Incorporating a Public Health Approach in Drug Law: Lessons from Local Expansion of Treatment Capacity and Access under California's Proposition 36

Abstract

A major state drug abuse initiative, California's Proposition 36 of 2000, mandated that adults convicted of drug possession be offered treatment in lieu of incarceration. While the law expanded public treatment for arrestees, the counties were given discretion in structuring their systems of care and procedures to manage clients. Using data from a study of key informants in eight counties, this article examines local planning to increase drug treatment capacity and manage clients’ access to treatment. In both these planning domains, it describes the counties’ strategies and concerns, reasons for their differences in approaches, and whether and how this state initiative, which explicitly incorporated treatment objectives into penal drug law, will shift the debate over drug abuse policy toward greater consideration of public health goals.

In recent years, with the policy debate intensifying in the United States over how society should respond to the use of illicit psychoactive drugs, the implementation of Proposition 36 in California is being closely watched. Passed in 2000, this ballot initiative mandated that adults in California who are convicted of possessing or using illegal drugs be offered substance abuse treatment in lieu of incarceration. The law also requires that the state appropriate a specified amount of funding toward this end. For the first time, a large number of adults arrested for drug use in a major state were to be automatically directed into community-based treatment. The intent was for arrested individuals’ need for treatment to take priority in determining society's intervention, rather than the penal considerations that previously had been foremost.

National observers have followed Proposition 36 and its aftermath with great interest. Some people have suggested that Proposition 36 might mark a turning point in substance abuse policy in the United States (Colker and Watson 2001; New York Times 2000; Nieves 2000). Its passage has been variously characterized as signaling the end of the decades-old dominance of law enforcement in policy toward drug users, that is, the “war on drugs” (Waters 2000); as proof that voters want to slow or reverse the steep climb of recent years in government expenditures on imprisonment (Butterfield 2001; Davenport 2001); as evidence of the public acceptance of addiction as an illness and support for treatment (Lichtblau 2001); and as an expression of public willingness to consider as a goal a move toward harm reduction or even the partial legalization of drugs (Wallace 2000).

On the heels of Proposition 36, several other states, including Washington, New York, New Mexico, Michigan, Hawaii, and Florida, considered drug policy reforms focusing on treatment as an alternative to incarceration, through legislation, ballot propositions, or executive initiatives (Drug Policy Alliance 2002; Drug Reform Coordination Network 2002a, b, c; Join Together Online 2002a, b, c). Comparable developments have taken place more recently in Maryland, Texas, and Kansas (Andrews and Schiraldi 2004; Drug Policy Alliance 2004). Some of these changes were spurred by states’ serious budget difficulties, which prompted them to look for alternatives to costly growing imprisonment (Butterfield 2004; New York Times 2004). Concerns about the equitability and proportionality of incarceration for drug users have also been raised. Indeed, the American Bar Association issued a report, at the urging of U.S. Supreme Court justice Anthony Kennedy, criticizing, on jurisprudential grounds, the current overreliance on prisons and calling for, among other reforms, the repeal of mandatory minimum sentences, more judicious incarceration of persons violating their probation or parole, and more diversion of arrestees away from punishment and into substance abuse treatment (American Bar Association 2004).

The drafters of Proposition 36 stated that they intended to provide treatment as an alternative to prison and to address substance abuse as a matter of public health rather than criminal justice (Hser et al. 2003; Marlowe et al. 2003). They also made fiscal arguments for reducing imprisonment (Appel 2004).

The idea for the proposition came at a retreat in 1997–98 for what was then called the Lindesmith Center (Abrahamson 2001). After merging with the Drug Policy Foundation, the center became known as the Drug Policy Alliance and has been a leading force in national- and state-level advocacy for drug law reform.

The general approach of Proposition 36 left many implementation details to be worked out by the 58 counties in California and the designated state “lead” agency, the Department of Alcohol and Drug Programs. Because California government is, both structurally and culturally, deeply decentralized, many of the proposition's administrative policies and practices were decided at the local county level. In this article we argue that the variations in these protocols and procedures may be important to determining the outcomes of Proposition 36 and assessing the relevance of California's legal experiment to other states. The article presents findings from an implementation study of eight counties just before Proposition 36 took effect. We focus on the issues that emerged in the local planning processes regarding the capacity of, and access to, publicly funded substance abuse treatment for persons found to have broken the drug laws. The concluding discussion addresses the implications of these findings for policy and research in both California and other states contemplating similar initiatives.

Provisions and Passage of Proposition 36

Proposition 36, formally known as the Substance Abuse and Crime Prevention Act of 2000 (SACPA), was passed by California voters on November 7, 2000. The initiative had sparked a fierce battle, and the decisive outcome was surprising. Despite the scant support of most elected officials and even many treatment providers, two-thirds of the counties voted in favor of Proposition 36, which took effect on July 1, 2001, with funding for five years, through June 2006.

Its provisions were as follows, per California Penal Code Section 1210 (also see California Board of Prison Terms 2001; and California Legislative Analyst's Office 2000a): Persons convicted of a “nonviolent drug possession offense” or parolees who committed a drug-related parole violation were to be placed on county probation and ordered by the court into as much as one year of community-based treatment plus up to six months of follow-up care. Persons convicted of drug sales or other offenses such as property crimes were excluded, as were persons with recent non-drug-possession felony or violent convictions and persons refusing treatment.

Under Proposition 36, persons who committed a first or second drug-related violation of their probation or parole, such as a new arrest for drug possession, were to have their treatment altered or intensified rather than face incarceration. The exceptions were persons who had committed non-drug-related violations or had been found to pose a danger to public safety or to be unamenable to treatment. After a third drug-related violation, persons would no longer be eligible for Proposition 36. Those who successfully completed their treatment and probation term under Proposition 36 could petition the court to set aside their recorded conviction, with certain exceptions.

Only treatment programs that had been certified or licensed by the state could receive clients under Proposition 36. For each of the first five fiscal years of the act (2001–02 through 2005–06), the state appropriated $120 million for implementation, along with $60 million for its start-up (during the last months of 2000–01), for a total of $660 million. The state's Department of Alcohol and Drug Programs allocated the money to the counties using a traditional population- and indicator-based formula and was to approve each county's proposed distribution of funds. The state was required to evaluate Proposition 36's effectiveness for each of the five years, including any reductions in crime or incarceration costs. (For the first-year report, see California Department of Alcohol and Drug Programs 2003; and Longshore et al. 2003.)

