Download Investment, Human Capital and Growth Kevin Murphy, PhD’86 January 17, 2013

Investment, Human Capital
and Growth
Kevin Murphy, PhD’86
January 17, 2013
U.S. Real Per Capita GDP 1889-2012
10.75
Log of Real Per Capita GDP
10.25
9.75
9.25
8.75
8.25
1880
1900
1920
1940
1960
1980
2000
Where Does Growth Come From?
• There are three primary sources of growth
• Investment in physical capital
• Investment in human capital
• Improvements in technology (knowledge)
• Primary goals of policy should be to
• Maintain the incentive for physical investment
• Provide an environment that fosters the growth of
human capital
• Provide rewards for innovation
How do People Fit into the
Economic Picture?
• People are important as both inputs and outputs
• Human capital is our most important input
• Accounts for roughly 65 percent of our productive capacity
• With increasingly mobile capital and technology, countries
will be increasingly defined by their human capital
• The production and maintenance of human capital is
our most important output
• Education
• Healthcare
• On the job training
Education Wage Premiums
2.3
Relative Wage
2.1
College Graduates
1.9
Graduate School
1.7
1.5
1.3
1965
1975
1985
1995
2005
Education Premiums by Gender
1.8
Relative Wage
1.7
1.6
1.5
1.4
Women
Men
1.3
1.2
1965
1975
1985
1995
2005
Education Premiums by Race
1.8
Relative Wage
1.7
1.6
1.5
Blacks
Whites
1.4
1.3
1965
1975
1985
1995
2005
Overall Rise in Wage Inequality for Men
Indexed Real Wage (1967=100.0)
180.0
160.0
10th Percentile
50th Percentile
140.0
90th Percentile
120.0
100.0
80.0
1966 1971 1976 1981 1986 1991 1996 2001 2006
Wage Growth by Percentile 1968-2004
0.60
Real Wage Growth
0.50
0.40
0.30
0.20
0.10
0.00
5
15
25
35
45
55
Percentile
65
75
85
95
Explaining Changes In Education
Returns Using Supply & Demand
• Growth in the college premium can be
explained by a very simple model
• Model based on Katz-Murphy 1992
• The model:
• Demand grows steadily over time
• Fluctuations in supply cause education premiums to
fluctuate
• Supply grows faster than demand  premium falls
• Demand grows faster than supply  premium rises
Supply Growth & Relative Wages
2.20
2.00
Actual Wage Ratio
Predicted Wage Ratio
1.80
1.60
1.40
1.20
1963
1973
1983
1993
2003
The Supply Response
• Growth in the college premium has generated
a predictable response – more people have
gone on to college
Wage Ratios & College Enrollment
0.50
1.8
Attendance
Wage Ratio
1.7
1.6
0.45
1.5
0.40
1.4
0.35
1965
1975
1985
1995
1.3
2005
Wage Ratio
Fraction of 20-25 Yr. Olds with Some
College
0.55
College Graduation by Gender (Age 30)
Mean GPA of High School Graduates,
High School Transcript Studies
DISTRIBUTION OF FIRST-YEAR UNDERGRADUATE GPA,
BEGINNING POSTSECONDARY STUDENTS LONGITUDINAL STUDY
The Fiscal Threat to Growth
4,000
4%
U.S. government surplus/deficit to
GDP (in %) signficantly worsened
Outlays escalated
3,000
2,500
2%
0%
-2%
-4%
-6%
2,000
Revenue
growth stayed
behind
1,500
1,000
-8%
-10%
-12%
Receipts, bln. US$, left
Sources: OECD, BEA, The Conference Board
Outlays, bln. US$, left
Government surplus/deficit to GDP, right
surplus/GDP or deficit/GDP (%)
receipts/outlays (bln. US$)
3,500
How do we get out of this?
The Canadian Experience 1990-2010
700
Canadian government "gets its house in order"
70%
600
Revenues grow at or
below nominal GDP
550
500
450
400
A 2-year pause in spending
growth (1996-97)
60%
Outlays grow at or below
population plus inflation
50%
40%
30%
350
20%
300
250
10%
200
0%
Receipts, bln. Can$, left
Outlays, bln. Can$, left
Sources: Bank of Canada, IMF, The Conference Board
Debt-to-GDP, right
government debt/GDP (%)
receipts/outlays (bln. Can$)
650
80%
Canada Debt Reduction 1995-2008
The Sources of Canada's GDP Growth and Debt Reduction, 1995-2008
6%
5%
P
4%
3%
g
I
2%
1%
e
N
r
e
r
0%
nominal GDP employment government
(N) +
growth (N+I+P)
expenditure
Inflation (I) +
growth(e)
productivity (P)
growth
government
revenue
growth(r)
Sources: IMF, The Conference Board Total Economy Database
g
r
e
5.2% =NOMINAL GDP growth = N + I + P
4.6% =revenue growth
3.6% =expense growth
λ
β
0.86 = ratio government growth to N + I = λ
0.88 = ratio revenue growth to GDP (N + I +P)
N
I
P
1.9% = employment growth
2.3% = inflation (GDP deflator)
1.0% = productivity (GDP - N - I)
U.S. Recent Historical Picture
The Sources of U.S. GDP Growth and Debt Reduction, 1995-2008
6%
P
g
r
e
5.5% =NOMINAL GDP growth = N + I + P
4.8% =revenue growth
5.2% =expense growth
I
λ
β
1.58 = ratio government growth to N + I = λ
0.87 = ratio revenue growth to GDP (N + I +P)
N
I
P
1.1% = employment growth
2.2% = inflation (GDP deflator)
1.8% = productivity (GDP - N - I)
5%
4%
3%
2%
1%
g
N
e
r
e
r
0%
nominal GDP employment government
(N) +
growth (N+I+P)
expenditure
Inflation (I) +
growth(e)
productivity (P)
growth
government
revenue
growth(r)
Sources: BEA, Federal Reserve, The Conference Board Total Economy Database
Applying the Canadian Model to the U.S.
