Episode 3: Analisa Packham
Analisa Packham
Analisa Packham is an Assistant Professor of Economics at Miami University.
Date: May 14, 2019
A transcript of this episode is available here.
Episode Details:
In this episode, we discuss Professor Packham’s work on the effects of syringe exchange programs:
"Are Syringe Exchange Programs Helpful or Harmful? New Evidence in the Wake of the Opioid Epidemic" by Analisa Packham.
Other research we discuss in this episode:
“Effectiveness of sterile needle and syringe programming in reducing HIV/AIDS among injecting drug users” — World Health Organization report.
“Needle exchange programs and drug injection behavior” by Jeff DeSimone.
“Needle exchange programs: Research suggests promise as an AIDS prevention strategy” — GAO report.
“Syringe exchange programs around the world: The global context” — GMHC report.
“The Effects of Naloxone Access Laws on Opioid Abuse, Mortality, and Crime” by Jennifer L. Doleac and Anita Mukherjee.
Transcript of this episode:
Jennifer [00:00:07] Hello and welcome to Probable Causation, a show about law, economics and crime. I'm your host, Jennifer Doleac of Texas A&M University, where I'm an Economics Professor and the Director of the Justice Tech Lab.
Jennifer [00:00:18] My guest this week is Analisa Packham. Analisa is an Assistant Professor of Economics in the Farmer School of Business at Miami University. Analisa, welcome to the show.
Analisa [00:00:27] Hi, thank you so much for having me. I'm really excited to be here.
Jennifer [00:00:31] So today, we're going to talk about a working paper that you just released about the effect of syringe exchange programs on a variety of outcomes, you look at HIV diagnoses, opioid abuse, crime. But to start out, could you tell us about your research expertise and what led you to become interested in this topic?
Analisa [00:00:49] Sure. So I am an applied microeconomist, and my research focuses primarily on evaluating health policy. So that means to date, I've done work on contraception access, nutritional assistance timing, and also a paper on physical education. So this recent paper is actually the first study I've done on policies to address the opioid crisis. And it certainly won't be the last because it's gotten me so much more interested in the complex issues surrounding the epidemic. But I really became interested in this topic of syringe exchange specifically for a couple reasons. So first of all, three years ago, I moved to rural Ohio. So I currently work at Miami University and I've really seen some of the devastation of the opioid crisis firsthand. And it's affected some of the people that I know and love. So I have somewhat of a personal investment in this issue and trying to figure out what kinds of policies can best address the crisis. And I also became interested in syringe exchange because I was advising an undergraduate student at Miami whose name is Katherine Wells, and she's from Portsmouth, Ohio. So Katie was really interested in SEPs because her hometown had recently opened a program. And so she was telling me about this, and we had many conversations about syringe exchange programs more broadly. And then Katie actually wrote an honors thesis about syringe exchange programs and eventually went to work for the Portsmouth Syringe Exchange program. And so that kind of facilitated more of an interest to me and the hands-on type of work that syringe exchange programs do every day. And I think another reason that I'm really interested in talking about syringe exchange is that really nothing has seemed to be successful thus far in addressing the increases in opioid related mortality that we're seeing across the U.S. So the CDC has really promoted SEPs as one of these foolproof ways to reduce blood borne illness with no downside and no unintended consequences. So according to agencies like the CDC, these programs really seem like a net win. And the promotions even caught the attention of media outlets like NPR and has been suggesting that more cities open SEPs. But as I kind of looked more into the research as an economist about SEPs, there's really little study of these programs in recent years and really even less work on the causal effects of SEPs on outcomes other than HIV, so things like health and crime, in these areas that choose to open the program. So I really think that there's a lot of work that needs to be done in this area.
Jennifer [00:03:28] Yeah, I agree, it's a super important topic. And as you've you've just alluded to, the public health community likes to remind economists who study topics like this that there's a long literature that they've worked on, on syringe exchange programs and similar harm reduction efforts. So tell us a little bit more about what that literature looks like and kind of what we what we do know what we knew before your paper came along.
