TPOR-60227\3120936.1
1331 N. California Blvd.
Suite 600
Walnut Creek, CA 94596
T 925 935 9400
F 925 933 4126
www.msrlegal.com
Matthew C. Henderson
Direct Dial: 925 941 3271
[email protected]
Offices: Walnut Creek / San Francisco / Newport Beach / Reno May 5, 2025
VIA E-MAIL
Riverside Planning Commission
City of Riverside
City Hall
3900 Main Street
Riverside, CA 92522
Email: [email protected]
Re: May 8, 2025 Planning Commission Meeting Agenda Item 4, Case No. PR-2025-001795 (Also Referred to as File No. 25-1637) Dear Honorable Members of the Planning Commission:
This letter is sent with respect to the above-referenced agenda item for the Planning Commission’s pending May 8, 2025 meeting.1 As set forth in a letter to the City dated April 23, 2025, this office represents three of the applicants for the City’s storefront retail cannabis business operating permits. The purpose of this letter is to correct several errors in the Memorandum prepared for the above-agendized item.
First, the Memorandum misrepresents the City Council’s actions on March 25, 2025.
The Memorandum states, “The City Council voted to modify the [Storefront Retail Commercial Cannabis Business] program.” That is incorrect. The minutes for the meeting clearly demonstrate that the Council directed staff to draft an ordinance and resolution making modifications to the program. That ordinance and resolution will be subject to a separate vote, which has not yet occurred. Thus, the Memorandum suggests that the City Council has taken a final action that it has not in fact taken.
This mischaracterization misleads the public, jeopardizes public faith in the City’s actions, suggests a precommitment by the City to an outcome without a fair public hearing, and potentially violates the Brown Act.
Second, the Memorandum states that “Staff has not received public comments regarding this project.” This is also incorrect. My partner Arthur Coon submitted an extensive comment letter, with exhibits, to the City on April 23, 2025, directed at the changes the City has proposed with respect to the Storefront Retail Commercial 1 The item is on the agenda as Case No. PR-2025-001795, but is elsewhere referred to in the online materials as File No. 25-1637. This letter is thus directed at and should be included in both file/case numbers.
Riverside Planning Commission
City of Riverside
May 5, 2025
TPOR-60227\3120936.1
Cannabis Business program. Obviously those changes extend to the matters the Planning Commission will be considering on May 8. Accordingly, we hereby request that the letter and exhibits be included in the file for Case No. PR-2025-001795/File No. 25-1637 (see footnote one) and to also be included as part of the administrative record for any action the Planning Commission and/or City Council may take with respect to the same. We have received confirmation from the Office of the City Clerk that the letter and attachments have been received; if this is incorrect, please notify me and I will have copies provided immediately.
Please do not hesitate to contact me if any of the foregoing is unclear or if you have any questions.
Very truly yours,
MILLER STARR REGALIA
Matthew C. Henderson
MCH:klw
cc: City Clerk ([email protected])
City Manager Mike Futrell ([email protected])
Interim City Attorney Rebecca McKee-Reimbold, Esq. ([email protected]) Principal Planner Matthew Taylor ([email protected]) Arthur F. Coon, Esq. ([email protected])
Dana Cisneros, Esq. ([email protected])
EXHIBIT 1
1 STOREFRONT RETAIL COMMERCIAL CANNABIS
BUSINESS PERMIT APPLICATION
questions m be submitted to the City by email to [email protected] and will be answered in the order received . Additionally, applicants can Application Packet
Application packets shall be organized in the same order as the checklist with page of this document on top. An application that is missing any of the required components .
2 Filing
15, PM, as noted on the City’s annabis ( above). Any incomplete application received may
terms for any such extension will be posted on the City’s annabis . Fees City Hall is located at 3900 Main Street, Riverside, CA 92522, and is open from 8:00AM to 5:00PM, Monday through Friday. All city business shall occur during these business hours. will receive a receipt of payment to upload with their respective application submittal.
Payment can be made anytime during normal business hours, excluding holidays.
Applicants that
proceed past Phase will
Phase 1: Application Review fee – $13,842
Pursuant to Riverside Municipal Code Chapter 5.77 ,
3 1. Applicant / Business Entity:
Applicant (Owner):
Business Entity (add DBA if applicable):
Mailing Address Street:
City/State/Zip Code:
Phone Number:
Email:
Date of Birth:
2. Primary Contact Information
Applicant’s Representative/Contact:
Representative / Contact Title:
Company DBA (if applicable):
Relationship to Applicant:
Mailing Address Street:
City/State/Zip Code:
Phone Number:
Email:
Date of Birth:
3. Business Entity Status
Check one:
Limited L ability Company
General Partnership
Limited Partnership
Corporation (or foreign corporation)
Limited Liability Partnership
Other:
Attach formation documents (e.g., articles of incorporation, operating agreements, partnership agreements, Applicant Name:
4 4. Cannabis Activity Denials, Suspensions, or Revocations Applicant(s) shall provide the type of license or permit applied for, the name of the licensing/permitting authority that denied, suspended, or revoked the application or license/permit, the date of denial, suspension, or revocation, the length of suspension, if applicable, and the basis of the denial, suspension, or revocation.
Check this box if the applicant has not had any prior commercial cannabis activity denials, suspensions, or revocations.
Denial, Suspension, or Revocation Details
License type:
Authority:
Date of denial:
Reason for denial:
Denial, Suspension, or Revocation Details
License type:
Authority:
Date of denial:
Reason for denial:
5 5. Criminal Convictions
convicted of any of the following:
A felony for hiring, employing, or using a minor in transporting, carrying, selling, giving away, preparing for sale, or peddling, any controlled substance to a minor; or selling, offering to sell, furnishing, offering to furnish, administering, or giving any controlled substance to a minor.
Code.
A felony involving fraud, deceit, or embezzlement.
Pursuant to Riverside Municipal Code Chapter 5.77.080, applicant acknowledges and agrees to require employees to submit to a background check within one month of being hired.
Applicant signature: Date:
Pursuant to Riverside Municipal Code Chapter 5.77.080, applicant acknowledges and agrees that any evidence of a conviction of any the offenses outlined above shall be grounds for denial or revocation of a cannabis business permit, and/or employment with a cannabis business permittee in the City of Riverside.
Applicant signature: Date:
6. State Law Compliance
Applicant acknowledges and agrees to meet the requirements of the state cannabis laws, including, but not limited to, track-and-trace, inventory, returns, destruction of products, waste management, environmental sustainability, records retention, and operational requirements.
Applicant signature: Date:
Yes No
Yes No
Yes No
Yes No
Yes No
6 8. Labor Peace Agreement
Pursuant to Riverside Municipal Code Chapter 5.77.130, applicants are required to attest that they have
an applicant has not yet established such an agreement, they must furnish a notarized statement expressing their commitment to promptly enter into a labor peace agreement and adhere to its terms within 30 days of receiving a Storefront Retail CCB permit.
Check this box if the applicant has not yet established a labor peace agreement and intends to submit a notarized statement.
Applicant attests to the ownership of an existing labor peace agreement and will abide by the terms of the agreement.
Applicant signature: Date:
7. Insurance Compliance
Pursuant to Riverside Municipal Code Chapter 5.77.330, applicants must obtain and provide liability insurance to Risk Management. Below is the requirement to obtain and maintain insurance in the amounts and types
as additional insureds by endorsement.
Applicants also must obtain and maintain Workers’ Compensation with limits as required by the Labor Code of the State of California.
the City’s insurance portal: https://RiversideCA.gov/Coiportal Applicant signature: Date:
General Liability Auto Liability
Required Limits $2,000,000 per Occurrence
$4,000,000 per Aggregate
$1,000,000 Combined Single Limit
(CSL)
Required
Endorsements Additional Insured Additional Insured
7 9. Cannabis Experience and Background
experience.
Include copies of permits, licenses, or other written forms of permission by a state, county, or municipal government entity identifying claimed experience. Copies of supporting documents do not count against your application page limits, only the business proposal has page limits.
List the license types and license numbers that the applicant currently holds from the State of California and/or all other out-of-state licensing authorities.
Check this box if the applicant does not have any prior commercial cannabis experience.
Full Legal Name:
Mailing Address Street:
City/State/Zip Code:
Phone Number:
Email:
Date of Birth:
Experience Details
Year(s) of Experience:
Mailing Address Street:
City/State/Zip Code:
License Type:
License Number:
Full Legal Name:
8 Full Legal Name:
Mailing Address Street:
City/State/Zip Code:
Phone Number:
Email:
Date of Birth:
Experience Details
Year(s) of Experience:
Mailing Address Street:
City/State/Zip Code:
License Type:
License Number:
Full Legal Name:
Mailing Address Street:
City/State/Zip Code:
Phone Number:
Email:
Date of Birth:
Experience Details
Year(s) of Experience:
Mailing Address Street:
City/State/Zip Code:
License Type:
License Number:
9 Pursuant to City of Riverside Municipal Code Chapter 5.77.330, and to the fullest extent permitted by law, the City of Riverside shall not assume any liability whatsoever with respect to having issued a commercial cannabis business permit or otherwise approving the operation of any commercial cannabis business.
Applicant hereby attests and agrees to defend (at Applicant’s sole cost and expense, and with counsel liabilities or losses which arise out of, or which are in any way related to, the City’s issuance of the commercial cannabis business permit, the City’s decision to approve the operation of the cannabis business or activity, the process used by the City in making its decision, or the alleged violation of any federal, state or local laws by the Applicant signature: Date:
Applicant hereby attests and agrees to reimburse the City for all costs and expenses, including but not limited to legal fees and court costs, which the City may be required to pay as a result of any legal challenge related to the City’s approval of the Applicant’s cannabis business permit or related to the City’s approval of a cannabis activity. The City, at its sole discretion, may participate at its own expense in the defense of any such action, but such participation shall not relieve the Applicant of any of the obligations imposed hereunder.
Applicant signature: Date:
11. Proof of Application Fee Payment
Application
ee in the amount and manner established by
Applicants shall submit a copy of the receipt provided by the City of Riverside Treasury as proof of payment for the pplication eview.
Applicant acknowledges and agrees to pay additional fees that may be incurred if selected for the second phase of the Storefront Retail CCB Permit application. Furthermore, the applicant also acknowledges and at the discretion of the City Manager.
Applicant signature: Date:
10 12. Business Proposal
All Storefront Retail CCB Permit applicants must prepare a narrative Business Proposal that will be used during the Merit-Based evaluation. The Business Proposal must 12-point font text that does not exceed a total of 50 single-sided pages that addresses topics outlined in the Evaluation Criteria i Business Plan/Financial Investment
Operations Plan
Safety/Security Plan
Labor, Employment, and Local Sourcing Plan
EXHIBIT 2
1 CITY OF RIVERSIDE STOREFRONT RETAIL COMMERCIAL
CANNABIS BUSINESS PERMIT – WRITTEN RANKING
D. Application Evaluation Process – Written Ranking
Pursuant to the City of Riverside Storefront Retail Commercial Cannabis Business Permit Procedure Guidelines and Application Review Criteria, below is a Written Ranking of the top fourteen (14) applications that have been reviewed, evaluated, and ranked by the City Review Panel.
Based on the stated Merit-Based Evaluation criteria and point system, the top fourteen (14) ranked applications – highlighted in GREEN – have received Provisional Approval.
RANK BUSINESS NAME APPLICANT ID
1 STIIIZY Riverside LLC 24
2 SGI Riverside LLC 25
3 C4TP Retail A Inc. 5
3 Riverside Community Retail LLC 17
5 Community Oriented Riverside Retail LLC 6
5 Riverside Responsible and Compliant Retail
LLC
19 7 Blaine St. RS LLC 4
8 OTC Riverside City LLC 16
9 Packs Riverside LLC 2
9 Riverside West Coast Retail LLC 18
9 The Artist Tree Holings LLC 23
12 Catalyst Riverside Equity LLC 33
13 Haven Riverside LLC 11
13 Catalyst Riverside LLC 32
EXHIBIT 3
EXHIBIT 4
“Pot Luck”
Santa Ana’s Monopoly on Licensed
Retail Adult-Use Cannabis
in Orange County
GRAND JURY 2020-2021
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page i
TABLE OF CONTENTS
SUMMARY ................................................................................................................................... 1 BACKGROUND ........................................................................................................................... 1 Federal Cannabis Laws ............................................................................................................ 1 Legislation in California ........................................................................................................... 1 REASON FOR STUDY ................................................................................................................ 2 METHOD OF STUDY ................................................................................................................. 3 INVESTIGATION AND ANALYSIS ......................................................................................... 3 Santa Ana City Council Says “Yes” ........................................................................................ 3 Dispensary Application and Licensing Process ...................................................................... 3 Cannabis Community Benefits Program ................................................................................. 4 Community Safety ..................................................................................................................... 4 Site Visits ................................................................................................................................. 5 Financial Impact........................................................................................................................ 6 Communication ......................................................................................................................... 7 COMMENDATIONS ................................................................................................................... 7 FINDINGS ..................................................................................................................................... 7 RECOMMENDATIONS .............................................................................................................. 8 RESPONSES ................................................................................................................................. 9 RESPONSES REQUIRED ..................................................................................................... 10 APPENDIX.................................................................................................................................. 11 APPENDIX 1 – Approved and Operating Adult-Use Cannabis Dispensaries .................. 11 APPENDIX 2 – Commercial Cannabis Application Process.............................................. 12 APPENDIX 3 – Phase 1 and Phase 2 Application and Permit Process.............................. 13 APPENDIX 4 – Commerical Cannabis Application Information...................................... 14 APPENDIX 5 – Santa Ana Ordinance No. NS-2959 ............................................................ 16 GLOSSARY................................................................................................................................. 20 BIBLIOGRAPHY ....................................................................................................................... 21 “Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 1
SUMMARY
Santa Ana was the first city in Orange County, California to approve the retail sale of Adult-Use Cannabis. This action has added significant revenue to the city with no reported increase in criminal activity. Through interviews and investigation, the Orange County Grand Jury has discovered that the revenue generated by the Retail Adult-Use Cannabis business has provided much needed funds to the City of Santa Ana. These funds have not only increased the city’s general fund account but have also been used for enhanced police services and code enforcement efforts as well as funding for additional youth programs through the Parks, Recreation, and Community Services Agency.
The Orange County Grand Jury does not express an opinion on the use of Cannabis.
BACKGROUND
Cannabis use has long been a subject of controversy in the United States. Once commonly grown for hemp, made from fibers from the plant and used in a variety of products such as rope and paper, cannabis was later discovered to have medicinal purposes and subsequently became a criminalized product. Over the last six decades there has been much debate and many propositions introduced to decriminalize and/or regulate cannabis and allow it to be legally sold and used for medicinal as well as recreational purposes.
Federal Cannabis Laws
Despite a cannabis legalization trend sweeping the country, the federal government still classifies cannabis as a Schedule 1 drug. The federal government considers drugs in this class to be some of the most dangerous.
A Schedule 1 classification puts cannabis in the same class as heroin, which means the federal government considers cannabis more dangerous than Schedule 2 drugs like cocaine and methamphetamine.
At the present time, the Unites States Department of Justice (“DOJ”) is not prosecuting most cannabis users and businesses that follow state and local cannabis laws, as long as those laws do not conflict with certain federal requirements. These requirements include preventing minors from using cannabis and preventing cannabis from being transported across state lines.
Legislation in California
Proposition 19 (1972) also known as “The California Marijuana Initiative” was a ballot initiative on the November 7, 1972 California Statewide Ballot. California became the first state to vote on a ballot measure seeking to legalize cannabis. If it had passed, the measure would have removed penalties in the State of California for persons 18 years of age or older for using, “Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 2
possessing, growing, processing, or transporting marijuana for personal use. The initiative was defeated by the voters with 66.5% No votes to 33.5% Yes votes.
Proposition 215 (1996) also known as “The Compassionate Use Act of 1996” made it legal under California law for individuals of any age to use cannabis for medicinal purposes.
Individuals must have a recommendation from a doctor to use medical cannabis. The act passed by a vote of 55.58% Yes votes to 44.42% No votes.
Proposition 19 (2010) also known as “The Regulate, Control, and Tax Cannabis Act of 2010” was defeated by the voters with 53.5% No votes to 46.5% Yes votes. If it had passed, Proposition 19 would have legalized various cannabis-related activities in California and authorized local governments to control these activities. In addition, the Proposition would have granted local governments the right to impose and collect cannabis-related fees & taxes, and authorized various criminal and civil penalties.
Proposition 64 (2016) also known as “The Adult-Use of Marijuana Act” passed by a vote of
57.13% to 42.87%. The measure:
Legalized adult use of cannabis for recreational, non-medical purposes Created a system for regulating Retail Adult-Use Cannabis businesses Imposed taxes on Retail Adult-Use Cannabis sales
Changed penalties for cannabis-related crimes
Once Proposition 64 was passed, cities in California were granted the opportunity to approve Retail Adult-Use Cannabis and begin the process of granting licenses to shops within their city limits.
REASON FOR STUDY
The selling of cannabis for “Adult-Use” or “recreational” purposes has been legal in the State of California since January 1, 2018 and yet, until July 2020, Santa Ana was the only city in Orange County that had approved licensing for this type of business. The Orange County Grand Jury (OCGJ) was interested in investigating how the decision to move forward with this licensing impacted Santa Ana and if there were any significant issues.
The Grand Jury felt it was important to investigate this matter in order to make the public aware of the potential gains or pitfalls other cities in the county might encounter should they move forward with Retail Adult-Use Cannabis licensing.
This report focuses only on the licensing and selling of Retail Adult-Use Cannabis in the City of Santa Ana and does not address medicinal sales, cultivation, distribution, or any issues related to the use of cannabis products.
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 3
METHOD OF STUDY
The OCGJ conducted extensive internet research on the legal status of Retail Adult-Use Cannabis, both nationwide and in California, by reviewing and analyzing relevant legislation as well as the numerous California propositions that culminated in the passage of Proposition 64. In addition, the OCGJ reviewed City of Santa Ana documents including staff reports, commission reports, Requests for City Council Actions, and ordinances that authorized and established conditions for the sale of cannabis products for Adult-Use.
The OCGJ also interviewed current and former Santa Ana City Council members and City employees who work in agencies directly involved with or impacted by the licensing and sale of Retail Adult-Use Cannabis products in Santa Ana. In addition, OCGJ interviewed professional experts not employed by the City, including licensed Retail Adult-Use Cannabis dispensary proprietors and employees. The OCGJ visited several licensed Retail Adult-Use Cannabis dispensaries in Santa Ana to observe the facilities, amenities, staff, clientele, and operations.
INVESTIGATION AND ANALYSIS
Santa Ana City Council Says “Yes”
In response to the November 2016 passage of Proposition 64 in California, the Santa Ana City Council began to consider licensing and regulating the retail sale of cannabis for adults. After multiple meetings and discussions and after thorough staff analysis, on October 17, 2017, the City Council introduced ordinance number NS-2929 for a first reading. The ordinance was identified as “An Ordinance of the City Council of the City of Santa Ana creating Chapter 40 of the Santa Ana Municipal Code, ‘Regulation of Commercial Cannabis’, to Regulate Commercial Cannabis Activities, excepting Medicinal Retail.”
On November 9, 2017, the ordinance was introduced for a second reading and approved with amendments by a vote of 5-0 (two City Council members were absent). Santa Ana has remained the only city in Orange County issuing business licenses and regulating the retail sale of AdultUse Cannabis for the years 2018, 2019, and 2020.
