Marijuana reform

In October 2022, President Biden announced three steps on marijuana reform. His stated goal was to reduce the negative consequences of criminal convictions for marijuana possession (while leaving the consequences for trafficking, marketing, and underage sales in place):

“As I often said during my campaign for President, no one should be in jail just for using or possessing marijuana. Sending people to prison for possessing marijuana has upended too many lives and incarcerated people for conduct that many states no longer prohibit. Criminal records for marijuana possession have also imposed needless barriers to employment, housing, and educational opportunities. And while white and Black and brown people use marijuana at similar rates, Black and brown people have been arrested, prosecuted, and convicted at disproportionate rates.”

President Biden immediately pardoned anyone with a prior federal conviction for simple possession of marijuana. Next, since most such convictions are at the state level, he urged state Governors to do the same. Finally, he asked the Secretary of Health and Human Services to reconsider how marijuana is scheduled under federal law, presumably to move it from a Schedule I drug (the classification of the most dangerous substances) to a less-restrictive schedule or even to legalize it.

In this survey, we asked the Criminal Justice Expert Panel to consider three statements about marijuana reform. Their responses are below. (Survey conducted in April 2023.)

Responses

Pardoning federal convictions for simple possession of marijuana will have meaningful social benefits that exceed any social costs. - participant responses