Analisa [00:03:52] Yeah, absolutely. So, as you mentioned, this isn't a particularly new topic. And public health folks have been really studying this question for decades. And my guess is that, you know, public health researchers and epidemiologists listening to me discuss this right now are thinking, come on, this question isn't new. It's already been solved. And in some ways, this is right. You know, so there are many studies looking at syringe exchange program clients and their outcomes after they visit a program. And generally, these studies to date find that syringe exchange programs are successful in reducing HIV and blood borne illness, which is similar to what I find. And so they imply then that these programs are very cost effective because it's fairly cheap to give out syringes and fairly expensive to treat diseases like hepatitis C and HIV. And I want to highlight here that more specifically, you know, the previous research on syringe exchange is largely correlational. So these studies focus on syringe syringe sharing, mostly during the AIDS crisis in the 1980s and 1990s, and find that the programs are associated with reductions in the spread of HIV and reduced syringe sharing behavior. But they also find that these programs are not correlated with an increase in the amount of drugs used by current drug users or any kind of increase in new drug users. And there has been some economic studies. So, for example, there was one in the Journal of Policy Analysis and Management from Jeff DeSimone. There's also a literature compilation of public health studies from a GAO report that syringe exchange programs result in fewer discarded syringes. And so the benefits of people excuse me, indicating that the benefits of SEPs are really through preventing the spread of of HIV and that these benefits are not limited to people who inject drugs. But again, I want to emphasize here that the data from these compiled studies is not really ideal. So it often includes small sample sizes, oftentimes just providing data from one syringe exchange clinic and also self reported data regarding individuals' drug use rates. And then they also don't typically consider spillover effects or externalities on those not directly treated or on other outcomes like other health and crime outcomes. Another thing I wanted to bring up is that many of these studies are comparing drug use in the US to other countries. So a lot of times you'll see studies comparing U.S. syringe exchange to Canadian syringe exchange or syringe exchange in New Zealand. But it's pretty problematic to address causality there because, for example, these countries like New Zealand might have other programs like fee-based services. And we know that other countries, including Canada, provide substance abuse treatment to injection drug users for free. So it might not necessarily be the fact that needle exchange doesn't increase drug use, but rather that these people just have access to other services on top of needle exchange.
Jennifer [00:06:56] So. So this gets to the heart of why this is a difficult empirical question to answer. But, you know, there are a lot of difficult questions to answer. Talk a little bit about why this has been such a, why there hasn't been any other research that's really been able to identify the causal effects of syringe exchange, particularly on opioid abuse, until you came along.
Analisa [00:07:21] Yeah. So I think first and foremost, one of the challenges with identifying causal effects of syringe exchange is that there is no directory kept of syringe exchange programs over time. So it's really hard to analyze syringe exchange openings or closings because there's just not really a dataset that can track these things. So there are some datasets that contain snapshots of syringe exchange program locations. But like I said, there's really no data tracking them over time. And so I'm going to be able to overcome that by kind of doing my due diligence in tracking down these different locations. But another thing that makes this really difficult is that in terms of outcomes, you know, the CDC makes annual county level HIV diagnosis available since 2008. But, the data are censored at the county level if you have a county with less than five counts. And if you really dig into that, you can see that this corresponds to about 75 percent of the data. So while the CDC offers these data, it's really not that useful for looking at changes in HIV over time.
Jennifer [00:08:33] And just to clarify, when you say censored, you mean that they're basically saying it's missing or if there are fewer than five cases, they basically don't release the count. So that in 75 percent of cases it's basically just a missing variable.
Analisa [00:08:45] That's right. So we just don't know if that county year observation is a zero or whether it's a four. Right.
Jennifer [00:08:53] Got it.
Analisa [00:08:54] Yeah. So, I mean, the thing is that states have been asked to maintain HIV case level data by the CDC since 2008, but they don't even always make this available to researchers or even publish this count these counts on their their county website. So data is a main challenge here. And then beyond that, setting effects like drug use, for example, is also really hard to get at with current data. So there aren't really good comprehensive measures of drug use across the U.S. over time. So generally we have to use things like hospitalizations or mortality data just as proxies for drug use. Any data on drug use thus far just kind of relies on ex post survey data, meaning that surveys will ask individuals whether or not they've used drugs in the past week or month or year, or they might follow clients of syringe exchange programs and ask if they've used drugs after entering the program. But these kinds of survey data can suffer from a lot of bias, and they might not be represented across the US. So, for example, if you think about asking somebody whether or not they use drugs, which is illegal behavior, there are certain incentives to maybe shade down your drug use. Or you may just not remember in the past year how many times you've used drugs. Or certainly anybody that is a drug user, we might think that they have a certain higher probability of mortality. And so that person isn't going to be in a survey sample for a long time horizon. So, you know, in that case, we would rather use something like mortality data, which is administrative and could at least show some instances of fatal misdoses, which we can think of as kind of a lower bound for drug use. And then the other issue, I would say, is that, you know, the opioid crisis is recent. This isn't something that's been going on for more than I'd say 20 years. And so we really have to wait for data to catch up to what's happening. And we need data recent enough to map out how these new syringe exchange programs are actually affecting people today. So this project is really a labor of love because I really have done a lot of effort to collect as much data as possible that exists that would give us a good sense of opioid related drug use and other health outcomes.
Jennifer [00:11:20] And so you're going to be looking at as syringe exchange programs open and close in various places across the United States and basically comparing trends in places that changes in trends in places that open a syringe exchange program with the trends in places that don't open a program. So kind of a difference in difference framework, is that right?
Analisa [00:11:44] Yes, that's exactly right.
Jennifer [00:11:46] OK, perfect. And so tell us about the these actual interventions and how how frequently these places open and whether you're seeing openings and closings, like how how often do they open and then close in a year? And just tell us more about the context that you're studying here.