Dispensary Application and Licensing Process
Santa Ana Ordinance number NS-2929 allows up to 30 Retail Adult-Use Cannabis stores within the City. As of April 15, 2021, there were 23 dispensaries open and operating in the City of Santa Ana (see Appendix 1). The ordinance sets forth operational standards, permit procedures, and an operating agreement to address collection of operating fees.
To ensure fairness and impartiality in the selection process, the City devised a system of “Retail Adult-Use Cannabis merit-based criteria and possible points” (see Appendix 2) to evaluate applicants who sought to operate a dispensary offering Adult-Use Cannabis products. This led to the Commercial Cannabis Application (Phase I and Phase II) and Permit Process (see Appendix 3), a comprehensive evaluation process that each Retail Adult-Use Cannabis dispensary applicant had to follow and pass before being allowed to open for business. “Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 4
After completion of the application process, each applicant was required to obtain a Regulatory Safety Permit (RSP) which is a permit issued by the City of Santa Ana pursuant to Chapter 40 of its municipal code. The RSPs issued for Phase I and Phase II had fees associated with each phase of the process. The fees charged were based on city processing costs as follows: Phase I, initially $1,690, was subsequently raised to $1,752; Phase II, initially $12,086, was later raised to $12,530.
The City’s Planning and Building Agency drafted a two-page Phase I and Phase II applicants’ information form advising interested parties of the steps in the application process (see Appendix 4).
Cannabis Community Benefits Program
All parties seeking a license to operate a cannabis dispensary in Santa Ana must submit a written operating agreement titled “Operating Agreement for Adult use (Non-Medicinal) Cannabis Retail Business.” One section of the agreement is entitled “Public Benefit.” In this section, the applicant for the proposed dispensary is encouraged to submit a “Community Benefit and Sustainable Business Practices Plan” (PLAN). The plan serves as a goodwill program sponsored by the dispensary for the benefit of the Santa Ana community.
A review of the PLANs submitted by the applicants revealed a wide variety of current community service projects such as sponsoring a local debate team, supporting a community garden, organizing and/or funding toy/clothing/food drives, diversion or prevention educational programs, and neighborhood clean-up efforts.
While these programs are a step towards goodwill in the community, there is no oversight or confirmation by the City that the dispensary operator is participating in the plan. It is up to each dispensary owner to decide how and to whom they will provide a “benefit.” There are no specific requirements in terms of money or volunteer hours that a dispensary must donate, and there is no obligation to provide proof of participation in the plan to the City.
The OCGJ has concluded that there should be some process in place to set standards for and document participation in the Community Benefits Program. In addition, the benefits provided by the Retail Adult-Use Cannabis dispensaries should be reported to the City Council and made available to the general public.
Community Safety
The OCGJ learned through interviews with City officials and staff that there had been more than 120 unlicensed dispensaries operating illegally in Santa Ana before ordinance NS-2929 went into effect. That number has since been reduced to “less than a handful,” due to enforcement efforts by the City, especially the Planning and Building Agency, including Code Enforcement, and the Santa Ana Police Department.
One third of the taxes and fees that the City receives from Retail Adult-Use Cannabis dispensaries is dedicated to Police and Code Enforcement. This money provides funding for a vice unit, including a sergeant and four officers. Other City agencies, such as Planning and “Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 5
Building and the City Attorney’s Office, also receive funding because of their role in ensuring Retail Adult-Use Cannabis compliance with regulations.
The reality is that shutting down the unlicensed, illegally operating dispensaries will increase business for the licensed facilities, thereby increasing the City’s tax revenues. Closing unlicensed facilities is a win-win for both the licensed dispensaries and the City of Santa Ana.
It should be noted that the licensed Retail Adult-Use Cannabis dispensaries must meet the quality standards for their merchandise that comply with requirements set forth by the State of California’s Bureau of Cannabis Control (BCC). Consumers have no such protections when products are purchased from unlicensed shops.
Ordinance number NS-2929, Section 1, subparagraph K, states: “The City of Santa Ana has a compelling interest in ensuring that cannabis is not sold in an illicit manner, in protecting the public health, safety, and welfare of its residents and businesses, in preserving the peace and quiet of the neighborhoods in which these businesses may operate, and in providing access of cannabis to residents.”
The OCGJ further learned through interviews with both Police Department and Code Enforcement staff that there has been no apparent increase in criminal activity in the areas surrounding the Retail Adult-Use Cannabis dispensaries. It should be noted that all licensed dispensaries are required to have security guards in place during hours of operations, and 24/7 video surveillance protecting their facilities, employees, and customers. Additionally, all dispensaries are required to secure all inventory in a locked safe, unless open for business.
Site Visits
During its investigation, the OCGJ interviewed cannabis dispensary owners and visited Retail Adult-Use dispensary sites. The OCGJ would like to note that it received full cooperation from the dispensary owners and staff while touring their locations.
The dispensary sites the OCGJ visited appeared to be well managed, with clean public areas that were adequately illuminated and well-appointed. The shelves were fully stocked with products.
All products were marked with California approval code stamps, indicating that the products conformed to quality control standards approved by the BCC.
Sites visited had the required security guards; the OCGJ noted that there were two security guards at each site visited. Each store had 24-hour video surveillance cameras covering the interior and exterior. One of the store owners mentioned that their external video surveillance is so extensive and of such high quality that they were able to assist law enforcement with investigations into criminal activities at nearby properties. Inside, stores were well furnished, had appropriate security doors, and all products were locked up at night in secured structures, vaults, or safes.
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 6
Financial Impact
In November 2018, residents of the City of Santa Ana approved Measure Y by a vote of 71% Yes to 29% No. Measure Y is a tax on Retail Adult-Use Cannabis businesses operating in the City that provides for a tax of 25 cents to 35 dollars based on the gross square footage of the business and a sales tax of up to 10% for retail sales. The City estimates that Measure Y will generate $11-14 million a year to fund city services.
On December 4, 2018, the Santa Ana City Council adopted Ordinance NS-2959 (see Appendix 5), establishing a “Cannabis Public Benefit Fund”. The fund derives almost all of its money from Measure Y and requires that the city allocate all money received from the sale of Retail AdultUse Cannabis as follows: one third to the General Fund, one third to Enforcement Services, and the final third to Youth Services.
Some of the City of Santa Ana Council members and staff interviewed by OCGJ informed the OCGJ that there is no true, viable oversight regarding disbursement and use of cannabis money received. The OCGJ learned that there is no clearly identifiable accounting for residents to see how this money is spent. Furthermore, the COVID pandemic has caused a shifting of money from previously planned programs to others.
Money received by the Retail Adult-Use Cannabis businesses for the last two years from the Measure Y tax has been in excess of $20 million. The General Fund money can be allocated to projects or programs in any City department, and expenditures from this fund cannot be specifically attributed to the Cannabis Public Benefit Fund. It has also been difficult to secure specific information about how the money for Enforcement Services has been used. Interviews with City staff indicated that various departments rely on Measure Y funds for their enforcement efforts. For example, the Police Department has funded the Vice Unit with Measure Y proceeds and the Planning and Building Agency, especially the Code Enforcement Division, also relies on Cannabis Public Benefit Fund money to staff some positions. However, the OCGJ has not received a clear breakdown of how the Enforcement Services money has been used by the various city agencies.
Expenditures related to Youth Services are much more transparent. The following programs and projects are anticipated to be funded from the Cannabis Public Benefit Fund and undertaken in the following year for the benefit of Santa Ana youth:
Library Services
Book/Techmobile
Digital Collection for Teens
Laptop Dispenser Kiosks
Wireless Hotspots
Library Improvements at the Newhope Branch Library
Library Playground at the Main Branch
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 7
Parks and Recreation
Anti-Drug Education Program
Fitness Courts
Goat Encounter at Santa Ana Zoo
Santa Anita Park Soccer Field Renovation
Splash pads for six City parks
Third Party youth programs
Traveling Zoo Exhibit
Year-Round Aquatics
Youth and Teen Excursion
Youth Programs and Services
Zoo and You Program
Community Development Agency youth paid internships
Contribution to the Deferred Action for Childhood Arrival Defense Fund The OCGJ, through its investigation of the use of Cannabis Public Benefit funds, concluded that the youth of Santa Ana have benefited greatly and will continue to benefit as a result of the decision to license Retail Adult-Use Cannabis dispensaries in the City.
Communication
During the investigation of the Retail Adult-Use sale of cannabis in Santa Ana, the OCGJ noted that there are several independent City departments involved in the licensing and regulation process. While the expertise of each department may be required to ensure compliance with all City ordinances and to process all necessary documents and fees, multiple points of contact can make it difficult to obtain information when needed.
COMMENDATIONS
The City of Santa Ana received more than $20 million in revenue during the first three years of licensed Retail Adult-Use Cannabis sales. Santa Ana was the first city in Orange County to begin licensing for these dispensaries and has remained the only city for several years, giving it a “monopoly” on Retail Adult-Use Cannabis business in Orange County.
The City of Santa Ana saw a significant drop in the number of illegal/unlicensed cannabis shops as the number of licensed dispensaries increased. This reduction in illegal/unlicensed shops has improved community safety for both consumers and residents.
FINDINGS
In accordance with California Penal Code Sections 933 and 933.05, the 2020-21 Grand Jury requires responses from each agency affected by the findings presented in this section. The responses are to be submitted to the Presiding Judge of the Superior Court.
Based on its investigation described here, the 2020-21 Orange County Grand Jury has arrived at the following principal findings:
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 8
F1. There is no clearly identifiable accounting of where all the Enforcement Services money received from Retail Adult-Use Cannabis licensing in accordance with Santa Ana Municipal Ordinance number NS-2959, section 13-203 was spent.
F2. The legalization of Retail Adult-Use Cannabis by the City of Santa Ana and the resulting increase in city revenue have allowed the city to significantly expand its youth services programs.
F3. There are multiple departments within the City of Santa Ana responsible for various aspects of Retail Adult-Use Cannabis oversight. The decentralized nature of the oversight within the City makes information difficult to obtain.
F4. The number of unlicensed cannabis dispensaries in Santa Ana has significantly declined since the business licensing of Retail Adult-Use Cannabis dispensaries.
F5. The Orange County Grand Jury did not receive evidence of an increase in crime as a result of the licensing of Retail Adult-Use Cannabis dispensaries.
F6. The Cannabis Community Benefits Program motivates local dispensaries to contribute funds and/or staff volunteer hours for the benefit of the City of Santa Ana.
RECOMMENDATIONS
In accordance with California Penal Code Sections 933 and 933.05, the 2020-21 Grand Jury requires responses from each agency affected by the recommendations presented in this section.
The responses are to be submitted to the Presiding Judge of the Superior Court.
Based on its investigation described herein, the 2020-21 Orange County Grand Jury makes the following recommendations:
R1. The Orange County Grand Jury recommends that the Santa Ana City Council require an annual report specifically detailing all Retail Adult-Use Cannabis money spent each fiscal year. This report should be presented to the Santa Ana City Council and made public. This should be completed by December 31, 2021, for fiscal year 2020-21, and by September 30 following each fiscal year thereafter. (F1) R2. The Orange County Grand Jury recommends that the City of Santa Ana designate an individual to oversee and provide centralized coordination of Retail Adult-Use Cannabis activity commencing January 1, 2022. (F3)
R3. The Orange County Grand Jury recommends that the City of Santa Ana create a process to report to the Santa Ana City Council and the residents of Santa Ana detailed information about the participation of Retail Adult-Use Cannabis dispensaries in the Community Benefits Program by December 31, 2021 and yearly thereafter. (F6) “Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 9
RESPONSES
The following excerpts from the California Penal Code provide the requirements for public agencies to respond to the Findings and Recommendations of this Grand Jury report:
§933
(c) No later than 90 days after the grand jury submits a final report on the operations of any public agency subject to its reviewing authority, the governing body of the public agency shall comment to the presiding judge of the superior court on the findings and recommendations pertaining to matters under the control of the governing body, and every elected county officer or agency head for which the grand jury has responsibility pursuant to Section 914.1 shall comment within 60 days to the presiding judge of the superior court, with an information copy sent to the board of supervisors, on the findings and recommendations pertaining to matters under the control of that county officer or agency head and any agency or agencies which that officer or agency head supervises or controls. In any city and county, the mayor shall also comment on the findings and recommendations. All these comments and reports shall forthwith be submitted to the presiding judge of the superior court who impaneled the grand jury. A copy of all responses to grand jury reports shall be placed on file with the clerk of the public agency and the office of the county clerk, or the mayor when applicable, and shall remain on file in those offices. One copy shall be placed on file with the applicable grand jury final report by, and in the control of the currently impaneled grand jury, where it shall be maintained for a minimum of five years. §933.05.
(a) For purposes of subdivision (b) of Section 933, as to each grand jury finding, the responding person or entity shall indicate one of the following:
(1) The respondent agrees with the finding.
(2) The respondent disagrees wholly or partially with the finding in which case, the response shall specify the portion of the finding that is disputed and shall include an explanation of the reasons therefor.
(b) For purposes of subdivision (b) of Section 933, as to each grand jury recommendation, the responding person or entity shall report one of the following actions:
(1) The recommendation has been implemented, with a summary regarding the implemented action.
(2) The recommendation has not yet been implemented, but will be implemented in the future, with a timeframe for implementation.
(3) The recommendation requires further analysis, with an explanation and the scope and parameters of an analysis or study, and a timeframe for the matter to be prepared for discussion by the officer or head of the agency or department being investigated or reviewed, including the governing body of the public agency when applicable. This timeframe shall not exceed six months from the date of publication of the grand jury report. “Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 10
(4) The recommendation will not be implemented because it is not warranted or is not reasonable, with an explanation therefor.
(c) However, if a finding or recommendation of the grand jury addresses budgetary or personnel matters of a county agency or department headed by an elected officer, both the agency or department head and the board of supervisors shall respond if requested by the grand jury, but the response of the board of supervisors shall address only those budgetary or personnel matters over which it has some decision-making authority. The response of the elected agency or department head shall address all aspects of the findings or recommendations affecting his or her agency or department.
(d) A grand jury may request a subject person or entity to come before the grand jury for the purpose of reading and discussing the findings of the grand jury report that relates to that person or entity in order to verify the accuracy of the findings prior to their release. (e) During an investigation, the grand jury shall meet with the subject of that investigation regarding the investigation, unless the court, either on its own determination or upon request of the foreperson of the grand jury, determines that such a meeting would be detrimental. (f) A grand jury shall provide to the affected agency a copy of the portion of the grand jury report relating to that person or entity two working days prior to its public release and after the approval of the presiding judge. No officer, agency, department, or governing body of a public agency shall disclose any contents of the report prior to the public release of the final report.
RESPONSES REQUIRED
Comments to the Presiding Judge of the Superior Court in compliance with Penal Code Section 933.05 are required from:
Responses are required from the following governing body within 90 days of the date of the publication of this report:
90 Day Required Responses F1 F2 F3 F4 F5 F6
Santa Ana City Council x x x x x x
90 Day Required Responses R1 R2 R3
Santa Ana City Council x x x
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 11
APPENDIX
APPENDIX 1 – Approved and Operating Adult-Use Cannabis Dispensaries “Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 12
APPENDIX 2 – Commercial Cannabis Application Process
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 13
APPENDIX 3 – Phase 1 and Phase 2 Application and Permit Process “Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 14
APPENDIX 4 – Commercial Cannabis Application Information “Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 15
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 16
APPENDIX 5 – Santa Ana Ordinance No. NS-2959
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 17
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 18
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 19
“Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 20
GLOSSARY
Adult-Use Terminology used for “recreational” use to distinguish from “medicinal” use.
BCC California Bureau of Cannabis Control
Cannabis Cannabis refers to a group of three plants with psychoactive properties, known as cannabis sativa, cannabis indica, and cannabis ruderalis. When the flowers of these plants are harvested and dried, it becomes one of the most common drugs in the world. It is also referred to as weed, pot, or marijuana.
Cannabis Public
Benefit Fund The fund created by ordinance NS-2959 for revenue received in accordance with Measure Y.
Community
Benefits Program A voluntary, good-will program for dispensary owners and employees to provide donations or volunteer hours to the City of Santa Ana.
DOJ United States Department of Justice
Dispensary A retail store or business that sells Cannabis products.
Medicinal Use specifically for medical purposes such as treating PTSD or for relieving nausea caused by chemotherapy.
OCGJ Orange County Grand Jury
PLAN Community Benefit and Sustainable Business Practices Plan Regulatory Safety
Permit (RSP) A permit used by the City of Santa Ana pursuant to Chapter 40 of its Municipal Code.
Unlicensed Also known as “illegal” or “black market.” Dispensaries not licensed or legally authorized to sell Cannabis products to the public. “Pot Luck”: Santa Ana’s Monopoly on Licensed Retail Adult-Use Cannabis in Orange County 2020-2021 Orange County Grand Jury Page 21
BIBLIOGRAPHY
California Marijuana Legalization, Proposition 19 (1972). (1972, Nov 7). Retrieved 2021, from BALLOTPEDIA:
https://ballotpedia.org/California_Marijuana_Legalization,_Proposition_19_(1972) California Proposition 215, the Medical Marijuana Initiative (1996). (1996, Nov 5). Retrieved 2021, from BALLOTPEDIA:
https://ballotpedia.org/California_Proposition_215,_the_Medical_Marijuana_Initiative_( 1996)
ORDINANCE NO. NS -2929. (2017, Oct 17). Retrieved 2021, from SantaAna Granicus:
http://santaana.granicus.com/MetaViewer.php?view_id=2&clip_id=1414&meta_id=5450 9 Ordinance NS- 2959, Establishing a Cannabis Public Benefit Fund. (2018, Dec). Retrieved 2021, from SantaAna Granicus:
https://santaana.granicus.com/MetaViewer.php?view_id=2&event_id=8261&meta_id=12 2354
Proposition 64: The Adult Use of Marijuana Act. (2016, Nov). Retrieved 2012, from California Courts The Judicial Branch of California: https://www.courts.ca.gov/prop64.htm Santa Ana, California, Measure Y, Recreational Marijuana Business Tax (November 2018). (2018, Nov 6). Retrieved 2021, from BALLOTPEDIA:
https://ballotpedia.org/Santa_Ana,_California,_Measure_Y,_Recreational_Marijuana_Bu siness_Tax_(November_2018)
Text of Proposition 19, the "Regulate, Control and Tax Cannabis Act of 2010" (California). (2010). Retrieved 2021, from BALLOTPEDIA:
https://ballotpedia.org/Text_of_Proposition_19,_the_%22Regulate,_Control_and_Tax_C annabis_Act_of_2010%22_(California)
EXHIBIT 5
EXHIBIT 6
EXHIBIT 7
EXHIBIT 8
EXHIBIT 9
EXHIBIT 10
Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 1
DRUG ENFORCEMENT AND POLICY CENTER
Effects of Drug Policy Liberalization on Public Safety:
A Review of the Literature
Maria M. Orsini, Drug Enforcement and Policy Center, Moritz College of Law, The Ohio State University
EXECUTIVE SUMMARY
After decades of criminalization, cannabis policy has liberalized rapidly throughout the U.S. in the 21st century. Thus, there has been considerable speculation regarding the effects these reforms would have on crime and public safety.
This review summarizes extant literature on the effect of drug policy liberalization on crime, traffic safety, law enforcement, and racial disparities. Overall, the literature suggests that cannabis legalization has resulted in some benefits to public health and public safety, even while some studies have produced mixed findings with regard to particular outcomes.
KEY FINDINGS
Overall, the literature suggests that the liberalization of drug laws confers some benefits related to public safety and public health.
Much of the literature regarding the impact of marijuana legalization on crime shows promising effects, including decreases in violent and property crime, reductions in drug-related arrests, and an improvement in crime clearance rates. Yet, a caution might be warranted given some of the criminal justice data limitations.
