A conversation with Mark Kleiman on July 2, 2013


Note: This set of notes gives an overview of the major points made by Mark Kleiman.


Good Ventures and GiveWell spoke with Mark Kleiman about drug policy and criminal justice reform. Topics discussed included approaches to reducing crime and incarceration rates, new markets for cannabis in Washington and Colorado, organizations working on crime and drug policy, and ideas for further research.

Reducing crime and incarceration rates

The US crime rate is the same as it was in 1965, but the country has 5 times as many prisoners. To get back to the 1965 prison numbers without crime going up will require doing a better job of punishing people and modifying their future behavior without confining them. Many people are concerned about this problem, but there are few ideas for how to improve the situation. Dr. Kleiman's suggestions for reducing crime and incarceration rates include:

  1. Reducing alcohol consumption and abuse through such measures as higher taxes on alcohol and prohibition of drinking among past abusers.
  2. Swift and certain sanctions to enforce conditions imposed on offenders not behind bars: on parole, on probation, and on pretrial release.

Reducing alcohol consumption

Drug and alcohol policy can have direct impacts on crime rates. Increasing the alcohol tax is a straightforward way to reduce crime and likely one of the most effective crime control policy steps that could be taken. It would require little administrative effort and would have an immediate effect on the homicide rate. There is also evidence that fetal alcohol exposure contributes to people ending up in prison. [This is an issue at exposure levels well below those that generate clinically diagnosed Fetal Alcohol Syndrome.] While raising alcohol taxes would be administratively easy, it would be difficult to do politically, because the vast majority of drinkers don’t suffer from alcohol-related problems and don’t feel that their drinking should be taxed more heavily, because public opinion opposes restrictions on freedom and opposes higher taxes, and because the alcohol industry would be strongly opposed to the change.

One way of dealing with problematic usage of drugs and alcohol would be to place a marker on the IDs of individuals with a history of problems that prevents them from purchasing the substance. Another would be to allow individuals to set themselves monthly quotas for alcohol (or other drug) purchases that they could only change with 30 days' notice. The latter approach would be a way of helping people commit to the plans that they wish to follow on reflection, while preventing them from reneging on commitments based on in-the-moment judgments.

Alcohol abuse is the most important public policy issue that Dr. Kleiman can think of that has no major advocate at the moment. While is a strong alcohol abuse prevention and treatment community, part of it has a “temperance” mindset rather than focusing on reducing abuse while allowing moderate use, while another part wants to treat the “disease of alcoholism” as if it were unconnected with normal social drinking patterns.

Swift and certain sanctions

A particularly effective way to prevent alcohol and drug use among abusers is "swift and certain" sanctions: frequently testing probationers for drugs or alcohol and implementing very short jail sentences or other punishments for failing on any occasion. People who get in trouble with drugs and crime tend to have trouble adjusting their current behavior to their long-term interests, so low-probability, high-severity punishments are not optimal. Swift and certain sanctions—mild but almost immediate—are therefore effective deterrents. The swift-and-certain model is relatively low cost and could be implemented on a national scale, unlike some other drug and alcohol treatment programs that are more expensive. Drug courts, for example, are unlikely to be implemented nationally because doing so would require too many judges and more treatment capacity than actually exists.

There have been 4 studies demonstrating the effectiveness of swift and certain sanctions programs. Swift and certain sanctions strongly outperformed mandated treatment in the DC Drug Court experiment. The HOPE program in Hawaii used swift and certain punishment to curtail illicit drug (mainly methamphetamine) use. The 90-day 24/7 Sobriety program in South Dakota used a similar system for alcohol and also had promising results, reducing DUI recidivism by 50% over two years among participants and reducing domestic violence countywide where it was implemented. The National Institute of Justice (NIJ) is currently funding replications of the HOPE program, and results should be available in roughly 2 years.

Despite the success of swift-and-certain-sanctions programs, there is limited funding for research or implementation of this model. For example, more research is needed to determine what the smallest effective punishment is.

