14.

452 Economic Growth: Lecture 8, Overlapping

Generations

Daron Acemoglu

MIT

November 22, 2011.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 1 / 53

 Growth with Overlapping Generations Growth with Overlapping Generations

Growth with Overlapping Generations

In many situations, the assumption of a representative household is
not appropriate because
1 households do not have an in…nite planning horizon
2 new households arrive (or are born) over time.
New economic interactions: decisions made by older “generations”
will a¤ect the prices faced by younger “generations”.
Overlapping generations models
1 Capture potential interaction of di¤erent generations of individuals in
the marketplace;
2 Provide tractable alternative to in…nite-horizon representative agent
models;
3 Some key implications di¤erent 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.
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 2 / 53
 Growth with Overlapping Generations Problems of In…nity

Problems of In…nity I

Static economy with countably in…nite number of households, i 2 N

Countably in…nite number of commodities, j 2 N.
All households behave competitively (alternatively, there are M
households of each type, M is a large number).
Household i has preferences:

ui = cii + cii+1 ,

cji denotes the consumption of the jth type of commodity by

household i.
Endowment vector ω of the economy: each household has one unit
endowment of the commodity with the same index as its index.
Choose the price of the …rst commodity as the numeraire, i.e., p0 = 1.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 3 / 53

 Growth with Overlapping Generations Problems of In…nity

Problems of In…nity II

Proposition In the above-described economy, the price vector p̄ such that

p̄j = 1 for all j 2 N is a competitive equilibrium price vector
and induces an equilibrium with no trade, denoted by x̄.

Proof:
At p̄, each household has income equal to 1.
Therefore, the budget constraint of household i can be written as

cii + cii+1 1.

This implies that consuming own endowment is optimal for each

household,
Thus p̄ and no trade, x̄, constitute a competitive equilibrium.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 4 / 53

 Growth with Overlapping Generations Problems of In…nity

Problems of In…nity III

However, this competitive equilibrium is not Pareto optimal. Consider

alternative allocation, x̃:
Household i = 0 consumes its own endowment and that of household 1.
All other households, indexed i > 0, consume the endowment of than
neighboring household, i + 1.
All households with i > 0 are as well o¤ as in the competitive
equilibrium (p̄, x̄ ).
Individual i = 0 is strictly better-o¤.
Proposition In the above-described economy, the competitive equilibrium
at (p̄, x̄ ) is not Pareto optimal.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 5 / 53

 Growth with Overlapping Generations Problems of In…nity

Problems of In…nity IV

Source of the problem must be related to the in…nite number of

commodities.
Extended version of the First Welfare Theorem covers in…nite number
of commodities, but only assuming ∑j∞=0 pj ω j < ∞ (written with the
aggregate endowment ω j ).
Here theonly endowment is labor, and thus pj = 1 for all j 2 N, so
that ∑j∞=0 pj ω j = ∞ (why?).
This abstract economy is “isomorphic” to the baseline overlapping
generations model.
The Pareto suboptimality in this economy will be the source of
potential ine¢ ciencies in overlapping generations model.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 6 / 53

 Growth with Overlapping Generations Problems of In…nity

Problems of In…nity V

Second Welfare Theorem did not assume ∑j∞=0 pj ω j < ∞.

Instead, it used convexity of preferences, consumption sets and
production possibilities sets.
This exchange economy has convex preferences and convex
consumption sets:
Pareto optima must be decentralizable by some redistribution of
endowments.
Proposition In the above-described economy, there exists a reallocation
of the endowment vector ω to ω ~ , and an associated
competitive equilibrium (p̄, x̃ ) that is Pareto optimal where x̃
is as described above, and p̄ is such that p̄j = 1 for all j 2 N.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 7 / 53

 Growth with Overlapping Generations Problems of In…nity

Proof of Proposition
Consider the following reallocation of ω: endowment of household
i 1 is given to household i 1.
At the new endowment vector ω ~ , household i = 0 has one unit of good
j = 0 and one unit of good j = 1.
Other households i have one unit of good i + 1.
At the price vector p̄, household 0 has a budget set
c00 + c11 2,
thus chooses c00 = c10 = 1.
All other households have budget sets given by
cii + cii+1 1,
Thus it is optimal for each household i > 0 to consume one unit of
the good cii+1
Thus x̃ is a competitive equilibrium.
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 8 / 53
 The Baseline OLG Model Environment

The Baseline Overlapping Generations Model

Time is discrete and runs to in…nity.

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 (t ) = u (c1 (t )) + βu (c2 (t + 1)) , (1)
u : R+ ! R satis…es the usual Assumptions on utility.
c1 (t ): consumption of the individual born at t when young (at date
t).
c2 (t + 1): consumption when old (at date t + 1).
β 2 (0, 1) is the discount factor.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 9 / 53

 The Baseline OLG Model Environment

Demographics, Preferences and Technology I

Exponential population growth,

L (t ) = (1 + n )t L (0) . (2)

Production side same as before: competitive …rms, constant returns to

scale aggregate production function, satisfying Assumptions 1 and 2:

Y (t ) = F (K (t ) , L (t )) .

Factor markets are competitive.

Individuals can only work in the …rst period and supply one unit of
labor inelastically, earning w (t ).

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 10 / 53

 The Baseline OLG Model Environment

Demographics, Preferences and Technology II

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 + r (t ) = R (t ) = f 0 (k (t )) , (3)

As usual, the wage rate is

w (t ) = f (k (t )) k (t ) f 0 (k (t )) . (4)

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 11 / 53

 The Baseline OLG Model Consumption Decisions

Consumption Decisions I

Savings by an individual of generation t, s (t ), is determined as a

solution to

max u (c1 (t )) + βu (c2 (t + 1))

c1 (t ),c2 (t +1 ),s (t )

subject to
c1 ( t ) + s ( t ) w (t )
and
c2 ( t + 1 ) R (t + 1) s (t ) ,
Old individuals rent their savings of time t as capital to …rms at time
t + 1, and receive gross rate of return R (t + 1) = 1 + r (t + 1)
Second constraint incorporates notion that individuals only spend
money on their own end of life consumption (no altruism or bequest
motive).
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 12 / 53
 The Baseline OLG Model Consumption Decisions

Consumption Decisions II

No need to introduce s (t ) 0, since negative savings would violate

second-period budget constraint (given c2 (t + 1) 0).
Since u ( ) is strictly increasing, both constraints will hold as
equalities.
Thus …rst-order condition for a maximum can be written in the
familiar form of the consumption Euler equation,

u 0 (c1 (t )) = βR (t + 1) u 0 (c2 (t + 1)) . (5)

Problem of each individual is strictly concave, so this Euler equation

is su¢ cient.
Solving for consumption and thus for savings,

s (t ) = s (w (t ) , R (t + 1)) , (6)

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 13 / 53

 The Baseline OLG Model Consumption Decisions

Consumption Decisions III

s : R2+ ! R is strictly increasing in its …rst argument and may be

increasing or decreasing in its second argument.
Total savings in the economy will be equal to

S (t ) = s (t ) L (t ) ,

L (t ) 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,

K (t + 1) = L (t ) s (w (t ) , R (t + 1)) . (7)

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 14 / 53

 The Baseline OLG Model Equilibrium

Equilibrium I
De…nition A competitive equilibrium can be represented by a sequence
of aggregate capital stocks, individual consumption and
factor prices,
fK (t ) , c1 (t ) , c2 (t ) , R (t ) , w (t )gt∞=0 , such that the factor
price sequence fR (t ) , w (t )gt∞=0 is given by (3) and (4),
individual consumption decisions fc1 (t ) , c2 (t )gt∞=0 are
given by (5) and (6), and the aggregate capital stock,
fK (t )gt∞=0 , evolves according to (7).
Steady-state equilibrium de…ned as usual: an equilibrium in which
k K /L is constant.
To characterize the equilibrium, divide (7) by
L(t + 1) = (1 + n ) L (t ),
s (w (t ) , R (t + 1))
k (t + 1) = .
1+n
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 15 / 53
 The Baseline OLG Model Equilibrium

Equilibrium II

Now substituting for R (t + 1) and w (t ) from (3) and (4),

s (f (k (t )) k (t ) f 0 (k (t )) , f 0 (k (t + 1)))
k (t + 1) = (8)
1+n
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
k (t + 1) = k (t ) = k , i.e.,

s (f (k ) k f 0 (k ) , f 0 (k ))
k = (9)
1+n
Since the savings function s ( , ) can take any form, the di¤erence
equation (8) can lead to quite complicated dynamics, and multiple
steady states are possible.
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 16 / 53
 The Baseline OLG Model Special Cases

Restrictions on Utility and Production Functions I

Suppose that the utility functions take the familiar CRRA form:
!
c1 ( t ) 1 θ 1 c2 ( t + 1 ) 1 θ 1
U (t ) = +β , (10)
1 θ 1 θ

where θ > 0 and β 2 (0, 1).

Technology is Cobb-Douglas,

f (k ) = k α

The rest of the environment is as described above.

The CRRA utility simpli…es the …rst-order condition for consumer
optimization,
c2 ( t + 1 )
= ( βR (t + 1))1/θ .
c1 ( t )
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 17 / 53
 The Baseline OLG Model Special Cases

Restrictions on Utility and Production Functions II

This Euler equation can be alternatively expressed in terms of savings

as
s (t ) θ βR (t + 1)1 θ = (w (t ) s (t )) θ , (11)
Gives the following equation for the saving rate:

w (t )
s (t ) = , (12)
ψ (t + 1)

where
1/θ (1 θ )/θ
ψ (t + 1) [1 + β R (t + 1) ] > 1,
Ensures that savings are always less than earnings.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 18 / 53

 The Baseline OLG Model Special Cases

Restrictions on Utility and Production Functions III

The impact of factor prices on savings is summarized by the following

and derivatives:
∂s (t ) 1
sw = 2 (0, 1) ,
∂w (t ) ψ (t + 1)
∂s (t ) 1 θ 1/θ s (t )
sR = ( βR (t + 1)) .
∂R (t + 1) θ ψ (t + 1)

Since ψ (t + 1) > 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.
Re‡ects counteracting in‡uences of income and substitution e¤ects.
Case of θ = 1 (log preferences) is of special importance, may deserve
to be called the canonical overlapping generations model.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 19 / 53

 The Baseline OLG Model Special Cases

Restrictions on Utility and Production Functions IV

Equation (8) implies

s (t )
k (t + 1) = (13)
(1 + n )
w (t )
= ,
(1 + n ) ψ (t + 1)
Or more explicitly,
f (k (t )) k (t ) f 0 (k (t ))
k (t + 1) = (1 θ )/θ
(14)
1/θ 0
(1 + n ) [1 + β f (k (t + 1)) ]
The steady state then involves a solution to the following implicit
equation:
f (k ) k f 0 (k )
k = .
(1 + n) [1 + β 1/θ f 0 (k ) (1 θ )/θ ]
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 20 / 53
 The Baseline OLG Model Special Cases

Restrictions on Utility and Production Functions V

Now using the Cobb-Douglas formula, steady state is the solution to
the equation
h i
(θ 1 )/θ
(1 + n) 1 + β 1/θ α(k )α 1 = (1 α)(k )α 1 . (15)

For simplicity, de…ne 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 ]

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 21 / 53

 The Baseline OLG Model Special Cases

Restrictions on Utility and Production Functions VI

Proposition In the overlapping-generations model with two-period lived

households, Cobb-Douglas technology and CRRA
preferences, there exists a unique steady-state equilibrium
with the capital-labor ratio k given by (15), this
steady-state equilibrium is globally stable for all k (0) > 0.

In this particular (well-behaved) case, equilibrium dynamics are very

similar to the basic Solow model
Figure shows that convergence to the unique steady-state
capital-labor ratio, k , is monotonic.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 22 / 53

 Canonical OLG Model Canonical Model

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).
Income and substitution e¤ects exactly cancel each othe: changes in
the interest rate (and thus in the capital-labor ratio of the economy)
have no e¤ect on the saving rate.
Structure of the equilibrium is essentially identical to the basic Solow
model.
Utility of the household and generation t is,

U (t ) = log c1 (t ) + β log c2 (t + 1) , (18)

β 2 (0, 1) (even though β 1 could be allowed).