Analisa [00:12:04] Yes, so many cities, as I've mentioned, have considered opening these programs in recent years due to the fact that the CDC and other medical organizations have promoted them as being sort of a net good. And so what I want to look at in this study is programs that have opened since 2008, so sort of what I think of as as the middle or high of the opioid crisis. And so in terms of that, I really want to estimate in this study if the openings of these programs have tradeoffs and estimate to what extent these programs, which, you know, we haven't even mentioned really what they do yet, but they provide injection drug users with sterile needles, hygiene kits and referral options for for drug counseling, really in this nonjudgmental setting. And so I just want to look at the openings, as you mentioned, how they can address the spread of blood borne illness. And so what I'm doing in terms of tracking openings is I'm looking at a snapshot of syringe exchange program clinics that the North American Syringe Exchange Network put out in 2016. And so contacting all these clinics, either directly or finding information from them online, I just look to see if I can find opening dates on the existence of those clinics that I saw on the snapshot in 2016 so that I can identify which of these are recent clinics. And so, as you mentioned, I'm going to be comparing the clinics that opened in the last 10 years with clinics that opened much previously to that. And to get a sense of how many clinics have opened, I identify about 90 counties that experienced openings in the past 10 years. So that corresponds to ninety five clinics. And over half of these clinics are opening in areas that previously did not have a clinic. So we're thinking about, for example, rural Ohio has opened a couple syringe exchange programs in the last 10 years. And as you mentioned, kind of one of the the potential limitations of the studies, I'm actually not going to be able to track closings. So all I have is to go off of is a snapshot in 2016 of programs that existed at the end of my dataset. And I kind of follow them backwards. But for example, if we think that a clinic opened in 2010 and closed in 2012, I won't be able to observe that clinic in my study.
Jennifer [00:14:30] We should talk a little bit more, I guess, about what syringe exchange programs actually do. So they they provide free syringes. And so this makes it cheaper and easier for people to use opioids and other injection drugs more safely. As an economist, how do you think through like, what's interesting to you about this? Is it just the the the impact on on blood borne illnesses or like how are you thinking about the potential mechanisms of what kind of impact these kinds of programs can have?
Analisa [00:15:03] Yeah, it's a good question. So really what I was thinking through here is first I am really thinking through how do these programs work on the ground floor. So it's nice that we can gather a bunch of administrative data, but it's even better if we can really like contact programs and ask them kind of what are your day to day operations look like? And what do the clientele look like that you're serving? So first and foremost, you know, I was able to get some descriptive data to just see what syringe exchange programs look like on the ground floor. So I gathered some client level data from just a small syringe exchange program in rural Ohio. And I think that that's really interesting. They're not going to be necessarily representative of all clients across the U.S. because we know that, of course, rural Ohio is very different from a place like, say, Philadelphia. But, you know, I'm interested in some insight into what are the backgrounds and demographics of visitors. And so I'm able to look at that first and foremost and see that visitors are about 38 years old, they're largely white, and about 80 percent use heroin. And about 16 percent have used fentanyl, which is about 80 times more powerful than morphine.
Analisa [00:16:11] So just given that information and going off of that, I think the next best question is since syringe exchange programs are focused on helping injection drug users get access to sterile needles and other supplies, how effective are they at actually getting them into drug treatment facilities or counseling for drug use in a way to start them on a path towards recovery? Or rather, are these programs simply just serving as a way to provide spillovers to the community by getting dirty needles off the street? And if the latter case is what's happening, then we might be worried that in focusing only on the issue of blood borne illness, that lowering the cost of injection drug use could actually increase overall opioid related use. And so in thinking about that or potentially more important question of how do syringe exchange programs affect blood borne illness, which is, number one, their intended goal, what they're actually trying to achieve is reducing cases of HIV and hepatitis C. So I want to look at that, obviously, as kind of the first order, but then thinking about if we do reduce the costs of using drugs, how does that affect drug use? And I'm going to end up proxying for that with mortality and opioid related hospitalizations. So just to kind of answer your your initial question, which is what is interesting here, I think really what's interesting is are syringe exchange programs achieving their intended goal of reducing HIV? And if so, are there any tradeoffs of that? So as an economist, we care a lot about unintended consequences. And so if we think that opioid related mortality is increasing as a result and obviously mortality has really high societal costs, then we might be worried that syringe exchange programs need other interventions on top of them to essentially reverse any kind of large increases in mortality in the wake of the opioid crisis.
Jennifer [00:18:08] So, like in the best case scenario, you have people who are addicted to drugs and this makes it safer for them to use them. But if there isn't - and I think a lot of advocates for syringe exchange programs and other harm reduction methods really hope that this will give people a chance to get treatment - but if if we don't see them actually getting treatment, then it might not have those benefits and might might have the unintended consequences as you're you're describing. So what are your main findings?
Analisa [00:18:42] Yeah. So first of all, I just wanted to address what you just said, because I think that's really important, which is that we do want to see syringe exchange programs. If they're touting this as a benefit of getting people into drug counseling, we would expect to see that at the client level. And so actually, what I find just from data just from one clinic is that only about one percent of clients are actually seeking treatment for referrals and drug counseling. So it doesn't seem that syringe exchange either has the bandwidth or ability, at least in rural settings, to get people to treatment. So I just wanted to think that I just wanted to mention that because I think that's an important starting point.
