International approaches to drug law reform: research

3. Case Studies

3.1 Australia

Key Points

• The dominant strategy in Australia for many decades has been a policy of criminalising the use and supply of illicit drugs.
• In recent years, some reforms have occurred including:
  • Legal decriminalisation of cannabis possession in some States and Territories, which have replaced criminal sanctions civil fines.
  • In other states, other forms of de facto reform have taken place. These generally involve interventions such as diversionary cautions or referral to treatment. However, these schemes have strict eligibility criteria and can exclude those who may need them most.
  • In some places, police have supported harm minimisation interventions by, for example, limited unwarranted patrols near needle exchanges.
• Research has indicated that drug use in Australia is relatively stable over time and cannabis decriminalisation has not increased consumption. However, over time the number of civil penalties issued has continued to increase, suggesting that police propensity to detect and formally process cannabis possession may have increased due to the lower stakes of the civil penalty – an effect known as "net widening".

Background and model

Commonly used illicit drugs such as heroin, cannabis, opium and MDMA were legal in Australia in the first half of the 20th century. These drugs were gradually prohibited, but usage continued and in the 1960s, despite prohibition, drug use increased as the culture became more socially liberal (Gotsis, Angus, & Roth, 2016).

Drug policy in Australia is governed by a combination of federal laws and regulations, and the National Drug Strategy 2017–2026, which pursues "demand reduction, supply reduction and harm reduction". The dominant strategy for many decades has been a policy of criminalising the use and supply of illicit drugs. Further control is devolved to states and territories, which also pass their own laws and policies.

The Queensland Government's Inquiry into Imprisonment and Recidivism (2020) notes that there is some inconsistency between evidence on harm and the scheduling of drugs in legislation – noting that ecstasy, a relatively low harm drug, is listed in schedule 1, while relatively high harm drugs like fentanyls, are in schedule 2.

In recent years, some reform has occurred in drug law and enforcement.

• Legal reform
  • South Australia, the Australian Capital Territory and the Northern Territory have decriminalised use and possession of cannabis, replacing criminal sanction with a system of civil fines.
  • No state or territory has undertaken legal reforms for illicit substances other than cannabis. Accordingly, many people do still appear in court for possession of small quantities of drugs.
  • No state or territory has implemented criminal law reforms legalising supply-side-related offences.
  • Defacto reform
  • All jurisdictions that haven't legally decriminalised cannabis have undertaken some kind of de facto reform, including various forms of diversionary approaches such as cautions, provision of information, referral to treatment, compulsory treatment and compulsory education.
  • In most jurisdictions, these types of de facto reforms have also been extended to some other illicit drugs.
  • All de facto decriminalisation schemes have relatively strict eligibility requirements. Typically, these include to admit the offence, not have been detected by police more than once or twice, and carry only a particular quantity of drug (e.g. 2 grams or less). Anyone who does not meet the requirements (or has exceeded the low limits for past referrals) is processed through the usual court mechanism. Such eligibility requirements can exclude those most marginalised and/or those most in need of diversion into treatment and rehabilitation. De jure schemes have fewer eligibility restrictions which increases program access and equity.

De facto law enforcement reforms in relation to other harm minimisation approaches have also been implemented in some places. In 1988 the Commissioner of New South Wales Police discouraged police officers from making unwarranted patrols near needle and syringe exchange programmes, in order not to discourage drug users from attending. Other police forces subsequently introduced similar instructions (Monaghan & Bewley-Taylor, 2013).

De facto and de jure reforms in Australia are summarised in the table below:

Reform type and jurisdiction	Drugs	Scheme	Response/ penalty	Number of referrals allowed	Response to non-compliance
De jure reforms
ACT	Cannabis	Simple cannabis offence notice (SCON)	Fine	No limit	May result in criminal penalty
NT	Cannabis	Cannabis expiation scheme	Fine	No limit	Debt to state, may result in criminal prosecution
SA	Cannabis	Cannabis expiation notice (CEN)	Fine, but option to pay via community service	No limit	Reminder notice, additional fee, automatic criminal conviction
De facto reforms
ACT	All illicit drugs	Police Early Diversion Program	Caution plus brief intervention	2 pervious	May result in criminal penalty
NSW	Cannabis	Cannabis cautioning scheme	Caution plus information	1 previous	Recorded and court advised if subsequently reoffends
NT	Other illicits	NT Illicit Drug Pre-Court	Assessment and compulsory treatment	No limit	May result in criminal penalty
QLD	Cannabis	Police diversion program for minor offences	Assessment	1 previous	May result in criminal penalty
SA	Other illicits	SA Police Drug Diversion Initiative	Assessment and referral	No limit	May result in criminal penalty
TAS	All illicit drugs	Police diversion	Caution and brief intervention (for third offence, assessment and compulsory treatment)	3 previous in last 10 years	May result in criminal penalty
VIC	Cannabis	Cannabis cautioning program	Caution and education, optional referral	1 previous	Nil
WA	Cannabis	Cannabis intervention requirement	Assessment and compulsory education	1 previous	May result in criminal penalty
WA	Other illicits	All drug diversion	Assessment and compulsory treatment	1 only	May result in criminal penalty
Reproduced from Hughes et al, 2016

Hughes et al (2016) concluded that although population drug use rates remain stable in Australia, the rates of drug use/possession detections are continuing to rise. This means more people who use drugs risk getting charged, convicted and imprisoned for minor quantities of drugs.

