 Thank you so much for being here today. My name is Emily Bobrow. I'm a journalist with The Economist and we have another fascinating panel. It's always humbling to discover just how bad we are at making decisions, how poorly we know ourselves and how little we can predict how we're going to behave in the future. Many people have all sorts of laudable goals for how they want to handle their finances. They want to pay off debts, maybe save for a rainy day, but oftentimes what happens when the money is in their hands is they behave just a little less cautiously, a little less the way they behave a little bit less like the people they want to be. So it's a little bit interesting and certainly a little disconcerting to discover that there are ways to nudge people to behave in the way that they say they want to behave, to make the kinds of choices they say they want to make for themselves. I'm particularly heartened by something called the Refund to Savings Initiative. This is a collaboration between Washington University and St. Louis, Duke University and Intuit, and it's about encouraging more financially insecure people to save their tax refunds. To talk a little bit more about the results of this experiment is my great pleasure to introduce some of the architects of the project. Our first panelist probably needs no introduction. It's now incredibly fashionable to talk about what idiots we are at making decisions, how poorly we know ourselves. And this has a lot to do with Dan Ariely's extensive research into what makes us predictably irrational, which was the name of the first of several bestselling books. Dan Ariely is now the James B. Duke Professor of Psychology and Behavioral Economics at Duke University, where he is also the founding member of the excellently named Center for Advanced Hindsight. Michal Greenstein-Weiss is a founder and director of the new Involve Center for Health Behavior Change at Washington University. She is a professor at the university's Brown School for Social Work and an associate director at the school's Center for Social Development. Michal has devoted much of her career to rethinking policies for promoting financial security and a number of her big research projects are about encouraging more people to save. David Williams is basically the chief tax wizard at Intuit, but before he was at Intuit, he was, you know, doing all sorts of fascinating tax policy work in the public sector, including 14 years as a U.S. Senate staffer and 13 years with the IRS. I wonder if we could start with Dan Ariely to talk a little bit about how your research has informed some of this refund to savings initiative. Sure. So let me tell you about three pieces of research that we have done in the last couple of years that will give you kind of the flavor for how behavioral economics fits and informs what we do. So the first thing I'll describe has nothing to do with savings, it actually has to do with spending. So imagine you come to a grocery store and it's kind of a standard grocery store where people come for lunch and they usually buy drink and a sandwich. And there's different drinks you can buy and different sandwiches so the amount of money varies, but on average people spend about $6. And one day you get a coupon that says spend $8 or more and get a dollar off. It's a nice coupon, but you do get to spend more than you usually would. And the question is, would this coupon affect your behavior? But now consider two versions of this coupon. Version one, you get three feet before you enter the store. Version two, you get six feet after you enter the store. So it's not very different, but one of them is before you enter the store and one is after you enter the store. Do you think there will be any difference? Of course I would describe it if it was for the same. What happens? The people who get the coupon outside of the door are dramatically influenced by it. The people who get the coupon once they've entered the store and had a couple of seconds inside are much less influenced by the coupon. And why is that? It's because this difference of crossing the door is a difference between having a general idea of, hey, I'm going to go and buy something today and having a very specific idea. The moment people have a very specific idea about what they're going to get, I'm going to get a Coke and this particular sandwich, you can't really change their mind anymore. You really have to get them beforehand. We call this process kind of the idea that you have implementation intentions. At some point you have a general understanding about what you want to get. And at some point they get consolidated into action plans. And when they get into action plans, it's really very hard to move people. By the way, we did a bit more bizarre condition. We gave people a coupon outside that said spend $4 or more, get a dollar off. What happened now? People started spending less. It's a crazy coupon because you could spend $4 or more. It doesn't really matter, but people kind of use the $4 as an anchor. Okay, this is what I'm going to try and spend. By the way, this phenomenon happened to all kinds of other things. We had people to give us their shopping list. Before they entered the store, they gave it at very high categories, like I'm going to get a drink or toothpaste or something like that. The moment they entered the door, the category became much more concrete. They said I'm going to get Coke, I'm going to get a can, I'm going to get whatever brand of toothpaste they wanted to get. So what this actually tells us is that the moment that we get people is incredibly crucial. People move from this notion of having general idea about how we want to behave to having concrete plans, and once they have concrete plans, it's a little too late. Not very too late. Okay, the second experiment I want to tell you is something we're just exploring now. So we're asking people to come to the lab and to simulate the first day on the job, so they kind of meet the HR people and they