Jennifer [00:19:17] Yeah, that's really interesting.
Analisa [00:19:19] Yeah. So in terms of the the main findings, so in this paper, I'm really interested in estimating effects for four main outcomes. As I've already mentioned, HIV rates I think is a really important first first order condition because that will give us an idea of the spread of blood borne illness as a result of syringe sharing, but also opioid and drug related mortality, which gives us a sense of how syringe exchange programs affect fatal overdoses and also opioid related hospital visits. That's going to include emergency room visits and inpatient stays, which hopefully gives us another measure of drug usage or at least a proxy for drug use that we think might be better than mortality. And then I'm also going to be looking at drug related arrests. And so that's going to include possession of opioids, as well as the sale of opioids and also arrests for theft, which I think can indicate drug use or have an added benefit of us thinking through maybe legal attitudes towards addiction within a community. So, for example, if we think that the community is OK with building a syringe exchange program, maybe this reflects a shift in attitudes towards addiction or injection drug users. And so we might also think that the legal community responds to that, that shift in attitudes as well.
Analisa [00:20:37] So if we just dig into the main results here. So what I basically find is that syringe exchange programs do lead to significant reductions in HIV rates. And this is really consistent with the previous literature. So I find that syringe exchange programs reduce HIV diagnosis in counties that have a new syringe exchange program opening by about 25 percent. And if we look at this effect spread out over three years, this effect grows to about 30 percent over time and the effects grow over time. So they're concentrated in the two plus years after the clinic opens, potentially due to the nature of the spread of disease. And this corresponds to about 30 fewer HIV cases per county with a syringe exchange program per year. And so if we think about what that means nationwide, that's about 6000 cases per year that we would have expected in the absence of a syringe exchange program. So syringe exchange programs do seem to achieve their intended goal of reducing syringe sharing and reducing blood borne illness.
Analisa [00:21:44] When we turned to looking at drug related mortality, however, I find that syringe exchange programs lead to an increase in drug related mortality, including opioid related mortality. And I also find that these effects are driven by heroin deaths, which might be unsurprising given that a large percentage of syringe exchange program clients are primarily using heroin. So the estimates indicate that a syringe exchange program opening corresponds to about a 13.5 percent increase in opioid related mortality. So this corresponds to about 13 additional deaths per county per year that we would have not predicted in the absence of the program. And importantly, I find that these effects are really driven by effects in rural areas and that the effects are most prevalent, really, one to two years after a program opens, but they fade out in much later years, so four plus years later. So I think that that suggests that syringe exchange programs might be affecting the most vulnerable visitors. So those that are already on a pathway towards drug related mortality, and especially for those in rural areas that might have financial or transportation barriers to obtaining treatment. And then I also find effects on opioid related hospitalization, which, as I mentioned earlier, might be potentially a better proxy for drug use than mortality because it doesn't pick up just the fatal cases of misdosing. So I find that a syringe exchange program opening increases opioid related emergency department visits in a state by 8.9 percent, and that corresponds to about over 1,400 additional visits per year. And so essentially what these findings indicate is that while syringe exchange programs are fairly successful at reducing blood borne illness in terms of HIV rates, they might unintentionally encourage more opioid use just by lowering the physical or even the networking costs of injecting drugs.
Jennifer [00:23:52] And if I'm remembering right, the graphs of the outcomes on opioid related E.R. visits and opioid related mortality suggest the mortality increases and then falls. But then but opioid related E.R. visits just keep going up. And so that, that struck me as suggestive evidence that, you know, over this time period, naloxone is also becoming more available, which saves your life if you're overdosing. So it might just be that we're getting better at saving the lives of people who are who are overdosing. But like but you're finding that more and more people are overdosing to begin with. Does that sound right?
Analisa [00:24:26] Yeah, that's absolutely right. So exactly what you said, which is that emergency room visits and I neglected to mention it earlier but also drug related arrests for possession of opioids is also increasing and keeps increasing over time. So while affects for mortality kind of fade out, we don't see those effects for emergency room visits fading out. So exactly what you're saying is correct. It just might be that, you know, potentially we're just getting better at saving lives. And I'm not here to say that that's a bad thing. But it might also indicate that drug use is not slowing down, so to speak, over time.
Jennifer [00:25:05] Right. So papers like this always include like a 100 page appendix with a million robustness checks. And of course, the exercise that economists are going through is trying to convince the reader that you're identifying the causal effect of these programs and not just some preexisting trends or, you know, people moving into the area in response to these programs or something like that. So talk a little bit about all the different tests you do to convince yourself and and your readers that it is that the effects you're identifying are those of these syringe exchange programs.
