individual-level demand for and production of health...

Individual-leveldemand for andproduction of

health.

Demand forHealth.Individual-level demand for and production

of health.Hurley, Chapter 5

Chris AuldEconomics 318

January 15, 2014


health.

Demand forHealth.

Hurley chapters we’re skipping (for now):

I Chapter 2: Equity and efficiency

I Chapter 3: Basics of markets

I Chapter 4: Methods of economic evaluation

I I will assume you already have an Econ 103 levelunderstanding of chapters 2 and 3, and we will returnto chapter 4 if time permits.


health.

Demand forHealth.

Chapter 5: Individual-level demand for andproduction of health.

I Our goal is to understand what determines the choicespeople make that affect their health.

I We will use theoretical models to help us understandthese choices and to interpret data.

I A foundational model we will focus on is due toGrossman (1972).


health.

Demand forHealth.

Basic ideas.

I We assume people have goals and tend to selectbehaviors to help them achieve those goals.

I One important goal is to be healthy, safe, andlong-lived, but this is not the only goal.

I We face tradeoffs when making choices that affect ourhealth.

I Graphs: PPF (next slide), equating marginal costs andbenefits.


health.

Demand forHealth.

C:\Users\auld\Desktop\Files\E317\Hurley_HealthEconomics_Figures_Tables.doc 3-84

Figure A5.1: Choosing the Optimal Combination of Health and Entertainment

•

•

•

Income

Leisure

365

Ymax

R

T

S

H1

H*

H3

E3

E* E1

TLS

TH TE

C

HC

Health

Entertainment

Health

Entertainment E*

H*

(ii) Constructing the Health-Entertainment PPF

(iii) The Optimal Combination of Health and Entrainment

••

•

•

PPCR

PPCS

PPCT

PPF

V

• V

V

YS

(i) Labour-Leisure Choice and allocation of time and money between health care and entertainment


health.

Demand forHealth.

Modeling health over the lifecycle.

I Off–the–shelf models of say, demand for apples do notcapture important aspects of health:

I Health is a stock, not a flow.

I Health is a derived demand: e.g., people want health,not health care.

I People produce health using multiple inputs, e.g., youmay buy an exercise bike, use your time on the bike,and eat some spinach after.


health.

Demand forHealth.

Grossman’s model.

I The model we will sketch here is due to Grossman(1971)

I Very highly cited paper. Framework and extensionscommonly used in theoretical and empirical research.

I Built in turn on Becker (1966) “household production”model.

I We will have this framework in mind when we discusstopics like smoking, obesity, illicit drugs, and health careuse.


health.

Demand forHealth.

Basic notions

I People’s goals are health and “home goods,” “bread,”or “entertainment.”

I Use time and resources (money) to produce thesegoals—tradeoffs!

I Dynamic model: takes into account how decisions todayaffect outcomes in the future.


health.

Demand forHealth.

Health capital.

I Very useful to think of health as like physical capital instandard theory.

I Health evolves over time. An increase or decrease inhealth does not immediately vanish.

Ht = Ht−1 − δtHt−1 + It−1 (1)

where δt ∈ [0, 1] represents (age–dependent) decay inhealth, Ht is health status at age t, and It is“investment” in health.

I (graphs—δt over time, an Ht profile over time)


health.

Demand forHealth.

Why do people value health?

1. Consumption: People directly value health becausethey enjoy life more when healthy.

2. Investment: Good health allows people to work morehours and to be more productive per hour worked.

3. Longevity: Good health increases length of life.


health.

Demand forHealth.

Health investment.

I We invest in health much like we think of firmsinvesting in capital stock.

I We use inputs (medical care, diet, exercise, time) toproduce an increment in health status (physical health,mental health).

I Health care in this model is a derived demand, care isvalued because it produces health, and people actuallywant health not health care.


health.

Demand forHealth.

Production of health

I We can model investment in health with a function

It = I (Mt ,THt) (2)

where M is market goods used to produce health andTH is time spent producing health. (graph-isoquants)

I Similarly, people produce “home goods” using time andmoney.

I How people choose to “produce” health depends on theprice of health-affecting goods and services and theopportunity cost of time.


health.

Demand forHealth.

Goals and constraints.

I People make choices today which affect them over therest of their lives.

I They try to make choices which give them desirablecombinations of health and “entertainment.”

I We can’t have everything we want: we only have somuch time and so much money.


health.

Demand forHealth.

Budget constraint.

I The value of the goods the person is able to purchasecannot exceed his income, where we think of bothpurchases and income over the life cycle.

I We take into account that healthier people tend to earnmore and to live longer.


health.

Demand forHealth.

Time constraint.

I Production of health depends not just on stuff (healthcare, food, etc) but also on time (e.g., exercise).

I If we measure time in days, the annual time constraint is

365 = TH + TB + TL + TW (3)

where

TH time spend producing healthTB time spent producing home goodsTL time lost to illnessTW time spent working


health.

Demand forHealth.

Time lost.

I We want to capture the effect of poor health onproductivity.

I One way to do this is to model poor health as causing adecrease in time one is able to spend working for pay orfor producing bread or health.

I (graph)

I Even if people only cared about money, and even iflength of life were fixed, people would still want toinvest in their health.


health.

Demand forHealth.

Summary of framework.

I People care about their health directly, because healthincreases length of life, and because health helps themearn income.

I People also have goals besides health (“bread”), andproduce health and “bread” using market goods andtime.

I Health decays over time and people invest in health toincrease their “stock” thereof.

