Health Economics

The production of health

Gerald J. Pruckner

Gerald J. Pruckner The production of health 1 / 23

The production of health

What is the contribution of health care to the health status of

the population?
What is the best way to produce and distribute health care?

Different features of life not only health care contribute

to our health!

The production function of health

The role of medicine and health care
The importance of schooling

Gerald J. Pruckner The production of health 2 / 23

Production function

The importance of marginal products:

Gerald J. Pruckner
investment of a marginal A
The production of health
C 1 billion 3 / 23
Marginal product and its importance

Gerald J. Pruckner The production of health 4 / 23

The historical role of health care and medicine

Stylized facts
Practitioner-provided medical interventions played a small role
in the historical decline in population mortality rates
Effective medicine is a fairly recent phenomenon (twentieth
century)
Other causes of mortality declines
public health measures
spread of knowledge of the sources of disease
improvements in environment (supply of foodstuffs
agricultural and industrial revolutions)

Gerald J. Pruckner The production of health 5 / 23

US Death rates: infectious diseases

Gerald J. Pruckner The production of health 6 / 23

Public health

Evidence that public health measures have contributed to the

decline in mortality rates, however,
Disputes whether public health was the major contributor
. . . it primarily affected the exposure to waterborne and
foodborne disease
. . . it cannot account for the large decline in airborne infectious
disease mortality
Public health measures
immunizations
quarantines
standards for sanitary water supplies and sewage systems
sanitary handling and treatment of foodstuff
These measures came into widespread use late in the era of
mortality declines

Gerald J. Pruckner The production of health 7 / 23

Nutrition

The introduction of corn and potatoes

Agricultural advances
new crops
crop rotation
seed production
winter feeding
improvements in farm implements
Improved nutrition has obviously reduced infectious diseases
Fogel (1986)
40 % of the historical decline in mortality rates due to
improved nutrition
the biggest rewards attributable to infants
the quality of the protein in the diet played a major role

Gerald J. Pruckner The production of health 8 / 23

What do we learn from the medical historian?

Controversy over the importance of different measures (public

health, nutrition, environment, . . . )
Consensus on the minor role of medical practice
Importance of medical research (insights into causes and
transmission of diseases, health-enhancing behavior, . . . )
Medical research accounted for almost one third of the
cost-savings to society from reduction in sickness or death
rates in the US between 1900 and 1975 (Mushkin 1979).
Effectiveness of the healthcare system in particular in treating
depression, pain, and discomfort.
Reduction or elimination of pain from migraine headaches,
terminal cancer, heart treatment, . . .
Substantial contribution of healthcare system in total over
recent decades.
Gerald J. Pruckner The production of health 9 / 23
Health production in modern times

The marginal product of healthcare is relatively small.

The most important factors influencing mortality are related
to socioeconomic status and lifestyle (Thornton 2002).
In many instances exceed incremental benefits incremental
cost.

Gerald J. Pruckner The production of health 10 / 23

Allokative Effizienz des Mitteleinsatzes: Kosten

Tabelle: Kosten pro gewonnenem Lebensjahr

Dekade Systembeitrag Systembeitrag

50 % 30 %
Alle Personena
1960-70 A
C 8.542 A
C 14.236
1970-80 A
C 9.687 A
C 16.144
1980-90 A
C 11.582 A
C 19.304
1990-00 A
C 20.768 A
C 34.613
2000-08 A
C 26.094 A
C 43.491
Quelle: Eigene Berechnungen
a
Preise 2009

Gerald J. Pruckner The production of health 11 / 23

Allokative Effizienz des Mitteleinsatzes: Nutzen

Leiter, Pruckner (2009, EARE): Zahlungsbereitschaft f

ur die
Vermeidung eines zusatzlichen Lawinenrisikos
WSLAUT A
C 2,12 mio. [2,03 2,21]