Analisa [00:25:39] Yeah absolutely. So just to back up a bit. So what I'm comparing in the study is I'm just looking at the effects of health and crime outcomes for counties that experience syringe exchange program openings in the last 10 years. And I'm comparing those counties over time to other counties that already had syringe exchange programs but didn't have any additional openings. And I think that, you know, that might be a good comparison group for for a few reasons. But we also need to be careful when we're identifying any kind of causal effect that there aren't, as you said, preexisting trends or something else driving this. So, first of all, we shouldn't expect the opening of a syringe exchange program to be random. So counties opening syringe exchange programs in the height of the opioid crisis might look very different in terms of demographics or locales or economic conditions than other counties across the US. So, for example, Portsmouth, Ohio, may look very different than, say, you know, a county in rural Texas during the opioid crisis. And so we want to really be careful not to just look at raw trends and mortality in these counties over time and just declare that the effect is causal once they open a clinic. Because it could really just be to due to a continuation that opioid mortality was going up before and after the program opening. So instead, as I mentioned, I'm going to compare counties with openings to those that had existing syringe exchange programs. And so what I'm going to do is I'm going to compare the trends in health and crime outcomes in these treatment and comparison groups over time. And I'm going to compare them before and after the syringe exchange program opening. So importantly, by comparing the counties with new syringe exchange programs to those with existing syringe exchange programs, the methodology isn't going to require these counties to be similar on observable characteristics. But it is going to require that trends and health outcomes and crime outcomes are following a similar trajectory prior to the opening of the clinic, meaning that the trends in opioid related mortality in these two areas look similar before the opening happens. And if then those trends diverge after a syringe exchange program opening, then we can attribute the increase in mortality, per se, to the opening of the clinic.
Analisa [00:28:08] So the idea is basically this. To get at the causal effect, I kind of am asking the question, what would a county's health outcomes have looked like in the absence of a syringe exchange program opening? And so I'm just going to compare the actual outcomes to what we could have predicted in the absence of a of the clinic. So, of course, we're worried that counties that choose to open a new program, they're experiencing maybe a new wave of injection drug use, and that's why they're opening the programs. So the counties might be opening a program, for example, because the opioid crisis has hit particularly hard and that locality has has now made a new push for anything that might reverse that trend. So in this case, estimates would then just be picking up a continuation of a trend of increasing mortality. We wouldn't say then that the effect is due to the syringe exchange program, but rather just that existing trend. Now, the good news is in in the data that I have since I have data before the openings, I can actually check if trends in the treatment and comparison group are following a similar trajectory prior to the opening. So I can control for observable differences statistically and then say that any divergence is attributed to that program. So what I find is that counties with existing syringe exchange programs and those with new syringe exchange programs do experience similar pre trends in the increases in opioid related mortality. So essentially what's happening here is, yes, opioid related mortality is increasing throughout these counties during the time of the study, but they're increasing in a similar way. And then counties with new syringe exchange programs are increasing even more sharply after the opening. So that's kind of the idea for the identification strategy. And I find that these trends are similar in all of the outcomes that I study, including emergency department visits for opioid related overdoses, trends on crime outcomes, and also opioid and drug related mortality.
Analisa [00:30:11] Now, as you mentioned, we might be also worried that there is alternative explanations or maybe omitted variable bias that might be driving the results. So I won't go through every single one of my appendix checks or robustness checks on this. But hopefully I can assuage you that there are many checks in the paper to show that the findings are not just driven by my ad hoc decisions of functional form or statistical modeling or or my defined comparison groups. So, you know, these checks would hopefully be some be reassuring, that it's not just my way of picking counties for the study or using what I call statistical voodoo. That's yielding these. So the other issue is that we might worry that counties with new syringe exchange programs are also trying lots of other policies at the same time to try to mitigate opioid use within their communities. So we know that over this time, states have passed a number of policies, including things like Medicaid expansion, as you mentioned earlier access to naloxone, and also things like doctor or pharmacist laws like tamper resistant prescription forms or prescription drug monitoring programs and other similar policies. And so I'm going to be able to control for those and show that these other laws are not attributing this increase in mortality or decrease in HIV rates.
Analisa [00:31:38] And then finally, we might just wonder, is this a population composition change? So maybe when a syringe exchange program opens, injection drug users come to the area more. And so any increases in emergency room visits or mortality is just due to the fact that more people are there and more people are using injection drugs in that county. And so it's just that we're pulling in people from surrounding areas. So I want to highlight here that that's actually not what I'm picking up because the data is going to indicate county of residence. And so not a current. So even if I come over, say, from Butler County, Ohio, to Hamilton County, Ohio, and I overdose in Hamilton County, because my address is still in Butler County, then I wouldn't be counted in the mortality statistics for Hamilton. So this should help us worry less about the concern of the migration, say, of injection drug users, artificially increasing emergency room visits, or mortality rates, or even like thinking about potential spillovers to other counties. But just to get at this more directly, I also checked for population population composition changes. So I find no evidence that syringe exchange program openings affect total population or subgroups over time. And then I also look at things like changes in risky sexual behavior or changes in just general mortality rates. So I look at total mortality, alcohol related and vehicle related mortality, and I don't find any evidence that there are changes in other death rates or in other STD rates. So it's not just that the population at the time of the syringe exchange program opening is just changing in these meaningful ways.
