Econ 702 Macroeconomic Theory Spring 2016

Econ 702 Macroeconomic Theory Spring 2016∗
José-Vı́ctor Rı́os-Rull
University of Pennsylvania
Federal Reserve Bank of Minneapolis
CAERP
April 26, 2016
∗
This is the evolution of class notes by many students over the years, both from Penn and Minnesota including Makoto
Nakajima (2002), Vivian Zhanwei Yue (2002-3), Ahu Gemici (2003-4), Kagan (Omer Parmaksiz) (2004-5), Thanasis
Geromichalos (2005-6), Se Kyu Choi (2006-7), Serdar Ozkan (2007), Ali Shourideh (2008), Manuel Macera (2009),
Tayyar Buyukbasaran (2010), Bernabe Lopez-Martin (2011), Rishabh Kirpalani (2012), Zhifeng Cai (2013), Alexandra
(Sasha) Solovyeva (2014), Keyvan Eslami (2015), and Sumedh Ambokar (2016).
1
Contents
1 Introduction 5
2 Review: Neoclassical Growth Model 5
2.1 The Neoclassical Growth Model Without Uncertainty . . . . . . . . . . . . . . . . . . 5
2.2 A Comment on the Welfare Theorems . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3 Recursive Competitive Equilibrium 8
3.1 A Simple Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.2 Adding Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2.1 Markov Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2.2 Problem of the Social Planner . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2.3 Recursive Competitive Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . 12
3.3 Economy with Government Expenditures . . . . . . . . . . . . . . . . . . . . . . . . 13
3.3.1 Lump Sum Tax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.3.2 Labor Income Tax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.3.3 Capital Income Tax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.3.4 Taxes and Debt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4 Adding Heterogeneity 15
4.1 Heterogeneity in Wealth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.2 Heterogeneity in Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.3 An International Economy Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2
4.4 Heterogeneity in Wealth and Skills with Complete Markets . . . . . . . . . . . . . . . 19
5 Some Other Examples 20
5.1 A Few Popular Utility Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.2 An Economy with Capital and Land . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
6 Asset Pricing: Lucas Tree Model 22
6.1 The Lucas Tree with Random Endowments . . . . . . . . . . . . . . . . . . . . . . . 22
6.2 Asset Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
6.3 Taste Shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
7 Endogenous Productivity in a Product Search Model 25
7.1 Competitive Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
7.1.1 Problem of Individuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
7.1.2 Firms’ Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
7.1.3 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
7.1.4 Pareto Optimality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
8 Measure Theory 32
9 Industry Equilibrium 35
9.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
9.2 A Simple Dynamic Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
9.3 Introducing Exit Decisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
9.4 Stationary Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3
9.5 Adjustment Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
10 Incomplete Market Models 43
10.1 A Farmer’s Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
10.2 Huggett’s Economy[3] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
10.3 Aiyagari’s Economy[1] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
10.3.1 Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
10.3.2 Aggregate Shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
11 Monopolistic Competition, Endogenous Growth and R&D 49
A A Farmer’s Problem: Revisited 54
4
Introduction
A model is an artificial economy. Description of a model’s environment may include specifying the
agents’ preferences and endowment, technology available, information structure as well as property
rights. Neoclassical Growth Model becomes one of the workhorses of modern macroeconomics because
it delivers some fundamental properties of modern economy, summarized by, among others, Kaldor:
1. Output per capita has grown at a roughly constant rate (2%).
2. The capital-output ratio (where capital is measured using the perpetual inventory method based
on past consumption foregone) has remained roughly constant.
3. The capital-labor ratio has grown at a roughly constant rate equal to the growth rate of output.
4. The wage rate has grown at a roughly constant rate equal to the growth rate of output.
5. The real interest rate has been stationary and, during long periods, roughly constant.
6. Labor income as a share of output has remained roughly constant (0.66).
7. Hours worked per capita have been roughly constant.
Equilibrium can be defined as a prediction of what will happen and therefore it is a mapping from
environments to outcomes (allocations, prices, etc.). One equilibrium concept that we will deal with
is Competitive Equilibrium1 Characterizing the equilibrium, however, usually involves finding solutions
to a system of infinite number of equations. There are generally two ways of getting around this.
First, invoke the welfare theorem to solve for the allocation first and then find the equilibrium prices
associated with it. The first way sometimes may not work due to, say, presence of externality. So the
second way is to look at Recursive Competitive equilibrium, where equilibrium objects are functions
instead of variables.
Review: Neoclassical Growth Model
We review briefly the basic neoclassical growth model.
The Neoclassical Growth Model Without Uncertainty
The commodity space is
L = {(l1 , l2 , l3 ) : li = (lit )∞
t=0 lit ∈ R, sup |lit | < ∞, i = 1, 2, 3}.
t
1
Arrow-Debreu or Valuation Equilibrium.
5
The consumption possibility set is
X (k 0 ) = {x ∈ L : ∃(ct , kt+1 )∞
t=0 s.th. ∀t = 0, 1, ...
ct , kt+1 ≥ 0, x1t + (1 − δ)kt = ct + kt+1 , −kt ≤ x2t ≤ 0, −1 ≤ x3t ≤ 0, k0 = k 0 }.

Q
The production possibility set is Y = t Yt , where
Yt = {(y1t , y2t , y3t ) ∈ R3 : 0 ≤ y1t ≤ F (−y2t , −y3t )}.
Definition 1 An Arrow-Debreu equilibrium is (x ∗ , y ∗ ) ∈ X × Y , and a continuous linear functional ν ∗

