How Substitutable are Marijuana Possession Arrests?

Many times in the history of the United States, rising prevalence of use of a drug has been responded to with increased arrests of people in possession of that drug. As Peter Reuter and Rob MacCoun pointed out some years ago, this makes the rise of marijuana possession arrests in the U.S. particularly puzzling: The rate of marijuana use was historically low and stable prior to the growth of arrests.

The trend began in the 1990s and has continued to the present day. It has been almost entirely an urban phenomenon, with the best example being New York City (For a wonderful series of charts on NYC’s arrest patterns, see this piece by Michael Keller).

No informed observer seems to believe that this rise in arrests resulted from heightened concern about marijuana use per se. Rather the explanations invoked range from the arrests being a side effect of stop-and-frisk policing, to them being a way for those police who are racially prejudiced to hassle people of colour, to them being a tactic to punish malefactors who have gotten away with something worse (e.g., someone who has repeatedly beaten a spouse who is too afraid to testify against the perpetrator). Some people are sure they know the explanation with absolute certitude, but if like me you admit the possibilities of your own ignorance and life’s complexity, please read on.

In studies of drug use, researchers employ the concept of “substitutes” for various drugs (i.e., a different drug with a similar effect). It works out more neatly in theory than in everyday life, but a well-founded aspect of the theory holds that removal of a drug from the market can result in its users seeking a substitute. For example, if a state tightens up regulations on doctor shopping, people addicted to opiate pain medication may start seeking a substitute, such as heroin. It would be useful to know whether this same phenomenon occurs not just for drugs, but for drug possession arrests.

The question of interest is whether legalisation or decriminalisation of marijuana (e.g., in a state or nation) would affect the total number of arrests for all crimes, or, whether any drop in marijuana possession arrests would simply be substituted for with other charges. That is, if you were a police officer bent on hassling young males of colour, would marijuana legalisation/decriminalisation affect you in any way, or could you just start charging the same people with disorderly conduct/vagrancy (followed by, if you were really in a bad mood, resisting arrest). Alternatively, if you were a policeman who took sympathy on a young male of colour caught with cocaine and marijuana by only charging on the latter, would you respond to marijuana legalisation by substituting the cocaine charge rather than letting the person go free?

The answers to these sorts of questions have huge implications for understanding the impact of marijuana legalisation/decriminalisation, both in terms of fiscal impact but also in terms of whether the population of users will have less, more or the same amount of contact with the criminal justice system. So, how would one get those answers?

My own approach would be to interview a random sample of 1000 police officers who had recently made a marijuana possession arrest. The arrest record would be examined by the officers and the interviewer (under strict confidentiality), with the interviewer asking “If you could not have arrested this individual for marijuana possession, what would you have done instead?”. Granted, some drug policy reform advocates do not care about data, but there are enough who do to make such a study of the substitutability of marijuana possession arrests of great value.

Author: Keith Humphreys

Keith Humphreys is the Esther Ting Memorial Professor of Psychiatry at Stanford University and an Honorary Professor of Psychiatry at Kings College London. His research, teaching and writing have focused on addictive disorders, self-help organizations (e.g., breast cancer support groups, Alcoholics Anonymous), evaluation research methods, and public policy related to health care, mental illness, veterans, drugs, crime and correctional systems. Professor Humphreys' over 300 scholarly articles, monographs and books have been cited over thirteen thousand times by scientific colleagues. He is a regular contributor to Washington Post and has also written for the New York Times, Wall Street Journal, Washington Monthly, San Francisco Chronicle, The Guardian (UK), The Telegraph (UK), Times Higher Education (UK), Crossbow (UK) and other media outlets.

33 thoughts on “How Substitutable are Marijuana Possession Arrests?”

  1. Excellent question. My guess (stress on guess, I’m not thinking of any particular data set or natural experiment) would be that the substitutes are imperfect — nothing is quite so convenient or easy to prosecute as possession of contraband. And in many if not most states, marijuana possession still carries a greater sanction, or at least potential sanction, than would some possible substitutes (disorderly conduct, loitering, etc.) If that guess is right, then the substitution effect you raise would undercut some but not all of the savings of various sorts.

    But one factual point: MJ use was more or less stable for an extended period, including years when arrests for marijuana were rising. But it has not been stable for the last decade. The following changes pertain to 2002 – 2010 since 2002 was the first year NSDUH replaced NHSDA, so we have a consistent survey instrument.

    Just a 15% increase in total users (and about half of that could be chalked up to growth in the total population).
    But growth has been uneven.
    People using in the past year but not past month: up only 5% (less than the population growth rate).
    People using 20+ times in the past month: up 52%.
    The latter group dominates quantity consumed, hours of intoxication, etc., so total HH self-reported days of MJ use were up 40% between 2002 and 2010.