Tabelle: Nutzen pro gewonnenem Lebensjahr WSLJ

Diskontsatz
3% 2%
Mittelwert A
C 74.257 A
C 62.334
Quelle: Eigene Berechnungen auf Basis von Leiter und Pruckner (2009)

Durchschnittsalter der Befragten: 39 Jahre

Lebenserwartung: 81 Jahre
Preise 2009
Gerald J. Pruckner The production of health 12 / 23
Morbidity studies

The Rand Health Insurance Experiment

One of the largest randomly controlled economic experiments
ever conducted
Effects of alternative health insurance policies were tested on
the demand for health care and on the health status
The greater the portion of their health care bill the less health
care individuals choose to purchase
Fully insured people purchased 40 % more health care as
compared to those who had to pay their own bill
However, this increase in services had almost no effect on
health status for the average adult

Gerald J. Pruckner The production of health 13 / 23

Lifestyle and environmental pollution

Empirical evidence on the influence of lifestyle on health

. . . what we do for ourselves often matters the most for our
health . . .
Smoking certainly causes ill health
Maternal cigarette smoking has a significant negative effect
on newborn birth weight (Rosenzweig & Schultz 1983;
Rosenzweig & Wolpin 1995)
It is well-known that environmental pollution causes ill health
and deaths in individuals (e.g. Cropper et al. 1997)

Gerald J. Pruckner The production of health 14 / 23

The role of schooling

Health status is significantly correlated with schooling

Two different theories on the role of schooling
better educated persons are economically more efficient
producers of health (they have the know-how to stay healthy;
they better know how to use medical and other market inputs
and their own time to produce health)
schooling and health are correlated only because they are both
related to one or more other factors: self-selection problem
People with low discount rates will tend to invest in both
education and health

Gerald J. Pruckner The production of health 15 / 23

The causality issue

Education Health
2

3 3

Confounding factors
- education of parents
- parents genes
- living conditions
-

Gerald J. Pruckner The production of health 16 / 23

Schooling and health empirical evidence

A natural experiment setting (Lleras-Muney 2002)

Compulsary education laws came into being in various places
at various times in the US in the 20th century
Birth cohorts from that era would have experienced different
levels of education but would have been similar in many other
respects
The survival pattern of these people allowed to detect a pure
influence of education on health
Conclusion: education has a clear, causal, and positive effect
on health

Gerald J. Pruckner The production of health 17 / 23

Screening in Austria: an IV approach

The main idea

Patients do not only self-select into screening but are also
examined simply because of their GPs recommendation.

Do GPs have an incentive to recommend screening?

Screening as a sensible way of secondary prevention. GPs
advocate it to their patients to improve their future well-being
(altruistic motive)
Financial incentives to sell screenings (financial motive)
can be sold to healthy patients
the screening fee for a GP in Austria is four times higher than
that for the first treatment in a quarter

Gerald J. Pruckner The production of health 18 / 23

The IV in detail
IV Screening exposure in zip code area
Screeningsi,z,t
Si,z,t =
Insurantsi,z,t

Gerald J. Pruckner The production of health 19 / 23

The effect of screening on health outcomes

Gerald J. Pruckner The production of health 20 / 23

Drugs for the nervous system

back

Gerald J. Pruckner The production of health 21 / 23

The effect of screening on health outcomes

Gerald J. Pruckner The production of health 22 / 23

Screening & Mortality

Changes in the probability of being alive in 2009.

Birth cohorts
1944-1953 1934-1943 1933 or before

OLS 0.01 0.02 0.04

(0.00) (0.00) (0.00)

IV 0.03 0.02 0.07

(0.03) (0.04) (0.12)

Coeff. of instrument 1.05 1.01 1.12

F Stat 44.36 26.32 34.46
Observations 99,008 85,723 92,745

Estimation method: Linear Probability Model. Data Structure: Cross Section.

Standard errors are robust but not clustered. Other controls: doctor, age,
and zip code area dummies; dummies for foreign nationality, academic degree,
sex and exemption of prescription charge.

Gerald J. Pruckner The production of health 23 / 23