I Poor health leads to time lost due to illness and loss inlifespan.


health.

Demand forHealth.

Predictions.

I The model predicts changes in how much, and which,inputs, people use to produce health over the lifespan.

I The model predicts how health and health care usagewill change as wage rates change and as educationchanges.

I Our first stab at thinking about how labor markets, theeducation system, and health are interrelated.

I We’ll consider predictions under the simplification thatthe consumption value of health is zero.


health.

Demand forHealth.


Figure 5.1: The Optimal Level of Health Capital

Figure 5.1. The optimal level of health capital in period t (Ht

*) is that level at which the marginal benefit of a unit of health capital (as indicated by the downward sloping demand curve) equals the marginal cost of health capital (as indicated by the supply curve, which is assumed to be constant with respect to health capital).

$

Health Capital in Year t

S = marginal cost of health capital = monetary cost + depreciation rate

D = marginal (monetary) benefit of health capital

Ht*


health.

Demand forHealth.

Aging.

I Physiological process of aging modeled as increase inrate of depreciation.

I As we age, must increase investment in health tomaintain a given stock of health.

I Generally optimal to choose lower health as we getolder.

I Whether health care demand goes up or down dependson elasticity of MEI schedule.


health.

Demand forHealth.


Figure 5.2: The Impact of Aging, Wages and Education on the Optimal Level of Health Capital under the Investment Demand Model (i) Aging

(ii) Wage Increase

$


S0 = MC$ + DR0

D

H*t,0H*t,1

Let Age0 < Age1 S0 = supply curve at age0 S1 = supply curve at age1 MC$ = monetary component of marginal cost of

health capital (which does not vary with age) DR = depreciation rate (which increases with age)

S1 = MC$ + DR1

$


S

D0

H*t,0 H*t,1

D0: wage = W0 D1: wage = W1 W0 < W1

D1


health.

Demand forHealth.

Wages.

I If the wage the person earns goes up, the opportunitycost of an hour spend ill goes up.

I Higher wages cause higher demand for health (graph).

I Labor market outcomes and health are theninter-related.

I Notice we are ignoring the possibility that higher healthcauses higher wages, and other reasons we may seecorrelations between health and income.


health.

Demand forHealth.

Education.

In this framework, education, or intelligence, or knowledge,or other characteristics which allow more health for giventime and money inputs are modeled,

It = I (Mt ,T Ht; E ) (4)

where E is a measure of education at time t, such that I ishigher when E is higher for given levels of money (M) andtime (TH). (graph)


health.

Demand forHealth.


Figure 5.2 cont’d: The Impact of an Increase in Depreciation Rate, Wages and Education on the Optimal Level of Health Capital, cont’d

(iii) Increase in Education

Figure 5.2. As a person ages, the depreciation rate of health capital increases, raising the marginal cost of health capital and decreasing the optimal level of health. As wage rate increases, the marginal benefit of health capital increases (because working time is now more valuable), causing the optimal level of health to rise. Finally, an increase in education makes a person more productive, increasing the marginal benefit health capital and the optimal level of health.

$


S

D0

H*t,0 H*t,1

E0 = initial education level E1 = new education level E0 < E1

D1


health.

Demand forHealth.

Education cont.

I More education leads to more health produced for anygiven combination of inputs, increasing demand forhealth.

I Whether education increases or decreases demand forhealth care depends on how large the effect of educationon health is (ie, the elasticity of the MEI schedule).


health.

Demand forHealth.


Figure 5.4: The Impact of Education on Demand for Health Capital

Figure 5.4. An increase in education causes a person’s productivity to improve, shifting the demand curve for health capital upward. A given change in productivity will cause a smaller increase in the optimal health capital for a person with inelastic demand (Ht

* to Ht^) than

for a person with elastic demand (Ht* to Ht

~). Given their increase in productivity, a person with inelastic demand may be able to achieve the new, only slightly higher, optimal level of health while reducing health care consumption, while the person with elastic demand must increase health care consumption to achieve the new, substantially higher optimal level of health.

$


Dinelastic

Ht*

Delastic

Demand curves before the increase in education Demand curves after the increase in education

D’inelastic

D’elastic

Ht^ Ht~


health.

Demand forHealth.

Other causal links between education and health.

I In this model more education makes people moreefficient at producing health, e.g., better able to chooseeffective treatments or behaviors.

I We will see later in the course that education may alsoaffect preferences over health and other goals (graph,PPF health/entertainment), and education may affecthealth behaviors for other reasons.

I Poor health may also affect education, which we ignorehere.


health.

Demand forHealth.

Assessing the model.

I Deliberately simplifies behavior: no uncertainty, onetype of health and one home good, “black box” healthproduction, and so on.

I But still richer and more realistic than basic treatments,e.g., models health over the life cycle, and explicitlymodels the allocation of time.

I Many extensions since 1972. e.g., what happens ifhealth changes have a component which is randomfrom the person’s point of view?

I Popular in part because lends itself easily to statisticalapplications.

I Many of the topics we will discuss later in the courseare modeled using variants of this approach.


health.

Demand forHealth.

Further reading.

Grossman, M. (1972) “On the concept of health capital andthe demand for health,” Journal of Political Economy80(2):223–55.Grossman, M. (2000) “The human capital model of thedemand for health,” in Handbook of Health Economics, A.Culyer and J . Newhouse (eds), Elsevier. (Also availablefrom: http://www.nber.org/papers/w7078.pdf).

http://www.nber.org/papers/w7078.pdf

individual-level demand for and production of health...

Documents