WEBVTT 1 00:00:02.910 --> 00:00:07.289 Abdullah Dogan: Welcome to the Tobacco Online Policy Seminar, or TAPS. 2 00:00:07.350 --> 00:00:29.930 Abdullah Dogan: Thank you for joining us today. I am Abdullah Dawan, a PhD student in Health Economics at the University of Missouri. Tubs is organized by Mike Pesko at University of Missouri, C. Shank at the Ohio State University, Michael Darden at Johns Hopkins University, Jamie Hartman Moss at University of Massachusetts Amherst. 3 00:00:29.930 --> 00:00:32.920 Abdullah Dogan: and Justin White at Boston University. 4 00:00:32.920 --> 00:00:36.810 Abdullah Dogan: Before we begin, a quick reminder that 5 00:00:36.810 --> 00:00:51.890 Abdullah Dogan: TAPS is accepting submissions for next season. If you are interested in presenting, we encourage you to submit a brief proposal on our website, tobaccopolicy.org, by October 12th. 6 00:00:52.480 --> 00:01:08.309 Abdullah Dogan: The seminar will be one hour with questions from the moderator and discussant. The audience may post questions and comments in the Q&A panel, and the moderator will draw from these questions and comments in conversation with the presenter. 7 00:01:08.620 --> 00:01:15.000 Abdullah Dogan: Please review the guidelines on TobaccoPolicy.org for acceptable questions. 8 00:01:15.170 --> 00:01:33.750 Abdullah Dogan: Please keep the questions professional and related to the research being discussed. Questions that meet the seminar series guidelines will be shared with the presenter afterwards, even if they are not read aloud. Your questions are very much appreciated. 9 00:01:34.100 --> 00:01:45.389 Abdullah Dogan: This presentation is being video recorded and will be made available along with presentation slides on the TAPS website tobaccopolicy.org. 10 00:01:46.810 --> 00:02:05.210 Abdullah Dogan: Before turning things over to today's moderator, we would like to take a moment to recognize Michael Darden, who will be stepping down from the Tufts Executive Board at the end of this season. Michael has served on the Executive Board for more than two years, and 11 00:02:05.210 --> 00:02:11.630 Abdullah Dogan: We are grateful for his time, dedication, and contributions to the seminar series. 12 00:02:11.760 --> 00:02:18.580 Abdullah Dogan: We are also pleased that he will continue to be involved with TAPS as a member of the advisory board. 13 00:02:18.740 --> 00:02:26.420 Abdullah Dogan: With that, I will turn things over to Michael Darden from Johns Hopkins University to introduce today's speaker. 14 00:02:28.570 --> 00:02:46.610 Michael Darden: Thanks so much, Abdullah, and thanks to the organizers. It's a great, great thing we're doing with TOPS, and so I'm proud to have been involved. Okay, so today we're going to conclude our summer 2026 season with a single paper presentation by Amit Suman entitled Distinguishing Legalization from Commercialization. 15 00:02:46.710 --> 00:02:53.729 Michael Darden: rapid retail cannabis rollout and adolescent tobacco and cannabis use in the United States. So the presentation was. 16 00:02:53.970 --> 00:02:58.770 Michael Darden: Selected via a competitive review process by submission through the TOPS website. 17 00:02:58.780 --> 00:03:23.739 Michael Darden: Dr. Ahmad Suman is a health economist at Economics for Health at the Johns Hopkins Bloomberg School of Public Health and a visiting fellow at the One Health Trust. His research evaluates effects of tobacco and nicotine tax and regulatory policies in the United States with a focus on e-cigarettes, other emerging nicotine products, and tobacco cannabis co-use. His wider work examines fiscal policies for health. 18 00:03:23.740 --> 00:03:29.520 Michael Darden: and their effects on alcohol and tobacco consumption in low and middle income countries and globally. 19 00:03:29.830 --> 00:03:38.739 Michael Darden: And his methods combine quasi-experimental policy evaluation, demand estimation, and simulation modeling. Dr. Suman, thank you for presenting for us today. 20 00:04:02.160 --> 00:04:03.700 Amit Summan: Hi, everyone. 21 00:04:04.660 --> 00:04:10.460 Amit Summan: It's my presentation sharing right now. It's good to go. 22 00:04:10.460 --> 00:04:12.389 Michael Darden: It looks great, thank you. Yep. 23 00:04:12.560 --> 00:04:23.650 Amit Summan: I'm very excited to present this work, looking at the effects of cannabis policy On adolescent tobacco and cannabis. 24 00:04:23.840 --> 00:04:27.109 Amit Summan: use in the United States. This is… 25 00:04:27.260 --> 00:04:33.290 Amit Summan: joint work with Sam Sturm, Michael Teenan, and Jeffrey Drope. 26 00:04:33.400 --> 00:04:35.920 Amit Summan: This is a working paper. 27 00:04:36.140 --> 00:04:38.979 Amit Summan: So, I look forward to everyone's comments. 28 00:04:39.490 --> 00:04:44.090 Amit Summan: And any suggestions on how to strengthen this analysis? 29 00:04:49.960 --> 00:04:58.589 Amit Summan: So, funding for this work, this is funded by Bloomberg Philanthropies, through Economics for Health at the Johns Hopkins. 30 00:04:58.750 --> 00:05:03.530 Amit Summan: Bloomberg School of Public Health, and also, the CC Foundation. 31 00:05:03.770 --> 00:05:05.650 Amit Summan: There are no conflicts of interest. 32 00:05:05.840 --> 00:05:08.390 Amit Summan: Or, industry funding to declare. 33 00:05:12.790 --> 00:05:14.030 Amit Summan: So… 34 00:05:14.500 --> 00:05:30.440 Amit Summan: In the past, 15 years, there's been, an increase in, states that have legalized and subsequently allowed for retail sales of. 35 00:05:30.550 --> 00:05:33.989 Amit Summan: Of adult use cannabis, 36 00:05:34.160 --> 00:05:51.170 Amit Summan: As of now, 24 states and DC have legalized, cannabis, and 23 have allowed for licensed retail sales. I'm going to be referring to this as recreational cannabis laws. I know some people prefer. 37 00:05:51.260 --> 00:05:57.079 Amit Summan: Increasingly adult use cannabis laws just keeping with previous literature. 38 00:05:57.370 --> 00:06:00.819 Amit Summan: That's how I'm going to be defining my acronyms. 39 00:06:00.950 --> 00:06:05.500 Amit Summan: What do we know about, you've, 40 00:06:05.780 --> 00:06:10.120 Amit Summan: Cannabis use and tobacco use, in the U.S. 41 00:06:10.320 --> 00:06:20.140 Amit Summan: Substance use has decreased generally. Cigarettes has plummeted in the US, cannabis use. 42 00:06:20.300 --> 00:06:28.929 Amit Summan: Has decreased, but it's still quite high in 2023, YRPS estimates show 14 to 16%. 43 00:06:29.410 --> 00:06:37.970 Amit Summan: e-cigarettes, I've reached a peak in 2019, and now it's kind of stabilized in, 44 00:06:38.310 --> 00:06:40.080 Amit Summan: You know, the mid-teens… 45 00:06:40.250 --> 00:06:50.520 Amit Summan: So cannabis use is still quite high along with e-cigarette use while cigarettes cigarette use has decreased. 46 00:06:50.900 --> 00:06:59.820 Amit Summan: So this audience, of course, is going to be interested in in tobacco. But as public health practitioners you would be interested in 47 00:07:00.240 --> 00:07:05.160 Amit Summan: In the deleterious effects of cannabis as well, 48 00:07:05.800 --> 00:07:11.700 Amit Summan: Cannabis, adolescence is a critical period of, development, 49 00:07:12.100 --> 00:07:16.239 Amit Summan: For the brains, cannabis can have negative effects. 