    Now, I don’t think that explains the MJ arrest trends, so it doesn’t undermine the “mystery” of why MJ arrests have risen or the realization that the # of MJ arrests is not just the amount of MJ use times some universal constant like pi or e. My sense is that the increase in MJ use may not come in the same demographic groups that have seen the increase in arrests; it may (back to guessing) come from delaying the maturing out of MJ use. (At least it does not appear to be explained primarily by increased use by youth.)

  2. or could you just start charging the same people with disorderly conduct/vagrancy (followed by, if you were really in a bad mood, resisting arrest)

    One could, but wouldn’t an increased racial skew in the number of arrests for such notoriously catch-all charges be just a bit more blatantly malicious? Presumably this would lead to at least some increased political costs for the guilty forces.

    I’m not altogether convinced by your proposed experiment – partly because it’s not obvious why the police officers would trust your promise of confidentiality, and partly because I’m not convinced that all officers making bullshit arrests are being cynically honest with themselves as to their actual motives to action. Therefore, I’d expect the results to understate the level of substitutability significantly. If this is correct, I think it would be hard to produce any result that would persuasively reject substitutability, and hence by your logic support decriminalization. How would you try to falsify your hypothesis of substitutability, given this pitfall?

    1. What do people get charged with in jurisdictions where marijuana possession has been decriminalized? Do they get charged with other contraband, or other types of offense entirely? (Note also that in many cases the stop and search are sufficient punishment, so actual arrest and booking may already be a special case.

    1. MikeM: I assume you mean the factor in the large increase in heavy use that Jon Caulkins notes, to which I say that is certainly possible and is also certainly testable. Another explanation could be a fall in price within the “downmarket”, which is where most heavy users buy their marijuana.

  3. It would be interesting to add a question on the officer’s own views on marijuana. Those who were dedicated drug warriors would be likely to find other charges to arrest all smokers, not just minorities.

    1. I like that idea, although I would expect very few cops on the beat would be “dedicated drug warriors” as you put it.

  4. Under the misbegotten doctrine of “civil forfeiture”, the arresting police are sometimes empowered to sieze cars, boats, cash associated with a marijuana arrest, to auction the siezed assets, and to keep the proceeds in the department, whether or not the arrestee is ultimately convicted (or in some cases, even charged) with drug posession.

    All the incentives in this situation encourage police corruption.
    I’d be surprised if this wasn’t a factor in the increased arrests for cannabis posession.

    1. Does anybody know of good resources on civil forfeiture rates by state? I’ve read about extreme levels of predatory forfeiture in Florida and Texas but I don’t really have any idea how common this sort of thing is throughout the country. We would want to know forfeiture rates and policies in NYC if we want to estimate the impact of forfeiture on drug arrests in NYC.

    2. Joel hanes: I think police are generally way better human beings and professionals than you are giving them credit for. Even apart from that, I don’t think your argument holds water because (a) the civil forfeiture laws were put in place in the 1980s not the 1990s, and (b) I am not sure civil forfeiture would drive a corrupt officer to arrest low-income males of color for pot possession on street corners. I doubt many of them have yachts…if a police officer just wanted to grab things, s/he would go for people with more assets.

      1. I suspect, either Joel or Keith is being naive. Civil forfeiture is a serious issue in driving up non-arrests by police because the initial forfeiture laws were written to allow a mere suspicion to shift the burden of proof on the suspect. 60 Minutes and others have done several profiles of the Forfeiture Alleys, particularly in the South, although some exist in the Midwest as well. To claim that forfeiture has no effect on arrest records is disingenuous. This is not, however, what Keith said–what I see is a claim that since average minority street mj smokers do not have much to forfeit, the “corrupt” cops would not bother with them. This is not the case. Forfeiture Alleys target largely minorities because they have few resources to fight the claims against them. It’s the volume vs. margin issue–yes, busting a drug dealer with homes and yachts would be more efficient, in terms of claiming forfeiture funds, but nickel-and-diming random drivers is also effective, albeit in a different way. And, at least the 60 Minutes story, clearly pointed that the stops target minorities. I have not seen similar data on the Forfeiture Alleys around casinos, but I’d be willing to bet the demographics would be similar. On the other hand, there is difficulty here. First, there is often no need to arrest to claim civil forfeiture and the arrest is almost never for possession–it’s for suspicion of distribution or trafficking or money laundering. The key word is “suspicion”, because if the case never goes to trial, it’s the suspect who has to initiate a court case. So I suspect that 1) civil forfeiture IS an issue, although 2) it’s not as significant as Joel believes, but 3) the attraction of “free money” is significant enough that Keith is easily wrong in his opinion of beneficence of “police”. I would say that Joel has overstated his case, not that his argument “doesn’t hold water”.

        1. ShadowFox: Keith is easily wrong in his opinion of beneficence of “police”

          Obviously, you are entitled to your own view, but I have known too many law enforcement officers to take such a negative perspective on them as a profession (and you have not proved your view simply by asserting it, even though many people in the drug advocacy blogosphere no doubt agree strongly with your negative view of law enforcement).

