Lecture 9: Reference-Dependent Preferences

Key Takeaways

okay so let me sort of just recap what we discussed last time fairly quickly and then we want to move to empirical evidence on a variety of settings we might not get through all of the slides that's fine in that case we'll just discuss some of this in recitation so let me sort of recap what in their 1979 article um were proposing based on a bunch of experiments and empirical evidence that they had collected so their theory is called prospect theory what they proposed versions of prospect theory is essentially versions of reference dependent utility have been used are used prominently now in economics so the first thing that they proposed was what matters what the carrier of utility is is changes rather than levels that's to say it doesn't matter for you necessarily how warm it is it matters kind of what's the change of temperature compared to yesterday it doesn't matter that much how much money people have in total it matters how much the changes relative to what they had previously more generally what matters for people is essentially a certain consumption or the like relative to reference point as opposed to an absolute terms second loss aversion this is losses loom larger than gains and we had some evidence of that in the last lecture that's to say if you lose some money or some consumption or anything else grades etc a lot relative to either your status quo or to your expectation looms larger is more important hurts more than a gain of the same magnitude in the positive direction and number three which you talked about quickly is what's referred to as diminishing sensitivity that is um people are risk averse in the gain domain but risk loving in the last domain that's to say um um if you are for example if you think about distance time chance and the like going from zero to one or from one to two from two to three is more important for you than going from like 100 to 101 101 102 and so on essentially uh the further you'll go away from like your status quo your reference point uh any marginal changes like diminishing right and that's true for both the gain domain and um uh the loss domain so these are sort of the main three characteristics related to what kind of first we call the value function we can think of this as essentially a version of the utility function there's a fourth characteristic which is probability weighting which we're not going to talk about at least for now are there any questions on these three things so far okay so now prospect theory is then what kahneman and versky were proposing they were essentially saying like instead of having concave utility which i showed you um uh last week instead your utility function may look like this and what are sort of the features of this utility function what what are the three key features that i just showed you how they're showing up in this function here yes uh c minus where is the changes right so the carrier of the utility function so what we have here is like c and r uh c is like consumption that could be anything that could be apples that could be bananas that could be uh sort of how much money people have available to consume overall the carrier of the utility function is not c itself it's c minus r so it's c relative to some reference point r that's exactly right so and if r is the status quo if r is how much do we have right now or the person has right now c minus r is a change relative to the status quo notice that r uh we're going to talk about this a little bit could be also other things it doesn't have to be necessarily the status quo it could be also people's expectation or their goals or their aspirations for the future right the key part is however what matters is the first thing i was saying like it's changes rather than levels changes relative to some reference point or consumption relative to some reference point that's number one yes yes uh the curve flattens out which is diminishing exactly the curve flattens out in both directions that's diminishing sensitivity it's essentially concave in the in the gain domain of the domain where like c is larger than r and it's convex in the last domain where c is smaller than r right and that's exactly the issue that essentially the first the marginal change going from like uh say one to two is relatively large compared to a marginal change going from 10 to 11 that's both going in the right direction in the gains domain and the left direction and the last domain uh yes um for the loss of version um the left side is cheaper than the right side exactly in particular there's like kink in and this function uh when you look at zero where essentially um the gain and loss domain was c equals r uh the gain and loss domain intersect uh there we have a kink in the in the value function and the utility function uh which essentially exactly is the last domain at the last aversion which is like going to the right it's less deep than going to the left but differently if you lose if you're sort of below the reference point that's more painful than being the same amount of uh units to be above the reference point that's number two that's essentially exactly the last aversion so i just wrote down all of that again let me repeat caro the utility has changes relative to the reference point excuse me rather than levels second there's loss of version there's a kink at zero in this function and three there's diminishing sensitivity which is concavity and gains and convexity in losses any questions on that now a key question here is i'm going to sort of flag this we're going to talk about this a little bit at the end of the lecture we're going to talk to this a little bit in the next problem set is kind of like how is the reference point determined and does it matter as i said before in kanaman firski's work a lot of the reference point is the status quo so they essentially postulated the status quo is what would really what matters originally i think people have moved towards saying like the reference point and this is what kossig and raven and others have written down in their models of reference dependence utility and and and recent more economics work is what really matters is