Thinking at the margin

This page in:

In a previous post, I had talked about problems due to power outages faced by retailers in India using Enterprise Survey’s data on 1,948 retail stores. I provided evidence showing that losses due to power outages (as a % of the annual sales of stores) and hours of power outage in a typical month were much higher for low-income lagging states compared with more developed leading states in India.

However, I had cautioned that to properly understand the rate of return on investments to improve power supply across regions, we should look at which states (leading vs. lagging) benefit more from an hour’s reduction in power outages. It is possible that this marginal benefit could be higher in the leading states even though they face much fewer power outages.

The figure below shows that losses as a % of annual sales to retail stores per hour of power outage are much higher in the initial hours of power outage. In other words, as stores face more and more hours of power outage, the additional negative impact becomes smaller and smaller.

Now add two more facts to the figure. First, stores in leading states face fewer hours of power outages than stores in the lagging states (47.6 vs. 126.3 hours per month). So this implies losses of about 0.245% per hour of power outages to stores in the leading states compared with 0.203% in the lagging states. Second, annual sales of stores in the leading states are about 1.6 times that in the lagging states. Putting all this together, the absolute benefit (in USD) from an hour’s reduction in power outage in the leading states is about 1.6 times the same in the lagging states.

In short, assuming that the cost of increasing power supply is the same across leading and lagging states, the leading states offer a better return to improvement in power supply than the lagging states. This finding must be balanced against concerns about regional equality, but this is separate from the issue of the efficiency of an investment. More broadly, we should be careful in highlighting the worst performing states or countries as in dire need of reforms. A proper impact analysis or marginal cost-benefit calculations are important.


Source: Enterprise Surveys. The figure is a partial scatter plot controlling for store-type fixed effects (traditional store, consumer durable store and modern format store), city fixed effects and floor area of the shop. The strong negative relationship shown does not depend on the stated controls.


Mohammad Amin

Private Sector Development Specialist

Join the Conversation

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