Is arrest for prohibited drug use a prelude to more serious offending?

Abstract

Although they constitute a significant fraction of the workload of most courts, very little research has been conducted on the criminal careers of those who commit minor offences. Such research is important for both theoretical and practical reasons. It is of theoretical importance because the criminal careers of those who commit minor offences may differ significantly from those who commit serious offences. It is of practical importance because the assumed rate of re-offending among minor offenders has a bearing on both the sentence imposed and the question of whether some offences should be decriminalised. Use or possession of a prohibited drug is a common minor offence – and one that many have argued should be decriminalised. Little is known, however, about the criminal careers of those convicted of this offence. We do not know what proportion are rearrested, what further offences (if any) they commit, or what factors affect the rate of offending among those who do have further contact with the criminal justice system. To answer these questions, we examined a cohort of 13,953 people whose first proven offence was for the use or possession of a prohibited drug and examined their criminal careers over an average period of 4.4 years (sd. = 3.4 years, range = 20.8 years). The majority (73%) had no further contact with the NSW criminal justice system. The most common offence among those who did re-offend was another drug possession offence. Significantly higher risks of re-offending were found among those living in areas in the lowest quartile of disadvantage and among those found in possession of cannabis. The implications of these findings are discussed.

Keywords

Cure fraction model drug possession re-offending drug type socioeconomic status

Introduction

Criminological theorists often assume that criminal behaviour arises out of some adverse psychological, social, or economic processes or conditions. This assumption may be valid for serious offences such as theft, robbery or serious sexual or physical violence, but it is unlikely to be true for all offences. An offender is simply a person who has breached the criminal law. Some criminal laws enjoy near-universal public support (e.g. the prohibitions against murder, robbery, theft, and sexual assault). Others are highly contested and/or breached by large segments of the general public. Possessing small amounts of cannabis is a criminal offence in most Australian States and Territories; however, more than 4 in 10 Australians oppose this law (Australian Institute of Health and Welfare, 2020). Driving under the influence of alcohol is a criminal offence, yet more than 50% of New South Wales (NSW) drivers report having driven while under the influence of alcohol at some stage of their lifetime (Centre for Road Safety, 2020). The patterns of offending and recidivism among those who commit widely contested, very common and/or relatively minor offences (e.g. illicit drug use, trespassing, offensive language) may be quite different from those who commit unusual or serious offences.

The theoretical preoccupation with serious crime has led to a similar focus in criminal career research. Since the publication of Criminal Careers and Career Criminals (Blumstein et al., 1986) an enormous body of research has accumulated on the correlates and causes of re-offending. Investigators have examined the correlates of re-offending for sex offenders, violent offenders, offenders who are mentally disordered, offenders with ADHD, and arsonists, to name a just few of the domains occupied by criminal career research. Researchers have endeavoured to determine the age of onset of offending, the peak age of offending, the average duration of a criminal career, the frequency of offending and whether offences are generalists or specialists. Nearly all of this research, however, has been focused on serious offences. Very little has been conducted on the criminal careers of those who commit minor offences, though they constitute a significant fraction of the workload of most courts. In Australia, for example, traffic offences, public order offences and drug possession offences account for 47% of the workload of the magistrates’ courts (Australian Bureau of Statistics, 2022).

Research into the criminal careers of minor offenders is important for practical as well as theoretical reasons. Courts making decisions on bail and sentence frequently have regard to the likelihood of the defendant/offender committing further offences. At present, defence counsel cannot call on any reliable statistical evidence bearing on the likelihood or seriousness of reoffending among those convicted of minor offences. Policy makers also have a strong interest in the criminal careers of those who commit minor offences. Criminal prosecution has some very undesirable consequences. They include social stigmatization, restrictions on travel overseas (Council of Europe, 2017) and damage to an individual's future earnings and employment prospects (Borland & Hunter, 2000; Graffam et al., 2004; Grogger, 1995; Nagin & Waldfogel, 1995; Newton et al., 2018; Waldfogel, 1994). Criminal prosecution is also expensive. The costs include the expenditure on policing, prosecution, sentencing and court administration. Some have argued that the costs of criminal prosecution outweigh the benefits for offences such as possession of a prohibited drug (e.g. Howard, 2020). The case for non-prosecution has recently been strengthened with the emergence of evidence that not prosecuting those who commit minor offences, such as illicit drug use, can actually reduce the risk of re-offending (Agan et al., 2021; Mueller-Smith & Schnepel, 2000).