Outcomes and impact

The research on Australia's overarching prohibitionist approach is consistent in concluding that prohibition contributes to harm rather than minimising it. For example:

• The Queensland Government's Inquiry into Imprisonment and Recidivism (2020) concluded that "criminalisation has created significant costs and unintended harms. It helped to create an illegal market worth at least $1.6 billion (with high levels of violence), made the quality of supply uncertain (resulting in increased morbidity and mortality), and impeded treatment of harmful use".
• Two studies have been done on the impact of police crackdowns in drug law enforcement.
  • Wan et al (2014) studied the impact of supplier arrests and seizures of heroin, cocaine and amphetamine-type substances over 10 years. They found no consistent effects between any of the supply reduction measures and police reports of theft, robbery and assault. The findings suggest that increases in seizures or supplier arrests indicate increased supply in the market making it easier to catch people, rather than signalling that a reduction should be expected. There was some evidence that 'very large-scale supply control operations do sometimes reduce the availability of illicit drugs for a period'.
  • Earlier, Maher and Dixon (1999) looked at the impact of ostensibly successful crackdowns generating "respectable" arrest and conviction rates. However, the authors conclude that "crackdowns, whether carried out in the name of law enforcement or quality of life, push markets in directions which are highly undesirable." They reported that much sale and consumption simply moved geographically, often to more dangerous locations, and that they adopted more dangerous transport methods, were less likely to carry clean injecting equipment, and increased other unsafe injection practices.
• Research comparing Australian jurisdictions with other countries has indicated that punitive approaches do not appear to impact consumption:
  • Martin et al (Martin, 2019) found that ecstasy, cocaine, methamphetamines and opioids are all significantly more expensive in Australia than in other countries, but conclude that this does not appear to been a deterrent, since drug use in Australia is relatively high. UNODC data suggest Australia has the second-highest prevalence of ecstasy use, the equal highest of cocaine, the fourth-highest of amphetamines and the fourth-highest of prescription opioids.
  • Degehardt et al. (2008) found that although the proportion of the Australian population imprisoned for drugs is substantially lower than in the US, "this does not appear to have deterred drug consumption—the United States consumes more illicit drugs per capita than Australia"

One possible explanation for this lack of deterrence is the low probability of getting caught. "About 90,000 drug offences are prosecuted each year in Australia while an estimated 2680 million drug offences are committed each year (2400 million cones, 40 million pills, 40 million hits, etc.) or about 1 prosecution per 30,000 offences; the cocaine rate is about one prosecution for every 110,000 lines of coke" (Jiggens, 2013).

Outcomes from cannabis expiation notice schemes seem to be mixed. Under this scheme, detected adult offenders can avoid prosecution by paying an expiation fee within a set timeframe. Penalties for single offences range from $50 to $150, but failure to pay expiation fees could lead to prosecution and the possibility of a conviction being recorded.

Evaluations of the scheme's impact found:

• The proportion of cannabis offences cleared by payment is relatively low, at below 50%. Christie and Ali (2009) report that " one of the most common outcomes for offenders who do not clear their CENs by paying the expiation fees is prosecution with a conviction being recorded. This has occurred in a system that, it could be argued, was designed to remove or reduce that risk of conviction. With many offenders not paying the fees, the absolute numbers of convictions being recorded for minor cannabis offences is probably greater than if the CEN system had not been introduced."
• Over time, the number of CENs issued has continued to increase, but research shows that cannabis use has remained relatively stable, indicating that police propensity to detect and/or formally process these crimes has increased.
• Notwithstanding the problems above, the expiation approach has been cost-effective, reducing enforcement costs without leading to increased cannabis use. Eastwood et al (2016) also reviewed the analytical literature on decriminalisation's impact on cannabis use in Australia. They found one study reporting a significant increase in cannabis use in states where it has been decriminalised, one study demonstrating a decrease in cannabis use after decriminalisation, and five studies showing that decriminalisation had no significant impact on the prevalence of cannabis use. They conclude that "Collectively, this would suggest that at the very least reform of the law and the ending of criminal sanctions for cannabis use has no or little impact on prevalence" (Eastwood, Fox, & Rosmarin, 2016)

3.2 Canada (cannabis legalisation)

Key Points

• Canada has allowed medically prescribed cannabis since 2001, and in 2018 the law was changed to create a fully legal, regulated market for recreational use.
• Retail cannabis is now available in all Canadian jurisdictions, with some states using government operated retailers and others using licensed private providers.
  
• This change was motivated by a desire to not just decriminalise cannabis consumers, but to also remove production and revenue from criminal enterprise, giving the government both control over quality assurance, and the ability to collect tax on this previously black market.
• It is too early to assess the change's impact with high confidence, however, the data available suggest that:
  • After a small initial rise, reported consumption returned to pre-reform rates and has been relatively stable since.
  • Canadians report accessing significantly less cannabis directly from the illegal market.

Background and model

In 2018 the law in Canada was changed to create a fully legal, regulated market for recreational use.

Canada's longstanding cannabis prohibition was first significantly altered in 2001, when medically prescribed cannabis was legalised. Various unsuccessful attempts were made in the following years at both liberalising and tightening the law, until the Cannabis Act was passed in 2018, allowing the creation of a legal, regulated market for recreational use.

The Canadian Government emphasised three key goals for this reform:

• protecting public health
• protecting young people
• reducing criminality associated with the illegal market

Reducing criminality associated with the illegal market has a number of motivations. The Final Report of the government's Task Force on Cannabis Legalization and Regulation (2016) notes concerns about the proceeds of the illegal cannabis trade enriching organized crime groups and funding other types of criminal activity, as well as the fact that criminalisation leaves potential tax revenue in the black economy rather than with the government; "With decriminalization [as opposed to legalisation] the production, distribution and sale of cannabis remain criminal activities. Thus, individuals remain subject to the potential dangers of untested cannabis. Criminal organizations continue to play the role of producer, distributor and seller, thereby increasing risk, particularly to vulnerable populations."

The key features of the new recreational model in Canada are:

• Provincial devolution:

Historically, Canadian provinces have taken their own independent approaches to alcohol regulation, and a similar devolved principle was applied to cannabis. There are national minimum requirements – individuals must be 18 or over, and possession cannot exceed of 30 grams – but provinces may place their own further restrictions on possession, cultivation, sale and use. Provinces are responsible for developing their own retail system, which may be through physical stores and/or federal mail.

• Staged implementation:

In the first year, the Act only allowed for herbal cannabis and oils to be sold, with edibles and concentrates subject to separate regulations, and not going on sale until the following year (Government of Canada, 2019)

• Promotion and packaging regulation:

Products may be branded, but the law prohibits certain activities such as event sponsorship or marketing aimed at young people.

Outcomes and impact

Because this law change only occurred in 2018, and was implemented in phases, it is too early to robustly assess impact, but there is no evidence that the police have faced an increase in issues.

The Transform Drug Policy Foundation note in their one year review of the law that "it will inevitably take time for the new system to bed in and for emerging problems to be addressed". Overall, they found that because cannabis regulation was devolved to provinces, "a 'patchwork' of regulatory models has now emerged across the country. Some provinces have been more successful than others and many have experienced 'teething' issues during early implementation", referring to issues with both preparedness and implementation speed (Transform Drug Policy Foundation, 2019). Media coverage has noted specific public criticisms of some local implementation programmes, but no formal evaluations are yet available.