Again f (k ) = k α.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 23 / 53

 Canonical OLG Model Canonical Model

Canonical Model II

Consumption Euler equation:

c2 ( t + 1 )
= βR (t + 1)
c1 ( t )

Savings should satisfy the equation

β
s (t ) = w (t ) , (19)
1+β

Constant saving rate, equal to β/ (1 + β), out of labor income for

each individual.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 24 / 53

 Canonical OLG Model Canonical Model

Canonical Model III

Combining this with the capital accumulation equation (8),
s (t )
k (t + 1) =
(1 + n )
βw (t )
=
(1 + n ) (1 + β )
α
β (1 α) [k (t )]
= ,
(1 + n ) (1 + β )
Second line uses (19) and last uses that, given competitive factor
markets, w (t ) = (1 α) [k (t )]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 .
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 25 / 53
 Canonical OLG Model Canonical Model

k(t+1)
45°

k*

k(t)
0 k(0) k* k’(0)

Figure: Equilibrium dynamics in the canonical overlapping generations model.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 26 / 53

 Canonical OLG Model Canonical Model

Canonical Model IV

Proposition In the canonical overlapping generations model with log

preferences and Cobb-Douglas technology, there exists a
unique steady state, with capital-labor ratio k given by
(20). Starting with any k (0) 2 (0, k ), equilibrium dynamics
are such that k (t ) " k , and starting with any k 0 (0) > k ,
equilibrium dynamics involve k (t ) # k .

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 27 / 53

 Overaccumulation and Policy Overaccumulation and Pareto Optimality

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
∞
∑ βtS U (t )
t =0

βS is the discount factor of the social planner, which re‡ects how she
values the utilities of di¤erent generations.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 28 / 53

 Overaccumulation and Policy Overaccumulation and Pareto Optimality

Overaccumulation II

Substituting from (1), this implies:

∞
∑ βtS (u (c1 (t )) + βu (c2 (t + 1)))
t =0

subject to the resource constraint

F (K (t ) , L (t )) = K (t + 1) + L (t ) c1 (t ) + L (t 1 ) c2 ( t ) .

Dividing this by L (t ) and using (2),

c2 ( t )
f (k (t )) = (1 + n) k (t + 1) + c1 (t ) + .
1+n

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 29 / 53

 Overaccumulation and Policy Overaccumulation and Pareto Optimality

Overaccumulation III

Social planner’s maximization problem then implies the following

…rst-order necessary condition:

u 0 (c1 (t )) = βf 0 (k (t + 1)) u 0 (c2 (t + 1)) .

Since R (t + 1) = f 0 (k (t + 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 di¤er from the
competitive equilibrium: planner is giving di¤erent weights to
di¤erent generations
In particular, competitive equilibrium is Pareto suboptimal when
k > kgold ,
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 30 / 53
 Overaccumulation and Policy Overaccumulation and Pareto Optimality

Overaccumulation IV

When k > kgold , reducing savings can increase consumption for

every generation.
More speci…cally, note that in steady state
1
f (k ) (1 + n )k = c1 + ( 1 + n ) c2
c ,

First line follows by national income accounting, and second de…nes

c .
Therefore
∂c
= f 0 (k ) (1 + n )
∂k
kgold is de…ned as
f 0 (kgold ) = 1 + n.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 31 / 53

 Overaccumulation and Policy Overaccumulation and Pareto Optimality

Overaccumulation V

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
ine¢ cient— it involves overaccumulation.
Another way of expressing dynamic ine¢ ciency is that

r < n,

Recall in in…nite-horizon Ramsey economy, transversality condition

required that r > g + n.
Dynamic ine¢ ciency arises because of the heterogeneity inherent in
the overlapping generations model, which removes the transversality
condition.
Suppose we start from steady state at time T with k > kgold .

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 32 / 53

 Overaccumulation and Policy Overaccumulation and Pareto Optimality

Overaccumulation VI

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:

∆c (T ) = (1 + n) ∆k > 0
∆c (t ) = f 0 (k ∆k ) (1 + n) ∆k for all t > T

The …rst expression re‡ects the direct increase in consumption due to

the decrease in savings.
In addition, since k > kgold , for small enough ∆k,
f 0 (k ∆k ) (1 + n) < 0, thus ∆c (t ) > 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.
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 33 / 53
 Overaccumulation and Policy Overaccumulation and Pareto Optimality

Pareto Optimality and Suboptimality in the OLG Model

Proposition In the baseline overlapping-generations economy, the

competitive equilibrium is not necessarily Pareto optimal.
More speci…cally, whenever r < n and the economy is
dynamically ine¢ cient, it is possible to reduce the capital
stock starting from the competitive steady state and increase
the consumption level of all generations.