It is difficult to reach an informed view on this issue, however, without information on the criminal careers of minor offenders. If most of those who commit minor offences go on to commit serious offences, it could be argued there is little to be gained by removing the possibility of criminal prosecution for minor crimes. If most of those who commit minor offences have no further contact with the court system, it could be argued that such offences are best dealt with by other means (e.g. infringement notices). Consider, for example, the prohibition against using or possessing (small amounts of) certain drugs. In the fiscal year 2020/21, more than 34,000 people in Australia were sanctioned by the courts for using or possessing a prohibited drug (Australian Bureau of Statistics, 2022). Ritter et al. (2013) estimated that Australian State and Territory spending on drug law enforcement in 2009/10 amounted to $770.8 million per annum (or $958 million in 2020 dollars). Since court appearances for drug use/possession in 2020/21 (36,126) exceeded the combined total of those for drug importation, trafficking, cultivation, and manufacture (12,620) by almost three to one (Australian Bureau of Statistics, 2022), we can safely assume much of the expenditure would have been directed towards those who use illicit drugs. Given the numbers involved, the cost of prosecution in terms of diminished offender employment and earning prospects is also likely to be large. The return on this investment, however, appears to be fairly limited. The available evidence indicates that the general deterrent effect of drug prohibition is small or non-existent (Donnelly et al., 2000; Hughes & Stevens, 2010; Single, 1989; Weatherburn et al., 2022). The few studies that have been conducted on the specific deterrent effect of sanctioning drug users find no effect (Green & Winik, 2010; Mitchell, 2016; Spohn & Holleran, 2002; Weatherburn & Yeong, 2021).

Many have argued on the basis of these facts that we should remove the sanctions against illicit drug use and possession (for a detailed discussion of the arguments, see MacCoun & Reuter, 2001). Policy makers contemplating this option, however, are hampered by the fact that there is almost no reliable information on the longitudinal pattern of contact between illicit drug users and the criminal justice system. We do not know whether a conviction for drug use and/or possession is just one minor element in a pattern of otherwise serious contact with the criminal justice system. We do not know what proportion of those convicted of drug use and/or possession has no further contact with the criminal justice system. We do not know which (if any) of those convicted of using or possessing a prohibited drug are likely to re-offend or transition to more serious offences. We do not know whether the risk or type of any further offending depends on the kind of drug defendants are prosecuted for having in their possession.

Information on these issues is of considerable theoretical, legal, and public policy importance. In this article, we investigate the criminal careers of 13,953 individuals in NSW, Australia, whose first proven offence involved using or possessing a prohibited drug. Our primary aim is to estimate the proportion that will re-appear in court for a further offence – however, we also seek answers to two other questions. The first concerns the types of offences committed by those who do re-offend. The second concerns the factors at the first court appearance that influence the likelihood that a person convicted of drug use and/or possession will have any further contact with the criminal justice system. Among these factors, we are particularly interested in one, namely whether the type of drug found in a person's possession by police has any bearing on the risk of reoffending. Our interest in this factor stems in part from the fact that the penalty imposed on persons convicted of drug possession varies significantly with the type of drug found in their possession (see Appendix), and in part from the possibility that the criminal careers of those who use illicit drugs may depend upon the kinds of drugs they consume. If this is true, the case for decriminalising illicit drug use may depend upon the type of drug found in a person's possession. In summary, the questions we seek to answer are:

What proportion of those whose first proven offence involves possession of (a small quantity of) a prohibited drug, is reconvicted of a further offence?

Among those who are reconvicted, what offences do they typically commit?

What effect does drug type have on the likelihood and speed of reconviction among those who are reconvicted?

Methods

The University of New South Wales (NSW) Human Research Ethics Committee granted approval for the study.

Data source

Data for the study were sourced from the NSW Bureau of Crime Statistics and Research re-offending database (ROD). ROD was established to facilitate and encourage further research into the factors influencing re-offending. Access to ROD data can be obtained on application to the NSW Bureau of Crime Statistics and Research at bcsr@justice.nsw.gov.au.

Each time a person in New South Wales is detained by police as a suspected offender (referred to hereafter as a CJS contact), a record is created in ROD containing details of the way the person was dealt with (viz. whether cautioned, referred to a youth justice conference or charged and required to appear in court), the charge or charges laid, the defendant's plea (guilty, not guilty, no plea), the court that dealt with the case (where applicable), and the court outcome (e.g. convicted/not convicted) and the penalty (type and quantum). Additional data collected at each CJS contact include the defendant's age; gender; Indigenous status; number and type of concurrent offences; and the socioeconomic status of the postcode in which each defendant lived at the time of their court appearance. Successive CJS contacts are linked using a matching algorithm based on surname, date of birth and criminal names index (a fingerprint-based identifier). Details of the matching procedure can be found in Hua and Fitzgerald (2006). The ROD database commenced operation in 1994 and is still operational.

Cohort

The cohort consisted of all those whose principal offence (i.e. the offence attracting the most severe penalty) at their first court appearance (hereafter referred to as their index contact) involved using or possessing a small quantity of a prohibited drug. A ‘small quantity’ under the NSW Drug Misuse and Trafficking Act varies from drug to drug but is, for example, no more than 1 g in the case of cocaine, methamphetamine, or heroin; no more than 0.25 of a gram in the case of MDMA (ecstasy) and no more than 30 g in the case of cannabis leaf. Cases of drug use/possession in NSW are dealt with in the Local Court. The maximum penalty for using or possessing a prohibited drug is 2 years in prison or a $2200 fine. The sample was restricted to people born during or after 1984 to ensure we had every offender's full criminal history. ROD contains the dates of death of all persons who have appeared in court and subsequently died in NSW. Those who died in NSW during the study are censored from their date of death. We are unable to adjust our results for deaths or re-offending that occurs outside of New South Wales. The total sample for the study consisted of 13,953 individuals.