There was little direct precedent across the nation for a potentially far-reaching drug policy shift such as Proposition 36. One other state, Arizona, had undergone a similar process, with the passage of Proposition 200 in 1996 and Proposition 300 in 1998 (Arizona Supreme Court 1999, 2001). While there were some similarities between the two states’ initiatives and the problems they were intended to address, the difference in scale defied comparison. For instance, in 1999, slightly more than 1,000 persons were treated under Proposition 200, of whom nearly three-fourths were in the metropolitan Phoenix area (Arizona Supreme Court 2001). In comparison, under Proposition 36, 45,000 persons were expected to participate annually (California Legislative Analyst's Office 2000b), of whom only one-third would be in the largest metropolitan area, Los Angeles.

Nearly all observers predicted that the implementation of Proposition 36 would present an enormous challenge as well as a major opportunity (Howard 2002; Riley et al. 2000). One concern was whether the funding was adequate for the projected demand for treatment (Krikorian 2001). Another concern was the will and capability of the 58 California counties to coordinate their health care and criminal justice agencies so that the actual access to treatment would resemble the law's directive. It was also widely agreed that Proposition 36 might lead to the greater development of the public treatment system. Local authorities were expected to expand their county's capacity to treat a new wave of clients referred from criminal justice. They also were required to develop procedures to admit clients into treatment to meet new legal deadlines and withstand the scrutiny of the proposition's advocates. Thus there were questions as to whether Proposition 36 would be implemented as it was written, what could be learned from this experiment, whether it would be declared successful, and whether the results would influence national policy.

Study Methods

This article is based on a study of the early implementation of Proposition 36 statewide and in eight California counties. It concentrates on alterations in the counties’ criminal justice and drug treatment systems in response to the changes in organizational responsibilities and client flows required by Proposition 36. We chose a qualitative approach as the most suitable for this exploratory, process-oriented research. The findings presented here are based on analyses of data gathered between Proposition 36's passage in November 2000 and its taking effect in July 2001. The core is a series of interviews conducted in eight counties in May and June 2001 with key informants responsible for the local implementation of the proposition.

Sample Selection and Size

The sample consisted of eight counties: Alameda, Fresno, Imperial, Orange, Sacramento, San Francisco, Santa Clara, and Solano, which together contain 27.3 percent of the state's population. We excluded Los Angeles because of its size and complexity, as well as resource constraints. These constraints also led us to exclude the smallest and most remote counties.

The county sample was selected to maximize two kinds of variability that we believed might affect Proposition 36 at the county level. First was the amount of local voter support for the law. As shown in Table 1, support for Proposition 36 in the sample counties ranged from 45.3 percent (rural Imperial) to 76.3 percent (urban San Francisco), compared with the overall statewide support of 60.9 percent. We also sought variability in the lead agency, as shown in Table 1, for implementing the law. In half the sample counties, the lead agency was the alcohol and drug department or its parent health or human service agency. This was the most common arrangement statewide. The remaining four sample counties had as their lead agencies the county executive office (Orange, Santa Clara), probation department (Solano), or probation and alcohol and drug administration serving jointly (Sacramento).

TABLE 1.

County Proposition 36 Lead Agency and Vote

County	Proposition 36 Lead Agency	Prop. 36 Vote (% for)
Alameda	Health Services	69.5
Fresno	Human Services	47.7
Imperial	Behavioral Health	45.3
Orange	County Executive	60.8
Sacramento	Health and Human Services and Probation	56.0
San Francisco	Health Services	76.3
Santa Clara	County Executive	62.5
Solano	Probation	55.4

Source: California Secretary of State 2001.

We also selected the county sample to reflect the general heterogeneity among California's counties with respect to population size and density, land, and economic resources. As shown in Table 2, the 2001 population of the eight counties varied from less than 150,000 (rural Imperial) to nearly 3 million (Orange), and in population density per square mile from 35 (Imperial) to more than 16,000 (San Francisco). They varied in land from fewer than 50 square miles (urbanized San Francisco) to nearly 6,000 square miles (Fresno). The per capita 1999 income ranged from $13,239 in high-poverty Imperial to $34,556 in wealthy San Francisco. The percentage of households living below the federal poverty line in 1999 varied from 7.5 percent in affluent “high-tech” Santa Clara to 22.9 percent in the agricultural center of Fresno.

TABLE 2.

Geographic and Demographic Characteristics of California Counties

County	Population (in 1,000s)a	Square Milesb	Population Density per Square Mile	Mean per Capita Incomec	Percentage Households below Federal Poverty Levelc
Alameda	1,458	738	1,976	$26,680	11.0
Fresno	816	5,963	137	$15,495	22.9
Imperial	146	4,175	35	$13,239	22.6
Orange	2,890	789	3,663	$25,826	10.3
Sacramento	1,269	966	1,314	$21,142	14.1
San Francisco	771	47	16,404	$34,556	11.3
Santa Clara	1,668	1,291	1,292	$32,795	7.5
Solano	404	829	487	$21,731	8.3
State Total	34,501	155,959	221	$22,711	14.7

^aU.S. Census Bureau 2003, 2001 data.

^bU.S. Census Bureau 2003, 2000 data.

^cU.S. Census Bureau 2003, 1999 data.

Before entering the field, the research team planned the selection criteria for the informants based on organization type and level of responsibility. The organizations selected were the courts (14 separate interviews conducted), district attorneys’ offices (eight), public defenders’ offices (nine), alcohol and drug treatment administrations (11), probation departments (eight), and, to the extent possible, local drug treatment organizations (three), as well as, if not already included, the designated county Proposition 36 “lead” agencies (three). We tried to interview officials who were as high as possible in the organization yet also were directly involved in planning Proposition 36. In some but not all cases this was the agency head. To gain permission and access, the research team asked the lead agency head in each selected county to participate in the study. Thereafter the assigned senior researcher worked with a designated liaison in each county to draw up a list of potential respondents in the targeted organizations. All the suggested informants were contacted, and occasionally they proposed an alternative or additional informant. Of all the organizations contacted in the eight counties, only one department executive refused to participate or allow subordinates to do so. We conducted a total of 58 interviews.

The interviews were semistructured, with each informant asked a series of standard, open-ended questions designed for that type of agency representative. The interviews elicited the interviewees’ opinions on a number of general topics and ones specific to that agency, including the current relationships between local treatment and criminal justice agencies; the organizational structure, process, and funding for Proposition 36's implementation; its perceived goals and outcomes; changes being made in protocols regarding clients and services; anticipated client flows and changes in the system of care; and perceived challenges, conflicts, and unresolved issues. We told the respondents that their statements would be on the record unless they specifically chose to respond to a particular question “off the record.” The interviews were conducted in person by four senior researchers and lasted for 45 to 60 minutes. The researchers took notes during the interviews, and immediately afterward they refined and expanded them and, as soon as possible, printed them. The project was conducted under the auspices of the committee for the protection of human subjects of the Public Health Institute, the researchers’ institution.