have to decide what they want to put into savings. And some people would tell them that their salary is $35 per hour, and some people would tell them it's $70,000 a year. These are equivalent amounts. When do you think people save more? Is there a difference? When they save more? The $70,000 a year, people save more between 2% and 3% more in 401K, which is a big difference. Now, you could say, well, maybe what happened is that people don't know how to do the math. So let's take the people who we tell them to make $35 an hour and say, please tell us how much you think it's per year, and let's take the people who do $70,000 per year and say, how much do you think it's per hour? No, people are wrong, but on average, they're okay with that setup and they get it right and the difference don't explain the difference. So we said maybe it's about stability. Maybe the people who get money per hour just think about their job is temporary. And maybe the people who make $70,000 a year think, oh, this job is going to be with us for a very long time. So we tried that. We gave them contracts that were either $35 an hour, but for a really long time, or $70,000 per year, but the contract can end at each point. So we had two versions, $35 temporary, $35 long-term, $70,000 long-term, $70,000 temporary. No, it's not that different either. We tried another version which we said to people, you're making $35, which means $70,000 per year. Even now, it made a difference. It was the first number that people had. And then we did another experiment in which we asked people, not just to say how much money they want to put in 401K, we asked them to separate the amounts into two savings accounts. We said, why don't you tell us how much you want to save in a short-term saving and in a long-term saving? Short-term saving, there was absolutely no difference between the two groups. The difference was only in long-term saving. And I think it's, again, it's about the mindset, right? It's not something that people are thinking kind of thoughtfully, this is a long-term job or a short-term job. The moment people have something that says the word year, they think more long-term and are willing to put more for retirement. Now, the sad thing, of course, is that almost all low-income jobs are framed as an hourly basis, which would be terrible. The last thing I'll tell you is an experiment we did also about mindset. In an experiment we carried out in Kenya, this is an experiment in the slum called Kibera. This is with people who make about, live on about $10 a week, very poor people. And we tried to get those people to save a little bit for rainy day for all the reasons we talked about today. And we teamed up with EMPESA, which is the electronic payment system in Kenya, belonged to SafariCom and with an investment bank. And the reason we did it is so that people could text money into their EMPESA account, but every night the money moved from the EMPESA account to the investment bank, and now it was really hard to get it back. So if you want to get the money back, you had to take a bus, go to the city, submit a form, wait an hour, get the money, take the bus back. We didn't want everything to become an emergency, but we know that emergency happened from time to time. So we gave the system to a lot of people. And then some people just got that system and nothing else. Some people got that system, and on top of it, they got the weekly reminder that it would be a good idea to try and save about 100 shillings, about $1 that week. Some people got the same text message, but as if it came from their kid, hi, mom, hi, dad, this is little Johnny, whatever the name of the kid was, try to save 100 shillings this week. By the way, there was no deception. These people knew that the kids don't have cell phones, and they knew that the message didn't really come from their kids. Another group of people got a 10% match. At the end of the week, save up to 100 shillings, we'll match it up to 10%. Another group got a 20% match. Two other groups also got 10% and 20% match, but they got what we call pre-match. What's pre-match? There's a principle in behavioral economics called loss aversion, that people hate losing more than they enjoy gaining. We said, what if we deposit the money, the 100 shillings, the 10 or the 20 shillings into their account in the beginning of the week, and we said, if you save 100 shillings, you get to keep it, you save 50, you get to keep half of it, you don't save anything, we take it all back. And then the last group got a coin about this size, and we etched the numbers of the weeks, 24 weeks for the program on the edges of it. We asked them to put the coin somewhere in their hut, and we asked them every week to take the coin and take a knife and scratch the number for that week and scratch it one way if they saved and a different way if they didn't save. And think about all of those methods and try to figure out which one you predict people would save the most. We could take a vote, but try to think to yourself which one you think people saved the most. We've also done a study on asking people to predict the results. And I can tell you when a large sample of people predict the results, they think that the 20% pre-match will be the most effective, and they think the kids and the coin will be the least effective. But what actually happens? Just giving the system to people with nothing else creates a little bit of savings, already good, right? Just making it a little hard to get the money back. Adding a weekly reminder helps substantially. Adding 10% at the end of the week helps a bit more. 20% at the end of the week, just like the 10%. 