Measures

A reoffence in this study was defined as reconviction for any offence that occurred after the date on which the first offence is recorded. The measure of disadvantage we employ is drawn from the Socio-Economic Indexes for Areas (SEIFA) developed by the Australian Bureau of Statistics (2011a). For the purposes of this study, SEIFA was coded as quartiles, with a SEIFA quartile of one indicating the poorest socio-economic status (highest social disadvantage) and SEIFA 4 the highest socio-economic status. The SEIFA value employed in the study is determined from their address at the first court appearance. Our controls were based on an exhaustive analysis of the correlates of reoffending in the ROD data by Stavrou and Poynton (2016). They include sex and age at the first court appearance, Indigenous status and number of proven drug offences at the first court appearance, number of concurrent offences at the first court appearance, the jurisdiction at the first court appearance, and remoteness of residence at the first court appearance. The latter is measured using the Australian Bureau of Statistics ARIA index (Australian Bureau of Statistics, 2018). Offences were coded using ANZSOC; the Australian and New Zealand Standard Offence Classification (Australian Bureau of Statistics, 2011b).

Statistical analysis

Most studies of recidivism employ logistic regression. In these analyses, a fixed follow-up period is defined, typically 1 or 2 years in duration, and an individual is counted as a recidivist if they commit a proven offence within that follow-up period. Our aim in this study is not to estimate the likelihood of re-offending over some limited period such as 1 or 2 years. We seek instead to estimate the likelihood of a person ever being convicted of another offence. The appropriate model for such estimation is a cure fraction survival model. Cure fraction models have an advantage over other survival models (e.g. Cox regression) in that they allow for the possibility that the outcome of interest (in our case re-offending) may never occur. They also provide information about the expected time to reconviction. In our context, the model generates estimates of the way in which various factors influence both the fraction who never re-offend and the rate of re-offending among those who do re-offend. Cure fraction models require an assumption about the underlying survival distribution. After trying various distributions and evaluating model fit, the logistic distribution was found to fit the data best and produce the lowest value of the Akaike Information Criterion (AIC).

Results

Sample description

The mean follow-up period was 4.9 years (median = 4.3 years, sd. = 3.4 years, and range = 20.8 years). Table 1 provides descriptive statistics for the study sample.

Table 1.

Descriptive statistics (independent variables).

	Freq.	Percent		Freq.	Percent
Age group at index contact			Remoteness
10–13	15	0.11	Major cities	11,949	85.64
14–18	1748	12.53	Inner regional	1648	11.81
19–23	8125	58.23	Outer regional	334	2.39
24–28	3290	23.58	Remote/very remote	22	0.16
29–33	774	5.55	Drug type
34–38	1	0.01	Amphetamines	1685	12.08
Aboriginal status			Barbiturates	162	1.16
Non-ATSI	12,509	89.65	Cannabis	2542	18.22
ATSI	176	1.26	Cocaine	1558	11.17
Unknown	1268	9.09	Ecstasy	7560	54.18
Gender			Hallucinogens	226	1.62
Female	3066	21.97	Opiates	105	0.75
Male	10,887	78.03	Other	115	0.82
SEIFA quartile			Proven drug offences
Highly disadvantaged	2157	15.09	1	12,471	89.38
Disadvantaged	2793	19.75	2 or more	1482	10.62
Advantaged	3462	24.58	Proven concurrent charges
Highly advantaged	5541	40.58	1	12,226	87.62
Court			2 or more	1727	12.38
Local Court	13,566	97.23
Children's Court	387	2.77

At the first court appearance, the modal age group was 19–23 and the median age was 21. The vast majority were non-Indigenous males, and most lived in an urban neighbourhood rated as ‘advantaged’ or ‘highly advantaged’. Most offenders were adults and dealt with in the Local Court, but a small percentage were juveniles and dealt with in the Children's Court. The majority lived in major cities and the most common drugs found in possession of those arrested were ecstasy, cannabis, amphetamines, and cocaine. Most offenders had just one drug offence and no non-drug offences.

Table 2 shows the distribution of the 27 most frequent offences at first re-offence among those who were convicted of a further offence. The 27 listed in Table 2 account for around 95% of all offences at first re-offence. The most common re-offence is another drug possession offence. What stands out about Table 2, however, is the wide variety of other offences committed by those who re-offend. In addition to drug offences, they include property damage, theft offences, violent offences, fraud offences, and public order offences. Drug offences account for less than half (40%) of all re-offences. The single most common category is driving offences, which collectively account for more than one-third (35.5%) of the offences in the table.

Table 2.

Principal re-offence profile of offending among drug offenders who were reconvicted.