Our analysis is also based on planning documents such as the minutes of statewide and county planning committee meetings, the counties’ plans for Proposition 36's implementation submitted to the state, presentations prepared for county boards of supervisors, and treatment and risk assessment instruments. We collected media reports of Proposition 36 in the state's major metropolitan newspapers and the primary newspapers in each of the eight counties, and newspaper reports of comparable developments in other states. The researchers also observed, and took notes on, state Proposition 36 committee meetings and conferences, and parallel local meetings.

Data Analysis Method

The data analysis for this project used a qualitative approach (Lincoln and Guba 1985; Lofland and Lofland 1995; Marshall and Rossman 1995). This method captures more detail than would, for example, statistical analyses of structured questionnaire responses, and can better inform future larger-scale or hypothesis-driven quantitative studies. Given that very little was known about how Proposition 36 was being implemented locally, this method was most suitable for capturing the details of local practices.

The written interview records, archival documents, and observational notes formed the raw data for our analysis, which we conducted using a more structured adaptation of the grounded-theory method (Glaser 1992; Glaser and Strauss 1967; Strauss and Corbin 1990). The senior researchers coded their own interviews and observational data for themes considered important to the analysis, using a list drawn from research questions and interview topics in the interview guides. Other recurring elements in the raw responses were later incorporated into this initial set of codes. Two outside consultants who were familiar with the interview guide reviewed these categorizations. The senior researchers then met to discuss their coded themes in an effort to reach consensus on the best “fit” of coded materials, potential aggregation or subdivision of codes, and additional data needing codes. Possible themes arising from archival data were also considered. One of the researchers drew up a second, refined set of codes and categorization, and all the researchers met to review both the initial categorization and the revised sets of codes and to agree on the analytic categories and the main patterns from the categorized material.

Findings

The findings presented here concern two primary goals of Proposition 36: expanding the state's capacity for public treatment and enhancing access to treatment for clients from the criminal justice system. The findings focus on both shared and differing trends observed across the counties that would be likely to affect the implementation of the law, and on possible contributing factors to the commonalities and differences in the patterns. We also point out future policy research directions suggested by the findings.

Expanding Treatment Capacity

During the lively debate over Proposition 36, proponents pointed to the scarcity of publicly funded treatments for drug arrestees, especially compared with the growing availability of incarceration, and they argued for the initiative's potential to increase treatment opportunities. The funds appropriated were largely intended to expand treatment capacity and eventually to reduce imprisonment and its rising costs (California Campaign for New Drug Policies 2002a, b). Hence, developing an adequate treatment capacity for the expected client flow was a key component of the initiative.

The three elements of planning for treatment capacity in the study counties were (1) expanding and funding total capacity, (2) deciding what types of care would be expanded, and (3) deciding what treatment models would be included.

California's existing public treatment system provided the context for understanding the expansion of capacity under Proposition 36. A decentralized and underdeveloped web of services had evolved under the constraints set by state-level funding and mandates, with management at the county level. Services were predominantly provided by nongovernmental organizations, mostly freestanding nonprofit entities specializing in alcohol and other drug abuse treatment. In October 2000, the recorded publicly funded drug treatment capacity in California was fewer than 80,000 “slots” (California Department of Alcohol and Drug Programs 2000), but the estimated need was for at least 50 percent more (Drug Abuse Research Center 2001). More than 150,000 clients were admitted each year (see Longshore et al. 2003). Annual public treatment expenditures were estimated at $550 million (Little Hoover Commission 2003). California was a typical state in relying on community-based providers and having a sizable unmet need (U.S. Substance Abuse and Mental Health Services Administration 2000). California was atypical, however, in its pronounced lack of centralized control over treatment, which added complexity to the planning and evaluation of Proposition 36.

These general characteristics of the state's treatment capacity on the eve of Proposition 36 were exemplified in the eight counties studied. Nearly all had contracts with several community-based providers that used their own treatment models; the various political relationships were complex; and the revenue streams were tortuous. Most of the larger counties (for county size, see Table 2) had attempted, often with mixed results, to establish a continuum of care, that is, to develop a range of treatment levels or intensities, in different settings and with separate elements or components, to serve populations with diverse needs and individuals at different stages of need. In every county, services had been shaped by multiple funding streams, including state-allocated federal block grants, public-fee reimbursements (e.g., Medicaid), and special programs (e.g., “drug courts”). These revenue streams gave counties considerable flexibility for both the long-term funding of providers and targeted program expansions (e.g., women's treatment). Capacity and coordination were limited by such constraints as low fee ceilings and narrow definitions of reimbursable services.

The eight counties differed greatly in service capacity. The first column of Table 3 shows the funding that the counties received from state and federal sources for treatment and prevention services on the eve of Proposition 36. At one extreme, urban San Francisco County received extremely large amounts of state and federal funds for substance abuse services relative to its population. In addition, it spent locally generated funds, including those committed under an ambitious initiative to achieve treatment “on demand.” San Francisco also shaped its mix of services rather than passively giving money to provider organizations to do so. For example, it requested and supported “harm reduction” activities in addition to abstinence-oriented treatment. While receiving a more average level of state and federal funding (Table 3) relative to its population (Table 2), the growing “high-tech” Santa Clara County also contributed county dollars to its treatment infrastructure and experimented with moving toward a centralized, data-driven, managed care model. Most of the other study counties spent only whatever state or federal funds they received (Table 3) and, except for the smallest counties, contracted for most services with community providers, often without standardized county requirements. In impoverished and sparsely populated Imperial County (Table 2), the dollars available for publicly funded services were so limited that there were no community-based providers; hence the county itself provided nearly all services.

TABLE 3.

County Proposition 36 Funding and Budgeting

County	Total State ADP AOD Funding, Fiscal Year 2001–02 (in 1,000s)a	State Proposition 36 Funding, Fiscal Year 2001–02 (in 1,000s)b	Percentage of Prop. 36 Funds Budgeted for Treatment- Related Activitiesc	Percentage of Prop. 36 Funds Budgeted for Direct Treatment Servicesd
Alameda	$13,815	$6,654	85	36
Fresno	$8,098	$4,128	75	43
Imperial	$1,498	$878	77	62
Orange	$22,709	$11,083e	73/89f	61/79f
Sacramento	$10,329	$5,083	61	37
San Francisco	$14,907	$6,671	95	78
Santa Clara	$14,702	$6,741	77	33
Solano	$3,391	$1,818	90	88

^aCalifornia Department of Alcohol and Drug Programs 2000. Includes California Department of Alcohol and Drug Programs (ADP) budgeted federal and state funds for alcohol and other drug (AOD) treatment and prevention for the counties for fiscal year 2001–02 excluding Proposition 36.