such that
P∞
1. x ∗ ∈ arg maxx∈X ,ν ∗ (x)≤0 t=0 β t u(ct (x), −x3t ),
2. y ∗ ∈ arg maxy ∈Y ν ∗ (y ),
3. and x ∗ = y ∗ .
Note that in this definition, we have added leisure. Now, let’s look at the one-sector growth model’s
Social Planner’s Problem:
max ∞ t
P
t=0 β u(ct , −x3t ) (SPP)
s.t.
ct + kt+1 − (1 − δ)kt = x1t
−kt ≤ x2t ≤ 0
−1 ≤ x3t ≤ 0
0 ≤ y1t ≤ F (−y2t , −y3t )
x =y
k0 given.
Suppose we know that a solution in sequence form exists for (SPP) and is unique. Clearly stating
sufficient assumptions on utility and production function, show that (SPP) has a unique solution.
Two important theorems show the relationship between CE allocations and Pareto optimal allocations:
Theorem 1 Suppose that for all x ∈ X there exists a sequence (xk )∞ k=0 , such that for all k ≥ 0,
xk ∈ X and U(xk ) > U(x). If (x ∗ , y ∗ , ν ∗ ) is an Arrow-Debreu equilibrium then (x ∗ , y ∗ ) is Pareto
efficient allocation.
Theorem 2 If X is convex, preferences are convex, U is continuous, Y is convex and has an interior
point, then for any Pareto efficient allocation (x ∗ , y ∗ ) there exists a continuous linear functional ν such
that (x ∗ , y ∗ , ν) is a quasiequilibrium, that is (a) for all x ∈ X such that U(x) ≥ U(x ∗ ) it implies
ν(x) ≥ ν(x ∗ ) and (b) for all y ∈ Y , ν(y ) ≤ ν(y ∗ ).
6
Note that at the very basis of the CE definition and welfare theorems there is an implicit assumption of
perfect commitment and perfect enforcement. Note also that the FWT implicitly assumes there is no
externality or public goods (achieves this implicit assumption by defining a consumer’s utility function
only on his own consumption set but no other points in the commodity space).
From the First Welfare Theorem, we know that if a Competitive Equilibrium exits, it is Pareto Optimal.
Moreover, if the assumptions of the Second Welfare Theorem are satisfied and if the SPP has a
unique solution then the competitive equilibrium allocation is unique and they are the same as the PO
allocations. Prices can be constructed using this allocation and first order conditions.
Show that
v2t v3t
= Fk (kt , lt ) and = Fl (kt , lt ).
v1t v1t
One shortcoming of the AD equilibrium is that all trade occurs at the beginning of time. This assump-
tion is unrealistic. Modern Economics is based on sequential markets. Therefore we define another
equilibrium concept, Sequence of Markets Equilibrium (SME). We can easily show that SME is equiv-
alent to ADE. Therefore all of our results still hold and SME is the right problem to solve.
Define a Sequential Markets Equilibrium (SME) for this economy. Prove that the objects we get from
the AD equilibrium satisfy SME conditions and that the converse is also true. We should first show
that a CE exists and therefore coincides with the unique solution of (SPP).
Note that the (SPP) problem is hard to solve, since we are dealing with infinite number of choice
variables. We have already established the fact that this SPP problem is equivalent to the following
dynamic problem (removing leisure from now on):
v (k) = max0 u(c) + βv (k 0 ) (RSPP)
c,k
s.t. c + k 0 = f (k).
We have seen that this problem is easier to solve.
What happens when the welfare theorems fail? In this case the solutions to the social planners problem
and the CE do not coincide and so we cannot use the theorems we have developed for dynamic
programming to solve the problem. As we will see in this course, in this case we can work with
Recursive Competitive Equilibria. In general, we can prove that the solution to the RCE coincides
with a sequential markets problem but not the other way around (for example when we have multiple
equilibria). However, in all the models we see in this course, this equivalence will hold.
A Comment on the Welfare Theorems
Situations in which the welfare theorems would not hold include externalities, public goods, situations
in which agents are not price takers (e.g. monopolies), some legal systems or lacking of markets which
7
rule out certain contracts which appears complete contract or search frictions. In all of these situations
finding equilibrium through SPP is no longer valid. Therefore, in these situations, as mentioned before,
it is better to define the problem in recursive way and find the allocation using the tools of Dynamic
Programming.
Recursive Competitive Equilibrium
A Simple Example
What we have so far is that we have established the equivalence between allocation of the SPP problem
which gives the unique Pareto optima (which is same as allocation of AD competitive equilibrium and
allocation of SME). Therefore we can solve for the very complicated equilibrium allocation by solving the
relatively easier Dynamic Programming problem of the social planner. One handicap of this approach
is that in a lot of environments, the equilibrium is not Pareto Optimal and hence, not a solution of
a social planner’s problem, e.g. when you have taxes or externalities. Therefore, the above recursive
problem would not be the right problem to solve. In some of these situations we can still write the
problem in sequence form. However, we would lose the powerful computational techniques of dynamic
programming. In order to resolve this issue we will define Recursive Competitive Equilibrium equivalent
to SME that we can always solve for.
In order to write the decentralized household problem recursively, we need to use some equilibrium
conditions so that the household knows what prices are as a function of some economy-wide aggregate
state variable. We know that if capital is Kt and there is 1 unit of labor, then w (K ) = Fn (K , 1) and
R (K ) = Fk (K , 1) . Therefore, for the households to know prices they need to know aggregate capital.
Now, a household who is deciding about how much to consume and how much to work has to know
the whole sequence of future prices, in order to make his decision. This means that he needs to know
the path of aggregate capital. Therefore, if he believes that aggregate capital changes according to
K 0 = G (K ), knowing aggregate capital today, he would be able to project aggregate capital path for
the future and therefore the path for prices. So, we can write the household problem given function
G (·) as follows:
Ω(K , a; G ) = max
0
u(c) + βΩ(K 0 , a0 ; G ) (RCE )
c,a
s.t. c + a0 = w (K ) + R(K )a
K 0 = G (K ),
c ≥0
The above problem is the problem of a household that sees K in the economy, has a belief G , and carries
a units of assets from past. The solution of this problem yields policy functions c(K , a; G ), a0 (K , a; G )
and a value function Ω(z, K , a; G ). The functions w (K ), R(K ) are obtained from the firm’s FOCs
8
(below).
uc [c(K , a; G )] = βΩa [G (K ), a0 (K , a; G ); G ]
Ωa [K , a; G ] = (1 + r )uc [c(K , a; G )]
Now we can define the Recursive Competitive Equilibrium.
Definition 2 A Recursive Competitive Equilibrium with arbitrary expectations G is a set of functions2

Ω, g : K × A → R, R, w , H : K → R+ such that:
1. given G ; Ω, g solves the household problem in (RCE),