The details of regulation vary by province/territory (Government of Canada, 2020).

• The legal age is set at 19 in all but two provinces.
• All provinces allow people to grow cannabis at home for recreational use, except Quebec and Manitoba. Where it is allowed, growth is consistently limited to four plants per household.
• Retail cannabis is now available in all jurisdictions – eight using government operated retailers, and the remaining five licensing private operators. Online sales are government operated in all jurisdictions except Saskatchewan and Manitoba.
• States are about evenly divided between those that allow use in public and those that don't. States that do generally have limitations in place similar to those that govern smoking tobacco in public.

Usage levels are a measure for considering public health implications, although the confluence of relevant factors is too complicated to robustly attribute causation to any single change. There is data available from the National Cannabis Survey, conducted by Statistics Canada from early 2018, although it must be treated with caution because:

• The law is too recent for trends to be clear yet, and we would expect to see a "novelty blip" at the point of change. The phased rollout further reduces comparability of quarterly results.
  
• People may have been less likely to disclose their use before law change.
  
• Prior to legalisation, some medical cannabis stores also sold recreationally illegally – this informal, quasi-legal grey market may have confused respondents in some areas.
  
• Sales recorded for medical purposes in July 2019 were lower than in July 2018, suggesting that at least some of those who previously relied on medical documentation were now accessing cannabis on the legal recreational market (Government of Canada, 2019).

With these caveats in mind, usage figures can nonetheless be informative. Previous national survey data showed a long history of high levels of cannabis consumption in Canada, and use has been rising for the preceding 4-5 years across almost all age groups (Statistics Canada , 2019). The two waves of data available for analysis since the law change:

• suggest use has remained relatively stable since the reform: "After a small rise in the first quarter of 2019, reported consumption went back down to pre-October 2018 levels in the second quarter.
• indicate a significant decrease in Canadians accessing cannabis directly from the illegal market or from friends and family.
• may give some evidence that those who already consumed cannabis prior to regulation are consuming more. Data from the first quarter of 2019 indicates that the number of occasional users did increase slightly compared to the same quarter in 2018. This, too, could potentially be linked to a novelty 'blip' and changes in survey honesty following the law change.
• recorded an increase in first-time users in the first quarter of 2019, with over half of these individuals aged 45 or older.
• Indicate ongoing challenges in encouraging consumers with an established supply source towards the legal market – reporting that the average cost of a gram of legal cannabis was $10.23 compared to $5.59 for illegal cannabis (Statistics Canada, 2019).

There has been no formal analysis yet on the impact the law change has had on other crimes or drug driving. Media coverage has generally indicated that police have not faced an uptick in problems.^[1]

In terms of cannabis specific offences, police-reported cannabis crime had been declining for several years due to being a relatively low police priority. Between 2017 and 2018, when the law change was made, cannabis related offences fell 29% overall (although import/export offences rose 22%). Import/export offences made up 21% of Cannabis Act offences in its first year, followed by possession of more than the maximum allowed (18%) (Statistics Canada, 2018).

3.3 Canada (Vancouver harm reduction)

Key Points

• In 2003, Vancouver opened North America's first supervised drug consumption facility. It initially operated under a special exemption from Canada's drug laws, and its legal status was affirmed by the Canadian Supreme Court in 2010.
• Vancouver police have adopted an organisation-wide policy that treats drug use as a public health issue and encourages police to use discretion and service referrals to achieve the best outcomes for people who use drugs.
• Since it began, there has never been a death at Vancouver's supervised injecting facility. Extensive evaluation has indicated that the facility decreases risk of fatal overdose, improves service users' safe injecting practices, increases uptake of addiction treatment, and reduces public nuisance issues.

Background and model

In 2003, Vancouver opened the first safe injecting facility in North America, called Insite, with a limited exemption from Canada's drug trafficking and possession laws to allow its operation. At the end of the three year pilot programme, the exemption was extended, but when the extension ended service users brought a court case which resulted in the British Columbia Court of Appeal ruling "While users do not use Insite directly to treat addiction, they receive services and assistance at Insite which reduce the risk of overdose that is a feature of their illness, they avoid risk of being infected or of infecting others by injection and they gain access to counselling and consultation that may lead to abstinence and rehabilitation. All of this is healthcare" (PHS Community Services Society v Canada (Attorney General), 2010). Ultimately, this was affirmed by the Supreme Court of Canada in 2010, who ruled that to "close Insite contravened the country's charter of rights by threatening the lives of injection drug users" (Small D. , 2012)

Vancouver now has an expanded harm minimisation approach across the city including multiple small scale injectable opioid programme sites, pilot drug checking service focusing on fentanyl in two sites with "legal authorizations for clients to be in possession of controlled substances and a clientele who regularly use opioids, stimulants, and other street drugs by injection" (Tupper, 2018) and projects to pilot automatic pill dispensing through secure atm-style machines intended to connect drug users easily to a clean, regulated supply (Fischer, 2018).

To support these programmes' success, the Vancouver Police Department developed and implemented an organisation-wide drug policy that frames drug use as a public health issue and promotes police practices that encourage people who inject drugs to access harm reduction services. (Landsberg, 2016). In particular, the policy encourages discretionary practices in street-level possession and use cases, and supports police referrals to safe injection facilities.

Outcomes and impact

Extensive evaluation has indicated that Vancouver's safe injecting facility decreases risk of fatal overdose, improves service users' safe injecting practices, increases uptake of addiction treatment, and reduces public nuisance issues. There is also evidence that relationships between drug users and police have improved, and that police now regularly refer people to the safe injecting facility.

A number of studies also shed light on the relationship between police and drug users in Vancouver:

• In 2001, prior to the reforms described above, a police exercise substantially increased police presence throughout the day in the vicinity of a Vancouver needle exchange. Researchers found that 27% fewer syringes were distributed in the first four weeks of the operation than in the four weeks prior (Wood, 2003). However, 9 years later, a qualitative study of policing practices in the neighbourhood where Insite is located during the 2010 Winter Olympic Games found that, despite higher police presence during this period, it did not reduce local drug users' access to health services or increase injection-related risk behaviour (Small W. , 2012).
• Police referral procedures have also played an important role. A 2008 study of Insite (De Beck, 2008), found that nearly 17 per cent of its clients reported being referred to the facility by police. In addition, the individuals the local police were referring to Insite were more likely to be engaged in sex work, to be frequent cocaine injectors, and to report discarding used syringes in public. By referring these particularly high-risk injection drug users to a health facility instead of a justice agency, local police help to reduce health harms.
• The authors also note that police referring injection drug users who are more likely to discard needles in public spaces to appropriate community services also serves to address public nuisance concerns.