^bCalifornia Department of Alcohol and Drug Programs 2001. Includes the counties’ estimated unexpended 2000–01 funds to be “rolled over” and expended in 2001–02. Alameda County: Version 1, ADP review pending. Solano, Imperial, and Fresno Counties: Version 1, ADP approved. Sacramento and Orange Counties: Version 2, ADP approved. San Francisco and Santa Clara Counties: Version 3, ADP approved.

^cIncludes all Proposition 36 activities except criminal justice (probation, courts, district attorney, public defender, sheriff) and countywide administration. Also see Ford and Smith 2001.

^dExcludes funds allocated to assessment and referral, case management, program administration and overhead, capital construction, ancillary literacy and vocational and counseling services, housing, client tracking, and program evaluation.

^eOnly $8,509,000 (77 percent of available state Proposition 36 funds) was budgeted.

^fThe first percentage is the allocation as a percentage of available funds; the second percentage is the allocation of budgeted funds.

Total Capacity: Planning and Funding for Growth

The state gave the counties guidelines to plan for Proposition 36. The state's lead agency developed broad administrative regulations to implement the initiative and required each county to submit its specific plan for state review. Each county was asked to project the types and costs of Proposition 36–related services for the first year in order to receive its money (for a summary of features of all county plans, see Ford and Smith 2001). The state also made general suggestions about making plans conform to the spirit of the law by focusing on the expansion of treatment rather than criminal justice. This state leadership, itself prompted by the law's political visibility, was pivotal in ensuring that the counties concentrated on improving their systems of care.

All eight counties planned to devote a majority of their new funding under Proposition 36 to supporting treatment-related activities, as opposed to expanding other activities such as criminal justice or general county coordination. As shown in the third column of Table 3, in seven of the eight study counties, as reported in the counties’ plans, 75 percent or more of the funds were to go to treatment-related activities in the first year of implementation. Only one county planned to dedicate considerably less, 61 percent, to treatment-related activities, and this county was publicly criticized by some Proposition 36 advocates. (“Treatment-related” activities, as reported in the county plans, were defined as all activities other than adjudication, prosecution, defense, probation, or countywide administration.) Many of our informants pointed out this treatment-oriented allocation pattern, attributing it to the law's clear intent and the strong statewide voter support. Many also commented how additional state-level developments, such as other counties’ allocations and the public reactions of Proposition 36 advocates, influenced their county's decision to allocate funds to treatment-related activities. Most respondents also expressed concern that their county might not have sufficient capacity to serve the expected additional client flow, especially by the second or third year of implementation.

Counties where voter support for Proposition 36 had been strong (see Table 1) tended to allocate a higher proportion of funds to treatment-related activities and to assign more client assessment and management functions to treatment agencies, as shown in Table 3. But there were exceptions to this pattern. The choice of local lead agency also seemed to affect allocation decisions. Those counties in which the lead agency was health care tended to allocate more money to treatment-related activities, but again there were exceptions to the pattern (see Tables 1 and 3). Counties also varied widely in the amount of funding to be devoted to direct treatment services (a category that excluded treatment-related administrative or ancillary activities; see the last column of Table 3).

Respondents from nearly all counties reported that Proposition 36 would have a sizable effect on their county's total treatment capacity. This was most pronounced in very small jurisdictions such as Imperial, where the $878,000 in new funds constituted an increase of more than 50 percent in what had been received annually for all public substance abuse services (see Table 3). Proposition 36 was expected to expand the county's total treatment capacity by more than 75 percent. The total treatment capacity of the five medium and large counties was budgeted to increase from between 20 and 75 percent. Even in San Francisco, whose local planners regarded the allotment of more than $6 million from Proposition 36 as relatively small in relation to its total treatment budget, these funds would still noticeably expand capacity.

Planning Levels of Care: Balancing Anticipated Needs and Existing Services

In planning for Proposition 36, counties also had to grapple with whether and how to change the mix of treatment types and levels. The counties’ plans for the first year of implementation reflected a balancing act between anticipating new treatment needs that might arise under Proposition 36 and relying on existing services to meet those needs, primarily outpatient care.

Proposition 36 represented a turning point in California's drug treatment not only because of the additional funding but also because of the state's decision to incorporate individual assessments of clients in the administrative regulations (California Department of Alcohol and Drug Programs 2002). Thus the counties wanted to create services that would accommodate clients’ individual needs based on initial screenings or assessments. This focus on the individual set Proposition 36 apart from the “one size fits all” treatment programs commonly used for criminal justice clients.

California's commitment to an assessment-driven placement of clients posed an immediate planning challenge. Little information about anticipated clients’ service needs was available for use in forecasting types or levels of care under Proposition 36. Among the eight study counties, only Santa Clara was already using centralized initial screening and placement for treatment admissions, and Sacramento was screening clients entering social services. Proposition 36 provided an impetus for the future collection and use of such data in planning treatment. For example, Solano County was considering the central screening of all clients admitted to treatment, not just those falling under Proposition 36.

In the short run, nearly all counties planned their largest expansion around medium-intensity outpatient treatment, centered on group counseling typically offered at least twice a week, with more limited (e.g., monthly) individual counseling. In most instances, counties planned a smaller expansion of the highest end of the continuum of care (i.e., the most intensive or specialized services) and just a slightly larger expansion of the least intensive or “early intervention” services.

Fresno County, for example, planned to offer outpatient treatment to individuals for nine months, supplemented by mandatory frequent “Twelve Step” self-help support meetings. For budget-planning purposes, nearly 75 percent of clients were expected to be assessed as eligible for this level of care. The other 25 percent would receive a less intensive 90-day course of weekly “drug education.” The projected small number of substance-dependent or very high-needs clients would be given residential treatment beds or “clean and sober” housing in conjunction with outpatient treatment. Special care such as methadone detoxification for opiate addicts, services for pregnant and parenting women, and “dual” treatment for persons also diagnosed with mental health disorders would be provided as well. For all but the drug education clients, six months of aftercare would be offered, composed of weekly self-help groups.

In most other counties, the basic planned capacity types were comparable to Fresno's, with the largest group of clients projected to require outpatient treatment, although the specifics varied. Imperial County developed a conceptual plan based on phases as well as levels of care, with all clients expected to progress through the phases in 12 months or less, from individually assessed starting points. Imperial County forecast a relatively higher-end client need than did most other counties: Nearly 50 percent were expected to begin their treatment phases at the intensive outpatient or residential levels, and 33 percent at the standard outpatient level; about 20 percent were estimated to need only drug education. Sacramento, by contrast, delineated three levels of care, with the largest planned expansion at the lowest end, the 90-day drug education course. The county anticipated assigning 70 percent of its clients here. Mid-level outpatient care was foreseen as needed by only 20 percent, and high-end residential or special services such as detoxification, by 10 percent.

