E&F - lecture 2

Question 1

What determines our demand for health care?

Answer 1

• The medico – technical model:
  o The doctor is in the lead, acting as a perfect agent
• The neo-classical model:
  o The patient/ consumer is in the lead, having perfect information
• The imperfect agency model:
  o Demand is determined by both doctors and patients as information is part of the transaction

Question 2

The medico-technical model

Answer 2

• Consumer demand is determined by medical experts based on objective needs
• Assumptions:
  o Health care providers act as perfect agents on behalf of their patients
  o Patients have uniform preferences and fully comply with the decisions made by their providers
  o Providers know with certainty the results of their decision
• Only one determinant of health care demand: need
• Thus: individual demand is completely (price-) inelastic; consumers are not sensitive to the price of medical services. Demand is fully determined by the medical expert.

Question 3

Critique on medico-technical model

Answer 3

• There is substantial empirical evidence that the assumptions of the medico-technical model are violated in practice:
  o Doctors are not perfect agents but also pursue their own, sometimes conflicting, interests
  o Doctors often do not know with certainty the effect of medical treatment
  o Consumers do not have uniform preferences for medical care, even if they would be equally well-informed as their doctors
  o Consumers do not fully comply with the decision of their doctors
  o Consumers are not insensitive to prices and income; they trade-off health care to other goods

Question 4

The neo-classical (basic economic) model

Answer 4

• Consumer demand is determined only by consumers themselves and these consumers maximize utility subject to a budget constraint
• Assumptions:
  o Consumers are sovereign and rational: they maximize utility and prefer more above less
  o Consumers have predetermined and ordered preferences; they know in advance what the difference is between goods and services, and they can trade-off the values
  o Consumers know with certainty the results of their consumption decisions; will know exactly the value of what they buy

Question 5

Critique on the neo-classical model

Answer 5

In health care assumptions about consumer behavior are violated:
* Consumers often are not sovereign but feel dependent on the physician’s judgment: information is part of the transaction!
* Consumers often do not have predetermined and well-ordered preferences with regard to medical care
* Consumers do not know with certainty the results of their consumption decisions: demand for health care does not equal demand for health
* Therefore, the demand curve (price) does not necessarily reflect the marginal value of health services to the consumer

Question 6

The imperfect agency model

Answer 6

• Information is part of the transaction between doctor and patient
• Demand is partly consumer-initiated and partly provider-initiated
• Providers act as an imperfect agent on behalf of their patients and may their information surplus to pursue their own interests (income, a=status, leisure), which may conflict with the patient’s interest
• As a result:
  o Patients demand curve may not reflect how they really value health services
  o Overprovision (supplier induced demand) or underprovision of care may occur, depending on physicians’ preferences and incentives
  o Different payment systems are likely to generate different outcomes

Question 7

Determinants of health care demand: utility theory

Answer 7

• Consumer demand is determined by consumers who maximize utility subject to a budget constraint
• Assumptions:
  o Consumers always prefer more above less of the same good  marginal utility of consumption >0
  o Consumers have predetermined and ordered preferences
  o Consumers know with certainty the results of their consumption decisions

Question 8

Key determinants of health care demand

Answer 8

• Needs (health status): position of indifference curve
• Wants (preferences): slope of indifference curve (substitution rate)
• Budget (income): position of budget restriction
• Prices of health care and other goods: slope of budget restriction

Question 9

price elasticity of demand formula

Answer 9

procentuele verandering in Q/ procentuele verandering in P

Question 10

• Warning: price elasticities are point estimates

Answer 10

o Price elasticity not only depends on the slope of demand curve (delta p/delta q) but also on the particular point on the demand curve (p,q)
o Makes it difficult to compare price elasticities across different health care settings

Question 11

elastic

Answer 11

change in price has huge impact on quantity

Question 12

inelastic

Answer 12

price decreases, impact relatively small on quantity

Question 12

what means -1 and 1 price elasticity

Answer 12

alles tussen -1 en 1 is inelastisch en alles daarbuiten is elastisch

Question 12

Problem: which prices?