Swift and certain sanctions are a genuine alternative to incarceration and could encourage judges and legislators to grant shorter sentences. An alternative but related approach would be to release convicts from jail early, under close monitoring and swift and certain sanctions.

Advocating for these policies

Both raising alcohol taxes and scaling up swift-and-certain-sanctions programs would likely have significant effects on crime and incarceration rates, but it would be more difficult to achieve higher alcohol taxes than to scale up swift and certain sanctions. The opposition to swift-and-certain-sanctions policies is a lot weaker than the opposition to raising alcohol taxes. While some people would prefer drug legalization or alternative criminal justice reforms to swift and certain sanctions, in general there is more skepticism than hostility towards the programs. Skepticism could be overcome with appropriate education efforts. Attempting to advocate for higher alcohol taxes would require fighting the alcohol industry lobby and changing public opinion on the status of alcohol as a dangerous drug and the need for higher taxes. Either could have a huge impact.

New cannabis markets in Washington state and Colorado

The opportunity

There is an immediate policy opportunity in helping Washington and Colorado to get their cannabis programs right, as both states are currently designing legal cannabis markets. Implementation of good policy in both states is important for the future of national drug and criminal policy. Washington hired a group of external consultants, including Dr. Kleiman, to advise it on how to implement its program, but the funding for that has run out. Colorado has not had outside help.

Without technical assistance or relevant advocacy, the cannabis market in Washington and Colorado is likely to end up looking like the alcohol market, where the industry is allowed to promote itself to heavy users with few policy constraints. Historically, alcohol went from being illegal to manufacture or sell to being an ordinary good that is entirely marketable (except to children). Ideally a middle ground could be found for cannabis. If not, there could be a backlash against cannabis legalization.

Generally speaking, the area of drug policy seems to have more low-hanging fruit than the area of criminal justice reform, though the latter is probably more important overall.

Dr. Kleiman's recommendations for Washington

Dr. Kleiman has been part of a group working with the state of Washington on their implementation of a market for cannabis. The work involved estimating the size of the market and making policy recommendations. Their recommendations included allowing home delivery and allowing individuals to set their own cannabis quota that can only be changed with 30 days' notice. A proposal that wasn't accepted but might be in the future was to limit cannabis production, and specifically to use THC as the basis for limiting it. An important aim was to get the state policy makers to focus on the outcomes of drug policy and not just the purchase process. With alcohol, regulators tend to focus on who can buy and sell alcohol, but with drugs and alcohol it is important to also consider the potential outcomes of any policy in terms of the costs and benefits to individuals and society.

The group proposed a number of potential projects that would require additional funding. One project would be to understand the process by which the cannabis market will bring in consumers: what they know about cannabis, how they acquire information, and what the state can do to communicate to users and dissuade abuse – including how to design labels and what to include on them, and what information should be distributed by the vendor, on a state website, or via other forms of communication. Cannabis is a complicated drug, so vendors need to be knowledgeable, and the state needs to figure out how to incentivize vendors not to market to drug abusers. The 20% heaviest users account for 80% or more of the volume of drugs consumed, so vendors have a strong incentive to market to abusers.

Colorado and Washington both plan to make the tax on cannabis a percentage of the sales price. It would be better to set the tax so as to keep prices similar to what they were on the illicit market, since higher prices would allow the illicit market to continue to exist, and lower prices would increase drug abuse. Since production costs will likely fall over time, Washington’s tax plan risks starting with prices too high and ending up with prices too low. Colorado’s tax starts at too low a level.


With the right push, smoking could be virtually eliminated in 20 years by getting people to use electronic cigarettes, which have substantially fewer health risks than traditional cigarettes.

