 you guys for being here. As you know, on the Senate calendar, we have S169 last year, which had been together with the governor. S169 included a number of revisions, one of which was a 24-hour waiting period for handgun purchases. At the time last year when the bill went through the Senate and the House, the data or research, the basis for the decision about 24 hours, while some people felt that there was a logic that would help save lives, others debated whether there was enough research to know whether it was recently or not. So as we contemplate what to do about S169 or how alternative paths to move forward, we thought it would make sense to have some professionals who are independent, who are viewed as effectively the top researchers on the data around waiting periods to come in and shed the best thinking and research of the day on the issues that the committees can make a more informed discussion. I'm very pleased to present Dr. Adluka from Harvard Business School, who wrote what is essentially the premier paper. I'll say it so they don't have to. On this issue, we're very generous to come up. They aren't being funded by any group or they can speak for themselves about the nature of their research and the impetus for it and hope that they'll be able to shed light on what they believe the research is on the safety impacts of waiting periods and let their data speak for itself and obviously that people ask as many questions. So I want to thank them though for making the haul up this morning for Boston and look forward to hearing what they have to say. Thank you for being here, Senator Fick Sears and Jenny Alliance. We are chairs of the Judiciary and Health Advisory Committees and thank you Senator Rash for arranging this testimony. We hear anything to do with firearms as conflicting evidence and happy to have some actual studies of waiting periods. The best we had when we were developing drafting the bill was a study out of the University of Houston that indicated there was some benefit but two really too few people in terms of both suicide and homicide. Please feel free to get started. Thank you Senators. It's a pleasure to be here with all of you. I'm just going to spend a few minutes up front giving a little bit of a background on who we are and why we're doing this work and generally what the data seems to suggest and then my colleague Mike will then spend a little bit more time on the actual study to give you a little bit more information and hopefully a lot more confidence on the rear with which the study was done. And as you think about any policy that you want to pass, certainly one of the questions that's going to be, boys are going to make any difference and our hope is that if there's one objective that we achieve with the time that we have together, it's that you'll have a lot of confidence that if this policy passes, if the objective is to reduce gun deaths, that that would be something that would be something that you can very much expect. Let me start by just saying again, my name is Deepak Mahotra. I've been at Harvard in faculty since 2002. So this is my 18th year. I'm an academic scholar and I work in a lot of different areas and I'm more into literature. Thank you. So I've been at Harvard since 2002. This event's a lot better. This is my 18th year and Mike Luca, he's been at Harvard Business School since 2011 and we've been colleagues and we've been doing this work together since a little after 2012 and what got us into the area of gun violence in general was what happened in Sandy Oak. When the shootings in Sandy Oak happened, neither one of us was studying the issue. The two of us had a meeting for lunch a couple of days after and everybody in the country was talking about what happened to those children and I know I have little kids and Mike does now and he did at the time. And after talking about the tragedy for about an hour, we were about to just go back to our offices and get back to work and one of us said something like, well, so what are we going to do about that? So what are we going to do about that? Just go back to work, right? And then we sort of paused and we thought, well, what could we do about it? And we didn't care to say maybe we should donate somebody some cause and we quickly assessed our assets and we realized that the one thing we do have going for ourselves is that we're academics, we know how to do research and we have research budgets that we're allowed to use however we think is important. So maybe we could spend some time just trying to understand what kind of models. We know we ended up whatsoever rather than let's just at least educate ourselves on this issue that it seems that we still often people have a lot of different theories, a lot of different models for why things work the way they do, different interests about how to solve these things. Let's just try to understand it. We spent the first couple of years just scouring the research and understanding what's we've done and hasn't been done, what's we propose, where there's evidence, where there's no evidence. And eventually we decided to look at the question of what might actually solve some of these problems, not totally, but what might make a difference. And we don't have any ties to any lobbying groups, we don't have any ties to any special interest groups. The research is not funded by any outside group. We started looking at waiting periods and the reason we started looking at waiting periods is because there's two questions, at least from our perspective, that always get asked when people think about whether or not any legislation should be passed in particular with regards to gun violence. And the two questions is one, is there any reason to leave this to work? Why pass legislation just to pass legislation? Why pass a law just to put one more thing on the books which will do nothing? So the first question is going to be, can this make a difference? And there was already research out there that suggested that theoretically waiting periods might make a difference. And Microsoft is a little more about why that is. What happens in the short amount of time when somebody is either feeling very depressed or very angry and all the kinds of decisions they might make at that time. The second question, which is