2. K 0 = H(K ; G ) = g (K , K ; G ) (representative agent condition),
3. w (K ) = Fn (K , 1) ,
4. and R (K ) = Fk (K , 1).
We define another notion of equilibrium where the expectations of the households are consistent with
what happens in the economy:
Definition 3 A Rational Expectations (Recursive) Equilibrium is a set of functions Ω, g , R, w , G ∗ ,

such that:
1. Ω(K , a; G ∗ ), g (K , a; G ∗ ) solves HH problem in (RCE),

2. G ∗ (K ) = g (K , K ; G ∗ ) = K 0 ,
3. w (K ) = Fn (K , 1),
4. and R (K ) = Fk (K , 1).
What this means is that in a REE, households optimize given what they believe is going to happen in
the future and what happens in the aggregate is consistent with the household’s decision. The proof
that every REE can be used to construct a SME is left as an exercise. The reverse turns out not to
be true. Notice that in REE, function G projects next period’s capital. In fact, if we construct an
equilibrium path based on REE, once a level of capital is reached in some period, next period capital is
uniquely pinned down by the transition function. If we have multiplicity of SME, this would imply that
we cannot construct the function G since one value of capital today could imply more than one value
for capital tomorrow. We will focus on REE unless expressed otherwise.
Note that unless otherwise stated, we will assume that depreciation rate δ is 1. R (K ) is the gross
return on capital which is Fk (K , 1) + 1 − δ. Net return on capital is r (K ) = Fk (K , 1) − δ.
2
We could add the policy function for consumption gc (K , a; G ).
9
Adding Uncertainty
Markov Processes
In this part, we want to focus on stochastic economies where there is a productivity shock affecting the
economy. The stochastic process for productivity that we are assuming is a first order Markov Process
that takes on finite number of values in the set Z = {z 1 < · · · < z nz }. A first order Markov process
implies
Pr(zt+1 = z j |ht ) = Γij , zt (ht ) = z i
where ht is the history of previous shocks. Γ is a Markov matrix with the property that the elements
of its rows sum to 1.
Let µ be a probability distribution over initial states, i.e.

X
µi = 1
i
and µi ≥ 0 ∀i = 1, ..., nz .
Next periods the probability distribution can be found by the formula: µ0 = Γµ.
If Γ is “nice” then ∃ a unique µ∗ s.t. µ∗ = Γµ∗ and µ∗ = limm→∞ (Γ)m µ0 , ∀µ0 ∈ ∆i .
Γ induces the following probability distribution conditional on z0 on ht = {z 0 , z 1 , ..., z t }:
Π({z 0 , z1 }) = Γi,. for z 0 = zi .
Π({z 0 , z1 , z2 }) = ΓΓi,. for z 0 = zi .
0
0
Γzz 0 z 0
P
Then,
PΠ(ht ) is the probability of history ht conditional on z . The expected value of z is z0
and z 0 Γzz 0 = 1.
Problem of the Social Planner
Let productivity affect the production function in a multiplicative fashion; i.e. technology is zF (K , N),
where z is the shock that follows a Markov chain on a finite state-space. The problem of the social
planner problem (SPP) in sequence form is
∞ X
X
maxt β t π(z t )u(ct (z t ))
{ct (z t ),kt+1 (z )}∈X (z t )
t=0 zt
s.t. ct (z t ) + kt+1 (z t ) = zt F (kt (z t−1 ), 1),
10
where zt is the realization of shock in period t, and z t is the history of shocks up to (and including)
time t. X (z t ) is similar to the consumption possibility set defined earlier but this is after history z t has
occurred and is for consumption and capital.
Therefore, we can formulate the stochastic SPP in a recursive fashion as

( )
X
V (zi , K ) = max0 u(c) + β Γij V (zj , K 0 )
c,K
j
0
s.t. c + K = zi F (K , 1),
where Γ is the Markov transition matrix. The solution to this problem gives us a policy function of the
form K 0 = G (z, K ).
In a decentralized economy, Arrow-Debreu equilibrium can be defined by:

∞ X
X
max β t π(z t )u(ct (z t ))
{ct (z t ),kt+1 (z t ),x1t (z t ),x2t (z t ),x3t (z t )}∈X (z t )
t=0 z t
X∞ X
s.t. pt (z t ).xt (z t ) ≤ 0,
t=0 zt
where X (z t ) is again a variant of the consumption possibility set after history z t has occurred. Ignore
the overloading of notation. Note that we are assuming the markets are dynamically complete; i.e.
there is complete set of securities for every possible history that can appear.
By the same procedure as before, SME can be written in the following way:
∞ X
X
max β t π(z t )u(ct (z t ))
{ct (z t ),bt+1 (z ,zt+1 ),kt+1 (z t )}
t
t=0 zt
X
s.t. ct (z t ) + kt+1 (z t ) + bt+1 (z t , zt+1 )qt (z t , zt+1 )
zt+1
= kt (z )Rt (z ) + wt (z t ) + bt (z t−1 , zt )
t−1 t
bt+1 (z t , zt+1 ) ≥ −B.
To replicate the AD equilibrium, here, we have introduced Arrow securities to allow agents to trade
with each other against possible future shocks.
However, in equilibrium and when there is no heterogeneity, there will be no trade. Moreover, we have
two ways of delivering the goods specified in an Arrow security contract: after production and before
production. In an after production setting, the goods will be delivered after production takes place
and can only be consumed or saved for the next period. This is the above setting. It is also possible
to allow the consumer to rent the Arrow security income as capital to firms, which will be the before
production setting.
11
An important condition which must hold true in the before production setting is the no-arbitrage
condition:
X
qt (z t , zt+1 ) = 1
zt+1
Describe the AD problem, in particular the consumption possibility set X and the production set Y .
Every equilibrium achieved in AD settings can also be achieved in a SM setting, by the relation where
qt (z t , zt+1 ) = p1t+1 (z t , zt+1 )/p1t (z t ),
Rt (z t ) = p2t (z t )/p1t (z t ),
and
wt (z t ) = p3t (z t )/p1t (z t ).
Check that from the FOC’s, the same allocations result in the two settings.
The problem above state contingent goods are delivered in terms of consumption goods. Instead of this
assume they are delivered in terms of capital goods. Show that the same allocation would be achieved
in both settings.
Recursive Competitive Equilibrium
Assume that households can trade state contingent assets, as in the sequential market case. We can
write a household’s problem in recursive form as:
( )
X
V (K , z, a) = max
0 0
u (c) + β Γzz 0 V (K 0 , z 0 , a0 (z 0 ))
c,k ,d(z )
z0
X
s.t. c + k 0 + d (z 0 ) qz 0 (K , z) = w (K , z) + aR (K , z)
z0
0
K = G (K , z)
a0 (z 0 ) = k 0 + d (z 0 ) .
Write the first order conditions for this problem, given prices and the law of motion for aggregate
capital.
Solving this problem gives policy functions. So, a RCE in this case is a collection of functions V , c, k 0 ,
d, G , w , and R, so that
1. given G , w , and R, V solves household’s functional equation, with c, k 0 and d as the associated
policy function,
12
2. d (K , z, K , z 0 ) = 0, for all z 0 ,
3. k 0 (K , z, K ) = G (K , z),
4. w (K , z) = zFn (K , 1) and R (K , z) = zFk (K , 1),