According to our respondents, the common reliance on outpatient treatment under Proposition 36 stemmed from its relative availability and the ease of expanding or reshaping it to meet new needs. Only in Imperial County did such publicly funded outpatient care have to be newly created. Everyone agreed that starting or extending more intensive or specialized services (especially residential care) would be far more difficult and costly.

There were widespread local concerns about relying on standard outpatient programs as the backbone of Proposition 36 treatment, and there were hopes of bolstering other types of capacity where possible. Many respondents doubted that the outpatient programs’ intensity of contact was sufficient for clients assessed as severe or chronic substance abusers. Others worried that standard outpatient care would fail for those clients needing supportive services such as housing and employment assistance. Criminal justice officials were skeptical about the outpatient programs’ ability to retain and monitor unmotivated clients at high risk for noncompliance. For all these reasons, most counties wanted to expand residential capacity, especially for short-term (e.g., 90-day) stays. As a cheaper and easier alternative, counties also planned to add “clean and sober” housing or peer-run “recovery homes” to supplement their outpatient treatment. These facilities would presumably enhance clients’ access to more intensive treatment contact, social services, and behavioral monitoring.

Planning Models of Care: A Limited Range

Nearly all counties planned to rely heavily on peer-oriented, abstinence-focused treatment models. In nearly every instance, Twelve Step meeting attendance was planned as the aftercare provided under Proposition 36, and required attendance at meetings of Alcoholics Anonymous, Narcotics Anonymous, or Twelve Step study was incorporated into most planned outpatient and residential programs. This continued the existing practice: In the eight counties, as throughout the state, services in most programs were based on the “social model” of recovery through nonclinical, peer-oriented, mutual self-help (Shaw and Borkman 1990). This paradigm was predominant for fiscal and philosophical reasons: Funding constraints, shortages of professional medical staff, and the beliefs of “recovering” staff had limited the public availability of other models of care, such as medical detoxification, inpatient treatment, individual psychotherapy, and pharmacotherapy. An important exception was San Francisco, which had endorsed a “harm reduction” model and sponsored medical and pharmacological approaches that did not focus solely on abstinence.

Debates over whether to include methadone maintenance for opiate addicts in the planned treatment continuums suggested the limited range of models under Proposition 36. Methadone maintenance was included in only half the study counties: Imperial, Orange, Sacramento, and San Francisco. Nearly all these counties planned to offer methadone to clients for a very short time, in some instances really comparable to long-term methadone detoxification. Although our respondents explained this decision as necessary to meet Proposition 36's one-year time frame for treatment, this decision was also consistent with the limited publicly funded methadone maintenance that the counties already offered. In nearly all counties, strong philosophical opposition to methadone maintenance had long hindered its development as part of overall treatment capacity. No county except San Francisco was even discussing other, less well-established pharmacotherapies.

Many study respondents expressed concern about the adequacy of local approaches. Several respondents predicted—some fearfully—that Proposition 36 would lead to public awareness of the limited effectiveness of current treatment models. This concern was expressed by both individuals who focused on treatment's ability to induce abstinence from illegal drug use and those who defined the goals of treatment more broadly. We will revisit this issue of treatment effectiveness later.

Future Issues in Capacity Expansion under Proposition 36

Proposition 36's requirement for treatment in lieu of incarceration can be seen as part of the national-level movement by health advocates for a greater public substance abuse treatment capacity. The emerging consensus is that considerable new public resources will be required to meet the needs of persons referred for treatment by the criminal justice system (on California, see California Legislative Analyst's Office 2001; and Howard 2002). One national prison and jail survey estimates that based on known offenses and self-reported histories, 70 percent of prison and jail inmates need substance abuse treatment (Belenko 2002b).

It will be important to follow Proposition 36's impact on treatment capacity in the state over time as state and local budget difficulties wax and wane. It also will be useful to track different counties’ changing capacities in light of their varying allocations of funds to treatment, treatment-related functions, and other activities. It will be useful to evaluate potential revenue streams and reimbursement mechanisms, both inside and outside this particular law, which might be used for new capacity. Some counties that projected sufficient treatment dollars for the first year predicted shortfalls by the end of the second year. Many advocates of the law's approach had hoped that lower imprisonment costs might provide a source of funds for additional community-based treatment capacity for persons who otherwise would be incarcerated (see, e.g., McVay, Schiraldi, and Ziedenberg 2004). But there is currently no way to identify and reallocate such savings. Furthermore, it is not yet clear whether Proposition 36 will greatly reduce overall imprisonment levels (see, e.g., the projections in Auerhahn 2004).

Capacity at higher levels of care requires ongoing examination. It already appears that the need for high-end treatment capacity is considerably greater under Proposition 36 than had been planned (Hser et al. 2003), with a national survey showing that the majority of inmates require intensive care (Belenko 2002b). Yet there is an endemic scarcity of intensive treatment. It will be important to see whether California's law spurs the development of such innovations as day treatment coupled with housing and support services tailored to clients’ needs, which might serve as an alternative to residential or inpatient care (see, e.g., Bedrick 1997). Successes and failures among the counties may contain additional lessons.

Finally, the appropriateness and quality of care under the initiative should be studied. There are indications that Proposition 36 might be strengthening the expectations that individual assessments will determine client placements and that data will be aggregated for program monitoring (see the report by Hser et al. 2003). An important indicator of Proposition 36's impact will be whether local treatment capacity is reshaped over the several years to meet clients’ needs, based on aggregated assessments, or whether the heretofore-prevalent customs of using existing capacity regardless of need and neglecting to use assessment data will persist. The counties’ actions to make their systems of care more suitable for their populations should be studied.

Quality improvement is needed in publicly supported alcohol and drug treatment (Fox, Egertson, and Leshner 1997; Join Together Online 2001a, b; Nelson 1996). While a full discussion of quality of treatment is beyond the scope of this article, we note the widespread reliance on a limited peer-support abstinence model. Although Proposition 36 was not explicitly aimed at enhancing the effectiveness of care, it may end up doing so (Bridging Research and Services 2001). The law's requirement that participating programs be certified or licensed by the state now involves only a minimal procedure, but it could open the door to meaningful statewide or federal standards. More immediately, the mandated statewide evaluation of Proposition 36 may help improve the quality of treatment. If the program is found to be ineffective, legislators are unlikely to fund the initiative beyond its five years.

Enhancing Access to Treatment

The adequacy of existing routes to community-based treatment for persons arrested for drug possession in California was a prominent subject of debate before the Proposition 36 ballot. A major goal of Proposition 36 was easing access to community-based treatment for individuals arrested for using drugs.