Research on the relationship between cannabis policy liberalization and changes in traffic safety has produced mixed results. Some differences in traffic outcomes depend on the specific type of policy change, with certain jurisdictions observing increases in fatal collisions and others observing decreases. While some research has found increases in positive cannabis tests among drivers, this was not determined to be a major threat to road safety; alcohol remains a greater hazard.
Research on the effects of marijuana policy changes on law enforcement suggests that police workload required greater resources following legalization. Qualitative research on officer perceptions indicated several concerns upon implementation. Findings suggest that more education, training, and resources would be beneficial.
While drug policy liberalization has the potential to reduce racial disparities, further changes are needed for those benefits to be achieved. Research shows that while drug arrest rates among adults decreased following policy liberalization, racial disparities persisted.
All-drug decriminalization is associated with reductions in problematic drug use and criminal justice overcrowding, declines in youth drug use, and other health and social benefits, as evidenced by evaluations of Portugal’s policy. In the U.S., preliminary evidence suggests that in Oregon, where all drugs were decriminalized in 2020, arrests for drug offenses have declined while the number of people receiving services has increased.
Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 2
INTRODUCTION
After decades of criminalization, cannabis policy has liberalized rapidly throughout the U.S. in the 21st century.
Following cannabis legalization in Colorado and Washington, legalization has gained momentum in many other U.S. states. While some states have fully legalized recreational cannabis for adult use, others have only decriminalized or legalized medical use. These reforms may have significant effects on public safety and public health, and therefore have become a topic of considerable academic interest. This review summarizes extant literature on the effect of cannabis policy liberalization on crime, traffic safety, law enforcement, and racial disparities, while also exploring future directions such as the decriminalization of all drugs.
There are many potential benefits that have been advanced as reasons to liberalize drug policy. Proponents of the liberalization of cannabis laws have argued that public safety would improve after legalization, in part because it enables police to focus on more serious crimes (Makin et al. 2019; Wu, Li, and Lang 2022). Advocates have also speculated that legalization would decrease the jail population and lessen the disproportionate incarceration of minorities (Stanton et al. 2020). Supporters of legalization also point out that it would reduce the underground market and attendant criminal activity, decreasing the number of people who come into contact with underground markets (Maier, Mannes, and Kippenhofer 2017). Furthermore, legalization is said to increase tax revenue and save money previously wasted on the enforcement of marijuana laws. On the other hand, opponents of legalization have asserted that if cannabis were legalized, crime would increase and youth use would rise, thus harming public safety. To address these questions, many studies have researched the effects of marijuana legalization on crime and public safety using a variety of measures. While early studies were less conclusive as it was too early to properly assess the effects of the reforms, more recent work increasingly demonstrates that marijuana legalization has a variety of benefits with regard to public safety.
EFFECTS OF CANNABIS POLICY LIBERALIZATION
Crime Rates
Due to the nature of drug prohibition, the drug trade has been linked with violence and crime. Prohibition creates an environment in which formal dispute resolution is not possible, sometimes leading drug-involved parties to resort to violence (Jacques et al. 2016). However, drug legalization has the potential to reverse this relationship between drugs and crime by curtailing the black market and reducing the burden on the criminal justice system. Researchers have assessed the relationship between cannabis policy liberalization and crime using various measures such as arrest rates and crime clearance rates.
One study found that the decriminalization of cannabis in five U.S. states was associated with a substantial reduction in drug-related arrests among both youth and adults (Grucza et al. 2018), indicating that the policy achieved its intended effects without adverse consequences. The researchers also found no association between decriminalization and an increase in youth cannabis use during the studied period. A more recent study focused on Washington State found that following legalization, marijuana arrest rates dropped dramatically among adults 21 and older (Firth et al. 2019).
Research also suggests that drug legalization may decrease violent crime. For example, one study found that the implementation of medical marijuana laws led to a decrease in violent crime in states that bordered Mexico (Gavrilova, Kamada, and Zoutman 2019). The study used Uniform Crime Reporting (UCR) data to examine homicides, assaults, and robberies, and found that with the introduction of medical marijuana laws, violent crimes in border states were reduced by between 5 and 12.5 percent. Additionally, the authors found that drug-law-related homicides decreased, which supports the theory that lessened drug market activity is associated with reductions in crime. Furthermore, the authors predicted that full legalization in states like Colorado and Washington would have an even larger impact. A later study focusing on Colorado found that the opening of medical and recreational dispensaries decreased violent crime in nearby neighborhoods with incomes above the median (Burkhardt and Goemans 2019). The authors also found a decrease in non-cannabis drug- and alcohol-related crimes near dispensaries. While they found that vehicle break-ins were elevated within a mile of dispensaries, they concluded that marijuana legalization had a net benefit with regard to crime rates. An additional study focusing on recreational Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 3 legalization in Washington and Oregon found that legalization likely caused a drop in crime (Dragone et al. 2019).
Specifically, the authors found that legalization resulted in a significant reduction in rape and property crime on the Washington side of the border compared to both the Oregon side and the pre-legalization years. Furthermore, while marijuana consumption increased, use of other drugs and alcohol decreased.
Researchers have also evaluated potential spillover effects of legalization in one state to the neighboring nonlegalized states. For example, one study used UCR data between 2003 and 2017 to examine the potential spillover effect from legalization in Colorado and Washington, focusing on the changes in crime rates in border counties in neighboring states, and found some evidence of a crime reduction effect (Wu, Boateng, and Lang 2020). Specifically, the authors found significant decreases in property crime, larceny, and simple assault in the region that includes six states near Colorado. They also noted that the effects of cannabis legalization on crime rates in nearby states may vary based on the state and type of crime.
While much of the literature observed decreases in crime, some studies found increases in crime rates after cannabis legalization. For example, using UCR data from 2007 to 2017 to examine the effect of marijuana legalization on crime rates in Oregon, one study found increases in crime rates for several types of offenses, including property and violent crime (Wu, Wen, and Wilson 2021). In another study pertaining to crime in Oregon, Wu and Willits (2022) found that the rate of simple assault had increased following legalization. However, they noted that their post-legalization time frame was fairly short and should be reassessed by future research.
Other research found no significant changes in crime following marijuana legalization. For example, using UCR data, Lu et al. (2021) conducted a quasi-experimental study to examine crime rates in Colorado and Washington. They found no statistically significant effects of marijuana legalization on violent or property crime. Similarly, data from several legalized states indicated that violent crime neither increased nor dropped dramatically following cannabis legalization (Dills et al. 2021). Overall, the literature exploring the relationship between liberalization of marijuana policies and crime suggests that legalizing marijuana is not a threat to public safety (French et al. 2022).
Finally, as proponents of marijuana legalization have asserted that it would enable police to focus on more serious crimes, some studies have used crime clearance rates to assess the impact of this policy change. Examining cannabis legalization in Washington and Colorado, researchers conducted time-series analysis of UCR data between 2010 and 2015 and found that some crime clearance rates had improved following legalization (Makin et al. 2019). In another study focusing on Oregon, which was also among the first states to legalize, researchers used UCR data from 2007 to 2017 to test the effects of legalization on clearance rates for violent crimes (Wu, Li, and Lang 2022).
They found that clearance rates increased for violent crimes and aggravated assault in Oregon counties compared to those in non-legalized states. This supports the idea that legalization of cannabis would help crime clearance rates by allowing officers to reallocate their attention and resources to more serious crimes.
Traffic Safety
Although some media reports and opponents of legalization have suggested an increase in traffic accidents following cannabis policy changes, researchers have found mixed results depending on the type of policy and measures analyzed. Typically, studies have used metrics such as fatal collisions and positive drug tests to evaluate changes in road safety. To assess the effect of legalization on traffic fatalities in Colorado and Washington, Hansen, Miller, and Weber (2020) used a synthetic control approach with data on fatal traffic accidents between 2000 and 2016. Despite media reports which exaggerated this threat, the authors found little evidence to support the idea that recreational legalization dramatically increased traffic fatalities. Specifically, synthetic control groups had similar changes in marijuana- and alcohol-related traffic fatality rates, as well as a similar change in overall traffic fatalities, despite not having legal marijuana.
On the other hand, some studies have found varying results depending on the type of policy change. For example, a 2015 report from the Governors Highway Safety Association cited studies with conflicting results; cannabis-positive fatalities rose slightly in Colorado after legalization, while there was no change in traffic fatalities in California after decriminalization (Dills et al. 2021). However, research has suggested a substitution effect whereby any increase in cannabis-related fatalities may be offset by a reduction in fatalities driven by alcohol consumption. Thus, an important consideration in evaluating public safety is the net effect when accounting for both of these measures (Dills et al. 2021). Similarly, Windle et al. (2022) also found a difference between medical and recreational legalization with Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 4 regard to fatal traffic collisions. Specifically, medical legalization was associated with a decrease in fatal collisions, whereas recreational legalization was associated with an increase. The authors speculate that this difference may be attributable to the population subgroups that participate in each market. Additionally, while they had limited evidence related to the impact of marijuana policy changes on alcohol-related road safety measures, they suggest that medical marijuana legalization may be associated with a decrease in positive alcohol tests among drivers. Another study found that no significant changes in marijuana-related traffic accidents occurred following medical legalization (Lee, Abdel-Aty, and Park 2018). However, the authors found increases following other types of marijuana policy changes such as decriminalization and full legalization.
A number of studies have also examined positive cannabis tests as an outcome. After a systematic review of the effect of cannabis legalization and decriminalization on road safety, Windle et al. (2022) found that recreational and medical legalization, as well as decriminalization, were all associated with an increase in positive cannabis tests among drivers. However, they determined that many of these studies were at risk of bias due to potential confounders and measurement error. Moreover, the authors emphasized that although more drivers may have tested positive for cannabis, this does not necessarily indicate impaired driving due to the way the body processes tetrahydrocannabinol (THC) and the length of time that it can be detected. Similarly, French et al. (2022) also reviewed several studies that examined marijuana policy liberalization and its effect on road safety. Some studies found that cannabis-involved driving increased in some jurisdictions following medical marijuana laws; however, one of these studies also concluded that cannabis-involved driving was not significantly related to fatal crashes across the country. Finally, Jones et al. (2019) evaluated the effect of legalized medical marijuana using data on drug and alcohol screens from the Arizona State Trauma Registry. Compared with prelaw projections, results showed small annual increases in the proportions of drivers who tested positive for alcohol or marijuana after the law was passed.
However, despite small increases in positive marijuana tests among drivers after the law compared to prelaw projections, alcohol remained a more prevalent threat to road safety.
Law Enforcement
Marijuana policy changes also affect law enforcement practices and resourcing needs. Police officer workload in the wake of recreational legalization in Washington State has been explored to assess the extent to which this reform affected calls for service, an understudied measure compared to crime or arrest rates (Makin et al. 2021). Makin et al. (2021) provide this more nuanced analysis of officer workload by conducting an interrupted time-series analysis of two cities near the Washington border. They found that recreational marijuana was associated with changes in police resourcing (i.e., increased calls for service) in the legalized state’s city compared to the neighboring city in a nonlegal state. This suggested greater police resourcing needs post legalization.
Another avenue of research has explored law enforcement perceptions of cannabis legalization. Survey research on police departments in states neighboring Colorado after legalization found that officers in nearby states largely disapproved of legalization (Ward, Lucas, and Murphy 2019). Qualitative results indicated that officers were concerned about increases in plant and edible cannabis, trafficking, a strain on resources, and perceived increases in youth use. Furthermore, officers with less favorable perceptions of cannabis perceived greater enforcement challenges. Similarly, Stanton et al. (2022) conducted semi-structured interviews and found that police professionals felt largely unprepared for marijuana legalization, were concerned about youth use, and believed that traffic safety was an issue as a result of legalization. They felt that civic education with regard to marijuana was underdeveloped, and that law enforcement lacked proper training and resources.
Racial Disparities
As past researchers have noted, the War on Drugs in the U.S. has been a central part of the rise in mass incarceration and has had a disproportionate impact on communities of color (Alexander 2010). Thus, one potential benefit of the liberalization of drug laws is that it may reduce racial/ethnic disparities in criminal justice system contact. To test this assertion, Firth et al. (2019) used National Incident Based Reporting System (NIBRS) data on marijuana-related arrests. They found that marijuana arrest rates among people over 21 fell dramatically after legalization of marijuana possession in Washington State, and that rates stayed at similar levels following the opening of the retail market. However, while marijuana-related arrest rates for both White and Black adults decreased, relative Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 5 disparities increased. African Americans previously had an arrest rate 2.5 times higher than the White arrest rate, but this increased to 5 times higher after the opening of the retail market. Similarly, recent research on Colorado and Washington has also found that while cannabis-related arrest rates generally declined after legalization, racial disparities persisted (Willits et al. 2022). Thus, while legalization lessens the absolute number of people who come into contact with the justice system overall, more needs to be done to specifically address racial disparities. Another study used jail booking data to assess whether legalization had an impact on jail populations (Stanton et al. 2020).
The authors found that jail population trends varied by county over time and also differed with regard to the impact on minorities and women. In terms of racial disparities, they concluded that there was little positive change. Since legalization did not seem to have a substantial impact on jail populations, the authors suggest that future research may benefit from using different measures to more thoroughly parse the effects of legalization.
Additional work has explored the impact of adult cannabis legalization on disparities among youth and changes in juvenile justice outcomes. Using data from Oregon between 2012 and 2018, Firth et al. (2020) found that juvenile allegation rates increased overall, with rates being highest for American Indian/Alaska Native and Black youth. Prior to legalization, Black youth rates were twice that of White youth rates, and this disparity decreased after legalization.
However, allegation rates for Black youth still remained greater than rates for White youth. The disparity between White and American Indian/Alaska Native remained the same before and after legalization. These changes are not explained by changes in juvenile cannabis use; thus, the authors conclude that changing adult regulations may have unintended consequences for youth.
Finally, other research has tested the effects of drug policy changes beyond those specific to cannabis. For example, to examine the effects of California’s Proposition 47, which reclassified felony drug offenses to misdemeanors in 2014, Mooney et al. (2018) used data on all drug arrests in California between 2011 and 2016. Overall, drug arrest rates declined across all racial/ethnic groups, indicating a deprioritization of drug law enforcement. Absolute disparities also decreased between Black and White felony drug arrest rates. However, results showed that relative disparities increased in part due to the preexisting composition of felony offenses by race/ethnicity and the specific offense types that were reclassified (e.g., drug possession instead of sale). Ultimately, they concluded that reducing penalties for drug possession could lessen disproportionate felony convictions among Black people, which may help alleviate racial/ethnic disparities in criminal justice exposure and improve inequalities related to social determinants of health.
EFFECTS OF BROADER DRUG DECRIMINALIZATION EFFORTS
While most U.S. drug policy liberalization thus far has involved marijuana laws, some jurisdictions outside the U.S. have decriminalized all drugs. One notable example is Portugal, which decriminalized all drugs for personal use in 2001. Rather than being sanctioned with criminal charges, drug possession is considered a public order or administrative offense, and people are referred to a panel that consists of medical professionals, lawyers, and social workers. Hughes and Stevens (2010) examined the effects of this policy by consulting Portuguese evaluative documents as well as conducting interviews with a purposive sample of people in politics, health, and criminal justice.
They found that decriminalization was associated with reductions in problematic drug use and criminal justice overcrowding. Notably, problematic drug use declined following implementation, and drug use among youth also declined in the long run despite a brief uptick immediately following the reform. While drug use may have also risen slightly among adults since the reform, it is debatable to what extent this is attributable to the reform since there are confounding factors. Proponents of the reform argue that this relationship may be spurious and at least partially a reflection of greater reporting of drug use due to the lessened stigma associated with drugs. While more research needs to be conducted to fully evaluate the effects of decriminalization, current evidence suggests that decriminalization is not associated with substantial increases in drug use and reduces the burden on the criminal justice system while facilitating a number of other health and social benefits (Hughes and Stevens 2010).
In the U.S., Oregon spearheaded a similar effort with Measure 110, which passed in November 2020, making Oregon the first state in the U.S. to decriminalize drug possession. While opponents feared it would be harmful, supporters maintained that shifting away from prohibition and toward decriminalization would have numerous benefits such as restoring individual liberty, removing one of the structures that enables police abuse, reducing government waste, and moving toward treating problematic drug use as a health issue rather than a criminal issue (Sutton 2020). While it is still too early following the reform to draw definitive conclusions about its efficacy, preliminary evidence suggests Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 6 that the reform is successful (Sutton 2022). As of one year after the policy change, more than 16,000 people had received services; arrests for drug offenses had fallen by about 60%; and housing, peer support, and other harm reduction services had become more accessible (Sutton 2022). Future research will reevaluate the effects of these reforms, and should also carefully consider which metrics are used to evaluate such reforms (Netherland et al. 2022).
CONCLUSION AND FUTURE DIRECTIONS
There has been substantial public discourse around the liberalization of cannabis laws in the U.S., particularly as medical and recreational legalization have garnered greater support among the public. As these reforms have become more common, researchers have been able to collect and analyze more data to test many of the arguments that preceded implementation. Analyses of various aspects of these reforms are ongoing as sufficient time is needed following policy changes before researchers can effectively assess their impact. However, the literature suggests that cannabis legalization has resulted in many benefits to public health and public safety (Todd 2018), even while some studies have produced mixed findings with regard to particular outcomes. While future research should continue to assess cannabis-related reforms, attention should also be paid to the effects of policies such as those implemented in Portugal and, more recently, Oregon, as it became the first state in the U.S. to decriminalize all drugs. This type of drug policy liberalization stands to produce even greater public health and public safety benefits, but more data are needed to thoroughly understand these reforms as we move toward novel and more humane drug policy frameworks.
Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 7
REFERENCES
Alexander, Michelle. 2010. The New Jim Crow: Mass Incarceration in the Age of Colorblindness. New York, NY: The New Press.
Burkhardt, Jesse and Chris Goemans. 2019. “The Short-Run Effects of Marijuana Dispensary Openings on Local Crime.” The Annals of Regional Science 63:163-189.
Dills, Angela, Sietse Goffard, Jeffrey Miron, and Erin Partin. 2021. “The Effect of State Marijuana Legalizations: 2021 Update.” Cato Institute 908.
Dragone, Davide, Giovanni Prarolo, Paolo Vanin, and Giulio Zanella. 2019. “Crime and the Legalization of Recreational Marijuana.” Journal of Economic Behavior and Organization 159:488-501.
Firth, Caislin L., Anjum Hajat, Julia A. Dilley, Margaret Braun, and Julie E. Maher. 2020. “Implications of Cannabis Legalization on Juvenile Justice Outcomes and Racial Disparities.” American Journal of Preventive Medicine 58(4):562-569.
Firth, Caislin L., Julie E. Maher, Julia A. Dilley, Adam Darnell, and Nicholas P. Lovrich. 2019. “Did Marijuana Legalization in Washington State Reduce Racial Disparities in Adult Marijuana Arrests?” Substance Use & Misuse 54(9):1582-1587.
French, Michael T., Julia Zukerberg, Tara E. Lewandowski, Katrina B. Piccolo, and Karoline Mortensen. 2022. “Societal Costs and Outcomes of Medical and Recreational Marijuana Policies in the United States: A Systematic Review.” Medical Care Research and Review 00(0):1-29.
Gavrilova, Evelina, Takuma Kamada, and Floris Zoutman. 2019. “Is Legal Pot Crippling Mexican Drug Trafficking Organisations? The Effect of Medical Marijuana Laws on US Crime.” The Economic Journal 129(617):375-407.