Pareto ine¢ ciency of the competitive equilibrium is intimately linked

with dynamic ine¢ ciency.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 34 / 53

 Overaccumulation and Policy Overaccumulation and Pareto Optimality

Interpretation

Intuition for dynamic ine¢ ciency:

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 a¤ecting
welfare of current generation.
Pecuniary externalities typically second-order and do not matter for
welfare.
But not when an in…nite stream of newborn agents joining the
economy are a¤ected.
It is possible to rearrange in a way that these pecuniary externalities
can be exploited.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 35 / 53

 Overaccumulation and Policy Overaccumulation and Pareto Optimality

Further Intuition

Complementary intuition:
Dynamic ine¢ ciency 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.
E¤ect 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.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 36 / 53

 Overaccumulation and Policy Role of Social Security

Role of Social Security in Capital Accumulation

Social Security as a way of dealing with overaccumulation

Fully-funded system: young make contributions to the Social Security
system and their contributions are paid back to them in their old age.
Unfunded system or a pay-as-you-go: transfers from the young
directly go to the current old.
Pay-as-you-go (unfunded) Social Security discourages aggregate
savings.
With dynamic ine¢ ciency, discouraging savings may lead to a Pareto
improvement.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 37 / 53

 Overaccumulation and Policy Fully Funded Social Security

Fully Funded Social Security I

Government at date t raises some amount d (t ) from the young,

funds are invested in capital stock, and pays workers when old
R (t + 1) d (t ).
Thus individual maximization problem is,

max u (c1 (t )) + βu (c2 (t + 1))

c1 (t ),c2 (t +1 ),s (t )

subject to
c1 ( t ) + s ( t ) + d ( t ) w (t )
and
c2 ( t + 1 ) R (t + 1) (s (t ) + d (t )) ,
for a given choice of d (t ) by the government.
Notice that now the total amount invested in capital accumulation is
s (t ) + d (t ) = (1 + n ) k (t + 1).
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 38 / 53
 Overaccumulation and Policy Fully Funded Social Security

Fully Funded Social Security II

No longer the case that individuals will always choose s (t ) > 0.

As long as s (t ) is free, whatever fd (t )gt∞=0 , the competitive
equilibrium applies.
When s (t ) 0 is imposed as a constraint, competitive equilibrium
applies if given fd (t )gt∞=0 , privately-optimal fs (t )gt∞=0 is such that
s (t ) > 0 for all t.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 39 / 53

 Overaccumulation and Policy Fully Funded Social Security

Fully Funded Social Security III

Proposition Consider a fully funded Social Security system in the
above-described environment whereby the government
collects d (t ) from young individuals at date t.
1 Suppose that s (t ) 0 for all t. If given the feasible
∞
sequence fd (t )gt =0 of Social Security payments, the
utility-maximizing sequence of savings fs (t )gt∞=0 is
such that s (t ) > 0 for all t, then the set of competitive
equilibria without Social Security are the set of
competitive equilibria with Social Security.
2 Without the constraint s (t ) 0, given any feasible
sequence fd (t )gt∞=0 of Social Security payments, the
set of competitive equilibria without Social Security are
the set of competitive equilibria with Social Security.
Moreover, even when there is the restriction that s (t ) 0, a funded
Social Security program cannot lead to the Pareto improvement.
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 40 / 53
 Overaccumulation and Policy Unfunded Social Security

Unfunded Social Security I

Government collects d (t ) from the young at time t and distributes to
the current old with per capita transfer b (t ) = (1 + n) d (t )
Individual maximization problem becomes
max u (c1 (t )) + βu (c2 (t + 1))
c1 (t ),c2 (t +1 ),s (t )

subject to
c1 ( t ) + s ( t ) + d ( t ) w (t )
and
c2 ( t + 1 ) R (t + 1) s (t ) + (1 + n ) d (t + 1) ,
for a given feasible sequence of Social Security payment levels
fd (t )gt∞=0 .
Rate of return on Social Security payments is n rather than
r (t + 1) = R (t + 1) 1, because unfunded Social Security is a pure
transfer system.
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 41 / 53
 Overaccumulation and Policy Unfunded Social Security