P
5. and z 0 qz 0 (K , z) = 1.
The last condition is known as the no-arbitrage condition (recall that we had this equation in the case
of sequential markets as well). To see why this is a necessary equation in the equilibrium, note that an
agent can either save in the form of capital, or Arrow securities. However, these two choices must cost
the same. This implies Condition 4 above.
Note that in a sequence version of the household problem in SME, in order for households not to
achieve infinite consumption, we need a no-Ponzi condition; a condition that prevents Ponzi schemes
is
at
lim Qt < ∞.
s=0 Rs
t→∞
This is the weakest condition that imposes no restrictions on the first order conditions of the household’s
problem. It is harder to come up with its analogue for the recursive case. One possibility is to assume
that a0 lies in a compact set A, or a set that is bounded from below.3
Economy with Government Expenditures
Lump Sum Tax
The government levies each period T units of goods in a lump sum way and spends it in a public good,
say medals. Assume consumers do not care about medals. The household’s problem becomes:
V (K , a) = max
0
{u (c) + βV (K 0 , a0 )}
c,a
s.t. c + a0 + T = w (K ) + aR (K )
K 0 = G (K ) .
A solution of this problem are functions ga∗ (K , a; G , M, T ) and Ω(K , a; G ) and the equilibrium can be
characterized by G ∗ (K , M, T ) = ga∗ (K , K ; G ∗ , M, T ) and M ∗ = T (the government budget constraint
is balanced period by period). We will write a complete definition of equilibrium for a version with
government debt (below).
Note that the equilibrium will be optimal. But if consumers cared about medals, the equilibrium will
not be optimal in general.
3
We must specify A such that the borrowing constraint implicit in A is never binding.
13
Define fˆ(K , 1) = f (K , 1) − M for the planner. Show that the equilibrium is optimal when consumers
do not care about medals.
Labor Income Tax
We have an economy in which the government levies tax on labor in order to purchase medals. Medals
are goods which provide utility to the consumers.
V (K , a) = max
0
{u (c, M) + βV (K 0 , a0 )}
c,a
s.t. c + a0 = (1 − τ ) w (K ) + aR (K )
K 0 = G (K ) ,
given M = τ w (K ).
Since leisure is not valued, the labor decision stays trivial. Hence, there is no distortion due to taxes
and CE is Pareto optimal. This will also hold when medals do not provide any utility to the consumers.
Is there any change in the above implications of optimality if the tax rate is a function of aggregate
capital?
Suppose medals do not provide utility to agents but leisure does. Is CE optimal now? Is it distorted?
What if medals also provide utility?
Capital Income Tax
Now let us look at an economy in which the government levies tax on capital in order to purchase
medals. Medals are goods which provide utility to the consumers.
V (K , a) = max
0
{u (c, M) + βV (K 0 , a0 )}
c,a
s.t. c + a0 = w (K ) + a(1 + r (K ) (1 − τ ))
K 0 = G (K ) ,
given M = τ r (K ) K and R (K ) = 1 + r (K ). Now, the First Welfare Theorem is no longer applicable;
the CE will not be Pareto optimal anymore (if τ > 0 there will be a wedge, and the efficiency conditions
will not be satisfied).
Derive the first order conditions in the above problem to see the wedge introduced by taxes.
Taxes and Debt
Assume that government can issue debt and use taxes to finance its expenditures, and these expenditures
do affect the utility.
14
A government policy consists of taxes, spending (medals) as well as bond issuance. When the aggregate
states are K and B, as you will see why, then a government policy (in a recursive world!) is
τ (K , B) , M (K , B) and B 0 (K , B) .
For now, we shall assume these values are chosen so that the equilibrium exists. In this environment,
debt issued is relevant for the household because it permits him to correctly infer the amount of taxes.
Therefore the household needs to form expectations about the future level of debt from the government.
The government budget constraint now satisfies (with taxes on labor income):
M(K , B) + R (K ) · B = τ (K , B) w (K ) + B 0 (K , B)
Notice that the household does not care about the composition of his portfolio as long as assets have
the same rate of return which is true because of the no arbitrage condition. Therefore, the problem of
a household with assets equal to a is given by:
V (K , B, a) = max
0
{u (c, M (K , B)) + βV (K 0 , B 0 , a0 )}
c,a
s.t. c + a0 = w (K ) [1 − τ (K , B)] + aR (K )
K 0 = G (K , B)
B 0 = H (K , B) .
Let g (K , B, a) be the policy function associated with this problem. Then, we can define a RCE as
follows.
Definition 4 A Rational Expectations Recursive Competitive Equilibrium, given policies M(K , B) and
τ (K , B) is a set of functions V , g , G , H, w , and R, such that
1. V and g solve household’s functional equation,