Participant	Vote	Confidence	Comment
Alex Tabarrok George Mason University	Agree	7	People convicted of possession in Federal Court may not be a random sample of simple users and may in fact have committed more serious crimes that the prosecutor couldn't prove--like convicting Capone of tax evasion. Still that's not the way the system is supposed to work and I approve overall of the change.
Aaron Chalfin University of Pennsylvania	Neutral/No Opinion	8	The pardon is largely symbolic as there are very few people who have federal convictions for marijuana possession.
David Kirk University of Oxford	Neutral/No Opinion	5	If there is a benefit, it is likely indirect by signaling a different federal position and approach on marijuana.
John MacDonald University of Pennsylvania	Neutral/No Opinion	10	This is a ceremonial pardon, as there are very few people with a federal conviction for simple possession only. On the other hand, there are no real social costs to public safety.
Megan Stevenson University of Virginia	Strongly Agree	10
Benjamin Hansen University of Oregon	Strongly Agree	10	There are vast economic and racial differences in prosecution for marijuana possession. Pardoning these convictions will help people to have more economic opportunities.
David Harding University of California - Berkeley	Strongly Agree	8
Christopher Blattman University of Chicago	Strongly Agree	8
David Weisburd George Mason University	Agree	7
John Donohue Stanford University	Agree	8
Tom Clark Emory University	Agree	7	Otis unclear how many will benefit from this move.
Michael Makowsky Clemson University	Strongly Agree	8	While I have some uncertainty regarding the magnitudes of the benefits from these pardons given the paucity of federal convictions, I am quite confident that the costs will be trivial, resulting in a meaningful net improvement in the lives of those with possession records and their families.
Stephen Billings University of Colorado - Boulder	Agree	9
Charles Loeffler University of Pennsylvania	Agree	8	The primary effect of this policy will be limited to assisting in the restoration of civic rights to individuals previously federally convicted of simple marijuana possession. This policy change is unlikely to lead to substantial changes in time served in prison or on community supervision among those formerly convicted of this federal offense.
Robert Apel Rutgers University	Agree	8	I don’t expect large-scale benefits, just because there are so few people convicted of marijuana possession (or any drug possession, really) in federal court. It’s a drop in the bucket compared to trafficking, but if we’re going to pardon marijuana possession we should pardon all federal drug possession, as a start. A more impactful result of federal pardons will be to reduce sentence lengths in federal prison by eliminating criminal history points for people with past possession convictions.
William Horrace Syracuse University	Agree	8	There may be short-run hiccups, but this policy makes sense in the long-run.
Randi Hjalmarsson University of Gothenberg	Neutral/No Opinion	8	Though I would think this benefits the pardoned individuals without introducing costs, I don't think the social benefits will be large, in part because simple possession of marijuana just does not make a large share of federal cases.
Peter Reuter University of Maryland	Strongly Agree	10	It is hard to identify any social costs.
Rosanna Smart RAND Corporation	Disagree	8	Given relatively few people are convicted of simple possession of marijuana at the federal level, this is unlikely to produce large benefits. Without expungement, labor market costs/barriers associated with a criminal conviction are unlikely to be meaningfully affected. While prior research shows federal signals can have important affects on state cannabis markets, this signal is likely not meaningful enough to generate large changes. Overall, both benefits and costs are likely small.
Ariel White Massachusetts Institute of Technology	Neutral/No Opinion	9	I don't perceive large social costs to this reform; I'm just unsure how large/"meaningful" the social benefits will be. A small number of people will be directly affected by the president's actions, and even the number that would be affected if all states were to follow the federal government's lead would be quite small relative to the size of the criminal legal system. This doesn't mean it's a bad idea! I just don't want to overstate its significance compared to other possible reforms.
Giovanni Mastrobuoni Collegio Carlo Alberto	Agree	8
Kevin Schnepel Simon Fraser University	Agree	9	While there is not a great deal of causal evidence suggesting large long-term benefits to expungement, any social costs associated with this reform are minimal. The largest benefits would likely be among individuals with very recent convictions and without convictions for other offenses.
Paul Heaton University of Pennsylvania	Agree	4
Morgan Williams, Jr. New York University	Agree	5
David Abrams University of Pennsylvania	Agree	4
Santiago Tobón Universidad EAFIT	Strongly Agree	10
Anna Bindler University of Cologne	Neutral/No Opinion	5	My understanding is that the pardoned offense would not be removed from the person's criminal record, but the pardon would appear in addition. I find it hard to say what the expected social benefits (or social costs) of this would be.
J.J. Prescott University of Michigan	Agree	8
Anna Harvey New York University	Strongly Agree	9	Conviction records for nonviolent offenses appear to impose large net social costs (Mueller-Smith and Schnepel 2021), so any step towards reducing the visibility/impact of nonviolent criminal records is likely to have positive social benefit.
Jillian Carr Purdue University	Strongly Agree	9
Jamein Cunningham Cornell University	Agree	5	Not sure if it will have meaning social benefits, but the benefits should exceed the cost. Although pardoning federal convictions is a step in the right direction - it does not reduce the social harm already inflicted by previous contact with the criminal justice system. Also, the federal government plays a much smaller role in convicting and incarcerating individuals for drug-related offenses.
Manisha Shah University of California - Los Angeles	Agree	2
Emily Owens University of California - Irvine	Agree	7	This will primarily be a change in the spirit of Federal policy; Federal convictions for simple marijuana possession are just not really something that happens very often.
Shawn Bushway RAND Corporation	Strongly Agree	10	I don't think the benefits will be large, but they will have a meaningful impact in some communities.
Robynn Cox University of California, Riverside	Strongly Agree	9
John Pepper University of Virginia	Agree	7
Felipe Goncalves University of California, Los Angeles	Agree	10	I support the policy, but the impact will be very small. There are few standalone federal convictions for marijuana possession.
Greg Midgette University of Maryland	Disagree	9	Simple marijuana possession convictions are very rare today and sentences are pretty short. So, there's some notional value to removing a rarely enforced law from the books, but it would affect few recently convicted people. People who were sentenced a long time ago already paid the direct penalty of incarceration and the secondary costs of being labeled a felon. A long delayed pardon has few costs, but benefits might accrue mostly to politicians.
Jennifer Doleac Texas A&M University	Neutral/No Opinion	9	There are very few such convictions, so this is mostly a symbolic move. But the costs should be minimal.

