Earmarked Taxes as a Policy Strategy to Increase Funding for Behavioral Health Services

Using publicly available information, we compiled details regarding the authorizing legislation, date of passage, design, revenue, and fiscal oversight of the earmarked behavioral health tax laws in California and Washington. As shown in Table 1, the taxes share only two broad features. First, both were signed into law in 2005. Second, both were passed with the overarching purpose of increasing funding for behavioral health services.

The two states differ in terms of the design, spending, and oversight of these earmarked taxes. Specifically, the state of California imposes an additional 1% income tax on the taxable portion of annual household incomes exceeding $1 million to furnish revenue in the state Mental Health Services Fund. In contrast, Washington permits each county (and city with a population of >30,000 in a county with a population of >800,000) to opt into the tax, which imposes a 0.1% sales tax increase to furnish revenue to county governments for behavioral health services. As of 2017, 28 of the 40 counties in Washington State and the city of Tacoma had opted to implement the tax. The differences between the two laws in tax structure are reflected in per capita revenue, with California generating more than double the per capita revenue compared with Washington ($56.50 versus $22.17).

Another striking difference between the taxes in California and Washington is the degree of flexibility related to behavioral health tax spending and fiscal oversight. In California, revenue from the tax is allocated to five mental health service components: community services and support, prevention and early intervention, capital facilities and technological needs, workforce education and training, and innovation. Each county is required to regularly submit spending reports to a statewide Mental Health Oversight and Accountability Commission, which oversees spending. However, counties have some flexibility and autonomy to make decisions about the specific mental health interventions across the five components. For example, the Los Angeles County Department of Mental Health uses tax revenue in the areas of prevention and early intervention to provide county mental health centers with funding and implementation support to train providers to deliver specified evidence-based treatments (4). All counties in California are required to ensure that decisions about tax spending are consumer engaged and reflect representation of diverse stakeholders (e.g., consumers of behavioral health services, caregivers, families).

In Washington, in contrast, each county or city that opts into the tax is offered tremendous tax spending discretion as long as the revenue is dedicated to “treatment services, case management, transportation, and housing that are a component of a coordinated chemical dependency or mental health treatment program or services.” The only service mandate is that implementing jurisdictions must establish and operate a therapeutic court for substance use disorder proceedings. Tax dollars can be used only to expand and implement new services, although funding can be used to support services for which federal funding has ceased. As a result of the discretion permitted by the authorizing legislation, the specific services funded through the tax appear to vary widely between counties. Implementing counties are not required to report how they spend tax dollars, and there is no requirement for fiscal oversight at the state level. However, some counties that have implemented the tax—such as King and Thurston—require, and make publicly available, detailed reports about tax spending.

Indicators of the impact of earmarked taxes can be assessed at multiple levels. These include the amount of revenue generated by the tax (policy level), the number of providers billing for services funded by the tax (system level), the number of providers offering evidence-based treatments through revenue from the tax (system level), and improvements in clinical outcomes and client or provider satisfaction linked to services funded by the tax (client and provider level). Although these indicators of tax impact are likely to be of great interest to policy makers and their constituents, it should be noted that they are more accurately characterized as outputs rather than outcomes. Determining the outcomes of earmarked taxes for behavioral health services requires comparison of what happens after the tax is implemented and the counterfactual scenario of what would have happened if the tax was not implemented. This can be achieved through quasi-experimental difference-in-difference designs. For example, in California, it would be possible to assess the effects of the tax by comparing the magnitude of changes in outcomes at behavioral health service settings that received revenue from the earmarked taxes versus similar settings in the same county that did not receive the tax revenue. In Washington, it would be possible to assess changes between similar counties that did and did not implement the tax.

In California, a small number of studies have assessed the impact of the tax in Los Angeles County. A mixed-method, quasi-experimental study assessed impacts of the tax in the county by comparing public mental health clinics with tax-funded “full-service partnerships” to clinics without these tax-funded services (5). The study found that the tax-funded services were associated with higher rates of service utilization, greater provision of recovery-oriented services, and improved client-provider working alliances. However, the tax-funded services were associated with higher levels of provider stress because providers were required to change practices to offer more coordinated and patient-centered care.

A RAND analysis of claims data assessed the reach of tax-funded services in Los Angeles (6). Among the main findings, tax-funded prevention and early intervention services were provided to 130,000 children and transition-age youths between 2012 and 2016, nearly 65% of whom were new clients—suggesting that the reach of these services might be attributable to the tax. A study of the reach of six tax-funded, evidence-based mental health treatments for children in Los Angeles receiving prevention and early intervention services sought to understand factors that affected the sustainment of these services within the context of the tax initiative (7). Results indicated that the reach of the services varied by the unit of analysis (e.g., system level versus provider level), highlighting the complexity of measuring tax impacts.