Rank	ANZSOC code	Offence	Freq.	%	Cum %
1	1041	Possess illicit drugs	1016	30.84	30.84
2	1431	Drink-driving	548	16.64	47.48
3	1411	Drive while disqualified	245	7.44	54.92
4	1434	Drive under the influence of drugs	170	5.16	60.08
5	213	Common assault	143	4.34	64.42
6	1022	Export illicit drugs	110	3.34	67.76
7	1219	Property damage	102	3.1	70.86
8	211	Serious assault resulting in injury	87	2.64	73.5
9	1412	Drive without a licence	82	2.49	75.99
10	1562	Resist or hinder police/justice official	61	1.85	77.84
11	412	Dangerous/negligent driving	58	1.76	79.6
12	831	Theft from the person	57	1.73	81.33
13	829	Theft (except motor vehicles)	51	1.55	82.88
14	1439	Regulatory driving offences	40	1.21	84.09
15	291	Stalking	37	1.12	85.21
16	823	Theft from retail premises	34	1.03	86.24
17	1313	Riot and affray	34	1.03	87.27
18	1626	Licit drug offences	32	0.97	88.24
19	911	Obtain benefit by deception	30	0.91	89.15
20	1121	Unlawfully possess weapons/explosives	29	0.88	90.03
21	212	Serious assault not resulting in injury	26	0.79	90.82
22	411	Drive under the influence	26	0.79	91.61
23	711	Break and enter	26	0.79	92.4
24	1099	Other illicit drug offences	26	0.79	93.19
25	1311	Trespass	19	0.58	93.77
26	1332	Offensive behaviour	17	0.52	94.29
27	532	Threatening behaviour	15	0.46	94.75

Multivariate analysis

Figure 1 shows the estimated cumulative proportion of re-offending for the average offender (viz. modal values on all covariates) over a 30-year time horizon.

Figure 1.

Estimated cumulative proportion re-offending.

At the 30-year mark, an estimated 70% have not yet been convicted of a further offence. Table 3 shows the results of the cure fraction analysis.

Table 3.

Parameter estimates of the mixture cure fraction model.

Function		Cure fraction	Scale
Variable	Category	Coefficient	Coefficient
ATSI status	Non-ATSI
	ATSI	−0.118*	0.443
	Unknown	−0.063	−0.300
Gender	Female
	Male	−0.123*	0.152
SEIFA quarter	Highly disadvantaged
	Disadvantaged	0.07*	−0.252
	Advantaged	0.089*	−0.229
	Highly advantaged	0.159*	−0.343*
Age group	10–18
	19–23	0.139*	−0.311*
	24–28	0.259*	−0.164
	29–38	0.307*	0.074
Remoteness	Major cities
	Inner regional	−0.012	−0.093
	Outer regional	−0.066	−0.154
	Remote/very remote	0.088	−0.001
Drug type	Cocaine
	Cannabis	−0.225*	0.246
	Opiates	−0.182*	1.408*
	Amphetamines	−0.16*	0.654*
	Hallucinogens	−0.140	−0.761
	Ecstasy	−0.006	−0.409*
	Barbiturates	−0.009	1.25*
	Other	0.021	0.798
Proven drug offences	1
	2 or more	−0.037	−0.819*
Proven concurrent charges	1
	2 or more	−0.070	0.553*
Court	Children's Court
	Local Court	−0.005	−0.267
Constant		0.492*	−6.628*
Function			Shape
			Coefficient
Constant			−0.076*
AIC		62146.02

The column labelled ‘Cure Fraction’ shows the effect of each variable on the estimated proportion who never re-offend. A significant positive coefficient against an offender/offence characteristic indicates that the proportion with this characteristic, which desists from offending is significantly higher than the proportion in the reference group (when all other variables are held constant). A significant negative coefficient against an offender characteristic indicates the reverse. To illustrate; the coefficient on Indigenous status is negative; indicating that the estimated proportion who commit no further offence is lower among Indigenous offenders than non-Indigenous offenders (all other factors held constant).

The column labelled ‘Scale’ shows the effect of each variable on the rate of re-offending among those who do re-offend. The interpretation of the parameter estimates is the opposite of that in relation to the cure fraction. A significant positive scale coefficient indicates a shorter time to reconviction, compared to the reference group, while a significant negative coefficient indicates a longer time to reconviction. The scale coefficient for Indigenous offenders, for example, is positive; indicating that Indigenous offenders who commit another offence tend to do so faster than non-Indigenous offenders who commit another offence (other things being equal).

Inspection of the cure fraction coefficients in Table 3 reveals that those associated with disadvantage are all positive and increasing. This means there is a lower risk of re-offending among those who are disadvantaged, advantaged, or highly advantaged than among those who are highly disadvantaged. Put more simply, the likelihood of a further offence increases with disadvantage. The cure fraction coefficients on drug type are (except for the ‘other’ group) all negative, although only those associated with cannabis, opiates and amphetamines are significant. Since the reference category for drug type is cocaine, it follows that those whose first offence involves cannabis, opiates or amphetamines are significantly more likely to re-offend than those whose first offence involves cocaine. Significantly higher risks of re-offending are also found among Indigenous, male, and younger offenders. There is no significant association between the cure fraction and either remoteness, number of proven drug offences, number of concurrent charges or court.

The scale coefficient pattern is broadly consistent with the pattern of cure fraction coefficients. Those who are highly advantaged and who do re-offend, tend to re-offend more slowly than those who are highly disadvantaged. The same is true for older offenders. As might have been expected from the cure fraction coefficients, those whose first offence involved opiates or amphetamines re-offended more quickly than those whose first offence involved cocaine. On the other hand, those whose first offence involved ecstasy, re-offend more slowly than those whose first offence involved cocaine. Those with more concurrent offences also offend more quickly. For reasons that are unclear, although there was no apparent difference between those found in possession of cocaine and those found in possession of barbiturates in terms of the estimated proportion of re-offending, those found in possession of barbiturates who do re-offend, tend to re-offend more quickly. Those whose first court appearance involved two or more drug offences, for some unknown reason, re-offended more slowly.