Thanks in part to a major resource commitment when Insite was established to support robust data collection and research, the safe injecting facility has been extensively evaluated using cohort samples of drug users in Vancouver, administrative data and topic-specific surveys. Findings in this area can broadly be grouped into those related to the impact on overdoses, access to treatment, injecting practices, crime/nuisance, and implementation and targeting.

• Impact on overdoses
  • A number of studies have consistently shown that Insite has helped to prevent overdose fatalities. Although no overdose at Insite has ever resulted in death, nonfatal overdose is fairly common, with roughly 13 for every 10,000 injections (Milloy, 2008). In an 18-month study in 2004-5, 87% of overdoses were treated just with oxygen. The authors concluded overall that Insite plays a role in successfully managing overdoses among people who inject drugs, and it is also likely that Insite reduces the burden on emergency services that would otherwise respond to overdose events.
  • Another study estimated how many non-fatal overdoses at Insite would have been fatal if they had happened in the community. The author's models suggest that between 8 and 51 deaths were averted in a four year period. The number of overdose deaths in the area had averaged about 35 over the preceding five years. The authors concluded that the care provided in the facility reduces the risk of death and improves public health in the area (Kerr, 2006).
  • Some critics have suggested that supervised injecting facilities may actually lead to more overdoses, if the safe space makes people feel more comfortable with the risk of taking a higher dose. An evaluation surveying more than 1,000 Insite users over two years tested this theory, but found that those who were at Insite for the majority of their usage were not more likely to overdose than those who mostly used outside the service (Milloy, 2008).
  • Qualitative research with 50 service users also supports these findings- people who used Insite reported that it addresses many of the environmental factors that drive the high rate of overdose among people who inject drugs. In particular, giving people the time to carefully and safely inject without rushing is noted as a key way in which InSite reduces overdose risk (Kerr, Small, Moore, & Wood, 2007).
• Impact on access to treatment
  • The Canadian cohort study found that attendance at Insite was associated with increased referral to addiction care centres, uptake of detox treatment and uptake of methadone maintenance (Wood E. , 2006). Contact with Insite's addictions counsellor further significantly increased a person's chances of enrolling in detox.
  • In the year after Insite opened, there was a 33% increase in detox service use, compared to the year prior. Moreover, Insite clients who entered detox were 1.6 times more likely to enroll in methadone treatment and 3.7 times more likely to enroll in other forms of addiction treatment. People who entered detox also visited Insite less frequently in the month after enrolling in detox services than in the month prior to enrolment (Wood, Tyndall, Zhang, Montaner, & Kerr, 2007).
  • Criticism that safe injecting facilities may make people in recovery more likely to relapse was also tested in a study. However, no change was found in rates of relapse into injection drug use among former users in the year after Insite opened compared to the year before, and fewer people started binge drug use. The authors conclude that the facility's benefits have not been offset by negative effects on drug use patterns in the area (Kerr T. , 2006)
  • People's access to other medical treatment also improved. A qualitative study of 50 service users found that nurses at InSite regularly provide care for injection-related infections and frequently connect drug users with off-site medical treatment by supplying referrals and arranging transportation. Service users reported that it was easier to get care for infections there than in conventional care settings because Insite nurses are experienced in working with drug users and because the facility is open late at night. Additionally, the authors note that if scaled this type of care has the potential to reduce emergency room use and hospitalization among local injection drug users (Small W. , 2008).
    

Impact on injecting practices

A large number of studies have looked at Insite's impact on injecting practices. They have consistently found:

• Reduced needle sharing:

After adjustment for relevant socio-demographic and drug use characteristics, use of the facility was associated with a marked reduction in syringe sharing (Kerr T. , 2005) These findings are supported by earlier results suggesting that attendance at Insite appears to reduce syringe sharing specifically in HIV risk situations (HIV+ people lending needles or HIV- people borrowing them), and analysis of Insite data combined with a similar study in Spain concluded that regular safe injection facility users reduce their likelihood of sharing syringes by 69% (Molloy, 2009).

• Improved knowledge and practice of other safety measures:

The education provided at Insite has also been found to positively impact people's use of safe injecting practices. A study comparing those who consistently visited Insite (for 25% or more of their injections) and those who used the facility less consistently found that consistent Insite attenders were substantially more likely to use sterile water, swab injection sites, dispose of syringes safely, and cook or filter their drugs. They were also less likely to rush during injections or share syringes. The authors conclude that InSite is helping to reduce some of the health risks associated with unsafe injecting (Stolts, 2007).

A 2006 survey, (Petrar, 2007) found that 75% of Insite Users reported injecting more safely as a result of visiting Insite. Qualitative research with 50 InSite clients supports these findings (Fast, 2008), with many people reporting that they did not know about the benefits of cleaning the skin with an alcohol swab prior to injecting, inserting the syringe bevel-side up, or other measures they could be taking to minimize health risks. Regular Insite users reported that they learned about these practices from nurses in a way that felt safe, supportive and unhurried. Importantly, study participants told researchers that the overall environment at the facility encouraged them to adopt the safer practices and to make a habit of using them both within and outside of the facility.

A small number of research participants did report that they had not received safer injecting education at the facility, indicating that Insite may not be meeting the educational needs of everyone who injects drugs there. Those people who use the facility often are more likely to interact regularly with nurses and receive educational messages that help protect health.

• Reduced public injecting:

A study of those Insite users who also continue to inject publicly was published in 2007. The results showed that those still injecting publicly were 3 times more likely to be homeless compared to other Insite users, and also 1.6 times more likely to require help injecting (a practice that is not permitted at Insite). People who continue to inject in public told researchers that the waiting times limit their use of the facility. The authors conclude that increasing availability through a program expansion may further help to reduce persistent risk behaviours and address community concerns about ongoing levels of public drug use, and that the restriction against assisting with injections at Insite may pose a barrier to use by some people who inject drugs (McKnight, et al., 2007).

