11Analyzing Medical and Dental Markets to Improve Financing Mechanisms*Paul Griffin, Julie L. Swann, Natalya ZlobinISyE, Georgia Institute of Technology (GT)Susan Griffin Oral Health, Centers for Disease Control*Supported by NSF grant DMI-02233642Outline• Background• Informational Asymmetries in Dental Insurance Market• Comparison of Dental and Medical Markets• Conclusions and Implications for Policy• Future Research23• Proportion of elderly with natural teeth increased from 40% in 1957 to over 75% in 1999 • Elderly may be at increased risk for dental decay• At present, elderly may have less access to preventive and routine dental services:• Medicaid typically only covers emergency care for adults and Medicare covers none• Less than 1 in 4 have dental insurance• Dental inflation rate has outpaced medical rate• Many dental schools have closed, leading to reduced suppliers of servicesBackground on Dental Care in U.S.4• How can access/care be improved at a reasonable cost?• Optimal financing mechanism to subsidize consumption of dental care among elderly depends on:• If adverse selection (people with poor dental health are more likely to purchase dental insurance) exists; then private insurance markets may not develop in absence of pooling mechanisms• If moral hazard (consumption of dental services increaseswith insurance) exists; then insurers would have to monitor dental providers• If pent-up demand (consumers delay services until periods of coverage) exists; then insurers need to account for variability in costsInformation Asymmetries Research35• Population Studied: U.S. adults aged 18+• Data Sources: 1999 MEPS & NHANES III • Variables:• Insured• MEPS: Reported having dental insurance in all 3 phases or had dental service paid for by insurance• NHANES III: Reported having dental insurance (1988-1994)• Health Status: Fair, Average or Very Good• MEPS: general health status in more than 2 phases• NHANES III: condition of natural teethMethods6• Used statistical testing (stratified) and logistic regression to answer following (level of significance = 5%):• Does insured differ by dental health status(controlling for age and income)• Do past year dental visits differ by insured (controlling for age, income, and dental health status)• Among those with past year dental visit do # of follow-up visits and annual dental expendituresdiffer by insured (controlling for age, income, and dental health status)Analysis47$350.18$520.4565+$368.00$499.9145 to 64$262.82$338.5718 to 44UninsuredInsuredAge in YearsMean Annual Dental Expenditures among Adults with Past Year Dental Visit (MEPS)IncreasingIncreasing81.10 (0.96-1.26)1.30 (1.13-1.57)1.61 (1.34-1.93)2.07 (1.75-2.44)HealthAVGVG0.43 (0.35-0.53)0.19 (0.17-0.22)Income ≤ 200% FPL0.26 (0.20-0.34)0.22 (0.19-0.26)Age ≥ 65NHANES III (teeth)MEPS (general health)Odds Ratio using 95% CIReference group: aged 18-64 years and reporting fair health.Adverse Selection: Is insured status negatively associated with dental health status?59Does insurance cause good health?3.59%1.84%3.66%1.20%Poor Health9.65%6.35%10.45%6.93%Fair Health27.55%22.95%27.29%20.73%Avg Health31.60%35.62%30.30%32.91%Good Health27.60%33.24%28.30%38.23%Very Good HealthPhase 2 = noPhase 3 = no(control)Phase 2 = noPhase 3 = yesPhase 1 = noPhase 2 = no(control)Phase 1 = noPhase 2 = yes• Working population (18-64) that went from uninsured in one phase to insured in the next phase (MEPS) had significantly higher initial health than the uninsured10• Approach• Compare dental expenses for people that changed from uninsured to insured to people insured in all rounds (control)• If pent-up demand exists, then the annual dental expenses will be higher (compared to control) for those who just become insured• Results• Average dental expenses for those who switched from uninsured to insured was $133.96 (SE=17.30) as compared to $205.54 (SE=8.67) for the control group.• Difference is significant at 1% levelIs there pent-up demand in the dental market?611Increased Consumption: Are dental visits positively associated with insured status?1.78 (1.53-2.08)2.45 (2.03-2.95)1.08 (0.92-1.27)1.35 (1.18-1.55)AVG HealthVG Health0.59 (0.50-0.70)0.58 (0.52-0.65)Income ≤ 200% FPL1.75 (1.47-2.08)1.98 (1.73-2.26)Age ≥ 651.96 (1.65-2.34)3.75 (3.38-4.16)InsuredNHANES III (teeth)MEPS (general health)Odds Ratio and 95% CIReference group: aged 18-64 years and reporting fair health.12Increased Consumption among Utilizers1.16 (0.96-1.40)1.28 (1.08-1.51)1.06 (0.84-1.34)0.94 (0.77-1.14)AVG HealthVery Good Health0.56 (0.49-0.63)0.81 (0.69-0.96)Income ≤ 200% FPL2.21 (1.91-2.56)2.01 (1.67-2.43)Age ≥ 653.42 (3.05-3.84)1.31 (1.13-1.51)InsuredExpenditures >median>1 Dental VisitOdds Ratio on MEPS data with 95% CIReference group: aged 18-64 years and reporting fair health.• Among those with dental visit are number of follow-up visits and annual dental expenditures positively associated with insured status?713• Adverse selection does not appear to exist in the U.S. dental insurance market• Consumption of dental services increases with insurance• Dental visits increase with insurance• Supplier induced demand (number of follow-up visits) although significant, does not appear to be high• Among adults with dental visit, average annual expenditures are higher for insuredInformation Asymmetry Conclusions14Analyzing Medical & Dental Markets• Medical and dental markets operate differently• Risk and caps lower in dental• Goals• Use economic theory to identify factors affecting insurance coverage and utilization• Compare presence of such factors in medical and dental markets • Predict their impact on medical and dental insurance• Test the predictions815Model and Predictions• Employment:• Positive association with coverage • More of an impact for medical • Income• Unclear a priori• Tax advantages versus risk aversion−=,',,1,,behaviorprovidermonitorandhealthascertaintoabilitysinsurerincomeagetfromcaredefertoabilityhealthemploymentfInsurancetttttWealth$W0WillEWU(W0)U(Will)EUUtilityExpected UtilityUtility of illness with and without insuranceW0= initial wealthWill= wealth after illnessp = probability of an episodeE(Wealth) = p*Will+ (1-p)*W0E(Utility) = p*U(Will) + (1-p)*U(W0)*Pauly, M. (1968). “The Economics of Moral Hazard: Comment,” The American Economic Review 58, pp 531 – 537.917Risk Theory•