expectations so what do you expect to consume or to have and so on that matters that sometimes coincides with the status quo for example if you have a house the status quo is that you have a house you probably assume that you have a house in the future uh these things coincide if you think about wages et cetera what seems to matter often is not so much how much what people's wages are right now what matters is like what wage gains and so on do they expect to receive in the future and they evaluate their future their their outcomes in the future uh not necessarily to like the current wage but rather what they expect they would get in the future okay and so here's an example and you'll have some problem sets questions again this is problem set three which is not posted yet but but will be uh of that kind which is you might have like reference dependent utility over uh this is just a very simple example over shirts and money so you have essentially two different domains you have essentially losses and gains over both of those domains you have a reference point which is rs is how many shirts you might have you have a reference point rm how much money you might have now what's important here is that essentially when you think about buying a shirt or or selling a shirt there's going to be two dimensions which you have to sort of consider if and when you think about the endowment effect of mugs and so on you have to consider not only the losses on gains and shirts but also the losses can gains in money so that's to say if you're trying to buy something you're going to get like a gain in shirts relative to the reference point if it's unexpected but you'll also have a loss in money similarly if you're going to like sell a shirt you're going to have like a loss in shirts but a gain and money on these things and interact now what's the value function the value function as i said before it's usually concave in the gain domain and convex in the in the last domain there's a kinkex zero it's deeper on the left uh then on the right usually we think uh the relative slope is about like two 2.5 okay so um so one version of that would be uh this function that i wrote down again there will be some problem such questions etc sort of to clarify uh that but one key question here is then uh what are the different domains and that's kind of like a question like of mental accounting we'll get to this like in the second half of the course the question kind of like what are the different categories do you have like shirts do you have like uh pants uh sweaters or is it just for clothes overall um or like when you think about like earnings and uh consumption et cetera is it daily consumption is it weekly consumption is it monthly consumption so there's lots of questions on like how to exactly specify this utility function these questions are mostly um unanswered in the literature so for now for us in our purposes we're going to essentially assume there is a value function given and then sort of work with with that any questions on that okay great so now we're going to talk about the number of different applications we talked a little bit about already about the endowment effect and about insurance i'm going to skip this um there's some of this already in in recitation and the problem sets we're going to particularly talk about labor supply and employment decisions essentially how much do people like to work in is effort or people's work decisions are they reference dependent uh we're gonna talk about finances what was mentioned last time already about investment decisions when do people sell and buy stocks we're going to talk about housing when do people decide to sell or buy a solar house and at what price uh we're gonna talk about sports in particular marathon running uh and golf um uh and a little bit about this some papers in domestic violence we're gonna talk very briefly about and then we talk a little bit about firms how do firms think about pricing um uh and so on and what what is sort of like the market response to to reference dependence that's to say given that we know there's lots of reference dependence on the world uh how should we now as a firm or like uh uh treating other people how should we think about um how does that affect our our own behavior or how maybe firms interact with us okay so let's start with um a labor supply let's start with a very simple um example suppose there's a worker in the following situation she can work freely choose how many hours she works every day and there are frequent temporary changes in her hourly wage now there's different relationships between wages and hours per day you could always work the same number of hours you could work more hours on the days when wages are high or you could work a few hours and days when the wages are high what are sort of standard theory say what should you do yes two two why is that um because that maximizes your expected value for time right so if if hours are effortful you usually don't like to work you should work during the hours when uh uh when your payment is the largest uh that's right so what about option number one why might option number one be optimal so what you're saying is exactly right uh depends a little bit on on something else and what is that like why might you choose number one anyway um habit farming is nice yeah there could be sort of habits exactly or it could just be that it's really effortful to work like 12 hours in one day maybe there's like kids at home or maybe there's like you know it's just really tedious to work at some point you just want to go home and do other stuff so it could just be that working beyond you know like say eight or nine hours per day is really tough to do so then you might say i'm always going to work eight hours i'd love to work more uh but that's difficult to do so essentially it's just to say if the effort function if the effort as a resp function of hours is convex and you might sort of say keep it the same surely what you don't want to do as number three working few hours and days and wages are high unless you know effort costs are particularly high on those days so it could be that it's like really super hot or it could be like it's super