Part 2

Marijuana’s Impact on the Correctional System

The remaining portion of the criminal justice system where marijuana may be a significant cost factor is in the corrections system, which includes people in jail and prison or placed on probation or parole. As shown in Table 3, this a very large system that includes over 7 million people on any given day. The difficulty is in establishing the number of persons who are under correctional supervision on any give day or have been touched by the system during a calendar or fiscal year solely due to a marijuana charge. Such data do not exist, making it somewhat speculative in terms of what proportion of the correctional system is allocated to the control, punishment and treatment of marijuana offenders.

It is known that within the state and federal prison systems, there are approximately 250,000 state and another 70,000 federal inmates incarcerated for drug crimes. It is not known what numbers of prisoners are incarcerated solely for marijuana crimes. But if one were to assume that the marijuana cases would reflect the number of persons arrested and convicted for such crimes, one would have to assume that at best no more than 1/5th of the drug convicted and defendants (or about 64,000 prisoners) are marijuana cases. This would represent about 4% of the entire US prisoner population. King and Mauer estimated that 27,900 state and federal prisoners are incarcerated solely for marijuana crimes.12

But even this estimate is probably too high as it does not take into account that persons convicted of marijuana possession or even low level sale of the drug rarely receive prison terms. Once again there are no national data to test this hypothesis, but we can look at California’s state correctional data to see how significant the marijuana conviction cases are with respect to the parole and prisoner populations.

Table 4 summarizes the results of analysis made by the author based on detailed data files provided by the California Department of Corrections. This table reports the number of prisoners and parolees who are either incarcerated or are on parole for the crimes of marijuana possession or sale. For both the prisoner and parole populations the numbers of such cases reflect less than 1% of the total.

A somewhat similar analysis was done for Louisiana with somewhat different results. Louisiana, unlike California, is not a marijuana decriminalization state, so it may be that the number of person incarcerated or placed on probation or parole will be higher. It was not possible to separate the marijuana crimes by sale and possession but I was able to include the very large number of persons on probation. Here the numbers of marijuana prisoners are proportionately larger (2.2%) but still reflect a relatively small proportion of the prisoner population. However, the marijuana crime category for the parole and probation populations – especially for the probation populations - is a sizable portion (nearly 10%) of the entire 60,000 plus supervised population. What these data show is that a larger portion of the correctional population of persons convicted solely for marijuana possession or sale are on probation and to a lesser extent parole. But these populations (parole and probation) are also the least expensive forms of corrections.

Another possible impact on prison and jail populations is the number of parolees or probationers who have their parole or probation terms violated for the use or sale of marijuana. Here again the data are somewhat lacking. It is known that 1/3rd of all prison admissions are for parole supervision violations. About half of the parole violations are for technical violations. A few states have been able to point out that at least another 1/3rd of prison admissions are probation violators although there are no data on what percent are technical violators let alone those being violated solely for marijuana use, possession or sale.13

A recent example in the District of Columbia illustrates how the violation of probation or parole based on marijuana use can serve to produce a prolonged period of incarceration. One case involved a young man who was paroled in 2002 after a lengthy prison term for car-jacking. One of his conditions of supervision was to refrain from drug use and to undergo weekly drug testing. After several weeks of his release from prison he began to use marijuana for recreational purposes. He then tested positive on several occasions which resulted in his parole being revoked. There were no other charges or violations brought against him. The parole board revoked his parole which resulted in him being incarcerated first in the DC jail system and later by the Bureau of Prisons for a total of approximately 5 months before being re-released to parole. His case will not show up as a marijuana case in the agency statistic even though his re-incarceration was solely caused by marijuana use.

Table 4
Number of California Prisoners and Parolees Sentenced for Marijuana Crimes
December 2004

Total Population149,889100.0%126,578100.0%
Marijuana Crimes    
• Possession290.0%730.1%
• Cultivation590.0%1330.1%
• Transportation for Sale6300.4%1,3821.1%
• Possession with Intent to Sell3820.3%6690.5%
• Total Marijuana Offenders1,1000.7%2,2571.8%

Table 5
Number of Louisiana Prisoners, Probationers and Parolees Sentenced for Marijuana Crimes – April 2005

Total Population37,868100.0%38,231100.0%24,219100.0%
Marijuana Crimes8622.2%3,6779.6%1,3125.4%

Jail is also a place where there may be additional costs for marijuana crimes. As suggested above, of the 755,000 marijuana arrests, the vast majority will be booked into a jail and will spend some amount of time incarcerated while awaiting the disposition of their cases. For example, the US Department of Justice reports that 80% of all defendants charged with felony drug possession crime are released within one week of their booking into the jail. While the length of stay may be short, the booking process is one of the most expensive and dangerous operations in a jail.14 Elimination of these cases would no doubt have some impact on costs and overall safety of the facility.