Grucza, Richard A., Mike Vuolo, Melissa J. Krauss, Andrew D. Plunk, Arpana Agrawal, Frank J. Chaloupka, and Laura J. Bierut. 2018. “Cannabis Decriminalization: A Study of Recent Policy Change in Five U.S. States.” International Journal of Drug Policy 59:67-75.
Hansen, Benjamin, Keaton Miller, and Caroline Weber. 2020. “Early Evidence on Recreational Marijuana Legalization and Traffic Fatalities.” Economic Inquiry 58(2):547-568.
Hughes, Caitlin E. and Alex Stevens. 2010. “What Can We Learn From the Portuguese Decriminalization of Illicit Drugs?” British Journal of Criminology 50:999-1022.
Jacques, Scott, Rosenfeld, Richard, Wright, Richard, and Van Gemert, Frank. 2016. “Effects of Prohibition and Decriminalization on Drug Market Conflict: Comparing Street Dealers, Coffeeshops, and Cafés in Amsterdam.” Criminology & Public Policy 15(3):843-875.
Jones, Jefferson M., Ruth A. Shults, Byron Robinson, Kenneth K. Komatsu, and Erin K. Sauber-Schatz. 2019. “Marijuana and Alcohol Use Among Injured Drivers Evaluated at Level I Trauma Centers in Arizona, 2008– 2014.” Drug and Alcohol Dependence 204(107539):1-6.
Lee, Jaeyoung, Ahmad Abdel-Aty, and Juneyoung Park. 2018. “Investigation of Associations Between Marijuana Law Changes and Marijuana-Involved Fatal Traffic Crashes: A State-Level Analysis.” Journal of Transport & Health 10:194-202.
Lu, Ruibin, Dale Willits, Mary K. Stohr, David Makin, John Snyder, Nicholas Lovrich, Mikala Meize, Duane Stanton, Guangzhen Wu, and Craig Hemmens. 2021. “The Cannabis Effect on Crime: Time-Series Analysis of Crime in Colorado and Washington State.” Justice Quarterly 38(4):565-595.
Maier, Shana L., Suzanne Mannes, and Emily L. Koppenhofer. 2017. “The Implications of Marijuana Decriminalization and Legalization on Crime in the United States.” Contemporary Drug Problems 44(2):125-146.
Makin, David A., Mikala R. Meize, Dale W. Willits, Mary K. Stohr, Craig Hemmens, Nicholas P. Lovrich, Alexis Nordman, and Duane Stanton. 2021. “The Impact of Recreational Marijuana Sales on Calls for Service: An Analysis of Neighbouring Cities.” Policing and Society 31(7):848-862.
Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 8 Makin, David A., Dale W. Willits, Guangzhen Wu, Kathryn O. DuBois, Ruibin Lu, Mary K. Stohr, Wendy Koslicki, Duane Stanton, Craig Hemmens, John Snyder, and Nicholas P. Lovrich. 2019. “Marijuana Legalization and Crime Clearance Rates: Testing Proponent Assertions in Colorado and Washington State.” Police Quarterly 22(1):31-55.
Mooney, Alyssa C., Eric Giannella, M. Maria Glymour, Torsten B. Neilands, Meghan D. Morris, Jacqueline Tulsky, and May Sudhinaraset. 2018. “Racial/Ethnic Disparities in Arrests for Drug Possession After California Proposition 47, 2011-2016.” American Journal of Public Health 108(8):987-993.
Netherland, Jules, Alex H. Kral, Danielle C. Ompad, Corey S. Davis, Ricky N. Bluthenthal, Nabarun Dasgupta, Michael Gilbert, Riona Morgan, and Haven Wheelock. 2022. “Principles and Metrics for Evaluating Oregon’s Innovative Drug Decriminalization Measure.” Journal of Urban Health 99:328-331.
Stanton, Sr., Duane L., David Makin, Mary Stohr, Nicholas P. Lovrich, Dale Willits, Craig Hemmens, Mikala Meize, Oliver Bowers, and John Snyder. 2022. “Law Enforcement Perceptions of Cannabis Legalization Effects on Policing: Challenges of Major Policy Change Implementation at the Street Level.” Contemporary Drug Problems 49(1):20-45.
Stanton, Duane, Xiaohan Mei, Sohee Kim, Dale Willits, Mary Stohr, Craig Hemmens, Guangzhen Wu, Ruibin Lu, David Makin, and Nicholas Lovrich. 2020. “The Effect of Marijuana Legalization on Jail Populations in Washington State.” The Prison Journal 100(4):510-530.
Sutton, Matt. November 3, 2020. “Drug Policy Action’s Measure 110 Prevails, Making Oregon the First U.S. State to Decriminalize All Drugs & Expand Access to Addiction and Health Services.” Drug Policy Alliance. Retrieved from https://drugpolicy.org/press-release/2020/11/drug-policy-actions-measure-110-prevails-making-oregonfirst-us-state. Sutton, Matt. February 1, 2022. “One Year of Drug Decriminalization in Oregon: Early Results Show 16,000 People Have Accessed Services through Measure 110 Funding & Thousands Have Avoided Arrest.” Drug Policy Alliance. Retrieved from https://drugpolicy.org/press-release/2022/02/one-year-drug-decriminalizationoregon-early-results-show-16000-people-have. Todd, Tamar. 2018. “The Benefits of Marijuana Legalization and Regulation.” Berkeley Journal of Criminal Law 23(1):99-119.
Ward, Kyle C., Paul A. Lucas, and Alexandra Murphy. 2019. “The Impact of Marijuana Legalization on Law Enforcement in States Surrounding Colorado.” Police Quarterly 22(2):217-242.
Willits, Dale W., Brittany Solensten, Mikala Meize, Mary K. Stohr, David A. Makin, Craig Hemmens, Duane L.
Stanton, and Nicholas P. Lovrich. 2022. “Racial Disparities in the Wake of Cannabis Legalization:
Documenting Persistence and Change.” Race and Justice 0(0):1-18.
Windle, Sarah B., Peter Socha, José Ignacio Nazif-Munoz, Sam Harper, and Arijit Nandi. 2022. “The Impact of Cannabis Decriminalization and Legalization on Road Safety Outcomes: A Systematic Review.” American Journal of Preventive Medicine 000(000):1-16.
Wu, Guangzhen, Francis D. Boateng, and Xiaodong Lang. 2020. “The Spillover Effect of Recreational Marijuana Legalization on Crime: Evidence from Neighboring States of Colorado and Washington State.” Journal of Drug Issues 50(4):392-409.
Wu, Guangzhen, Yongtao Li, and Xiaodong (Eric) Lang. 2022. “Effects of Recreational Marijuana Legalization on Clearance Rates for Violent Crimes: Evidence from Oregon.” International Journal of Drug Policy 100(103528):1-6.
Wu, Guangzhen, Ming Wen, and Fernando A. Wilson. 2021. “Impact of Recreational Marijuana Legalization on Crime: Evidence from Oregon.” Journal of Criminal Justice 72(101742):1-11.
Wu, Guangzhen and Dale W. Willits. 2022. “The Impact of Recreational Marijuana Legalization on Simple Assault in Oregon.” Journal of Interpersonal Violence 0(0):1-22.
Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 9
APPENDIX
Alexander, Michelle. 2010. The New Jim Crow: Mass Incarceration in the Age of Colorblindness. New York, NY: The New Press.
Alexander’s book argues that the criminal justice system in the U.S. functions as a redesigned racial caste system. The underpinning of the War on Drugs was used as a justification for the criminal justice system to target Black men and destroy communities of color. While no longer acceptable to discriminate on the basis of race, it remained possible to discriminate against people who have criminal records. Thus, a new system of racial control was created wherein people who have contact with the justice system can face discrimination in areas including but not limited to employment, housing, education, and public benefits.
Burkhardt, Jesse and Chris Goemans. 2019. “The Short-Run Effects of Marijuana Dispensary Openings on Local Crime.” The Annals of Regional Science 63:163-189.
The authors examine legalization of marijuana use and distribution of criminal activity. They use a difference-in-differences design to test the effect of marijuana dispensary openings on local crime rates in Denver, Colorado. They find that the opening of dispensaries decreases violent crime rates in abovemedian-income neighborhoods. They also find that non-marijuana drug-related crimes decrease within a half-mile of new dispensaries, but do not increase within a half-mile to one mile of new dispensaries. Finally, they find that vehicle break-ins increase up to a mile away from new dispensaries.
Dills, Angela, Sietse Goffard, Jeffrey Miron, and Erin Partin. 2021. “The Effect of State Marijuana Legalizations: 2021 Update.” Cato Institute 908.
The authors review arguments around both support and opposition to legalization. They discuss how violent crime has neither increased nor dropped dramatically following cannabis legalization. Additionally, they review the effects of legalization on marijuana use, other drug use, road safety, and economic outcomes.
They conclude that there is minimal support for strong claims made by both opponents and supporters, with the exception of significant increases in tax revenue. They emphasize existing data limitations and the importance of future research for understanding the full impact of legalization.
Dragone, Davide, Giovanni Prarolo, Paolo Vanin, and Giulio Zanella. 2019. “Crime and the Legalization of Recreational Marijuana.” Journal of Economic Behavior and Organization 159:488-501.
This study focuses on recreational legalization in Washington and Oregon to assess the impact on crime.
Overall, they find that legalization likely causes a drop in crime. Specifically, the authors find that legalization resulted in a significant reduction in rape and property crime on the Washington side of the border compared to both the Oregon side and the pre-legalization years. While marijuana consumption increased, use of other drugs and alcohol decreased.
Firth, Caislin L., Anjum Hajat, Julia A. Dilley, Margaret Braun, and Julie E. Maher. 2020. “Implications of Cannabis Legalization on Juvenile Justice Outcomes and Racial Disparities.” American Journal of Preventive Medicine 58(4):562-569.
The authors assess changes in juvenile marijuana criminal allegation rates and racial disparities in Oregon following legalization for adults. The study finds that juvenile allegation rates increased overall, with allegation rates being highest for American Indian/Alaska Native and Black youth. Prior to legalization, Black youth rates were twice that of White youth rates, and this disparity decreased after legalization. However, for Black youth, allegation rates still remained greater than rates for White youth. The disparity between White and American Indian/Alaska Native remained the same before and after legalization. These changes are not explained by changes in juvenile cannabis use; thus, the authors conclude that changing adult regulations may have unintended consequences for youth.
Firth, Caislin L., Julie E. Maher, Julia A. Dilley, Adam Darnell, and Nicholas P. Lovrich. 2019. “Did Marijuana Legalization in Washington State Reduce Racial Disparities in Adult Marijuana Arrests?” Substance Use & Misuse 54(9):1582-1587.
This study used data on adult marijuana-related arrests in Washington State following legalization to assess changes in adult arrest rates as well as changes in racial disparities. They found that marijuana arrest rates among people over 21 fell dramatically following legalization of possession, and this stayed roughly the Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 10 same after the opening of the retail market. However, they found that while arrest rates for all adults decreased, relative disparities in arrest rates increased.
French, Michael T., Julia Zukerberg, Tara E. Lewandowski, Katrina B. Piccolo, and Karoline Mortensen. 2022. “Societal Costs and Outcomes of Medical and Recreational Marijuana Policies in the United States: A Systematic Review.” Medical Care Research and Review 00(0):1-29.
This article reviews the literature pertaining to effects of marijuana legalization on crime, road safety, employment, other drug use, tobacco use, alcohol use, and many other domains. With regard to criminal activity, the review finds that, overall, marijuana legalization is not a threat to public safety. The authors also detail numerous mixed findings regarding traffic safety following policy changes, concluding that in general marijuana legalization does not reduce road safety.
Gavrilova, Evelina, Takuma Kamada, and Floris Zoutman. 2019. “Is Legal Pot Crippling Mexican Drug Trafficking Organisations? The Effect of Medical Marijuana Laws on US Crime.” The Economic Journal 129(617):375- 407.
This study shows that medical marijuana laws lead to decreases in violent crime in U.S. states bordering Mexico. This reduction effect is strongest for crimes related to drug trafficking and for counties close to the border. They also find that medical marijuana laws in inland states are associated with crime reduction in the closest border state. This is consistent with the theory that decriminalization of cannabis would lead to less violent crime in markets traditionally controlled by Mexican drug trafficking organizations.
Grucza, Richard A., Mike Vuolo, Melissa J. Krauss, Andrew D. Plunk, Arpana Agrawal, Frank J. Chaloupka, and Laura J. Bierut. 2018. “Cannabis Decriminalization: A Study of Recent Policy Change in Five U.S. States.” International Journal of Drug Policy 59:67-75.
The authors examine the associations between cannabis decriminalization and arrests and youth cannabis use in five U.S. states that decriminalized between 2008 and 2014. Using federal crime statistics on cannabis possession arrests, as well as use data from Youth Risk Behavior Surveys between 2007 and 2015, the authors use a difference-in-differences framework. They find that decriminalization in five states resulted in substantial decreases in cannabis possession arrests for both adults and youth. Results also indicated no increase in youth cannabis use during the studied period.
Hansen, Benjamin, Keaton Miller, and Caroline Weber. 2020. “Early Evidence on Recreational Marijuana Legalization and Traffic Fatalities.” Economic Inquiry 58(2):547-568.
The authors examine effects of marijuana legalization on traffic fatalities in Colorado and Washington. They use a synthetic control method to compare outcomes in these states to other states that had similar economic and traffic trends before legalization. They use records on fatal traffic accidents between 2000 and 2016. They find that the synthetic control groups had similar changes in marijuana- and alcohol-related traffic fatality rates, as well as overall traffic fatality rates, despite not legalizing marijuana for recreational purposes. They explain that their estimates show little evidence to support the idea that recreational marijuana legalization caused traffic fatalities to increase dramatically as some media reports (e.g., Colorado news articles) conjectured.
Hughes, Caitlin E. and Alex Stevens. 2010. “What Can We Learn From the Portuguese Decriminalization of Illicit Drugs?” British Journal of Criminology 50:999-1022.
This paper examines drug decriminalization in Portugal which was implemented in 2001. The authors use Portuguese evaluative documents and interviews with key informants in the field. They also compare criminal justice and health effects with Spain and Italy. Contrary to opponents’ arguments, Portuguese decriminalization resulted in less problematic use and criminal justice overcrowding, as well as a reduction in drug-related harms.
Jacques, Scott, Rosenfeld, Richard, Wright, Richard, and Van Gemert, Frank. 2016. “Effects of Prohibition and Decriminalization on Drug Market Conflict: Comparing Street Dealers, Coffeeshops, and Cafés in Amsterdam.” Criminology & Public Policy 15(3):843-875.
This study examines rates of victimization, retaliation, and legal mobilization in three retail drug markets in Amsterdam, the Netherlands. The three markets include the legal alcohol trade of cafés, the decriminalized Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 11 market of coffee shops which sell cannabis, and the illegal street market. Data include interviews with 50 sellers in each market. Results indicate, unsurprisingly, that illicit dealers have the highest rates of victimization and retaliation, and the lowest rates of legal mobilization. Prohibition creates an environment in which formal dispute resolution is unavailable, and therefore victimization and retaliation are more common.
Jones, Jefferson M., Ruth A. Shults, Byron Robinson, Kenneth K. Komatsu, and Erin K. Sauber-Schatz. 2019. “Marijuana and Alcohol Use Among Injured Drivers Evaluated at Level I Trauma Centers in Arizona, 2008– 2014.” Drug and Alcohol Dependence 204(107539):1-6.
This study evaluated the effect of legalized medical marijuana using data on drug and alcohol screens from the Arizona State Trauma Registry. Compared with prelaw projections, results showed small annual increases in the proportions of drivers who tested positive for alcohol or marijuana after the law was passed.
However, despite small increases in positive marijuana tests among drivers after the law compared to prelaw projections, alcohol remained a more prevalent threat to road safety.
Lee, Jaeyoung, Ahmad Abdel-Aty, and Juneyoung Park. 2018. “Investigation of Associations Between Marijuana Law Changes and Marijuana-Involved Fatal Traffic Crashes: A State-Level Analysis.” Journal of Transport & Health 10:194-202.
The authors examine five types of marijuana policy changes and their effect on fatal traffic crashes. They analyze policies in five states which had different policy shifts: prohibition to medical legalization; prohibition to decriminalization; decriminalization which adds medical legalization; medical legalization to full recreational legalization; and both decriminalization and medical to full legalization. They find that while there were no significant changes in a move to medical marijuana legalization from prohibition, there were increases in marijuana-related crashes associated with all of the other policy shifts.
Lu, Ruibin, Dale Willits, Mary K. Stohr, David Makin, John Snyder, Nicholas Lovrich, Mikala Meize, Duane Stanton, Guangzhen Wu, and Craig Hemmens. 2021. “The Cannabis Effect on Crime: Time-Series Analysis of Crime in Colorado and Washington State.” Justice Quarterly 38(4):565-595.
The authors use a quasi-experimental multi-group interrupted time-series analysis to assess whether UCR crime rates in Colorado and Washington State were influenced. Results show that marijuana legalization and sales have had minimal to no effect on major crimes in Colorado or Washington. There were no statistically significant long-term effects of recreational cannabis laws or retail sales on violent or property crime rates in these states.
Maier, Shana L., Suzanne Mannes, and Emily L. Koppenhofer. 2017. “The Implications of Marijuana Decriminalization and Legalization on Crime in the United States.” Contemporary Drug Problems 44(2):125- 146.
The authors use 2014 UCR data to explore the relationship between decriminalization and marijuana legalization (recreational and medical) and crime rates and arrests for “drug abuse violations.” They found that the legal status of cannabis in states did not significantly predict property or violent crime rates.
Makin, David A., Mikala R. Meize, Dale W. Willits, Mary K. Stohr, Craig Hemmens, Nicholas P. Lovrich, Alexis Nordman, and Duane Stanton. 2021. “The Impact of Recreational Marijuana Sales on Calls for Service: An Analysis of Neighbouring Cities.” Policing and Society 31(7):848-862.
Previous research neglected a nuanced analysis of officer workload, specifically calls for service received and initiated by police. This study uses an interrupted time series to analyze data from two border-straddling cities: Pullman, WA, and Moscow, ID. Recreational marijuana is associated with changes in police resourcing in the legalized state setting and the magnitude of those changes vary based on call type.
Makin, David A., Dale W. Willits, Guangzhen Wu, Kathryn O. DuBois, Ruibin Lu, Mary K. Stohr, Wendy Koslicki, Duane Stanton, Craig Hemmens, John Snyder, and Nicholas P. Lovrich. 2019. “Marijuana Legalization and Crime Clearance Rates: Testing Proponent Assertions in Colorado and Washington State.” Police Quarterly 22(1):31-55.
This study uses 2010-2015 UCR data to conduct time-series analysis on the offenses cleared by arrest to create monthly counts of violent and property crime. The authors found no negative effects of legalization on Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 12 crime clearance rates. They also found that some crime clearance rates had improved. In sum, the study suggests that legalization resulted in improvements to some clearance rates.
Mooney, Alyssa C., Eric Giannella, M. Maria Glymour, Torsten B. Neilands, Meghan D. Morris, Jacqueline Tulsky, and May Sudhinaraset. 2018. “Racial/Ethnic Disparities in Arrests for Drug Possession After California Proposition 47, 2011-2016.” American Journal of Public Health 108(8):987-993.
This study examines the effects of California’s Proposition 47 on racial/ethnic disparities in drug arrests.
Proposition 47 reclassified felony drug offenses to misdemeanors in 2014. The authors use data on all drug arrests in California between 2011 and 2016. Drug law enforcement was likely deprioritized, as evidenced by a decline in total drug arrest rates. The authors conclude that reducing criminal penalties for drug possession could reduce racial/ethnic disparities in criminal justice exposure. They also posit that reducing criminal penalties would have implications for improving health inequalities related to social determinants of health.
