Week 12

Lectures 23 & 24

The Economy in the Long Run: Economic Growth

Reference: Bernanke, Olekalns and Frank Chapters 11,
12
Key Issues
Economic Growth
Real GDP per capita
The production function Cobb-Douglas
Growth accounting

Economic Growth
Economic growth has delivered large increases in
standard of living (at least for many people)
Concerned with the long-run evolution of potential real
output (y*)
In AD-AS model changes in AD have no effect on longrun growth rate of GDP (need a different model)

Measuring Economic Growth

Conventional to use real GDP per capita as a measure of
a countrys living standards and stage of economic
development
Real GDP per capita:

We know from Week 1, that GDP is not a perfect

measure of economic wellbeing; however it does seem
to be positively related to life expectancy, infant health
etc.

Aug-2011

Jul-2010

Jun-2009

May-2008

Apr-2007

Mar-2006

Feb-2005

Jan-2004

Dec-2002

Nov-2001

Oct-2000

Sep-1999

Aug-1998

Jul-1997

Jun-1996

May-1995

Apr-1994

Mar-1993

Feb-1992

Jan-1991

Dec-1989

Nov-1988

Oct-1987

Sep-1986

Aug-1985

Jul-1984

Jun-1983

May-1982

Apr-1981

Mar-1980

Feb-1979

Jan-1978

Dec-1976

Nov-1975

Oct-1974

Sep-1973

$A, cvm

Real GDP per Capita for Australia (1973-2012)

70000

60000

50000

40000

30000

20000

10000

Growth Fact 1
There are large differences in the level of real GDP
per capita across countries
$US (2005 prices)
US

Japan

2010 41,376 31,453

China South Korea Chad

7,746

26,614

1,332

Cross-country inequality in averages

5

Growth Fact 2
An important cause of this inequality is the fact that
countries grow at different rates.
Average % per annum

60-90

US

Japan

1.4

5.0

China South Korea Chad

2.4

6.0

-1.7

Small Differences in Growth Rates Have Big Level Effects

Relatively small differences in growth rates, maintained
over many years, can make large differences in living
standards.
Country Growth
Rate
A
1%
B
B-A

2%

Initial
Income
1000
1000
0

Income after x years

5
25
35
1051 1284
1419
1105
54

1649
365

2013
594
7

Convergence and Catch-up

Idea that relatively poor countries will have higher
growth rates, than rich countries and eventually catchup in terms of the level of per-capita GDP.
If we look at the set of all countries in the world, then
there is no evidence of convergence in per-capita GDP
But, for some groups of countries there is evidence of
convergence

Convergence in Rich Countries

6

Annual growth % (1950-2000)

0
0

2000

4000

6000

8000

10000

12000

14000

GDP-per capita in 1950 (1995 USD)

JA, IT, GER, FR, UK, CA, AU, US

9

Decomposing Changes in GDP per Capita

Real GDP per capita:

Re-write real GDP per capita as:

GDP GDP
N

POP
N
POP

where N is employed workers.

The right-hand-side is:
Average labour productivity (GDP/N) (times)
Share of population in employment (N/POP)
10

Growth in Australian Real GDP per Capita

Real GDP per capita can only grow if labour productivity
grows, and/or the employment share grows
From 1965 and 2006, real GDP per capita grew by 108%.
share of population working grew from 66% to 74%;
as more women entered the workforce
average labour productivity increased by 87%

11

Influences on Labour Productivity (y/N)

Physical Capital per worker
Includes the stock of machines, tools, plant and
equipment, buildings and structures
Increases in quantity and quality of physical capital
increase labour productivity
But, there are usually diminishing marginal returns to
capital.
12

Influences on Labour Productivity (y/N)

Human Capital
Refers to education and training, skills and talents of
labour
Can be formal education or schooling or on-the-job
training
Think of N as basic or unskilled labour input. Human
capital accumulation raises quality of N
13

Influences on Labour Productivity (y/N)

Land and Natural Resources
Productive land, energy and mineral resources have
potential to improve labour productivity e.g. Australia
But,
Not essential to high labour productivity e.g Japan,
Singapore
Resource curse Abundant natural resources
impede economic growth e.g. African countries
14

Influences on Labour Productivity (y/N)

Technology
Stock of ideas and knowledge
Development of new technologies is viewed as the
primary source of economic growth
Combination of:
- Primary research and
- Development (applied research)
15

Influences on Labour Productivity (y/N)

Management and Entrepreneurs
Development of new firms and the way firms are
organised
Bill Gates and Microsoft, e.g. Entrepreneur
Just-in-time inventory, theory of management
Internal to a firm
16

Influences on Labour Productivity (y/N)

Political and Legal Environment
External to the firm
Includes the system of property rights, degree of
political stability
What are the incentives for people to produce goods or
ideas?

17

Policies to Promote Economic Growth

Are there policies that can increase a countrys rate of
economic growth?
- Support for (basic) human capital accumulation
- Encourage saving and investment
- Support for research and development (particularly
fundamental research)
- Secure system of legal and property rights

18

Limits to Economic Growth

Can economic growth continue indefinitely without
depleting natural resources and causing massive change
to the global environment? Are there limits to growth?
Some possibilities:
Growth in real GDP can be in the form of new or
higher quality products
Higher levels of real GDP per person are associated
with lower levels of pollution
Market prices can adjust for shortages in resources
19

Aggregate Production Function

Production functions will be familiar from
microeconomics where we can represent the output of
an individual firm as a function of its inputs.
No. of computers produced by a firm depends on:
Number of workers
Quantity of capital
In macroeconomics we represent aggregate output (GDP)
by a production function.
20

Aggregate Production Function

Assume that level of real output (y) depends upon three
things:
aggregate labour input (l)
aggregate capital stock (k)
the state of technology (A)
Formally
= (, )
The exact form of the function f (.) is not specified

21

A Picture
Y
= (, )

k and A are fixed

l (labour)

22

Another Picture
Y
= (, )

l and A are fixed

k (capital)

23

Property 1: Constant Returns to Scale (CRS)

= (, )
Suppose we simultaneously double (2) both k and l, this
leads to a doubling of output.
(, )
Double
(2, 2)
CRS implies

2 (, ) = 2

24

Which of the following production functions exhibits CRS?