2. w (K ) = F2 (K , 1) and R(K ) = F1 (K , 1),
3. g (K , B, K + B) = G (K , B) + H(K , B),
4. Government’s budget constraint is satisfied
H (K , B) = R (K ) B + M (K , B) − τ (K , B) w (K ) ,
h
5. and, government debt is bounded; i.e. there exists some B̄ so that for all K ∈ 0, ke , H (K , B) ≤
B̄.
Adding Heterogeneity
In the previous section we looked at situations in which recursive competitive equilibria (RCE) were
useful. In particular these were situations in which the welfare theorems failed and so we could not use
the standard dynamic programming techniques learned earlier. In this section we look at another way
in which RCE are helpful, in models with heterogeneous agents.
15
Heterogeneity in Wealth
First, let us consider a model in which we have two types of households that differ only in the amount
of wealth they own. Say there are two types of agents, labeled type 1 and 2, of equal measure of 1/2.
Agents are identical other than their initial wealth position and there is no uncertainty in the model.
The problem of an agent with wealth a is given by
01 02 0
V K 1 , K 2 , a = max

u (c) + βV K , K , a
c,a0
1
K + K2
1
K + K2

0
s.t. c + a = R a+W
2 2
0i i 1 2

K = G K , K , i = 1, 2.
Note that (in general) the decision rules of the two types of agents are not linear (even though they
might be almost linear); therefore, we cannot add the two states, K 1 and K 2 , to write the problem
with one aggregate state, in the recursive form.
Definition 5 A Rational Expectations Recursive Competitive Equilibrium is a set of functions V , g ,

R, w , G 1 , and G 2 , so that:
1. V solves the household’s functional equation, with g as the associated policy function,
2. w and R are the marginal products of labor and capital, respectively (watch out for arguments!),
3. representative agent conditions are satisfied; i.e.
g (K 1 , K 2 , K 1 ) = G 1 (K 1 , K 2 ),
and
g (K 1 , K 2 , K 2 ) = G 2 (K 1 , K 2 ).
Note that G 1 (K 1 , K 2 ) = G 2 (K 2 , K 1 ) (why?).
This is a variation of the simple neoclassical growth model; what does the growth model say about
inequality?
In the steady state of a neoclassical growth model, Euler equations for the two types simplify to
u 0 c 1 = βRu 0 c 1 , and u 0 c 2 = βRu 0 c 2 .

Therefore, we must have βR = 1, where

1
K + K2

R = FK ,1 .
2
16
Finally, by the household’s budget constraint, we must have:
kiR + W = ci + ki,
where k i = K i , by representative agent’s condition. Therefore, we have three equation, with four
unknowns (k i and c i ’s). This means, this theory is silent about the distribution of wealth in the steady
state!
Heterogeneity in Skills
Now, consider a slightly different economy where type i has labor skill i . Measures of agents’ types,
µ1 and µ2 , satisfy µ1 1 + µ2 2 = 1 (below we will consider the case where µ1 = µ2 = 1/2).
The question we have to ask ourselves is, would the value functions of two types remain to be the
same, as in the previous subsection? The answer turns out to be no!
The problem of the household i ∈ {1, 2} can be written as follows:
V i K 1 , K 2 , a = max u (c) + βV i K 01 , K 02 , a0

c,a 0
1
K + K2
1
K + K2

0
s.t. c + a = R a+W i
2 2
K 0j = G j K 1 , K 2 , j = 1, 2.

Notice that we have indexed the value function by the agent’s type; the reason is that the marginal
product of the labor supplied by each of these types is different.
We can rewrite this problem as

0 0 0
V i (K , λ, a) = max i

0
u (c) + βV (K , λ , a )
c,a
s.t. c + a0 = R (K ) a + W (K ) i
K = G (K , λ)
λ0 = H (K , λ) ,
where K is the total capital in this economy, and λ is the share of one type in this wealth (e.g. type
1).
Then, if g i is the policy function of type i, in the equilibrium, we must have:
G (K , λ) = g 1 (K , λ, λK ) + g 2 (K , λ, (1 − λ) K ) ,
and
H (K , λ) = g 1 (K , λ, λK ) /G (K , λ) .
17
An International Economy Model
In an international economy model the specifications which determine the definition of country is an
important one; we can introduce the idea of different locations or geography; countries can be victims
of different policies; trade across countries maybe more difficult due to different restrictions.
Here we will see a model with two countries, A and B, such that labor is not mobile between the
countries, but with perfect capital markets. Two countries may have different technologies, F A (KA , 1)
and F B (KB , 1); therefore, the resource constraint in this world would be
C A + C B + K 0A + K 0B = F A K A , 1 + F B K B , 1 .

(Therefore, while the product of the world economy can move freely between the two countries, once
installed, it has to be used in that country.)
The first question to ask, as usual, is what are the appropriate states in this world? As it is apparent
from the resource constraint and production functions, we need the capital in each country. But, also,
we need to know who owns this capital. Therefore, we need an additional variable as the aggregate
state; we can choose λ, the share of country A in total wealth. But, why not the share of this country
in the total capital in each of these countries? Since, at the point of saving, the capital in the two
countries is perfect substitute. In other words
FkA K A , 1 = FkB K B , 1 = R A = R B ,

where we have incorporated the depreciation into the production function. This follows from a no
arbitrage argument.
As a result, country i’s problem can be written as:

0A 0B 0 0
V i K A , K B , λ, a = max i

u (c) + βV K , K , λ , a
c,a0
s.t. c + a0 = R K i a + w i K i

K 0j = G j K A , K B , λ , j = A, B

λ0 = Γ K A , K B , λ .

Notice that, given K A , we know that, from our earlier no-arbitrage argument, that
R K A = FkA K A , 1 = FkB K B , 1 .

Therefore, we can infer K B , and R K B . As a result, we don’t need to keep track of K B any longer.
Moreover, the wage rate is given by
w A K A = FnA K A , 1 .