not necessarily a question scientists always find themselves asking, but I think it's important for everybody in this room, which is even if we could show that there is an impact, that we could save lives by having something like a waiting period, what about the idea that there's segmental parties that we need to protect and we should be taking guns away from folks, etc. And what waiting periods do is that, as we'll show, they can save lives and we can show that empirically while respecting the Second Amendment rights of people by taking guns away from a person without any more restrictions on people's right to own a country. So that made it a compelling thing for us to study. And it turns out that the results are pretty strong. We'll get into those in just a moment. But in addition, the majority of Americans and the same pattern holds in Vermont from the data to the little bit of what we've seen that's on it. Majority of Democrats, majority of Republicans, majority of gun owners, majority of non-gun owners support the idea of waiting periods probably for the same reasons that I just articulated, which is that by using just a brief delay between the decision to purchase a gun and then having that gun delivered to you or picked up by you, we can reduce homicides and suicides. And notably, the polling data that's been out there is before we even had the strong evidence that we're going to share with you. So even if intuitively people have thought that this could be a good idea, but now we're at a point where we can make a much more compelling case that says, well, your intuition, if this was your intuition, happens to be quite accurate and so with that, I'm going to hand this over to Mike and he's just going to go through some slides that provides a little bit more context onto what we're studying and then digs in a little bit more into the actual research. And from our point of view, we welcome any and all questions. Absolutely anything that you want to ask about, you should feel free. Don't censor yourself in any way or pattern you have. If we can't answer your question, we'll of course let you know that. But our hope is that by the time we're done in this session, there's no question that's still left in your mind about the results, about the empirical evidence, about what about this or what about this, anything like that. We hope that the time we have together will help resolve all those issues. And with that, I'll hand it over to Mike and then soon after that we'll go to the questions. So the research that we're going to be talking about today comes from a recent paper that was published in PNAS. The papers on the impact of and about waiting periods on gun homicide and on gun suicides. We'll just start with a little bit of background on this. As Debak was mentioning, when we started thinking about this issue, it became clear that gun deaths are a major issue in the United States. So according to the CDC, there are about 40,000 gun deaths across the U.S. in 2017. There's a mix of outsides and suicides that are in account for this number. So about two-thirds of gun deaths are suicides that we did plot out. We're going to power the moms in comparison to the rest of the United States in terms of suicides and all of a sudden the suicide rate is kind of on the higher side of this, which I suspect is part of what prompted this bill to begin with. We'll walk a little bit through the rationale for a waiting period. So as Debak had mentioned, when we went into this study, we didn't just want any type of gun policy. We were thinking about things that we thought might potentially be our support, but also things that we thought private research suggests might have an impact on behavior. So to give a sense of what we have in the back of our mind to motivate the study of this, there's growing research within psychology and behavioral economics on the nature of emotions and in particular the fact that emotions can be transient. So you can feel the rest now, but not 36 hours of marrow, 48 hours of marrow. You can feel angry and ready to commit violence, and that can go away in retirement. So there are some references in here that are also in the paper that was distributed if people want to read more about this. Now, once we start thinking about the transient component of emotional states, you can imagine that violence or self-harm has potential to be more likely in hot states and kind of outside of the context of gun violence. There's been some research on this, for example, looking at domestic violence in hot states versus cold states. Now, thinking about how as a policymaker you might design a policy about this, you can imagine that even when they go a little bit of a waiting time, a little bit of a period to allow somebody to cool off and transition from a hot state to a cold state could have a persistent impact on how they're all being of people. So thinking of this, we started looking at the landscape of waiting periods. And thinking about waiting periods as something that has the potential to just kind of reduce your access to a firearm at the moment when you're in a hot state and still kind of allow you to get this 48, 24, 36 hours at once. So we wanted to see what the data would say about the causal effect of waiting periods on gun deaths. Because even though we had seen research suggests this might be effective, we wanted to get direct evidence on the policies that have been implemented over time that have implemented waiting periods. And what we realized is that many states have already implemented or removed waiting periods over time. You see that for one reason or another, there are 44 states that have had evidence of waiting periods implemented over the last few decades. Now some of these were changed intentionally. Thinking about implementing a waiting period and some had a waiting period that was put into place to support something else like to provide a time for background checks or other things. And what we wanted to do in our analysis is to take advantage of all the data we could find on changes in waiting periods to get a sense of do waiting periods impact gun deaths. So we took this and used a couple of different methods that are common in the toolkit that economists have developed in recent decades for trying