tedious to work on those days and you might say then you don't want to do it but assuming that's not the case you you kind of want to avoid this because for the same number of hours you're going to make less money um overall right and sort of here's a concrete example suppose wages five hours an hour on day one and ten hours a day on day two so there's three strategies you can work eight hours in both days you can work six hours on day one nine hours on day two or the opposite nine hours on day one and six hours on on day two so if you do that you can do you sort of calculate how much that is you can essentially work eight hours on both days you get 120 hours you can work six hours and nine uh uh uh uh on on days one and two which makes 120 hours at dollars as well or you can essentially do um the opposite which makes you 105 dollars this is what i was saying like options three doesn't make a lot of sense um unless effort costs are particularly high on certain days when the wages are high we're assuming that away for now um and so now option two and this is what you were saying essentially saves you an hour overall you work only 15 hours instead of 16 hours and you make the same amount of money now unless sort of the effort court unless the ninth hour is extremely costly for you to do you might not want to to do that okay now why might you do uh something else uh instead or what sort of where does reference dependence come in here yes i see on the high wage day that i've made more than the less waste day so i feel like stopping maybe and why am i why would you stop like what's what's what's causing you to stop uh [Music] yeah and so yes and but so what's the what are you evaluating or what's the or what happens for example if you work only six hours on on a different day like how much do you make on the six hour day i guess which would be uh maybe thirty dollars right and so what are you comparing that to i guess so suppose you on average you want to make a certain amount of money which is 50 60 and so on now if on sundays you make a lot of money and on sunday you make very little money you might sort of evaluate that separately and say on that day i'm essentially in the last domain i'm like below my target or below my expectation and so if you evaluate your utility that way you might sort of not want that because it feels essentially you're below a certain threshold it feels kind of like a loss relative to your expectation and so on and you might be inclined to work a lot of hours instead on the days when you make a lot of money why might you stop yes you might have like a target that you expect to hit every day to cover your expenses and if you feel like working less you can still reach that target right so if your target your reference point is essentially a a certain number you might reach that target quickly because your wage is pretty high on that day now once you reach that target you might say well like now your utility function is relatively not very steep anymore relative to being in the last domain so it's like flatter so then you might just stop relatively soon because essentially any marginal earnings are not really valuable for you anymore okay any questions on that or comments so why do we want the wage changes to be temporary here what's an issue here when you sort of try to look at this in the data suppose i had like persistent wage changes yeah so you can have like a frame of reference yeah you want to kind of keep the reference constant so in some sense if wage changes are permanent uh then you get essentially income effects you'll be a lot richer overall if they're only temporary essentially you can sort of argue that essentially your expectation should be the same and once you reach once you have a lot more money then your classical model should actually say that there should be income effects the neoclassical model only says essentially the inner classical model says you should essentially aggregate all of your income and say it doesn't really matter like whether you earn it in monday or tuesday wednesday or thursday you should look at like how much are you earning overall depending whether you earn sufficiently much um uh you're gonna work fewer or more hours now if you earn like a lot because your wages doubles or whatever you might actually work few hours not because you reach a reference point on a given day it's just because you got a lot richer and then you decide it's not worth for you to work that much so we're kind of trying to avoid that we're trying to have only permanent temporary changes which is to say forgiven wealth overall um on any given day it shouldn't matter whether you earn the money on monday or on tuesday so essentially if the wage happens to be really high on tuesday you should be working more earning more on tuesday as opposed to on monday uh now what i was saying is if you reference the pendant you might actually care about this you might care about on monday you didn't reach your target therefore you want to work more on tuesday you reach your target very quickly and you work uh uh less even though the wage is actually really high okay now strategy one might be optimal even in the neoclassical model if effort costs uh are convex this is what i was saying is if it's really sort of costly for you to work nine hours you might say i always work like eight hours i wanna have a certain amount of money for my children and so on um so the extra hour if that's really really painful you might not want to do that uh we don't think that's necessarily the case in in so many uh situations um the question is like can we really say that strategy three that's the strategy of doing the opposite of working like essentially a lot of hours and wages are low and few hours when wages are high can we really say it's a mistake well it depends on kind of like whether the effort costs are correlated with wages so if cab drivers for example make really high wages on some days and low wages on other days it could just be on low low uh wage day it's really um uh it's much less effortful to drive around so you would do like more hours it turns out when you actually ask a cab drivers they actually prefer a busy day is