50 00:07:16.420 --> 00:07:21.640 Amit Summan: On cognitive effects on on teens that use. 51 00:07:21.850 --> 00:07:32.600 Amit Summan: And also, earlier use is a risk factor for more problematic use, cannabis use disorders. 52 00:07:32.760 --> 00:07:33.949 Amit Summan: In the future. 53 00:07:36.530 --> 00:07:41.969 Amit Summan: What do we know about cannabis and nicotine co-use? 54 00:07:42.410 --> 00:07:54.270 Amit Summan: Well, it seems most research suggests that, they're actually complements Or at best, they're unrelated. 55 00:07:54.850 --> 00:07:56.850 Amit Summan: A meta-analysis. 56 00:07:56.990 --> 00:08:04.100 Amit Summan: From 2019, found that, adolescents who vape were four times 57 00:08:04.630 --> 00:08:11.880 Amit Summan: More likely to use cannabis than those who did not vape the 2023. 58 00:08:12.710 --> 00:08:18.630 Amit Summan: YRBS study also found that 61% of students who vaped in the past month also used cannabis. 59 00:08:24.950 --> 00:08:33.960 Amit Summan: So, as I said, there are increasing number of states that, that are legalizing. 60 00:08:34.250 --> 00:08:42.450 Amit Summan: Cannabis use and then subsequently retail. So these are often seen as a single policy event. 61 00:08:42.710 --> 00:08:45.790 Amit Summan: But they're very distinct policy events. 62 00:08:45.980 --> 00:08:56.649 Amit Summan: And one doesn't guarantee that the other is going to take place immediately or in the near future. There might be a large gap. 63 00:08:56.930 --> 00:09:05.239 Amit Summan: between legalization and retail. So there are different decisions that policymakers face. 64 00:09:05.400 --> 00:09:08.689 Amit Summan: One is, are we going to legalize? 65 00:09:09.260 --> 00:09:11.659 Amit Summan: Are we going to allow for retail sales? 66 00:09:11.850 --> 00:09:16.669 Amit Summan: And, if we are allowing for retail sales, how quickly are we going to 67 00:09:16.990 --> 00:09:22.079 Amit Summan: Allow for retail sales, once we legalize. 68 00:09:22.240 --> 00:09:26.530 Amit Summan: And there's quite a bit of variation, in states 69 00:09:26.940 --> 00:09:31.629 Amit Summan: Legalizing to the time that they, 70 00:09:31.870 --> 00:09:33.420 Amit Summan: implement retail. 71 00:09:33.570 --> 00:09:41.370 Amit Summan: So there's a gap of right now of states that have implemented from. 72 00:09:41.770 --> 00:09:52.980 Amit Summan: From no… no gap at all between legalization and retail, all the way to, to 50, a gap of 51 months. 73 00:09:53.420 --> 00:09:58.919 Amit Summan: So past studies, you know, they typically just focused on. 74 00:09:59.370 --> 00:10:09.629 Amit Summan: On the legalization aspect, but they haven't really separated, What happens… After legalization, 75 00:10:10.170 --> 00:10:12.990 Amit Summan: And how, you know, what the policy path is. 76 00:10:20.500 --> 00:10:23.619 Amit Summan: So, you know, what are some of the… 77 00:10:24.040 --> 00:10:28.309 Amit Summan: Different ways in which, these policies can affect. 78 00:10:30.380 --> 00:10:32.550 Amit Summan: affect adolescent use. 79 00:10:33.080 --> 00:10:50.950 Amit Summan: Or even, you know, adult use. So, over here, I have, some possible mechanisms. You know, these, these might overlap. One might be stronger in, you know, after legalization versus retail, but there are a number of things that can happen. 80 00:10:51.150 --> 00:10:55.270 Amit Summan: So of course there's a change in perceived risk. 81 00:10:55.440 --> 00:11:00.129 Amit Summan: And use norms and peer norms shift. 82 00:11:00.930 --> 00:11:14.050 Amit Summan: There might be a change in parental monitoring and parental attitudes towards, towards use, after legalization. 83 00:11:14.300 --> 00:11:26.250 Amit Summan: There might be a change in the social supply. There's typically home cultivation that's allowed after legalization. There's also been studies that shown that. 84 00:11:26.720 --> 00:11:30.250 Amit Summan: Yeah, so the separate… these are separate policies, there's… 85 00:11:30.370 --> 00:11:40.739 Amit Summan: Some states might have decriminalization earlier, but, legalization may allow for this kind of gray area where, 86 00:11:41.110 --> 00:11:45.510 Amit Summan: There's reduced enforcement of increasing laws. 87 00:11:45.690 --> 00:11:56.359 Amit Summan: And for example, when California legalized, there were lower penalties for underage possession. 88 00:11:56.500 --> 00:12:00.120 Amit Summan: And, decreases in arrests. 89 00:12:01.010 --> 00:12:14.790 Amit Summan: Once, retail opens, now you actually have, You know, you have prices, you have increased availability, a higher retail density, you have marketing. 90 00:12:15.190 --> 00:12:18.389 Amit Summan: There might be a displacement of illicit markets. 91 00:12:18.560 --> 00:12:25.060 Amit Summan: But also, There's stricter… . 92 00:12:25.770 --> 00:12:36.800 Amit Summan: There's perhaps more barriers of entry as these legal markets replace these illicit markets with age verification and also compliance inspections. 93 00:12:37.710 --> 00:12:47.339 Amit Summan: So… I already discussed that there's some co-use evidence of these two being complements 94 00:12:48.060 --> 00:12:53.110 Amit Summan: There might be shared access channels as well. 95 00:12:53.280 --> 00:12:58.790 Amit Summan: So, vape shops, selling cannabis. 96 00:12:59.020 --> 00:13:12.350 Amit Summan: And vice versa. And then there might just be shared use norms as well. So perhaps, you know, if legalization 97 00:13:12.490 --> 00:13:21.410 Amit Summan: Causes, the perceived risk to cannabis use to be lower amongst adolescents. 98 00:13:21.680 --> 00:13:25.240 Amit Summan: I may also do the same for other types of substances. 99 00:13:38.100 --> 00:13:42.879 Amit Summan: So, these are states that have YRBS data, and you can see. 100 00:13:43.430 --> 00:13:53.660 Amit Summan: The gap, between, between legalization And retail by state. 101 00:13:54.310 --> 00:13:59.799 Amit Summan: So, 12 stores opened, within 102 00:14:00.160 --> 00:14:04.299 Amit Summan: 12 states opened retail within 14 months. 103 00:14:04.490 --> 00:14:08.120 Amit Summan: of, of legalization. 104 00:14:08.970 --> 00:14:13.979 Amit Summan: And seven took between 18 months and four years. 105 00:14:15.990 --> 00:14:21.479 Amit Summan: Virginia, doesn't have any retail sales, it'll have retail sales 106 00:14:21.900 --> 00:14:25.219 Amit Summan: In, July of next year. 107 00:14:27.530 --> 00:14:31.929 Amit Summan: So there are some states that YRBS doesn't. 108 00:14:32.410 --> 00:14:43.709 Amit Summan: It doesn't, survey, so those include, oregon and Washington, which legalized. 109 00:14:44.070 --> 00:14:52.290 Amit Summan: And Colorado is an early legalizer, but data isn't, it didn't participate in YRBS in. 110 00:14:52.750 --> 00:14:55.759 Amit Summan: In, 3 of the 7 waves. 111 00:15:01.030 --> 00:15:04.889 Amit Summan: So I mentioned some of the past literature. 