functioning as per plan beyond this period, then the economic growth is expected to be 1.63 percent in the current fiscal year 2019-2020. Thus, our estimate is quite close to the estimations provided by various other organisations, such as the World Bank and Economic Intelligence Unit.

If we calculate the GDP of the first 16 days after the shutdown, we still observe that a significant share of economic activity is missing. The production and demand for electricity are still lagging. Our estimates show that every day we are losing around Tk11.08 billion in terms of nominal GDP and if this trend continues, our real GDP growth will cut down further to 0.71 percent for this current fiscal year. 

These estimates show that we need to think more about the decision of opening up the economic activities as there are other costs due to Covid-19 positive cases in addition to the inevitable economic losses. Our hospital system is grossly underprepared and oversaturated with the Covid-19 patients at present. 

However, successful case studies exist around the world; many countries were able to contain the virus and manage the cases within about three months after an effective shutdown was imposed, including in many developing countries with similar GDP per capita as ours. We also need to think about longer-term implications- if we can't control the outbreak, the economy will not recover, and both exports and domestic demand will suffer. 
Analysing data of Bangladesh's GDP we have found that feeding 50 million of the poorest for the next six to twelve months is possible using our existing resources. Voluntary contributions from affluent households to their poor people are also commonly evident during the crisis, and we could mobilise more with contribution from the relatively well-off people. Our economic performance in the recent past should give us enough courage to make a bold step to provide generous transfers for the poor. Overall, we have the resources and capacities which we need to manage efficiently and fairly. We could improve our management in dealing with both economic and health crises. We can then come out stronger, both health and economy-wise.

Asad Islam is Professor, Economics Department, Monash University, Australia and Hashibul Hassan is Assistant Professor, Department of Finance, Jagannath University, Dhaka.