actually preferred when they're customers as opposed to just driving around and looking for people so we don't think that's actually true but in principle you would have to sort of think about uh that now uh there's a long literature uh starting with camera at all on uh cab drivers a lot of that essentially is pre uber like uh collecting trip sheets um from cab drivers now there's lots of like uber data that's like essentially much more powerful in some ways so there will be probably more papers using uber and lyft etc data on that um but what camera doll did at the time was they essentially looked at a typical cab drivers that rent their cab for 12 hour periods for a fixed fee and within this 12 hour window a driver can choose the hours freely so you just get your cab that's not yours you rent it for the day and then you can essentially choose how many hours you want to work and their wages how much money they make in any given hour varies by a lot there's like you know uh the weather varies there's subway breakdowns conferences and so on and so forth there's lots of variation in how much money you make on a given day so then they have like trip sheets that look at like how long cab drivers work and their overall earnings and so they can essentially back out the wages from from each day and then look at like how much do people work on uh days when wages are high versus wages are days when wages are low and then they essentially find the basic finding and that's a finding that's sort of been contested in the literature and many people are found are confirmed subsequently or and others have not but i think eventually people have sort of settled on this hours are negatively correlated with wages so when wages in particular are unexpectedly high uh cap drivers tend to work a few hours and again the and this is like not for permanent changes but for transitory changes if surprisingly on a given day people get more money then [Music] drivers work few hours and that's very hard to explain for the neoclassical model because essentially like you shouldn't care about like whether you make the money on monday on tuesday as i said you should care about how much you money you make overall and how many hours you uh work and so it's very hard to explain this and sort of the explanation uh that cameradoll and others um we're testing or or arguing is essentially this has to do with um reference dependence and so what's being evaluated in a reference dependent way or how do we think about this yeah i have a question yeah um how did they rule out the possibility that like maybe there's like um reverse causality maybe i'll say where none of the cab drivers want to work that much because of supply and demand but they just go up so so that's to say like effort costs are high so usually it's to do with like other drivers so let me let me actually get the so let me defer this for one second i have like a slide on like confounds and then we can see whether it addresses your question and then get get back to that but that's a good question so uh what is being evaluated in reference dependent way here what are people looking at in terms of where is the reference point in the evaluation here yes um i guess if you're a cab driver you kind of say oh i want to make like this much money today and then you just kind of like work until you feel like you've made it up and then you just stop right exactly you have to pay your fixed fee for the day or for the month or whatever but like there's an implicit uh fixed fee for the day so you kind of want to make at least that much money to make not a loss you probably have some positive target in some way and saying like i want to make uh pay back like my um my fee plus you want to make some money for the day and minus sort of expenses and once you reach that target you are in the gain domain below that you are in the last domain right and so the daily income essentially and that's essentially sort of like uh money after paying back the fee but you could also get uh it's essentially your uh uh what's being evaluated in the reference append network what's the reference point it's some daily target that you have often it's like expectation and so on and sort of essentially how much you think you will make and then what's the feature of the value of a function that explains the phenomena well it's loss of version if you're like falling short of the target essentially your marginal utility how much when you drive another hour or another trip the marginal utility that you get to see it then the value function is very steep below the target is very high once you reach the target it's very flat and then essentially you you you tend to stop does that make sense yeah okay great so um the main takeaway here is although if drivers often stop at their daily income target um drivers with a higher wage reach their targets faster and they work few hours again that's like variation uh uh uh within drivers um across days and so that there's lots of subsequent work and debate regarding this finding that the debate is still ongoing for example one reason paper looks at like tips that people uh drivers get unexpectedly so sometimes drivers get like large tips sometimes get small tips it depends a lot like when in the day they receive the tip if they receive it really early versus late if they sort of get to their target and essentially the target seems to be adjusting over time uh but overall this is sort of getting a little bit at your question um uh where we think it's not sort of aggregate um supply but the debate is over are still ongoing but overall we sort of think that lots of labor supply decisions when people have daily decisions of how many hours to work and their wages vary depend on uh reference points uh and might sort of potentially be at least um suboptimal or put differently they could people would could work few hours um and trying to sort of adjust their overall amounts you can actually when for a while like this you can ask your uber lyft drivers what they're doing and so on and see whether they're a reference dependent uh a driver so they have reference dependent preferences now what are sort of potential alternative hypotheses one