to get a sense of whether an estimate is caused more or not. So I'll just walk a little bit through the methodology to give a little bit of sense of what's going on underneath the hood in our paper and what we were looking at in the data. Of course, when thinking about that, if you were just to look at states that have waiting periods that compare cross-sectionally to states that don't, you might say, well, there are other things that are different across those states. So one way to get around this is our first approach, which is called a difference in differences strategy. So essentially what we wanted to do is find the timing on which different states either implemented or removed a waiting period. And then essentially what we look at is either looking for jumps or drops in gun deaths in the states that had a change in their waiting period policy relative to the states that had no changes. So notice what that allows us to do is we're no longer just kind of comparing cross-sectionally across states, but now we're kind of looking for the discrete changes that happen after each date has implemented this. And essentially what we're doing is a variant of taking the average of factor across the states out of the implementation of that policy and controlling for anything else that might have been happening over time. In this first strategy, we look at data from 1970 to 2014 and all 44 of the states that have been implementing waiting periods for this. Beyond controlling for the specific state and for the specific time, what we do is we put in a host of control variables for other things that might have been changing, such as demographics at the state level. That might have also been affecting the number of gun deaths to convince ourselves that this is a causal estimate that we're picking up on. So often what those results are, that's sort of like the first basic approach that we took. That is it. Well, one possible concern with a difference in a different strategy is that this may have been adopted at a time when you are looking to do other things as well. So maybe we're not just picking up on waiting periods, but there's always the potential that something else is going on. And we did our best to control for all of the other main things we think might have been going on, but we wanted to take a second approach to sort of cross-validate our results for our main estimation strategy. I just want to clarify exactly why we need the second strategy to be just doubly short. Somebody could argue that isn't it possible that a state passed a waiting period law right when gun deaths were going to go down anyway? There's something else going on across these 45 years which is correlating with the decision specifically for waiting period law, not other laws that we might have looked at, but waiting period laws just happen to show up right now for completely other reasons. A state wasn't going to see a decline in suicides or homicides. Now, to solve that problem, you know, some people might say, well, that's a little bit hard to actually have not been composed yet, but let's do it seriously. It could be true. Ideally, what you want to do in a situation like that is you want to experiment, force some states to take a waiting period and not force other states to take a waiting period, and so it happens. And what you're essentially doing is saying nobody chose to take a waiting period. Experimentally, they were just forced to do it from the outside, but that doesn't usually happen. You only actually want experiments with states in our country. But it turned out that in this case, an experiment was essentially run where from the outside, some states were forced to have a waiting period, even those that would not have wanted it themselves. It's what we call a natural experiment. Nobody designed it to run an experiment, but naturally that's what happened in the 1990s because of the rate yet, and that's what we're going to get into, which is what happens when nobody chose a waiting period, but it was implemented from the outside upon them. Do you still see that same effect? So it's not a choice correlation, right? So as Deepak was suggesting, essentially, nobody were looking for things that could help to approximate an experiment. So essentially something that's going to give us a plausibly exogenous variation in the implementation of a waiting period in some states, but not in other states. And the Brady Act did a version of this, and let me kind of walk through a couple of the details of it. So the Brady Act had to focus on background checks, but while implementing the background checks, they also posed a five-day waiting period for some number of years during the process. So this is during the Brady Interim period, while they were sort of getting against their check system in place. Now the thing that allowed us to get an extra variation on this is the fact that not every state was subject to the new waiting periods that were being imposed. So some states already had a waiting period, so they would have been exempt from it. Some states were exempt altogether, because they already had a background check system, that method for a period. So that gave us two sets of states that weren't getting this new waiting period and a bunch of other states that were getting in your waiting period. And all of this was sort of just imposed from this federal act, so not treated by other state-specific things that might have been passed at the same time. So what we did is then we run that analysis for about a 10-year span to sort of tightly focus on the period before and after the changes in those states to cross-validate against our other results, looking at the data from 1970 to 2014. And again, what we're going to be looking for there is also a version of the difference in differences, but exploiting this natural experiment. We're going to look at the change in gun deaths in states that were subject to the new waiting period relative to the change in gun deaths that were not subject to the new waiting period that was imposed on other states. And once you have the data and these two different identification strategies, we are allowed to get a better sense of what's the causal impact of putting our waiting period into place. And again, so what we found is