ECO5201: Macro I

Lecture 13

Anshumaan Tuteja
Assistant professor of Economics
Ashoka University

October 17, 2024

Lecture outline

* Welfare Theorems

* Overlapping Generations model

* The Baseline OLG Model

– Environment
– Consumption Decisions
– Equilibrium

* Canonical OLG Model

– Canonical Model

* Overaccumulation and Policy

– Overaccumulation and Pareto Optimality

* Conclusions

1 / 32
Welfare Theorems I
We have discussed several models with centralized and decentralized systems. Let’s do
a quick review:

▶ Centralized setup involves a single optimization problem by the social planner

▶ Subject to resources in the economy.
▶ Takes into account any externalities.
▶ This is the pareto optimal equilibrium.

▶ De-centralized setup involves optimization for each sector

▶ Existence of markets for factor inputs.
▶ Externalities are not taken into account.
▶ This is the competitive equilibrium.

▶ Under certain conditions, the two solutions match! They are the topic of
discussion in two famous welfare theorems in Economics

▶ First Welfare Theorem: Every competitive equilibrium is Pareto Optimal.

2 / 32
Welfare Theorems II
▶ Proof of the First Welfare Theorem based on two intuitive ideas.
1. If another allocation Pareto dominates the competitive equilibrium, then it must be
non-affordable in the competitive equilibrium.
2. Profit-maximization implies that any competitive equilibrium already contains the
maximal set of affordable allocations.

▶ In an infinite horizon environment with infinite number of households, we assume:

▶ Local non-satiation
▶ Finite endowments

▶ Second Welfare Theorem: Every Pareto Optimum can be supported as a

competitive equilibrium.

▶ In an infinite horizon environment with infinite number of households, we assume:

▶ Convexity of consumer preferences (well-behaved indifference curves)
▶ Convexity of production set
These assumptions are required to ensure existence of a CE, which is not
guaranteed with non-convexity
3 / 32
Welfare Theorems III
▶ The conditions for the Second Welfare Theorem are more difficult to satisfy than
those for the First.

▶ Also the more important of the two theorems: stronger results that any Pareto
optimal allocation can be decentralized.

▶ The idea that they match is testament to Smith’s invisible hand.

▶ Motivates many to look for the set of Pareto optimal allocations instead of
explicitly characterizing competitive equilibria.

▶ Real power of the Theorem in dynamic macro models comes when we combine it
with models that admit a representative household.

▶ Enables us to characterize the optimal growth allocation that maximizes the

utility of the representative household and assert that this will correspond to a
competitive equilibrium.
4 / 32
Lecture outline

* Welfare Theorems

* Overlapping Generations model

* The Baseline OLG Model

– Environment
– Consumption Decisions
– Equilibrium

* Canonical OLG Model

– Canonical Model

* Overaccumulation and Policy

– Overaccumulation and Pareto Optimality

* Conclusions

5 / 32
Overlapping generations models
▶ How to model households in infinite horizon?
1. “infinitely lived” or consisting of overlapping generations with full altruism linking
generations→infinite planning horizon
2. overlapping generations→finite planning horizon (generally...).
▶ So far, we have looked at models with representative households living finite or
infinite horizons
▶ In many situations, the assumption of a representative household is not
appropriate because
1. households do not have an infinite planning horizon
2. new households arrive (or are born) over time.
▶ Tractable model but inconsistent with overlapping dynasties
▶ A structure of overlapping generations implies older generation affect prices faced
by younger generation

6 / 32
Growth with Overlapping Generations
1. Capture potential interaction of different generations of individuals in the
marketplace;
2. Provide tractable alternative to infinite-horizon representative agent models;
3. Some key implications different from neoclassical growth model;
4. Dynamics in some special cases quite similar to Solow model rather than the
neoclassical model;
5. Generate new insights about the role of national debt and Social Security in the
economy.

7 / 32
Problems of Infinity
▶ An example can show that the First Welfare Theorem does not apply in OLG
models
▶ This is because there are infinite number of HHs and infinite number of
commodities
▶ Source of the problem must be related to the infinite number of commodities.
▶ Extended version of the First Welfare Theorem covers infinite number of
commodities, but only assuming ∞ ∗
P
j=0 pj ωj < ∞ (written with the aggregate
endowment ωj ).
▶ Here theonly endowment is labor, and thus pj∗ = 1 for all j ∈ N, so that
P∞ ∗
j=0 pj ωj = ∞

8 / 32
Lecture outline

* Welfare Theorems

* Overlapping Generations model

* The Baseline OLG Model

– Environment
– Consumption Decisions
– Equilibrium

* Canonical OLG Model

– Canonical Model

* Overaccumulation and Policy

– Overaccumulation and Pareto Optimality

* Conclusions

9 / 32
The Baseline Overlapping Generations Model
▶ Time is discrete and runs to infinity.
▶ Each individual lives for two periods.
▶ Individuals born at time t live for dates t and t + 1.
▶ Assume a general (separable) utility function for individuals born at date t,

U = u (c1,t ) + βu (c2,t+1 ) , (1)

▶ u : R+ → R satisfies the usual Assumptions on utility.