We examine three aspects of planning for how individuals would gain access to treatment under the initiative: (1) designing initial pathways to ensure treatment, (2) placing clients in a level of care based on their needs rather than on their risk to others, and (3) maintaining control over continuing access to treatment.

The previously existing pathways and roadblocks to community-based treatment provide a context for considering Proposition 36's potential impact on access to treatment for persons convicted of drug possession. These routes were built by criminal justice officials wanting to maximize the efficiency of the courts and the ability of the correctional system to control defendants. Indeed, most opponents of Proposition 36, which included numerous law enforcement associations, district attorneys, and judges, argued that the initiative would change the existing pathways to treatment that provided needed coercion and control of individuals ordered to treatment by the courts (Californians United Against Drug Abuse 2002a, b). They pointed specifically to the power of prosecutors or the courts to find defendants ineligible for court-ordered treatment or to remove individuals from court-ordered treatment and incarcerate them. The opponents contended that these discretionary powers were safeguards that would be threatened by Proposition 36. They also expressed fear that Proposition 36 might slow the handling of cases, which in turn would delay treatment.

In California, as elsewhere, for some years defendants could be referred to community-based treatment by the courts, most commonly as a condition of receiving probation. Persons charged with first-time drug possession could be “diverted” from a criminal conviction if they agreed to participate in an intervention, per California Penal Code 1000. In most counties this consisted of brief drug education or required Twelve Step participation. While such diversion had been widely used, many of our respondents stated that it had been unsuccessful, because the mandated interventions were minimal and the rates of failing to comply with them were high. This mechanism therefore apparently did not ensure that drug defendants would receive treatment. The respondents expected that Proposition 36 would provide a second chance for the large numbers who failed after diversion to receive treatment.

A second previous route to treatment for defendants was the more recent innovation of “drug courts” (Belenko 1999, 2002a). Like Proposition 36, drug courts were intended for individuals who were convicted of drug charges and willing to enter treatment in order to avoid imprisonment. Drug courts unquestionably provided access to treatment for those who participated, but the numbers served were few. Unlike diversion or Proposition 36, drug court usually involved selective gatekeeping, permitting broad prosecutorial and judicial discretion in rejecting defendants or quickly incarcerating those who failed to comply. Furthermore, a substantial effort by both defendants and court officials was usually required, such as numerous court dates and drug tests. Our informants predicted that defendants eligible for both Proposition 36 and drug court would choose the former, and some suggested that drug court be redesigned as a backup treatment opportunity for defendants ineligible for or failing under Proposition 36. (On Proposition 36 versus drug court in California, see Wolf 2004.)

Before Proposition 36, counties in California differed in the degree to which persons arrested for drug use had access to community-based treatment through the courts. Differences in philosophical beliefs and organizational arrangements, as were common in local justice systems in many states (Church 1982; Eisenstein, Flemming, and Nardulli 1988), influenced local procedures. Before Proposition 36, defendants facing identical drug possession charges and with identical criminal histories might routinely be offered probation with treatment in one county but sentenced to state prison in another (for variations in drug imprisonment rates in California counties, see Males, Macallair, and Jamison 2002).

All eight counties in the study used diversion and had at least one drug court, but the differences among them were likely to affect Proposition 36's impact on the defendants’ access to treatment. In San Francisco, which had an unusual “pre-plea” drug court that functioned like a diversion program to encourage defendants to enter treatment, informants expected that Proposition 36 might not dramatically improve arrestees’ access to treatment. Alameda County's largest drug court had an unusually high caseload but offered minimal treatment. Officials here hoped that most defendants would enter Proposition 36. Some drug courts, such as Santa Clara's, were held up as criminal justice models of intensive control and treatment, and their officials feared losing visibility under Proposition 36. Other drug courts, such as Fresno's, appeared willing to change in order to implement Proposition 36.

Designing an Initial Pathway

The initial pathways planned by counties for clients under Proposition 36 would ensure relatively generous eligibility for, and prompt access to, treatment. First, most prosecutors developed protocols that closely followed the letter of the law in setting broad legal eligibility (see Kennedy 2001). District attorneys who had opposed Proposition 36 reported to us that they interpreted the statewide vote as a clear mandate for expanding drug defendants’ access to treatment. In fact, several intended to adopt criteria flexible enough to include minor offenders falling into “gray areas” of strict legal eligibility who, in their view, would benefit from treatment. Public defenders uniformly and strongly supported Proposition 36 and said that they intended to advise most of their eligible clients to participate.

A second aspect of planning the initial access to treatment was swiftness. Speedy referral to care has been shown to make it more likely that persons ordered by the court to report to a treatment program will actually do so (Belenko 1999, 2002a). However, the legal system normally creates delays because of postponed hearings and adversarial arguments. Other county agencies in both the justice and health systems have also been characterized by long waits for service. The study found that Proposition 36 spurred the counties to develop streamlined procedures, particularly in time lines for convicted offenders to be assessed for their need of treatment. Each county created interagency protocols and allocated resources among the prosecutors, courts, probation departments, assessment agencies, and treatment organizations. Across the counties, the initial pathways to treatment under Proposition 36 were planned as concurrent or quickly sequential processes: determination of legal eligibility by the prosecutor, defense motions and pleas, court conviction, probation risk assessment, treatment need assessment, and placement in a program.

Notwithstanding the generous eligibility and swiftness of the planned initial access to treatment that we observed across the counties, the degree of the processing sites’ co-location and the centralization of the initial procedures differed. Each county already had courts, probation offices, and treatment agencies in various locations. In planning sites and procedures, local decision makers grappled with weighing clients’ accessibility, efficiently using resources, and offering “command and control” functions. In some instances, they chose decentralized sites for convenience or better service; in other instances, they chose to create a “one-stop shop.”

Fresno and Alameda counties planned particularly centralized initial procedures for managing clients under Proposition 36. Fresno, which already had a single consolidated courthouse, assigned all Proposition 36 defendants to the same courtroom, prosecutors, and defense attorneys, so there would be maximum uniformity of legal procedure. Furthermore, each defendant in Fresno would be screened for both treatment need and safety risk at one facility, jointly run by the health care and probation departments. In Alameda County, defendants were to appear in one of two major regional courthouses and be assessed by the treatment staff located there. In contrast, Imperial County planned to continue its decentralized adjudication of defendants in the small local branch courts that served large, sparsely populated regions. And Orange and Solano Counties planned for clients to receive their treatment need and safety assessments at separate offices. While “one-stop shops” appeared to offer the considerable benefits of client convenience and staff coordination, some respondents observed that co-locating treatment need and probation assessment staff might give clients the impression that treatment was part of the process of punishment rather than an independent service.