ultimately there are very similar results across the two different sets of analyses, and across them we found that the waiting period seemed to reduce gun homicide by about 70 percent. And this is very controlling for other things, and also using this natural experiment created by the brilliant. Then we found that while the estimates kind of were wide over, and you were looking at suicides, the data suggests that waiting periods reduced gun suicides by about 7 to 11 percent. Now, thinking about this, in cross-validities we're controlling for other factors like background checks, other gun policies, demographic trends. If you look into the paper, you'll see we have a number of other falsification exercises to kind of convince ourselves that this is actually waiting periods that we're picking up on, and not some nails. So what we have this is kind of convinced us that in that waiting period seemed to have a causal effect on the number of guns at the state level. And then finally, in the paper, we wanted to do a little bit of a calibration to say, okay, so this is having an effect, but kind of how big is this effect? So is this a meaningful number to be picking up, or is this just kind of a statistical artifact? And I think kind of looking at the percent change already comes to the fact that this is a big impact on the total number of gun deaths, but we can also think about this in terms of the number of lives saved if it's were to be implemented nationwide or in a given state. So currently 15 states have waiting periods. Then to give a sense of the magnitude here, if all states were to implement a waiting period, our data suggests that this would save an additional 900 homicide and an additional 950 suicides per year. So I'm just going to stop there, and this kind of is what's contained in the paper, and I'm pointing out that everybody has a draft of the paper, and we also have the one-page summary that we have distributed as well. Just to verify one thing, 15 states that have waiting periods, there's two categories of those. Some states specifically have a waiting period, not unlike what you've been considering. Other states have a de facto waiting period because they have some other law in place that creates a waiting time before something starts, there's a permanent system, for example, and that takes x number of days, so that essentially creates a waiting period, even while those can act as a waiting period because the mechanism underlying this we suspect is that cooling off period. So you may have a specific waiting period, you may have some other, I think that acts like a waiting period in some states, and any questions there? I have a couple, others may, if it's okay. Absolutely. My first question is, is there any magic to the number of days for a waiting period? We in our bill, the Governor vetoed, we had 24 hours, and that was a compromise between members of the Senate Judiciary Committee in developing a bill originally called for 48. There's other bills that call for 72 that are staged with much longer waiting periods. Is there any magic to the number of days? So I can tell you what we know and what we don't know on this. In the range, we could only study the things that have been tried. The range was typically from a low of about three days to a high of about 10 days. Now, we didn't find any difference across that range. So the one short answer, the complete answer would be to say the lengthier ones aren't having a better impact on the shorter ones. The more complete answer would be to say but that's clearly true in the range that we saw. But we can necessarily extrapolate beyond that. So when you have to sort of think about well, what's the underlying mechanism? If the ideas cool off, is 72 hours significantly more than 48 hours or more than those, for example, enough to cool off? So that's what the data shows. I can tell you, sort of anecdotally, the US House, there's a bill now based on this research in the US House of Representatives. They decided to go with 72 hour one. They had some same questions, which is, is there a magic number or like a minimum number that you need? And we can't answer it any more than to say we didn't see any difference across the lens, but there's only a certain range that we could have evaluated because those are the things that were in the data. Second, yeah, I've got another second question that you started out with a majority of Americans support waiting periods. Where there would be gun owners, you listed a whole group. And what I heard from a lot of, and I'm generally not supported bills that are like the one that I did vote for, that the Governor vetoed. I did that because I believe that would make a difference, but many of my constituents said to me, but this, you know, you may think this is okay, but what you just started was just looking pretty slow. And once you do this, then you get the better or worse yelling, yes, we're going to take your guns away. And that was kind of a reaction from many of those folks who had a real fear of what was happening. Is there any, did you get into any other gun legislation? For example, Bands On, Weapons, Bands On, Magazine, Clips, that sort of thing. The waiting period. Let me say a couple things about this. First, this is a problem that we all face in society. Which is this whether it's legitimate or not, this mistrust that exists, which is to say find the will to do certain things, but how do we know you're not going to take this and then just do a lot of other things that are maybe not right, maybe that are not supported apparently, maybe. And yes, this is a problem that actually this is exactly what society need to confront and I think you confront every day. As scientists we don't have an answer to that. We can only tell you what would save lives. But the idea that your constituents fairly or unfairly legitimately or for other reasons you believe that the downstream risk is that even though we did something really good today, may lead you to one day do something worse, is something that we can't solve with this kind of research. And that is a problem that we need to all confront, is to say does that mean we do nothing? And that is squarely in your world in less than hours. I have a view on that, I think we should do what we can but