Pardoning state convictions for simple possession of marijuana will have meaningful social benefits that exceed any social costs. - participant responses

Participant	Vote	Confidence	Comment
Alex Tabarrok George Mason University	Strongly Agree	8	More state simple possessions are simple possessions.
Aaron Chalfin University of Pennsylvania	Agree	7	If one takes the normative position that marijuana use has few social costs (or, at a minimum, that a series of pardons won't stimulate further use), then it seems likely that any benefits will outweigh the few costs that there are. A number of states have already decriminalized marijuana so concerns about the effect of pardons on use are mostly moot. In those states, there is an especially strong reason to see pardons as reasonable given that marijuana possession is no longer a crime.
David Kirk University of Oxford	Neutral/No Opinion	5	Few people in state prison for simple possession. If the government wants to reduce the use of incarceration further, it needs to think about its approach to violence.
John MacDonald University of Pennsylvania	Neutral/No Opinion	10	Very few people end up in prison or jail on possession alone without significant other criminal histories. However, given that marijuana is essentially de-crimininalized in half of US states and widely used, it makes little sense to allow a conviction for possession to remain on someone's record.
Megan Stevenson University of Virginia	Strongly Agree	10
Benjamin Hansen University of Oregon	Strongly Agree	10	There are vast economic and racial differences in prosecution for marijuana possession. Pardoning these convictions will help people to have more economic opportunities.
David Harding University of California - Berkeley	Strongly Agree	8
Christopher Blattman University of Chicago	Strongly Agree	8
David Weisburd George Mason University	Agree	7
John Donohue Stanford University	Agree	8
Tom Clark Emory University	Strongly Agree	9	This would be a much bigger event than the federal version.
Michael Makowsky Clemson University	Strongly Agree	9	State convictions of simple possession being far more common than federal, I think the benefits from pardoning state convictions stand to be significant. I am quite confident that the costs will be trivial, resulting in a meaningful net improvement in the lives of those with possession records and their families.
Stephen Billings University of Colorado - Boulder	Agree	9
Charles Loeffler University of Pennsylvania	Agree	8	Pardons of past convictions for simple possession of marijuana will have a first order effect on voting/jury service. Only when paired with expungement or other criminal record clearance policies is it likely to have a larger social impact. That behind said, the costs of pardoning are primarily administrative.
Robert Apel Rutgers University	Strongly Agree	8	There is domain similarity with expungements, which have evidence for quite substantial social benefits—employment and wage gains, and even lower recidivism. Unfortunately, uptake for these kinds of initiatives is low, so they should be automatic to remove any administrative burden. If pardons (along with decriminalization, if not legalization) are too high a political hurdle, then reclassification from felony to misdemeanor should be prioritized.
William Horrace Syracuse University	Agree	8
Randi Hjalmarsson University of Gothenberg	Agree	7
Peter Reuter University of Maryland	Strongly Agree	9	The loss of benefits and barriers to employment consequent on such convictions has never made any sense.
Rosanna Smart RAND Corporation	Agree	7	This would likely affect a much broader set of individuals, although the extent to which meaningful social benefits derive will depend on the extent to which convictions are fully expunged, processes for doing so (e.g., automatic or individual must apply), and the share of people with only simple possession of marijuana convictions on their records. Social costs will also depend on these factors (e.g., costs for verifying eligibility and sealing or expunging records) but are likely minimal.