112 00:15:06.430 --> 00:15:12.970 Amit Summan: I'm just highlighting some key literature here. There's, you know, there's a lot of other studies. 113 00:15:13.560 --> 00:15:22.850 Amit Summan: looking at associations, and then also, doing state, evaluations. I'm just looking at 114 00:15:22.950 --> 00:15:30.410 Amit Summan: Studies that have looked at multiple states and have used these national surveys. 115 00:15:30.510 --> 00:15:41.969 Amit Summan: And… What's typical here, and especially beyond, the studies shown here, is that there's a coding where, 116 00:15:42.320 --> 00:15:50.590 Amit Summan: Basically, you have legalization. So you have, RCL, if, if the state implemented. 117 00:15:51.000 --> 00:15:55.509 Amit Summan: A cannabis law and then after that treatment, 118 00:15:55.750 --> 00:16:08.170 Amit Summan: There's… it doesn't distinguish between what other path, that state may have taken, right? Whether they implemented retail along with legalization, or they implemented retail 119 00:16:08.400 --> 00:16:10.480 Amit Summan: Much later. 120 00:16:10.680 --> 00:16:19.680 Amit Summan: So, there's been, you know, two primary data sets, used in these first set of studies. The YRBS, 121 00:16:19.920 --> 00:16:23.469 Amit Summan: Data set… So that's what, 122 00:16:24.360 --> 00:16:29.900 Amit Summan: Anderson and these two, studies by Colley and colleagues use. 123 00:16:30.240 --> 00:16:36.920 Amit Summan: And basically, they see, anderson sees a slight decrease in, 124 00:16:37.210 --> 00:16:44.420 Amit Summan: And cannabis, prevalence, 1.4 percentage points. No change when using. 125 00:16:44.530 --> 00:16:49.530 Amit Summan: the national Yrps. Kali looks at 126 00:16:50.190 --> 00:16:56.389 Amit Summan: Looks at both intensity and prevalence, and sees, a decrease 127 00:16:56.570 --> 00:17:04.200 Amit Summan: In intensity, no change in prevalence, and then also sees a decrease in our latest study of 128 00:17:04.339 --> 00:17:08.569 Amit Summan: The latest study of e-cigarettes use. 129 00:17:09.470 --> 00:17:16.299 Amit Summan: In a study just a couple of years before that, they actually see an increase In e-cigarette use. 130 00:17:16.720 --> 00:17:20.669 Amit Summan: Then there are all these NASDAQ studies, so both of these studies 131 00:17:21.230 --> 00:17:25.620 Amit Summan: By Anderson colleague, they also include the COVID, 132 00:17:26.040 --> 00:17:29.759 Amit Summan: A COVID period, where BS wave. 133 00:17:30.190 --> 00:17:36.250 Amit Summan: Then these NASDAQ studies, they're a bit earlier, pre-COVID. 134 00:17:37.060 --> 00:17:40.909 Amit Summan: And, they seem to find, 135 00:17:42.090 --> 00:17:49.369 Amit Summan: No change, all the way from no change to, you know, one study finding a slight increase 136 00:17:49.700 --> 00:17:53.710 Amit Summan: And they mostly focus on, 137 00:17:54.340 --> 00:18:02.260 Amit Summan: On cannabis here, the only study that looks at Tobacco is, 138 00:18:02.530 --> 00:18:09.550 Amit Summan: The last study in that first panel there by Wenberger, which shows a decrease in. 139 00:18:09.720 --> 00:18:19.210 Amit Summan: Cigarette prevalence, and and also co-use between cigarettes and, and, cannabis. 140 00:18:19.550 --> 00:18:27.079 Amit Summan: This… there's a few of these… so, a few of these studies have also included some sort of retail. 141 00:18:27.640 --> 00:18:28.970 Amit Summan: Indicator? 142 00:18:29.310 --> 00:18:41.130 Amit Summan: on top of on top of Rcl. And they tend to show a decrease in prevalence. 143 00:18:42.830 --> 00:18:51.250 Amit Summan: The first two studies, then Hollingsworth and colleagues, they find no change. 144 00:18:51.720 --> 00:18:53.150 Amit Summan: And 145 00:18:53.540 --> 00:18:56.960 Amit Summan: And then the one increase is, 146 00:18:57.880 --> 00:19:00.860 Amit Summan: is the NASDAQ study, which finds, 147 00:19:01.040 --> 00:19:05.210 Amit Summan: 1.5 to 1.9 percentage point increase. 148 00:19:05.470 --> 00:19:12.360 Amit Summan: If we look at meta-analyses which include, you know, which would include these state valuations and some 149 00:19:12.630 --> 00:19:23.620 Amit Summan: Associational studies, there's, there's a slight increase after, legalization in cannabis use. Doesn't distinguish between. 150 00:19:24.260 --> 00:19:29.150 Amit Summan: retail in these different environments. So, a policymaker is, you know. 151 00:19:29.780 --> 00:19:31.400 Amit Summan: Policymakers. 152 00:19:32.700 --> 00:19:40.550 Amit Summan: looking at these studies, and basically they want to know, well, should we legalize? What's going to be the effect of legalization? 153 00:19:40.660 --> 00:19:44.790 Amit Summan: Should we implement retail? What's going to be the effect on retail? 154 00:19:45.690 --> 00:19:50.060 Amit Summan: And what's the ideal gap? 155 00:19:50.220 --> 00:19:54.529 Amit Summan: And that's not very clear amongst all these studies. 156 00:19:54.730 --> 00:20:01.380 Amit Summan: You know, what the distinct policy effects are of these decisions. 157 00:20:04.740 --> 00:20:07.979 Amit Summan: So basically, what we do is, 158 00:20:08.840 --> 00:20:17.299 Amit Summan: We asked this question, what is the effect of cannabis legalization and sales, but what's also the effect of. 159 00:20:17.610 --> 00:20:18.860 Amit Summan: Of the gap. 160 00:20:19.400 --> 00:20:26.199 Amit Summan: Does… Does the gap matter? And also, 161 00:20:27.150 --> 00:20:34.380 Amit Summan: Were the effects spillover effects on cigarettes, e-cigarettes and potential co-use and polysubstance use? 162 00:20:35.770 --> 00:20:39.760 Amit Summan: A third thing we look at is also intensity. 163 00:20:44.550 --> 00:20:48.609 Amit Summan: So, as I mentioned, to the contributions here. 164 00:20:48.740 --> 00:20:57.400 Amit Summan: Is one, being very clear about, what the effects are of these distinct policy windows. 165 00:20:57.600 --> 00:21:02.329 Amit Summan: What's the effect of, a fast rollout? 166 00:21:03.750 --> 00:21:14.980 Amit Summan: And, you know, in the future, we were looking at, what the effects are with this slow rollout, so it's future work with, the 2025 wave. 167 00:21:15.820 --> 00:21:23.300 Amit Summan: And then also, were the effects by, subgroup, and then… 168 00:21:23.480 --> 00:21:34.520 Amit Summan: Another contribution is the data extension, which allows for data post-COVID, and also. 169 00:21:34.780 --> 00:21:40.740 Amit Summan: second, a second post-retail wave, for Illinois and Michigan. 170 00:21:45.210 --> 00:21:49.609 Amit Summan: So the data we use is the YRBS data. 171 00:21:50.160 --> 00:21:58.090 Amit Summan: This is a survey that's conducted in high schools. 172 00:21:58.360 --> 00:22:08.399 Amit Summan: The way we construct this data, so, basically each, each wave there's a combined, combined data set that's, 173 00:22:08.670 --> 00:22:23.690 Amit Summan: That's published and available and that combined data set has historical data, but the historical data may vary from year to year. So basically we construct this by. 174 00:22:25.030 --> 00:22:30.240 Amit Summan: Allowing for the maximum number of state wave observations. 