Most counties planned for clients to obtain treatment by presenting for care at existing or expanded programs. The more populous or urban counties generally planned numerous treatment sites, although Sacramento County planned to concentrate treatment in only three locations. In the less populous or less densely populated counties, such as Imperial and Fresno, most programs were in the county seat. Again, the different degrees and patterns of centralization suggest the complexity of balancing proximity, choice, efficiency, and oversight.

Placement in Needs-Driven Levels of Care

A second key issue in planning access to treatment was placing clients in a setting and a service appropriate to their needs. State regulations directed that the treatment needs of each Proposition 36 client be assessed before the client was referred for treatment. Most counties planned to conduct in-person assessments using a standardized substance abuse–related screening instrument. Specialized staff were to perform this function under the county's auspices.

This initial treatment need assessment could increase the likelihood that clients would be placed in a level of care based on their need rather than their criminal record. This represents an important break from earlier methods of determining what kind of care drug defendants would receive. Customarily, decisions about placement were influenced by judges, prosecutors, and probation officers whose main consideration had been the defendants’ perceived risk to public safety, often measured in terms of personal instability or criminal history. These officials often demanded a “secure” placement, such as a residential setting, as an alternative to incarcerating the defendant. By contrast, placement based on the client's assessed treatment need would use residential care only when more intensive substance abuse treatment was indicated. Strong tensions of the sort that had arisen in preelectoral debates about Proposition 36 emerged in negotiations between court agencies and health care officials over planning the appropriate referral for clients.

The counties chose different strategies for determining the placement of clients in care. In Alameda, Orange, and Santa Clara Counties, health care respondents reported arguing—in some instances, successfully—with reluctant judges and prosecutors to allow treatment need assessors to determine placement based on need. In some counties with strong criminal justice stakeholders, or with weak public support for Proposition 36, some respondents expected that placement decisions might continue to be influenced by these stakeholders’ criteria for placement.

Structuring Continuing Access to Treatment: Who Controls and How?

A third access issue was the clients’ continuing ability to secure treatment while under the jurisdiction of Proposition 36. This included their right to seek a new referral if the initial treatment referral was inappropriate or if they were failing to comply with a program's requirements, and the right to remain in treatment even if they were found to be continuing to use illicit drugs.

A controversial provision of Proposition 36 was that clients who were not succeeding in their initial placement or who were continuing to use drugs be referred again to treatment rather than be incarcerated. Two re-referrals were to be required before a client could be removed from Proposition 36 by the court and resentenced under existing legal guidelines. The initiative's supporters successfully argued that because addiction is a chronic medical condition characterized by relapse, re-referral to care is an appropriate and cost-effective management approach (Institute of Medicine 1990; McLellan 2001). Opponents had contended that treatment should be a revocable legal privilege, a “carrot” that required the “stick” of threatened incarceration. (For the argument for the effectiveness of such threats, see Anglin and Hser 1990; and Farabee, Prendergast, and Anglin 1998.) Many treatment providers relied on the threat of penal sanctions to bolster their clients’ retention rates. All sides agreed that most Proposition 36 clients would continue to use drugs at least sporadically, that many would not adhere to all program requirements, and that some might not complete the initial planned course of treatment. Nevertheless, the initiative's clear legal intent was to reduce the use of incarceration to control clients. Thus the battle over managing client compliance and re-referral under Proposition 36 was expected to be fierce.

Our respondents in all study counties agreed that Proposition 36 would make it more difficult for agencies to use jail or prison to control their clients and thus would require new approaches. Nearly all prosecution, court, and probation officials said, however, that they intended to observe the law scrupulously and allow their clients at least “three bites of the apple” to continue treatment in the event of unfavorable drug test results or other problems.

Differences were found in counties’ planned client case management procedures. In San Francisco and Alameda Counties, client management was assigned to the treatment function. This decision might be expected in the two counties with strong votes for Proposition 36, health care serving as lead agencies, and large treatment-related allocations (see Tables 1 and 3). By contrast, in Sacramento, Solano, Fresno, and Orange Counties, probation officials would supervise clients and make recommendations for treatment re-referral to the court. This was not unexpected in the two counties, Solano and Sacramento, where probation departments served as the lead or co-lead agency, or in Fresno County, where there had been weak voter support for Proposition 36 (Table 1). Imperial County devised a third choice in case management, designating an interagency health and probation team to review periodically their clients’ treatment plans and probation reports for potential re-referral at the end of each treatment phase. Under Santa Clara County's plan, clients whose needs were judged to be high would receive special case management from contracted treatment organizations, and those assessed as being at high public safety risk would receive more intensive probation supervision.

The counties’ contrasting plans regarding methods and sites of client control and re-referral could affect their clients’ continued access to treatment. Many health officials, client defense advocates, and Proposition 36 supporters were afraid that greater involvement of probation officers in supervision could result in the quick labeling of clients as failures and be dismissed from Proposition 36. We address the merits of this argument in the following section.

Future Issues in Access to Treatment under Proposition 36

Our findings suggest important points for research and policy regarding access to drug treatment for criminal justice clients. First, to see whether the rules on legal eligibility are too open to varying interpretation, it will be important to examine whether and how discretion in the criminal justice system influences defendants’ initial access to treatment under Proposition 36. Individuals charged with drug possession are commonly charged with other offenses as well, which may render them ineligible for Proposition 36. Although prosecutors told us that they might be willing to drop some minor disqualifying charges, it is not clear whether they would consider reducing the common disqualifying charge of drug possession for sale to the charge of simple drug possession in order to make defendants eligible for Proposition 36 (for official guidelines, see California District Attorneys Association 2004). Another potential legal barrier that should be monitored is whether district attorneys or judges are not approving eligible defendants who have earlier convictions for violent offenses, despite Proposition 36's requirements that they approve them unless the conviction is recent. Overall, the implementation of Proposition 36 by prosecutors historically favorable to punishment, and by judges and probation officers partial to determining placements themselves, can be expected to be decisive in influencing whether drug offenders gain access to treatment under Proposition 36. Evaluators and policymakers should examine the numbers and proportions of potentially eligible arrestees who are actually found to be eligible for Proposition 36, both statewide and across counties with different philosophical approaches and organizational relations.

It will be important to determine how counties’ varying logistics, such as the schedules and locations of appointments and services, encourage or discourage clients from appearing for treatment in a timely way. Tracking variations in the rates and timing of clients’ appearances over time by locality could help the entire alcohol and drug treatment field identify and replicate the most effective procedures for receiving clients from the criminal justice system. The first-year statewide Proposition 36 evaluation found that the initial entry rates into treatment were positively associated with service co-location and fewer required appointments (Longshore et al. 2003). A subsequent, more detailed evaluation in one major county found that despite local efforts to streamline the path from conviction to treatment, clients continued to be at the highest risk for reoffending in the brief period between assessment of need and entry into treatment (Wiley et al. 2004).