Netherland, Jules, Alex H. Kral, Danielle C. Ompad, Corey S. Davis, Ricky N. Bluthenthal, Nabarun Dasgupta, Michael Gilbert, Riona Morgan, and Haven Wheelock. 2022. “Principles and Metrics for Evaluating Oregon’s Innovative Drug Decriminalization Measure.” Journal of Urban Health 99:328-331.
February 2022 marked one year since the enactment of Oregon’s measure which decriminalized possession of small amounts of all drugs, and the authors outline principles and metrics for evaluating the policy’s success. They interviewed people who use drugs in Oregon to understand how best to assess Measure 110. They caution against rushing to judgment too early as some news outlets began to prematurely report on the number of citations issued. The authors also advise researchers to take into account the complexities of policy implementation and to involve those who are most directly affected by such policies.
Stanton, Sr., Duane L., David Makin, Mary Stohr, Nicholas P. Lovrich, Dale Willits, Craig Hemmens, Mikala Meize, Oliver Bowers, and John Snyder. 2022. “Law Enforcement Perceptions of Cannabis Legalization Effects on Policing: Challenges of Major Policy Change Implementation at the Street Level.” Contemporary Drug Problems 49(1):20-45.
This study produces qualitative findings regarding the experience of law enforcement in a context of novel cannabis legalization. The authors find, through semi-structured interviews, that police professionals felt unprepared for marijuana legalization, were concerned about youth use, and believed that traffic safety was an issue as a result of legalization.
Stanton, Duane, Xiaohan Mei, Sohee Kim, Dale Willits, Mary Stohr, Craig Hemmens, Guangzhen Wu, Ruibin Lu, David Makin, and Nicholas Lovrich. 2020. “The Effect of Marijuana Legalization on Jail Populations in Washington State.” The Prison Journal 100(4):510-530.
Advocates have argued that legalization would lead police to focus their resources on other matters, thus decreasing the jail population and lessening the disproportionate incarceration of minorities. To test this assertion, the authors used jail booking data to conduct interrupted time-series regression models. They found that jail population trends differ among counties across time and regarding minorities and women.
Regarding racial/ethnic disproportionate impact, they found that there was little positive change.
Sutton, Matt. November 3, 2020. “Drug Policy Action’s Measure 110 Prevails, Making Oregon the First U.S. State to Decriminalize All Drugs & Expand Access to Addiction and Health Services.” Drug Policy Alliance. Retrieved from https://drugpolicy.org/press-release/2020/11/drug-policy-actions-measure-110-prevails-making-oregonfirst-us-state.
In this Drug Policy Alliance press release, Sutton summarizes Oregon Measure 110, the first all-drug decriminalization measure in the nation. Measure 110 refocuses efforts on public health rather than criminal sanctions, prioritizing health over punitive responses. Benefits of this measure include expanding access to evidence-based treatment and harm reduction services. The Oregon Criminal Justice Commission also speculated that this measure would decrease racial disparities in drug arrests.
Sutton, Matt. February 1, 2022. “One Year of Drug Decriminalization in Oregon: Early Results Show 16,000 People Have Accessed Services through Measure 110 Funding & Thousands Have Avoided Arrest.” Drug Policy Alliance. Retrieved from https://drugpolicy.org/press-release/2022/02/one-year-drug-decriminalizationoregon-early-results-show-16000-people-have. Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 13 This Drug Policy Alliance press release summarizes preliminary results of drug decriminalization in Oregon.
The press release marked the one-year anniversary of Oregon’s groundbreaking reform. While it is still too early following implementation to draw definitive conclusions about its efficacy, preliminary evidence suggests that the reform is successful. As of one year after the policy change, more than 16,000 people had received services; arrests for drug offenses had fallen by about 60%; and housing, peer support, and other harm reduction services had become more accessible.
Todd, Tamar. 2018. “The Benefits of Marijuana Legalization and Regulation.” Berkeley Journal of Criminal Law 23(1):99-119.
This article summarizes work that describes how marijuana-related offenses (e.g., possession, cultivation, distribution) dropped in states that legalized marijuana. Low-level court filings in Washington fell by 98% between 2011 and 2015. In Colorado, there was an 81% drop around the same time. Oregon and Alaska saw similar effects. The author points out that these states created jobs instead of arresting people.
Legalization also did not lead to increased youth use rates. Five years after legalization, arrests for alcohol DUIs and other drugs have declined in Washington and Colorado. Emerging research indicates that cannabis legalization is also associated with reductions in opioid overdose deaths and untreated opioid use disorder. Finally, states are acquiring hundreds of millions in taxes that are funding important public policy priorities such as education, school construction, behavioral health, and alcohol and drug prevention and treatment.
Ward, Kyle C., Paul A. Lucas, and Alexandra Murphy. 2019. “The Impact of Marijuana Legalization on Law Enforcement in States Surrounding Colorado.” Police Quarterly 22(2):217-242.
The goal of this study was to explore how recreational marijuana legalization affected law enforcement duties in police departments in neighboring states. The authors use a survey method. They find that law enforcement in states bordering Colorado disapproved of legal marijuana. The perceived impact on law enforcement fell into four categories: increases in plant and edible marijuana, concerns with marijuana trafficking, a strain on resources, and a perceived increase in juvenile use. Officers with less favorable perceptions of cannabis perceived higher enforcement challenges.
Willits, Dale W., Brittany Solensten, Mikala Meize, Mary K. Stohr, David A. Makin, Craig Hemmens, Duane L.
Stanton, and Nicholas P. Lovrich. 2022. “Racial Disparities in the Wake of Cannabis Legalization:
Documenting Persistence and Change.” Race and Justice 0(0):1-18.
The authors use UCR data from Colorado and Washington to explore racial disparities following cannabis legalization. Their findings show that there was a general decline in arrests for marijuana for most racial groups; however, racial disparities persisted after legalization. This finding was especially pronounced in Colorado. Thus, while legalization generally had a positive effect by decreasing contact with the criminal justice system, more work needs to be done to address racial disparities.
Windle, Sarah B., Peter Socha, José Ignacio Nazif-Munoz, Sam Harper, and Arijit Nandi. 2022. “The Impact of Cannabis Decriminalization and Legalization on Road Safety Outcomes: A Systematic Review.” American Journal of Preventive Medicine 000(000):1-16.
This article reviews the impact of cannabis decriminalization and legalization on road safety. The authors found that decriminalization, medical legalization, and recreational legalization were all associated with increases in positive marijuana tests among drivers. However, while this may reflect increases in marijuana use, it does not necessarily correspond to impaired driving. With regard to fatal traffic collisions, they found that medical legalization was associated with a decrease, while recreational legalization was associated with an increase.
Wu, Guangzhen, Francis D. Boateng, and Xiaodong Lang. 2020. “The Spillover Effect of Recreational Marijuana Legalization on Crime: Evidence From Neighboring States of Colorado and Washington State.” Journal of Drug Issues 50(4):392-409.
This study uses UCR data from 2003 to 2017 and a difference-in-differences analysis to examine the potential spillover effect of recreational marijuana legalization in Colorado and Washington, especially focusing on changes in the rates of various crimes in border counties of neighboring states (compared to nonborder counties of those states). Results reflect some evidence of a spillover crime reduction effect of Electronic copy available at: https://ssrn.com/abstract=4294150 DRUG ENFORCEMENT AND POLICY CENTER | THE OHIO STATE UNIVERSITY 14 legalization, as indicated by the significant decreases in property crime, larceny, and simple assault rates in the Colorado region that includes six neighboring states. Results also suggest that the effects of marijuana legalization on crime in neighboring states vary based on the crime type and state.
Wu, Guangzhen, Yongtao Li, and Xiaodong (Eric) Lang. 2022. “Effects of Recreational Marijuana Legalization on Clearance Rates for Violent Crimes: Evidence from Oregon.” International Journal of Drug Policy 100(103528):1-6.
The authors use a difference-in-differences analysis and synthetic control method to examine the effect of recreational cannabis legalization on clearance rates for various violent crimes in Oregon. Using UCR data from 2007 to 2017, they find some evidence suggesting a favorable impact of legalization on violent crime clearances. Specifically, there were significant increases in clearance rates for overall violent crimes and aggravated assault in Oregon counties (compared to those in non-legalized states). Results also indicated that the positive effect of legalization on violent crime clearance rates may lessen over time.
Wu, Guangzhen, Ming Wen, and Fernando A. Wilson. 2021. “Impact of Recreational Marijuana Legalization on Crime: Evidence from Oregon.” Journal of Criminal Justice 72(101742):1-11.
The authors use UCR data from 2007 to 2017 and difference-in-differences analysis to examine the impact of recreational marijuana legalization on the rates of various serious crimes in Oregon. Results show some evidence of a crime-exacerbating effect of recreational cannabis legalization, as indicated by increases in the rates of multiple types of serious crimes (relative to non-legalized states), including property and violent crime overall, and other crimes such as burglary, larceny, motor vehicle theft, and aggravated assault.
Wu, Guangzhen and Dale W. Willits. 2022. “The Impact of Recreational Marijuana Legalization on Simple Assault in Oregon.” Journal of Interpersonal Violence 0(0):1-22.
This study examines the impact of cannabis legalization on simple assault, which is considered a less serious form of violent crime. The authors use UCR data from 2007 to 2017 and a quasi-experimental research design. Results indicated that counties in Oregon had increases in the simple assault rate following legalization relative to rates in the 19 non-legalized states.
Electronic copy available at: https://ssrn.com/abstract=4294150 EXHIBIT 11
EXHIBIT 12
EXHIBIT 13
EXHIBIT 14
1 1. Introduction
The impact of liberalizing marijuana laws on crime is a subject of great political and scholarly debate. Advocates for policy reform in states considering liberalization laws, both medical marijuana laws (MMLs) allowing for dispensaries as well as policies promoting retail sale for recreational purposes, suggest that bringing marijuana markets out of the shadows of the black market is a clear net public safety gain.1Indeed such a position is supported by scholarly work seeking to identify a causal link between marijuana use and violence, but not finding any (Arseneault et al. 2000; Mulvey et al. 2006). Nonetheless, law enforcement agencies in jurisdictions that have already adopted dispensary systems for medical marijuana claim that these dispensaries are inextricably connected to crime (California Police Chiefs Association 2009; Ingold and Lofholm 2016; Powers 2014).
The difficulty in reconciling these two positions can to some extent be comprehended by understanding the different mechanisms through which marijuana liberalization laws might potentially influence crime. First, there is the obvious impact of legitimizing what was previously an illegal market. By transitioning marijuana transactions from illegitimate exchanges that had to be actively enforced to legitimate transactions, there is an immediate reduction in the burden of enforcement assuming the legal market replaces the black market (Miron and Zwiebel 1995; Shepard and Blackley 2005). Law enforcement and the Courts may then transition resources to other, arguably more violent, types of crimes (Huber, Newman, and LaFave 2016; Miron and Zwiebel 1995). Second, there is the potential for liberalization laws to influence crime rates through an increase in marijuana (mis)use (e.g. psychopharmacological crime), to the extent that marijuana use induces criminogenic behavior. While there is no clear causal link between 1 For example, see the “Issues” webpage for the Marijuana Policy Project: https://www.mpp.org/issues/. 2 marijuana use and criminogenic behavior, there is suggestive evidence for a positive correlation between use and property crime (Green et al. 2010; Pacula and Kilmer 2003).2 A third potential mechanism through which liberalization laws could plausibly influence crime, which might also explain the positive correlation between use and property crime, is that these liberalization ordinances enable the creation of new brick and mortar and delivery businesses that, because of the federal prohibition and banking laws that prevent (until recently) debit cards from being accepted in stores, operate entirely on a cash basis, creating strong targets for burglaries or thefts (California Police Chiefs Association 2009).
A whole new body of scholarly work has emerged exploiting the natural experiment created by new state laws that liberalize the sale of medical marijuana to examine this relationship.
As of November 2016, medical marijuana laws have been passed by 28 states plus the District of Columbia. The enactment of state laws since 1996 provide an opportunity to empirically test the effect of regulated markets on outcomes of interest. Studies have evaluated outcomes such as marijuana use (Chu 2014; Harper, Strumpf, and Kaufman 2012; Hasin et al. 2015; LynneLandsman, Livingston, and Wagenaar 2013; Pacula et al. 2015; Wall et al. 2011; Wen, Hockenberry, and Cummings 2015), crime rates (Chu and Townsend 2017; Gavrilova, Kamada, and Zoutman 2017; Huber, Newman, and LaFave 2016; Morris et al. 2014; Shepard and Blackley 2016), and other health outcomes (Anderson, Rees, and Sabia 2014; Chu 2015; Smart 2015). These studies all use a difference-in-difference methodology where the treatment is a change in a state law that loosens restrictions on the sale of marijuana. 2 Another important factor to consider is marijuana use and victimization, although any evidence of a causal link pointing to an increase in victimization has been inconclusive (Markowitz 2005; Office of National Drug Control Policy 2013).
3 The concern with these state-level studies is that many states, particularly the early adopting states, actually defer to local entities when it comes to regulating marijuana supply and production, which leads to variation in treatment within states (Dilley et al. 2017; Freisthler et al. 2013). For example, in Colorado and Washington State, which legalized the sale of recreational marijuana in 2014, various types of policies exist at the community level and a significant portion of the population live in communities where the sale of recreational marijuana is not allowed (Colorado Department of Revenue 2016; Dilley et al. 2017). Moreover, medicinal marijuana laws within a state may differ on important elements, such as bans on dispensaries and cultivation (Pacula, Boustead, and Hunt 2014). Therefore, studies that generalize a MML across the state are ignoring heterogeneity within their treatment sample, possibly leading to the inconsistent findings in the MML literature (Pacula et al. 2015). Given the localized nature of crime and the importance of levels of aggregation, this (mis)measurement of the treatment dosage is especially problematic for estimating effects on crime rates (Freeman, Grogger, and Sonstelie 1996; Hipp 2007).
The objective of this study is to investigate whether a particular element of MMLs, namely allowance for dispensaries, affects local crime and other indicators of marijuana misuse (i.e. driving under the influence). We are also able to identify other dimensions of MMLs, such as allowance for home cultivation, but due to little variation in these other dimensions, we focus on allowances for dispensaries.3 Moreover, we estimate effects on different types of crime, as well as arrests indicating marijuana use, to better understand the mechanisms driving the results. By utilizing a novel dataset that codifies elements of MMLs across local jurisdictions within California, we capture heterogeneity on the treatment variable that is present in other studies.
Moreover, by focusing on local variation within a single state with a long experience with the 3 We still control for whether a jurisdiction has limitations on home cultivation in all of our models. 4 policy, we are able to implicitly control for changing state norms and marijuana use that could be independently associated with marijuana-involved crime. Other state-level factors that could bias estimates of crime rates across states, such as depenalization of marijuana, are also implicitly controlled in our analysis (Huber, Newman, and LaFave 2016). This analysis can help inform policies at the local level, where regulation is usually enacted, that better balance safety and access to medical marijuana.
We find no evidence that ordinances allowing for marijuana dispensaries lead to an increase in crime. In fact, we see some evidence of a reduction in property crime along with an increase in DUI and misdemeanor marijuana arrests, pointing to possible increases in misuse of marijuana that do not result in more crime. Supplementary analyses indicate that the significant effects may be driven by pre-existing trends, so we limit our conclusions to the fact that counties allowing dispensaries did not experience an increase in crime. The rest of this paper proceeds as follows: Section 2 provides some background into the literature on regulation of marijuana markets and crime, Section 3 describes the methods used for the analysis, Section 4 provides results, and Section 5 concludes with a discussion of policy implications and limitations.
1. Background
2.1. Why Might Dispensary Laws Affect Reported Crime Rates and Arrests? It is clear that explicitly writing into law that entities are permitted to engage in retail distribution of medical marijuana reduces the criminal justice risks of supplying marijuana.
Theoretically, we might expect this to increase availability and access to marijuana, which could increase demand at both the extensive, and potentially the intensive margins. Indeed there is consistent evidence that laws on-the-books explicitly permitting entities to produce and distribute medical marijuana increase non-medical use of marijuana among adults (Hasin et al. 2017; Pacula 5 et al. 2015; Wen, Hockenberry, and Cummings 2015). The laws appear to have no general impact on youth marijuana use (Choo et al. 2014; Hasin et al. 2015; Lynne-Landsman, Livingston, and Wagenaar 2013; Anderson, Hansen, and Rees 2015; Shu-Acquaye 2016), although there is some evidence from studies accounting for the relative size of these evolving marijuana market that larger and/or more competitive markets do in fact influence youth marijuana use (D'Amico et al. 2015; Smart 2015); It is a priori unclear, however, what effect this increased use among adults may have on community-level violent and property crime or driving under the influence.
Moreover, the replacement of a black market by a new cash-based business may or may not lead to a change in reported robberies, burglaries, and thefts (California Police Chiefs Association 2009). Although we cannot hypothesize on the overall directional change, by applying the Goldstein (1985) typology of drug crime, we consider the potential mechanisms driving changes in crime.
The first set of crimes in the typology is those committed due to intoxication, or psychopharmacological crimes. As was summarized in a recent report by ONDCP, there is little evidence for a causal link between marijuana intoxication and pharmacological crime (Office of National Drug Control Policy 2013, 14). Marijuana has been linked to correlates of violence such as development of psychosis disorders, aggression later in life, and delinquent behavior (Arseneault et al. 2000; Hall and Degenhardt 2008; White and Hansell 1998); but laboratory studies have not found a link between cannabis and violence (Moore and Stuart 2005) and there is reason to believe that marijuana use alone may lower the propensity to commit an aggressive act (Ostrowsky 2011). If marijuana is a substitute for alcohol, then increased availability of marijuana through retail outlets may lead to substitution away from alcohol, thereby reducing crime that would otherwise be associated with alcohol intoxication (Carpenter and Dobkin 2010; Carpenter 6 2007) and DUIs (Anderson, Hansen, and Rees 2013). However, if alcohol and marijuana are complements (Pacula 1998; Williams et al. 2004), it is possible that their joint consumption could lead to more aggressive behavior than alcohol or marijuana alone. Therefore, on net, we are a priori ambivalent towards the expected directional change (if any) in reported crime and DUI arrests due to pharmacological criminality, resulting from legal dispensaries or looser rules on cultivation.
A second type of crime in the typology is “economic-compulsive” crimes caused by those in need of income to pay for a drug habit (Goldstein 1985). We can expect that legalization of marijuana, even for medical purposes, will reduce the price of obtaining marijuana, and indeed there is limited evidence showing that potency has risen while prices for potency-adjusted fixed amounts have fallen (Anderson, Hansen, and Rees 2013; Sevigny, Pacula, and Heaton 2014).
Substantially larger price declines have been observed with full legalization (Smart et al. 2017).
Overall, we would expect a minimal increase in income-producing property crime driven by economic-compulsive behavior as a result of legalizing dispensaries.