To get a clearer picture of the effect of different offender/offence characteristics on the proportion that never re-offend, we use the coefficients from the model shown in Table 3 to estimate the proportion who never commit another offence for four distinct groups of offenders varying in age at index appearance, Indigenous status, level of disadvantage and drug type. This is done by taking the constant value for the cure fraction in Table 2 and then adding or subtracting (depending on the sign of the variable) the coefficients associated with the factor of interest and the characteristics of the offenders. For example, the constant for the cure fraction in the model is 0.492. If we wish to know the cure fraction for an offender aged 19–23 at their first court contact with other characteristics at the reference group (i.e. non-ATSI, female, highly disadvantaged, resided in major cities, dealt with by Children's Court, cocaine for the principal offence, one proven concurrent charges and one proven drug offence at index contact), we simply add the coefficient for this variable value (0.139) to 0.492 to obtain a revised cure fraction of 0.631.

Table 4 shows the cure fraction for distinct groups of offenders with other characteristics fixed at their modal values. The first panel shows the effect of varying ages. The remaining panels show the effects, respectively, of variations in Indigenous status, level of disadvantage and drug type.

Table 4.

Estimated percentage not committing another offence by offence/offender characteristics.

Base case	Variable of interest	Category	Cure fraction
Non-ATSI	Age group at index contact	10–18	51.73%
Male		19–23	65.60%
Resided in area with high socioeconomic advantage		24–28	77.60%
Resided in major cities		29–38	82.42%
Ecstasy for the principal offence at index contact
Had one proven drug offence at index contact
Had one proven concurrent charge at index contact
Appeared in Local Court at index contact

Aged 19–23 at index contact	Aboriginal status	Non-ATSI	65.60%
Male		ATSI	53.80%
Resided in area with high socioeconomic advantage		Unknown	59.31%
Resided in major cities
Ecstasy for the principal offence at index contact
Had one proven drug offence at index contact
Had one proven concurrent charge at index contact
Turned up in Local Court at index contact
Base case	Variable of interest	Category	Cure fraction
Non-ATSI	SEIFA quartile	Highly disadvantaged	49.69%
Aged 19–23 at index contact		Disadvantaged	56.67%
Male		Advantaged	58.59%
Resided in major cities		Highly advantaged	65.60%
Ecstasy for the principal offence at index contact
Had one proven drug offence at index contact
Had one proven concurrent charge at index contact
Turned up in Local Court at index contact
Non-ATSI	Drug type	Cocaine	66.16%
Aged 19–23 at index contact		Barbiturates	65.23%
Male		Ecstasy	65.60%
Resided in area with high socioeconomic advantage		Hallucinogens	52.18%
Resided in major cities		Amphetamines	50.13%
Had one proven drug offence at index contact		Opiates	48.01%
Had one proven concurrent charge at index contact		Cannabis	43.65%
Turned up in Local Court at index contact

There is an inverse relationship between the level of disadvantage and the estimated proportion who cease re-offending, with those in the highly advantaged category having a cure fraction of around 66% compared with 50% for those in the highly disadvantaged group. After fixing other factors in the modal category, the cure fraction for those whose drug possession offence involved cocaine (66%) is markedly higher than the cure fraction for those whose drug possession offence involved cannabis (44%). The cure fractions for other drugs are intermediate between these two extremes. The cure fraction increases with the age at the first court appearance, reaching over 80% among those whose first appearance in court occurs when they are 29 to 38. The cure fraction for Indigenous offenders, however, is about 12 percentage points lower than for non-Indigenous offenders.

Discussion

We had three aims in this study. The first was to estimate the reconviction risk among those whose first proven offence involves possession of a prohibited drug. The second was to determine the offence profile among those who are reconvicted. The third was to examine the effect of various factors, including drug type, on the likelihood and speed of reconviction. Our estimates indicate that the majority (73%) of those convicted of illicit drug use and/or possession are not (and will not be) convicted of any further offence. Among those who do re-offend, most (31%) will be reconvicted (at first instance) of either a drug use/possession offence or some other fairly minor offence (see below).

The fact that a substantial proportion of those convicted of drug possession has no further contact with the court system runs counter to the popular stereotype of illicit drug users as prone to violence and/or to committing income-generating crimes such as burglary or robbery. This is not surprising. The evidence linking illicit drug use to property and violent crime comes from studies of frequent or dependent drug users. The majority of illicit drug users, however, are not and will not become dependent drug users. Anthony et al. (1997), for example, found that dependence among non-medical users of illicit drugs ranged between 3.7% for users of inhalants and 23.1% for users of heroin. The corresponding proportions for other drugs were: 9.1% (cannabis); 16.7% (cocaine), 11.2% (other stimulants); and 4.9% (psychedelics). Similar findings have emerged from the 2001–2002 US National Epidemiologic Survey of Alcohol and Related Disorders (Moss et al., 2014). The Australian National Strategy Survey includes a question asking current (last year) drug users how frequently they use their preferred drug. Sixty-three percent of those who report using cannabis say they use it once a month or less frequently. Only 17% of those who report using methamphetamine say they used it once a week or more. The percentages of ecstasy and cocaine users who report weekly use of ecstasy and/or cocaine were each less than 5% (Australian Institute of Health and Welfare, 2020).