Jennifer [00:33:20] Yeah, and I would want to talk a little bit more about the crime data, too, and the crime results. I think one of the challenges in studying this topic is, as you've mentioned, is, is getting data on what exactly you you're hoping to see. Right? Ideally, what we'd want to see in the data is all risky uses of opioids. And that, of course, is not available. You know, the surveys try to pick that up. But you mentioned all the problem with surveys. And so you can use mortality as a proxy. You can use overdoses as a proxy. Some people also use arrests for for possession or sale of drugs as a proxy for drug use. I think you make a really nice case in the paper that most of the places where a syringe exchange program is going to open work with law enforcement so that law enforcement isn't just like waiting outside the door and arresting everybody. Right? Like that would completely defeat the point of this kind of program. And so I think you make a good case in the paper that I believe you find increases in crime, although they're not huge, but that those would be real, like you would expect those to be biased downwards and law enforcement sort of backing off and treating this as a health crisis. So so yeah talk a little bit more about like what you find on the crime results and how you think about the the impact on arrests and theft, I guess, as well, in relation to the broader context of how law enforcement's dealing with this problem.
Analisa [00:34:44] Yeah, it's a good question. So the way that I look at crime outcomes in the paper is I'm using FBI uniform crime reports. And so those data include arrests for drug related crimes. So this would include things like opioid related drug sales or opioid related drug possession. And then I also do look at theft, not drug related theft, just theft in general, because there's some suggestive evidence that theft does change when when drug use also changes. So I look at that in terms of thinking about kind of property crimes. And essentially what I'm finding here is that if we look at just total drug arrests after a syringe exchange program opens, you see an increase of about 16.4 percent. So these effects are not visible in the first year that the program opens, but increases over time. And I find that these effects are really driven by drug possession rates. So I see no evidence that drug sale arrests increase after a syringe exchange program opening. So, as you mentioned, you know, we might worry that by building a syringe exchange program, you now have kind of a locus for cops to patrol if they're looking to increase arrests for drug offenses. And so basically what I'm finding is that it doesn't seem that because I don't see any effects on drug sales, it doesn't seem that police are just simply patrolling around these areas and catching kind of drug pushers in a way in the act. But in fact, we do see increases in drug possession and those are mostly concentrated in the first couple years after a syringe exchange program opens. So it does seem to suggest that that individuals are being caught with with drugs after at a higher rate after a syringe exchange program opens. That does seem to suggest a couple things. One, not it doesn't seem to be the case that when these programs open, there's like this new legal leniency. But it also doesn't seem to be like too much of a lashback. So the effects aren't particularly large. So we can think about then that it doesn't seem to be the case that cops are simply just hanging out, as you mentioned, outside of the program, but rather that individuals are maybe simply just doing more drugs and so they're more likely to get caught for these sorts of crimes now.
Jennifer [00:37:11] So you've mentioned a couple times that the effects seem to be driven by rural areas. Talk a little bit about how your effects overall are varying in different places. So urban versus rural, whether you find regional effects, whether effects vary by age or race or gender. All your your heterogeneity analyses.
Analisa [00:37:31] Yeah. So I do I am able to look at these effects across both types of individuals and types of counties just based on on the detail of the data. So, for example, if we split the analysis by urban and rural counties with syringe exchange programs, I find that syringe exchange programs are really highly effective at reducing HIV rates in urban counties with really smaller effects and actually statistically insignificant effects in the analysis in rural areas. So importantly, this could be due to the fact that blood borne illness can just spread more easily in cities due to high population density, or the fact that encountering a used needle is just more likely for a non drug user in an area with a lot of people or foot traffic. But importantly, the results for opioid related mortality are driven almost entirely by rural areas. So what I find is that opioid related mortality increases by 25 percent in rural counties. And this is really important because rural areas are less likely to have substance abuse treatment available. And also distance to a hospital might just be a major barrier in preventing fatal overdoses or just reversing the symptoms of overdose. So if we think that syringe exchange programs lower the cost of getting injection drug use supplies and therefore lower the cost of just using injection drugs but can't seem to be able to refer or counsel people in these rural areas because there are these other financial or transportation barriers, then we might be worried that syringe exchange programs are just likely to affect this vulnerable population and that maybe we should focus either having other programs in rural areas or focus our efforts on syringe exchange programs in urban areas where we think that HIV is not only more prevalent, but also can be more easily spread.
Analisa [00:39:32] So I also look at other types of areas of the country that might be most affected by the opening of a syringe exchange program. So, for example, I separate out the states that the CDC has identified as the top states affected by the opioid crisis. And effects are about double in this area. So those individuals that are really struggling or those states that are really struggling to address the opioid crisis are seeing the biggest effects of syringe exchange programs in those areas. So if we turn to looking at individuals most affected by syringe exchange programs, I find that these these results really aren't that surprising because I basically just find that syringe exchange programs affect not only the people that are most likely to use them, but also the most likely that in the data we've seen experience the harshest effects of the opioid crisis. So this includes individuals that are age 30 to 49. And I find effects for white individuals. So the effects that I find are pretty much entirely driven by white Americans. So I find no evidence that syringe exchange programs have adverse effects for those in their 20s or for African-Americans or even for Hispanic Americans. So what we're seeing here is that, again, syringe exchange programs are really kind of having the biggest effects on those that have experienced the opioid crisis, I guess, or those that are in the demographics that are more likely to have experienced the opioid crisis and also in individual for individuals that live in in rural areas, most specifically.