The third category of crime is “systemic crimes,” or those associated with the provision and distribution of drugs in black markets. There is very limited evidence of violence attributed to illicit retail marijuana markets, although a recent study has found that counties bordering Mexico in states that passed MMLs have experienced a decrease in violent crime by decreasing the financial incentives of drug trafficking organizations (Gavrilova, Kamada, and Zoutman 2017; Reuter 2009). Nevertheless, any possible violence would have likely been mitigated in California given that home cultivation was allowed in most counties for many years before dispensary laws.4 4 Our models will control for whether the county had any restrictions on self-cultivation. 7 However, the presence of dispensary store-fronts may, themselves, lead to a change in both violent and property crimes in a given area, although again the direction is theoretically ambiguous. On the one hand the sale of marijuana, even for medicinal purposes, is illegal to the federal government. Therefore, no banks with a national charter are willing to provide credit or regular services to dispensaries that sell these goods. This has meant that most dispensaries must operate on a cash-basis, and they have a lot of cash (California Police Chiefs Association 2009; McDonald and Pelisek 2009). On the other hand, dispensaries have lots of security in and around them because of their cash business and highly desirable product. They often are zoned in areas that previously had empty buildings, and so by moving into these areas and bringing their own security systems they provide more “eyes on the ground” which can deter crime.
Spatial models that measure the density of dispensaries in a given area are an effective way to test the effect of store fronts on crime, but the results from these have been quite mixed. A few correlational studies have found a positive relationship with crime either in the immediate area (Contreras 2016) or in adjacent neighborhoods (Freisthler et al. 2016). A recent study exploiting an exogenous shock that led to closings of dispensaries in Los Angeles County, though, found that these closures actually led to an increase in crime in the immediate vicinity (Chang and Jacobson 2017). The authors argue that the increase in crime was a result of a loss of “eyes on the street” being provided by the dispensaries that were forced to close. Overall, we cannot say whether we expect a change in the distribution of marijuana caused by the legalizing dispensaries to have a positive or negative effect on reported crime in California.
Because theory does not provide any clear guidance on anticipated effects of these laws, it has been viewed an empirical question. A recent set of studies examine the relationship between MMLs and crime by exploiting variation in uptake across states and using Part I reported crimes 8 at the aggregate level. Morris et al. (2014) and Shepard and Blackley (2016) both use a differencein-difference methodology, with the former employing a sample of all states in the period between 1990-2006 and the latter a sample of only the eleven states that make up the Western Census Region between 1997 and 2009 .5 Chu and Townsend (2017) adopt a similar approach while measuring crime at the city policy agency level to improve measurement, but still rely on a statelevel classification of MMLs. Huber (2016) add information about whether states have depenalized marijuana to their difference-in-difference model, arguing that depenalization has an effect on nondrug crime by shifting enforcement resources. Finally, Gavrilova, Kamada, and Zoutman (2017) employ a difference-in-difference-in-difference approach to study the added effect of MMLs on crime in counties bordering Mexico compared to inland adopting counties (where MMLs are measured at the state-level). These studies have mostly found very little evidence of a relationship between uptake of medical marijuana laws and reported crime,6 with the exception of the Huber study that estimated a 5% significant decrease in robberies, larcenies, and burglaries.
The current paper addresses important limitations of the prior studies. First, prior studies that exploit state-level policy variation assume the treatment (exposure to medical marijuana dispensaries) is homogenous across the state. It is clear from recent work that this is not the case, as many local jurisdictions choose to adopt bans on medical marijuana dispensaries (Dilley et al. 2017; Pacula et al. 2015). Crime rates are also not constant across a state, and in fact are very localized, which raises uncertainty as to whether variation in aggregated crime rates observed at the state level are being driven by the same or different jurisdictions in which medical marijuana dispensaries are allowed (Dilley et al. 2017; Freeman, Grogger, and Sonstelie 1996). Our study is 5 They argue that because up until 2009 only four states outside of the Western Region had passed a MML, a sample of only western states provides a more similar control group. 6 The Gavrilova study finds a significant decrease in violent crime in Mexico-bordering counties with MMLs, but a negligible insignificant effect on violent crime in inland counties with MMLs. 9 able to explicitly address this concern by examining more localized measures of crime and dispensaries at the county level. Second, prior state analyses frequently omit other relevant policy variables that are also changing during this time period, such as cannabis depenalization, that might also be important for determining crime and arrest rates (Huber, Newman, and LaFave 2016). Our study overcomes this limitation by exploiting variation within one very large state, thereby holding constant across our treated and control counties changes in other state-specific laws (Shepard and Blackley 2005).
2.2 California Experiment: Medical Marijuana Laws across California Counties In 1996, California was the first state to pass a law allowing for the legal possession and cultivation of marijuana for medicinal purposes. The initiative changed a section of the Health and Safety code to protect patients who used marijuana with the recommendation from a physician from state prosecution. Passed through a ballot initiative, Proposition 215 (later to be known as the Compassionate Use Act (CUA)), did not address any channels through which marijuana might be supplied or obtained due in large part because of its explicit contradiction with federal law, which still maintained a strict prohibition on marijuana for medicinal purposes by classifying it as a Schedule 1 drug.
A few factors encouraged county and city jurisdictions in California to start crafting their own medical marijuana regulations, creating the variation over time we exploit in this study. First, the lack of specificity in the CUA regarding the production and distribution of marijuana left local governments with the authority to adopt whatever regulations they felt was appropriate, as there was no state pre-emption of any local regulations (Freisthler et al. 2016). Second, the ambiguity of the state law meant that the distribution of marijuana within the state remained illegal unless localities specifically addressed the issue.
10 These two factors paved the way for subsequent policy decisions since the passage of the CUA in 1996 that have affected medicinal marijuana regulation in California and solidified the role of counties and cities to create their own laws related to medical marijuana dispensaries.
Statute SB 420, which provided legal protection to marijuana dispensaries operating within the state as of January 1, 2004, gave local jurisdictions the autonomy to decide whether and how to permit dispensaries. While it exempted the “collective or cooperative cultivation” from prosecution, it left it to local jurisdictions the authority to implement and regulate them (State of California. October 12, 2003, §1(b), (c)). The “Ogden memo,” published in October, 2009, strengthened the incentive localities had to develop clear regulations over dispensaries, as it specified that the Federal government would not prioritize prosecuting patients or caregivers that were acting in clear compliance with state laws (Ogden 2009). As California law delegated these authorities to local jurisdictions, this memo signaled to city and county governments that local ordinances regarding dispensaries would be binding.7 As demonstrated by the increase in counties after 2004 and 2009 with dispensary laws shown in Figure 1, the evolution of these rulings and decisions has spurred the variation in local policy that is currently observed within the state of California today.
7 The authority of local governments to regulate dispensaries in their jurisdiction was reinforced in 2013 after the Supreme Court of California ruled in the case of Riverside vs. Inland Empire Patients Health and Wellness Center (56 Cal. 4th 729 [2013]) that the city of Riverside had the right to abolish marijuana dispensaries through zoning laws.
11 2. Data and Methods
3.1. Data
City and County Ordinance Data
This study uses a database of medical marijuana legal provisions adopted across all 58 counties of California as well its most populous 14 cities (those with a population exceeding 200,000 residents). The database is based on legal analysis of the language in the public law versions of county ordinances that were adopted between the period January 1, 1997 through December 31, 2014. The year 1997 is used as a starting point because California’s statewide policy passed in November of 1996.
The legal database includes jurisdictions’ (dis)approval on provisions related to the distribution of cannabis supply-related products. By December 31, 2014, 28% (16 out 58) of jurisdictions had made legally effective a county ordinance permitting co-operatives or dispensaries.8In order to limit the subjectivity of the database associated with subsequent implementation of the provisions and to ensure every jurisdiction’s county ordinances were assessed along the same criteria, e.g. as written in public version, this study does not include successive interpretations of courts or policies established by regulatory bodies.
An indicator variable was created that is defined as “1” for counties that explicitly allow dispensaries in a given year and “0” otherwise. This is a reasonable definition because none of the state-level statutes or court rulings explicitly allow for dispensaries. Inevitably, some counties changed their policy throughout the year and we have created an annual dataset, so we use the law 8 The peak number of dispensaries open in one year is actually 18 during 2013, but two counties stopped allowing for dispensaries the following year.
12 in place for the majority of the year.9 Additionally, because home cultivation is allowed by the Compassionate Use Act, we assume that home cultivation is allowed with no limits unless explicitly stated. We create a variable that identifies whether the county has placed any limits10 on cultivation for the given year. To make it easier to interpret along with the dispensary variable, the indicator variable is defined as “1” when there are no limits placed by the county on home cultivation and “0” when the county either explicitly prohibits or places any sort of limit on home cultivation.
A significant complication to the analysis is that a county ordinance applies to the unincorporated part of a county, which is the area of a county that does not pertain to an incorporated city. Incorporated cities may create different laws than the county they are nested in, which apply to residents within the city limits. Estimation of impacts of just county ordinances ignoring the specific ordinances of the cities incorporated within them could therefore lead to biased results. This is why in addition to the county ordinances, the research team also completed the same categorization procedure of ordinances for the 14 incorporated cities in California with a population larger than 200,000. Doing so meant that in most counties we would capture the laws that applied to the greatest share of the county population in each county.
While crime offense data are available at the level of police agencies within counties, our main analysis will be conducted at the county level because it mitigates the problem of agency jurisdiction borders and because more control variables are available at the county level. The distinction between counties and cities, however, means we have to be careful to account for the 9 The data on ordinances is at the monthly level, but because crime offense data should be interpreted at the annual level we collapse the ordinance data annually (very little variation is lost by collapsing the ordinance data to the annual level). We also show as a robustness check the results from estimating the treatment variable as the fraction of the year in which the policy was in place (for example, 0.5 if dispensaries were allowed for 6 months. 10 We categorize this variable as “any limits” because it is rare for counties to place limits on cultivation. Counties that place any limits, then, should be different than those that do not regulate cultivation. 13 treatment of cities that had laws different than that of the county (and for the possibility of differences for those cities in which we do not have information). Therefore, we construct a county-level policy indicator using the following rules: (1) the county unincorporated area policy is used if we don’t have information on any city within a county, or if the city information is consistent with the county; (2) when information on a city within a county is available and contradicts that of the county policy, we use the county or city policy that applies to the larger share of the full county population.11 As a sensitivity analysis, we collect data on police agencylevel reported crime for the 14 cities and the unincorporated areas of each of the 58 counties; and run a similar analysis using ordinances for these 72 independent jurisdictions to ensure that our results are not driven by different laws within jurisdictions in a county. We also conduct robustness checks using other methods of classifying the treatment variable.
Figure 1 displays the distribution of counties in each year that have a policy that allows for dispensaries, using the policy definition described above. It shows that the change in this element of the MMLs was a gradual process that, with the exception of 2 counties (San Francisco and Santa Clara), starts in 2004 after SB420 passed. By 2010, 12 of the 58 counties allow dispensaries and in 2013 18 counties do so. Finally, by the end of our sample period in 2014, 16 of the 58 counties allow for dispensaries. One may note that the number of counties allowing for dispensaries can decrease, and this is a function of the fact that counties that allow dispensaries may either (a) include sunset provisions, or (b) subsequently pass ordinances that disallow them. Our indicator variable reflects these subsequent changes as well.12
11 There are 3 counties (Alameda, San Diego, and Los Angeles) that have 2 cities within its boundaries that fit the criteria of a population of 200,000 or more. In these cases, we use the city with the larger population. 12 This is complicated in our event study analyses, but we account for it by adding a dummy variable to the model indicating years post-treatment when the county no longer allows for dispensaries. 14 Crime Incident Data
The second source of data we have compiled for use in this study are the number of total offenses reported to police by type of crime, for each county and each year in our sample period.
The data on reported offenses for the seven types of index crimes are pulled from the State of California Department of Justice (Criminal Justice Statistics Center 1997-2014) website. The California DOJ publishes raw county-level data13 from the information it receives from each police agency. We also create a variable for violent offenses that corresponds to the Uniform Crime Report (UCR) Part I violent crimes: homicide, rape,14 robbery, and aggravated assault; and property offenses refers to UCR Part I property crimes: burglary, larceny/theft,15 and motor vehicle theft.
Since the UCR is based on the Hierarchy Rule, only the most severe crime is counted per incident. The importance of this for our purposes is that if marijuana has an effect on the severity of crimes, we may observe this as a change in crime; although no change in the actual number of incidents. For example, where two offenses (e.g. aggravated assault and theft) occurred during an incident; this incident will be recorded as an aggravated assault. If marijuana results in a fall in pharmacological crime (such as aggravated assault), but still affects economic-crimes (theft), we would observe a decrease in aggravated assault and an increase in thefts. While in this scenario the former is true (there is indeed a fall in assault), the latter is not true; the offense of theft occurred in both incidents.
13 Raw data means that no imputation procedures are used to account for possible missing values. California does not conduct a state-wide version of the National Crime Victimization Survey, meaning that reported crime-offense reports is the only source for measuring the level of crime. 14 We don’t show results for effects on rape because there is no reason to believe dispensaries would have an effect and agencies in California were allowed to start using the new expanded definition in January 2014. 15 Larceny/theft includes both felony and misdemeanor crimes. The classification for felony theft in California changed in 2011 and it was not possible to go back and re-classify all previous felony offenses into misdemeanors, so the state decided to include all larceny and theft crimes, regardless of monetary value, under felony property crime.
15 Final Dataset
We have created a dataset of aggregate crime by year and county and merged it with the ordinance data to create a panel dataset from 1997 to 2014. Figure 2, Panels A, B, and C, track how reported violent offense, property offense, and DUI arrest rates, respectively, have changed over our sample time period for counties that started allowing dispensaries at any period between 1997 and 2014 and those that never allowed dispensaries during that same period. The vertical axes signify the two years (2004 and 2009) that begin an “episode” when more counties start entering the treatment group, as well as a year (2011) when various important state criminal justice policies are passed.16 The first important pattern to note is that crime has decreased in the state as a whole since 1997, regardless of whether dispensaries were allowed. This downward trend in crime per capita is consistent with the decline in crime that has been observed throughout the United States. Secondly, it appears that there are parallel trends throughout most of the pretreatment time period between the two groups of counties, with the exception that counties that never allowed dispensaries had a larger decline in violent and property offense rates between 1999 and 2001 than counties that allowed dispensaries at any point in time, and DUI arrests between the two groups start to converge around 2009. Nevertheless, because jurisdictions start allowing dispensaries at different times, it is difficult to draw conclusions about the relationship between dispensaries and crimes from these broad state trends.
We also collect data on variables at the county level that have been shown in the literature to influence the crime rate. These variables include the one-year lagged unemployment rate ( ), the average per capita income, the density of alcohol outlets per capita (Gruenewald and Remer 2006), and the county population density (Shepard and 16 These include decriminalization of marijuana as well as AB 109, a major policy that led to a shift in resources among all law enforcement agencies.
16 Blackley 2005). Lastly, we include an indicator for 2011 and later, the year that California both decriminalized recreational marijuana use and substantially changed its criminal justice system through a process that has been termed “Public Safety Realignment.” The unemployment rate comes from the Bureau of Labor Statistics, the per capita income from the Bureau of Economic Analysis, the alcohol outlets from the California Department of Alcohol and Beverage Control, and the land area and population from the United States Census Bureau.
Table 1 shows the summary statistics for the different types of crime we are analyzing and the independent variables used in our model. One will note that most of the total property crime is made up of larceny/thefts and most of the total violent crime is made up of aggravated assaults.
3.2 Empirical Strategy
To test whether allowing dispensaries affects aggregated criminal activity, we analyze the impact of local dispensary laws on UCR reported violent and property crime rates as well as DUI and marijuana-related arrest rates.17 Using the county-year as our unit of analysis, we will capture the effect from a change in dispensary allowance through a staggered difference-in-difference approach. Counties become part of the treated group at different times as they adopt laws throughout our sample period, and the changes resulting from adoption are compared to a control group that never adopts. All our model specifications include county fixed effects, as we are confident that there are unique unobservable county characteristics, which may cause a spurious correlation between crime rates and policy adoption. To account for the fact that there are trends in crime and arrest rates that are common across counties, we also include in the model a continuous (annual) time variable and a second order term. We choose this specification, over the 17 DUIs include driving under the influence of any substance that may impair driving, so driving under the influence of marijuana is included in these figures..
17 more common method of including year dummy variables, to preserve more power after observing a clear quadratic trend in all crime.18 Finally, we control for various time-variant county characteristics, described in the previous section, that may be correlated with both changes in crime and a county’s propensity to adopt an ordinance allowing for dispensaries.
Our preferred specification is one that also adds county-specific time trends to the model.
If counties across the state had differing pre-treatment trends, this specification helps create a better fit of the data. Studies that examine crime as an outcome across states, including in the MML literature, have argued for the inclusion of these jurisdiction-specific trends (Chu and Townsend 2017; Gavrilova, Kamada, and Zoutman 2017; Raphael and Wint ). California is very diverse with counties that differ in economic, political, and demographic characteristics; creating differences in crime trends one would usually associate with states. Moreover, there were differential impacts of the Great Recession and Public Safety Realignment across counties because of these different characteristics, resulting in differential crime trends that we can see when we look at counties individually.
The model specification is represented by the equation
log ( ) = + + + + + + + +
where represents the logarithm of the reported crimes per 100,000 residents19 of crime type i for county c in year t. Our main treatment variable is represented by , an indicator for whether county c in year t allows for dispensaries. controls for the county-specific variation, Time and Time2control for state trends over the study’s time-period, and accounts for the county-specific trend (we will show results based on different functional forms used to model the trend). Cult 18 We run all the models including year dummies as well to ensure that the coefficients are not affected by this choice.
19We ran a variety of tests for model fit and found that this model best described the data generating process.
Additional models were also tested and available upon request. 18 controls for whether the county had a restriction in place on amount or location with regards to cultivation and Xict represents a vector of county time-varying covariates that have been shown to be associated with crime rates in the literature.20 The coefficient of interest, , estimates the average effect in reported offenses for counties that allowed dispensaries compared to those that did not.21 Finally, our models are robust to clustered standard errors.
A primary assumption in the difference-in-difference methodology is that of pre-policy parallel trends in outcomes, or that there are no variables in the error term correlated to the outcome as well as the decision for a jurisdiction to adopt a dispensary policy. If this type of policy endogeneity were occurring or if pre-policy trends in crime between the treated and untreated groups differed for other reasons, we’d expect the trend for policy-adopting jurisdictions to change before the passage of an ordinance, leading to a biased coefficient of the treatment variable. One advantage from our technique is that the treatment is staggered over time, mitigating the probability that something happened at the state level that affected both crime and county-specific entry into treatment. Moreover, as ordinances are legislative processes, it is likely that many factors are attributable to the passage that have nothing to do with changes in crime (Williams and Bretteville-Jensen 2014). Finally, dispensaries were adopted by large and small, urban and rural counties, which mitigates the concern that counties adopting dispensaries are inherently different.
As a check that the parallel trends assumption holds and to explore possible dynamic effects of treatment, we complement our average effect model with an event study. The event study 20 The covariates used are: An indicator for when California decriminalized marijuana starting in 2011, the density of alcohol outlets in the county, log of per capita income for the county, the lagged unemployment rate in the county, and the log of the population density in the county. 21 While there is a wide range in populations and urban density in counties across California, which could lead to variance in the error term that is not constant across observations, we decided not to incorporate a weighted least squares regression. Models incorporating a WLS regression, testing various weights, did not improve estimates for homoskedasticity, so we did not feel that we fully understood the structure of the variance component to properly adjust for it. Results from these tests are available upon request. 19 disaggregates Dit into a set of dummy variables indicating whether a county-year observation represents a certain number of years before or after treatment. The model specification will be the same as that shown above, to account for other state and county characteristics and trends that affect county-specific crime rates.
The event study allows for identifying potential endogeneity if there are significant effects in the years leading up to policy adoption. For example, it picks up effects from the years preceding the passage of the law if suppliers sense that the county legislature or law enforcement are amenable to dispensaries and start to operate before an ordinance is officially put in place. The event study model also addresses a limitation from our model in which the average effect may mask differences in the development stage of dispensary operations after implementation (Meer and West 2015). It may be the case that there is a lag in observed effects as development of dispensaries takes place in the first few years.