I have a knowledge that some people have legitimate fears that they would also need to be addressed and that trust does need to be built. And you do need to sort of say you know what, we're going to take a stand and really focus on the thing where we have either very strong empirical evidence or very good theory that suggests this should work but we're not going to just willy-nilly throw out ideas. But I'm thinking more for me and maybe for us but not for some, this is just a problem in society. Did we look at, no, what we started with was to what we control for other laws to make sure that there's no effect of something else that we're picking up. But the reason we went through waiting periods is because there was theory to suggest behavioral science, psychology suggested that something like this could make a difference and that lets you think about whether this would actually work out or not. There's people out there who do research sometimes they publish it in academic journals, often they just put it out there. And what you'll find is that every study they ever do points in the same direction. Every study they will do will either show there should be no restrictions on guns or there should all be restrictions on guns. In the scientific community we don't consider that science. If you have two or three studies in a world that happen to show one thing, maybe that happens to be what you can focus on. But if everything that you say is in one direction, you may be a perfectly good person and you mean well and you really believe this is the case but at some point you're losing credibility as a scientist. Our primary identity is as academics and scientists. If we were to find a gun policy that we studied that actually we could show invariably would make no difference we would be just as motivated to put that out there. And the reason is not just a statistic, going back to your initial question, that's what allows us to build that trust in society. That when we think something works, we put it out there when we think something doesn't work, we put that out there. The other project we're thinking about working on is to scour the earth for all the best research and let people know we're interested in maybe doing something. Which laws seem to help? And which laws interned? Which gun restrictions seem to actually save lives? Which ones there's no data on but might? Which ones seem not to work? Which rules seem to save lives? And which would not? And so your question is a question that we've heard a lot. We don't have great answers but we have a great empathy for it. But I think when you talk to this waiting period which is extremely helpful in terms of art, my final question really is, oh I'm sorry. So I'll just add one more thing to that, which is a little bit of extra complex. People say, oh how many new gun bills are going to be proposed or enacted? So we have looked at some data on this and actually there have been a lot of gun bills that are proposed and enacted over time and we've looked across states. There are thousands of proposals that have come out across the US over the last 40 years or so. But the thing that we started thinking about which is part of our motivation in this project is that we can't be more systematic about thinking about what the evidence says about which of these are likely to be effective. So rather than just kind of saying there are 10 things that people can try to focus on if you were applying two or three of the most evidence and sort of triangulate on that part of the world, that sort of well let us down the path of getting interested in waiting periods. Final question for me is, there were figure 900 homicides, 950 suicides, additional savings of lives. In Vermont our studies indicate roughly 90% of the firearm deaths are suicide versus 10% murder and most of those murders are intimate partners or people well-owned to each other. They're not necessarily focused at which indication might be impulsive by the way. But we luckily have a very low murder rate but a very high suicide rate and I'm wondering if it would if there's any indication one way or the other regarding those numbers what Vermont might expect that we would institute a waiting period. Well the best we could say is we need to look at it nationally first if it's 17% reduction in gun homicides and less than that almost let's say roughly half of that in terms of suicides at first cut one might think well okay so the reduction in homicides is going to be greater but because more deaths are suicides that 7 to 11% actually adds up to at least as many as the homicides if not more and I suspect something similar would be expected in Vermont which is to say if the number is 7 to 11% in Vermont that's going to come for a lot of those deaths. Now I mean that still leaves 90% so this is not a magic wand this is not a get rid of everything but it's not just the individual impact every one of those deaths has an impact on the entire state so the impact is the number per year multiplied by all of the impact it has on other folks but we can't be precise about the exact number it would be in any one year in one state this is the estimate. Thank you very much Senator White and then Senator Booth I just had a question if you could give more examples maybe of the de facto waiting period if you had some or as in which states have those or is it just one example you pointed out is permitting that's how I think it would be if you have a permitting system for handguns that ends up requiring some time so I think back in the day the idea of meeting a background check before there was the reason the Brady Act even came in was because these background checks were not a thing and the only reason the Brady Act even imposed waiting periods was because they didn't make sure there was time to do a background check so for a lot of background check was another example of exactly what you're asking for but I think those were probably in the tube main now we have some background checks in many places so that was a reason some people removed their waiting periods so maybe I shouldn't speak to this there was a lot of focus on background checks in the 90s and it had an impact on legislation etc but when an instant background checks became more common states then removed