▶ c1 : consumption of the individual born at t when young (at date t).
▶ c2 : consumption when old (at date t + 1).
▶ β ∈ (0, 1) is the discount factor.
▶ Exponential population growth,

Lt = (1 + n)t L0 . (2)

10 / 32
Demographics, Preferences and Technology I
▶ Production side same as before: competitive firms, constant returns to scale
aggregate production function, satisfying Assumptions 1 and 2:

Yt = F (Kt , Lt ) .

▶ Factor markets are competitive.

▶ Individuals can only work in the first period and supply one unit of labor
inelastically, earning wt .
▶ Assume that δ = 1.
▶ k ≡ K /L, f (k) ≡ F (k, 1), and the (gross) rate of return to saving, which equals
the rental rate of capital, is

1 + rt = Rt = f ′ (kt ) , (3)

▶ As usual, the wage rate is

wt = f (kt ) − kt f ′ (kt ) . (4)

11 / 32
Consumption Decisions I
▶ Savings by an individual of generation t, st , is determined as a solution to
max u (c1,t ) + βu (c2,t+1 )
c1,t ,c2,t+1 ,st

subject to
c1,t + st ≤ wt
and
c2,t+1 ≤ Rt+1 st ,
▶ Old individuals rent their savings of time t as capital to firms at time t + 1, and
receive gross rate of return Rt+1 = 1 + rt+1
▶ Since u (·) is strictly increasing, both constraints will hold as equalities.
▶ Thus first-order condition for a maximum can be written in the familiar form of
the consumption Euler equation,
u ′ (c1,t ) = βRt+1 u ′ (c2,t+1 ) . (5)
▶ Solving for consumption and thus for savings,
st = s (wt , Rt+1 ) , (6)
12 / 32
Consumption Decisions II
▶ s : R2+ → R is strictly increasing in its first argument and may be increasing or
decreasing in its second argument.
▶ Total savings in the economy will be equal to

St = st Lt ,

▶ Lt denotes the size of generation t, who are saving for time t + 1.

▶ Since capital depreciates fully after use and all new savings are invested in capital,

Kt+1 = Lt s (wt , Rt+1 ) . (7)

13 / 32
Equilibrium I
Definition A competitive equilibrium can be represented by a sequence of aggregate
capital stocks, individual consumption and factor prices,
{Kt , c1,t , c2,t , Rt , wt }∞ ∞
t=0 , such that the factor price sequence {Rt , wt }t=0
is given by (3) and (4), individual consumption decisions {c1,t , c2,t }∞ t=0
are given by (5) and (6), and the aggregate capital stock, {Kt }∞ t=0 ,
evolves according to (7).
▶ Steady-state equilibrium defined as usual: an equilibrium in which k ≡ K /L is
constant.
▶ To characterize the equilibrium, divide (7) by Lt+1 = (1 + n) Lt ,

s (wt , Rt+1 )
kt+1 = .
1+n
▶ Now substituting for Rt+1 and wt from (3) and (4),

s (f (kt ) − kt f ′ (kt ) , f ′ (kt+1 ))

kt+1 = (8)
1+n
14 / 32
Equilibrium II
▶ This is the fundamental law of motion of the overlapping generations economy.
▶ A steady state is given by a solution to this equation such that kt+1 = kt = k ∗ ,
i.e.,
s (f (k ∗ ) − k ∗ f ′ (k ∗ ) , f ′ (k ∗ ))
k∗ = (9)
1+n
▶ Since the savings function s (·, ·) can take any form, the difference equation (8)
can lead to quite complicated dynamics, and multiple steady states are possible.
▶ Suppose that the utility functions take the familiar CRRA form:
1−θ 1−θ
!
c1,t −1 c2,t+1 −1
U= +β , (10)
1−θ 1−θ

where θ > 0 and β ∈ (0, 1).

▶ Technology is Cobb-Douglas,
f (k) = k α

15 / 32
Restrictions on Utility and Production Functions I
▶ The CRRA utility simplifies the first-order condition for consumer optimization,
c2,t+1
= (βRt+1 )1/θ .
c1,t
▶ This Euler equation can be alternatively expressed in terms of savings as

st−θ βRt+1
1−θ
= (wt − st )−θ , (11)

▶ Gives the following equation for the saving rate:

wt
st = , (12)
ψt+1

where
−(1−θ)/θ
ψ ≡ [1 + β −1/θ Rt+1 ] > 1,
▶ Ensures that savings are always less than earnings.