175 00:22:30.610 --> 00:22:38.269 Amit Summan: From the 2011 to 2023 paper, 2023, period. 176 00:22:38.500 --> 00:22:44.990 Amit Summan: So this results in 46 states, 1.1 million, High school students. 177 00:22:45.140 --> 00:22:56.160 Amit Summan: So it's conducted every spring, winter semester from January to June. In the COVID period, it was 178 00:22:56.540 --> 00:23:00.299 Amit Summan: conducted in, in the fall semester. 179 00:23:00.780 --> 00:23:07.229 Amit Summan: And it's unbalanced. Unfortunately, in 2023, there was a drop in, 180 00:23:07.560 --> 00:23:10.749 Amit Summan: And the number of states that, that participated. 181 00:23:10.990 --> 00:23:12.649 Amit Summan: In this data set. 182 00:23:13.840 --> 00:23:21.159 Amit Summan: The 20… 2015 was the first year that they included, The e-cigarette question. 183 00:23:23.670 --> 00:23:27.340 Amit Summan: So, how can we look at cannabis, cigarettes. 184 00:23:27.910 --> 00:23:32.570 Amit Summan: E-cigarettes, so this is past 30-day, use. 185 00:23:34.020 --> 00:23:35.089 Amit Summan: And then. 186 00:23:35.520 --> 00:23:42.830 Amit Summan: I hope this is not confusing, so we have tool use, which is cannabis and cigarettes, so that's… so we can have a longer, 187 00:23:43.700 --> 00:23:47.769 Amit Summan: A longer period of study, and because 2015 188 00:23:48.010 --> 00:23:53.139 Amit Summan: is when e-cigarettes were allowed, we have a polysubstance 189 00:23:53.300 --> 00:23:56.049 Amit Summan: Use indicator, which is, 190 00:23:56.240 --> 00:24:02.030 Amit Summan: 2 or more of cannabis, e-cigarettes and cigarettes. Additionally, we look at intensity. 191 00:24:02.630 --> 00:24:04.069 Amit Summan: Which is defined as? 192 00:24:04.240 --> 00:24:13.279 Amit Summan: Greater than 20 days of use of cigarettes or e-cigarettes in the past 30 days, greater than 20 times for cannabis. 193 00:24:13.610 --> 00:24:17.469 Amit Summan: So basically, the YRBS 194 00:24:17.590 --> 00:24:22.970 Amit Summan: That's the number of days of use in the past 30 days for cigarettes and e-cigarettes. 195 00:24:23.190 --> 00:24:27.920 Amit Summan: But for some reason they have for cannabis they have 196 00:24:28.740 --> 00:24:32.110 Amit Summan: Only the number of times within the past 30 days. 197 00:24:32.470 --> 00:24:36.140 Amit Summan: And the the. 198 00:24:36.680 --> 00:24:54.840 Amit Summan: it's 20 to 40 times is the second highest category, and then the highest category is greater than 40 times. So our daily, cannabis is, is actually more than daily, 40 times… greater than 40 times in the past 30 days, and then for cigarettes and e-cigarettes, it's. 199 00:24:55.280 --> 00:24:57.749 Amit Summan: It's, every single day. 200 00:25:03.020 --> 00:25:13.780 Amit Summan: So… There's, three ways… Actually four ways that we code our Trient variable. 201 00:25:14.080 --> 00:25:22.179 Amit Summan: So one is what we call the just Rcl, which is basically 202 00:25:22.460 --> 00:25:34.779 Amit Summan: You're… it's equal to zero before legalization, and then it switches on once you legalize, but once you, have retail sales, you leave the sample. 203 00:25:35.040 --> 00:25:45.690 Amit Summan: Then you have RCL+, which is pretty much how these other studies have, the conventional kind of treatment, which is 204 00:25:46.000 --> 00:25:55.110 Amit Summan: Once you legalize, you're assigned as treated, and it doesn't matter what happens after that. 205 00:25:55.430 --> 00:26:02.500 Amit Summan: Then the additional, Treatment we look at is, 206 00:26:02.740 --> 00:26:11.380 Amit Summan: What we call rapid recreational cannabis retail, and that's defined. 207 00:26:11.550 --> 00:26:20.450 Amit Summan: principally is, as… Legalize retail sales within 18 months of legalization. 208 00:26:21.050 --> 00:26:24.879 Amit Summan: And then you, of course, may have a slow retail state. 209 00:26:25.260 --> 00:26:37.550 Amit Summan: Which we don't look at here, just because there are very few states that, that have that, but we plan on looking at that, once we use the 2025, YRBS. 210 00:26:38.080 --> 00:26:43.340 Amit Summan: Data, which is Basically, 211 00:26:43.530 --> 00:26:47.259 Amit Summan: I guess greater than 18 months for the, for the, 212 00:26:48.160 --> 00:26:52.899 Amit Summan: For the main kind of RSCR measure we're looking at. 213 00:26:54.970 --> 00:26:58.569 Amit Summan: There's… Yeah, fourth, 214 00:26:58.790 --> 00:27:07.929 Amit Summan: measure we look at where we have RCR, but the previous wave is… has a legal baseline. So, based… depending on 215 00:27:08.210 --> 00:27:20.829 Amit Summan: when the… the state, allowed for retail sales, and when they legalized relative to, when the YRBS wave was, was conducted. 216 00:27:21.540 --> 00:27:34.729 Amit Summan: There… there might be some pre-waves that have prohibition, some pre-waves that have, legalization, so that's a separate, kind of treatment that we… 217 00:27:35.070 --> 00:27:36.440 Amit Summan: Look at. 218 00:27:42.200 --> 00:27:50.550 Amit Summan: So… These are the different, states in our YRPS data, so there are 9 states. 219 00:27:51.300 --> 00:27:53.259 Amit Summan: That are included in our… 220 00:27:53.450 --> 00:28:00.470 Amit Summan: Main analysis, Colorado drops out because it doesn't have a pre-wave. 221 00:28:06.560 --> 00:28:10.770 Amit Summan: So, the… The estimator we start with is, 222 00:28:11.380 --> 00:28:13.750 Amit Summan: It is commonly used in, in… 223 00:28:14.790 --> 00:28:18.560 Amit Summan: In the past literature is, the… 224 00:28:19.600 --> 00:28:29.190 Amit Summan: Two-way fixed effects difference-in-difference model, where the outcome is, Past 30-day prevalence. 225 00:28:30.100 --> 00:28:39.960 Amit Summan: those 5 outcomes that I mentioned in the previous slide by state and wave, and that the Policy indicator… 226 00:28:40.180 --> 00:28:49.619 Amit Summan: He's going to be one of those three, policies in, treated state waves, 227 00:28:49.810 --> 00:28:53.769 Amit Summan: And then the covariates we use, they're… 228 00:28:54.030 --> 00:29:05.140 Amit Summan: Pretty, standard, so we actually build on On, on colleagues, controls… so there's a… 229 00:29:05.860 --> 00:29:11.680 Amit Summan: Tobacco and e-cigarette controls, cigarette excise tax, e-cigarette excise tax. 230 00:29:11.860 --> 00:29:15.290 Amit Summan: Smoke-free laws for cigarettes and e-cigarettes. 231 00:29:15.530 --> 00:29:21.410 Amit Summan: Flavored e-cigarette bans, Tobacco 21 laws, PR excise tax. 232 00:29:23.050 --> 00:29:30.059 Amit Summan: Medical cannabis laws, cannabis decriminalization dummies. 233 00:29:30.490 --> 00:29:35.690 Amit Summan: And then also a COVID activity limitations index. 234 00:29:36.140 --> 00:29:42.730 Amit Summan: Only for 2021, and then also, the unemployment rate. So. 235 00:29:43.000 --> 00:29:47.649 Amit Summan: The additional controls we add on top of, 236 00:29:48.740 --> 00:29:53.369 Amit Summan: Cauli studies are the, flavor restrictions. 