Jennifer [00:41:10] Yeah. The finding about the rural urban divide and the potential mechanism that access to substance abuse treatment is driving that difference is really striking and keeps coming up in this conversation. So, as you know, I have a paper with Anita Mukherjee that looks at the effects of increasing access to naloxone. And we basically find that making opioid use safer increases rates of opioid abuse. But we also find suggestive evidence that access to substance abuse treatment can help mitigate that unintended consequence. So both of our papers now are suggesting that we really need to get serious about increasing access to substance abuse treatment in local communities and finding ways to get people to actually engage in that treatment. So and, of course, you could test directly for whether places with less substance abuse treatment available are differentially affected by syringe exchange programs. I can't remember. Do you do that in the latest version of the paper, or are you just inferring that from the urban rural difference?
Analisa [00:42:02] Yeah, it's definitely on the to do list. So I think it's a look at counties that have substance abuse treatment facilities directly. I haven't started that analysis yet. But essentially what I do find looking at a paper by Isaac Swensen is that rural areas are much less likely, I can't remember the exact percent, I want to say something like two times less likely to have substance abuse treatment. So what I do think I'm picking up there is truly an effect on substance abuse treatment. But it is a good question. I have definitely made plans to look at that directly, but I haven't included that in the latest draft.
Jennifer [00:42:39] The beauty of working paper is you keep iterating. But yeah, I agree. It sounds very likely that that's that's the driving factor here. So the results are obviously a little bit of a mixed bag. Right? I mean it's a classic case of tradeoffs existing in a policy space. So you're finding that these programs reduce HIV rates, but they increase opioid related mortality and opioid related overdoses and opioid abuse more broadly. So can you say anything about the net effect of this policy? So on average, have these programs increased or decreased mortality? And I guess another question that that I know has come up in conversation about this paper is kind of a little bit more depressing. To what extent do you think the increase in opioid related mortality itself is what's causing the decline in HIV rates that you're finding?
Analisa [00:43:29] Yeah, it's kind of a grim question, but I think the benefit of being an economist is that we can take that in and still use the skills that we have to to think about those important questions. So what I can say about this is really, you know, in the early days of the AIDS crisis, contracting HIV was essentially a death sentence. So those that were infected expected to live only about eight years longer. But now that's really not necessarily the case. So there's dozens of drugs available to treat HIV and they really suppress suppress the chances of stage 3 HIV or AIDS. And today, people with HIV are likely to have a lifespan that's generally the same as uninfected people. So importantly, that we might think that if those statistics really rely on individuals regularly taking medication and also having access to health care to get the medication in the first place. So if we think that injection drug users are less likely to access HIV testing and know that they have HIV and or get medication, we can expect that these folks are are more likely to have a shorter lifespan. So that being said, you know, some studies find the injection drug users aren't any less likely to take the medication than people in the general population. So that could be some more of an empirical question. But, you know, getting back to the original question, in the case of my study, I find that syringe exchange programs reduce HIV diagnosis by about 30 cases per county per year, but increase opioid related deaths by about 13 per county per year. So just first of all, for the increase in mortality to be driving the changes in HIV, it would have to mean that nearly 50 percent of the clients in syringe exchange programs contra yeah contract HIV at some point. But if we look at just the patient data that I'm able to get from the small rural clinic in Ohio, only about one percent of those users have HIV through the duration of the data. So in my opinion, it seems unlikely that the increase in mortality is really driving this reduction in HIV and is probably a very small proportion of the reduction.
Analisa [00:45:42] So what I think is actually more convincing than just looking at the proportion of the decrease in HIV is that, you know, what I was mentioning earlier is that effects are concentrated in the study in different areas for HIV and opioid related mortality. So I find that HIV falls in urban areas, but increases in mortality are driven by the rural areas. So to me, what that indicates is that people that are living in these different locations are affected by these programs differently based on these financial or transportation hurdles to medical care or drug counseling or as you mentioned earlier, substance abuse treatment facilities. So to me, what this suggests is that syringe exchange programs just have fewer unintended consequences in urban areas where they're more easily able to provide additional services and really have a great potentially greater success in getting people into drug counseling because there are just close services available. So if we take those results together on net, it seems to suggest that syringe exchange programs do have large costs associated with increased overdoses compared to the benefits of reducing blood borne illness. But the results that are differential across rural and urban areas, have kind of some, I think, stark policy implications. So potentially using existing networks of clinics and infrastructure of syringe exchange programs to offer additional treatment in rural areas could potentially curb some of these large societal costs. So to sum that up, what I'm trying to say here is that it doesn't seem like the increase in mortality is causing the decrease in HIV, but rather just that syringe exchange programs are having differential effects based on where they are.
