Pmak Session4

Macroeconomics 2 (PMAK)
Discussion Session 4
Bopjun Gwak
Goethe University Frankfurt
June 4, 2018
Bopjun Gwak (Goethe University Frankfurt) PMAK - Discussion Session 1 June 4, 2018 1 / 32
Today’s Agenda
1 Levels Accounting
2 R&D-based endogenous growth model
3 Business Cycle Facts
1 Levels Accounting
Levels Accounting
Question: Why are some countries so much richer than others?

Production function of country i with human capital-augmented labor
input (Hi ):
Yi = Kiα (Ai Hi )1−α
α
1−α
Yi Ki Hi
How do we get Li = Yi Li Ai ? (Lecture 4, s.13)
α
Kiα (Ai Hi )1−α

Yi Ki Hi
= = Ai
Li Li Ai Hi Li
α
! 1−α α
Ki1−α
1−α
Hi Ki Hi
= Ai = Ai
(Ai Hi )1−α Li Kiα (Ai Hi )1−α Li
1−α α
Ki Hi
= Ai
Yi Li
Levels Accounting
Production function of country i: Yi = Kiα (Ai Hi )1−α

Human capital-augmented labor input:
Hi = eφ(Ei ) Li
where
I φ(.): a not yet specified function
I Ei : average years of schooling in country i
I Li : labor input in country i
If Ei = 0, then φ(0) = 0 and Hi = e0 Li = Li

Hi
hi = Li = eφ(Ei )
Levels Accounting
How do we specify φ(E)?
Production function of country i: Yi = Kiα (Ai eφ(Ei ) Li )1−α
Wages:
∂Yi
wi = = (1 − α)Kiα (Ai eφ(Ei ) Li )−α Ai eφ(Ei )
∂Li
= (1 − α)Kiα A1−α
i e(1−α)φ(Ei ) L−αi
Consider the wages of two workers (1,2) in the same country i with
different levels of education:
w1,i = (1 − α)Kiα Ai1−α e(1−α)φ(Ei,1 ) L−α
i ; w2,i = (1 − α)Kiα A1−α
i e(1−α)φ(Ei,2 ) L−α
i
w1,i
= e(1−α)[φ(E1,i )−φ(E2,i )] ⇔ ln(w1,i )−ln(w2,i ) = (1 − α)[φ(E1,i )−φ(E2,i )]
w2,i
⇒ We can compare wages of workers in one country to detect a
shape for φ(E).
Level Accounting
Hall and Jones (1998) describe φ(Ei ) as piecewise linear: For the first
4 years of education the rate of return is 13.4 %, for the next 4 years
10.1% and for education beyond 8 years 6.8 %.

0.134 Ei
 if 0 ≤ Ei ≤ 4
φ(Ei ) = φ(4) + 0.101(Ei − 4) if 4 < Ei ≤ 8

φ(8) + 0.068(Ei − 8) if 8 < Ei

1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Hall and Jones (1998) assume the same function φ(Ei ) for all
countries. Do you think this is plausible?
Levels Accounting
GDP per capita in country X relative to country Z can be decomposed

as follows α
yX (KX /YX ) 1−α hX AX
= α
yZ (KZ /YZ ) 1−α hZ AZ
where y is output per worker, K /Y is the capital-output ratio, h is
human capital-augemented labor (per worker) h = eφE , E are the
years of schooling, and A is the productivity level.
In log-differences:
α
ln(yX /yZ ) = ln(κX /κZ ) + φ(EX − EZ ) + ln(AX /AZ ),
1−α
KX KZ
where κX ≡ YX and κZ ≡ YZ .
Levels Accounting
Hall and Jones (1998) find that productivity differences are the most
important factor to explain the differences in the levels of output per
worker between countries.
Output per worker in the five rich countries is about 30 times higher
than in the five poorest countries (figures for 1988).
I Differences in capital intensity and educational attainment across
countries are only about 2 times higher in the rich countries compared to
the poor.
I Productivity over 8 times higher in the rich countries.
However, differences in education matter as much as capital intensity!
Does educationTHE
matter for economic growth?
EMPIRICS OF ECONOMIC GROWTH 425
(1). Test for unconditional convergence:
TABLE III
TESTS FOR UNCONDITIONALCONVERGENCE
Dependent variable: log difference GDP per working-age person 1960-1985
Sample: Non-oil Intermediate OECD

Observations: 98 75 22
CONSTANT -0.266 0.587 3.69
(0.380) (0.433) (0.68)
ln(Y60) 0.0943 -0.00423 -0.341
(0.0496) (0.05484) (0.079)
R2 0.03 -0.01 0.46
s.e.e. 0.44 0.41 0.18
Implied X -0.00360 0.00017 0.0167
(0.00219) (0.00218) (0.0023)
Note. Standard errors are in parentheses. Y60 is GDP per working-age person in 1960.
Table III from Mankiw, Romer and Weil (1992): ”A contribution to the empirics of economic
essentially zero. growth“,
There Quarterly
is no tendency
Journal of for poor countries to grow
Economics.
faster on average than rich countries.
Table III does show,
Bopjun Gwak (Goethe University Frankfurt)
however, that there is a significant
PMAK - Discussion Session 1 June 4, 2018 10 / 32
426 QUARTERLYJOURNAL OF ECONOMICS
Does education matter for economic growth?
TABLE IV
(2). Test for conditional convergence:
TESTS FOR CONDITIONALCONVERGENCE

CONSTANT 1.93 2.23 2.19
(0.83) (0.86) (1.17)
ln(Y60) -0.141 -0.228 -0.351
(0.052) (0.057) (0.066)
ln(I/GDP) 0.647 0.644 0.392
(0.087) (0.104) (0.176)
ln(n + g + 8) -0.299 -0.464 -0.753
(0.304) (0.307) (0.341)
R72 0.38 0.35 0.62
s.e.e. 0.35 0.33 0.15
Implied X 0.00606 0.0104 0.0173
(0.00182) (0.0019) (0.0019)
Note. Standard errors are in parentheses. Y60 is GDP per working-age person in 1960. The investment and
population growth rates are averages for the period 1960-1985. (g + 8) is assumed to be 0.05.
Table IV from Mankiw, Romer and Weil (1992)

Bopjun Gwak (Goethe University Frankfurt) PMAK - TABLE V 1
Discussion Session June 4, 2018 11 / 32
Does education matter forTABLE
economic
V
growth?
(3). Test for conditional convergence
TESTS FOR with education:
CONDITIONAL CONVERGENCE

CONSTANT 3.04 3.69 2.81
(0.83) (0.91) (1.19)
ln(Y60) -0.289 -0.366 -0.398
(0.062) (0.067) (0.070)
ln(I/GDP) 0.524 0.538 0.335
(0.087) (0.102) (0.174)
ln(n + g + 8) -0.505 -0.551 -0.844
(0.288) (0.288) (0.334)
ln(SCHOOL) 0.233 0.271 0.223
(0.060) (0.081) (0.144)
R2 0.46 0.43 0.65
s.e.e. 0.33 0.30 0.15
Implied X 0.0137 0.0182 0.0203
(0.0019) (0.0020) (0.0020)
Note. Standard errors are in parentheses. Y60 is GDP per working-age person in 1960. The investment and
population growth rates are averages for the period 1960-1985. (g + 8) is assumed to be 0.05. SCHOOL is the
average percentage of the working-age population in secondary school for the period 1960-1985.
Table V from Mankiw, Romer and Weil (1992)

Does education matter for economic growth?
What else could be relevant for economic growth?

Health (life expectancy)
Financial conditions (access to credit)
Institutions, democracy
Openness, Trade policies
...
2 R&D-based endogenous growth model
R&D-based endogenous growth model
The production of final goods:
Yt = Ktα (At LYt )1−α , 0 < α < 1
The production of new technology:
At+1 − At = ρAφt LλAt , ρ > 0 and λ > 0
Capital accumulation
Kt+1 = (1 − δ)Kt + sYt , 0 < δ < 1 and 0 < s < 1
Population growth: Lt+1 = (1 + n)Lt
Allocation of labor across different activities: Lt = LYt + LAt
LAt = sR Lt , 0 < sR < 1
The production of new technology - φ
At+1 − At = ρ Aφt LλAt , ρ > 0 and λ > 0
φ > 0: ”standing on shoulders”, existing technology facilitates the

production of new technology.
φ < 0: ”fishing out”, it becomes more difficult to create new ideas
since the easiest ideas are discovered first.
0 < φ < 1: decreasing technological progress (growth rate of At ).

φ > 1: increasing technological progress.
The production of new technology - λ
At+1 − At = ρAφt LλAt , ρ > 0 and λ > 0
λ < 1: ”stepping on toes”, the more research there is, the more
probable that firms are working on the same ideas.
λ = 1: no ”stepping on toes” effect.
Technological Dynamics
The rate of technological progress ...
At+1 − At
gt ≡ = ρAφ−1
t LλAt (1)
At
can be written as a first-order difference equation
gt+1 = (1 + n)λ gt (1 + gt )φ−1 . (2)
Semi-endogenous Growth Model
0 < φ < 1 and n 6= 0

Growth rate of technology:
gt+1 = (1 + n)λ gt (1 + gt )φ−1 (3)
In the long-run the growth rate of knowledge converges monotonically

to gse = gt+1 = gt
λ
g = (1 + n) 1−φ − 1. (4)
With n = 0 the long-run growth rate of knowledge converges
monotonically to zero
λ
g = 1 1−φ − 1 = 0. (5)
Endogenous Growth Model
φ = 1 and n = 0
Constant growth rate of technology:
g = ρLλAt (6)
Scale effect: If sR is higher, then the growth rate of GDP (per worker)
in the long run is higher.
3 Business Cycles: The short run
Defining Business Cycles
Business Cycles
Fluctuations in the aggregate economic activity
A cycle consists of periods of expansion, followed by periods of
contractions in aggregate economic activity (trough - peak - trough)
A full business cycle lasts for more than one year
Business cycles are recurrent but not periodic
Recesssions
Negative growth of real GDP for at least two quarters
Measuring Business Cycles
GDP can be divided into a growth component gt and a cyclical

component ct
g
Yt = Yt Ytc ,
g
log(Yt ) = log(Yt ) + log(Ytc ) .
| {z } | {z } | {z }
yt gt ct
with E(Ytc ) = 1 and E(ct ) = 0.