18
Thus, we may rewrite the problem as
V i (K , λ, a) = max u (c) + βV i (K 0 , λ0 , a0 )

0
c,a
s.t. c + a0 = R (K ) a + w i (K )
K 0 = G (K , λ)
λ0 = Γ (K , λ) ,
where K is the total capital in the world, and R and w i can be derived as
R (K ) = FkA (αK , 1) ,
w A (K ) = FnA (αK , 1) ,
w B (K ) = FnB ((1 − α) K , 1) ,
and α is the solution to the following equation:
FkA (αK , 1) = FkB ((1 − α) K , 1) .
Note that the relative price of capital in the two countries is 1 since they offer the same marginal
product and we are in the deterministic case. Now, we can define the equilibrium as
Definition 6 A Recursive Competitive Equilibrium for the (world’s) economy is a set of functions, V i ,
g i and w i , for i ∈ {A, B}, and R, G , and Γ, such that the following conditions hold:
1. V i and g i solve the household’s problem in country i (i ∈ {A, B}),
2. G (K , λ) = g A (K , λ, λK ) + g B (K , λ, (1 − λ) K ),
3. Γ (K , λ) = g A (K , λ, λK ) /G (K , λ),
4. w i is equated to the marginal products of labor in each country,
5. and, R is equal to the marginal product of capital (in both countries).
Heterogeneity in Wealth and Skills with Complete Markets
Now, let us consider a model in which we have two types of households that care about leisure differ
in the amount of wealth they own and labor skill. Also, there is uncertainty and Arrow securities like
we have seen before.
Let A1 and A2 be the aggregate asset holdings of the two types of agents. These will now be state
variables for the same reason that K 1 and K 2 were state variables earlier. The problem of an agent
19
i ∈ {1, 2} with wealth a is given by
( )
X
V i z, A1 , A2 , a = max Γzz 0 V z 0 , A01 , A02 , a0 (z 0 )

0 0
u (c, n) + β
c,n,a (z )
z0
X
0 0
a (z ) qz 0 z, A1 , A2 = R (z, K , N) a + W (z, K , N) ni

s.t. c +
z0
A0i (z ) = G i z, A1 , A2 , z 0 ,
0
i = 1, 2, ∀z 0

N = H z, A1 , A2

A1 + A2
K= .
2
Let g i and hi be the asset and labor policy functions be the solution to this problem. Then, we can
define the RCE as below.
Definition 7 A Recursive Competitive Equilibrium with Complete Markets is a set of functions V i ,

g i , hi , G i , R, w , H, and q, so that:
1. V i , g i and hi solve the problem of household i (i ∈ {1, 2}),

2. H (z, A1 , A2 ) = 1 h1 (z, A1 , A2 , A1 ) + 2 h2 (z, A1 , A2 , A2 ),
3. G i (z, A1 , A2 , z 0 ) = g i z, A1 , A2 , Ai , z 0 i = 1, 2, ∀z 0

1 2 0
P
4. z 0 qz 0 (z, A , A , z ) = 1,
5. G 1 (z, A1 , A2 , z 0 ) + G 2 (z, A1 , A2 , z 0 ) is independent of z 0 (due to market clearing).

6. R and W are the marginal products of capital and labor.
Write down the household problem and the definition of RCE with non-contingent claims instead of
complete markets.
Some Other Examples
A Few Popular Utility Functions
Consider the following three utility forms:
1. u (c, c − ): this function is called habit formation utility function; utility is increasing in con-
sumption today, but, decreasing in the deviations from past consumption (e.g. u (c, c − ) =
2
v (c) − (c − c − ) ). In this case, the aggregate states in a standard growth model are K and C − ,
and individual states are a and c − . Is the equilibrium optimum in this case?
20
2. u (c, C − ); this form is called catching up with Jones; there is an externality from the aggregate
consumption to the payoff of the agents. Intuitively, in this case, agents care about what their
neighbors consume. Aggregate states in this case are K and C − . But, c − is no longer an
individual state.
3. u (c, C ): the last function is called keeping up with Jones. Here, the aggregate state is K ; C is
no longer a pre-determined variable to appear as a state.
An Economy with Capital and Land
Consider an economy with capital and land but without labor; a firm in this economy buys and installs
capital. They also own one unit of land, that they use in production, according to the production
function F (K , L). In other words, a firm is a “chunk of land of are one”, in which firm installs its
capital. Share of these firms are traded in a stock market.
A household’s problem in this economy is given by:
V (K , a) = max
0
{u (c) + βV (K 0 , a0 )}
c,a
s.t. c + a0 = R (K ) a.
K 0 = G (K )
On the other hand, a firm’s problem is
Ω (K , k) = max
0
{F (k, 1) − k 0 + q (K 0 ) Ω (K 0 , k 0 )}
k
s.t. K 0 = G (K ) .
Ω here is the value of the firm, measured in units of output, today. Therefore, the value of the firm,
tomorrow, must be discounted into units of output today. This is done by a discount factor q (K 0 ).
A Recursive Competitive Equilibrium consists of functions, V , Ω, g , h, q, G , and R, so that:
1. V and g solve household’s problem,
2. Ω and h solve firm’s problem,
3. G (K ) = h (K , K ), and,
4. q (G (K )) Ω (G (K ) , G (K )) = g (K , Ω (K , K )).
One condition is missing in the definition of the RCE above. Find it! [Hint: it relates the rate of return
on the household’s assets to the discount rate of firm’s value.]
21
Asset Pricing: Lucas Tree Model
We now turn to the simplest of all models in term of allocations as they are completely exogenous,
the Lucas tree model. We want to characterize the properties of prices that are capable of inducing
households to consume the endowment.
The Lucas Tree with Random Endowments
Consider an economy in which the only asset is a tree that gives fruit. The agents problem is
( )
X
V (z, s) = max
0
u (c) + β Γzz 0 V (z 0 , s 0 )
c,s
z0
0
s.t. c + p (z) s = s [p (z) + d (z)] ,
where p (z) is the price of the shares (to the tree), in state z, and d (z) is the dividend associated with
state z.
Definition 8 A Rational Expectations Recursive Competitive Equilibrium is a set of functions, V , g ,

d, and p, such that
1. V and g solves the household’s problem,

2. d (z) = z, and,
3. g (z, 1) = 1, for all z.
To explore the problem more, note that the first order conditions for the household’s problem imply:
p (z 0 ) + d (z 0 )
X
uc (c (z, 1)) = β Γzz 0 uc (c (z 0 , 1)) .
z0
p (z)
As a result, if we let uc (z) := uc (c (z, 1)), we get:

X
p (z) uc (z) = β Γzz 0 uc (z 0 ) [p (z 0 ) + z 0 ] .
z0
Derive the Euler equation for household’s problem.
Notice that this is just a system of n equations with unknowns {p (zi )}ni=1 . We can use the power of
matrix algebra to solve it. To do so, let:
 
p (z1 )
p :=  ...  ,
 
p (zn )
22
and
 
uc (z1 ) · · · 0
uc :=  .. ... ..
.
 