(a) y = k + l
(b) y = kl
(c) y = k/l

25

Which of the following production functions exhibits CRS?

(a) y = k + l
(b) y = kl
(c) y = k/l

2k+2l = 2(k+l) = 2y
2k2l = 4kl = 4y
2k/2l = k/l = y

Yes
No
No

26

Property 2: Diminishing Marginal Product

y A f (k , l )

Suppose we change one input, but hold all of the others

fixed.
Marginal products of labour and capital are:
positive
y
MPk
A f k (k , l ) 0
k

y
MPl
A f l (k , l ) 0
l

27

Diminishing Marginal Product

diminishing
k MPk

l MPl

28

Demand/Marginal Product for Labour

MPl

MPl

l
Marginal product curve gives demand for labour.
Labour will be employed up to point where:

MPl

W
P

29

Demand/Marginal Product for Capital

MPk

MPk

k
Marginal product curve gives demand for capital.
Capital will be employed up to point where: MPk r

30

Cobb-Douglas Production Function

y A f (k , l )

Specific Form
y Ak l1

31

Marginal Products
1

y Ak l
How does y change if we change l?
Partial derivative
y

l
1 1
(1 ) Ak l (1 ) Ak l
(1 ) Ak l
l
l
l
y
y
MPl (1 )
l
l

Marginal product of labour

32

Marginal Products
Marginal product of labour
MPl (1 )

y
(1 ) APl
l

Marginal product of capital

y
MPl APk
k

Marginal product = Average Product Exponent

33

A Result: Factor Incomes Exhaust Total Output

Suppose labour and capital are paid their marginal
products.
(Real) labour income =

W
y
l MPl l (1 ) l
P
l

(Real) capital income =

y
r k MPk k k
k

labour income + capital income = y

(1 ) y y y
34

Contributions to Economic Growth

Suppose we believe that the Cobb-Douglas function
provides a good model for aggregate output in an
economy.

y Ak l1
It must be the case that all changes in Y can be
accounted for by changes in labour, capital or
technology.
Growth accounting
35

A Result on Logarithms
Suppose

z xa
Take logs of both sides
log z a log x
Take the difference of both sides
log z a log x
z
log z
Note that
z
z
x
a
So
z
x
36

Decomposing Output Growth

Production function for output in period (t):
yt At kt lt1

Use the above result to write

yt yt 1 At At 1
kt kt 1
lt lt 1

(1 )
yt 1
At 1
kt 1
lt 1

or
yt At
k
l

t (1 ) t
yt 1 At 1
kt 1
lt 1

37

Contributions to Output Growth

yt At
kt
lt

(1 )
yt 1 At 1
kt 1
lt 1

In any period output growth is due to:

growth in technology;

At
At 1

growth in capital (weighted by );

kt

kt 1

growth in labour (weighted by 1-);

lt
(1 )
lt 1
38

Estimating the Growth Rate of Technology

While we potentially have direct measures of output,
labour and capital, this is not so for technology.
However write

At yt
kt
lt

(1 )
At 1 yt 1
kt 1
lt 1

Given data on the growth rates of ouput, labour and

capital and a value for , we can estimate the growth
rate of technology (sometimes called TFP = Total Factor
Productivity or MFP = Multi-Factor Productivity)
39

Estimating
We showed earlier that
MPk

Now re-arrange to get

y
k

MPk k

If capital is paid its marginal product

rk

so the left-hand side is just capitals share of total output,

and we can use this figure as an estimate for .
40

Decomposition of Output Growth for Australias Market

Sector
Annual Growth Rates
Period
y/y
k/k + (1-) l/l
A/A
1965-1970
5.1
4.0
1.1
1970-1980 2.9
1.4
1.6
1980-1990 3.2
2.4
0.8
1990-2000 3.2
1.7
1.5
2000-2005 2.8
2.2
0.6
1965-2005 3.3
Source: ABS

2.1

1.2

41

Decomposition of Output Growth for Australias Market

Sector (Updated)
Annual Growth Rates
Period
y/y
k/k + (1-) l/l
A/A
1965-1970
5.1
3.9
1.2
1970-1980 3.1
1.6
1.5
1980-1990 3.1
2.4
0.7
1990-2000 3.4
2.1
1.3
2000-2010 3.3
3.1
0.1
1965-2010 3.4
2.4
1.0
Source: ABS

42

2.0
% per annum

TFP (or MFP) Growth in Australia 1966-2011

10.0

8.0

6.0

4.0

0.0

Jan-2010
Jan-2008
Jan-2006
Jan-2004
Jan-2002
Jan-2000
Jan-1998
Jan-1996
Jan-1994
Jan-1992
Jan-1990
Jan-1988
Jan-1986
Jan-1984
Jan-1982
Jan-1980
Jan-1978
Jan-1976
Jan-1974
Jan-1972
Jan-1970
Jan-1968
Jan-1966
-2.0

-4.0

-6.0

43