ct can be interpreted as the percentage deviation of output from its
trend
g
ct = log(Ytc ) = log(Yt ) − log(Yt ).
How to measure gt ?
The Hodrick-Prescott (HP) Filter
T
X T
X −1
min (yt − gt )2 +λ [(gt+1 − gt ) − (gt − gt−1 )]2
g
t=1 t=2
| {z } | {z }
A B
g
A: minimize the deviation of trend yt to actual output yt .
g
B: minimize the variation in the slope of yt (the slope of the long-run
growth component should not vary much).
The higher λ the higher the penalty for slope differences:
λ = 0: trend equals the original series; λ → ∞: linear trend (for
quarterly data: λ = 1600).
US Real GDP
Trend Component
16.8
16.4
log
16
15.6
15.2
1970 1975 1980 1985 1990 1995 2000 2005 2010
Cyclical Component
5
percentage dev. from trend
−1
−3
−5
1970 1975 1980 1985 1990 1995 2000 2005 2010
Data: St.Louis Fed FRED and SPF survey; US. Bureau of Labor Statistics.
Studying Business Cycles
When studying business cycles we are interested in describing

movements of the cyclical component of output and its subcomponents,
e.g. consumption or investment. Typical questions are:
What would be a normal value for ct ? By how much does ct fluctuate?
Is the cyclical component of some variable X moving in the same
(opposing) direction as the cyclical component of GDP, i.e. is it
cyclical (countercyclical)?
Does some variable X start to rise/fall before (after) GDP does, i.e. is
it a leading (lagging) indicator?
US Private Consumption and Investment
Consumption and Output
5
Yc
3 Cc
1
percent
−1
−3
−5
1970 1975 1980 1985 1990 1995 2000 2005 2010
Investment and Output

20 Yc
Ic
10
percent
−10
−20
1970 1975 1980 1985 1990 1995 2000 2005 2010
Data: St.Louis Fed FRED and SPF survey; US. Bureau of Labor Statistics.
Some Business Cycle facts:
Investment is much more volatile than GDP (about 3 to 4 times). It is

the most unstable component of aggregate demand.
For most countries, private consumption is less volatile than GDP.
Employment is also less volatile than GDP.
Consumption, investment, employment and imports are cyclical, i.e.
comove with GDP.
Real wages tend to be weakly cyclical.
Employment is a lagging variable, i.e. tends to rise after GDP does.

Pmak Session4

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pmak Session4

Uploaded by

Copyright:

Available Formats

Macroeconomics 2 (PMAK)

Goethe University Frankfurt

2 R&D-based endogenous growth model

3 Business Cycle Facts

Question: Why are some countries so much richer than others?

Production function of country i: Yi = Kiα (Ai Hi )1−α

If Ei = 0, then φ(0) = 0 and Hi = e0 Li = Li

GDP per capita in country X relative to country Z can be decomposed

Dependent variable: log difference GDP per working-age person 1960-1985

Sample: Non-oil Intermediate OECD

Dependent variable: log difference GDP per working-age person 1960-1985

Sample: Non-oil Intermediate OECD

Table IV from Mankiw, Romer and Weil (1992)

Dependent variable: log difference GDP per working-age person 1960-1985

Sample: Non-oil Intermediate OECD

Table V from Mankiw, Romer and Weil (1992)

What else could be relevant for economic growth?

The production of final goods:

Yt = Ktα (At LYt )1−α , 0 < α < 1

The production of new technology:

At+1 − At = ρAφt LλAt , ρ > 0 and λ > 0

Kt+1 = (1 − δ)Kt + sYt , 0 < δ < 1 and 0 < s < 1

Population growth: Lt+1 = (1 + n)Lt

Allocation of labor across different activities: Lt = LYt + LAt

LAt = sR Lt , 0 < sR < 1

The production of new technology:

At+1 − At = ρ Aφt LλAt , ρ > 0 and λ > 0

φ > 0: ”standing on shoulders”, existing technology facilitates the

0 < φ < 1: decreasing technological progress (growth rate of At ).

The production of new technology:

At+1 − At = ρAφt LλAt , ρ > 0 and λ > 0

The rate of technological progress ...

gt+1 = (1 + n)λ gt (1 + gt )φ−1 . (2)

0 < φ < 1 and n 6= 0

gt+1 = (1 + n)λ gt (1 + gt )φ−1 (3)

In the long-run the growth rate of knowledge converges monotonically

GDP can be divided into a growth component gt and a cyclical

with E(Ytc ) = 1 and E(ct ) = 0.

When studying business cycles we are interested in describing

Investment and Output

1970 1975 1980 1985 1990 1995 2000 2005 2010

Investment is much more volatile than GDP (about 3 to 4 times). It is

You might also like