16 / 32
Restrictions on Utility and Production Functions II
▶ The impact of factor prices on savings is summarized by the following and
derivatives:
∂st 1
sw ≡ = ∈ (0, 1) ,
∂wt ψt+1
 
∂st 1−θ st
sR ≡ = (βRt+1 )−1/θ .
∂Rt+1 θ ψt+1
▶ Since ψ(.) > 1, we also have that 0 < sw < 1.
▶ Moreover, in this case sR > 0 if θ > 1, sR < 0 if θ < 1, and sR = 0 if θ = 1.
▶ Reflects counteracting influences of income and substitution effects.
▶ Case of θ = 1 (log preferences) is of special importance, may deserve to be called
the canonical overlapping generations model.

17 / 32
Restrictions on Utility and Production Functions III
▶ Equation (8) implies
st
kt+1 = (13)
(1 + n)
wt
= ,
(1 + n)ψt+1
▶ Or more explicitly,

f (kt ) − kt f ′ (kt )
kt+1 = (14)
(1 + n) [1 + β −1/θ f ′ (kt+1 )−(1−θ)/θ ]
▶ The steady state then involves a solution to the following implicit equation:

f (k ∗ ) − k ∗ f ′ (k ∗ )
k∗ = .
(1 + n) [1 + β −1/θ f ′ (k ∗ )−(1−θ)/θ ]

18 / 32
Restrictions on Utility and Production Functions IV
▶ Now using the Cobb-Douglas formula, steady state is the solution to the equation
h
(θ−1)/θ i
(1 + n) 1 + β −1/θ α(k ∗ )α−1 = (1 − α)(k ∗ )α−1 . (15)

▶ For simplicity, define R ∗ ≡ α(k ∗ )α−1 as the marginal product of capital in

steady-state, in which case, (15) can be rewritten as
h i 1−α
(1 + n) 1 + β −1/θ (R ∗ )(θ−1)/θ = R ∗. (16)
α
▶ Steady-state value of R ∗ , and thus k ∗ , can now be determined from equation
(16), which always has a unique solution.
▶ To investigate the stability, substitute for the Cobb-Douglas production function
in (14)
(1 − α) k (t)α
k (t + 1) = . (17)
(1 + n) [1 + β −1/θ (αk(t + 1)α−1 )−(1−θ)/θ ]
▶ In this particular (well-behaved) case, equilibrium dynamics are very similar to the
basic Solow model 19 / 32
Lecture outline

* Welfare Theorems

* Overlapping Generations model

* The Baseline OLG Model

– Environment
– Consumption Decisions
– Equilibrium

* Canonical OLG Model

– Canonical Model

* Overaccumulation and Policy

– Overaccumulation and Pareto Optimality

* Conclusions

20 / 32
Canonical Model I
▶ Even the model with CRRA utility and Cobb-Douglas production function is
relatively messy. Many of the applications use log preferences (θ = 1).
▶ Changes in the interest rate (and thus in the capital-labor ratio of the economy)
have no effect on the saving rate: Income and substitution effects exactly cancel
each other.
▶ Utility of the household and generation t is,

U = logc1,t + βlogc2,t+1 , (18)

▶ β ∈ (0, 1) (even though β ≥ 1 could be allowed).
▶ Consumption Euler equation:
c2,t+1
= βRt+1
c1,t
▶ Savings should satisfy the equation
β
st = wt , (19)
1+β
▶ Constant saving rate, equal to β/ (1 + β), out of labor income for each individual.
21 / 32
Canonical Model II
▶ Combining this with the capital accumulation equation (8),
st
kt+1 =
(1 + n)
βwt
=
(1 + n) (1 + β)
β (1 − α) [kt ]α
= ,
(1 + n) (1 + β)
▶ Second line uses (19) and last uses that, given competitive factor markets,
wt = (1 − α) [kt ]α .
▶ There exists a unique steady state with
  1
∗ β (1 − α) 1−α
k = . (20)
(1 + n) (1 + β)
▶ Equilibrium dynamics are identical to those of the basic Solow model and
monotonically converge to k ∗ .
22 / 32
Figure: Equilibrium dynamics in the canonical overlapping generations model.