10% in the beginning of the week helps some more. 20% in the beginning of the week, again, just like the 10%. So loss of version works, but 10% or 20% don't matter. By the way, people think that 10% or 20% will make a big difference, but it doesn't. It's not about the financial incentives. Kids, by the way, were indistinguishable from the 10%, 20% with loss of version. And then the coin was the big surprise because it increased savings dramatically above everything else. And of course, the question is why was the coin so successful? As Emily mentioned, our research center is called the Center for Advanced Hindsight. And we use the name also to remind ourselves that it's very easy to see a result and say, oh, I knew that all along. But we need to remember all the cases in which we thought we knew we had a different prediction. So what happened? So the coin was the most successful. And when you look at the days of the week, everybody saved on Thursday, which is the day that we sent people the text message. And the coin was slightly higher on Thursday than the other conditions. But what the coin really got its benefit was all the other days of the week. And all the other days of the week, in the no-coin condition, people saved nothing. In the coin condition, they saved from time to time. And again, this is all about mindset, bless you. This is all about mindset. Imagine that you have kind of a universe of things that you could be thinking about. Are you thinking about all of them? Of course not. You think about the subset of them. Which subset? Whatever comes to mind. What is it that in your environment reminds you about savings? Nothing, right? And if you live in Kibera, certainly nothing in your environment reminds you about savings. You think about spending. All of a sudden, we had something in people's environment that from time to time reminded them about savings. They thought about it from time to time and took some action. Now if you think about all of those results, if you think about the question about the supermarket and mindset, you think about money per hour versus per year, you think about the coin, these are all very delicate effects, right? They all basically affect about, are we catching people at the right moment? Are we catching people at the right mindset? Can we create the right mindset? Can we do it at the right moment in the right mindset? And hopefully with those ideas, we can get people to behave differently. Final words. So this is it? That's simultaneously disheartening and encouraging. It's good to know that these are things that we can correct for even if it's a shame that we constantly make these bad decisions. Michal, I wonder, or David. Sorry about that. David, why don't you walk us through some of this project and how these lessons are being put to action? Okay. However, I want to start with a quiz because you haven't had one in a while would be my guess. How many people in the audience took microeconomics at some point in your life? I would say that's a majority. And it may have changed because when I took it, it was a long, long time ago. You remember the concept of, at the time, forgive me, it was the rational man, right? Everything was about at the margin, what would people do and you could see the indifference curves and people behaved in this amazingly rational way, which we all had to learn and study. It was a prerequisite for graduating and now you've learned it's crap, right? Not completely crap, but I think even as we did, I still recall the time thinking, who really thinks that way and do I need to start thinking that way? And I'm in a course that's teaching me how to think that way and the reality of the matter is people do not behave rationally all the time. And in fact, as a matter of course, we have built public policies in this country that make assumptions about how people should behave. And we build programs and we administer them based on how that works. And what I love about refund-to-savings and I will talk about some of the details in a moment, you'll get the big details, I'll just cover the fun stuff, is that we've really recognized that while we have outcomes that we want, we want people to save, to plan, to be ready for their financial lives. The methods that we've used so far have not gotten us where we want to be. And so we need to break out of that box and that's what I love about what Dan's talking about and it's why my company and why I personally believe so much in the notion of that moment, that in this case, the tax time moment, as a way of leveraging someone's intellectual bandwidth, if you will, the cognitive ability to actually think about their futures and things that we think are important. And when I say we, policymakers, think are important. And with that context, let me talk a little bit about refund-to-savings because it specifically focuses on that tax time moment as a way of leveraging that instant when someone is actually thinking about money coming in and maybe they've actually gotten at least a broader sense of their finances as a way of encouraging folks to save. And that was the question that we began with as a team which was how can we build financial security, specifically savings, for families and how can we do it in a way that's scalable. And I want to emphasize scalable. We talked a bit about VITA today which is an incredibly important program. Our company has supported it for years. I worked on it when I worked at the IRS. But I would submit it as very difficult to scale VITA. And you heard, I don't know where she is if she's still here, but our colleague from New York talked about the amazing work they're doing to try and drive and build VITA. But it still is hard to reach all of those people. And that's in a place like New York where there's an active government trying to drive the program. I worried when I worked at the IRS about folks in rural areas who didn't have access to those kinds of resources. So scalable and scalability matters a lot to me. McCall brought this incredible background and expertise in asset building for low-income families. Dan joined and really brought his background and you can see his thinking about how people actually behave and how we can drive and change behavior. And our company basically brought its insights and the platform. And our platform, just to be clear, we build a