Solar power for street vendors

Although energy poverty is usually considered a rural phenomenon, rapid urban growth in developing countries is aggravating the problem of electricity shortages in urban areas

With rapid urbanization across the developing world, the phenomenon of urban energy poverty is increasingly significant. The number of people living in urban areas of developing countries is growing rapidly, and urban planners face major difficulties in providing growing urban populations with basic energy access, such as electricity for lighting. Much of the urban growth is concentrated in slums with poor infrastructure and most urban dwellers in developing countries are employed in the informal economy. In India for example, as of 2013, over 65 million people reside in slums (Census of India), and as many as 400 million Indians work in the informal sector. In addition to the low quality of housing and basic facilities among the urban poor, the lack of property rights raises barriers to better energy access by preventing citizens from registering and investing in improvements to their properties and businesses.

We are currently conducting a randomized controlled trial on using solar power to alleviate energy poverty among urban street vendors. For the intervention, we partnered with an Indian NGO, Nidan, to set up centralized charging stations in urban marketplaces of Patna, the capital of Bihar. The technology was provided by a local company, PowerGreen Renewables. Through a centralized charging station, local vendors can charge the batteries of their portable lights using solar power. The station is operated by a local entrepreneur. The goal of the intervention was to provide access to 500 vendors, with another 500 vendors in a control group. Each centralized charging station was to serve a cluster of about 25 vendors.

The project began with a baseline of a random sample of 1,000 vendors in 24 marketplaces in Patna and the satellite town of Hajipur. The baseline confirmed that the study population is deprived and provided a motivation for the intervention. Only 64% of the vendors were able to read Hindi and 67% reported being vendors simply because they had no other options. The average daily lighting expenditure was above 10 Indian rupees per day (USD~0.167) and 80% of the vendors considered improved lighting a top priority for enhancing their business. In other words, the vendors paid a high daily price for an inadequate lighting solution.

Problems with centralised charging stations

After the baseline was finished, we began setting up centralized charging stations in a randomly chosen group of marketplaces. Over a period of six months, we faced many challenges with the centralized charging station model. Of the 10 marketplaces that we approached, we were only able to set up and successfully operate a station in three. Even in these three marketplaces, the daily cost of solar lighting to the vendors was 15 Indian rupees. While the quality of lighting improved, the cost did not decrease for most vendors. About 25% of the vendors in the areas where the centralized charging stations were set up chose to use solar power. The reason for the unexpectedly high cost was the need to cover not only the technology cost, but also to finance the payments collection and the daily operating cost to the local entrepreneur.

In seven marketplaces, we were unable to set up a station. In one of them, a station was initially set up but the local entrepreneur responsible for it neglected his responsibilities and the station had to be dismantled. In six marketplaces, we were unable to set up a station. There were four obstacles to setting up the stations:

• Local strongmen, who operated diesel generators in the marketplaces, were opposed to the use of solar power. They were afraid of competition. Even though we tried to negotiate with the strongmen, they refused to replace their diesel generators with solar power. Since solar power would have offered only an incremental improvement over diesel and the strongmen had already bought a diesel generator and invested in wiring, the strongmen chose not to consider solar power.
• Finding local entrepreneurs proved to be difficult. The street vendors that Nidan approached considered the operation of the station a distraction from their business, requesting high daily payments. This increased the cost of operating the station significantly, deterring vendors from subscribing to the service.
• Battery charging created a moral hazard problem. Since the vendors would rent the batteries and lights, they had limited incentives to maintain them properly. Misuse caused losses that were not covered by the warranty.
• Physical barriers, such as the lack of safe sites for installation, also caused delays in installations.

If the charging stations had been unambiguously and significantly superior to diesel generators, we believe these problems could have been overcome. However, the incremental benefits of solar power were not sufficiently high to offset the high transaction costs of operating the centralized charging stations. As a result, we decided to stop the intervention.

Since the problems with the centralized charging station model were associated with the mode of operation, as opposed to solar power itself, we are now experimenting with another intervention. We now sell standalone solar lights with individual batteries and solar panels to a randomly sampled treatment group of vendors. This revised intervention is at the early stages, but so far we have managed to avoid the aforementioned problems. The standalone devices are not considered a major threat to business by local strongmen, no local entrepreneur is required, moral hazard disappears because vendors by the devices for ownership, and there is no need to locate sites for large solar panels. We are also overcoming liquidity constraints by allowing vendors to pay for the devices in installments.

Electric grid connections

The general lessons from this intervention are threefold. First, although seemingly appealing, the centralized charging station model carries hidden transaction costs that have not been identified in previous research. Second, there is a world of difference between energy poverty interventions in rural and urban contexts. In our experience, the urban context features a more complex set of informal rules, power relations, and vested interests that have to be considered in the design of interventions. Finally, municipal corporations in Patna and elsewhere should give serious consideration to the possibility of providing street vendors with regular electric grid connections. Since the electricity grid already exists in the marketplaces we studied, the marginal cost of extending electricity access is lower than in many rural areas. For such an intervention, the main challenge would be to register and formalize the currently informal street vendors so that they could pay a monthly fee for their basic electricity access.