Although the majority of those convicted of drug use/possession were not reconvicted in our study, just over a quarter (27%) were reconvicted. We cannot tell from our data whether those who re-offended eventually progressed to committing more serious offences. If their first re-offence is any guide, however, there is little sign of such progression. Sixty percent of those who re-offended were convicted of either another drug possession offence (30.84%), drink-driving (16.64%), driving while disqualified (7.44%) or driving under the influence of drugs (5.16%). Fewer than one in 10 were convicted of violent or threatening behaviour. Fewer than one in 100 were convicted of a break and enter offence. None were convicted of robbery.

Significantly higher risks of reconviction were found among Indigenous, male, younger offenders and those living in areas within the lowest quartile of disadvantage (i.e. the most disadvantaged). These findings have been observed in other studies (Daniel et al., 2009; Stavrou & Poynton, 2016). The findings concerning drug type, however, are new and, in some cases, surprising. One might have expected those who use drugs commonly associated with involvement in property or violent crime, such as opioids or methamphetamine, to have the highest risk of further offending, while those who use drugs not typically associated with involvement in crime, such as cannabis, to have the lowest rate of re-offending. The results indicate that, even after controlling for age, gender, remoteness of residence, level of disadvantage, Indigenous status and number of concurrent offences, those found in possession of cannabis are the most likely to re-offend, while those found in possession of cocaine, barbiturates and ecstasy are the least likely to re-offend. The differences are substantial. As can be seen from the bottom half of Table 4, the cure fraction (viz. the estimated percent who will not re-offend) for those found in possession of cocaine and barbiturates are 66% and 65%, respectively, while the cure fraction for those found in possession of opiates or cannabis are 48% and 44%, respectively.

The high cure fraction for those found in possession of cocaine, barbiturates or ecstasy may be due to the fact that these drugs are either expensive (cocaine), difficult to obtain (barbiturates) and/or consumed at locations where they are obtained (e.g. ecstasy at rave parties), and therefore not in the possession of a user for every long. Users of these drugs, moreover, may cease consuming them while they are still comparatively young. Any or all of these factors would tend to lower the risk of detection. Cannabis, by contrast, is an inexpensive drug and one that is regularly used by a large proportion of the population, especially those who reside in more disadvantaged areas (NSW Bureau of Crime Statistics and Research, 2022a). This, and the fact that disadvantaged communities tend to be policed more aggressively (e.g. more frequently subjected to stop and search, more often subjected to intense surveillance) than those who live in better-off neighbourhoods (Braga et al., 2019; Maher & Dixon, 2001; Murphy et al., 2020; Sarre, 2005) would expose cannabis users to a higher risk of apprehension than those living in wealthier areas where cocaine use predominates (NSW Bureau of Crime Statistics and Research, 2022a). It would therefore be imprudent to consider the type of drug found in a person's possession as an indicator of the likelihood of further offending.

Australian support for legalising personal cannabis, ecstasy, and cocaine has risen significantly since 2013. Support for taking no action, or issuing a caution or a warning, has also grown for those found in possession of cannabis, ecstasy, and heroin (Weatherburn et al., 2021). In the introduction to this article, we noted that the prosecution of illicit drug users consumes a substantial proportion of magistrate court workload, yet there is little evidence that such prosecution exerts a substantial general or specific deterrent effect. The low overall rate of reconviction revealed in this study adds further weight to arguments for the decriminalisation of drug use/possession or to the imposition of less onerous sanctions. Although we have focussed on the criminal careers of those convicted of prohibited drug possession, there are many other minor offences worthy of close examination. To take just one example, in 2020/21 more than 18,000 adults and more than 1300 children appeared in an Australian court charged with committing a public order offence (Australian Bureau of Statistics, 2022). If NSW is any guide, many of these offences involve nothing more than offensive language (NSW Bureau of Crime Statistics and Research, 2022b).

More than half a century ago, Morris and Hawkins (1971) criticised what they described as the ‘over-reach of the criminal law’ and recommended the removal of a number of criminal offences from the statute books, including drunkenness, drug possession and disorderly conduct. It is to be hoped that the current study will encourage others to take a closer look at the outcomes of those charged with minor offences.

Footnotes

Acknowledgements

We thank the reviewers for their constructive and helpful feedback on the initial draft of this paper.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Don Weatherburn

Appendix

Average fine by drug type (NSW)
Drug type	Average fine ($)
Narcotics	1415
Cocaine	486
Amphetamines	459
Ecstasy	421
Cannabis	375

Source: NSW Bureau of Crime Statistics and Research (2021).

References

Agan

A. Y.

Doleac

J. L.

Harvey

(2021). Misdemeanour prosecution. National Bureau of Economic Research. Working Paper 28,600. https://doi.org/10.3386/w28600.