product called TurboTax which is a do-it-yourself tax software product which reaches roughly 30 million people, families, filed through that product. Many people don't know that we also have a tax pro product that files roughly the same amount, 25 to 30 million returns come through the tax pro product. And so we look at both of those as opportunities to think about leverage to encourage people to save and particularly in the TurboTax product. And we spend a lot of time on insights. So the notion of refund to savings is what can we do in the tax experience to trigger people's thoughts and get them to save money? Very, very simple. Big or small? We'll talk a little bit about that. We've done this work for several years and tried to innovate around things like the anchoring that Dan was talking about. There are social proofing. There are lots of ways one can use behavior economics and we've done some experience with them. I'm going to let McCall go into the details. Let me just talk at a high level and with apologies to David Letterman, our top 10 list. Because you haven't had one of those in a while because he's off the air. So these are the 10 things that we learned and I'll cover them relatively quickly with some of the insights that we've gotten through the work so far. And to express a little bit of excitement about the work that is coming, particularly in partnership with Treasury around a new product called MyRA which I'm sure you will hear more about from our next speaker. So top 10, we'll start with 10. One, the population that we were looking at was far more financially vulnerable than we realized. I think most people in this room would know that. But when you talk about outside folks are like, I don't get it. We've got the ability to get your hands on $2,500, whatever it is in case of an emergency. The bottom line is a lot of people in this country shockingly don't. And we can sit in this room, what a Clinton called the ceiling audience. Was that what we are, the ceiling audience? And we may well be, but the reality is that many folks who we are really looking at trying to influence do not have the financial capability that we thought they did even a little bit. The income households are very volatile. Their incomes are volatile. And we saw really a graphic demonstration of that. I would add and submit that it's not only income that's volatile, it's family composition that's volatile. For those of you who know how the earned income credit is structured you know that there are particular rules about when a qualifying child can be claimed and that also applies for a variety of other federal tax benefits. The assumption in the tax code is that people have sort of stable financial lives and stable families. Well that just doesn't happen to be the case. You'll hear more about that for me at another venue at another time, but I think what we've learned is being able to predict all that is really tough. I'm going, unsecured debt, but not secured debt is associated with lower savings rates. Who knew? That one's pretty obvious. I think anchoring works really well. So the notion of saying 25% works very well in prompting people to save and a corollary to that is that higher numbers actually work quite well. So not a bad thing to do. Most people who save say they save for emergencies. So that notion of why are you putting money away? Well, many of us might say well it's for our financial futures. Particularly in this population we are looking at folks who are saying I'm saving because something bad is going to happen. I don't know what it is, but I need to be ready for it. This is one that's important and it's a learning that we had as a company and I think it's insightful which is if people open their emails and that's an if because we don't get high email rates they are more susceptible to savings interventions. So it's a self selection obviously you see the email, they open they are more open to actually the prompts and other things that we can do to prime them if you will to make that savings decision. Believe it or not we found that people do save portions of their refund. Not in gigantic amounts and not in numbers that perhaps might resonate in this room but as a bottom line message refund to savings works and we found ways to make it work and it resulted in a lot of savings with very low touches and I want to just talk about that for a second a prompt in software or with a person may not seem like a lot. It's not a heavy lift for us, it does matter when we build it because it is about how we provide service to our customers but if a minor change in wording can generate a million dollars in savings maybe folks won't think a million dollars is a lot but you know two words equals a million dollars that's a lot of leverage if you really think about how we can get people to save and frankly I don't think that we're going to solve the savings crisis or problem that we're talking about just through what we're doing but this is one component that I think can contribute to our overall solution and lastly and I think this is actually critically important some of the behaviors about saving that we are able to generate persist after tax time and you're going to hear from McHaul how that happens the bottom line is that some of it is persistent which I think is another goal we want people to continue to save and so we've been able to show that that's possible using the behavioral economics principles and the design that McHaul and her team put together for us so with that let me turn it over to you well I'm sorry Emily I just know there's a presentation coming okay thank you so I'm a professor I will stand and use a power point for my presentation so for us okay so I will continue talking with you about you know what we have done what we have learned from 2015 I will build on what Denner really say and the Williams and go to our kind of high level and early learning from the 2015 year but because we started with quizzes and because I know there are many