Ariel White Massachusetts Institute of Technology	Neutral/No Opinion	6
Giovanni Mastrobuoni Collegio Carlo Alberto	Agree	8
Kevin Schnepel Simon Fraser University	Agree	9	While there is not a great deal of causal evidence suggesting large long-term benefits to expungement, any social costs associated with this reform are minimal. The largest benefits would likely be among individuals with very recent convictions and without convictions for other offenses.
Paul Heaton University of Pennsylvania	Agree	6
Morgan Williams, Jr. New York University	Agree	6
David Abrams University of Pennsylvania	Agree	4
Santiago Tobón Universidad EAFIT	Strongly Agree	10
Anna Bindler University of Cologne	Neutral/No Opinion	5
J.J. Prescott University of Michigan	Strongly Agree	9
Anna Harvey New York University	Strongly Agree	9	Conviction records for nonviolent offenses appear to impose large net social costs (Mueller-Smith and Schnepel 2021), so any step towards reducing the visibility/impact of nonviolent criminal records is likely to have positive social benefit.
Jillian Carr Purdue University	Strongly Agree	9
Jamein Cunningham Cornell University	Agree	7	Pardoning individuals is just one step. Decriminalizing marijuana possession will only have positive spillover effects if local police do not shift patrolling activities to target quality-of-life offenses. If so, a less punitive approach to marijuana would likely be offset by a more aggressive approach to combating other criminal offenses. Therefore, decriminalizing marijuana should also be accompanied by new policing strategies to deter serious offenses.
Manisha Shah University of California - Los Angeles	Agree	2
Emily Owens University of California - Irvine	Strongly Agree	9	This would be a much bigger deal in states that have not yet done this, although again, it depends on how what such convictions have meant for people in practice. In terms of incarceration (versus police contact), generally these convictions primarily create/increase criminal history, which then has its own social and CJ consequences.
Shawn Bushway RAND Corporation	Strongly Agree	10	There are a relatively small group of people for whom their only conviction is a misdemeanor MJ possession charge. But, for those people, this could be very meaningful.
Robynn Cox University of California, Riverside	Strongly Agree	9
John Pepper University of Virginia	Agree	7
Felipe Goncalves University of California, Los Angeles	Agree	7	Many more people have state-level convictions for marijuana possessions, so the social benefits are potentially very large. My best guess on the potential social costs is that they would be small. The early literature on marijuana legalization has found small impacts on crime, and there may even be reductions if police shift their enforcement to more serious offending.
Greg Midgette University of Maryland	Agree	7	Criminal penalties toward simple possession probably have little deterrent or incapacitation effect on more serious crimes. About half of U.S states have not yet decriminalized cannabis possession, so many more people stand to benefit from state reform than a comparable federal reform.
Jennifer Doleac Texas A&M University	Agree	4	This won't reduce incarceration rates much, and it's currently unclear whether expunging/sealing such records has any benefit. (Initial evidence suggests there is little/no benefit in terms of employment, especially to people with less-recent convictions. Preventing the conviction in the first place would be more helpful.) But there should also be very little cost to this. Maybe there are some benefits that outweigh minimal costs?