23 / 32
Lecture outline

* Welfare Theorems

* Overlapping Generations model

* The Baseline OLG Model

– Environment
– Consumption Decisions
– Equilibrium

* Canonical OLG Model

– Canonical Model

* Overaccumulation and Policy

– Overaccumulation and Pareto Optimality

* Conclusions

24 / 32
Overaccumulation I
▶ Compare the overlapping-generations equilibrium to the choice of a social planner
wishing to maximize a weighted average of all generations’ utilities.
▶ Suppose that the social planner maximizes
X∞
βSt Ut
t=0
▶ βS is the discount factor of the social planner, which reflects how she values the
utilities of different generations.
▶ Substituting from (1), this implies:
∞
X
βSt (u (c1,t ) + βu (c2,t+1 ))
t=0
subject to the resource constraint
F (Kt , Lt ) = Kt+1 + Lt c1,t + Lt−1 c2,t
▶ Dividing this by Lt and using (2),
c2,t
f (kt ) = (1 + n) kt+1 + c1,t + .
1+n
25 / 32
Overaccumulation II
▶ Social planner’s maximization problem then implies the following first-order
necessary condition:
u ′ (c1,t ) = βf ′ (kt+1 ) u ′ (c2,t+1 ) .
▶ Since Rt+1 = f ′ (kt+1 ), this is identical to (5).
▶ Not surprising: allocate consumption of a given individual in exactly the same way
as the individual himself would do.
▶ No “market failures” in the over-time allocation of consumption at given prices.
▶ However, the allocations across generations may differ from the competitive
equilibrium: planner is giving different weights to different generations
▶ In particular, competitive equilibrium is Pareto suboptimal when k ∗ > kgold ,

26 / 32
Overaccumulation III
▶ When k ∗ > kgold , reducing savings can increase consumption for every generation.
▶ More specifically, note that in steady state

f (k ∗ ) − (1 + n)k ∗ = c1∗ + (1 + n)−1 c2∗

≡ c ∗,

▶ First line follows by national income accounting, and second defines c ∗ .

▶ Therefore
∂c ∗
= f ′ (k ∗ ) − (1 + n)
∂k ∗
▶ kgold is defined as
f ′ (kgold ) = 1 + n.

27 / 32
Overaccumulation IV
▶ Now if k ∗ > kgold , then ∂c ∗ /∂k ∗ < 0: reducing savings can increase (total)
consumption for everybody.
▶ If this is the case, the economy is referred to as dynamically inefficient—it
involves overaccumulation.
▶ Another way of expressing dynamic inefficiency is that

r ∗ < n,

▶ Suppose we start from steady state at time T with k ∗ > kgold .

28 / 32
Overaccumulation V
▶ Consider the following variation: change next period’s capital stock by −∆k,
where ∆k > 0, and from then on, we immediately move to a new steady state
(clearly feasible).
▶ This implies the following changes in consumption levels:

∆cT = (1 + n) ∆k > 0
= − f ′ (k ∗ − ∆k) − (1 + n) ∆k for all t > T


∆ct

▶ The first expression reflects the direct increase in consumption due to the decrease
in savings.
▶ In addition, since k ∗ > kgold , for small enough ∆k, f ′ (k ∗ − ∆k) − (1 + n) < 0,
thus ∆ct > 0 for all t ≥ T .
▶ The increase in consumption for each generation can be allocated equally during
the two periods of their lives, thus necessarily increasing the utility of all
generations.

29 / 32
Interpretation
▶ Intuition for dynamic inefficiency:
▶ Individuals who live at time t face prices determined by the capital stock with which
they are working.
▶ Capital stock is the outcome of actions taken by previous generations.
▶ Pecuniary externality from the actions of previous generations affecting welfare of
current generation.
▶ Pecuniary externalities typically second-order and do not matter for welfare.
▶ But not when an infinite stream of newborn agents joining the economy are affected.
▶ It is possible to rearrange in a way that these pecuniary externalities can be exploited.
▶ Complementary intuition:
▶ Dynamic inefficiency arises from overaccumulation.
▶ Results from current young generation needs to save for old age.
▶ However, the more they save, the lower is the rate of return and may encourage to
save even more.
▶ Effect on future rate of return to capital is a pecuniary externality on next generation
▶ If alternative ways of providing consumption to individuals in old age were
introduced, overaccumulation could be ameliorated.

30 / 32
Lecture outline

* Welfare Theorems

* Overlapping Generations model

* The Baseline OLG Model

– Environment
– Consumption Decisions
– Equilibrium

* Canonical OLG Model

– Canonical Model

* Overaccumulation and Policy

– Overaccumulation and Pareto Optimality

* Conclusions

31 / 32
Conclusions
▶ Overlapping generations of the more realistic than infinity-lived representative
agents.
▶ Models with overlapping generations fall outside the scope of the First Welfare
Theorem:
▶ they were partly motivated by the possibility of Pareto suboptimal allocations.
▶ Equilibria may be “dynamically inefficient” and feature overaccumulation:
unfunded Social Security can ameliorate the problem.
▶ Overaccumulation and Pareto suboptimality: pecuniary externalities created on
individuals that are not yet in the marketplace.

32 / 32