Lecture Note 3

Yang Tang

September 13, 2017

1 / 16
Outline

I Diamond Overlapping Generation Model

I Two-sector Neoclassical Growth Model

2 / 16
Diamond Overlapping Generation Model

3 / 16
Main Features

I This is an exogenous economic growth model, but saving

decision is endogenous.

I This is a discrete time model.

I In term of welfare analysis, this model has different results

compared with Ramsey model.

4 / 16
General Environment

I Time is discrete with t = 0, 1, 2....

I Assume that each individual lives two period. Lt individuals

are born in period t. Population grows at rate n:
Lt = Lt1 (1 + n).

I Let c1t and c2t denote the consumption in period t of young

and old individuals. The life-time utility of an individual born
at t is given as:
1 1
c1t 1 c2t+1
Ut = +
1 1+1
I Young individuals inelastically provide 1 unit of labor. Old
individuals do not engage in production.

5 / 16
Production

I Capital and labor are the input of producing output each period:
Yt = F (Kt , At Lt ).
I Markets are competitive, and thus effective-labor and capital earn their
marginal product:
F (Kt , At Lt )
rt =
Kt
F (Kt , At Lt )
wt =
At Lt
I In the beginning of t = 0, the capital owned by the old and the labor
supplied by the young are combined to produce output. At then end of
t = 0, the old consume their capital income then die. The young divide
their labor income between consumption and saving. They carry their
saving to the next period when they are old.
I Capital stock in period t + 1 is thus:

Kt+1 = (wt At c1t )Lt

6 / 16
Individual Behavior
I Individuals life-time budget constraint can be written as:
c2t+1
c1t + = At wt
1 + rt+1
I Individuals optimization problem can be characterized as
follows:
1 1
c1t 1 c2t+1
max +
c1t ,c2t+1 1 1+1
c2t+1
s.t c1t + = At wt (2.45 budget
1 + rt+1 constraint)
I The following can be derived by solving the problem above:

c2t+1 1 + rt+1

= (2.47)
c1t 1+

7 / 16
The Dynamics of the Economy

I It can be solved that:

(1 + )1/
c1t = At wt (2.54)
(1 + )1/ + (1 + rt+1 )(1)/
I Since st = Kt+1 , therefore capital stock accumulate over time
as follows:

(1 + rt+1 )(1)/
Kt+1 = At wt Lt
(1 + )1/ + (1 + rt+1 )(1)/
F (Kt+1 ,At+1 Lt+1 ) F (Kt ,At Lt )
where rt+1 = Kt+1 and wt = At Lt

8 / 16
Logarithmic Utility and Cobb-Douglas Production

I Suppose utility function and production function take the

following forms:

u(c) = log (c)

F (Kt , At Lt ) = Kt (At Lt )1

where technology grows at a constant rate of g .

Kt
I define kt = A t Lt , then we have:

1 1
kt+1 = (1 )kt
(1 + n)(1 + g ) 2 +

9 / 16
Diagram
Steady state kt+1 = kt = k is globally stable.

10 / 16
Two-sector Neoclassical Growth Model

11 / 16
Motivation

I If consumption and capital goods are two types of goods

I If both labor and capital are inputs to produce both goods
I How labor and capital are allocated across two sectors to
sustain a long-run growth?

12 / 16
General Environment
I Time is discrete and infinite with t = 0, 1, 2...
I Population size grows at rate n: Lt+1 = Lt (1 + n).
I There are two production sector to produce consumption and
capital goods, respectively.
I Both capital and labor are input for production in each sector,
and production function takes form:
1
Yt = AK1t N1t
1
Xt = BK2t N2t

where
I Yt and Xt denote the output of consumption and capital
goods, respectively.
I K1t and K2t are capital goods used in consumption and capital
goods sector, respectively.
I N1t and N2t are labor used in consumption and capital goods
sector, respectively.
13 / 16
Representative Agents Optimization Problem
I Representative agent is endowed with K0 unit of capital goods in t = 0.
I The representative agents optimization problem can be characterized as
follows:

X
max t log(ct )
{K1t ,K2t ,N1t ,N2t }
t=0 t=0
s.t. ct = Yt
Kt+1 = Kt (1 ) + BK2t N2t
1

I Normalize consumption goods price to be 1, and denote capital goods

price to be pt . In equilibrium, we must have:

AK1t1 N1t
1
= Bpt K2t1 N2t
1

(1 )AK1t N1t = Bpt (1 )K2t N2t
I Therefore, we have:
K1t K2t
=
N1t N2t
I Denote t to be the fraction of capital (labor) used in consumption goods
production sector.

14 / 16
Solving the Optimization Problem
I Representative agents optimization problem can be rewritten as follows:

X
max t log(t AKt Nt1 )
{Kt+1 ,t }
t=0 t=0

s.t. Kt+1 = Kt (1 ) + (1 t )BKt Nt1

I Set up the langrange gives:

X
L= t {log(t AKt Nt1 ) + t [Kt (1 ) + (1 t )BKt Nt1 Kt+1 ]}
t=0

I First order condition with respect to t and Kt+1 gives:

1
= t BKt Nt1 (1)
t
1 1 1
t [ + t+1 (1 ) + t+1 (1 t+1 )BKt+1 Nt+1 ] = (02)
Kt+1
I From (1) we have:
1
t = (3)
t BKt Nt1

15 / 16
Solving the Optimization Problem

I Replace (2) into (1) gives:

1 1 1 t+1
= [ + + ]
t BKt Nt1
t+1 BKt+1 1
Nt+1 t+1 Kt+1 Kt+1

I Denote kt = Kt
Nt
, then we have:

1 1 1 t+1
(1 + n) = [ + + ]
t Bkt
t+1 Bkt+1 t+1 kt+1 kt+1
I On Balanced growth path, kt = k and t = , then we have:

1+n 1 (1 )
= [ + + ]
Bk Bk k k
k(1 + n) = k(1 ) + B(1 )k
I The above equations have unique solution (k , ).

16 / 16