Moving marijuana from a Schedule I drug to a less-restrictive schedule or legalizing it at the federal level would have meaningful social benefits that exceed any social costs. - participant responses

Participant	Vote	Confidence	Comment
Alex Tabarrok George Mason University	Strongly Agree	9
Aaron Chalfin University of Pennsylvania	Agree	6	The feds are less important in this arena than the states but marijuana enforcement is probably a poor use of federal resources. Major gangs and traffickers, who should receive attention from federal law enforcement, usually do not traffic in marijuana so I don't anticipate that this will compromise many investigations.
David Kirk University of Oxford	Agree	4
John MacDonald University of Pennsylvania	Strongly Agree	10	The less restrictive schedule is long overdue.
Megan Stevenson University of Virginia	Strongly Agree	10
Benjamin Hansen University of Oregon	Strongly Agree	10	Marijuana's scheduling is one of the chief obstacles in helping to reduce the negative externalities marijuana can have. As a schedule I drug, we are limited in the medical research we can do. Some people drive high, and this is a negative externality. Others drive having recently smoked in the last few days or weeks. Current tests don't distinguish between the two. At the same time, I worry about big tobacco taking over the industry, so care and appropriate regulation is needed.
David Harding University of California - Berkeley	Strongly Agree	6
Christopher Blattman University of Chicago	Strongly Agree	8
David Weisburd George Mason University	Strongly Agree	10
John Donohue Stanford University	Agree	8
Tom Clark Emory University	Strongly Agree	8
Michael Makowsky Clemson University	Strongly Agree	5
Stephen Billings University of Colorado - Boulder	Strongly Agree	9
Charles Loeffler University of Pennsylvania	Neutral/No Opinion	8	It is difficult to forecast the full knock-on effects of rescheduling or descheduling. Utilization will almost certainly increase with secondary impacts on everything from impaired driving to workplace accidents. At the same time, the costs of enforcement will decline somewhat as resources are reallocated to enforcement of other criminal statutes.
Robert Apel Rutgers University	Strongly Agree	8	Schedule I is reserved for substances with no accepted medical use and a high potential for abuse. It makes zero sense for marijuana to be on that list, for other drugs to be listed in a lower Schedule (e.g., methamphetamine), and for alcohol to not be listed at all. There have been no meaningful social benefits to it being listed in Schedule I, so it should not require proof of benefits from removing it. Law enforcement could maybe even boost clearance rates of more serious crimes.
William Horrace Syracuse University	Strongly Agree	8
Randi Hjalmarsson University of Gothenberg	Neutral/No Opinion	5
Peter Reuter University of Maryland	Agree	6	Not comfortable with having two such different and distinct policy options in one question. Schedule 1 is a relatively simple decision. Legalizing is much more complex. I am generally in favor on the basis of evidence from legalization so far but believe that long-term effects are still hard to predict.
Rosanna Smart RAND Corporation	Neutral/No Opinion	2	I can’t answer this because the HOW here is so important. The impacts of rescheduling are going to be fundamentally different than the impacts of federal legalization that (for example) allows commercial cultivation and retail sales, interstate commerce, etc. The likely effects of federal legalization itself will hinge critically on how the legal regime is designed and regulated as these specifics will shape price/availability, potency/product variety, and the functioning of illegal markets.
Ariel White Massachusetts Institute of Technology	Neutral/No Opinion	2	I think we know much less about the possible impacts of legalization
Giovanni Mastrobuoni Collegio Carlo Alberto	Agree	5
Kevin Schnepel Simon Fraser University	Agree	9	From evaluations of state-level marijuana legalization, it does not appear to be associated with increased social costs from increased use or negative changes in economic or social behavior.
Paul Heaton University of Pennsylvania	Neutral/No Opinion	8
Morgan Williams, Jr. New York University	Neutral/No Opinion	4
David Abrams University of Pennsylvania	Strongly Agree	5
Santiago Tobón Universidad EAFIT	Strongly Agree	10
Anna Bindler University of Cologne	Agree	3
J.J. Prescott University of Michigan	Agree	8
Anna Harvey New York University	Agree	7	Many nonviolent misdemeanor charges are for simple marijuana possession, and these charges are racially disparate. There is some evidence that prosecuting these charges has social costs (Agan et al 2023), and no evidence of corresponding social benefits.
Jillian Carr Purdue University	Agree	8
Jamein Cunningham Cornell University	Agree	8	Legalizing marijuana at the federal level has implications beyond the criminal legal system; for instance, it would open up the cannabis industry to banking, investment opportunities, and potential trade markets.
Manisha Shah University of California - Los Angeles	Agree	2
Emily Owens University of California - Irvine	Strongly Agree	10	At minimum, this would make it easier to conduct credible research on the costs and benefits of marijuana consumption, so society could have a more clear picture of the implications of expanded use.
Shawn Bushway RAND Corporation	Disagree	8
Robynn Cox University of California, Riverside	Strongly Agree	7	Evidence is still relatively new here but so far evidence on recreational legalization and MML laws suggest legalization has either no effect or may even reduce crime. Although Blacks make up a disproportionate share of marijuana arrests, it is not clear whether passage of such color-blind laws will actually decrease racial disparities in arrests for marijuana. Finally, marijuana has medicinal purposes so it is debatable whether it should be a schedule I drug to begin with.
John Pepper University of Virginia	Neutral/No Opinion	3
Felipe Goncalves University of California, Los Angeles	Agree	7
Greg Midgette University of Maryland	Agree	9	We've probably learned enough from states with recreational markets to be able to design a market for recreational cannabis that is much less harmful than the legal market for alcohol. Rescheduling cannabis could spur federal research funding, and in turn replace some of the pseudoscience that currently informs consumer decisions. Potential crime, (e.g. robbery, fraud) associated with paper cash-based markets could also be mitigated without huge expenditures on private security.
Jennifer Doleac Texas A&M University	Agree	5	Prosecuting people for non-violent misdemeanors increases recidivism (see Agan et al, 2022), but that doesn't mean legalizing those behaviors would have the same benefits (perhaps the arrest & court appearance are sufficient punishment). I also worry that many simple possession arrests are an effort to get a particular person off the street; legalizing marijuana could simply mean police officers arrest them for something else (disorderly conduct, trespassing). This would reduce benefits.