Jennifer [00:47:29] Yeah, and it's interesting to think about how the tradeoffs change here over time, right. So in during earlier decades when HIV was the main crisis, an intervention that reduces the incidence of HIV is totally is has huge social benefits, even if maybe it increased drug use a little bit. It's you know, you're probably still on net saving a huge number of lives. And then you kind of fast forward to the present day when the cost of, as you said, the cost of contracting HIV is actually much lower and perhaps very low. And but now we're super worried about anything that encourages drug use. And so suddenly we're in this this this context where the cost benefit of this type of program is much more heavily weighted toward the costs. And so finding ways to mitigate those costs seems really crucial here.
Analisa [00:48:22] Yeah, I totally agree.
Jennifer [00:48:24] So are there other policy implications that you have in mind for your paper or the other working in this area, what you know, when policy makers come to you and ask, what should they do? Should they shut down their syringe exchange programs? Should they should they do something else? Like, what's your what's your response?
Analisa [00:48:40] Yeah. So, you know, I think that a lot of people have looked at this paper thus far and thought, wow, this author must just be so against syringe exchange or have some sort of agenda or even go as so far as to suggest that maybe I've been paid by syringe exchange critics or that I'm an advocate against syringe exchange, or that the study implies that we should stop funding syringe exchange altogether. But I really want to focus on the fact that in reality, what the data shows here is that surrender exchange programs have tradeoffs that we really need to consider and that they do seem to be very effective in reducing blood borne illness. And so if the main public health policy concern is getting used needles off the street, then syringe exchange programs maybe should be funded. But it's not to say that we should just throw all this money at syringe exchange without thinking about whether or not they're the best tool to reduce addiction, especially in these rural and high poverty areas that might have less access to substance abuse treatment facilities. So really, for anyone who truly cares about getting injection drug users the treatment that they might need, then we need more research to show what interventions work instead of just really investing in programs that might have some upside and some of the outcomes that we study, but could also have these tradeoffs that we really care about and might actually be something that's super costly in terms of societal welfare.
Analisa [00:50:12] Now, that being said, you know, syringe exchange programs are meant to provide a safe and judgmental free zone for injection drug users to get sterile supplies and to potentially talk to a friendly face. So the goal of syringe exchange is to reduce blood borne illness, and it's to allow injection drug users access to someone who could suggest counseling or referral if they need it. And so, again, I want to emphasize that surge exchange programs are actually really successful at doing that. They achieve their intended goals, but they're not effective in addressing opioid addiction. So where we think that injection drug users might lack the resources necessary to get treatment, I just want to say that from a policy standpoint, it seems that we might need some sort of two pronged or multipronged approach of syringe exchange programs plus something like medically assisted treatment. And that seems to be a more promising way, both for reducing syringe sharing and to get dirty needles off the street, but also provides an avenue of treatment for those that need it. And just making sure that individuals in low income or in rural areas can get treatment if either at a facility or some sort of outpatient clinic if they need it.
Jennifer [00:51:25] Yeah, I totally agree that we need a lot more research on what works to get people into substance abuse treatment and to improve their long term outcomes. Do you have other research frontiers in mind here? What are the next questions aside from that one that need to be answered, that you yourself or others are going be working on in the years ahead?
Analisa [00:51:45] Yeah, so when we started, I mentioned this isn't going to be my my last paper on the opioid crisis because I just think that this topic is so important and that there really are so many more questions to be answered. So first of all, you know, I'm unable to look at drug use directly here. I think that's one of the major kind of criticisms or limitations of the paper. So really, the most important question that still remains is how much can syringe exchange programs affect injection drug use? And the other question that I've kind of hinted at is how effective can these programs be when they're combined with other interventions? So can we estimate things like spillover effects in terms of used needles to non injection drug users? Can we look at things like hepatitis C, which is more prevalent than HIV, but we don't currently have county level data on outcome. And then, you know, more broadly, I think that future research can just do so much more to answer the question of how can we combat the opioid crisis. So to date, 120 people are dying per day of opioid related overdoses. And we know that things like supply side policies, like pharmacists and doctor restrictions and prescription drug monitoring programs have really been ineffective at slowing rates of opioid related mortality. So if we think that demand-side policies like syringe exchange programs or like substance abuse treatment facilities are the way to go, then we need to invest more in that and see if those are a great avenue of trying to slow rates of opioid related mortality. And then also, I'm really interested in thinking about how medically assisted treatment or new prescription opioid antagonists might be an important way forward to slowing increasing rates of opioid related mortality as well.
Jennifer [00:53:34] My guest today has been Analisa Packham of Miami University. Analisa, thanks so much for doing this.
Analisa [00:53:39] Thanks so much. I appreciate it.
Jennifer [00:53:46] You can find links to all the research we discussed today on our website, probablecausation.com. You can also subscribe to the show there or wherever you get your podcasts to make sure you don't miss a single episode. Big thanks to Emergent Ventures for supporting the show. Our sound engineer is Caroline Hockenbury. Our music is by Werner and our logo is designed by Carrie Throckmorton. Thanks for listening and I'll talk to you in two weeks.