Fairburn et al also note in their qualitative research that in many cases, the first time a woman uses an injection drug, someone else— usually an older male— injects the drug for her. By learning how to inject themselves, women rely less on men and gain more control over the circumstances of their own drug use. They are then more likely to practise safer habits when injecting, such as using clean needles (Fairburn, Small, Shannon, Wood, & Kerr, 2008).

• Impact on crime, nuisance and the community
  • A rigorous quasi-experimental study found that the number of people injecting in public, number of publicly discarded syringes, and amount of injection related litter all declined significantly in the 12 weeks after Insite opened, compared to the weeks before (Wood, et al., 2004).
  • More broadly, crime rates in the year before Insite opened were compared with crime rates in the year after, using Police data on drug trafficking, assaults, robberies, vehicle break-ins, and thefts. Researchers found no statistically significant changes in rates of drug trafficking, assaults, or robberies, and a drop in vehicle break-ins and thefts. The results of this study provide evidence that Insite's presence has not contributed to an increase in drug-related crime in surrounding neighbourhoods.
• Implementation and service user targeting
  • Due to its extensive evaluation, much can be learned about implementation and service user targeting from the Insite model.
  • Multiple cohort and survey samples have shown that InSite is attracting those at highest risk, as is intended (Wood, et al., 2005) (Wood, et al., 2006). The researchers concluded that the facility was successfully attracting:
    • people who are at elevated risk of blood-borne disease or infection
    • people at elevated risk of overdose
    • people who were using publicly or unsafely disposing of syringes
    • Marginalised people including homeless people, sex workers, younger people and indigenous people.
  • A survey of over 1000 service users also identifies some of the most common obstacles to using the service. These were most frequently travel time to Insite, the facility's limited operating hours, and the waiting time to use the facility (Petrar, 2007).

3.4 Denmark

Key Points

• Since a law change in 2011, Denmark now has five Drug Consumption Rooms across three municipalities.
• Overdose deaths in Denmark have been falling since consumption rooms were introduced. Evaluations indicate that DCR's have reduced the number of overdose deaths in the cities that have them.
• There is also evidence from evaluations of the Copenhagen facility that crime, violence and publicly discarded syringes have all decreased in the area.
• Research has highlighted the important role consumption rooms have played in helping people access health, social and addiction services.
• Research on police attitudes also suggests that the advent of decriminalization zones around consumption rooms has caused more police to view drug users as people in need of police protection rather than as police targets.

Background and model

Denmark has pursued harm-reduction policies such as opioid substitution treatment (OST) and needle exchange programmes for many years, but politicians and health authorities initially opposed drug consumption rooms (Ministry of the Interior and Health, 2003) (Ministry of the Interior and Health, 2010). However, in 2011, in an act of civil disobedience, Danish NGOs started two mobile drug consumption rooms in retired ambulances, staffed by volunteer nurses and doctors (Ege, 2012), and the next year Denmark passed legislation to allow municipalities to establish drug consumption rooms.

The amendment to Denmark's law on psychedelic substances had three stated intentions:

• to reduce the number of overdose deaths,
• to improve life situations for people who use drugs by connecting them to the healthcare system, treatment facilities and social services,
• to reduce the nuisance of public drug taking to surrounding neighbourhoods (Ministry of Health and Prevention, 2011).

The bill amending the law provides that "within the vicinity of the DCRs, police will not prosecute people who are over the age of 18 years, who possess drugs for their own use, and who have a prolonged use of and addiction to illicit drugs" This can be seen as a form of de facto decriminalisation of drug use, since possession remains an offence, but in practice the law is not fully enforced.

The key features of the Danish DCR model are summarised here from Keppel et al (2019):

• Consumption rooms follow two delivery models:
  • Integrated units, typically part of a shelter with additional services such as counselling, laundry and shower facilities, or a health clinic.
  • Mobile unit, with relatively limited space, primarily only functioning as a hygienic, safe place for injecting.
• DCRs are financed by the municipalities and managed by NGOs^[2].
• All DCRs are staffed with registered nurses or nursing aides, who work in a team alongside social workers and social educators. All staff members have advanced first aid training and are trained in the effects and side effects of the most commonly consumed drugs, but no additional formal training is required to work at a DCR. The healthcare professionals are mainly responsible for intervention and treatment in cases of severe intoxication.
• To access DCRs, clients must:
  • register, create an alias that will be used for future entry, and agree to house rules.
  • accept that staff will intervene if they overdose
  • indicate at every entry what drug they plan to consume
  • not be a minor or pregnant
  • not deal or trade within the facility
  • no client is allowed to provide assistance to peers

In addition to providing safe, clean space for consumption, DCRs also provide information on drug ingredients and potency, refer clients to a variety of treatment and health services, and may also issue clients with Naloxone. The staff address barriers to care for injection-related injuries, provide low-threshold nursing attention on site, and connect with off-site medical service (Small, Wood, Lloyd-Smith, Tyndall, & Kerr, 2008).

Outcomes and impact

There are now five DCRs, spread across three municipalities in Denmark, all of them in permanent locations with integrated services except for one mobile unit in Copenhagen. By 2016, within five years of the law change, there were 3600 service users registered at DCRs (Toth, Tegner, Lauridsen, & Kappel, 2016).

The Danish ministerial evaluation of DCRs is not available in English, but Kappel et al (2016) report that it indicates "DCRs in Denmark has had an effect on reducing the number of deaths by overdose in the cities that have implemented the DCR programme". Although we cannot confidently attribute causation, it is noteworthy that Denmark's overdose deaths peaked in 2011, the year the DCRs commenced, and have been declining since. Other impacts have been reported in the media – news articles (such as Boffey, 2013 and Busby, 2018) on the Copenhagen DCRs have reported that:

• Year on year, burglaries in the wider area are down by about 3%, theft from vehicles and violence down about 5%, and possession of weapons also down.
• Up to 10,000 syringes used to be picked up off the streets of Vesterbro every week before the room opened. Since the launch of the room, the quantity of drug paraphernalia collected from gutters, playgrounds, stairwells and doorways in the area has halved.

Their research concluded that the "humanizing approach of DCR staff, combined with the provision of facilities and tools for drug consumption, appear to promote a sentiment of social acceptance among DCR clients" and noted that this was consistent with other studies of DCRs (eg (Rance & Fraser, 2011). They further found that the approach empowered the DCR clients towards feeling more "like citizens rather than scummy junkies", and concluded that this empowerment is in fact "the most important feature of DCRs… which ultimately paves the way for both the successful prevention of overdoses, as well as for steering clients towards utilizing both social and health services, including drug rehabilitation facilities."