          1. Keith :

            I too know law enforcement officers. Most are wonderful people.

            But most is not all, and there are many places in the United States where the local constabulary comprises people you should not trust.

            You’ve heard of Joe Arpaio ? He’s an outlier, but it’s a long tail.

            None of the fine men and women that you and I know would have committed the Danziger Bridge shootings in the aftermath of Katrina — but someone did, and those men were cops.

          2. Joel — We agree then, as your view is where I started with the post, by acknowledging that some police officers abuse their authority. I am pushing back only against those who seem to extend this reality into generalized cop-bashing.

      2. “……if a police officer just wanted to grab things, s/he would go for people with more assets.”

        You mean assets held by people with good lawyers.

  5. Why not look at what happened in Massachusetts when possession of small amounts of marijuana was decriminalized in 2009? Here’s the first relevant site I found: the DOJ’s Easy Access to FBI Arrest Statistics: 1994-2009. Filter on Massachusetts, and we see that total arrests went from 24,395 to 23,639 from 2008 to 2009; Drug abuse violations went from 21,571 to 12,622; Disorderly conduct went from 9,012 to 8,838; Drunkenness 8,206 to 7,793; and “police harassment” arrests were extremely low (13->10 for “Curfew and loitering”; 19->20 for “Vagrancy”). For both years the “coverage indicator” was 98% so the numbers should be comparable.

    Anybody know of a site with more detailed information about the results of MJ decriminalization in Massachusetts?

    1. Urg, bad copy/paste! I meant to write that total arrests went from 170,990 to 159,256.

      1. Good link to data David, thanks. It looks like arrests went down further than drug arrests, which could mean a secular trend was at work, or that fewer drug arrests spills over into fewer arrests for everything.

        1. What changed here in Massachusetts was as much the status of marijuana as grounds for a further search and warrant check as the actual disposition. A ticket or throw-away and lecture was probably as common as an actual arrest before the law was changed – at least here in the burbs.

          1. Your observation gibes with what the research in this area suggests: Decriminalisation is often a formalisation of what police have been doing ad hoc anyway.

  6. You might be able to run this experiment by comparing cities that de-prioritize M.J. arrests.
    San Francisco, Berkeley, and probably others put M.J. offenses low on the priority list, since they can’t legally say they are not enforcing it.

    See if the hypothetical substitution effect exists.

    1. MobiusKlein: That’s a good idea, Seattle would be another site, as would that Colorado city (forget which) which has legalized, at least rhetorically.

      1. Just not sure what the proper control group would be.
        Same city over different time intervals,
        or similar city with different enforcement protocols.

        Neither is such a great control. If you could do it by precinct level….

  7. Keith,

    A methodological note, if you get to do the survey. Assuring people of strict confidentiality is all well and good, but survey respondents often doubt it. A randomized response (when properly explained) reassures respondents that even if their information somehow does get out, they can’t be tagged for the response.

    It has some efficiency costs, because the actual sample size is a fraction of the apparent sample size. On the other hand, taking away a reason to lie reduces that source of bias.

    1. Always wondered about that technique: If they don’t trust you to maintain confidentiality, why should they trust your explanation?

  8. OK… must ask: Have demographics of mj use changed over time? For other (“harder”) drugs, at any given time, demographics are usually fairly distinct. Is this the case for mj use and has there been a shift just before or since the spike in arrests occurred. It seems this can give indirect information on what is going on.

  9. Another question: What did prohibition era booze dealers turn to for a living when booze became legal again?

  10. I would add that the misperception that pot is legal – especially prevalent in California post-Prop 215 – may raise the arrest rate statistic without necessarily implicating the usage statistic simply by making public what was previously kept private.

    In my own county (one vertex of the Emerald Triangle), people who are too lazy or cheap to get 215 cards will still carry their “medicine” around in their car with them. (As an aside, I really hope that people who advocate for the legalization of pot will at least have the good sense to acknowledge that there is a time and a place, and that while driving in your car is not that time and place.) Anyway, they’ll get pulled over for something, and then, because pot has such an obvious and pervasive odor, the cop will have probable cause to search the vehicle. Police officers with good PC are almost always going to search – even if they don’t care about pot, they might find something that they do care about.

    Although there are racist cops and cops who hate potheads, there are also a couple of far less sinister reasons why this might subsequently turn into an arrest: 1) The person is a jerk to the cop, and what would have ended up as a seizure and a warning turns into a night in jail; or 2) The cop is low on his or her stats for the month, doesn’t want to hear about it from his or her sergeant, and ends up writing the person a citation.

    1. People who advocate for the legalization of marijuana DO acknowledge that there is a time and a place, and are, in fact, pushing for regulation of marijuana – something the black market does not do. It’s also something that’s very hard for the states to do when the feds keep interfering with their attempts to regulate.

      1. Well, I’m not going to argue with that; life in my county would be a lot easier if the feds would live up to Obama’s 2008 campaign promise.

Comments are closed.