people here who are either running tax time program doing research on tax time or thinking about policies you know encouraging people to save a tax time I'm just curious how many of you have already filed your own taxes please stand up let's give them a round of applause a round of applause we need to see you guys okay so while there are still two months left into the tax season at this time last year already half of the R2S participant filed their taxes so two months left and already more than half of them were done so why is it because tax time is a really important financial moment and that's why we are all here today and that's why we are talking about it and here is some number to kind of suggest that it's really important financial moment but it is especially important for low income people so again looking at our data and what we have learned from this initiative last year in 2014 we have found that the average refund was equivalent to 2.5 months of housing for our people and also to 1.3 months of income average refund so many such important and large components of their earnings and that's why we started the refund to saving initiative and then we are trying to test behavioral economics we are using the platform of the free file and that allows us to have access to very large scale of people so we are doing our intervention every year with about half million households so we are very fortunate in that way to be able to have a really rigorous testing and we use randomized control trial and we are really trying to learn how we can develop low cost and low touch and scalable saving interventions and to be eligible to participate in the free file online there are several that you can see here in red people have to on less than $31,000 a year or qualified to ATC or beyond military it's a little bit hard to read so I'll just tell you so as you can expect how participants are having very very low incomes they are relatively young but they do have a relatively high federal tax return because of the qualification for ATC and other so what we have done, what we have done this year what we are doing to try to encourage them to save in 2015 we really built on one of major behavioral economics component choice architecture right, choice architecture is described as people trying to engineer choices to impact people behavior in the most predictable way so we are all familiar with the aisle of the grocery store we are faced there with many temptations candy for our stomach and candy for our brain who married whom in Mexico but what if instead we'll put some more we'll have the better healthier choice of food or more intellectual readings for example so this is we have built on this technique in our intervention in 2015 we took these techniques of choice architecture of changing the choices to try to encourage more savings so and it's also we intervening in the golden moment so we intervene when people are finished filing their taxes and they're getting to this point when they need to decide what they do with your money and it's a golden moment the money is yours but not quite in your head so that's when we intervene they get to this end of the finishing file they're getting the screen that say how you would like to get your refund and usually what they see and what control group have seen in 2015 is just three simple options do you want to direct deposit into your bank account or get a paper check or split your refund there is no strong saving emphasis right so what we have done for our treatment group is we have changed that we engineered this screen to start with two very strong saving options one saying would you like to save your entire refund into a saving account that's kind of the first option and the second would you like to save part of your refund into a saving account so this is what we did in terms of changing the presentation of the choices on the screen we also introduced motivational prompt we also have three different treatment groups and for each one we use a different prompt one we're talking about emergencies the other one we're talking about retirement and the last one we're talking about future so the final screen looks something like that this is an example of one of the intervention group looking on future orientation you can see that in the red is the prompt that people are getting encouraging to save for future goals and there is some interactive prompt on the side they are also getting this choice architecture the new engineering screen for us so what we have learned we have learned just by changing this just by this simple change to the screen we are able to really statistically significant but also in a very meaningful way increase how much people deposit into a saving account so you see here with more than 5% increase in how much people in the treatment group deposit into a saving account compared to the control group and that's true across all the different intervention all the different treatment group and that's also true for the amount of how much they save so the same result we see in substantial and meaningful statistically significant increase on how much they save and here you see on average how much to save but people who were able to get to save more usually save their entire refund so they save $2,000 more here is just include also people who didn't save so there was zero in what led to the statistics but the impact is quite substantial we compared to our control group the treatment group had 51 increase in the number of filers who deposit into savings and 46% increase in the average amount that deposit into savings now imagine you receive a large sum of money like 1.3 times of what you make in a typical month how long it will take us to convince you to save part of this money the people in the intervention group spend only 3 to 4 seconds seconds more than people in the control group where they decided where to put their money so just a matter of seconds when you can engineer you know people to spend more you know