cowell

Mental Health and Substance Abuse Block Grants and the Demand for MH/SA Services in the Private Sector: 

Mental Health and Substance Abuse Block Grants and the Demand for MH/SA Services in the Private Sector P.O. Box 12194 · 3040 Cornwallis Road · Research Triangle Park, NC 27709 Phone: 919-541-8754 · Fax: 919-541-6683 · cowell@rti.org · www.rti.org RTI International is a trade name of Research Triangle Institute. Presented at AcademyHealth Annual Research Meeting San Diego, CA June 8, 2004 Presented by Alexander Cowell, Ph.D.

Contributors: 

Contributors Coauthor Jeremy Bray, Ph.D. Other Contributors Janet Cummings Kay Miller Shu Wen Ng Eva Witt Dana Messick Government Project Officer Ronald Manderscheid, Ph.D. This research was funded by the Center for Mental Health Services as part of the Medicare, Medicaid, and Managed Care Analysis (MMMCA) Project, contract #: 280-99-0701

Previous Research: 

Previous Research No study has examined the association between block grant funding and private sector MH/SA expenditures

Link Between Block Grants and Private Sector Expenditures: 

Link Between Block Grants and Private Sector Expenditures An increase in block grant funding may: Decrease private sector expenditures Claimants may seek care in the public sector instead of the private sector (crowd out effect) Increase private sector expenditures Improved infrastructure may increase the demand for MH/SA services in both the public and private sector (crowd in effect) Have no effect The public and private sectors may serve completely separate populations

Objectives: 

Objectives To estimate the association between block grant funding and the proportion of private sector fee-for-service (FFS) claimants seeking MH/SA care To estimate the association between block grant funding and average MH/SA payments per MH/SA claimant in the FFS private sector

Data: 

Data 1997 FFS MarketScan® data from the MMMCA project database Mental health and substance abuse block grant funding allocated by the federal government to states Other state-level variables controlling for state differences

Methods : 

Methods Two Outcome Variables: Probability of making an MH/SA claim (logit) Conditional on making a claim, log of MH/SA payments (regression) Explanatory variables of interest: Log of state per capita MH block grant dollars Log of state per capita SA block grant dollars Control Variables Demographics (age, gender) Months of eligibility Chronic medical conditions Health insurance plan generosity State-level policy variables and income per capita

Methods (cont’d): 

Methods (cont’d) All analyses weighted to represent the full fee-for-service MarketScan® claimant population Estimate elasticities E.g. if the elasticity of the probability of seeking MH/SA care with respect to the MHBG = 1 ... ...then a 1% increase in the MHBG is associated with a 1% increase in the probability of seeking care

Probability of Seeking MH/SA Care: Estimated Elasticities: 

Probability of Seeking MH/SA Care: Estimated Elasticities

Interpretation: 

Interpretation What is the effect of a 10 percent increase in per capita block grant funding on proportion of FFS MarketScan® private sector seeking care? MHBG from $.95 to $1.05 per capita: 2.65 percent decrease in claimants or approximately 2,882 claimants SABG from $4.60 to $5.06 per capita: 7.75 percent decrease in claimants or approximately 8,428 claimants

Conditional MH/SA Payments: Estimated Elasticities: 

Conditional MH/SA Payments: Estimated Elasticities

Interpretation: 

Interpretation What is the effect of a 10 percent increase in per capita block grant funding on FFS MarketScan® private sector expenditures? MHBG from $.95 to $1.05 per capita: 18.9 percent decrease in expenditures or approximately $59 per claimant SABG from $4.60 to $5.06: 15.2 percent increase in expenditures or approximately $47 per claimant

Future Directions: 

Future Directions Conduct similar analyses in the public sector (Medicaid, Medicare andlt;65, Medicare 65+) Use multiple years of data to examine within state changes and possible lag effects Conduct separate analyses for MH claimants and SA claimants Expand analyses to all enrollees, not just claimants

Future Directions (cont’d): 

Future Directions (cont’d) To examine causality need stronger theoretical foundation Need a better sense of how block grant funds are really spent within state

Slide15: 

Additional Information

Previous Research: 

Previous Research Increased federal MH and SA block grant allocations increase state spending on MH/SA programs (Jacobsen andamp; McGuire, 1996) SA block grant funding has no significant effect on state or local SA expenditures (Gamkhar andamp; Sim, 1999)

Data: 

Data Other state-level variables controlling for state differences Per capita arrest rates for DUI offenses, drug offenses, violent crimes, and property crimes Jail sentences and fines for marijuana offenses Beer and cigarette taxes State per capita income

Characteristics of Analysis Sample: 

Characteristics of Analysis Sample 328,906 claimants; 108,754 of whom sought MH/SA care (33%) Mean per capita MH block grant allocation is $0.95 Mean per capita SA block grant allocation is $4.60 Mean annual MH/SA payments per MH/SA claimant are $310 58% female Mean age is 36

Total Cost Implications: 

Total Cost Implications 10% increase in MHBG MHBG increases by $41 million A total decrease in MH/SA payments of approximately $7 million 10% increase in SABG SABG increases by $130 million A total increase in MH/SA payments of approximately $2 million

Discussion: 

Discussion Why are there differences in the results for MH and SA block grant funding? Relative magnitude of MH and SA block grants Differences in use of MH and SA block grant funds Need to look at changes in number of people served in public sector as well as private sector

Limitations: 

Limitations Results are obtained from cross-state comparisons, not from within-state changes Omitted variables bias Models are correlations, not causal Current payment variable does not separate MH payments from SA payments Simulations used means Claimants only, not enrollees What about the public sector?