Anthony

J. C.

Warner

L. A.

Kessler

R. C.

(1997). Comparative epidemiology of dependence on tobacco, alcohol, controlled substances, and inhalants: Basic findings from the National Comorbidity Survey. Retrieved from: https://psycnet.apa.org/record/1994-45545-001. Accessed 8/9/2021.

Australian Bureau of Statistics (2018). Australian Statistical Geography Standard (ASGS): Volume 5 - Remoteness Structure. Canberra. Retrieved from: https://www.abs.gov.au/websitedbs/d3310114.nsf/home/australian + statistical + geography + standard + (asgs). Accessed 8/9/2021.

Australian Bureau of Statistics (2022). Criminal Courts, Australia 2020-21. Retrieved from; https://www.abs.gov.au/statistics/people/crime-and-justice/criminal-courts-australia/latest-release&hash;:∼:text = Most&per;20defendants&per;20(468&per;2C349&per;20or&per;2092,Children’s&per;20Courts&per;2C&per;20a&per;201&per;25&per;20increase. Accessed 18/3/2022.

Australian Bureau of Statistics (2011a). Census of Population and Housing: Socio-Economic Indexes for Areas (SEIFA). Cat. no. 2033.0.55.001. Canberra. Retrieved from: https://www.abs.gov.au/ausstats/abs@.nsf/mf/2033.0.55.001. Accessed 8/9/2021.

Australian Bureau of Statistics (2011b). Australian and New Zealand Standard Offence Classification (ANZSOC). Cat. No. 1234.0. Canberra. Retrieved from: https://www.abs.gov.au/ausstats/abs@.nsf/mf/1234.0. Accessed 8/9/2021.

Australian Institute of Health and Welfare (2020). National Drug Strategy Household Survey 2019. Drug Statistics series no. 32. PHE 270. Canberra: Australian Institute of Health and Welfare. Retrieved from: https://www.aihw.gov.au/reports/illicit-use-of-drugs/national-drug-strategy-household-survey-2019/contents/summary. Accessed 8/9/2021.

Blumstein

Cohen

Roth

Visher

(1986). Criminal careers and career criminals. Vol. 1. National Academies Press.

Borland

Hunter

(2000). Does crime affect employment status? Economica, 67(265), 123–144. https://doi.org/10.1111/1468-0335.00199

10.

Braga

A. A.

Brunson

R. K.

Drakulich

K. M.

(2019). Race, place, and effective policing. Annual Review of Sociology, 45(1), 535–555. https://doi.org/10.1146/annurev-soc-073018-022541

11.

Centre for Road Safety (2020). Drink driving attitudes and behaviours. Summary Report. Retrieved from: https://roadsafety.transport.nsw.gov.au/search_results_iframe.html?q = drink + driving + attitudes + &hash;gsc.tab = 0&gsc.q = drink&per;20driving&per;20attitudes&per;20&gsc.page = 1. Accessed 18/3/2022.

12.

Council of Europe (2017). Costs and unintended consequences of drug control policies. Report by the expert group on possible adverse effects and associated costs of drug control policies. Retrieved from: https://rm.coe.int/costs-and-unitended-consequences-of-drug-control-policies/16807701a9. Accessed 18/3/2022.

13.

Daniel

J. Z.

Hickman

Macleod

Wiles

Lingford-Hughes

Farrell

Araya

Skapinakis

Haynes

Lewis

(2009). Is socioeconomic status in early life associated with drug use? A systematic review of the evidence. Drug and Alcohol Review, 28(2), 142–153. PMID: 19320699. https://doi.org/10.1111/j.1465-3362.2008.00042.x

14.

Donnelly

Hall

Christie

(2000). The effects of the cannabis expiation notice system on the prevalence of cannabis use in South Australia: Evidence from the national drug strategy household surveys 1985–95. Drug and Alcohol Review, 19(3), 265–269. https://doi.org/10.1080/713659382

15.

Graffam

Shrinkfield

Lavelle

Hardcastle

(2004). Attitudes of employers, corrective services workers, employment support workers, and prisoners and offenders towards employing ex-prisoners and ex-offenders. Report to the Criminology Research Council. Grant 26/02-03.

16.

Green

D. P.

Winik

(2010). Using random judge assignments to estimate the effects of incarceration and probation on recidivism among drug offenders. Criminology: An Interdisciplinary Journal, 48(2), 357–387. https://doi.org/10.1111/j.1745-9125.2010.00189.x

17.

Grogger

(1995). The effect of arrests on the employment and earnings of young men. The Quarterly Journal of Economics, 110(1), 51–71. https://doi.org/10.2307/2118510

18.

Howard

(2020). Special Commission of Inquiry into crystal methamphetamine and other amphetamine-type stimulants. Volume 1. https://www.dpc.nsw.gov.au/publications/special-commissions-of-inquiry/the-special-commission-of-inquiry-into-the-drug-ice/.

19.

Hua

Fitzgerald

(2006). Matching court records to measure reoffending. In Crime and justice bulletin (Vol. 95, pp. 1–12). NSW Bureau of Crime Statistics and Research. Retrieved from: https://www.bocsar.nsw.gov.au/Publications/CJB/cjb95.pdf. Accessed 8/9/2021

20.

