SAPRP

Proposition 36 had a significant and instantaneous impact on treatment referrals in California.
There were over 44,000 Proposition 36 treatment referrals in FY2001 and this figure hovered around 50,000 in the subsequent years (Longshore et al., 2003; Urada and Evans, 2008a). Proposition 36 clients accounted for almost 25% of admissions to treatment in California in FY2006 (California Department of Alcohol and Drug Programs, 2007).

Approximately 71% of Proposition 36 referrals entered treatment.
Of the 50,732 Proposition 36 referrals in FY2006, 43,219 were assessed and 36,221 were placed in treatment (Urada and Evans, 2008a). Compared to other admissions, Proposition 36 clients were more likely to be "men, first-time admissions, treated in outpatient drug-free programs, employed full-time, and users of methamphetamine or marijuana" (Hser et al., 2003).

Surprisingly little is known about what happened to the nearly 30% of referrals who failed to show up for assessment or treatment. Evans and Hser (2008) examined Proposition 36 referrals who did and did not enter treatment in five counties (n=1,335 and 124, respectively). They found that the untreated group tended to be younger, more criminally active, and less motivated for treatment. As for reasons for not entering treatment, 32% reported they were rearrested before entering, 24% reported they had no desire for treatment, and 11% reported that they were assigned to a program that that was too far away.

For those Proposition 36 referrals who eventually entered treatment, the completion rate was approximately 33%.
This completion rate is typical for criminal justice referrals to drug treatment (Urada and Evans, 2008b) and very similar to the rate for everyone discharged from substance-abuse treatment in California in 2005 (31%; Office of Applied Studies, 2008). Since almost 30% of Proposition 36 clients never entered treatment, the treatment completion rate for all Proposition 36 offenders was approximately 24%.

Proposition 36 clients may have displaced clients who were entering treatment voluntarily.
Among the treatment providers interviewed in five counties, Hser and colleagues (2007b) found that one-third reported a decrease in treatment availability for non-Proposition 36 clients in the second year. The authors also found that voluntary treatment admissions in California decreased between the first and second years. While Proposition 36 led to a modest increase in treatment staff, it did not lead to a significant increase in the number of new facilities in the first two years (Hser et al., 2007b). Whether or not this potential displacement occurred in subsequent years has yet to be determined.

Those eligible for Proposition 36 were more likely to be arrested for drug and property offense compared to a matched sample.
Longshore et al."s (2006) individual-level analyses found that those eligible for Proposition 36 in its first year were "more likely to have a new arrest for drug or property offenses during the ensuing 30-months than a pre-SACPA-era comparison group of similar offenders who would have been eligible for SACPA." One explanation for this is that Proposition 36 offenders spent more time in the community than those in the comparison groups. Similar analyses by Urada and Hawken (2008) based on those eligible for Proposition 36 in its second and third years also found an increase in the drug and property arrests; however, the differences were not as large.

Another study by Worrall and colleagues (forthcoming) which analyzed arrest trends in Orange County, California yielded similar results. The authors found some evidence to suggest that robbery and paraphernalia arrests increased after Proposition 36 was implemented. It should be noted, however, that paraphernalia arrests were treated as drug arrests in Orange County, thus potentially promoting such arrests.

Proposition 36 saved taxpayer dollars in the short-run, primarily because of avoided incarceration costs.
California taxpayers saved $2-$2.50 for every $1 spent on Proposition 36, largely attributable to avoided incarceration costs (Longshore et al., 2006; Hawken et al., 2008). This is based on a cost-benefit study comparing Proposition 36 referrals with a group that would have been eligible for the program if it had existed before implementation. The study focused on the costs associated with criminal justice, treatment, and health care. The Legislative Analysts Office (2000) projected that the passage of Proposition 36 would lead to a "one-time avoidance of $450 - $550 million in capital outlay for new prison construction"; however, it is impossible to prove that the prison would have been constructed had Proposition 36 been rejected.

Whether or not Proposition 36 saves taxpayers money in the long run is a vital question deserving of additional research. Future studies should also incorporate the victimization costs associated with the increase in property offenses as well as other costs and benefits (e.g., the costs associated with being incarcerated or having a family member incarcerated).

A report by the Arizona Supreme Court (2006) finds evidence that Proposition 200 has saved taxpayer funds as well. They report over $14 million was avoided in prison costs and almost $6 million was spent on probation and treatment services; however, a full cost-benefit analysis from the perspective of the taxpayer has not been conducted.