237 00:29:53.600 --> 00:29:57.489 Amit Summan: Along with the harmonized e-cigarette tax. 238 00:29:57.940 --> 00:30:15.419 Amit Summan: And then also the COVID, activity limitation. So we have clustered standard errors, so we present these models with clustered standard errors, and then, also with wild cluster bootstrap p-values, given the small number of, 239 00:30:16.500 --> 00:30:18.399 Amit Summan: The treated clusters we have here. 240 00:30:23.580 --> 00:30:33.480 Amit Summan: So our main model, so that first model is to see how we compare with previous literature. 241 00:30:33.800 --> 00:30:41.490 Amit Summan: But our main model is the staggered difference-in-difference model from CALAWEE and Santa Ana. 242 00:30:42.610 --> 00:30:46.810 Amit Summan: So, as we know that, two-way fix effects. 243 00:30:46.910 --> 00:30:52.419 Amit Summan: Might not be ideal, when there is staggered 244 00:30:52.670 --> 00:31:01.509 Amit Summan: Adoption of a policy, you may have bad comparisons, you may compare. 245 00:31:03.660 --> 00:31:06.290 Amit Summan: Already treated states, 246 00:31:07.640 --> 00:31:16.479 Amit Summan: You may compare earlier tree estates with later, sorry, later tree estates with earlier tree estates, and, that may… 247 00:31:16.700 --> 00:31:21.919 Amit Summan: Allow for biases in the estimate, negative weights. 248 00:31:22.450 --> 00:31:29.850 Amit Summan: And even a potential opposite sign of the actual effect. So, These… 249 00:31:30.130 --> 00:31:40.189 Amit Summan: there are a number of staggered DID models that we can use. This is the principal model that we use, is the Callaway-Santana model. Our main 250 00:31:40.660 --> 00:31:45.850 Amit Summan: Our main, estimate looks at, 251 00:31:46.620 --> 00:31:50.660 Amit Summan: Never treated states as the control. 252 00:31:51.210 --> 00:31:56.790 Amit Summan: And, we present both the dynamic effects . 253 00:31:57.600 --> 00:32:03.779 Amit Summan: And the average effects over time. There are no covariates. 254 00:32:04.220 --> 00:32:12.470 Amit Summan: Used in these models, as is typical, but we also have a sensitivity analysis where we include some covariates. 255 00:32:12.620 --> 00:32:16.620 Amit Summan: That might be fixed, before… 256 00:32:17.410 --> 00:32:24.050 Amit Summan: Before treatment and won't be affected by, by the actual intervention. 257 00:32:25.810 --> 00:32:29.600 Amit Summan: Sorry, I see I might be going a bit over in this first section. 258 00:32:30.160 --> 00:32:34.060 Amit Summan: But… I'll try to finish my methods. 259 00:32:35.020 --> 00:32:44.030 Amit Summan: So the key identifying assumption is parallel trends that Absent, the policy, 260 00:32:44.150 --> 00:32:49.590 Amit Summan: Trends in use between treated and never treated states. 261 00:32:51.200 --> 00:32:54.580 Amit Summan: Would have moved in parallel, so we test this. 262 00:32:54.710 --> 00:32:59.490 Amit Summan: And with event study plots looking at single-year violations. 263 00:32:59.990 --> 00:33:10.189 Amit Summan: Before the treatment, single year differences in treated and control states. And then 264 00:33:10.510 --> 00:33:14.290 Amit Summan: We also do this joint test. 265 00:33:14.580 --> 00:33:18.720 Amit Summan: To make sure that we're not missing anything from the, 266 00:33:19.370 --> 00:33:24.700 Amit Summan: That the, the single year violations may may miss. 267 00:33:25.800 --> 00:33:28.889 Amit Summan: And the other assumption is no anticipation. 268 00:33:30.390 --> 00:33:37.069 Amit Summan: Before… before the, before the policy is actually implemented. 269 00:33:39.720 --> 00:33:44.069 Amit Summan: So sub-sample analysis we do is, we look at effects by 270 00:33:44.490 --> 00:33:54.590 Amit Summan: By sex, by age group, 14, 16, and 17 to 18 plus, by ethnicity and race, so white, black, and 271 00:33:54.720 --> 00:33:58.869 Amit Summan: Hispanic or Latino group subsamples. 272 00:33:59.300 --> 00:34:04.940 Amit Summan: And we drop any subsamples that are based on less than 50 respondents. 273 00:34:05.460 --> 00:34:14.309 Amit Summan: For intensity, I mentioned the different intensity outcomes. We do those for Each subsample as well. 274 00:34:14.580 --> 00:34:17.969 Amit Summan: I sent in sensitivity analysis. We also look at. 275 00:34:18.330 --> 00:34:24.579 Amit Summan: The, gap… changed the gap from 18 months to 12 and 24 months. 276 00:34:24.920 --> 00:34:35.990 Amit Summan: We also look at… Not yet treated as a control, changing from treated. We dropped 2023. 277 00:34:36.210 --> 00:34:38.320 Amit Summan: And. 278 00:34:38.530 --> 00:34:49.049 Amit Summan: the 2023 wave, just to see how we compare with, previous studies, and also because 2023 has, much fewer states included in it. 279 00:34:49.590 --> 00:34:55.049 Amit Summan: Then we also look at the effects of including controls in the Galois-Santana models. 280 00:34:55.350 --> 00:35:02.210 Amit Summan: And then we also do stack DID, and also leave one 281 00:35:03.030 --> 00:35:05.959 Amit Summan: Out analysis, dropping each treated state. 282 00:35:06.650 --> 00:35:09.920 Amit Summan: Sorry, I think I went a bit long there, 283 00:35:10.800 --> 00:35:17.249 Amit Summan: But… I'll take any questions or any comments so far. 284 00:35:17.430 --> 00:35:30.169 Michael Darden: So let's introduce our discussant. So today we're going to have Dr. Sarah Stith, who's an associate professor of economics from the University of New Mexico. She has researched 285 00:35:30.170 --> 00:35:40.790 Michael Darden: cannabis for more than a decade, including work on state and local cannabis policies and substitution between cannabis and conventional medications. So, Dr. Stith, any comments so far? 286 00:35:41.220 --> 00:35:53.679 Sarah Stith: Just, this has been great, really interesting work. I have sort of broader comments that I think are best left for the end, so I think I'll let Amit go ahead and continue with this interesting work. 287 00:35:55.010 --> 00:36:05.539 Michael Darden: Great. I think in the interest of time, we'll do that, so let's try to get the results in the next 10 or so minutes. So, Dr. Suman, take it away. 288 00:36:06.870 --> 00:36:07.599 Amit Summan: Thank you. 289 00:36:17.480 --> 00:36:32.409 Amit Summan: Okay, in the, YRBS data in our study period, we see a general decrease in. 290 00:36:33.580 --> 00:36:48.400 Amit Summan: Cannabis use, with… The difference between, so, it seems like… These RCR, states are tracking. 291 00:36:48.590 --> 00:36:57.659 Amit Summan: Never legalized states, pretty well here. There is, higher cannabis 292 00:36:57.810 --> 00:37:02.779 Amit Summan: use in these States, but the trends seem pretty similar. Cigarettes. 293 00:37:03.050 --> 00:37:10.559 Amit Summan: It's the case for most of these substances, and also our, polysubstance and dual use. 294 00:37:12.230 --> 00:37:17.899 Amit Summan: Outcomes, trends seem pretty similar, and 295 00:37:18.520 --> 00:37:23.840 Amit Summan: And we see decreases in cannabis and cigarettes, as I mentioned. 