Tobacco regulation has generated a substantial illegal market: the smuggled tobacco market is of the same order of magnitude as the methamphetamine market. The illegal market in tobacco is not violent but it could become so if enforcement increases and the illicit market grows. This creates a strong tradeoff between public health demands for tighter regulation and higher taxes on the one hand and public-safety concerns about developing new and possibly violent illicit markets on the other. Most of the appropriate analysis and research has yet to be done, and the public-health-dominated tobacco-control movement is strongly inclined to ignore the issue.

International drug policy

US international drug policy is focused on reducing drug flows and putting drug traffickers in prison, not on reducing harm and violence. Drug law enforcement can shape the behavior of drug traffickers; making violence something that attracts enforcement attention could reduce the violence. Such a change would require a different set of policies and would be controversial, but the potential benefits are very great.

Potential research projects

Dr. Kleiman believes that clean water and female literacy are likely the most promising humanitarian causes globally, but within drug and crime policy the most promising area for funding in terms of its long-term benfeits is research on self-command. Swift and certain sanctions and similar programs may be able to improve individuals' self-command in the long-term. For example, the 24/7 sobriety program in South Dakota has noticeably reduced DUI rates of participants long after the treatment ended, so it is possible that such interventions can improve people's decision-making. It’s possible that other programs, for example the Good Behavior Game and Nurse-Family Partnership, may influence self-command. Many life outcomes are affected by self-command, so if there were a way to improve people's capacity to manage their own lives, many people could be made better off. Research could search for characteristics that relate to self-command, are predictive of the individual's outcomes, and are malleable to policy intervention.

There is surprisingly little research on the possibly beneficial effects of illicit drugs on the body or the mind. Research on illicit drugs is rare in the US because it is difficult to get permission to use illicit drugs in benefit-oriented or harm-reduction studies. Such research is possible in other countries, such as Israel, the UK, or the Netherlands. Research would allow us to better understand the significance of varying chemical compositions of drugs. Cannabis is a complex drug and states should be able to require labeling that explains the effects of each variety, e.g. the consequences of different THC/CBD ratios, but there is little research on these effects.

Organizations involved in drug policy and criminal justice

The Drug Policy Alliance is a controversial player in the pro-drug legalization community. They are more committed to advocacy than to research and the research they do tends to be driven by their policy agenda. The Open Society Foundations work on drug policy tends to be tightly aligned with the Drug Policy Alliance views. The Open Society Foundations work on criminal justice reform tends to focus on reducing overpunishment.

The Pew Foundation's staff assist state governments with long-term planning on how to reduce the state's prison population, with a fairly non-political approach. It is very good at what it does. The Pew Foundation does not fund much research.

The Arnold Foundation is a new and potentially large research funder in this space. They have been working to create a predictive tool to determine which arrestees could be let out pre-trial, and have begun to fund some work on swift and certain sanctions.

The Ford and MacArthur Foundations are involved on the edges of crime and drug policy. MacArthur has been considering getting involved with drug policy in a serious way. If they got involved, it would be a big change for the cause.

The Smith Richardson Foundation makes relatively small grants, but are very supportive of their researchers. Dr. Kleiman suggested that GiveWell and Good Ventures partner with them on this issue. They are politically conservative, which does not have an impact on the people they fund but does give more credibility to their results in some circles.

The National Institute of Justice funds research, but has limited resources (their core annual budget is around $20 million). They provided $250,000 for the HOPE project study, which funded the first stage of the research, but wasn't enough to do long-term follow up, interview the probationers after the program, or to measure health, education, and employment data or cortisol stress.

Other research funders include the Annie E. Casey Foundation and the California Wellness Foundation. The National Institute on Drug Abuse and the National Institute on Alcohol Abuse and Alcoholism fund research with a pharmacological perspective. The Robert Wood Johnson Foundation used to have a drug research program with a biomedical focus, but that ended a decade ago. The Sloan Foundation has not provided much funding for drug or crime policy research but some of the work in this area has qualities that might appeal to them.

Other people for Good Ventures to talk to

Reading suggestions