spend more into a saving account and that's translate to a really large impact again as Dave said this year we even seen a bigger impact than previously this year we found that we were able without this intervention generate over 21,000 more new savers and that's translate to 35.6 million of new savings that we wouldn't have seen without this intervention so in conclusion I only had like a very short of time just kind of share with you the high high level but from this kind of high level result we see that choice architecture seems to be very powerful we found a substantial increase in the rate of savings and how much people have saved the amount saved emergency prompt seems to be the most effective although the other prompts were also effective and we do know that people in this population face with many many emergencies and we find that low cost low touch interventions are very effective so what next for us we also part of the refund to saving initiative invest a lot of time developing this household financial survey we administrative household survey very long at the end like very detailed at the end of the when they are filling their taxes and it's include like a deep look on balance sheet assets and debt to save and what is retirement saving and use of government programs et cetera very detailed and we had at the end of the taxes and we had almost 24,000 households filling this survey and we did a six months follow up with 8,840 participants that completed that and will be analyzing this data now and we also as you will hear from Melissa we have worked this year on trying to test how we can get people at tax time use the MyRA to how we can use behavioral economics to get people at tax time open a MyRA and Melissa will co-ed their remark shortly after me after this panel we'll talk a little bit more about the old treasury work with regarding to MyRA and we also are testing right now in the field new interventions which will be our 2016 intervention so with that I want to acknowledge all our partner and go back to my seat and open it back to Emily I want to make one comment about the time difference the control condition and the experimental condition also had different amount of text on them and you would expect people to take more time the fact that people take so little time doing turbo tax to start with and how little time was added in the experimental condition I think should get us to rethink the kind of interventions we have so in no way do I want to say that doing your taxes in turbo tax is not completely fun and exciting and emotional people just want to stay doing it for longer but it does seem that people are rushing through the process and maybe it's not the right time for interventions where people have to read more or think more carefully because one of the important things about choice architecture is you want to engineer your intervention to the type of attention, mindset and so on and I think for me this has been such an interesting surprising data that I think would inform our future interventions to be more mindless and less thoughtful so let me just respond to that so we've made a business out of reducing cognitive burden essentially I don't know how many of you have ever tried to use turbo tax but the notion of reducing the friction the mental friction that one has to go through to complete one's taxes is something that has built an amazing business and so this notion of adding a couple seconds actually matters a lot to us and it has to be the right couple of seconds and it has to appeal to people but it also affects millions of people at once and whether it's in turbo tax or whether it's applied more broadly you heard a discussion about apps and technology earlier this is not just a tax time application it is the way of the future for people who are engaging electronically and so it may seem almost trivial that you've just added a couple seconds I don't know that they're rushing through it the amount of burden it takes to answer a question to which they know it's substantially lower than oh I have to go look that up or I have to count that or I have to get the right rule to make the answer and that's what makes this so powerful just minor tweaks can result in significant behavioral changes anyway thank you so much all three of you once again it's discouraging to discover just how much the architecture of our choices informs our decisions but it is encouraging that there are folks like you actually ensuring that we make decisions that we're proud of it seems like we probably have some questions now for our panel now seems like a good time to open it up to the floor can I just mention one more thing is so you know we're trying to use choice architecture for good sadly there's some terrible examples out of people using it for bad maybe the worst one is a company called wanga I don't know if any of you know it's this company they're not allowed to they're a payday lender they're not allowed to work in the US because they have higher rates than our own payday payday lending so that's bad but what they do which is I admire their creativity and I hate the business model but it's an iPhone app in which you ask for money you have two sliders how much money do you want and for how long do you want it and what they've learned is that people who push the sliders all the way to the right very quickly are really bad credit risk so they just give them the money but if you think about this as an idea it basically says that we can track people's thinking process in the online environment in a way that could tell us something about people's intentions so they're using it for evil but I think there might be some interesting ways to think about this David Martzall with Center for Economic Progress simple question but I'm really interested between Mikhail and yourself and David at what point does this experiment and we're now two or three years into it become universalized within TurboTax and what are the limitations in doing that and maybe