Predictions: Standard Model from Rightward Shift in Supply

• Expenditures up or down depending on elasticity of demand curve
• Fee falls
• Quantity rises
• Income per doctor probably declines

• Expenditures rise
• Shift in demand curve at least equal to shift in supply curve-->fee steady or increase, quantity per doctor about constant
• Income per physician constant

• An increase in MD supply reduces travel and waiting times, travel time because more doctors causes doctors to locate in more remote areas; waiting time a little harder to see.
• Also possible that length of visit increases and other practice amenities increase
• This causes the demand curve to shift for reasons of quality of care.

• Frequently consumed/frequently produced: little opportunity for inducement
• Infrequently consumed/frequently produced: more opportunity for inducement
• Infrequently consumed/infrequently produced: greatest opportunity for inducement

• How is target income set?
• How does physician decide between increased quantity and increased fees? Why would physician ever raise quantity?
• Raising fees involves less not more work? Reinhardt idea: order more "ancillary services"--call this "defensive medicine."
• Example PSA test.
• No limits on inducement set by the market?

• Jeffrey Harris: As markets grow larger, get increase in extent of product differentiation. At start of tobacco market, only had Camels, Chesterfields, Lucky Strike. As market expanded got menthol, filter 120mm cigs. New products-->higher prices. But opposite happened for e.g., color TV.
• Boardman et al. (1983): Will have development of boutique practices in large markets, eg., Rodeo Drive (Los Angeles), Madison Avenue (New York City)

Pauly-Satterthwaite Model

• Question addressed: Why are fees higher where physician-population ratio is higher? (e.g., fees in Manhattan much higher than Bergen County, NJ)
• Answer: More consumer ignorance about fees when physician density is higher.
• Monopoly power conferred on physicians by consumer ignorance makes demand curve facing physician more inelastic.

• In P-S Model, person asks, Do you know Dr. Johnson? In big city have to ask more people to get information on him.
• But Phelps says: a) If the person wants to find out about the fee, why not just ask the physician? b) Could change the question. Ask friend who his/her MD is. Then ask about this MD’s fees.

Empirical Results on Induced Demand from Phelps

• Physician location decisions—discuss later
• Econometric studies of physician fees
• Comparisons of utilization of physician families versus other families: Bunker and Brown (1974) and Hay and Leahy (1982)
• Hickson et al. study

• Residents in clinic as subjects, 1/2 paid on fee-for-service and half paid salary
• Patients randomly assigned to residents
• Patient load established so that on average resident income same for two groups
• Results: FFS 4.9 scheduled visits per yr. v. 3.8 for salaried; FFS 3.6 actual visits per yr. v. 2.9 salaried.
• Results, cont.: FFS failed to schedule necessary care 4% of time v. 9% for salaried residents. But FFS scheduled excess well-care visits for 22% of their patients v. 4% for salaried. Difference was statistically significant.
• Implication: More unnecessary care under FFS, but also less failure to provide needed care