Unfunded Social Security II

Only s (t )— rather than s (t ) plus d (t ) as in the funded

scheme— goes into capital accumulation.
It is possible that s (t ) will change in order to compensate, but such
an o¤setting change does not typically take place.
Thus unfunded Social Security reduces capital accumulation.
Discouraging capital accumulation can have negative consequences
for growth and welfare.
In fact, empirical evidence suggests that there are many societies in
which the level of capital accumulation is suboptimally low.
But here reducing aggregate savings may be good when the economy
exhibits dynamic ine¢ ciency.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 42 / 53

 Overaccumulation and Policy Unfunded Social Security

Unfunded Social Security III

Proposition Consider the above-described overlapping generations

economy and suppose that the decentralized competitive
equilibrium is dynamically ine¢ cient. Then there exists a
feasible sequence of unfunded Social Security payments
fd (t )gt∞=0 which will lead to a competitive equilibrium
starting from any date t that Pareto dominates the
competitive equilibrium without Social Security.

Similar to way in which the Pareto optimal allocation was

decentralized in the example economy above.
Social Security is transferring resources from future generations to
initial old generation.
But with no dynamic ine¢ ciency, any transfer of resources (and any
unfunded Social Security program) would make some future
generation worse-o¤.
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 43 / 53
 OLG with Impure Altruism Impure Altruism

Overlapping Generations with Impure Altruism I

Exact form of altruism within a family matters for whether the

representative household would provide a good approximation.
Parents care about certain dimensions of the consumption vector of
their o¤spring instead of their total utility or “impure altruism.”
A particular type, “warm glow preferences”: parents derive utility
from their bequest.
Production side given by the standard neoclassical production
function, satisfying Assumptions 1 and 2, f (k ).
Economy populated by a continuum of individuals of measure 1.
Each individual lives for two periods, childhood and adulthood.
In second period of his life, each individual begets an o¤spring, works
and then his life comes to an end.
No consumption in childhood (or incorporated in the parent’s
consumption).
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 44 / 53
 OLG with Impure Altruism Impure Altruism

Overlapping Generations with Impure Altruism II

No new households, so population is constant at 1.

Each individual supplies 1 unit of labor inelastically during is
adulthood.
Preferences of individual (i, t ), who reaches adulthood at time t, are

log (ci (t )) + β log (bi (t )) , (21)

where ci (t ) denotes the consumption of this individual and bi (t ) is

bequest to his o¤spring.
O¤spring starts the following period with the bequest, rents this out
as capital to …rms, supplies labor, begets his own o¤spring, and
makes consumption and bequests decisions.
Capital fully depreciates after use.

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 45 / 53

 OLG with Impure Altruism Impure Altruism

Overlapping Generations with Impure Altruism III

Maximization problem of a typical individual can be written as
max log (ci (t )) + β log (bi (t )) , (22)
ci (t ),b i (t )

subject to
ci (t ) + bi (t ) yi ( t ) w (t ) + R (t ) bi (t 1) , (23)
where yi (t ) denotes the income of this individual.
Equilibrium wage rate and rate of return on capital
w (t ) = f (k (t )) k (t ) f 0 (k (t )) (24)
R (t ) = f 0 (k (t )) (25)
Capital-labor ratio at time t + 1 is:
Z 1
k (t + 1) = bi (t ) di, (26)
0

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 46 / 53

 OLG with Impure Altruism Impure Altruism

Overlapping Generations with Impure Altruism IV

Measure of workers is 1, so that the capital stock and capital-labor
ratio are identical.
Denote the distribution of consumption and bequests across
households at time t by [ci (t )]i 2[0,1 ] and [bi (t )]i 2[0,1 ] .
Assume the economy starts with the distribution of wealth (bequests)
R1
at time t given by [bi (0)]i 2[0,1 ] , which satis…es 0 bi (0) di > 0.
De…nition An equilibrium in this overlapping generations economy with
warm glow preferences is a sequence of consumption and
bequest
n levels for each household, o∞
[ci (t )]i 2[0,1 ] [ i ( )]i 2[0,1 ]
, b t , that solve (22) subject to
t =0
(23), a sequence of capital-labor ratios, fk (t )gt∞=0 , given by
(26) with some initial distribution of bequests [bi (0)]i 2[0,1 ] ,
and sequences of factor prices, fw (t ) , R (t )gt∞=0 , that
satisfy (24) and (25).
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 47 / 53
 OLG with Impure Altruism Impure Altruism