that's a David question to ask but you can tell him what to do but I'd just be curious because of that 30 million people that's a huge number of people to move the scale needle from where you've started so it's a great question there are a couple things I'd say one is we haven't really nailed down we continue to look for ways to be more effective at this and so that is part of what's coming next I think the real focus for us and you're going to hear about this from Melissa I don't want to steal too much of her conversation is can we actually trigger folks to use savings vehicles that are I won't say ideal but I would say MyRA is a pretty impressive product for those of you who aren't aware of it you'll hear more about it but the challenge we have now is that in the last several years we've incorporated choices about what one can do with one's refund where's Tim Flocki we've added at the behest of Fred Goldberg and a whole coalition of people that leads us to Doorway to Dreams we've added the ability to split a refund we've also added the ability to save US savings bonds it turns out the take up rates on those are remarkably low unfortunately and so the question that we're actually now starting to explore is with the right savings vehicle or with a savings vehicle can we actually move people just saying they're going to save but to a vehicle that actually has some of the features of having to drive to another village and wait an hour in line but perhaps at least creates a framework and an architecture that includes a savings vehicle that isn't a cash account and so that for us is the next frontier you can imagine for the financial folks in the room their interest in opening financial savings products of their own but we are focused on ways in which we can help stimulate that so more to come I'm not sure when we will scale it but I think when we have I don't believe there's a silver bullet but when we have something that we think is ready to scale across not just us but across industry I think that's when we're going to pull that trigger there's another interesting thing when in the TurboTax process have we passed the right mindset so if you use TurboTax and the question is at what point do people basically said okay this is the money I'm expecting and I already have plans for it so we've been talking about when is the right time to do it and right now it's a refund step but should we do it much earlier should we do it first thing when you go into TurboTax or at least get people to think about savings all around or you have this little counter on the side that you expected to have as a refund should we start by having from the beginning this is how much money you have going to savings and to as a refund all along there's lots of ways to try and make this intervention I think more powerful and we're trying to work on some of those Morningstar's HelloWallet division I wanted to ask if you'd given any thought to the feasibility of reducing people's refunds by getting them to withhold less money and subsequent years and then doing an intervention at that point because it seems to me that the large refund might just be too much of a temptation for people but if they increase their paychecks by better aligning their withholdings with their actual tax burden and then redirected that money to MIRA or another kind of savings account that might be a good form of pre-commitment I know that's more difficult for the EIC population than people because of the refundable nature of it but I wasn't just wondering if you'd given any thought to those kind of interventions so so it's interesting when we started working with Intuit there was lots of engineers at Intuit that were trying to get people to have zero refund they thought this is the right answer and people basically resist resist very strongly this refund actually in the first year when we did the household survey even before we did anything at TurboTax we just asked as a survey questions people particularly low income individuals really wanted self-control mechanisms they really wanted something out of their hands now does it have to go to the government can it go in some other ways I think it can go some other ways we just don't have this mechanism yet but if anything I think we should until we have a good mechanism that people could deduct money and can't approach I think we should increase deductions so that people have more money at the end of the year let me make just two comments about that one I won't have a quiz this time but I'm willing to bet that a substantially large number of people in this room are hoping they'll get at least a refund I mean in this room with all of you who've had microeconomics and I'm guessing some finance you still want a little bit of a refund and that is behavior we saw at the IRS all the time more specifically the earned income tax credit used to offer an advanced feature now there's a debate about and in fact I was just at a conference with David and others about how you could resurrect that because it was killed several years ago and I would tell you having looked at that for a long period of time there were a number of administrative problems with it not the least of which was the burden of administering it went on the employer but psychologically all of the people who might have been eligible resisted because of things we've just talked about volatility of income and volatility of family structure meaning uncertainty about the outcomes and so I think to Dan's point unsurprising that large numbers of people particularly the low income said I want to know that I'm getting a refund and so I think that's probably not where we're going to go immediately Thank you so much David I think that that brings our panel to a close but I want to thank Dan Ariely, David Williams and the Hall of Greenstein Weiss thank you so much for walking us through refund to savings and for informing us of ways to compensate for the ways that were predictably irrational I guess Thank you very much