296 00:37:24.050 --> 00:37:29.930 Amit Summan: Also, a potential leveling off of the e-cigarettes after the peak. 297 00:37:33.940 --> 00:37:40.010 Amit Summan: getting to the… results… So, over here, 298 00:37:42.050 --> 00:37:49.000 Amit Summan: The first two columns present, the, 299 00:37:50.560 --> 00:37:59.539 Amit Summan: two-way fixed effects, results. The first, parentheses is the standard error, and the second parentheses has the. 300 00:38:00.170 --> 00:38:03.920 Amit Summan: the wild bootstrap p-values. 301 00:38:04.770 --> 00:38:12.129 Amit Summan: So we see nothing significant for two-way fixed effects models. 302 00:38:12.820 --> 00:38:17.280 Amit Summan: When it comes to the Callaway Santana, the great. 303 00:38:18.740 --> 00:38:29.269 Amit Summan: The grayed out text there, that means that it did not pass, past parallel pretense. 304 00:38:32.130 --> 00:38:35.680 Amit Summan: So… Basically, the only… 305 00:38:36.120 --> 00:38:43.679 Amit Summan: The only significant effect we find that passes pre-trends is RCR, and we see that there's. 306 00:38:43.820 --> 00:38:46.839 Amit Summan: A decrease about, 10%. 307 00:38:47.460 --> 00:38:53.609 Amit Summan: In, in… Prevalence of past 30-day cannabis use. 308 00:38:53.820 --> 00:39:01.900 Amit Summan: When we use both the never treated and then not yet treated states as, 309 00:39:02.110 --> 00:39:03.270 Amit Summan: Pest control. 310 00:39:03.710 --> 00:39:07.949 Amit Summan: For RCL Plus, there is, 311 00:39:08.350 --> 00:39:12.640 Amit Summan: There's a decrease, but pretense are not satisfied. 312 00:39:24.370 --> 00:39:27.349 Amit Summan: So, if we're looking at the other substances. 313 00:39:27.940 --> 00:39:35.030 Amit Summan: We see no effect, and, and, and, unfortunately, 314 00:39:36.360 --> 00:39:40.949 Amit Summan: Free Trends is not satisfied for the RCL and RCL plus. 315 00:39:41.190 --> 00:39:51.550 Amit Summan: Estimates for… For, most substances. For RSCR, it is, and there's… but there's no… no significant effect. 316 00:39:52.890 --> 00:40:02.590 Amit Summan: I also introduced the RCR, and we limit that to, states that have a legal baseline, 317 00:40:03.190 --> 00:40:09.400 Amit Summan: Rather than states that have prohibition. So those are 6 of the 9 states. 318 00:40:09.620 --> 00:40:14.620 Amit Summan: And, we do see significant effects, but there's no, 319 00:40:14.910 --> 00:40:17.540 Amit Summan: But they don't sell… satisfy pre-trends there. 320 00:40:22.440 --> 00:40:26.960 Amit Summan: Are these, sensitive to the… to the gap that we use? 321 00:40:29.440 --> 00:40:42.169 Amit Summan: So, the cannabis result is, is, is robust to changing the gap to 12 months and, 24 months. Nothing changes for, the other states. 322 00:40:43.130 --> 00:40:46.149 Amit Summan: And then I also changed. 323 00:40:47.180 --> 00:40:50.480 Amit Summan: Look at the RCR legal baseline. 324 00:40:50.840 --> 00:40:57.669 Amit Summan: Group, and what's interesting there is, so there is… 325 00:40:58.180 --> 00:41:02.159 Amit Summan: A decrease found for e-cigarettes for this group. 326 00:41:04.240 --> 00:41:12.680 Amit Summan: For 18 months, but the pre-trends are not satisfied, but for the 24-month period, pre-trends are satisfied and it's significant. 327 00:41:12.920 --> 00:41:14.599 Amit Summan: There's a decrease as well. 328 00:41:20.940 --> 00:41:25.489 Amit Summan: So, these are the event study plots here. 329 00:41:26.890 --> 00:41:31.059 Amit Summan: And you can basically see what I mentioned previously. 330 00:41:31.870 --> 00:41:36.099 Amit Summan: with RCL plus not satisfying pretense. 331 00:41:38.610 --> 00:41:46.559 Amit Summan: There's, if we're talking about dynamic effects, there seems to be… these are very short-term effects, 332 00:41:47.400 --> 00:41:56.410 Amit Summan: And hopefully with the 2025 wave, we'll see the effects, you know, four years out and then also eight years out for RCL+. 333 00:41:56.760 --> 00:42:06.890 Amit Summan: But there is some indication that these effects may grow, pro over time for RCR. 334 00:42:11.940 --> 00:42:18.720 Amit Summan: So this is offense site plots for RCR for all the five outcomes. 335 00:42:20.650 --> 00:42:24.620 Amit Summan: But of course, only cannabis has a significant effect. 336 00:42:31.950 --> 00:42:36.020 Amit Summan: So this is some sensitivity analysis with. 337 00:42:36.180 --> 00:42:43.199 Amit Summan: The stacked, DID, so we… We look at, 338 00:42:43.540 --> 00:42:49.980 Amit Summan: The effect from legal… from legalization From prohibition to legalization? 339 00:42:50.250 --> 00:42:53.250 Amit Summan: And find no significant effect, but if we look at 340 00:42:53.390 --> 00:43:00.710 Amit Summan: So this is somewhat comparable to, to the RCR. 341 00:43:01.950 --> 00:43:09.050 Amit Summan: There's a slightly, higher Average effect there with, 342 00:43:09.190 --> 00:43:13.869 Amit Summan: with the DID, but mostly the stack DID, but it's mostly consistent. 343 00:43:24.710 --> 00:43:26.849 Amit Summan: I'm sorry, I'm just going to… 344 00:43:26.980 --> 00:43:32.180 Amit Summan: It looks like my thing just unplugged there. Just give me one second, please. 345 00:43:55.800 --> 00:43:59.399 Amit Summan: Sorry, that extension cord's very unreliable. 346 00:44:02.430 --> 00:44:03.920 Amit Summan: Stack DID. 347 00:44:05.180 --> 00:44:10.620 Amit Summan: What's interesting is we actually find significant effects here. 348 00:44:10.800 --> 00:44:22.950 Amit Summan: Significant reductions for, for multiple substances, for e-cigarettes, 349 00:44:23.380 --> 00:44:26.769 Amit Summan: And the dual use, which is not surprising, 350 00:44:27.370 --> 00:44:31.020 Amit Summan: If, both cannabis and e-cigarettes are, 351 00:44:31.590 --> 00:44:37.560 Amit Summan: We're decreasing, and then even the polysubstance use can be driven by the decline there. 352 00:44:37.780 --> 00:44:39.660 Amit Summan: With, e-cigarettes. 353 00:44:47.700 --> 00:44:54.420 Amit Summan: With Leaf One analysis, you're dropping each of the treated states? 354 00:44:54.960 --> 00:44:56.740 Amit Summan: For 3 states. 355 00:44:56.900 --> 00:45:03.380 Amit Summan: The results, I mean, there are so few treated states here, but you can see that, 356 00:45:03.770 --> 00:45:14.250 Amit Summan: For three of the states, the results become marginally significant. I guess when Michigan's out, they don't quite reach that marginally significant. 357 00:45:14.850 --> 00:45:25.549 Amit Summan: And value, but this is the sensitivity of the results to, this analysis here. 358 00:45:35.120 --> 00:45:46.759 Amit Summan: So there's nothing when we look at, the sex subsamples, you know, the coefficients are. 359 00:45:47.460 --> 00:45:57.919 Amit Summan: We're all, the same direction, but, we only find significance for, for the male students. 360 00:45:58.040 --> 00:46:01.770 Amit Summan: When it comes to, 361 00:46:02.390 --> 00:46:11.239 Amit Summan: when it comes to cannabis decreases. So they're possibly driving some of the you know, the the main cannabis results here. 362 00:46:11.670 --> 00:46:16.650 Amit Summan: With females, it's marginally significant. 363 00:46:18.240 --> 00:46:21.289 Amit Summan: Since we've got the 10% level. 