Fieldwork for Kappel et al's study also observed that the trust forged between the staff and clients was evidenced in their willingness to heed staff's advice on being cautious with their intake, and attribute this as a success of the service. They also note that DCRs appear to function as bridges to other health and social services, including drug treatment, but that it is difficult to measure impact because people use aliases to access the DCR, and are then need to follow up the referral themselves.

The model is also reported to have had an impact on policing practices. According to qualitative research by Kammersgaard (2019) "decriminalization zones that were established with the DCRs in 2012 enabled the police to abstain from traditional drug law enforcement techniques that focus on use reduction through punishment and the confiscation of illegal substances. Instead the police were able to direct their attention to the wider harm experienced by people who use drugs". Kammersgaard further suggests that this policy change led to police viewing people who use drugs primarily as people in need of police protection rather than as offenders who pose a threat to their community.

3.5 Netherlands (cannabis 'coffee shop' market)

Key Points

• Cannabis has been decriminalised for personal use since 1971. The current day 'coffee shop' model has developed through a dialogue between informal police tolerance policies and legislative/regulatory reform.
  
• Today, coffee shops may sell cannabis as long as they are licensed and adhere to a range of regulations, including limits on the volume that can be traded, a minimum age of 18, and not selling alcohol or contributing to public nuisance.
  
• The consumer side of this model has been largely successful, and cannabis use in the Netherlands is about average for Europe despite its laws being significantly more liberal than most.
  
• However, because the law only permits and regulates small scale, consumer transactions, the cultivation and wholesale supply to coffee shops remains unregulated and criminal enterprise is significantly involved.

Background and model

In the early 1970s, two government committees on drug law (Hulsman, 1971 and Baan, 1972) recommended attempting to separate cannabis from the wider drug scene, to reduce the extent to which young people who experimented with cannabis were exposed to more harmful substances.

Since 1971, cannabis has been decriminalised for personal use, and the Ministry of Health has had primary responsibility for drug policy. A further law change in 1977 sanctioned tolerant policing practices that had emerged as informal policy in the previous decade, giving authorities the ability to creatively interpret the new legislation in their local areas. Public prosecutors subsequently deprioritised small scale cannabis offences, focusing almost exclusively on large operations and harder drugs (De Kort, 1995).

'Coffee shops', where people could buy cannabis more or less free from the risk police enforcement, emerged as an unintended consequence of this change, and the Government responded with restraint, applying a pre-existing but informal set of tolerance criteria: that coffee shops must only sell small quantities, and not serve young people, advertise or cause nuisance. However, due to this unclear legal status, enforcement was inconsistent, with some places raided while others were tolerated (De Kort, 1995).

The number of coffee shops continued to grow and eventually in 1991 formal regulation was brought in which:

• Set a minimum age of 18
• Decreased the maximum transaction amount to 5 grams per person per day
• Limited daily trade to 500g

These regulations have been strengthened over time, for example tightening restrictions on premises licensees, and giving local municipalities the power to close down coffee shop's deemed to cause a public nuisance.

In 2012, the government piloted a new approach, known as the 'weed club pass'. This required coffee shops to operate on a small membership-only model, and only residents of the Netherlands could become members (Aanwijzing Opiumwet, 2012) (although, as discussed below, this was not successful and the pilot has not been rolled out more widely).

Outcomes and impact

There is no evidence that the relatively free availability of cannabis has led to substantial increases in consumption. Despite having significantly more liberal laws than its neighbours, cannabis use in the Netherlands is on par with the European average (Grund & Breeksema, 2017).

However, while the consumer side of the coffee shop model has been largely successful, important lessons can be drawn from the Netherlands' experience of what is termed the "back door problem". The regulated coffee shop model means sales of small quantities (ie, retail to consumers – the 'front door') is exempt from prosecution, but cultivating larger quantities for supply (ie, wholesale supply to coffee shops – the 'back door') remains illegal (De Kort, 1995). This came about, according to former Prime Minister Van Agt, because 'there was no parliamentary majority for decriminalising the supply of cannabis or other drugs in 1976, and the Dutch government did not want to risk diplomatic or economic problems with neighbouring countries and the international community'.

The compromise reached gave consumers safe access in regulated environments, certainly lowering harms to them, but the vast majority of research agrees that the fact cultivation and supply to coffee shops is not regulated is the main source of the negative side effects currently seen in the Netherlands, particularly the involvement of organised crime and the lack of quality control. Investment in criminal enforcement has not been effective at removing these problems. At the same time, the Dutch government has felt limited in their ability to regulate the back door because of the potential diplomatic and economic consequences (Grund & Breeksema, 2017).

Through the 1990s and early 2000s periodic police crackdowns on suppliers and home growers had the unanticipated consequence of significantly increasing criminal organisations' involvement in cannabis cultivation (Belackova, Maalste, Zabransky, & Grund, 2015). This problem has arguably been exacerbated by the fact that the Ministry of Health has primary responsibility for drug policy, and do not see this type of commercial regulation or criminal law enforcement as their role.

The 2012 pilot in southern provinces, of clubs restricted to small memberships of only Dutch residents, was not found to be very successful. Much of the evaluation literature is not available in English, but Grund and Breeksema summarise it: "Nuisance from street drug sales and feelings of lack of security increased (SSC Onderzoek en Informatie, 2012). Local consumers refused to register at coffee shops, citing privacy concerns (Wouters and Korf, 2011; Nijkampand Bieleman, 2012) and many regulars abandoned the shops, resorting to illegal markets instead (Maalsté and Hebben, 2012)." Consequently, soon after the pilot was introduced, the requirement that shops be limited-membership only was abolished, while the requirement to only serve Dutch residents was devolved to local municipalities, many of which do not enforce it (Opstelten, 2012).

3.6 Netherlands (harm minimisation)

Key Points

• The Netherlands has developed a comprehensive health-based harm minimisation approach, with many of their most notable policy developments arising from informal or experimental practices that were subsequently codified by the government. Needle exchanges, safe injecting facilities and heroin assisted treatment are all examples of this.
  