their waiting period and originally put them in to facilitate the background check based on reading and what our researchers suggested they were having their own independent effect that you now lost by removing them they weren't just helping do a background check even if you have a background check on top of that a waiting period actually saves incremental life so thank you for coming first of all I was interested to hear that your work started after standing hook my interest in the subject spiked to I think the same is true for many people around the room it's striking to me that when these incidents happen and then nationally there is no legislation passed we tend to say nothing happened that time nothing happened again this time nothing happened this time but there has been a change and your research is a really crucial part of that especially since the federal government has been prohibitive research in many ways or inhibitive I did want to ask the most surprising thing about your findings for me was the 17% reduction in gun and homicide I hadn't been expecting that there would be an almost double effect on homicide because I knew that suicides were a much larger component of your world deaths do you have any personal theories about why waiting periods are so much more effective in preventing the homicide as opposed to because as you say a cooling off period emotions change over that period of time obviously radically different sense of emotions but any thoughts on the differences between the percentage of the bank as is confirmed that we're just speculating at this point if you have any theories if you ask an academic if they have any theories we can take out the entire whole bunch of theories many have tested it so I didn't have any I've known this data for so long it's so hard to reflect back before you knew the outcome what your expectation was I don't think we have any priors that were strong one way or the other somebody's angry somebody's angry at someone else they're able to obtain a firearm in short order the opportunity and the anger is still there that they might use it when you're depressed and everybody's been at some little point but think about being at that point where you're willing to anger like that level of hope might happen and you're able to fight a firearm that's a different notion could it be that the person who is in a depressed state eventually gets to that same depressed state again in some future time and so maybe you stop them once or they don't do it but maybe the next time it happens they're still able to act on it maybe that's different than what happens in a very very important moment in terms of the likelihood that they will be just as angry and willing to take those actions maybe there's a difference between those two kinds of things but that level of mechanism we have frameworks that don't have that kind of data so we really speculate about that if you want to add something to that yeah, so I agree with that we don't really have the data to say and I share a bit of the intuition not going into it we didn't have strong buyers about where we'd see the larger effect I guess another component that made us have a big impact would be going in because other things are changing in homicide there may be a shorter time frame during which you have the situational opportunity to commit the homicide that the woman is looking to convey so there are other things going on that make it harder for us to sort of defer what Lisa brought up with the validation we did also is thinking about the mechanism and I mentioned this before there are other studies for instance looking at domestic violence that have found and that trans-storied emotions make a difference and that's kind of like say a level of violence so for example that we'll look at are the jams being from all games and not expecting wins and losses is the type of exercise people kind of look at and see spikes in violence so it didn't surprise us per se that we saw kind of an impact on actions toward others as well the one thing that we'll say is sort of implied earlier which is at the end of the day it's about 10% of the suicides and for maybe the reason that we were just regulating but more generally this isn't the only answer there more needs to be done to help with this problem and that never means don't do anything that you can do but this is not to say all right we'll check the box on suicides or even guns suicides specifically but hopefully this is something that saves very specific individual lives I have a question so and I apologize I haven't had an opportunity to read full of publication but as you were going through and thinking about the other variables that contribute to suicide and trying to eliminate some of those variables so it was only the waiting period can you talk a little bit about what you discovered, what you eliminated how you eliminated and maybe there are some things that are outstanding that a point is in a new direction but not to say that the waiting period isn't the right direction at this point but some of the other maybe social or emotional contributors to suicide and homicide that you found so one kind of dis-clarification point that I would make about the empirical approach that we took is that our approach wasn't to kind of say let's kind of have one by one control for other things that we can do but more defined as something that led to the implementation of a waiting period that was unlikely to change all these other things that were going on so I'm thinking about the Brady Act example finding the states where we could kind of narrow in on the couple of years before the couple of years after and sort of start to see this Brady and Poe's waiting period was associated with the sudden changes in the number of suicides so kind of looking for a clausically exogenous variation in the implementation of a waiting period and you say okay it's much less likely that kind of demographic changes or so far emotional conditions are changing in a way that will be correlated with states that were differentially impacted by the Brady Act so in a sense it's kind of looking for these changes that were almost as good as an experiment in terms of states that got it and then sort of looking at the just before and just after