364 00:46:25.800 --> 00:46:34.480 Amit Summan: For subgroups, what we see is that, it's the younger students and. 365 00:46:34.840 --> 00:46:39.139 Amit Summan: White students that, that seem to be, 366 00:46:39.690 --> 00:46:44.730 Amit Summan: Potentially driving some of these decreases in use. 367 00:46:47.210 --> 00:46:48.480 Amit Summan: For cannabis. 368 00:46:50.790 --> 00:47:00.640 Amit Summan: For intensity, we don't find any significant effects for intensity for any of the treatments that we explore. 369 00:47:04.070 --> 00:47:11.080 Amit Summan: And then finally, we do subgroup analysis by… We do intensity by subgroup. 370 00:47:11.310 --> 00:47:20.239 Amit Summan: Over here, I've only included… so there are different ways to do this. So basically, I just had, you know. 371 00:47:20.520 --> 00:47:29.810 Amit Summan: My kind of decision rule for just including results here was, was, 372 00:47:30.450 --> 00:47:40.689 Amit Summan: Significance at, 18 months, and then also, significance at, 373 00:47:41.090 --> 00:47:52.559 Amit Summan: in one of the other gap periods, so either 12 months or 24 months. If it was only significant, you know, at 12 months or 24 months, I did not show those results. 374 00:47:52.810 --> 00:47:59.659 Amit Summan: Also, if the results rest on Fewer than 4 states. 375 00:48:00.270 --> 00:48:08.000 Amit Summan: I didn't show those results here. So Given… 376 00:48:08.420 --> 00:48:13.850 Amit Summan: Yeah, just because there are so many subgroups and intensities we're looking at and substances. 377 00:48:16.140 --> 00:48:25.050 Amit Summan: Given those qualifications, these are the subgroup intensity results we find. 378 00:48:25.240 --> 00:48:29.490 Amit Summan: No change in prevalence but potentially. 379 00:48:29.900 --> 00:48:36.259 Amit Summan: increases in, in frequency of use amongst, 380 00:48:36.910 --> 00:48:38.880 Amit Summan: Hispanic students? 381 00:48:39.560 --> 00:48:45.449 Amit Summan: For cannabis and then also there's a decrease in, 382 00:48:47.620 --> 00:48:54.000 Amit Summan: In taping frequency by, by students who keep vaping. 383 00:48:54.870 --> 00:49:01.429 Amit Summan: And no decrease found for prevalence in, of, of use. 384 00:49:02.870 --> 00:49:10.859 Amit Summan: Amongst, students who keep vaping, or sorry, amongst all students, but amongst students who keep vaping, there's a… there's a… 385 00:49:11.260 --> 00:49:17.190 Amit Summan: Sorry, who keep using cannabis, there's a decrease in heavy use. 386 00:49:22.930 --> 00:49:28.590 Amit Summan: So… What we basically find is that, 387 00:49:29.370 --> 00:49:33.330 Amit Summan: The decrease is coming at the, 388 00:49:34.390 --> 00:49:41.500 Amit Summan: At the retail, step, there's nothing happening, really, when it comes to legalization. 389 00:49:41.980 --> 00:49:48.040 Amit Summan: At least in this analysis, but there's a decrease . 390 00:49:48.470 --> 00:49:55.340 Amit Summan: For states who rapidly, implement retail, after legalization. 391 00:49:55.650 --> 00:49:57.320 Amit Summan: About a 10%. 392 00:49:57.490 --> 00:50:06.529 Amit Summan: Decrease in use… There isn't robust evidence of of, 393 00:50:06.750 --> 00:50:14.359 Amit Summan: Changes in tobacco, but there is some weak evidence that e-cigarette use is decreasing. 394 00:50:14.730 --> 00:50:16.979 Amit Summan: And there are also changes in, 395 00:50:17.660 --> 00:50:22.490 Amit Summan: In dual and polysubstance use, so decreases. 396 00:50:22.620 --> 00:50:32.160 Amit Summan: For subsample groups, There might be some indication that the effects are concentrated amongst males. 397 00:50:32.300 --> 00:50:36.880 Amit Summan: And younger and white students. 398 00:50:39.300 --> 00:50:49.770 Amit Summan: The… Change in intensity of use in Hispanic students might be alarming, but There isn't, 399 00:50:50.380 --> 00:50:52.890 Amit Summan: A lot of changes seen in, 400 00:50:53.330 --> 00:50:56.679 Amit Summan: In the intensity of, of use generally. 401 00:51:02.210 --> 00:51:10.100 Michael Darden: So why don't we… why don't we go to our, thank you, Dr. Suman. Why don't we go to Dr. Stith to get us, have some comments? 402 00:51:11.410 --> 00:51:36.410 Sarah Stith: Thank you so much, Amit. This was really interesting. You know, it's an interesting research question with counterintuitive findings. We see this increase in access to cannabis, and then for some reason, the youth are reducing use. And I think this… and I think that your treatment measure is really a very interesting step towards trying to deal with the issue of how do we measure cannabis legalization, because to me, this is a long, sort of continuous process that starts off 403 00:51:36.410 --> 00:51:49.859 Sarah Stith: With decriminalization, it works through this whole medical legalization system into the rec legalization. There are all sorts of different characteristics of these laws. And even once the first store opens, how the market evolves from there can vary enormously. 404 00:51:49.860 --> 00:51:57.320 Sarah Stith: And so, what I would encourage you to do is to take this idea of the rapid commercialization and try to dig into what 405 00:51:57.320 --> 00:52:19.739 Sarah Stith: Specific policy characteristics are driving your different outcomes across these states. So, for example, when I think about different ways that cannabis legalization and the rollout of stores gets delayed from my own experience with these laws, is that one, you get efforts by the legislature, the executive and judicial branches to try to block ballot initiatives passed by citizens. 406 00:52:19.740 --> 00:52:22.080 Sarah Stith: And try to prevent these stores from ever opening. 407 00:52:22.080 --> 00:52:29.149 Sarah Stith: So in Virginia, the Republican governor just refused to sign the regulations that were needed to set up the retail stores. 408 00:52:29.150 --> 00:52:44.460 Sarah Stith: And now they have a Democrat governor, she's agreed to sign it, but she wants to change the tax rate, and so it's all gotten delayed again. Then there are other states where you see a slow rollout, but it has to do with sort of more in the post-store period, so like in Illinois. 409 00:52:44.460 --> 00:52:57.819 Sarah Stith: The license fees were something like $60,000 for a dispensary, whereas in New Mexico, for $2,500, you can have a vertically integrated micro-license. And so, in New Mexico, the market opened up and it exploded. There's, like. 410 00:52:57.820 --> 00:53:04.600 Sarah Stith: three dispensaries on a block in some places in Albuquerque, whereas in Illinois, the whole south side of Chicago had like one dispensary. 411 00:53:04.600 --> 00:53:23.339 Sarah Stith: So there's a really huge differences. And I would encourage you to try to sort of tie it to the youth access. So if we're thinking about, you know, the medical variable, the medical legalization, were, 18 year olds allowed to buy without a chaperon if they had a medical license, for example, the home cultivation components, I think are really important. 