• The Netherlands has very low rates of problem drug use, and arrests for minor possession are extremely rare.
• Drug users in the Netherlands also tend to use safer practices (for example, a very low proportion of opiate users inject), and this has led to relatively high survival rates and longer life expectancy for people who use heroin in the Netherlands.
• Due to substantial investment in sheltered housing, integrated drug treatment, public mental health care, services for the homeless and criminal justice interventions, most problematic drug users now live in supported housing where they receive welfare and treatment.

Background and model

The Netherlands tried unsuccessfully to dismantle the organised crime groups importing heroin in the 1970s and 80s, but its prevalence continued to grow. As street level policing crack downs drove heroin dealing indoors, many local municipalities supported tolerating these so-called "street addresses" as a way of keeping drug-related nuisance off the streets.

In some cities, the police, drug services, urban planners, and neighbourhood organisations actively collaborated with dealers and proprietors of street addresses, with an unwritten understanding that they would be tolerated as long as they sold only consumer amounts and did not advertise, cause nuisance or serve young people. This lasted until a crackdown in the 1990s closed street addresses again, causing the heroin and crack cocaine market and users to return to the streets (Grund & Breeksema, 2017).

The Netherlands has developed a comprehensive health-based harm minimisation approach, with the most notable policy developments coming out of informal or experimental practices that were subsequently codified by the government. For example:

• Needle distribution began informally in the 1970s, and continued to grow through grass roots organising, until official needle and syringe exchange programmes were introduced (Blok, 2008) (Grund, et al., 1992). By 2012 there were 150 needle exchange programmes in the Netherlands. (EMCDDA, 2012).
• As early as 1974, off-the-record drug injection rooms were available in two Amsterdam drop-in centres, and one in a church basement in Rotterdam began in 1982 (Blok, 2008). These activist projects laid the groundwork for the official safe consumption facilities that opened in most large Dutch cities 1995. By 2012 there were 37 drug consumption rooms across the country, targeting injectors and smokers (Schatz & Nougier, 2012)(EMCDDA, 2012).
• Heroin Assisted Treatment was introduced in 1996 as a scientific trial and, after a favourable evaluation, was registered as a legal medication for 'chronic, treatment-resistant heroin-dependent patients' in 2006 (Blanken, et al., 2010) (Fischer, et al., 2007).

An official agreement with the public prosecution service ensures that anyone possessing illicit drugs at an official drug-testing service will not be arrested or prosecuted (EMCDDA, 2017)

Outcomes and impact

The largely tolerant, harm-minimisation focussed approach to drugs in the Netherlands has not given rise to high prevalence.

• The Netherlands has the lowest level of problem drug use in the EU (Van Laar & Van Ooyen-Houben, 2009)
• The overall prevalence of drug use in the general population is below the EU and USA averages(ECMDDA, 2012; Van Laar et al., 2014).

Arrests and criminal records for use or minor possession are extremely rare in the Netherlands compared to other European states (3 per 1,000 users, compared to 44 per 1,000 users in Austria) (Room, 2008).

The research also indicates a number of other ways in which Dutch drug policy has contributed to reducing harms. Notably, the Netherlands has one of the lowest rates of injecting amongst opiate users – only about 7% of all people in treatment for heroin dependence were or are injecting users. This has served to reduce Dutch opiate and crack cocaine users' risks of both overdose and HIV, which has contributed to relatively high survival rates of people who use heroin (Grund & Breeksema, 2017). Treatment for heroin dependence increasingly includes elements of geriatric care, and the EMCDDA recently complimented the Netherlands for pioneering senior citizens homes for the aging group of heroin consumers (EMCDDA, 2015).

The Netherlands also provides important lessons in the unintended consequences of intensive law enforcement. Research suggests that both of the major policing crackdowns attempted in the last 50 years led to increasing harms:

• Schreuder and Broex (1998) found that policing crackdowns in the 70s did not stop the escalation in heroin. Instead, they argue that along with other environmental trends, stringent policing contributed to pushing street scenes out of the city centre, diffusing heroin into more working-class neighbourhoods.
• A large body of research suggests that as street addresses were closed by a crackdown in the 90s, drug users were pushed into a much less favourable risk environment. Without the protective environment and social control of the street addresses, people smoking cocaine in the streets became increasingly prone to its negative side effects, and the street market scene became larger, more volatile, and more harmful to the people in it (Blanken, Barendregt, & Zuidmulder, 1999).

Finally, the success of the Dutch approach has relied on government investing significant resources in a comprehensive and integrated harm reduction, treatment and social support system targeting people with drug problems, the homeless and chronic psychiatric patients (Grund & Breeksema, Drug Policy in the Netherlands, 2017). Since 2006, coordinated investments have been made in:

• sheltered housing
• integrated drug treatment
• public mental health care
• services for the homeless, and
• criminal justice interventions

Consequently, most problematic street drug users now live in sheltered or supported housing where they receive welfare, medical care and tailored drug treatment, or consume their drugs in on-site drug consumption rooms (Schatz, Schiffer, & Kools, 2010). Those who continue to cause nuisance or engage in crime are subjected to various criminal justice interventions, including compulsory treatment and other forensic psychiatric interventions (Van Laar, 2015)

3.7 Portugal

Key Points

• In 2001, following expert medical and social strategic advice, Portugal decriminalised the purchase, possession and use of all illicit drugs.
• The change went beyond depenalisation, which removes custodial sentencing, but did not amount to full legalisation, as the production, manufacture and large-scale distribution of illicit drugs remain a criminal offence.
• Studies have generally found that the change did not significantly increase consumption or drug prices. Moreover, analysis estimates that decriminalisation has decreased the social costs of drug use by 18%.
• Initially, decriminalisation had a clear direct impact on prison populations in Portugal, with fewer people incarcerated and fewer incarcerated people reporting drug use. Incarceration rates began to rise again in 2008, although the profile of incarcerated offenders is now different and it is likely other external trends contributed to this.
• There is some evidence of "net-widening" leading to increasingly low level offenders being brought before commissions. This is an important implementation lesson in ensuring laws are explicitly designed to prevent mission creep.

Background and model

At the end of the 1990s, Portugal was noted for high rates of problematic drug use and heroin market. This led to significant increases in infectious diseases, particularly HIV and AIDS, and drug related deaths with a peak of 369 in 1999 (Hughes & Stevens, 2010) (Van Het Loo & Van Beusekom, 2002).