Sources of Projected U.S. GDP growth and Debt Reduction
P
g
I
N
Note: Assuming a cyclically adjusted 19% revenue-to-GDP ratio
Sources: Federal Reserve, The Conference Board.
e
r
Improvements in Health
(Based on Murphy & Topel 2007)
Basic Results
• Historical improvements in life expectancy
have been very significant – improvements
in longevity from 1970 to 2000 were worth
roughly $95 trillion (or about $3.2 trillion
per year) to U.S. citizens
• Improvements in life expectancy have
contributed about as much to overall
welfare as have improvements in material
wealth
Figure 6a. Gains from Increased Longevity for Males 1970-2000
$500,000
Males1990-2000
Males1980-1990
$400,000
Males1970-1980
$300,000
$200,000
$100,000
$0
0
-$100,000
10
20
30
40
50
60
70
80
90
100
Aggregate Gains
Aggregate Gains from Increased
Longevity 1970-2000
Aggregate Gains (Billions of $2004)
1970-1980 1980-1990
1990-20 1970-2000
Males
$26,699
$15,471
$19,153
$61,323
Females
$20,515
$9,067
$4,440
$34,022
Total
$47,214
$24,538
$23,593
$95,345
Estimated Gains Net of the Increase in Health Expenditures
Gross Gains (from Table 5)
Increase in Expenditures
Gains Net of Expenditure Growth
Expenditure Increase as a % of Gains
1970-1980
1980-1990
1990-2000
1970-2000
$47,214
$8,206
$39,008
17.4%
$24,538
$14,928
$9,611
60.8%
$23,593
$11,591
$12,001
49.1%
$95,345
$34,725
$60,620
36.4%
Longer Term Changes
• Recent improvements are reflective of longer
term gains in longevity
• Gains were actually somewhat greater in
earlier decades using a fixed valuation profile
(like fixed basis GNP accounting)
• Gains have become increasingly concentrated
at older ages in recent decades
Figure 5: Cumulative Value of Longevity Gains Since 1900: Men and
Women in 2000
$1,400,000
$1,200,000
Males
Females
Gain (1996 Base Year)
$1,000,000
$800,000
$600,000
$400,000
$200,000
$0
1900
-$200,000
1910
1920
1930
1940
1950
Year
1960
1970
1980
1990
2000
The Good News & the Bad News:
• The Good: People value gains in health and
longevity a lot
• The Bad: Medical care is not cheap
• While life extension added 3 trillion per year
in value, medical expenditures grew by
roughly 1 trillion per year
• For the oldest groups expenditures grew by
more than the value of life extension
Balancing the Costs & Benefits
• In thinking about medical advances we must
consider both sides of the equation
• Progress is important
• Controlling costs is important
• Controlling costs raises the value of medical
advances
• Cost containment and medical progress
complement one another
A Simple Example
• 200 billion dollar “war on cancer”
• 50% probability of success – 50% probability
of total failure
• Success = 10% reduction in cancer death rates
• Based on Murphy & Topel – value of success ≈
$5 trillion
• What about costs of care?
Costs of care
• Two scenarios:
• “good” outcome = treatment adds 2.5 trillion (50% of
value) to costs of care
• “bad” outcome = treatment adds 10 trillion (200% of
value) to costs of care
• Assume each scenario is equally likely
• Three potential outcomes:
• 50% chance of “Failure” = -$200 billion
• 25% chance of “Good Success” = +$2.3 trillion
• 25% chance of “Bad Success” = -$5.2 trillion
• Expected gain = -$825 billion
What matters in this calculation?
• Costs of research are small by comparison to
costs and benefits (making them $100 billion
or $300 billion has little effect)
• Probability of success matters some but not
much
• Expected costs of care matter a lot
• Question: What can we do to improve the
situation?
• Answer: Make good care decisions!
Example Continued
• Improve care system = don’t implement if
costs of care are high
• Chance of “failure” now 75%
• But expected gain now +$425 billion
• Bottom line: appropriate cost containment
RAISES the value of research by eliminating
the major downside
• The potential downside to research is not
failure but unaffordable “success”
How do we get there?
• Best solution: improve incentives and
decisions in the delivery system – research will
follow
• Second best: change the direction of research
to look only for lower cost solutions
• Both enhance the case for more research
What does it take?
• Improve incentives for doctors and patients to
control costs
• Use technologies appropriately – not all or
nothing – many treatments will be cost
effective for some patients not for others
• Focus on treatments with low incremental
costs – reduces problem of over use
Potential Pitfalls
• Behavioral change is not free – people value
behavior as well as health – people value eating and
even smoking
• Behavioral change that mitigates gains in longevity
does not diminish the value of progress and maybe
increases it
• Behavioral factors increase in importance as care
moves out of the hospital and into the household
• Patient inputs make education more important and
inequality a bigger issue
Important Policy Questions
• How do we take advantage of growing
demand for education and skills?
• Increasing investment in higher education
• Improving education at lower levels
• This is a long term project
• How do we take advantage of potential gains
from medical advance?
• Balancing the costs and benefits
• Improving insurance – delivery system
• Talking advantage of scalable technologies and the
world-wide growth in incomes