412 00:53:23.340 --> 00:53:47.640 Sarah Stith: And then I think, you know, there's also issues, too, where sometimes the delays is because the law was really poorly designed. So New York, we saw that no one qualified for the equity licenses. So that delayed everything opening up. And so I just I think this rapid rollout, you're on to something. You're finding different results for these different groups of states. And I would just encourage you to try to dig into sort of what are the components you think are most affecting? 413 00:53:47.700 --> 00:54:01.060 Sarah Stith: this youth use, now that you've found there's this negative effect, which is sort of counterintuitive, too. But so let me start off with just that comment on the treatment variable, and then I have some more thoughts, sort of, on the mechanisms, maybe, and kind of where those are going. 414 00:54:04.160 --> 00:54:06.190 Amit Summan: Thank you so much. 415 00:54:06.630 --> 00:54:16.690 Amit Summan: for your comment, and I've also reviewed some of your work before this talk, and I enjoyed reading some of that. 416 00:54:17.710 --> 00:54:23.680 Amit Summan: I think… so I'm hoping, you know, with, so I agree with everything you say. 417 00:54:23.840 --> 00:54:27.650 Amit Summan: I'm hoping, you know, as we… as we… 418 00:54:27.790 --> 00:54:35.299 Amit Summan: get more data, and perhaps I go into different surveys, and, you know, the 2025 YRBS data, I hope. 419 00:54:35.470 --> 00:54:40.159 Amit Summan: You know, a lot of these state-specific, as we get more treated states. 420 00:54:40.350 --> 00:54:44.240 Amit Summan: Some of these state specific things that are happening. 421 00:54:45.230 --> 00:54:52.310 Amit Summan: kind of average out, but they're very, very important statistically. But they're very, very important. 422 00:54:54.390 --> 00:55:00.709 Amit Summan: And I guess, you know, that's stuff that more and more studies now are kind of looking at. 423 00:55:01.040 --> 00:55:04.949 Amit Summan: Effects of enforcement. 424 00:55:05.100 --> 00:55:20.310 Amit Summan: effects of retail density. That's not something we kind of do here, but I think it's important to look at. So I, you know, I'll see how we can look at some of these other, types of outcomes and kind of look at some of these 425 00:55:20.580 --> 00:55:23.700 Amit Summan: These mechanisms. 426 00:55:23.810 --> 00:55:26.460 Amit Summan: Because… 427 00:55:27.090 --> 00:55:44.880 Amit Summan: Yeah, as you mentioned, the policy pathways and what's happening behind the scenes, beyond that, oh, this state just implemented retail or legalized at this point, there's a lot going on, you know, behind the scenes. 428 00:55:44.980 --> 00:55:58.219 Amit Summan: That some of these numbers might hide, and yeah, there's the politics of it, and… and, you know, what's actually happening, and what's… what's the actual, availability. 429 00:55:59.020 --> 00:56:04.599 Amit Summan: you know, for youth when it comes to… on the street or wherever, right? 430 00:56:04.730 --> 00:56:13.370 Amit Summan: So that's all very interesting, and I want to continue looking at how we can explore those things statistically. 431 00:56:17.300 --> 00:56:32.370 Sarah Stith: Amit, I was also thinking that maybe, you know, you could already kind of build off your leave one out analyses you've already done to kind of get an idea of who's driving these effects, and then dig deeper into those states to kind of see what might be the policy levers, rather than just kind of blindly trying to code up a whole bunch of 432 00:56:32.370 --> 00:56:40.859 Sarah Stith: stuff. And to be fair, I've talked to two prominent econometricians about this whole issue of how to measure this, and their main response has been, I feel your pain. 433 00:56:40.860 --> 00:56:41.650 Sarah Stith: So… 434 00:56:41.760 --> 00:56:51.569 Sarah Stith: This is complicated and certainly a difficult econometric question. I'm hoping, actually, I see there's some comments in the Q&A that maybe some of the audience have some good suggestions as well. 435 00:56:52.590 --> 00:57:05.930 Michael Darden: Yeah, I mean, I can ask a few of these. So, one is, does your stacked different diff specification include covariates, or does it match the kind of base specification you see with Callaway Santana using no covariates? 436 00:57:07.000 --> 00:57:11.150 Amit Summan: It includes covariates and then also the stacked. 437 00:57:12.870 --> 00:57:14.500 Amit Summan: DID weights. 438 00:57:15.240 --> 00:57:17.000 Amit Summan: From recent literature. 439 00:57:17.920 --> 00:57:22.369 Michael Darden: Okay, great. And then, so, like, how should we think about… 440 00:57:22.440 --> 00:57:40.169 Michael Darden: some of the adjacent states to RCL states. So, you know, there's a whole literature on kind of cross-border shopping, but in the context of tobacco, but here, it's, like, especially interesting because there might be legal stores across state lines, and Ben Hansen has these great 441 00:57:40.170 --> 00:57:52.879 Michael Darden: papers that show that, like, people are willing to commit a felony to drive to another state and bring it back, rather than go to, like, the street corner and buy, decriminalized marijuana. So, how do you think about adjacent states? 442 00:57:54.920 --> 00:58:09.140 Amit Summan: Yeah, that's a great point. You know, I guess in in future analysis, we kind of thought about this along with, you know, a number of different things we can do is just dropping out. 443 00:58:10.420 --> 00:58:22.770 Amit Summan: states, bordering states that haven't legalized from the control, so perhaps we can look at that, in the future. I do wonder what the effect is, like, is it increasing 444 00:58:23.460 --> 00:58:25.790 Amit Summan: You know, is it having, 445 00:58:25.940 --> 00:58:33.480 Amit Summan: A different effect on these control states that they're driving over to get… to purchase, 446 00:58:33.900 --> 00:58:41.820 Amit Summan: These cannabis products, if that's the case, then that would inflate our, the… the… 447 00:58:41.980 --> 00:58:48.399 Amit Summan: the negative, value we have here. So, that would be interesting to check. I'm wondering… 448 00:58:48.900 --> 00:58:55.510 Amit Summan: You know, again, I would have to think about the mechanisms a bit, why the control states 449 00:58:56.120 --> 00:58:59.309 Amit Summan: Why they would be able to benefit from this, whereas 450 00:58:59.410 --> 00:59:03.639 Amit Summan: Those in mistreated states have not. 451 00:59:03.700 --> 00:59:21.289 Amit Summan: Maybe their illicit market supply is actually better for them to get, you know, they're not getting ID checked and stuff like that. Maybe that still exists for them, so perhaps it's better for them to stay, but I'm just… you know, this is all conjecture, so that's something I can look at. 452 00:59:23.570 --> 00:59:36.299 Michael Darden: Okay, great. I think we're just about out of time, so I want to thank you for your presentation, and thank Dr. Stith for a wonderful discussion, and I will kick it back to Abdullah, who will take us out. 453 00:59:39.600 --> 00:59:53.510 Abdullah Dogan: We are out of time. Thank you to our presenter, moderator, and discussant. Finally, thank you to the audience of 128 people for your participation. Have a top-notch weekend!