Learning by riding and Learning by seeing - why subsidies for public use vehicles goes farther

Our paper titled "Benefits of electrifying app-taxi fleet – A simulation on trip data from New Delhi" was just published in the journal Transportation Research Part D: Transport and Environment. New Delhi, the capital of India is usually in the news for a few weeks this time of the year because this is time the air quality tends to be the worst and so in a way it seems timely that this paper just came out.

A salient aspect of this study is it is based on actual trip data (thanks to Ola Mobility Institute) and one of the first such for an Indian city if I am correct. Using a dataset comprising ∼ 730,000 app-taxi trips spanning ∼ 15 million kilometers (km) in Delhi, India, we estimate that: i) ∼ 23000 BEVs with 200 km range and a network of 3000 50 kilo-Watt chargers could satisfy 100% of Delhi’s daily app-taxi demand, and relative to a compressed natural gas (diesel) fleet eliminate 180 (700) and 0.14 (70) metric tonnes of vehicle tail-pipe NOX and particulate matter emissions respectively annually and reduce lifecycle greenhouse gas emissions by 15% (27%) per km. We show that the impact on the electricity demand of a city is small and charging demand is not coincident with the times of the day when electricity demand peaks. 

A broader policy lesson is that when we subsidize clean and socially beneficial technologies, directing those subsidies towards adoption in settings where the technology or investment will be used publicly will yield greater returns. Otherwise it even risks being concentrated among the wealthy who may not use it much and in fact even prove counter-productive for mass adoption because it may be seen as a rich person's toy. This is because more people will be able to see it in operation,  experience it and learn from it. Therefore, a bus or even a taxi which is on the road a lot more relative to a private use vehicle and hence seen and experienced by more people might come to be seen as not just cool and clean but reliable and functional technology and thus help drive private adoption voluntarily with less subsidies. This is what I see as learning by riding and learning by seeing in contrast to learning by doing which we talk of on the manufacturing side, the idea that as firms make more of something they get better at it and are able to make something for less cost. In the case of taxis, the focus of this paper, the drivers are also poorer and hence helping them adopt is more equitable.