Answer 12

• In health care consumers often do not pay the full market price because of health insurance (or government subsidies)
• Empirical studies therefore typically use out-of-pocket prices (e.g., co-payments)
• But out-of-pocket price elasticities (or coinsurance elasticities) may be biased due to selection effects

Question 13

How to deal with selection bias?

Answer 13

1. Correct for relevant background variables (health)
   a. Problem: omitted variable bias  you may not measure price sensitivity correctly
2. Quasi experimental methods (for example between different regions)
   a. Problem: control and experimental groups may not be comparable at baseline
3. Randomized controlled experiment (randomly assigned to a group)
   a. Problem: social experiment often face ethical constraints

Question 14

Two important randomized experiments

Answer 14

1. RAND health insurance experiment (1971-1982)
   a. 7700 selected individuals (< 65 years of age; six regions)
   b. Randomly assigned to 4 cost-sharing levels: 0,25,50,95% out-of-pocket
   c. Income dependent limit on out-of-pocket expenses (up to 1000 dollars per year)
   (NB resulting in nonlinear coverage!)
2. Oregon health insurance experiment (2008)
   a. Natural experiment: lottery for low-income uninsured to get Medicaid insurance
   b. Price variation between randomly selected treatment group and a control group

Question 15

income elasticity of demand

Answer 15

percentage change in quantity demanded / percentage change in income

IE kleiner dan 0 –> inferior good
IE groter of gelijk aan 0 –> normal goods
IE tussen 0 en 1 –> necessary goods
IE groter dan 1 –> luxury goods

Question 16

What’s different about health care demand?

Answer 16

• Demand for health care is derived from the demand for health: health care has no utility in itself, but only if it can improve health
• Demand for health care is often not well defined: people are uncertain about what and how much care they need
• Health care demand is not independent form health care supply:
  o Doctors can “shift the demand curve” for health care (supplier-induced demand)
• Price often plays a limited role as a determinant of health care demand:
  o In many countries basic health care is perceived as universal right
  o People typically have to pay only part of the price and do not know the true price
• Health care is not only a consumption but also an investment good

Question 17

Why overestimation?

Answer 17

• Healthy individuals are more likely to:
  o Have lower demand for health care
  o Choose higher copayments
• Therefore, the effect of higher copayments on demand for health care, consists of:
  o A true price effect (lower demand because care is more expensice)
  o A health effect (lower demand because consumers are healthier)
• Thus, the true price elasticity of demand is smaller (in absolute terms) than the estimated out of pocket price elasticity

Question 18

Price elasticities at different levels

Answer 18

• Price elasticities can be measured at different levels of aggregation
• For instance, you can measure the impact of price change
  o On the total demand of physician services
  o On the demand for services of an individual physician
• Price elasticities for the total demand for physician services are typically: smaller than price elasticities of demand for individual physicians

Question 19

The impact of income on demand for health care

Answer 19

• Income is another potentially important determinant of health care demand
• The impact of income is can be analyzed at different levels of aggregation: e.g. at the individual level, the regional level and the country level
• The impact of income on health care demand may be strongly related to the Prescence and extent of health insurers
• That’s one of reasons why the impact of income on demand may substantially differ across rich and poor countries

Question 20

The use of income elasticities

Answer 20

• Income elasticities are an important tool for measuring the impact of income on health care demand
• Income elasticates indicate how important a good or service is for consumers
• Income elasticities at the individual and at the country level
• Income elasticities of health care demand at the country level are typically larger than 1
• In countries with generous health insurance, income elasticities of health care demand are more relevant at the country level

Question 21

Why are individual income elasticities smaller in high income than in low income countries?

Answer 21

• In high-income countries people tend to have (much) better health insurance coverage than in low-income countries
• Hence, in high-income countries most people have good access to health care irrespective of their income (income is not a binding constraint on health care demand)
• If in high-income countries people’s income increases this does not have much impact on their health care demand,
  in contrast to low-income countries where people have to pay most health care expenses out-of-pocket