In Washington State, few evaluations have explicitly focused on the effects of the tax. However, quasi-experimental studies have assessed the impact of services that were exclusively funded by the tax. By extension, the impact of these services is attributable to the tax. For example, a quasi-experimental evaluation of a tax-funded family treatment drug court in King County—which used propensity score matching to create a control group—found that the program improved outcomes for both parents and children (8). Some tax-implementing counties track indicators of the reach and impact of tax-funded programs and include these data in spending reports. Thurston County, for example, reports the number of people served by each tax-funded program (e.g., outpatient services were provided to 303 youths in 2018) as well as the percentage of program participants who meet the program goal (e.g., graduate from chemical dependency court). King County reports this information in addition to pre-post changes in mental illness symptoms among tax-funded program participants.

The field of implementation science offers guidance to help ensure that earmarked behavioral health taxes result in expanded access to evidence-based treatments. Earmarked taxes can align with several “outer-setting” implementation strategies (i.e., techniques to enhance the adoption and sustainment of an innovation) identified in the Expert Recommendations for Implementing Change compendium. These strategies include providing access to new funding, changing incentive structures, and mandating the use of evidence-based treatments, such a through contract requirements. Through these strategies, behavioral health taxes have potential to address well-established system- and provider-level barriers to the implementation of evidence-based treatments (3). However, the success of these strategies is reliant upon the successful implementation of the tax itself, and there is little evidence for policy—as opposed to clinical or organizational—implementation strategies (9).

Despite the dearth of evidence on policy-focused implementation strategies, the EPIS (Exploration, Preparation, Implementation, Sustainment) framework—which is widely used in implementation science (10)—can inform the design of earmarked taxes for behavioral health services. In the exploration phase, as legislators consider and draft tax proposals, it is important that tax design be aligned with the policy preferences of key stakeholders—such as constituents, consumers, mental health service providers, and behavioral health system leaders. A key question in tax design is the extent to which revenue must be spent on specific services, such as evidence-based treatments (as in California), as opposed to giving broad discretion about the services that can be funded by the tax (as in Washington). The former design is likely to produce the greatest benefits for population behavioral health, but success is contingent upon the capacity and flexibility of the local behavioral health systems and workforce to deliver these services. Thus, in the preparation phase in EPIS, it is important to assess readiness for tax implementation and develop plans and implementation support to capitalize on strengths (e.g., openness to system innovations) and address gaps (e.g., lack of trained service providers) to facilitate delivery of tax-funded services.

In particular, the authorizing legislation for earmarked taxes should consider designating funds for providing continuous support for organizational leaders and direct service providers during the implementation and sustainment phases. These types of support could include education about tax spending and reporting requirements, ongoing training for services funded by the tax, and other implementation strategies (e.g., facilitation, audit and feedback, enhanced training in evidence-based treatments) that have demonstrated effectiveness in improving the reach and fidelity of evidence-based treatments. This type of continuous support is particularly important given the high rates of staff turnover in public behavioral health systems.

During the implementation phase in EPIS, routine monitoring of the implementation process is essential to support needed adaptations to implementation strategies. Because taxes will be implemented in diverse and constantly changing local behavioral health system environments, it is important that authorizing legislation be broad enough to allow for adaptions to spending and service requirements for different populations and contexts. Focused attention to these outer- and inner-setting structures, processes, and support activities at the outset of tax design will increase the likelihood of success and potential for public health impact in the sustainment phase.

There are many important gaps in knowledge related to the use of earmarked taxes for behavioral health services. A first step is to conduct a legal mapping study to identify and describe all legislation in the United States that authorizes earmarked taxes for behavioral health services. Subsequent lines of inquiry include policy process research to examine factors that influence the passage of such taxes and advocacy frames that are most effective in building support for earmarking taxes for behavioral health services; outcome studies that use rigorous quasi-experimental designs (e.g., difference-in-difference designs with propensity score matching, interrupted time-series analysis) to assess service delivery, clinical, and population behavioral health outcomes; implementation studies that use experimental designs (e.g., stepped wedge, hybrid implementation effectiveness) to determine the effects of policy implementation strategies; and economic research to evaluate return on tax dollar investments.

Earmarked taxes are a policy strategy to address growing public demand for enhancement of behavioral health service systems across the United States. California and Washington offer two contrasting examples of how such taxes have been designed, implemented, and evaluated. Future research is needed to understand how such taxes can be optimized to maximize improvements in access and outcomes for people who need behavioral health services.