question one issue could be like liquidity constraints this is like for example if you just don't have enough cash and you have to pay back for uh your fee for the day or for the next day you might want to not work only very few hours on a given day uh it turns out that drivers who own their own medallion exhibit the same patterns they could be things like fatigue it's let's just say it's really tedious to work on certain days when wages are high we don't think that's going on in part because drivers themselves they say it's actually easier to drive with more passengers again recent papers in fact can also control for this and so on so we don't think it's actually um fatigue and this is i think what you're saying uh the last one is unobserved shocks so some shocks that affect all drivers labor supply at the same time example you know there are some days in which all drivers get the flu uh fewer drivers will work and those who will work uh work few hours and those who work get uh get higher wages we that's a little bit trickier for them to to to to roll out in part there's other papers other studies afterwards that sort of try to get at this usually um um yeah for this specific data that's hard to do i think for the other data where people have essentially not just daily so this is essentially trip sheets so they have what they're using mostly is like daily wages overall they don't even have the wage they only have the um the overall um uh earnings and then the hours and then they sort of impute the wage dividing the two which causes some other trouble but once you have like uber data once you have like specific trips in particular also tips and so on you can look at on a given day uh when i get like a high tip versus a low tip uh uh how do you can you can look at like you can predict essentially how much i'm gonna earn on a given day so suppose you you predict that i'm gonna uh earn hundred dollars in a given day now i have like say fifty sixty dollars now i get like a large trip and like a tip and so on that gets me over that threshold you can look at where their stop you can look at the exact same thing when i have like only twenty dollars and they get a large ship to a stop and so on so you can essentially control for all of that and then do within a driver comparisons like for trips uh that happen to be large or small that you happen to get in a given hour and then you can sort of deal with like overall market conditions i think even better in the future there will be sort of like experiments and so on where you can look at when uber and lyft try to incentivize their workers and so just say essentially sometimes pay people more and unless randomly sort of explicitly randomly because they want to kind of learn about how to best incentivize the drivers and then you can control for everything because it's like explicitly random whether a certain driver gets like high versus low uh uh wages or sort of trip fares yeah okay um are any questions on the um on the uh labor supply yes um i'm a bit confused about the unobserved shocks because like if you get like a really large tip that doesn't necessarily like predict that the rest of the day you have higher wages if you kept working okay like you just get this really high tip and then you might be like oh today's a lucky day i can stop early but if i had worked more this day i wouldn't necessarily be making continuously more than normal right so in the unobserved in the large tip example the assumption is exactly or what they show in the paper this is subsequent work not this specific work what they exactly show in this type of paper these are sort of random events in a sense it's really precisely not predictive of your future uh uh earnings it's random now one interpretation that you have is to say well it could be that you kind of have some expectations about your future earnings that could be particularly high or low it's like i got my lucky day and so on and so forth that is hard to rule out and sometimes if you had rational expectations uh that's hard to explain for the neoclassical model what's harder to rule out is to say you know i now think like i got my lucky draw for the day and that's it and i'm not gonna uh i get any lucky draw again let's just call it a day that is hard to rule out but what is um hard to explain for the neoclassical uh model i think we can sort of reject is to say if you just think this is a shock which really in reality it is and you just happen to get a bunch of money on given day uh just randomly uh that is then the reference dependent model can sort of explain why you stop early because you it just puts you tips gets you over the reference point the neoclassical model should just say depends on like essentially how many hours you want to work on a given day it's nothing to do with a target because essentially again it doesn't matter whether you make money on monday or tuesday or wednesday you should just care about like your overall amount of money that you make overall but that's a good question okay so now um a second paper is on the housing market so what is the natural reference point for housing markets or like how do we think about housing prices or how do you think about i guess a few of you will own a house but how would you think about if you if you owned a house how do you think about uh selling a house what is a natural reference point yes uh whatever it was for or like how it changes yeah so exactly so uh the previous purchase price it turns out it's a very very salient thing for owners people really know how much they paid for their house it's like a huge expense in their life they really remember it was like three hundred thousand five hundred thousand a million et cetera they know exactly how much they paid and that even asked people people know exactly how much they uh the majority of people know exactly how much they paid for their um home now one claim is that loss of version makes people unwilling to sell their houses at a loss and so what they would then do is they ask essentially for higher prices if a loss relative to like their their purchase price and let me just show you um exactly what i mean by that uh gina risa and meier have um