Some California counties limited the use of Proposition 36 funds for narcotic replacement therapy (NRT).
In their five-county study, Hser and colleagues (2003) found that one did not include methadone maintenance programs for Proposition 36 and another neither funded methadone detoxification nor maintenance programs for anyone. Klein and colleagues (2004) conducted key informant interviews in eight counties and found that methadone maintenance would only be covered in four counties, but nearly all would offer methadone for a "very short amount of time." Given the strong body of evidence on the effectiveness of methadone treatment (Rettig & Yarmolinsky, 1995), this is a consequential decision. Indeed, Hawken et al. (2007) found that opiate-using Proposition 36 referrals who were placed in NRT had better criminal justice outcomes than those who were not.

Proposition 36 clients were less likely to be placed into residential care than other criminal justice referrals.
Farrabee et al. (2004) noted that this was the case in the six months after Proposition 36 was implemented. Hawken et al. (2007) looked at all California treatment admissions from 2001 to 2005 and found this was true even after controlling for client-level demographics and drug use patterns. It is unclear, however, why Proposition 36 referrals were less likely to enter residential treatment compared to other criminal justice referrals. Findings from stakeholder analyses found that some providers used a "fail-outpatient first" approach with Proposition 36 clients before allowing them to enter residential treatment (Hawken, 2008b). One hypothesis for the difference in residential uptake is that this "fail-outpatient first" approach is more common with Proposition 36 clients than other criminal justice referrals. Another hypothesis is that there is already an established pipeline between existing residential programs and those criminal justice officials who make non-Proposition 36 referrals. Given the importance of residential treatment for some clients, more work should be done developing and testing these hypotheses.

Additional analyses by Hawken et al. (2007) found that young Hispanic males were less likely to be placed into long-term residential treatment than White males after controlling for a host of factors. Further, this Hispanic-White difference was larger for Proposition 36 referrals than for other criminal justice referrals. Interestingly, the Proposition 36 difference declined with age.

Frequent urine testing for Proposition 36 clients by treatment providers increased the probability of remaining abstinent from illegal drugs.
Drug testing is often used by treatment providers and criminal justice agencies to monitor treatment compliance and deter consumption in a testing and sanctions regime. Based on a random sample of 1,104 Proposition 36 participants in five counties, Hser and colleagues (2007a) found that drug testing by treatment providers increased the probability of abstinence from illegal drugs after controlling for several individual-level characteristics and the time-invariant county characteristics; however, it did not have a statistically significant effect on re-arrests, employment, or treatment success (defined by the combination of drug abstinence, employment, and no crime). Since the sample is based on those who had entered treatment, we cannot extrapolate these results to all individuals eligible for Proposition 36.

There was no evidence that testing by the criminal justice system influenced abstinence, employment, or re-arrest, but this should not be interpreted as evidence that drug testing in the criminal justice system cannot work. To learn more about the ability of drug testing to influence behavior, future studies should include information on the actual and expected consequence associated with testing positive. Studies should also attempt to measure the use of substances not usually tested to determine if the individuals are substituting illegal drugs with something else (e.g., alcohol).

Many of the individuals incarcerated for low-level drug charges before the implementation of Propositions 36 and 200 had lengthy criminal histories.
Riley et al. (2005) found that before Proposition 200 those incarcerated for a low-level drug offense had an average of 17.1 prior offenses on their records. They also found that imprisoned non-sales drug offenders in California had more severe criminal histories than imprisoned sales offenders, leading them to conclude: "The findings support prosecutors contention that low-level offenders receiving prison sentences have more serious and extensive criminal histories than the low-level conviction label has suggested."

Additional analyses by Hawken (2008a) reveal that 25% of Proposition 36 offenders accounted for 80% of the follow-up crime costs. Based on this, UCLA recommends using readily available assessment tools to identify high-risk offenders and managing them differently; however, Hawken cautions that "Responding appropriately to the supervision and treatment needs of high-risk participants will be a challenge given the limited funding available for Proposition 36."

After Proposition 200 was implemented, those imprisoned for low-level drug offenses in Arizona had more extensive and severe criminal records.
Part of this is attributable to the diversion of those with less severe criminal histories to treatment, but analyses by Webb and Rodriguez (2006) also revealed that "arrestees with more extensive criminal histories were more likely to have the severity of charges increase rather than decrease, as seen pre-Proposition 200."

Arizona probationers with medium-level violations received longer prison sentences after the passage of Proposition 200.
Rodriguez and Webb (2007) note that medium-level violators primarily include those probationers who tested positive for drugs or failed to attend drug treatment. It is unclear why this group received longer incarceration sentences post-implementation, but Rodriguez and Webb (2007) offer a potential explanation: "violators who failed to comply with treatment-related provisions of probation may be seen as most deserving of longer sentences given they had the opportunity to receive substance abuse treatment." This is indeed a topic worthy of additional research attention.

Key Results: Drug Treatment for Drug-Abusing Criminal Offenders: Insights from California's Proposition 36 and Arizona's Proposition 200