Overlapping Generations with Impure Altruism V

Solution of (22) subject to (23) is straightforward because of the log

preferences,
β
bi (t ) = yi (t )
1+β
β
= [w (t ) + R (t ) bi (t 1)] , (27)
1+β

for all i and t.

Bequest levels will follow non-trivial dynamics.
bi (t ) can alternatively be interpreted as “wealth” level: distribution
of wealth that will evolve endogenously.
This evolution will depend on factor prices.
To obtain factor prices, aggregate bequests to obtain the capital-labor
ratio of the economy via equation (26).
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 48 / 53
 OLG with Impure Altruism Impure Altruism

Overlapping Generations with Impure Altruism VI

Integrating (27) across all individuals,

Z 1
k (t + 1) = bi (t ) di
0
1 Z
β
= [w (t ) + R (t ) bi (t 1)] di
1+β 0
β
= f (k (t )) . (28)
1+β
R1
The last equality follows from the fact that 0 bi (t 1) di = k (t )
and because by Euler’s Theorem, w (t ) + R (t ) k (t ) = f (k (t )).
Thus dynamics are straightforward and again closely resemble Solow
growth model.
Moreover dynamics do not depend on the distribution of bequests or
income across households.
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 49 / 53
 OLG with Impure Altruism Impure Altruism

Overlapping Generations with Impure Altruism VII

Solving for the steady-state equilibrium capital-labor ratio from (28),
β
k = f (k ) , (29)
1+β
Uniquely de…ned and strictly positive in view of Assumptions 1 and 2.
Moreover, equilibrium dynamics again involve monotonic convergence
to this unique steady state.
We know that k (t ) ! k , so the ultimate bequest dynamics are
given by steady-state factor prices.
Let these be denoted by w = f (k ) k f 0 (k ) and R = f 0 (k ).
Once the economy is in the neighborhood of the steady-state
capital-labor ratio, k ,
β
bi (t ) = [w + R bi (t 1)] .
1+β
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 50 / 53
 OLG with Impure Altruism Impure Altruism

Overlapping Generations with Impure Altruism VIII

When R < (1 + β) /β, starting from any level bi (t ) will converge to

a unique bequest (wealth) level

βw
b = . (30)
1 + β (1 R )

Moreover, it can be veri…ed that R < (1 + β) /β,

R = f 0 (k )
f (k )
<
k
1+β
= ,
β

Second line exploits the strict concavity of f ( ) and the last line uses
the de…nition of k from (29).
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 51 / 53
 OLG with Impure Altruism Impure Altruism

Overlapping Generations with Impure Altruism IX

Proposition Consider the overlapping generations economy with warm

glow preferences described above. In this economy, there
exists a unique competitive equilibrium. In this equilibrium
the aggregate capital-labor ratio is given by (28) and
monotonically converges to the unique steady-state
capital-labor ratio k given by (29). The distribution of
bequests and wealth ultimately converges towards full
equality, with each individual having a bequest (wealth) level
of b given by (30) with w = f (k ) k f 0 (k ) and
R = f 0 (k ).

Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 52 / 53

 Conclusions Conclusions

Conclusions
Overlapping generations of the more realistic than in…nity-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 ine¢ cient” and feature
overaccumulation: unfunded Social Security can ameliorate the
problem.
Declining path of labor income important for overaccumulation, and
what matters is not …nite horizons but arrival of new individuals.
Overaccumulation and Pareto suboptimality: pecuniary externalities
created on individuals that are not yet in the marketplace.
Not overemhasize dynamic ine¢ ciency: major question of economic
growth is why so many countries have so little capital.
Daron Acemoglu (MIT) Economic Growth Lecture 8 November 22, 2011. 53 / 53