This chapter presents the following content: Empirical estimates of demand from the literature, practice problems, the RAND health insurance experiment, example: interpreting results from a regression on abortion demand.
Demand for Medical Services Part Health Economics Professor Vivian Ho Fall 2009 These notes draw from material in Santerre & Neun, Health Economics, Theories, Insights and Industry Studies Southwestern Cengate 2010 Outline Empirical estimates of demand from the literature Practice problems The RAND Health Insurance Experiment Example: Interpreting results from a regression on abortion demand Estimating Demand for Medical Care Quantity demanded = f( … ) out-of-pocket real price income time costs prices of substitutes and complements tastes and preferences profile state of health quality of care Empirical Evidence Demand for primary care services (prevention, early detection, & treatment of disease) has been found to be price inelastic Estimates tend to be in the -.1 to -.7 range A 10% in the out-of-pocket price of hospital or physician services leads to a to 7% decrease in quantity demanded Ceteris paribus, total expenditures on hospital and physician services increase with a greater out-of-pocket price Empirical Evidence (cont.) Demand for other types of medical care is slightly more price elastic than demand for primary care Consumers should be more price sensitive as the portion of the bill paid out of pocket increases Out-of-Pocket Payments in the U.S 1970 1980 2000 2007 National health expenditures ($b) $74.9 $253.4 $1,353.2 $2,241.2 33.2% 22.9% 14.2% 12.0% % out of pocket Hypothesis: Consumers are more price sensitive if they pay a larger % of the health care bill The fall in the % of out-of-pocket payments may explain the rapid rise in health care costs Out-of-Pocket Payments in the U.S Total Expenditures and % Paid Out-of-Pocket, 2007 Hospital care Physician Services Prescription Drugs Nursing Home Care Dental $696.5 476.6 227.5 190.4 $95.20 3.3% 10.4% 20.9% 18.5% 44.2% Hypothesis: Consumers are more price sensitive if they pay a larger % of the health care bill Higher hospital and physician expenditures may be due to the low % paid out-of-pocket Out-of-Pocket Payments in the U.S (cont.) The previous slides argue that: insurance coverage expenditures But it may be the opposite: expenditures insurance coverage We cannot identify a causal effect using just this data Empirical Evidence (cont.) Studies which have examined price and quantity variation within service types have found that: The price elasticity of demand for dental services for females is -.5 to -.7 The own-price elasticity of demand for nursing home services is between -.73 and -2.4 Empirical Evidence (cont.) At the individual level, the income elasticity of demand for medical services is below +1.0 The travel time elasticity of demand is almost as large as the own-price elasticity of demand Little consensus on whether hospital care and ambulatory physician services are substitutes or complements International Estimates of Income Elasticity Are health care expenditures destined to consume a larger portion of GDP as GDP grows? Regression Analysis Sample - developed countries Ln(Real per capita Ln(Real per = + health expenditures) capita income) Estimates of range between 1.13 and 1.31 + Applying Demand Theory to Real Data • Demand analyses in health care must take insurance into account • Demand analyses are critical in shaping managerial and public policy decisions The Rand Health Insurance Experiment A large, social science experiment to study individuals’ medical care under insurance A large sample of families were provided differing levels of health insurance coverage Researchers then studied their subsequent health care use The Sample • 5,809 individuals, under 65 • sites (Dayton OH, Seattle WA, Fitchburg MA, Charlston SC, Georgetown County SC, Franklin County MA) • 1974 – 1977 • Cost : $80 million Insurance Plans in the Experiment Free fee-for-service (FFS) - i.e., no coinsurance 25% copayment per physician visit 50% copayment per physician visit 95% copayment per physician visit Insurance Plans in the Experiment Individual deductible - $150 deductible for physician visits; all subsequent visits free HMO - Not the same as free fee-for-service - Since HMO receives a fixed annual fee, it seeks to limit physician visits Table 3.3 Sample Means for Annual Use of Medical Services per Capita Plans* Face-to- Outpatient Inpatient Total Face Visits Expenses Dollars Expenses (1984 $) (1984 $) (1984 $) Probability Using Any Medical Service Free 25% 4.55 340 409 749 86.8 3.33 260 373 634 78.8 50% 3.03 224 450 674 77.2 95% 2.73 203 315 518 67.7 Individual * The chi-square test was used to test the null hypothesis of no difference among deductible 3.02 235 373 608 72.3 the five plan means In each instance, the chi-square statistic was significant to at least percent level The only exception was for inpatient dollars Source : Willard G Manning et al “Health Insurance and the Demand for Medical Care : Evidence from a Randomized Experiment,” American Economic Review 77 (June 1987), Table Results (cont.) No statistically significant difference in inpatient (hospital) expenses by insurance type Does NOT necessarily imply inelastic demand for hospital services Experiment included $1,000 cap on out-ofpocket medical expenses; 70% of hospital admissions costs $1,000 + O As coinsurance ‘s, probability of ANY use ‘s Results (cont.) Own Price Elasticity of Demand All Care Copay 0-25% Copay 25-95% • - 0.10 - 0.14 Outpatient Care - 0.13 - 0.21 As consumers’ copayments drop, demand for medical care becomes more price inelastic The data confirms the theory Results (cont.) Free fee-for-service (FFS) versus HMO coverage No difference in physician visits found But only 7.1% of HMO patients admitted to hospital, versus 11.2% of FFS patients • HMO patients cost 30% less than FFS patients on average • HMO’s save money relative to FFS Health Implications The experiment verifies that coinsurance demand for medical care What are the implications for health outcomes? i.e restraining medical care expenditures is not the only objective we care about, especially for the poor Health Implications (cont.) Poor adults (lowest 20% of income distribution) with high blood pressure experienced clinically significant improvement under free FFS plan, but not in cost sharing plan Similar findings for myopia, dental health Free FFS only improves health outcomes in specific cases versus cost-sharing If want to restrain costs and maintain health, targeted programs at these health problems is more cost-effective than free care for all services Was it worth it? Rand Health Insurance Experiment cost $80 million Initial results published in 1981 In the next years, # of insurance companies with first-dollar coinsurance for hospital care increased from 30% to 63% # of insurance companies w/ annual deductible of $200 + per person ‘d from 4% to 21% Estimated cost saving from ‘d demand for medical care = $7 billion Government sponsored studies often yield important Conclusions Our economic model of demand provides hypotheses that we can test with real data Although it is difficult to measure the quantity of medical services demanded and economic variables, both price and income effects are important determinants of the demand for medical care ... that: The price elasticity of demand for dental services for females is -. 5 to -. 7 The own-price elasticity of demand for nursing home services is between -. 73 and -2 .4 Empirical Evidence (cont.)... (cont.) Own Price Elasticity of Demand All Care Copay 0-2 5% Copay 2 5-9 5% • - 0.10 - 0.14 Outpatient Care - 0.13 - 0.21 As consumers’ copayments drop, demand for medical care becomes more price... of the health care bill The fall in the % of out-of-pocket payments may explain the rapid rise in health care costs Out-of-Pocket Payments in the U.S Total Expenditures and % Paid Out-of-Pocket,