Supply preferences and productive electricity use constraints among micro and small enterprises in Bihar, India
40% of Bihar micro-firms stay unconnected despite good grid supply
Read the published paper ↗doi.org/10.1016/j.esd.2022.01.004
Whether supply is what constrains small firms
Evidence on how firms use electricity and what supply they prefer comes mostly from larger manufacturing and agro-processing firms. Much less is known about micro and small enterprises (MSEs), which provide livelihoods for many in both towns and villages. Bihar is a useful setting: its electricity supply was historically poor and the sector has recently been reformed.
We ask what electricity supply attributes MSEs prefer, and what constrains their productive electricity use. The distinction matters for policy design. If reliability and cost are the constraint, supply-side investment is the remedy. If the constraint lies in what firms can do with electricity once they have it, supply improvements alone will not be enough. Answering this requires measuring both the supply firms currently receive and the trade-offs they would make between competing supply offers.
A conjoint experiment with 696 enterprise owners
We surveyed 696 MSE owners across Bihar using stratified random sampling. We select districts, and then villages and towns within them, with probability weighted by the number of non-farm employed adults, keeping the urban and rural shares of each district. Enumerators then selected every Nth enterprise outward from the centre of each enumeration area. The COVID-19 pandemic interrupted data collection in March 2020, so the realised sample is not technically representative of all of Bihar.
A conjoint experiment asks respondents to choose between two hypothetical offers whose features are randomly assigned, which reveals how much each feature matters. Each respondent made this pre-registered choice twice, yielding 2,784 observations. Offers varied in supply hours (6, 12, 18 or 24 per day), blackouts (0, 1, 5, 10 or 15 per month) and voltage-fluctuation days (the same range). They also varied in capacity (lights and phone charging only, or any appliance), source (grid or microgrid) and cost (1–30 rupees per kWh). Standard errors are clustered at the enterprise level. We use Bayesian Additive Regression Trees, a flexible machine-learning estimator, to explore how effects differ across firm characteristics.
Supply is adequate; capabilities appear to constrain use
We find that reported grid supply is quite reliable. Grid-connected firms report 18.7 hours of electricity per day on average, with 1.3 blackout days and 0.6 voltage-fluctuation days per month. Yet only 60% of sampled enterprises have a grid connection, 23% use off-grid sources alone, and 17% use no electricity at all. Among firms without a grid connection, 53% cite poor affordability, 35% no need and 17% an inability to connect. Three-quarters of all firms consume less than 1 kWh per day.
In the conjoint experiment, firms respond modestly to supply attributes in the expected direction. A drop in availability from 18 to 12 hours reduces the predicted likelihood of accepting an offer by 7.4%-points [95% CI: 2.0 to 12.7], while a rise from 18 to 24 hours has no significant effect. An increase in blackout days from 1 to 5 reduces it by 9.5%-points [3.7 to 15.3]; the same change in voltage-fluctuation days has a small, statistically insignificant effect. Doubling the price from 5 to 10 rupees per kWh reduces acceptance by 9.3%-points [2.6 to 15.0]. By far the largest effect comes from restricting use to lighting and phone charging, at 28.3%-points [24.0 to 32.8]. Firms show no preference for the grid over a microgrid.
Look beyond supply towards use stimulation
These results point away from a supply-side diagnosis. Firms care more about whether the supply can run the appliances they need than about an extra hour or one blackout fewer. Their consumption stays low even where reliability is already quite good. Asked about business problems, both grid-connected and other electrified firms rated timely payment by customers, access to credit and corruption as the most pressing, ahead of electricity.
Overall, electricity reliability and costs are unlikely to be a serious business constraint for most MSEs in Bihar. A minority of larger firms may still respond to lower electricity costs, and price rises would hurt smaller firms most. The main challenge appears to be firms' limited capabilities to buy appliances and adapt their production processes. Policy reform in Bihar aiming to encourage productive electricity use should look beyond supply improvements towards stimulating productive use. Because firms accept microgrid supply, decentralised renewables can also serve many of them alongside the grid.
@article{pelz_supply_2022,
title = {Supply preferences and productive electricity use constraints among micro and small enterprises in {Bihar}, {India}},
volume = {67},
copyright = {All rights reserved},
issn = {09730826},
url = {https://linkinghub.elsevier.com/retrieve/pii/S0973082622000059},
doi = {10.1016/j.esd.2022.01.004},
language = {en},
urldate = {2023-01-28},
journal = {Energy for Sustainable Development},
author = {Pelz, Setu and Aklin, Michaël and Urpelainen, Johannes},
month = apr,
year = {2022},
pages = {151--162},
}