Review and Evaluation of the Substance Abuse and Mental Health Services Block Grant Allotment Formula
Jan 1, 1997
More Refined Measures Would Shift Substance Abuse Funds to Smaller, More Rural States
Each year, the federal government distributes block grants totaling $1.2 billion to the states for substance abuse services, representing roughly one-quarter of the total state expenditures on publicly funded services. Throughout the history of this program, begun in 1981, Congress has debated the most equitable way to distribute these allocations among states. A new RAND study reports that the adoption of empirically derived measures of the need for services and refined measures of their costs would substantially shift block grant funds for substance abuse to smaller, more rural states.
In 1992, Congress reauthorized the substance abuse block grant program that helps states provide treatment services for alcohol and drug abuse for all ages, and programs targeted at youth to prevent alcohol, drug, and tobacco use. Because discussions of the new law raised questions about its equity, Congress mandated a study to assess the equity of the existing allocation formula at the same time.
By using data that more explicitly measured the three components of the existing allocation formula, researchers from RAND's Drug Policy Research Center found that measures for two of the components—the population in need and the cost of services—could be improved. The third component, each state's ability to raise revenues to meet its costs of delivering services, is currently based on states' total taxable revenues and was found to be a good indicator of fiscal capacity.
Currently, need calculations are based on the shares of population in two age groups: 18- to 24-year-olds and those 25 to 64. Young adults living in urban areas are double-counted, reflecting an arbitrary judgment about the relative significance of substance abuse in urban areas. These calculations could be improved by taking into account the predictors of need for substance abuse treatment identified by the National Household Survey of Drug Abuse (NHSDA). RAND's analysis of the survey determined that the need for substance abuse treatment was best predicted by multiple social and demographic characteristics. Males, white non-Hispanics, those who are separated or unmarried, adults between 18 and 44, high school dropouts, and residents of western states are more likely to need treatment services. The resulting alternative measure of need in each state used a weighted summation of these predictors based on Census data. Contrary to the logic of double-counting young adults in urban areas, residence in a rural area was found to be predictive of a greater probability of need for substance abuse services because across states more people need treatment for alcohol abuse than for drug abuse, and because residents of rural communities bear a higher risk of alcohol abuse than urban residents.
Another improvement in the formula could be made by taking into account state differences in need for prevention services. Federal regulations require that at least 20 percent of a state's allocation go to fund prevention programs, although states may direct a larger proportion to prevention if they wish. After first defining the population needing prevention as youth aged 12-20, researchers calculated how many young people were likely to need prevention services by analyzing predictors of risk for substance use among youth and developing state indicators of prevention need using Census data on these predictors. Analyses of NHSDA data identified these predictors as being male and an older adolescent, having a lower family income, not attending school full time, living alone, and working.
At present, the cost of services is determined by an equation based on wages and rents from Census data. These costs are scaled according to the average fees of practicing physicians. However, this calculation is not sensitive to the specific occupational composition of the labor needs of substance abuse services or to cost differences in rural and urban settings.
To evaluate the current measure, researchers examined data on the actual provision of substance abuse services from the National Drug and Alcoholism Treatment Unit Survey (NDATUS). The findings showed that 81 percent of the cost of substance abuse treatment is from labor, rather than facilities. This is a higher proportion than for the physician services on which the current formula is based. Also, NDATUS analysis revealed that states systematically spend more to provide services in rural areas because, unlike cities, they do not benefit from economies of scale.
|Changes in State Allocations||80% Treatment, 20% Prevention Split|
Consistent with their charge from Congress, researchers did not undertake the task of recommending an alternative allocation formula. Proposing an alternative for implementation would require the identification of reliable sources of data for annual updates of state allocations, a task that should be pursued as a next step in making allocations more equitable. Instead, the study's alternative measures were intended as tools to evaluate the equity of the current approach.
A comparison of the current and alternative formulas reveals that the existing formula for allocating block grants shortchanges smaller and more rural states. If block grants were distributed according to the alternative measure used for analysis, 22 percent of the allocation for substance abuse services would have shifted across states in 1992, assuming an 80/20 split between treatment and prevention needs. Using this formula, nine states would lose 20 percent or more of their block grant funding, and 18 would gain 20 percent or more. The figure shows that more rural states would receive larger allocations, whereas some urban states would experience relative declines. Although such a correction seems severe, it is important to remember that Congress has traditionally implemented allocation changes only through budget increases from year to year, rather than through cuts to states' actual allocation amounts.
Researchers also analyzed mental health grants, which were separated from substance abuse allocations by the 1992 law. Although still important to states, these grants represent a much smaller share of what states spend on mental health services than the substance abuse block grants. The current mental health formula is based only on age as a need indicator and does not take other important predictors into account. However, the need for mental health services is higher for women, those who are disabled or receiving welfare, and those without a high school diploma. When these predictors are considered, the proportion of the population needing mental health services tends to be higher in southeastern states and lower in northwestern states. If the measure of need were adjusted for these factors—and if the cost-of-services calculation were refined according to actual delivery costs as with substance abuse—5 percent of mental health block grant funds would shift across states. Those living in states with a higher percentage of their populations in poverty would benefit. It is notable that even with major changes in the formula, allocations would shift only slightly.
Since other federal block grant formulas are not currently based on empirical measures of factors deemed to be important for ensuring equity, this model for evaluating such formulas may be usefully applied to these programs. For example, the current welfare block grant program is based on the dollar amounts distributed to states in 1996, before the program was implemented. Over time, this formula is likely to become increasingly problematic and may need to be refined to reflect changes in states' population needs, costs of providing services, or the ability of taxpayers to pay for services.
In the final analysis, the structure of the current substance abuse formula can be described as fundamentally sound and even conceptually sophisticated. RAND's analysis showed that as currently structured, the formula's fundamental purpose is to equalize the rates taxpayers in different states have to pay to support a standard level of service. Nonetheless, inequities have arisen because of limitations of the measures of formula components; these could be mitigated by more refined analysis of available data. In addition, the formula is limited in that it considers service needs of the total population—rather than just those of the uninsured who receive block grant services—and does not take into account other sources of federal funds for substance abuse services such as Medicaid. Now that more federal block grant programs with much simpler allocation formulas are being introduced, this study of the substance abuse grants may provide guidance for their improvement as well.