boston condominium data from 1990 to 1997. um luckily for them there's lots of variation or fluctuations in the housing market during that time so the housing market went up a lot and then it went down a lot and went up a lot now suppose you have two sellers a and b who both want to sell their house or their home in 1994. now what we can do is we can look at these two people who suppose they have very similar houses based on observable characteristics and location and so on and so forth so once they have really comparable houses we can now look at people who seller a purchased their home really early on in 1989 that person happens to be quite unlucky because they just bought the house really at the peak of uh i guess the housing boom at the time or seller number b who purchased in 1991 uh uh that that person was relatively lucky the bot essentially had a price where at a time where the housing market was relatively low and uh appreciated quite a bit so now we can look at these two people and ask the question uh is seller a or seller b more likely to sell their house and and the hypothesis is that seller a will sort of view this as like a loss this seller will essentially just say i'm losing money here i don't want to sell and that seller might essentially ask for like a higher listing price and wants to like more money for this house because they don't want to make a a loss on that uh sale well seller b says i'm actually gaining money anyway so let's just sort of like uh post whatever you think is actually the expected market price might be happy to to sell at a lower price compared to seller a okay so one thing we can look at essentially are listing prices how much do people want for the houses effect second thing we can look at is like actual sales prices you know are you now like selling this at a higher lower price and number three we can look at like how long is the house on the market how long does it actually take to sell it that's a quite costly thing to do to have your house on the market for quite a while because essentially often people then don't already move somewhere else or they don't kind of live in the house because they have to sort of show it and make it available for showings and so on so it's a costly thing to do you really don't want to you have your house on the market for like several months but it's sort of the broad idea claire what we're trying to do okay so what predictions um do you want to test but you want to test whether house owners are reluctant to sell their house when the current market price is below the purchase price so the ideal specification is essentially the following we look at the list price on the left-hand side and then we are going to run a regression that looks at some constant that's just kind of like time trends et cetera plus a beta which is the coefficient of interest or one coefficient of interest is of the actual market price how much is the thing worth and then delta on the loss is like how much do you lose relative to your purchase price now if people were not reference dependent what should we find what would be the predictions like the neoclassical model what should we find here yes exactly delta should not matter it shouldn't matter how much you lost or gained in that house uh what should matter is like what is the actual market value you should essentially try to be willing to sell it or not beta could be you know it doesn't have to be one it could be like everybody pays above the actual market for some housing bubble or whatever beta could be like 1.1 or whatever you might be uh that depends on essentially sort of aggregate conditions and so on but delta really should not matter when i'm trying to sell you a house you should ask like what's the actual market value and i should um essentially then based on that uh uh put it under market um on that listing price uh but what might be the case is that the loss actually um matters now one problem here is that the actual market value is not observable right so i don't actually know what the market value is because that's like endogenous that's part of like the transaction so what can i do instead or how do i do this yes by asking people like would you like would you take this trade what they say right so you can look at actually the market outcome overall you can look at like who buys it and what's the actual purchase price now that might also be endogenous to like the listing price right so if you think you know the people who are lost a verse are listing their houses too high relatively compared to what the market value actually is it might actually be sold at a higher price so that could that's hard to interpret it could just be that if you sell it at a high if you list it at a very high price and wait for a long time you also sell it at a higher price but doesn't mean that it's actually worth that much it just means that like you happen to find a buyer who happens to be willing to pay a lot and that's more likely when you wait for a long time which is not necessarily optimal because it's actually quite costly to do that but you're saying something else which is like asking them about what they think but what are the characteristics that we could look at or what data could we look at like if you had like zillow data or data on like essentially a bunch of these apps where you can look at houses what data could you get yes you'd look at like the houses like in the neighborhood like similar houses exactly what redfin and zillow et cetera do these days is essentially they have these algorithms that try to predict uh the actual market sales price and what they tend to do essentially is look at exactly as you say surrounding houses that are similar in some characteristics they look at the number of they look at the square footage they look at the number of uh bathrooms and rooms in general they look at like sort of like location and so on and so forth and then you can predict um uh how much uh and sort of they look at also time trends and so on how much is the actual market value but it's essentially fundamentally it's like a prediction exercise you can try to predict uh what the actual value is and that's exactly what uh genovese and meyer