years to identify what the effect was so I would think that imagine if you wanted to figure out what you should know about safety belt law one approach to thinking about studying safety belts would be to say well let's compare across 30 that can save lives in a car accident and see where sort of safety belts rank on that and other things like you know a little better signage on the screen and if somebody was interested in saying would safety belts save lives they would design a study kind of like we did that would say okay we know safety belts are less likely to die than people that don't but how do we know it's not that they're just safer drivers we tend to also wear safety belts so let's control for their safety record in accidents well how do we know that maybe wrong people are the ones that aren't the ones that are usually so let's control for that so you shouldn't change the way that other explanations that were really driving the saving of lives and making you confused thinking it was safety belts when you cross out all those other alternatives you say oh actually yes drinking driving is bad and yes being a bad driver is bad but in top of all of that safety belt saved lives okay now if you're confident that that might be worth pursuing that's how we did this one a different study with probably different data you would probably need different data would be to say okay well what other things cause somebody to commit suicide right well what other things might be good interventions really at that moment of despair but that's not what we did because it wasn't like controlling for something that would affect the waiting period implementation of a law so it would be a different kind of study that we didn't quite do so we don't have but there are there's experts out there whose job is to look at those factors right just to say what drives them to make that decision what kinds of interventions help how do you how do we do those things that just doesn't have to be us thank you I have two questions for you actually one of the things that we hear frequently is that the measures that we take to try to save lives also you know in infringing on the rights of gun owners who buy guns for some defense or for sporting purposes were there any is there any way that from your data that you can tell if sales for those purposes were ever say affected by this waiting period so our study doesn't look at sales but we were discussing this earlier there is another study and we can get to that study by Professor Glazier they looked at different data that wasn't the same kind of analysis we did but they used a different kind of survey data and I believe their research showed that implementation of waiting periods do not reduce gun ownership now is gun ownership the perfect measure of any particular sale at a particular moment I don't think you can get that granular we can't be attending one specific incident the second thing I'll say about what you've mentioned is hunting some of these are handgun waiting periods is what we looked at and a lot of them think it's not with handguns so that is usually carved out in some ways but that doesn't mean that those aren't equally relevant it's just that's what we looked at and the last thing I'll say about this is that the self-defense element of what you asked is it possible that there's someone out there who doesn't already want a gun wanted to buy a gun who was feeling threatened I think they had been able to get a gun within that same hour or whenever they went wouldn't then have been able to protect themselves because somebody was going to attack them that same day and they would have had that gun at that time and they would have been able to use it in self-defense so you don't quantify those probabilities but is it still possible of course it's possible but we don't have very many anecdotes of those being the problem but the nice thing about this is that they already met all of that out if that wasn't happening at any high level we wouldn't find results of saving lives we'd see that waiting period increase homicides if there might be there's probably a situation where if only a person wasn't wearing a safety belt they would have survived because if anything else ran out of a safety belt or the people that came to save their lives and couldn't get them out of it there's always individual incidents that we can't pick up in fact it seems to clearly be in the other direction that's why we're finding the impact on homicides going down rather than up and suicides going down the bill in question was S169 it only applied to handguns not rifles so following up what you just said about handguns is there any evidence of your study regarding waiting periods divided not all of the firearms so maybe some percent I think say for example homicides they're done by handguns I think a high percentage of suicides have the same so what we studied actually suggests that the percentages we've told you should relate to the bill you have but we were also talking about handguns now if somebody wanted to also add other guns percentage could only go up or down it may not go up we don't know we didn't study what happens if you you know address all of that address all of that or anything but we did look at that and it looks like you're not looking at that so the percentages shouldn't be that high I'm sorry is there any way for did you break down the reduction in suicides and homicides by factors like age and gender race were there any particular cohorts that reduced these acts more than others so we control for those I think all of the things you just mentioned race, age so we have that in there to see if we can create something I don't think we picked anything up but one of the things that happens is when you start losing degrees of freedom the statistic of power to be able to look at it goes down the more you try to cut the data into smaller components because you don't have enough data in each category the data that we have often was at the level of state demographics rather than each individual so we aren't looking at the likelihood that somebody within the age range of this commits suicides so we don't have that in this analysis that would be an additional analysis this controls for well maybe the states that have