Estimating the Impact of Decriminalization on Government Expenditures

How does one use the above information to determine the cost benefits from decriminalizing marijuana? We know that a large number of persons are arrested each year but that marijuana cases represent a relatively small proportion of the entire universe, and therefore, costs of police and related court processing tasks. Within the correctional system, there appears to be an even smaller number of persons sentenced to prison for marijuana crimes. For probation and parole the numbers seem to vary by state, but it’s clear that the largest numbers are going to be probationers. However, the costs of probation supervision are well below the costs of jail or prison. Finally, as illustrated above, there are several ways that marijuana use can lead to incarceration via the violation of supervision process.

There is no question that United States is spending large amounts of money on its criminal justice system. The current estimate is that the United States is spending about $167 billion a year to fund its federal, state, and local police, the courts and corrections agencies. In 1982 the costs were only $36 billion thus representing a 165 percent increase (in constant dollars).15 As shown in Figure 1, the largest proportion of these costs support police operations ($50.7 billion) followed state correctional agencies ($38.4 billion). The question to be addressed is; What proportion of these costs can be attributed to marijuana crimes? Or more directly, would these costs be reduced if marijuana were decriminalized?

Figure 1

Expenditure for justice functions varies by level of government

In fiscal year 2001 States spent $38.4 billion for corrections. Local governments spent the most for police functions -- $50.7 billion.

Source: Bauer, L. and S. Owens (2004). Justice Expenditures and Employment in the United States, 2001. WDC: BJS

Understanding Fixed, Dynamic, and Marginal Costs of Criminal Justice Agencies

Most cost benefit analyses fail to understand that the vast majority of the criminal justice costs are “fixed” or “static” and do not vary appreciably by the volume of activities, tasks or incidents undertaken by the agencies. This is because most of the agency costs are largely linked to agency personnel costs (salary and fringe benefits) which reflect 70-75% of a criminal justice agency budget and do not vary by marginal changes in workloads. Unless one can demonstrate that these costs vary by the number of persons arrested, prosecuted, defended, convicted, sentenced, supervised and incarcerated, there is little reason to believe that decriminalization of marijuana would have a fiscal impact. Or put differently, payrolls would have to be reduced, prisons closed or court-rooms emptied for real money to be saved.

A good analogy would be the airline industry. It costs a certain amount of money to fly a plane and operate the necessary support services (maintenance, flight attendants, pilots, fuel, and booking agents). These are so-called fixed costs that do not vary by the number of passengers that get on the plane. There are so called “marginal” cost savings that are real and need to be accounted for. In the example of airlines, certain non-fixed costs such as meals, beverages, and fuel used based on the plane’s passenger manifest and freight load.

The analogy can be applied to the business of criminal justice. Most of the previous studies that examine cost savings for decriminalization as well as any other major criminal justice reform rely on assumptions that serve to exaggerate cost savings. These studies and models often assume a direct relationship between fluctuations in events and caseloads and the costs that support public agencies and organizations that process cases and people. For example, if a particular police agency that has 7,500 officers makes 10,000 arrests for marijuana out of a total of 100,000 arrests, decriminalization advocates have argued that the police agency’s budget would be reduced by 10%. In reality, the police budget would remain largely the same unless 10% of the 7,500 officers were terminated. The only costs that would be realized could be linked to things like overtime costs.

One could argue that unless decriminalization had occurred, criminal justice costs would have been higher. This falls into the very speculative category of “averting” future costs that have yet to be incurred. There is some merit to this perspective but demonstrating cause and effect is quite problematic.

Furthermore, given the large amount of discretion enjoyed by police and the courts in terms of how resources are deployed, it is quite possible that as the emphasis on one crime is reduced, resources are re-directed to other crimes. Indeed, this is one of the major justifications for decriminalization of marijuana – not so much that it would reduce criminal justice costs. Rather it would allow the criminal justice system to focus on more serious crimes.

To better understand this phenomenon, one needs to examine the example of California’s decriminalization reforms in the 1970s. California on two occasions in the 1970s enacted major legislative changes designed to lessen the reach of the criminal justice system on marijuana use. In 1972, then Senator George Duekmejian successfully sponsored legislation that allowed persons charged with minor drug offenses and other misdemeanor crimes to be diverted from future criminal prosecution if they agreed to participate and complete a pretrial diversion program. In 1975, the California legislature enacted Senate Bill 95 which took effect on January 1, 1976 and had the following provisions:

  1. No arrest or booking for individuals apprehended in possession of small amounts of marijuana;
  2. No jail or incarceration for persons convicted of possession of small amounts of marijuana;
  3. Furnishing of small amount of marijuana for no consideration is treated as simple possession, not sale;
  4. Transportation of small amounts of marijuana is treated as simple possession, not felony transportation;
  5. Elimination of life-long criminal records for marijuana possession arrests and convictions, and placing a two-year limit on the retention of such records and the use of such records against individuals arrested and convicted of specified offenses;
  6. Abolition of recidivist penalties for simple possession, giving away and transporting small amounts of marijuana.