Hughes

Stevens

(2010). What can we learn from the Portuguese decriminalization of illicit drugs? The British Journal of Criminology, 50(6), 999–1022. https://doi.org/10.1093/bjc/azq038

21.

MacCoun

R. J.

Reuter

(2001). Drug war heresies: Learning from other vices, times, and places. Cambridge.

22.

Maher

Dixon

(2001). The cost of crackdowns: Policing Cabramatta’s heroin market. Current Issues in Criminal Justice, 13(1), 5–22. https://doi.org/10.1080/10345329.2001.12036213

23.

Mitchell

(2016). The effect of drug arrest on subsequent drug offending and social bonding. Journal of Crime and Justice, 39(1), 174–188. https://doi.org/10.1080/0735648X.2015.1087145

24.

Morris

Hawkins

(1971). The honest politician’s guide to crime control. Sun Books.

25.

Moss

Chen

(2014). Early adolescent patterns of alcohol, cigarettes, and marijuana polysubstance use and young adult substance use outcomes in a nationally representative sample. Drug and Alcohol Dependence, 136, 51–62. https://doi.org/10.1016/j.drugalcdep.2013.12.011

26.

Mueller-Smith

Schnepel

K. T.

(2021). Diversion in the criminal justice system. The Review of Economic Studies, 88(2), 883–936. https://doi.org/10.1093/restud/rdaa030

27.

Murphy

Madon

Cherney

(2020). Reporting threats of terrorism: Stigmatisation, procedural justice and policing Muslims in Australia. Policing and Society, 30(4), 361–377. https://doi.org/10.1080/10439463.2018.1551393

28.

Nagin

Waldfogel

(1995). The effects of criminality and conviction on the labour market status of young British offenders. International Review of Law and Economics, 15(1), 109–126. https://doi.org/10.1016/0144-8188(94)00004-E

29.

Newton

Day

Giles

Wodak

Graffam

Baldry

(2018). The impact of vocational education and training programs on recidivism: A systematic review of current experimental evidence. International Journal of Offender Therapy and Comparative Criminology, 62(1), 187–207. https://doi.org/10.1177/0306624X16645083

30.

NSW Bureau of Crime Statistics and Research (2021). Personal communication. Ms Stephanie Ramsey. 6^th April 2021.

31.

NSW Bureau of Crime Statistics and Research (2022a). Incidents of Drug Offences: Drug Offences from January 2021 to December 2021. http://crimetool.bocsar.nsw.gov.au/bocsar/

32.

NSW Bureau of Crime Statistics and Research (2022b). Criminal Courts Statistics: July 2016 to June 2021. https://www.bocsar.nsw.gov.au/Pages/bocsar_court_stats/bocsar_court_stats.aspx

33.

Ritter

McLeod

Shanahan

(2013). Government Drug Policy Expenditure in Australia-2009/10. Drug Policy Modelling Program Monograph 24. https://ndarc.med.unsw.edu.au/sites/default/files/ndarc/resources/24&per;20Government&per;20drug&per;20policy&per;20expenditure&per;20in&per;20Australia&per;20-&per;202009_10.pdf.

34.

Sarre

(2005). Police and the public: Some observations on policing and Indigenous Australians. Current Issues in Criminal Justice, 17(2), 305–313. https://doi.org/10.1080/10345329.2005.12036357

35.

Single

(1989). The impact of marijuana decriminalization: An update. Journal of Public Health Policy, 10(4), 456–466. https://doi.org/10.2307/3342518

36.

Spohn

Holleran

(2002). The effect of imprisonment on recidivism rates of felony offenders: A focus on drug offenders. Criminology: An Interdisciplinary Journal, 40(2), 329–357. https://doi.org/10.1111/j.1745-9125.2002.tb00959.x

37.

Stavrou

Poynton

(2016). The revised Group Risk Assessment Model (GRAM 2): Assessing risk of reoffending among adults given non-custodial sanctions. Crime and Justice Bulletin 197. Sydney: NSW Bureau of Crime Statistics and Research. Retrieved from: https://www.bocsar.nsw.gov.au/Publications/CJB/Report-2016-GRAM2-Group-Risk-Assessment-Model-CJB197.pdf. Accessed 8/9/2021.

38.

Waldfogel

(1994). The effect of criminal conviction on income and the trust “Reposed in the Workmen”. The Journal of Human Resources, 29(1), 62–81. https://doi.org/10.2307/146056

39.

Weatherburn

Alexeev

Livingston

(2021). Changes in and correlates of Australian public attitudes toward illicit drug use. Drug and Alcohol Review. 116(6), 1472–1471. https://doi.org/10.1111/add.15298

40.

Weatherburn

Darke

Zahra

Farrell

(2022). Who would try (or use more) cannabis if it were legal? Drug and Alcohol Review, 41(2), 386–395. https://doi.org/10.1111/dar.13360

41.

Weatherburn

Yeong

(2021). Does correctional supervision of amphetamine users reduce the risk of re-offending? Addiction, 116(6), 1472–1481. Epub 2020 Dec 13. PMID: 33067834. https://doi.org/10.1111/add.15298