weighting periods tend to have a younger population where they have a different racial demographic where they have a different economic situation so this controls to make sure we're not erroneously again erroneously suggesting what weighting periods are doing to work differently so that's what this one does but we aren't able to get into which deaths specifically within suicides or on the sides for the ones that can be used and which by the way I think is a very interesting and important question especially as you think of additional ways that maybe have nothing to do with legislation for solving crime or solving suicides Other questions? Senator Bennett Good morning, thank you for coming one particular note your data was talking about three to ten day weighting periods our bill is calling for 24 hours assuming the cooling off period is the underlying driver of the conversation would it be fair then to say that the lower weighting period of 24 hours would not produce the same results as those within the three to ten day window you had I think what we could say with more certainty is that a shorter weighting period what would be better than the lower weighting period because if the logic is what we think it is but it doesn't necessarily mean it would be worse so if the weighting period was five minutes maybe we would all get to it that seems too short and a weighting period being ten days instead of three didn't seem to make any additional benefit but where exactly you draw the line is not entirely wrong now I think if we look at other research it might help us to adjudicate this so I think there's a lot of research on suicides and you can go on with to look it up or ask us to look it up I guess which suggests that when people decide to make a decision to commit suicide they usually have to live the day so if there's other auxiliary research it might help us think about ok well maybe delay it by a day or two days might be sufficient but the only thing we can say with confidence is what's in our data and then we have to make some educated guesses about where the line should or shouldn't be drawn but did I get another question that you're asking? I didn't assuming the theory works that the cooling off period and the longer the cooling off period is the greater the possibility of addressing problems that I understand but you have not actually applied your research to a 24 hour waiting period in addition to Vermont's actual numbers of suicides and homicides well we haven't applied it to nobody has one for our period before so that wasn't in the data so we couldn't apply it in regards to Vermont but we looked at it ok that's another question just oh with respect to those waiting periods of three to ten days that you were analyzing did you also take into account any other programming that might have been developed in that period of time for instance zero suicide I don't know if you know anything about that program but were there any analysis that was conducted with respect to programming co-joined with Sheryl's waiting period so two answers to that I don't know if that's appropriate but two answers to your question one we did put in other policies that got in legislation that may have because often people don't just propose one got policies that impose multiple so we added other policies to make sure it wasn't something else that's driving the effect unfortunately we're not looking at those in normal waiting periods we've done that for exactly that reason which is we can't possibly know every other thing a state was thinking, doing going through implementing which we thought we had approached number two that was precisely the reason we had the Brady Act natural experiment because the Brady Act suddenly showing up and as myself is not to consider from the outside forcing a state cap waiting period that they weren't planning to have it eliminates that correlation that people that were waiting periods were also doing a bunch of these other things at the same time the beauty of that second approach the need for that second approach is exactly to address the kinds of questions that you're raising now to hear fantastic questions and to often research misses those pieces that they couldn't just get other things that are happening alongside that you didn't pick up so that's why you wanted an experiment if you can't have an experiment sometimes you get lucky in the world where there is an experiment for you and that's what we exploited in our research frankly thank you so much for the track great day soon you draw we draw it was a beautiful project it was a beautiful project he was complaining about the company he had but you've probably been in each other's company for a long time I'm not exactly sure fascinated by that by how you started this conversation and there is another gentleman who is part of your team who is obviously not here from UCLA and his involvement in this is similar so Chris by the way I could not say enough great things about Chris as a scholar and as a human being he was a PhD student at the time he started this work at Harvard and he really helped us go from the transition we made from trying to understand before we just jump in trying to do what I'm saying really understanding the layout when we were transitioning to actually doing the academic work that led to this paper Chris was a huge part of the team he was the one who really helped us understand what previous things had maybe not been in the research well and how to code the data better and so he is one of the people a partner I would say just as an aside we have a lot of these papers that get published in the economics world we often have to just alphabetically name the authors so just as somebody shows up last and doesn't show up in a room like this doesn't mean they're not in one of the great circles Chris Faulkner he's great in fact we discussed with him before we came here today what we're interviewing and got his thoughts and any things that he thought that we should assess as well thank you so much we're going to take a break unless there's other questions we're going to take a break until 1130 and meet in the committee room what you want to do well thank you senate health and welfare will return to room 17 right now and we'll probably adjourn by 1115