Lecture2 SolowModel
Lecture2 SolowModel
Simon Mongey
Winter, 2021
1
This lecture
• Questions
2
Growth facts
1. Share of income to capital and labor are approximately constant
Labor payments Capital payments
≈ 0.65 , ≈ 0.35
GDP GDP
3. Capital per worker and output per worker grow at constant rates
3
1. Labor share
Labor share was previously constant but has recently fallen. Lots of research on this
currently!
4
2. Capital / Output ratio
Capital/Output and the growth rates of Capital/Worker and Output/Worker are
approximately constant.
5
3. Capital- and Output- per worker
Capital/Output and the growth rates of Capital/Worker and Output/Worker are
approximately constant.
6
Solow model - Environment
• Technology
- Firms operate CRS technology
1−α
Yt = F (Kt , Nt ) = Ktα (At Nt )
U (Ct , Nt ) = Ct
• Resources
- Goods: Yt = Ct + It .
7
Competitive factor pricing - Wt
- Firms are competitive
- Profit maximization problem taking prices as given
Yt = F (Kt , At Nt ) (1)
Kt+1 = (1 − δ)Kt + It (2)
Yt = Ct + It (3)
Ct = (1 − ξ)Yt (4)
12
Per-worker
- Production function
Kt Kt
Yt = F (Kt , At Nt ) = At Nt F , 1 = At Nt f
At Nt At Nt
yt = f (kt )
kt+1 (1 + γA )(1 + γN ) = (1 − δ)kt + it
yt = ct + it
ct = (1 − ξ)yt
13
Solution
• Using this we have four equations in four unknowns {kt+1 , yt , ct , it }
yt = f (kt )
kt+1 (1 + γA )(1 + γN ) = (1 − δ)kt + it
yt = ct + it
ct = (1 − ξ)yt
• Solving these
1−δ ξ
kt+1 = kt + f (kt ) = g(kt ) (∗)
(1 + γA )(1 + γN ) (1 + γA )(1 + γN )
14
Steady-state
15
Steady-state
• Is there some k that solves k = g(k)?
1−δ ξ
k= k+ f (k)
(1 + γA )(1 + γN ) (1 + γA )(1 + γN )
ξ
k= f k [Fixed point equation]
γA + γN + δ
α
• In the case that f (k) = k : [Function only of parameters! {ξ, γA , γN , δ, α}]
1
1−α
ξ α ξ
k= k → k=
γA + γN + δ γA + γN + δ
• Comparative statics
- Increasing in ξ, decreasing in γA , γN , δ.
- Convex in ξ, less so if α is smaller → Increasing in α
16
Transition dynamics around steady state
• Is it stable?
1−δ ξ
kt+1 = g(kt ) = kt + f (kt )
(1 + γA )(1 + γN ) (1 + γA )(1 + γN )
• Using k̄ = g(k̄)
0
g (k̄)k̄ b kt − k̄
kt+1 ≈
b kt , kt :=
b
g(k̄) k̄
So if εg,k < 1, then the gap from steady-state shrinks |k̂t+1 | < |k̂t |.
∗ Show that this is stable iff the elasticity εf,k < 1. What parameter
determines this? Conclude that growth b kt is higher the further away
from steady-state. Compare U.S. vs. Japan after WWII
17
Transition dynamics around steady state
1−δ ξ
kt+1 = g(kt ) = kt + f (kt )
(1 + γA )(1 + γN ) (1 + γA )(1 + γN )
• Linearize around kt
So if εg,k < 1, then growth rate declines: |∆ log kt+1 | < |∆ log kt |.
∗ Show that if yt = ktα , then |∆ log yt+1 | < |∆ log yt |.
18
Capital per worker
19
Steady state
20
Transition dynamics
21
Steady state - Intuition
22
Dynamics following a shock
23
Dynamics following a shock
24
Positive implications - Growth facts
1. Share of income to labor is constant and Wt grows at constant rate
Wt Nt Wt+1 (1 − α)Yt+1 /Nt+1
=1−α , = = γA
Yt Wt (1 − α)Yt /Nt
Kt Kt /At Nt k
= =
Yt Yt /At Nt f (k)
3. Capital per worker and output per worker grow at constant rates
25
Positive implications - Growth facts
1. Share of income to labor is constant and Wt grows at constant rate
Wt Nt Wt+1 (1 − α)Yt+1 /Nt+1
=1−α , = = γA
Yt Wt (1 − α)Yt /Nt
Kt Kt /At Nt k
= =
Yt Yt /At Nt f (k)
3. Capital per worker and output per worker grow at constant rates
28
Golden rule
29
Data
Are countries all converging to the same k?
30
Data
Are countries all converging to the same k?
31
Data
Are countries all converging to the same k?
32
Data
Are countries all converging to the same k?
33
Covered
• Growth facts
• Next
- TA - More practice linearization
- Lecture - Endogenizing savings ... the neo-classical growth model.
34