do they do some more fancy things that try to get at like other unobservable characteristics and so on but the essence of this is exactly a prediction exercise where they say let's just look at uh the characteristics of this house let's try to predict what the market value is and then look at the loss which is essentially the difference between the previous selling price and the expected selling price truncated from zero because otherwise it's a gain and then we can look at uh does it really seem that like when people are in the lost domain when their losses positive are they now um selling their house or trying to sell their house at a higher um price yes how would you account for something like recent renovations like since the last listing and relative to any other similar units or causes and nearby right so um that's tricky to do and i don't think they have this in their data so what you'd have to assume here and that's perhaps reasonable is to say that when you look at sort of this picture that um seller a and zellerb did not do differential renovation depending on when they bought the house right so if you said it's fine if people have done renovations for example this is what i was saying about the beta on the actual market value if you systematically underestimate or overestimate how much people renovate and so on or maybe you know the housing market is really hot or whatever that's okay the main issue is like that can't be correlated with the uh uh the loss here so if it's the case that people who may who lost a bunch of money in terms of like their housing prices sort of like tanked uh if those people have done more or less renovations then you're in trouble with your estimates yeah is the expected selling price the estimation of actual work value correct okay yes um so now what what generation meyer find is and there's more detail to that but essentially the main finding is and that's a fairly solid one there's a bunch of different specifications is a 10 increase in a perspective loss if this loss coefficient the difference between the expected selling price and the uh purchase price uh if that's 10 percent higher than essentially the list price is 2.5 to 3 percent uh 3.5 percent higher and then these effects so like people list the price the house at a higher price if the house is at a loss compared to like a similar house who's not at a loss or like it was in the game domain these effects then translate into higher uh sales prices and the lower hazard rate of sale that's to say people actually sell it at a higher price so it's in fact you know it's hard to say here what's optimal versus not it could be that overall people are like uh you know it's actually a good thing to do that depends a lot on essentially like your opportunity costs of of money in terms of like or you know your how costly is it for you to keep the the house on the market for a while but essentially um uh people have a lower hazard rate of sales that's to say it just takes them a lot longer to sell the house that tends to be very costly to do so but you know if you you know just otherwise would have your money in the bank and you have another place to stay and you don't really care maybe then that's fine uh but we surely have like real effects in terms of like uh people list their houses as high at higher prices people sell them at somewhat higher prices and people also keep them longer on the market okay yeah is this the same analysis if it's gains in the value of the house so essentially what we're doing is we're implicitly comparing losses versus gains right so so that gains the gains here would be in the actual market value already as it is so it's so but essentially what you implicitly do is you create when comparing implicitly this is what i was saying here when you look at this picture implicitly what we're doing is we look at people who are um losing money compared to people who are gaining money and explicitly implicitly asking like is the increase in the gain sort of predictive of that sorry the the increase in the loss is a predictive of like your uh listing price now it's hard to do this at the same time for for gains because essentially an increase in the market price overall is this like that sort of culinaire in short essentially that's hard to separate you could do the same analysis and you wouldn't find that for the gains in part because essentially explicitly what we're doing is increasing losses to gains okay uh okay so then there's another piece of evidence that sort of tells us perhaps this is not optimal if you look at people who are owner occupied compared to investors so they're people who essentially invest in houses and they sell houses and they sell lots of houses over time those people have much lower uh endowment effects if you want for houses so they exhibit this behavior a lot less but people who live in that house who bought that house for them it's mostly like the only house they live in it's the main purchase that they have for them it's essentially their house for which they don't want to make a losses for them these effects are twice as large compared to the investors and that's perhaps some evidence that uh this is not optimal behavior in a sense of the sort of the professional investors they presumably know pretty well uh how to price their houses so if they do this behavior less uh presumably that's that's a sign that that there's some form of a mistake here um going on second there's some evidence in john list has some work on this overall is to say that experience can mitigate reference dependent effects essentially if you do a lot of trading if you sell a lot of houses and so on then you might still feel losses of losses and gains but you might sort of have a lot more experience with this you kind of know that this is happening sometimes and you might be less prone to these types of effects because you kind of know that you shouldn't be doing this you shouldn't sort of have your feelings of losses and gains get in the way of making profits so there's some people who would argue that this is very much consistent with like markets over time or exposures to markets and several predictions experience makes some of these effects go away and john list has some separate evidence on this on traders of cards and so on okay so now next we're going to talk a little bit about finance and stocks so interestingly yeah sorry on the previous um study why is this a reference dependence on information um owner and i think that my house is worth this amount and i'm just really anchored to believing that that's the problem right so one thing you could say is like that uh so what you always have to make the argument is like people who are losses compared to who are at gains so you look at your investor a and b they might have like additional information on uh how much the house is worth so it could be for example that seller a knows a lot about this house it's like very beautiful and so much light and this and that um and therefore they paid a lot therefore they have private information that it's worth a lot therefore they listed at a really high price so there's some specification here that look at this what you see is this is like columns two four and six which is uh the residual from the last sales price what is that that's essentially the difference between at the time when previously the house was sold what was the prediction of the market price then and how much was it sold so it's kind of like how much did you over uh uh how much did you overpay if you want relative to what we expected at the time presumably that's reflective again using the same prediction method so if you thought you know it's really beautiful and lots of windows and this and that lots of light and really quiet and so on if you overpaid at the time that should then sort of pre predictive of the listing price and sort of controlling for that then should make this go away what you see however is this is um there's some of that perhaps going on if you compare for example columns one and two you see that the effects goes down a little bit but it's still there quite a bit what's this is why i was saying like 25 to 35 percent uh that is exactly right that's a big concern and there's a bunch of sort of robustness et cetera checks in this specific study but that's exactly right there could be sort of unobservable information that the uh owner might have about the house that is not in zillow or in any sort of return etc predictions that's available for the public that's a great question but i think it's uh to the extent that that takes care of it it's sort of uh the authors have thought about that uh yes um on the following slide when you talk about the differences it's how do you control for like the selection bias about the people that may be the ones repeatedly selling houses exhibit this last and therefore like stay in the market versus a change in those individuals behavior right that's a great question so the question you're asking is essentially to say and that's in fact a great um sort of a segue into like behavioral finance which is to say suppose there's some people who are really sophisticated they don't have certain uh behavioral biases maybe they're not philosophers that makes you like a better investor say and therefore you stay in the market overall i think from this observation that i have here i cannot tell you is it like select is it just is it just that people um is it experience or is it selection so the question kind of is when people are investors do they do the effects of reference dependence of gains and losses go over go away over time essentially like maybe the first second third time i feel like really a loss in terms of making a bad investment but in number house number 20 i'm just like that's like as usual and i shouldn't really care very much is it that this goes away or is it that like the people who are particularly lost a verse and sort of essentially uh engage in this type of behavior in terms of like uh over uh listing too high of a price um for losses these are sort of like that investor in the housing markets and they sort of are essentially driven out of the market so that specific setting um i don't think we can necessarily account for that i think in the uh studies by john list it's very much people argue it's about experience but again there could also some part could also be a selection i think so now in the so i think in some sense either way um uh uh i think the the evidence that the investors are doing this behavior last sort of tells us something about like this is probably not at least financially optimal for you to engage in this type of behavior it might be privately optimal in some sense if you really feel bad at like selling your house at a loss you should probably list it at a higher price because that sort of limits you your losses that's just what the utility function looks like it's not necessarily a and the sub-optimal in the sense of like uh how you feel afterwards it might be sub-optimal in terms of like how much money you make or how much money you have eventually uh on how much you pay for keeping your house on the market and so on for an extended period of time um okay so does that answer your question no yeah okay so um so behavioral finance is an interesting field because for quite a while economists thought that sort of neoclassical assumptions are in fact most likely to hold in financial markets and why is that why do we think what and some of this already i mentioned but why are financial markets particular why might i think that financial markets are particularly efficient yes well you might think that financial markets are very competitive and so it's actually the ones who can get rid of their behavioral biases that benefit exactly so it's very much like sort of the chicago economics assumption is to say so uh financial markets are extremely competitive if now i'm an investor who has various behavioral biases presumably i'm going to lose some money one way or the other well if markets are really competitive in the long run i cannot stay in this market without sort of being a driven out so essentially the market favor sort of results oriented rational and selfish behavior so people are not rational and so on and so forth will be eliminated from the market eventually there's two parts to that that's true like across firms that's to say lik