Whether participants have access to placement based on their treatment needs will also merit tracking throughout the years of Proposition 36. Statewide and local evaluators and policymakers should decide whether placements were appropriate in light of the clients’ assessed treatment needs, whether they were not intensive enough (e.g., owing to budget difficulties eroding capacity at the higher levels of care), or whether in some instances they were too intensive (e.g., owing to these defendants’ assessed safety risk). Such data on the appropriateness of placements could help interpret the eventual results of the ongoing statewide evaluation of client success rates under Proposition 36, since appropriate treatment is likely to be a crucial contributing factor in the treatment's outcome. County differences in the appropriateness of treatment placements may help explain any eventual county differences in treatment outcomes.

Last, an analysis of rates of clients’ violations and terminations from Proposition 36 should shed light on the strengths and weaknesses of the law's re-referral provisions and the counties’ different re-referral practices. It is possible that some counties—for example, those whose probation departments manage clients—may increasingly sentence to prison those clients who fail their treatment. Alternatively, supervision that is oriented toward rehabilitation or coordinated with health care could result in the quick identification and re-referral of faltering clients. Furthermore, no monitoring of clients’ progress could result in their invisibly “blowing” their three legal chances for alternatives to incarceration. Perhaps the only safe prediction is that the continuing access to treatment intended under Proposition 36 will require dedicated resources, close attention, and creative incentives.

Conclusions

Our planning data from eight counties implementing Proposition 36 contain important questions concerning policy actions to increase the viability of substance abuse treatment as a large-scale intervention for persons arrested for illegal drugs. This concluding discussion touches on some broader scientific and policy issues for future initiatives of this kind in the United States.

Our analysis focused on decisions affecting treatment capacity and client access under Proposition 36. But this exploratory study is limited in that some data may prove to be unrepresentative or some key issues unidentified or misinterpreted. However, the material does present a baseline and identifies the variations in the implementation of Proposition 36 that may well influence its effectiveness.

The planning discussions and decisions that we studied reveal underlying policy tensions spanning implementation issues such as the proportion of funding to allocate for treatment, whether to include programs such as methadone maintenance in the continuum of care, which defendants to make eligible for Proposition 36, and how to respond to clients’ noncompliance with treatment. In these exchanges over the benefits and drawbacks of various resource allocations, procedures, and outcomes, we have concluded that Proposition 36 is opening important, previously concealed fault lines between public health and criminal justice approaches to policy (see, e.g., Szalavitz 2001). The debates in California revealed basic differences over how to define the problem being addressed by, and the appropriate expectations for, an initiative such as Proposition 36.

There are many possible ways to define the goals of a policy action such as Proposition 36, and hence there are different measures by which to evaluate its impact. In recent years, scientific evaluations of treatment initiatives for criminal justice clients have often concentrated nearly exclusively on client outcomes: whether the “treatment works” using the strict yardsticks of abstinence from illegal drug use and absence of re-arrest. These fairly narrow goals reflect the legacy of the domination of criminal justice approaches to substance abuse treatment. The argument often takes the form of criminal justice advocates asking whether treatment without the threat of incarceration would result in acceptable rates of participation, completion, and absence of relapse. In response, proponents of public health approaches contend that medically acceptable standards for clients’ compliance with treatment (McLellan 2001) are substantially lower than most current legal expectations of clients’ compliance. Many study respondents expressed the view—some with fear, others with hope—that clients’ outcomes in treatment may be held to a very high standard by legal officials and the public in judging the initiative's success. In this regard, Proposition 36's supporters, including some study respondents, note that the imperfect effectiveness of current treatment is often cited but rarely compared with the effectiveness of punitive interventions, such as prison. They also note that a health-oriented approach to client outcomes would include objectives of improved functioning or reduced incidence of problems related to substance use among clients, in addition to their abstaining from illegal activities. In a public health model, the evaluation would also focus on such societal impacts as advancing treatment capacity and quality, reducing incarceration, and generating public cost savings, in addition to individual outcomes.

There are also scientific and policy questions regarding how to improve the effectiveness of treatment for illicit drug users. Health services research suggests that a treatment's effectiveness—whether something “works” for its clients—is highly dependent on the quality of care and the patient's satisfaction. But many evaluations of public substance abuse treatment reflect the criminal justice framework and focus less on improving treatment and more on the contributory role of sanctions (see, e.g., Anglin and Hser 1990; and Farabee, Prendergast, and Anglin 1998). This belief in the possible therapeutic value of penal deterrence has meant that increases in treatment capacity and access have tended to be accompanied by a deepening criminalization and coercion of substance abusers (Hoffman 2000; Klein 1983; Noble 1991; Speiglman 1994, 1997).

A beneficial consequence of Proposition 36 in its early months was the rare public debate that it sparked over the relationship between drug abuse treatment and criminal justice. Previous treatment programs for arrested clients, such as diversion and drug court, generally grew invisibly inside the specialized provinces of agency administrators and experts, rarely gaining media attention or public awareness. Proposition 36 created noteworthy new treatment capacity, unprecedented access to treatment for persons convicted of drug possession, dedicated funding, and mandatory evaluation of client success rates and overall costs. The scope of the experiment may lay the groundwork for the continuing debate over the broad goals of drug policy.

A stable policy shift to incorporate public health goals into an arena previously dominated by legal concerns is by no means assured, however. Political feasibility dictated that under the initiative, individuals must proceed through the adjudication and conviction process and be under legal supervision before and while they obtain treatment. This requirement ensures the continuation of the criminal justice system's considerable control over clients’ access to treatment and, to a lesser extent, its influence on the resources and types of treatment. To date, public support for Proposition 36 among Californians has remained strong, with large majorities indicating in a recent survey that they would vote for it again (National Council on Crime and Delinquency 2004).

Proposition 36 has offered an opportunity for ideas from public health to reshape policies regarding illicit drug users. Until now, the illegal status of the drugs, defined as public safety threats by criminal codes, has made it difficult to bring to the fore the pros and cons of various approaches to obtaining public health benefits. Proposition 36 modestly shifts the foundation by incorporating the goals of expanding the capacity of and access to treatment into the language of the law itself. This article identified a number of issues regarding treatment capacity and access to care that may be illuminated by further research into societal benefits and costs. In a large and heterogeneous state such as California, many questions about processes and services are enriched by an examination of the local contexts and variations. Research can identify specific successes and failures and the factors involved, thus expanding and refining the choices presented to the public. Those engaged in drug-related policymaking and research should give priority to investigating what will be required for an experiment such as Proposition 36 to succeed.