3.3 Local Jurisdictions and Sensitivity Checks
As mentioned in the section describing the data, the analysis incorporates a measure of crime and arrests at the county level even though it is not always the case that a dispensary ordinance applies throughout the entire county. We address this issue by also running our difference-in-difference model at the ecological level of police-agency jurisdictions, where the reported crime rates should reflect 100% of the geographical area defined by our treatment variable. We collected agency-level data on the reported offenses by crime type from the UCR database for each of the 14 cities for which we have ordinance information and for the unincorporated areas of each of the 58 counties. For the cities, we use offense data that are reported by the police department of the city (e.g. reported offenses according to the Los Angeles Police Department to measure crime in Los Angeles City). For the unincorporated parts of each county, 20 we use offense data reported by the Sheriff’s department of the county. This leaves us with crime rates from a total of 72 independent jurisdictions, which match the 72 jurisdictions for which we have information on ordinances regarding allowance for dispensaries.
We follow the same empirical model from the main analysis, where the explanatory variable of interest is now an indicator for whether dispensaries are allowed in each jurisdiction, for the 72 independent jurisdictions over the 18-year sample period. One complication of running the analysis at the level of individual police-agency jurisdictions is that the covariates used in the regressions from the previous section are not available at this geographic level. Nevertheless, this should not affect the results because the variations across years within jurisdictions for variables that affect crime rates are minimal and are mostly absorbed by the controls that exploit the panel data structure. In the current model, we incorporate dummies indicating the independent jurisdictions to control for unobserved heterogeneity across individual jurisdictions and the same continuous time variables as above to control for state-level changes. Lastly, we estimate standard errors robust to clustering at the county-level, as even across two independent agencies, there may be correlation within the same county.
We also apply other sensitivity checks that address less serious, but important, concerns.
First, we present results removing certain counties that may be different than the rest because they adopted a dispensary policy very early, even before the passage of SB 420. Second, we conduct robustness checks related to the issue of differential city ordinances contained within a county by estimating the model using other methods to choose the applicable ordinance for the jurisdiction.
Third, we address the difficulty in properly identifying the amount of time during a year in which the policy was active by presenting results of a model measuring the main treatment variable using fractional years based on the month that the policy went into effect. Finally, we estimate a model 21 with a sample consisting of only control counties and treated counties with a policy in place five or more years. Similar results to the main analysis would point to consistent effects on crime over the treatment period.
3. Results
4.1 County Level Crime and Arrests
We present in Table 2 the results of the average effect on overall violent and property crime based on specifications with no county-specific time trend (columns (1) and (4)), as well as with county-specific trends using linear (columns (2) and (5)) and quadratic functional forms (columns (3) and (6)).22 The first important result to observe is that our estimates are sensitive to an inclusion of the county-specific time trend, as it leads to an increase in the magnitude of the coefficients for both overall violent and property crimes. For property crimes, it changes a roughly zero effect size to at least a partially significant coefficient. The choice of functional form for the county-specific trends is less important, with coefficients that are roughly similar across the different specifications. As we found differential property crime trends in some counties in supplemental analyses (not reported here), we have greater confidence in models that adjust these series for the county-specific time trend.
We find no significant impact of dispensaries on violent crime in any of our models. Table 3 shows that even when we disaggregate by crime type, none of the violent crimes (columns 1-3) are affected by dispensary laws. The consistency of findings regardless of inclusion or exclusion 22 We also ran the model using a cubic functional form for the county-specific time trend, but don’t show the results for simplicity, as they are very similar to the quadratic functional form model. 22 of the county-specific time trend is reassuring, but not surprising in light of the more consistent trends observed across counties in these measures.
For property crimes, we see no effect from adopting dispensaries in the model excluding county-specific time trends. However, the model incorporating a linear trend shows a 5.1% statistically significant drop in reported property offenses during the years in which counties allowed for dispensaries, while the quadratic specifications shows a 6.3% decrease that is statistically significant. Further decomposing these results, Table 3 shows that the effect on property crime appears to be driven by a decrease in thefts.
Only a brief discussion of the other covariates is warranted. As previously mentioned, due to very lax regulations on cultivation, the variable for cultivation regulations only measures whether there were any explicit limits set by a county. While Table 2 does show a sharp drop in violent crime of almost 10% in counties that didn’t restrict cultivation, when county-specific trends are not included, this relationship becomes insignificant with the inclusion of time trends. The effects from the other covariates included in the model are difficult to interpret due to limited variation once the fixed effects and time trends are controlled for.23 Table 4 shows results for the effects on variables that may be informative with regards to marijuana (mis)use. We see a very strong and robust effect on DUI arrests, as adopting dispensary laws was associated with at least a statistically significant 7.7% increase in DUI arrests. 24 This effect increases when county-specific time trends are included in the model, with the preferred specification indicating a significant increase in DUIs of 9.1%. As DUIs in California (at least 23 Supplemental analyses not shown here reveal that nearly all of the variation in our other descriptors (more than 90%), with the notable exception of unemployment, can be captured by fixed effects and county-specific time trends. 24 While the increase in DUIs may be a result of changes in enforcement in counties that allowed for dispensaries, it is unlikely that there is a high correlation between the timing of dispensary laws and changes in DUI enforcement.
Many factors impact enforcement, and cultivation of marijuana was allowed in almost all counties well before dispensaries opened (Williams and Bretteville-Jensen 2014). 23 during the study period) apply to any substance use, this increase may have been a result of more marijuana-impaired driving arrests. This is equivalent to 65 more DUI arrests per 100,000 residents on average per year,25 as a result of dispensaries. Arrests for felony and misdemeanor marijuana arrests are noisy due to important changes across the state that led to an overall large drop in both types of arrests statewide. Our results demonstrate a significant increase in misdemeanor arrests with our preferred specification, though, which does reinforce the evidence of possible increases in marijuana misuse.
The event study analysis results, demonstrated in the panels in Figure 3 where the graphs show the effect of each individual year relative to the passage of a law, can help in interpreting the results described above. Note that the sample is not perfectly balanced; many counties adopted dispensary laws later in the sample period so they did not have as many years of post-treatment observations. The tails in the figures below, the values -3 and 4 on the x-axis, represent dummy variables that incorporate all the years before or after, respectively, relative to the year of adoption (0 value on the x-axis).
Panels A of Figure 3 show that for overall violent crime, the failure to observe an effect is not due to a violation of the parallel trends assumption. The effect sizes of for violent crime consistently include 0 in both the pre- and post-policy periods and do not demonstrate any clear trends. Panels B, C, and D, on the other hand, demonstrate pre-existing trends for property crime and DUI and misdemeanor marijuana arrests. Moreover, it appears from the left tail of the figures that, historically, counties that adopt dispensary laws have higher property crime rates and lower DUI and misdemeanor arrests than non-adopting counties, and that regression towards the mean was occurring before dispensaries were allowed. Possible policy endogeneity makes it difficult to 25 We took the average across non-adopting years for counties that would eventually adopt dispensaries for this calculation because counties that adopted dispensary laws had lower DUIs on average (see Figure 2). 24 measure the magnitude of any possible overall and dynamic effects, though the trends continuing past zero, even if not significant, point to the significant effects shown in Table 2.
4.2 Sensitivity Checks
In this section, we show the results from a variety of sensitivity checks that account for limitations to our main analysis. Each row in Table 5 shows the coefficient for the “allows dispensary” variable of a different analysis, with regressions run for property and violent crime, as well as DUIs, presented in the columns. We show these three outcomes because our main analysis has not demonstrated any significant effects on specific types of these crimes, with the exception of theft, which seems to track the property crime variable.2627 The results for our first sensitivity check, shown in the first row of Table 5, represent the average effect of allowing dispensaries when variables are measured at the police-agency jurisdiction level. We see that the coefficient magnitudes are similar to those of the main analysis, even though the DUI arrests and property crime variables are no longer significant. This may occur because our new unit of analysis is smaller, leading to more variation from year to year and noisier data. Overall, these results do not contradict those of the main analysis.
In the next two rows, we check for whether how we define the treatment variable changes our findings. “Unincorporated County” means that we identify the treatment based only on the county (i.e unincorporated part of the county) law even if a city exists within the county with a different law, and “City Always” defines a variable that uses the city law (if available) to identify treatment regardless of whether the unincorporated population is larger. The following row shows the results of a model allowing for the treatment variable to be a fraction if an ordinance was passed 26 We also ran these models on theft crimes and find similar results to those shown for property crime. 27 We use county-specific time trends instead of agency-specific because county rates should have smoother trends.
This decision has no impact on the results shown.
25 after January of that year. The two rows labelled “No San Francisco” and “No Santa Clara” show the results of analyses that exclude each of these counties. These two counties adopted dispensaries very early on, even before the enactment of SB 420, which might indicate something unique about them.28 Moreover, given the changing trend in crime over our sample period, the timing of their “post-intervention” may impact the results (even after adjusting for county-specific linear trends) in addition to the higher leverage demonstrated by San Francisco due to it experiencing more years of treatment. Finally, the last row presents the effects of dispensary laws when we restrict the treatment sample to counties with laws for five or more years. All of these sensitivity checks point to the same findings as our main analysis, indicating a significant increase in DUI arrests and decrease in reported property crime offenses. While the analysis excluding San Francisco leads to an insignificant coefficient for property crime, it is still negative and similar in magnitude to the other models.
4. Discussion
California is experimenting with opening recreational marijuana retail stores, which will make it the largest state (in population and size) to do so. Again, localities will get to decide where and how many stores are allowed to open in each of their jurisdictions. Insights from the opening of medical marijuana dispensaries may be useful for better understanding the likely impacts of opening these recreational stores, and could serve to help police agencies and the courts and correctional systems prepare.
This study improves upon the work conducted thus far evaluating the impact of retail medical marijuana stores on crime. We use a novel longitudinal local ordinance database that allows us to assess the extent to which types of violent, property, and substance abuse crime rates 28 In fact, Santa Clara County is unique in that it stops allowing dispensaries to operate after 3 years and then adopts a new ordinance allowing for dispensaries in 2011.
26 are impacted over time with the decision by local jurisdictions to allow dispensaries to open.
Consideration of local variation within a state where substantial differences exist in allowances is crucial but had been previously ignored in the literature. Moreover, by examining variation within a single state, we can account for important statewide changes that are also important for driving marijuana use and potentially crime, including rules related to cultivation and decriminalization.
Evidence from our statistical analysis of a quasi-experimental setting finds no impacts on any type of violent crime, although counties adopting local ordinances did potentially experience a small decrease in property crime and increase in DUI arrests. Due to evidence of pre-existing trends, it is not possible to make a conclusive statement about the magnitude of these effects.
Our study is not without its own limitations, however. A clear problem is that our policy indicator is not capturing the actual exposure to the law for the residents in a county, since cities within counties can adopt conflicting ordinances. Our analysis at the police-agency level suggests that, at the very least, we are not missing increases to reported crime due to incongruence in treatment exposure. It also does not provide enough evidence to refute our findings of increases in DUI arrests. Moreover, when we measure our treatment variable using two alternative methods, we find similar results.
Second and relatedly, our study does not empirically assess the impact of having many versus few dispensaries within a jurisdiction (i.e. the “intensive margin”). Studies focusing on dispensary density and crime in the immediate vicinity, though, have not been much more definitive, finding no effect on any crime (Kepple and Freisthler 2012), a negative relationship with property crimes (Chang and Jacobson 2017), and small increases on property and violent crimes in adjacent areas (Freisthler et al. 2016). We do know that within California, counties differed substantially in their approach to dispensary allowances, with some jurisdictions 27 significantly limiting the total number of dispensaries allowed from the beginning and others not imposing any thresholds until much, much later. The lack of annual store-front data (pertaining to density) makes it more difficult to interpret dynamic effects, as there is scant research on the length of time we should expect for dispensary laws to be fully implemented within a jurisdiction and whether there are threshold effects in terms of total number of open dispensaries. There is also little to no information about delivery services, and laws associated with delivery services. The impact of delivery services may cause property crimes to rise in areas outside of the immediate vicinity of the dispensaries, thereby influencing property theft crimes in jurisdictions outside of those choosing to adopt the policy.
Third, a significant limitation in all difference-in-difference analyses is that there is no direct mechanism to test whether the treatment variable is correlated to an unobserved variable that affects the outcome, leading to a violation of the parallel trends assumption. We performed an event study analysis that did not refute our conclusions in the case of violent crimes, although there was evidence of policy endogeneity for property crime and DUI arrests. Until the policy endogeneity is explicitly addressed, the magnitude of the true effect on these outcomes cannot be easily determined.
Our study appears to reinforce the conclusions from other studies that fail to find an increase in the type of crime predicted by law enforcement. We find no effects on burglary, robberies, or assaults, which are the types of crimes one would expect if dispensaries were prime targets as a result of their holding large amounts of cash. It is important to note, though, that it may merely be the case that crime is such a localized effect that there is too much variation even within our treatment exposure aggregated to the city or county level (Hipp 2007). Pertaining to our findings of potentially decreasing property crime rates, there is a theoretical reason for why 28 dispensary store-fronts may decrease crime. Dispensaries may open in otherwise desolate areas, creating foot traffic, or “eyes on the street,” that makes these areas safer (Chang and Jacobson 2017).
We do find some interesting preliminary results with respect to the relationship between dispensaries and DUIs. Anderson, Hanson, and Rees (2013) find that MMLs in Colorado led to a substitution away from alcohol use, but the potential positive relationship between dispensaries and DUI arrests we find in our analysis suggests that either increases in marijuana-impaired driving exceeded reductions in alcohol impaired driving (a hypothesis we find highly unlikely) or that the opening of dispensaries induced use of both substances among those who were willing to drive impaired (more likely). The latter interpretation would have important ramifications for crime rates, given the known association between using alcohol together with other illicit substances and violent behavior (Office of National Drug Control Policy 2013). It is possible that our null results mask an increase in violent crime due to concurrent use of marijuana and alcohol, which is being offset by other mechanisms such as a decrease in pharmacological crimes due to an increase in marijuana use alone.
As we can only measure an aggregate effect, future research should attempt to tease out the effects on crime due to different mechanisms and actions of local actors. While some attributes of dispensaries may have led to a reduction in crime compared to the status quo, other aspects may have promoted crime. Moreover, the effect on crime rates will depend on other actions taken on by the local policymakers, dispensary owners, and law enforcement. For example, dispensaries may have adopted home delivery methods, which would reduce the potential number of victims near dispensaries. There may have also been specific actions taken by police that prevented an increase in crime rates, and these should be identified. Further research that identifies elements of 29 MMLs along with more specific aspects of implementation can help policymakers respond with actions that address crime-promoting aspects of allowing for retail dispensaries.
Our findings indicate that policymakers should be careful in how they regulate the presence of dispensaries, while not jumping to the conclusion that dispensaries are clearly crime generating hot-spots. Similarly, while police are right to be wary about potential crime effects from the introduction of cash-dependent businesses, our results demonstrate that current policy has not led to a wave in crime (even if this may be due to actual police practices). Our findings suggest that it is possible to regulate these markets and find a common ground between safety and access to medical marijuana. Natural experiments like the one being undertaken in California will only further help researchers better understand exactly how to find this ideal common ground. 30 5. References
Anderson, D. Mark, Benjamin Hansen, and Daniel I. Rees. 2013. "Medical Marijuana Laws, Traffic Fatalities, and Alcohol Consumption." Journal of Law and Economics 56 (2):333- 369. doi: 10.1086/668812.
Anderson, D. Mark, Daniel I. Rees, and Joseph J. Sabia. 2014. "Medical Marijuana Laws and Suicides by Gender and Age." American Journal of Public Health 104 (12):2369-2376. doi: 10.2105/AJPH.2013.301612.
Anderson, Mark D., Benjamin Hansen, and Daniel I. Rees. 2015. "Medical Marijuana Laws and Teen Marijuana Use." American Law and Economics Review 17 (2):495-528. doi:
10.1093/aler/ahv002.
Arseneault, L., T. E. Moffitt, A. Caspi, P. J. Taylor, and P. A. Silva. 2000. "Mental disorders and violence in a total birth cohort: Results from the dunedin study." Archives of General Psychiatry 57 (10):979-986. doi: 10.1001/archpsyc.57.10.979.
California Police Chiefs Association. 2009. White Paper on Marijuana Dispensaries.
Carpenter, Christopher, and Carlos Dobkin. 2010. "Alcohol regulation and crime." In Controlling crime: Strategies and tradeoffs, 291-329. University of Chicago Press.
Carpenter, Christopher
Driving Laws." The Journal of Law and Economics 50 (3):539-557. doi: 10.1086/519809.
Chang, Tom Y., and Mireille Jacobson. 2017. "Going to pot? The impact of dispensary closures on crime." Journal of Urban Economics 100:120-136. doi:
http://dx.doi.org/10.1016/j.jue.2017.04.001. Choo, Esther K., Madeline Benz, Nikolas Zaller, Otis Warren, Kristin L. Rising, and K. John McConnell. 2014. "The Impact of State Medical Marijuana Legislation on Adolescent Marijuana Use." Journal of Adolescent Health 55 (2):160-166. doi:
https://doi.org/10.1016/j.jadohealth.2014.02.018. Chu, Yu-Wei Luke. 2014. "The effects of medical marijuana laws on illegal marijuana use." Journal of Health Economics 38:43-61. doi:
http://dx.doi.org/10.1016/j.jhealeco.2014.07.003. Chu, Yu-Wei Luke. 2015. "Do Medical Marijuana Laws Increase Hard-Drug Use?" The Journal of Law and Economics 58 (2):481-517. doi: doi:10.1086/684043.
Chu, Yu-Wei Luke, and Wilbur Townsend. 2017. Joint Culpability: The Effects of Medical Marijuana Laws on Crime. SSRN.
Colorado Department of Revenue. 2016. Annual Report 2016. edited by Colorado Department of Revenue.
Contreras, Christopher. 2016. "A Block-Level Analysis of Medical Marijuana Dispensaries and Crime in the City of Los Angeles." Justice Quarterly:1-27. doi:
10.1080/07418825.2016.1270346.
D'Amico, Elizabeth J., Jon M. Houck, Sarah B. Hunter, Jeremy N. V. Miles, Karen Chan Osilla, and Brett A. Ewing. 2015. "Group motivational interviewing for adolescents: Change talk and alcohol and marijuana outcomes." Journal of Consulting and Clinical Psychology 83 (1):68-80. doi: 10.1037/a0038155.
Dilley, Julia A., Laura Hitchcock, Nancy McGroder, Lindsey A. Greto, and Susan M. Richardson. 2017. "Community-level policy responses to state marijuana legalization in Washington State." International Journal of Drug Policy 42:102-108. doi:
10.1016/j.drugpo.2017.02.010.
31 Freeman, Scott, Jeffrey Grogger, and Jon Sonstelie. 1996. "The Spatial Concentration of Crime." Journal of Urban Economics 40 (2):216-231. doi: https://doi.org/10.1006/juec.1996.0030. Freisthler, Bridget, Nancy J Kepple, Revel Sims, and Scott E Martin. 2013. "Evaluating medical marijuana dispensary policies: Spatial methods for the study of environmentally-based interventions." American journal of community psychology 51 (1-2):278-288.
Freisthler, Bridget, William R. Ponicki, Andrew Gaidus, and Paul J. Gruenewald. 2016. "A microtemporal geospatial analysis of medical marijuana dispensaries and crime in long beach California." Addiction:n/a-n/a. doi: 10.1111/add.13301.
