Skip to content

Cool Car, (Wo)Man!

Peer effects in EV adoption.

Obsessing over cool new gadgets, clothes, movies, music, and books has kept many a conversation going. In the classroom we still teach an economic model assuming rational decision making in a world of perfect information, yet the reality of how we learn about cool new stuff has a lot to do with social networks. You go to a birthday party and someone tells you about their shiny new Tesla! You come to work on a Monday morning and the person in the next cubicle over has a new iPhone. Your MBA students tell you that the new Taylor Swift album has dropped. The response is often “I want one too!” 


From an energy economist’s perspective, we have been thinking about how to get new and more efficient technologies into the hands of people. Meredith and Duncan have been thinking about heat pumps. Today I am, again, thinking about EVs. If we think that the future of the transportation sector is electric, then getting these cool machines into the hands (or under the butts of) people is key. One hypothesis has been that spillover effects are important. If Severin gets a Tesla 3 and brings it to work, Max sees it and wants one too, and is more likely to buy a Tesla than that Shelby GT500 he has been ogling. That makes intuitive sense. But as our readers know, we demand cold hard numbers. 

Figuring out the size of this effect is hard. If I observe the number of Teslas going up more in one neighborhood than another over time, I cannot figure out whether this is a peer effect or simply that one neighborhood is getting richer (or something else is changing in a different way). What we would like is some natural experiment that provides variation in when people choose to buy a car and the information about what is out there in terms of options.  Sebastian Tebbe, a recent visitor to our department, who is on the job market this year, wrote a really cool paper doing just that. 

Sebastian uses the fact that some drivers do not adopt a new car at random points in time, but rather at 36 month intervals – when their lease expires. And the actual date varies across different drivers. This is super clever in the first place. But what makes this paper blog-worthy is the second part. Sebastian is able to use data on Swedish households to figure out who is in your family, who lives in your neighborhood and who is in your work network (people employed by the same firm at the same location). “Yeah, Max, but there is no way he knows what people drive in those networks.” Sit down. Take a deep breath. He. Does. This is an absolutely bonkers amount of data work. He knows over time what cars enter Max’s family and work life – that are not his! He then combines this with the somewhat random lease renewals to check whether folks at the time of a lease renewal that have a larger share of EVs and hybrids in their networks are more likely to buy such a car themselves. Mind blown by empirical awesomeness.

So what do we learn from the paper (Yes, this is leases in Sweden, so hold you external validity hästar):

  1. The paper shows (to me) convincing evidence of significant peer effects in the adoption of electric vehicles in all three networks – family, neighborhood and workplace. How big? One more EV in the neighborhood leads to 0.114 more EVs adopted by neighbors! Good morning my neighbors! The corresponding number for workplaces is 0.077 and for families is 0.014.
  2. If you do the calculations per capita, the family effect is biggest. So if Auntie Megan buys a Tesla 3, it drives up the probability that grandpa is going to not get another Volvo V8 but pick up a Polestar instead. 
  3. The effects Sebastian estimates indicate persistent incremental demand. These effects are multiplicative over time, which means that these networks have a powerful multiplier effect. 
  4. The paper suggests that the shift towards EVs is strongest away from Diesel vehicles, which is great, as Diesels are nasty dirty engines from both a local and global pollutant point of view. 

Of course research on how to roll out new technologies using social networks is not new. There is some really cool experimental stuff done by folks like Kyle Emerick, Betty Sadoulet and Alain de Janvry in the context of new rice varieties in agriculture. The problem here is the same and the findings are not dissimilar. Networks are a powerful accelerator for the introduction of new technologies. I am now going to put down my computer and read some more papers on this stuff (much of which not surprisingly are written in fields other than economics).

Keep up with Energy Institute blog posts, research, and events on Twitter @energyathaas.

Suggested citation: Auffhammer, Maximilian, “Cool Car (Wo)Man!”, Energy Institute Blog, UC Berkeley, October 24, 2022,

Maximilian Auffhammer View All

Maximilian Auffhammer is the George Pardee Professor of International Sustainable Development at the University of California Berkeley. His fields of expertise are environmental and energy economics, with a specific focus on the impacts and regulation of climate change and air pollution.

%d bloggers like this: