Read on Twitter
Many have asked me about the effects of altruism in economic-epidemic models (in about every talk I give!). I've thought hard about this for a number of years and there are several things to consider. 1/8
#EconTwitter #EpiTwitter https://twitter.com/MiettinenTopi/status/1335269551857152000
#EconTwitter #EpiTwitter https://twitter.com/MiettinenTopi/status/1335269551857152000
Turns out that there are three issues that are important: (i) how you model such preferences, (ii) the nature of the externalities that people's decisions have on others and (iii) people's ability to coordinate/impact aggregate population dynamics. 2/8
(i) Standard approach is to assume that agent i maxes ui(xi,xj)+A uj(xi,xj) with (xj,xj) an allocation and A>0 the altruism parameter; nests complete selfishness and social preferences as special cases. But this is not only approach. See Section 7 in https://drive.google.com/file/d/1_hB-SzbpRQfHNETePKTsl5mTus43AfXo/view 3/8
(ii) When only infection externalities present, equilibrium moves from usual Nash outcome to first best as A increases. When competing externalities added, altruism can lead to more infection. I explain details at 11.45 in this talk 4/8
Thing about altruism is that it's bi-directional. While i cares about j, j also cares about i. This can have surprising consequences as case of self-abnegating preferences shows in paper linked above. Such strong altruism can have i act as if he is j and j as if she is i. 5/8
In other words, it's as if they swap preferences and thus equilibrium essentially that of Nash equilibrium with selfish preferences, but the altruists are selfish on the other's behalf! 6/8
(iii) All this is for settings with few people. When people are small in population, another issue arises. Even if everyone is altruistic, equilibrium is still not socially optimal. Reason is that small individuals are "price takers" and cannot influence aggregate dynamics. 7/8
Even if you care about others as much as utilitarian planner, effects of your decisions cannot influence aggregates, while planner can. So individuals need ability to coordinate actions in addition to altruism. This is our decomposition theorem in https://drive.google.com/file/d/1jGi8g4Ls93gxTQWIpSv9yF-gbu2GFlY4/view 8/8