Many observers have concluded that the California legislation has been very successful in terms of reducing arrests and saving large amounts of money. Aldrich et al.16, found that after S.B. 95 took effect, the number of marijuana arrests declined from a pre 1976 rate of approximately 100,000 per year to about 25,000 per year. Using a proportionate costs benefit model, the authors concluded that criminal justice expenditures declined by 24 percent from 1974 to 1984 with a culminative savings of $360 million. They conclude that:
“ It is rare that a single legal change has such an immediate and drastic effect on arrests and enforcement costs in a state.”

But what the study did not do was look at the total arrest and expenditure data for the same period. As shown in Table 6 the total number of arrests, police officers and, more importantly, costs have steadily increased over the same time period. There is some suggestion that the number of police dropped in 1976 and remained stable from 1977-1978 which may be attributed to SB 95. What one must consider is whether law enforcement and other criminal justice agencies did not reduce their agency costs, but rather, shifted their increasing resources and costs to other matters and/or relabeled certain behaviors that used to be recorded as a marijuana arrest to some other crime category.

For example, a law enforcement officer often has several choices in terms of labeling the arrestee with a specific charge. In the case of a marijuana arrest it is often the case that other behavior (e.g., reckless driving, trespassing, etc.) triggered the arrest and that marijuana was only later discovered after searching the person. With the advent of SB 95 the same person is arrested but on a non-marijuana charge.

It is conceded that one can not prove the case of “re-labeling” by law enforcement. Nor can it be denied that there has been an impact on marijuana drug arrests. The question is whether the effect has been as great as the advocates of SB 95 claim?


The debate on whether to decriminalize marijuana remains a hotly debated topic. Unfortunately the debate is rarely informed by data that can be used to help parents, voters and politicians make an informed decision on whether decriminalization will provide substantial savings and not lead to more use and more crime. With respect to the dual questions of decriminalization's impact on use and crime, there seems to be broad consensus based on scientific data that it will have little, if any, impact. Marijuana already is a widely used substance with over 26 million annual users. And despite increases in marijuana use since the early 1990s, the crime rate has plummeted over the same time frame. If there was a marijuana-crime link, it is not having its expected impact on crime.

Whether decriminalization would have a significant impact on the criminal justice system remains an open question. While representing a significant number of arrests (over 750,000) each year, the proportion of marijuana arrests as compared to the total universe of arrests is quite low. Imprisoned persons who are incarcerated only for marijuana are comparatively few. Further, the criminal justice system’s capacity to reconstitute itself and actually expand in the face of declining crime rates illustrates just how difficult it is to generate actual savings. One should not be surprised to find that if marijuana were legalized, the net effect to the taxpayer in criminal justice savings would be negligible.

Table 6
California Arrests, Police Officers, and Expenditures

Annual Change
Annual Change


This is not to say that efforts to decriminalize marijuana should be abandoned. Certainly the nearly three quarter million Americans who are arrested and booked into jail each year see the value of such a policy. There are also many hidden costs associated with marijuana use especially for those nearly 5 million persons on parole and probation supervision who are subject to incarceration if they do what 25 million Americans do on a regular basis – smoke pot.

Finally, while the numbers of persons incarcerated for pot may be low and criminal justice costs associated with marijuana use is not as great as some suggest, it makes no sense from a public policy perspective to add to the workload of an already over-extended criminal justice system. Our money and resources would be better spent on far more pressing social and public safety issues.

There are several studies that claim serious crimes actually increase as police and the courts pay greater attention to drug crimes.17 This perspective simply acknowledges what is obvious to the most common observations of police patrol behavior. Typically when a police officer makes an arrest, it will result in an officer being off the street for several hours in order to complete the arrest report. In my recent study of the Washington DC police department, I found that the average time it took an officer to complete the required paper work was approximately 7.7 hours from the point of arrest.18 During this time the officer’s presence to detect and deter other crimes that may be occurring was essentially eliminated.

This is the real benefit of decriminalization. Actual financial savings to taxpayers will only happen if police, probation and parole officers are laid off, and, court rooms and prisons closed. This is unlikely to occur simply due to marijuana being decriminalized. But it may halt the ever increasing and expanding amounts of money being spent on marijuana enforcement and force the criminal justice system to re-allocate its resources to more serious crimes.

Help Fund this Report
Join the Discussion