Discussion Section: Understanding Incentives, Part 1, featuring Steven Levitt

[Music] so one of the most important things to any economist is understanding incentives okay and I understood that early and in our rooming house at Harvard it turned out that each person when they assigned rooms got to pull one number out of the earth okay and they used a rule that the highest number of the entire rooming group was the order in which you picked okay so it's obvious when you know that if you want to have the biggest possible group right so even though they're really only for us that wanted to live together I convinced my roommates to group up with like 18 guys okay so of course 18 we had 18 draws out of a hundred and twenty numbers we were maybe like the second at the Rivermen group so we had 18 people so we got to pick five rooms but we got to pick five rooms all right at the top so that's all fine that's a good use of economic sector that's gaming the system the problem was once you have 18 guys and you got to divide up into brooms it's a really thorny problem because different people want to live with different folks and we didn't have a v-0 Chris couldn't solve the sorting problem of how to put 18 guys into four different rooms so I said look this is obvious I'm gonna build a computer program and each guy's gonna get a hundred points to assign he can assign those across any of the other 17 roommates he wants and and in addition to that there's going to be for two people you can put like negative twenty that won't come out of your hundred okay and what I'll do is I'll then have all the ratings and I'll use a computer program that that maximizes the joint utility subject to the constraint that we're gonna have 18 guys spread across you know four rooms three or four and and I'm one of six and I said everyone's worried about about you know it being secrets that are gonna destroy the data at the end so no one will see it'll run it'll say the answer destroyed okay so I've ran this program and it actually worked everything's a bit richer your works before you throw away the data okay so it actually worked I can't believe I actually got the 18 guys trusted me enough to go in and type on the data and I looked at the answers and I knew instantaneously I had made a terrible mistake because basically what it did is it took the for guys that nobody likes who actually hated each other worse than anyone else did and everyone kind of knew at them and it put those four guys in the one place and it puts the forward the six guys who really everybody liked they got put together and then there was the other mix and as soon as I saw it I knew exactly what had happened and I knew it was a terrible solution and I knew that as soon as people saw it everyone's feelings are gonna be hurt so I simply turned off the computer and said sorry guys that it crashed I don't know what happened I was outside the data but but was actually pretty smart one of the smarter things I did because the mini non-economists would have fallen prey to the Sun cost fallacy which is I'd spent days program another thing everyone typed in the data and even though knowing the answer was a disaster I would have stuck with it it was really interesting because I had that immediate insight which is damn I thought through everything I thought but there's always a case when you start messing with public policy or prediction where as soon as it happy like oh god there's that one thing I didn't think about which is which is obvious right it's to the fact if everybody hates four guys you got to put them together cuz it's the least costly way so in my life I was analytical but I'm not saying it wasn't analytical in class I mean I was a fabulous so if you had to redo that exercise what would you do you know it's hard because you really somehow you got to give may be some minimum utility of each guy or something like that but one thing yeah you eventually learn as you get learn about economics I've entered interpersonally utility comparisons are really a really thorny way to go and these sorting problems I don't know what would you do you know I think you do have to come up with some you know well you know some pricing system I think you know and one of the problems is obviously do you really want to allow explicit prices where you actually allow people to bring money into the equation but you know if you don't anyway I think you started in the right place thinking about well I'm going to give people a budget I'm gonna let them spend their budget and although I'm not so convinced this solution you came up with a bunch of terrible no it was efficient exactly it was efficient but when we go I know when you want to talk later about what economists have to learn from the real world and I think this was an example for me where I could tell already from having listened to the discussions that efficiency alone wasn't going to solve this problem because these guys you know there was an incentive compatibility constraint here these guys absolutely these four guys would kill each other if they live together and we all knew they couldn't live yeah I think the other side of it though is you have this inherently linear model there it was you know they didn't you didn't take account of the other side of the economics which is if I live with four people and there's two guys I like and one guy I don't like well I can I don't have to deal with everybody right and and you know the poor guys in that one room had nobody that they could deal with right and so no matter how they allocated their time they weren't going to have a good time yeah and I think that's something you would have to build into your algorithm to recognize there are slots in every room and maybe having three other bad guys to live with is a disaster much worse than having one guy you don't want to live with in each room yeah really good fun you know and let alone the feedback effects that happen between if I don't like you and you don't like me that could be a total disaster it's probably efficient in a very narrow sense but probably not in that broader sense of taking account of the kind of endogenous allocation of time we would say in economics the fact that I don't have to deal with everybody equally and having some good options can be better can be valuable in and of itself I you know I think that's a good so how did you guys allocate them then we had an all-night session of going back and forth and why don't you point captains and do like we used to pick based on teachers whatever you know your four captains and room captains you know take the guys who scores the highest on your list make them room captains and say okay those four room captains are going to pick I might have gone with the captain approach you know yeah I especially I think is I think what's always tough about the captain approach is the guys the last two guys are standing there last one we pick feel bad but in this setting you could actually do it all right exactly captain the three captains in the room were the four captains room nobody else yeah but even figure out who the four captains would be is he's a he's that tough problem well we made it we made it work I think in the end what happened was some of the more popular people just said hey look this is who we're living with you guys got a deal we got a deal with it there's a certain amount of power for people I don't know if that point if week really I think we couldn't really walk away I mean I think we had already committed to being in this group so I'm not sure what I'm gonna true what the default was if we simply couldn't agree but I guess I'm already seeing Steve Levitt kind of in these early days right I mean I'm you're an economist you think like an economist but on the other hand this you don't see economics is kind of living inside this classroom or living inside this box is and you know so I that's why I challenge the notion that you weren't analytical and you mentioned analytical in the sense of doing options and things like that but there's another dimension which is were you analytical in terms of how you organize the world around you that is thought about well cheese how is that why is the world looked away it does why do I see can I understand how things work [Music]

Channel: Becker Friedman Institute at UChicago - BFI

Views: 2,779

Rating: 5 out of 5

Keywords:

Id: z03O2-4mXLc

Length: 8min 24sec (504 seconds)

Published: Thu May 31 2018