In 1998 the Commission for a National Anti-Drug Strategy brought together legal, medical, and social professionals and recommended that decriminalisation of drug use and possession would allow the government to focus on prevention, harm reduction, treatment, and helping people to maintain their social connections (Domoslawski, 2011). Law enforcement and health experts viewed criminalisation as part of the problem rather than the solution because people with addictions were deterred from seeking treatment by the stigma of criminalisation (Hughes and Stevens, 2010).

In 2001, Portugal decriminalised the purchase, possession and use of all illicit drugs. The change went beyond depenalisation, which removes custodial sentencing as an option for low-level drug offenders, but did not amount to legalisation, as the production, manufacture and large-scale distribution of illicit drugs remain a criminal offence.

• Low-level offenders are now dealt with administratively by an informal 'Dissuasion Commission', which determines an appropriate non-custodial sanction. The Commission's work regionally, and their purpose is not to punish the offender but to encourage treatment and rehabilitation.
• The commission determine whether the person suffers from addiction or not, and then choose from a range of consequences.
• If the person suffers from addiction, commission is not able to mandate treatment but can suspend a penalty on the condition that an offender agrees to get treatment.
• Commissions may also impose fines or restrictions (from one month to three years) on people, depending on a range of factors including the type of drug, whether it was used in public, and how often the person uses.
• The Commissions come under the Ministry of Health rather than Ministry of Justice and include a treatment professional, social worker and lawyer, and are supported by a range of agencies such as treatment, health, employment, child protection, social services and schools.

Investments in drug treatment, harm reduction and social reintegration were also expanded at the same time. Crucially, the Government moved away from providing harm reduction interventions on a small scale based on local risk reduction and short-term aid, to a systematic approach across the whole country (Leite, 2019).

Outcomes and impact

In 2017, 11,329 people were involved in commissions, a new high watermark and an 80% increase over the period of a decade. In 2017, around 10% of those before commissions were classed as addicts (SICAD, 2019). For those classed as addicts, the most common ruling is a suspension to allow the perpetrator to undertake treatment. For those not considered addicted, the majority receive a provisional suspension. Between 2007 and 2012, only 4-6% appeared before a commission again, and 88% of those only registered one relapse in a year (SICAD, 2013).

Studies have generally found little impact on usage or drug prices:

• Reviews of the Portuguese experience of decriminalising all drugs have not found evidence of an increase in drug use (Greenwald, 2009) (Hughes & Stevens, 2010) Depending on the study examined, there was either no increase or only a very small increase in adult drug usage.
• In line with this, Felix and Portugal (2017) found that prices for cocaine and opiates did not decrease following decriminalization. They conclude that "drug decriminalization seems to have caused no harm through lower illicit drugs prices, which would lead to higher drug usage and dependence. This evidence contrasts with the commonly held belief that drug decriminalization would necessarily lead to a dramatic increase in usage rates.

Studies have indicated a number of benefits in terms of drug harms and costs:

• An ex-post analysis of Portugal's decriminalisation of all drugs estimated that the total social cost of drug use decreased 18 per cent (Goncalves, Lourenco, & Nogueira da Silva, 2015). This was mainly driven by reductions in indirect health costs (29 per cent), non-health related indirect costs (24 per cent) and non-health related direct costs including the criminal justice system (17 per cent). There was an increase in direct health costs of 9 per cent associated with providing prevention, treatment and harm reduction services.
• Greenwald (2009) and Hughes & Stevens (2010) drew similar conclusions. "The reforms did lead to a reduction in drug-related harms (both problematic drug use and youth drug use declined), and criminal justice system costs: While small increases in drug use were reported by Portuguese adults, the regional context of this trend suggests that they were not produced solely by the 2001 decriminalization. We would argue that they are less important than the major reductions seen in opiate-related deaths and infections, as well as reductions in young people's drug use. The Portuguese evidence suggests that combining the removal of criminal penalties with the use of alternative therapeutic responses to dependent drug users offers several advantages. It can reduce the burden of drug law enforcement on the criminal justice system, while also reducing problematic drug use." (Hughes & Stevens 2010, p.1018)
• An economic analysis of the effects concluded that the Portuguese approach was not harmful and contributed to a reduction in seizures of heroin and cocaine, number of drug offences, drug-related deaths and the incidence of HIV among drug addicts (Felix, Portugal, & Tavares, 2017).
• Opioid substitution treatments and injecting rooms have been tried in many jurisdictions and these approaches have been associated with reduced overdose hospitalisations and deaths, public injecting and discarding of needles (EMCDDA 2018).

Initially at least, it was possible to discern a direct impact of decriminalisation on incarceration rates in Portugal. Between 2000 and 2008 there was a fall in the imprisonment rate from 126 per 100,000 to 102, and the number of offenders who reported using heroin before entering prison also fell from 27 per cent in 2001 to 13 per cent in 2007 (Hughes and Stevens, 2010). Since 2008, however, there has been a notable rise in imprisonment again, and rates are now comparable to those before the reform. Again, the extant English language literature does not seem to offer an explanation for this trend, although the Council of Europe report on European prisons 2005-2015 indicates that during that period, at least, the proportion of prisoners sentenced for assault and battery as well as for sexual offences increased, while the proportion serving sentences for homicide, robbery, theft and drug offences decreased. It may therefore be that other trends in some combination of crime, detection, enforcement and sentencing have contributed to re-filling Portugal's prisons with a different profile of offender.

The Portuguese experience offers a number of important implementation lessons:

• Laqueur (2014) highlights the way in which Commissions are increasingly being required to deal with younger, non-addicted people for cannabis-related offences, which amounted to 76% of cases in 2009, and notes that such cases create a strain on scarce resources. This gradual increase in young, non-addicted people coming before commissions for small amounts of cannabis appears to be a similar phenomenon to the "net widening" effect seen in Australian cannabis expiation schemes, and underscores the need to consider ways to prevent this mission-creep in law enforcement, such as explicitly excluding certain substances or groups from enforcement measures, or by fully legalising and regulating trade.
• Strain on resources has also been identified by others as a factor impeding the work of the Commissions (Pinto, 2010) . An evaluation of the 2008-2013 action plan identified that a lack of quorum was responsible for delays in the system (SICAD, 2013; Laqueur, 2014). Pinto also highlighted that people being referred for treatment were often already undergoing treatment.