Gavrilova, Evelina, Takuma Kamada, and Floris Zoutman. 2017. "Is Legal Pot Crippling Mexican Drug Trafficking Organisations? The Effect of Medical Marijuana Laws on US Crime." The Economic Journal. doi: 10.1111/ecoj.12521.
Goldstein, Paul J. 1985. "The Drugs/Violence Nexus: A Tripartite Conceptual Framework." Journal of Drug Issues 15 (4):493-506. doi: 10.1177/002204268501500406.
Green, Kerry M., Elaine E. Doherty, Elizabeth A. Stuart, and Margaret E. Ensminger. 2010. "Does heavy adolescent marijuana use lead to criminal involvement in adulthood? Evidence from a multiwave longitudinal study of urban African Americans." Drug and Alcohol Dependence 112 (1–2):117-125. doi: http://dx.doi.org/10.1016/j.drugalcdep.2010.05.018. Gruenewald, Paul J, and Lillian Remer. 2006. "Changes in outlet densities affect violence rates." Alcoholism: Clinical and Experimental Research 30 (7):1184-1193.
Hall, Wayne, and Louisa Degenhardt. 2008. "Cannabis use and the risk of developing a psychotic disorder." World Psychiatry 7 (2):68-71. doi: 10.1002/j.2051-5545.2008.tb00158.x.
Harper, Sam, Erin C Strumpf, and Jay S Kaufman. 2012. "Do medical marijuana laws increase marijuana use? Replication study and extension." Annals of Epidemiology 22 (3):207-212.
Hasin, D. S., A. L. Sarvet, M. Cerdá, and et al. 2017. "Us adult illicit cannabis use, cannabis use disorder, and medical marijuana laws: 1991-1992 to 2012-2013." JAMA Psychiatry 74 (6):579-588. doi: 10.1001/jamapsychiatry.2017.0724.
Hasin, Deborah S., Melanie Wall, Katherine M. Keyes, Magdalena Cerdá, John Schulenberg, Patrick M. O'Malley, Sandro Galea, Rosalie Pacula, and Tianshu Feng. 2015. "Medical marijuana laws and adolescent marijuana use in the USA from 1991 to 2014: results from annual, repeated cross-sectional surveys." The Lancet Psychiatry 2 (7):601-608. doi:
http://dx.doi.org/10.1016/S2215-0366(15)00217-5. Hipp, John R. 2007. "Block, Tract, and Levels of Aggregation: Neighborhood Structure and Crime and Disorder as a Case in Point." American Sociological Review 72 (5):659-680. doi:
10.1177/000312240707200501.
Huber, Arthur, Rebecca Newman, and Daniel LaFave. 2016. "Cannabis Control and Crime:
Medicinal Use, Depenalization and the War on Drugs." The BE Journal of Economic Analysis & Policy 16 (4).
Ingold, John , and Nancy Lofholm. 2016. "Medical-marijuana dispensaries' effect on crime unclear." The Denver Post, March 12. http://www.denverpost.com/ci_17178820. Kepple, Nancy J., and Bridget Freisthler. 2012. "Exploring the Ecological Association Between Crime and Medical Marijuana Dispensaries." Journal of Studies on Alcohol and Drugs 73 (4):523-530.
Lynne-Landsman, Sarah D, Melvin D Livingston, and Alexander C Wagenaar. 2013. "Effects of state medical marijuana laws on adolescent marijuana use." American Journal of Public Health 103 (8):1500-1506.
32 Markowitz, Sara. 2005. "Alcohol, Drugs and Violent Crime." International Review of Law and Economics 25 (1):20-44. doi: https://doi.org/10.1016/j.irle.2005.05.003. McDonald, Patrick R., and Christine Pelisek. 2009. "L.A.'s Medical-Weed Wars." LA Weekly.
Meer, Jonathan, and Jeremy West. 2015. "Effects of the Minimum Wage on Employment Dynamics." Journal of Human Resources. doi: 10.3368/jhr.51.2.0414-6298R1.
Miron, Jeffrey A., and Jeffrey Zwiebel. 1995. "The Economic Case Against Drug Prohibition." The Journal of Economic Perspectives 9 (4):175-192. doi: 10.2307/2138396.
Moore, Todd M., and Gregory L. Stuart. 2005. "A review of the literature on marijuana and interpersonal violence." Aggression and Violent Behavior 10 (2):171-192. doi:
https://doi.org/10.1016/j.avb.2003.10.002. Morris, Robert G., Michael TenEyck, J. C. Barnes, and Tomislav V. Kovandzic. 2014. "The Effect of Medical Marijuana Laws on Crime: Evidence from State Panel Data, 1990-2006." PLoS ONE 9 (3):e92816. doi: 10.1371/journal.pone.0092816.
Mulvey, Edward P, Candice Odgers, Jennifer Skeem, William Gardner, Carol Schubert, and Charles Lidz. 2006. "Substance use and community violence: a test of the relation at the daily level." Journal of consulting and clinical psychology 74 (4):743.
Office of National Drug Control Policy. 2013. Improving the Measurement of Drug-Related Crime. Wishington, DC: Executive Office of the President.
Ogden, David W. 2009. Memorandum for Selected United States Attorneys. edited by Office of the Attorney General U.S. Department of Justice.
Ostrowsky, Michael K. 2011. "Does Marijuana Use Lead to Aggression and Violent Behavior?" Journal of Drug Education 41 (4):369-389. doi: 10.2190/DE.41.4.c.
Pacula, Rosalie L., Anne E. Boustead, and Priscillia Hunt. 2014. Words Can Be Deceiving: A Review of Variation among Legally Effective Medical Marijuana Laws in the United States. In Journal of Drug Policy Analysis. Pacula, Rosalie L., and Beau Kilmer. 2003. Marijuana and crime: Is there a connection beyond prohibition? : National Bureau of Economic Research.
Pacula, Rosalie L., David Powell, Paul Heaton, and Eric L. Sevigny. 2015. "Assessing the Effects of Medical Marijuana Laws on Marijuana Use: The Devil is in the Details." Journal of Policy Analysis and Management 34 (1):7-31. doi: 10.1002/pam.21804.
Pacula, Rosalie Liccardo. 1998. "Does increasing the beer tax reduce marijuana consumption?" Journal of Health Economics 17 (5):557-585. doi: https://doi.org/10.1016/S0167- 6296(97)00039-8. Powers, Scott. 2014. "Sheriffs take lead against legalizing medical pot." Orlando Sentinel, May 1. http://articles.orlandosentinel.com/2014-05-01/news/os-medical-marijuana-opponentsemerge-20140423_1_amendment-2-florida-sheriffs-association-law-enforcement. crime*." Journal of Law and Economics 44 (1):259-283.
Reuter, Peter. 2009. "Systemic Violence in Drug Markets." Crime Law and Social Change 52 (3):275-284. doi: 10.1007/s10611-009-9197-x.
Sevigny, Eric L., Rosalie Liccardo Pacula, and Paul Heaton. 2014. "The effects of medical marijuana laws on potency." International Journal of Drug Policy 25 (2):308-319. doi:
http://dx.doi.org/10.1016/j.drugpo.2014.01.003. Shepard, Edward M, and Paul R Blackley. 2005. "Drug Enforcement and Crime: Recent Evidence from New York State*." Social Science Quarterly 86 (2):323-342. 33 Shepard, Edward M., and Paul R. Blackley. 2016. "Medical Marijuana and Crime: Further Evidence From the Western States." Journal of Drug Issues 46 (2):122-134. doi:
10.1177/0022042615623983.
Shu-Acquaye, Florence. 2016. "The Role of States in Shaping the Legal Debate on Medical Marijuana." Mitchell Hamline Law Review 42:697-755.
Smart, Rosanna. 2015. "The Kids Aren't Alright, But Older Adults Are: How Medical Marijuana Market Growth Impacts Adult and Adolescent Substance-Related Outcomes." doi:
http://dx.doi.org/10.2139/ssrn.2574915. Smart, Rosanna, Jonathan P. Caulkins, Beau Kilmer, Steven Davenport, and Greg Midgette. 2017. "Variation in cannabis potency and prices in a newly legal market: evidence from 30 million cannabis sales in Washington state." Addiction 112 (12):2167-2177. doi:
10.1111/add.13886.
Wall, Melanie M, Ernest Poh, Magdalena Cerdá, Katherine M Keyes, Sandro Galea, and Deborah S Hasin. 2011. "Adolescent marijuana use from 2002 to 2008: higher in states with medical marijuana laws, cause still unclear." Annals of epidemiology 21 (9):714-716.
Wen, Hefei, Jason M. Hockenberry, and Janet R. Cummings. 2015. "The effect of medical marijuana laws on adolescent and adult use of marijuana, alcohol, and other substances." Journal of Health Economics 42:64-80. doi:
http://dx.doi.org/10.1016/j.jhealeco.2015.03.007. White, Helene Raskin, and Stephen Hansell. 1998. "Acute and long-term effects of drug use on aggression from adolescence into adulthood." Journal of Drug Issues 28 (4):837-858.
Williams, J., Rosalie Liccardo Pacula, Frank J. Chaloupka, and Henry Wechsler. 2004. "Alcohol and marijuana use among college students: economic complements or substitutes?" Health Economics 13 (9):825-843. doi: 10.1002/hec.859.
Williams, Jenny, and Anne Line Bretteville-Jensen. 2014. "Does liberalizing cannabis laws increase cannabis use?" Journal of Health Economics 36:20-32. doi:
http://dx.doi.org/10.1016/j.jhealeco.2014.03.006.
34 6. Figures
Figure 1: Number of Counties that Allow Medical Marijuana Dispensaries, by Year Note: This figure represents the method of using a county’s ordinance unless there is a city in the county that has the largest share of the population in the county. 0 5
10 15
20 Number of Counties
35 Figure 2: Crime Rates per 100,000 residents, by Whether County Ever Allows Dispensaries Panel A: Total Violent
Panel B: Total Property
36 Panel C: DUI Arrests
Notes: The dashed line represents the counties that ever allow dispensaries in the sample period.
The solid line represents counties that up until the end of the sample period, had never allowed dispensaries. The vertical lines represent the years 2004 and 2009 because these were important transition years.
37 Figure 3: Event Study
Notes: “Year 0,” indicating year when policy adopted, is omitted from the regression to provide an excluded category. Counties never adopting a law have a 0 for all indicator variables. Dummy variables for 3+ years pre- and 4+ years post-intervention are represented by -3 and 4, respectively.
Regressions also include the covariates from the model described in Section 3.2. An indicator variable in the model controls for county-year observations for periods after an adopting county no longer allows for dispensaries.
38 7. Tables
Table 1: Summary Statistics
Mean Standard Deviation
All Violent Crime Rate 433.3 156.3
Homicide Crime Rate 4.3 2.2
Robbery Crime Rate 85.7 82.7
Assault Crime Rate 312.7 109.3
All Property Crime Rate 2,866.1 980.0
Burglary Crime Rate 735.8 225.8
Theft Crime Rate 1,768.1 695.8
Motor Vehicle Theft Crime Rate 362.3 219.1
DUI Arrests 749.4 346.1
Felony Marijuana Arrests 59.5 61.0
Misdemeanor Marijuana Arrests 140.6 110.7
Alcohol Outlet Density 32.4 27.6
Per Capita Income 34,852.5 11,069.5
Unemployment Rate 9.0 3.1
Population Density 659.1 2,298.9
All rates are calculated per 100,000 residents. Alcohol outlet density is calculated as the number of outlets per 10,000 people in the county. Population density is calculated as the number of people per square mile of land area in the county. All violent crime rate includes rape crimes even though we do not study the effect on rape crimes alone. 39 Table 2: Effect of Allowing Dispensaries on Violent and Property Crime Rates Violent Offenses Property Offenses (1) (2) (3) (4) (5) (6) Allows Dispensaries 0.006 -0.020 -0.013 0.008 -0.051 -0.063* (0.065) (0.071) (0.064) (0.029) (0.025) (0.019) Cultivation-No Limits -0.096 -0.047 -0.048 -0.044 -0.056 -0.049 (0.043) (0.054) (0.055) (0.037) (0.038) (0.037) Alcohol Outlet Density -0.000 -0.000 0.000 0.001 -0.000 -0.000 (0.001) (0.002) (0.001) (0.001) (0.001) (0.001) Post CJ Reforms -0.137* -0.154** -0.151** 0.081* 0.084* 0.072* (0.042) (0.044) (0.043) (0.026) (0.025) (0.026) Population Density -0.871 1.495 1.161 -0.072 0.912 1.591** (0.432) (1.292) (0.894) (0.197) (0.630) (0.416) Per Capita Income -0.300 0.043 -0.139 -0.209 -0.122 -0.026 (0.364) (0.267) (0.301) (0.296) (0.182) (0.158) Unemployment Rate 0.003 0.004 0.006 0.002 0.001 0.004 (0.005) (0.004) (0.004) (0.003) (0.003) (0.003) R2 0.22 0.43 0.41 0.52 0.66 0.67 County- specific trends None Linear Quadratic None Linear Quadratic + p<0.0167; * p<0.0083; ** p<0.00167 (p-values are adjusted for multiple testing using the Bonferroni adjustment for 6 tests) Outcome variable is in log scale of the per/100,000 population. All models include county fixed effects and a continuous time trend (including 2nd order term). Standard errors (in parentheses) are adjusted for clustering within counties. The county ordinance is based on the ordinance of the major city when that city has a population greater than 200,000 and a population greater than the unincorporated area. Each regression has 1,044 observations and covers the period from 1997-2014. 40 Table 3: Effect from Allowing Dispensaries, by Crime Type Homicide Robbery Assault Burglary Theft MV Theft (1) (2) (3) (4) (5) (6) Allows Dispensaries 0.249 0.299 -0.047 -0.016 -0.060* -0.098 (0.185) (0.247) (0.077) (0.026) (0.029) (0.075) Cultivation-No Limits -0.449 -0.064 -0.027 -0.033 -0.078+ -0.081 (0.330) (0.167) (0.063) (0.045) (0.041) (0.121) Alcohol Outlet Density -0.016 -0.001 -0.000 0.000 -0.000 -0.004 (0.013) (0.002) (0.002) (0.001) (0.001) (0.004) Post CJ Reforms 0.129 -0.090 -0.134* 0.002 0.104** 0.263 (0.292) (0.180) (0.054) (0.035) (0.031) (0.176) Population Density 6.447 8.287+ 1.681 -0.650 1.176* -2.146 (4.161) (4.306) (1.501) (0.829) (0.508) (3.986) Per Capita Income -0.263 -1.616 -0.045 0.016 -0.282 0.120 (3.372) (1.061) (0.298) (0.221) (0.258) (1.274) Unemployment Rate -0.001 -0.006 0.003 0.012** -0.002 0.001 (0.050) (0.020) (0.005) (0.004) (0.003) (0.020) R2 0.08 0.11 0.46 0.51 0.64 0.09 + p<0.1; * p<0.05; ** p<0.01 Outcome variable is in log scale of the per/100,000 population. All models include county-specific linear time trends, as well as fixed effects and a continuous time trend (including 2nd order term). Standard errors (in parentheses) are adjusted for clustering within counties. The county ordinance is based on the ordinance of the major city when that city has a population greater than 200,000 and a population greater than the unincorporated area. Years since adoption is calculated from January of the first year dispensaries were explicitly allowed. Each regression has 1,044 observations and covers the period from 1997-2014. 41 Table 4: Effect of Allowing Dispensaries on Marijuana-Related Arrests DUI Arrests Felony Marijuana Misdemeanor Marijuana (1) (2) (3) (4) (5) (6) (7) (8) (9) Allows Dispensaries 0.077* 0.091* 0.088** -0.022 -0.015 -0.049 0.087 0.215 0.125 (0.024) (0.026) (0.025) (0.092) (0.083) (0.084) (0.120) (0.106) (0.080) Cultivation-No Limits 0.034 0.070 0.085 -0.113 -0.135 -0.090 -0.383 -0.241 -0.095 (0.032) (0.039) (0.038) (0.106) (0.100) (0.092) (0.174) (0.133) (0.098) Alcohol Outlet Density 0.001 0.001 0.000 0.002 -0.000 0.001 0.003 -0.002 -0.003 (0.001) (0.001) (0.001) (0.004) (0.003) (0.002) (0.003) (0.002) (0.003) Post CJ Reforms -0.198** -0.195** -0.204** -0.337** -0.434** -0.429** -1.628** -1.666** -1.714** (0.030) (0.032) (0.033) (0.061) (0.089) (0.102) (0.163) (0.187) (0.161) Population Density -0.657* -1.715 -1.701** -0.064 -5.463 -1.413 2.343 -0.341 -2.521 (0.226) (0.817) (0.463) (0.622) (2.955) (1.560) (1.217) (1.704) (1.522) Per Capita Income -0.177 -0.151 0.042 1.075 2.611 2.608 -3.052 -2.206 -0.679 (0.146) (0.206) (0.231) (0.785) (1.570) (1.773) (2.784) (2.377) (1.599) Unemployment Rate -0.003 -0.003 -0.002 0.033 0.053* 0.048 -0.010 -0.001 0.003 (0.004) (0.005) (0.005) (0.014) (0.018) (0.019) (0.018) (0.019) (0.015) R2 0.47 0.58 0.57 0.03 0.16 0.15 0.55 0.66 0.68 County- specific trends None Linear Quadratic None Linear Quadratic None Linear Quadratic + p<0.0111; * p<0.0055; ** p<0.0011 (p-values are adjusted for multiple testing using the Bonferroni adjustment for 9 tests) Outcome variable is in log scale of the per/100,000 population. All models include county fixed effects and a continuous time trend (including 2nd order term). Standard errors (in parentheses) are adjusted for clustering within counties. The county ordinance is based on the ordinance of the major city when that city has a population greater than 200,000 and a population greater than the unincorporated area. Each regression has 1,044 observations and covers the period from 1997-2014. 42 Table 5: Sensitivity Checks Violent Offenses Property Offenses DUI Arrests (1) (2) (3) City Agencies -0.036 -0.042 0.160 (0.067) (0.026) (0.145) Unincorporated County -0.027 -0.068+ 0.056 (0.085) (0.031) (0.038) City Always -0.034 -0.054 0.090** (0.075) (0.026) (0.028) Partial Year -0.022 -0.052 0.099** (0.088) (0.031) (0.030) Removing San Francisco -0.008 -0.036 0.097** (0.074) (0.022) (0.027) Removing Santa Clara -0.033 -0.059 0.087* (0.081) (0.028) (0.030) Only Counties 5+ years -0.170+ -0.088 0.089 (0.069) (0.044) (0.041) + p<0.0333; * p<0.0067; ** p<0.0033 (p-values are adjusted for multiple testing using the Bonferroni adjustment for 3 tests) City Agencies refers to using city agency crime rates and laws. Unincorporated County refers to analysis using laws according to county, even when larger cities exist. City Always refers to analysis always choosing law applicable to city within county if available. Partial Year refers to using fractions for treatment variable, in case a law was passed after January. Removing San Francisco, Removing Alameda, and Removing Santa Clara refers to excluding each of these counties, individually, from the analysis. Only Counties 5+ Years refers to analysis with sample of non-adopters and counties that had law in place 5+ years. Outcome variable is in log scale of the per/100,000 population. All models include county fixed effects, a continuous time trend (including 2nd order term), and county- specific linear time trends. Standard errors (in parentheses) are adjusted for clustering within counties. Each regression (except City Agencies) has 1,044 observations and covers the period from 1997-2014 EXHIBIT 15