. .
0 · · · uc (zn )
Then
 
p (z1 ) uc (z1 )
uc .p =  ..
,
 
.
p (zn ) uc (zn )
and
 
z1 uc (z1 )
uc .z =  ..
,
 
.
zn uc (zn )
Now, rewrite the system above as
uc p = βΓuc z + βΓuc p,
where Γ is the transition matrix for z, as before. Hence, the price for the shares is given by
uc p = (I − βΓ)−1 βΓuc z,
or
 −1
uc (z1 ) · · · 0
.. .. .. −1
p=  (I − βΓ) βΓuc z.
 
. . .
0 · · · uc (zn )
Asset Pricing
Consider our simple model of Lucas tree with fluctuating output; what is the definition of an asset in
this economy? It is “a claim to a chunk of fruit, sometime in the future”.
If an asset, a, promises an amount of fruit equal to at (z t ) after history z t = (z0 , z1 , ... , zt ) of shocks,
after a set of (possible) histories in H, the price of such an entitlement in date t = 0 is given by:
XX
p (a) = qt0 (z t ) at (z t ) ,
t z t ∈H
where qt0 (z t ) is the price of one unit of fruit after history z t , in today’s “dollars”; this follows from
a no-arbitrage argument. If we have the date t = 0 prices, {qt }, as functions of histories, we can
replicate any possible asset by a set of state-contingent claims, and use this formula to price that asset.
23
To see how we can find prices at date t = 0, consider a world in which the agent wants to solve
(∞ )
X X
max
t
βt πt (z t ) u (ct (z t ))
ct (z )
t=0 zt
∞ X
X ∞ X
X
s.t. qt0 (z t ) ct (z t ) ≤ qt0 (z t ) zt .
t=0 zt t=0 ht
This is the familiar Arrow-Debreu market structure, where the household owns a tree, and the tree can
yield z ∈ Z amount of fruit in each period. The first order condition for this problem implies:
uc (zt )
qt0 (z t ) = β t πt (z t ) .
uc (z0 )
This enables us to price the good in each history of the world, and price any asset accordingly.
What happens if we add state contingent shares into our recursive model? Then the agent’s problem
becomes:
( )
X
V (z, s, b) = max
0 0 0
u (c) + β Γzz 0 V (z 0 , s 0 , b 0 (z 0 ))
c,s ,b (z )
z0
X
0
s.t. c + p (z) s + q (z, z 0 ) b 0 (z 0 ) = s [p (z) + z] + b.
z0
A characterization of q can be written as:
q (z, z 0 ) uc (z) = βΓzz 0 uc (z 0 ) .
We can price all types of securities using p and q in this economy.
To see how we can price an asset, consider the option to sell tomorrow at price P, if today’s shock is
z, as an example; the price of such an asset today is
X
q̂ (z, P) = max {P − p (z 0 ) , 0} q (z, z 0 ) .
z0
The American option to sell at price P either tomorrow or the day after tomorrow is priced as:
X
q (z, P) =
e max {P − p (z 0 ) , q̂ (z 0 , P)} q (z, z 0 ) .
z0
Similarly, a European option to buy at price P the day after tomorrow is priced as:
XX
q̄ (z, P) = max {p (z 00 ) − P, 0} q (z 0 , z 00 ) q (z, z 0 ) .
z0 z 00
24
Note that R (z) = [ z 0 q (z, z 0 )]−1 is the gross risk
P
P free rate, given today’s shock is z. The uncondi-
tional gross risk free rate is then given as R = z µz R(z) where µ∗ is the steady-state distribution
f ∗
of the shocks defined earlier.

h 0 0i
The average gross rate of return on the stock market is z µ∗z z 0 Γzz 0 p(zp(z)
)+z
P P
and the risk premium
is the difference between this rate and the unconditional gross risk free rate.
Use the expressions for p and q and the properties of the utility function to show that risk premium is
positive.
Taste Shocks
Consider an economy in which the only asset is a tree that gives fruits. The fruit is constant over time
(normalized to 1) but the agent is subject to preference shocks for the fruit each period, θ ∈ Θ. The
agent’s problem in this economy is
( )
X
V (θ, s) = max0
θu (c) + β Γθθ0 V (θ0 , s 0 )
c,s
θ0
0
s.t. c + p (θ) s = s [p (θ) + d (θ)] .
The equilibrium is defined as before; the only difference is that, now, we must have d (θ) = 1. What
does it mean that the output of the economy is constant, and fixed at one, but, the tastes in this
output changes? In this settings, the function of the price is to convince agents keep their consumption
constant.
All the analysis follows through, once we write the FOC’s characterizing price, p (θ), and state contin-
gent prices q (θ, θ0 ).
Endogenous Productivity in a Product Search Model
Let’s model the situation where households need to find the fruit before consuming it;4 assume that
households have to find the tree in order to consume the fruit. Finding trees is characterized by a
constant returns to scale (increasing) matching function M (T , D),5 where T is the number of trees
and D is the shopping effort, exerted by households when searching. Thus, the probability that a tree
finds a shopper is M (T , D) /T ; total number of matches, divided by the number of trees. And, the
probability that a unit of shopping effort finds a tree is M (T , D) /D.
We further assume that M takes the form D ϕ T 1−ϕ , and denote the probability of finding a tree
by Ψd (Q) := Q 1−ϕ , where 1/Q := D/T is the ratio of shoppers per trees, capturing the market
4
Think of fields in The Land of Apples, full of apples, that are owned by firms; agents have to buy the apples. In
addition, they have to search for them as well!
5
What does the fact that M is constant returns to scale imply?
25
tightness; the more the number of people searching, the smaller the probability of finding a tree. Then,
Ψt (Q) := Q −ϕ . Note that, in this economy, the number of trees is constant, and equal to one.6
The household’s problem can be written as:

( )
X
V (θ, s) = max0 u (c, d, θ) + β Γθθ0 V (θ0 , s 0 ) (1)
c,d,s
θ0
s.t. c = dΨd (Q (θ)) (2)
c + P (θ) s 0 = P (θ) [s (1 + R (θ))] , (3)
where P is the price of tree relative to that of consumption, and R is the dividend income (in units of
tree). d is the amount of search the individual household exerts to acquire fruit.
We could also write the household budget constraint in terms of the price of consumption relative to
1
that of the tree P̂(θ) = P(θ) as c P̂(θ) + s 0 = s (1 + R (θ)). Also, let u (c, d, θ) = u (θc, d) from here
on. If we substitute the constraints into the objective and solve for d, we get the Euler equation for
the household. Using the market clearing in equilibrium, the problem reduces to one equation and two
unknowns, P̂ and Q (other objects, C ,D and R, are known functions of P̂ and Q). We still need
another functional equation to find an equilibrium. We now turn to one way of doing so.
Derive the Euler equation of the household from the problem defined above.
Competitive Search
Competitive search is a particular search protocol of what is called non-random (or directed) search.
To understand this protocol, consider a world consisting a large number of islands. Each island has
a sign that reads two number, W and Q. W is the wage rate on the island, and Q is a measure of
market tightness in that island, or the number of workers on the island divided by the number of job
opportunities. Both, workers and firms have to decide to go to one island. In an island with higher
wage, the worker might be happier, conditioned on finding a job. However, the probability of finding
a job might be low on the island, depending on the tightness of the labor market on that island. The
same story holds for the job owners, who are searching to hire workers.
In our economy, both firms and households search for specific markets indexed by price P̂ and market
tightness Q. Given the market tightness Q in a market, price of consumption in that market is
6
It is easy to find the statements for Ψt and Ψd , given the Cobb-Douglas matching function;
1−ϕ
D ϕ T 1−ϕ

T
Ψd (Q) = = = Q 1−ϕ ,
D D
−ϕ
D ϕ T 1−ϕ

T
Ψt (Q) = = = Q −ϕ .
T D
The question is, is Cobb-Douglas an appropriate choice for the matching function, or its choice is a matter of simplicity?
26
determined by the interaction of demand and supply. Therefore, we may write the price of consumption
as a function of the market tightness. We denote this price by P̂ (Q). Agents can go to any such
market and choose how much effort to put in that market. A firm chooses which market to operate in
to maximize its profit. Competitive search is magic. It does not presuppose a particular pricing protocol
(wage posting, bargaining) that other search protocols need.
Problem of Individuals
An individual who holds s shares of the tree, when the (aggregate) state of the economy is θ, solves
the following recursive problem:7
( )
X
V (θ, s) = max 0 u (θc, d) + β Γθθ0 V (θ0 , s 0 ) (4)
Q,c,d,s
θ0
c
s.t. d = (5)
Ψd (Q)
P̂ (Q) c + s 0 = s (1 + R) , (6)
If we substitute the constraints in the objective, we can rewrite this problem as:
( )
X
V (θ, s) = max u (θdΨd (Q) , d) + β Γθθ0 V θ0 , s (1 + R) − P̂ (Q) dΨd (Q) . (7)
Q,d
θ0
Fist order conditions, for the maximization problem on the right hand side of this functional equation
are:
1.
Ψd (Q) θuc (θdΨd (Q) , d) + ud (θdΨd (Q) , d)

X
− P̂ (Q) Ψd (Q) β Γθθ0 Vs θ0 , s (1 + R) − P̂ (Q) dΨd (Q) = 0,
θ0
and,
2.
θdΨ0d (Q) uc (θdΨd (Q) , d)

h i X
− P̂ (Q) dΨ0d (Q) + P̂ 0 (Q) dΨd (Q) β Γθθ0 Vs θ0 , s (1 + R) − P̂ (Q) dΨd (Q) = 0.
θ0
7
Note that, individual is assumed to be choosing only one market in each period. In principle, this needs not be the
case. However, since all households and firms are identical, in the equilibrium only one market is operational. Hence,
relaxing this assumption would not change anything. Also note that, R and Q, and thus P̂(Q) are implicit functions
of θ.
27
We can rewrite Condition 1 as:
ud (θdΨd (Q) , d)
θuc (θdΨd (Q) , d) +
Ψd (Q)
X
= P̂ (Q) β Γθθ0 Vs θ0 , s (1 + R) − P̂ (Q) dΨd (Q) . (8)
θ0
To get rid of Vs , look at the initial household problem, (4). Let the multiplier on the budget constraint
be λ. Get rid of d from the problem by using (5) in the objective. Applying the envelope theorem, we
get:
Vs (θ, s) = λ (1 + R) , (9)
where λ is the Lagrange multiplier on Constraint (6). The first order condition with respect to c implies:

c 1 c
λP̂ (Q) = θuc θc, + ud θc, .
Ψd (Q) Ψd (Q) Ψd (Q)
If we substitute this into Equation (9), we get:

1 (1 + R)
Vs (θ, s) = θuc (θc, d) + ud (θc, d) , (10)
Ψd (Q) P̂ (Q)
where the derivatives on the right hand side are evaluated at the optimal values. As a result, we may
write (8) as:
ud (θdΨd (Q) , d)
θuc (θdΨd (Q) , d) +
Ψd (Q)
(1 + R 0 ) 0

X
0 0 0 0 1 0 0 0 0
= P̂ (Q) β Γθθ0 θ uc (θ d Ψd (Q ) , d ) + 0
ud (θ d Ψd (Q ) , d ) . (11)
θ 0 P̂ (Q 0) Ψd (Q )
Condition 2 above can also be simplified further as:

1 h
0 0
i X
θuc (θdΨd (Q) , d) = 0 Ψd (Q) P̂ (Q) + P̂ (Q) Ψd (Q) β Γθθ0
Ψd (Q) θ0

× Vs θ0 , s (1 + R) − P̂ (Q) dΨd (Q)
" #
P̂ 0 (Q) Ψd (Q) X
= P̂ (Q) + β Γθθ0
Ψ0d (Q) θ 0
(1 + R 0 ) 0

0 0 0 0 1 0 0 0 0
× θ uc (θ d Ψd (Q ) , d ) + ud (θ d Ψd (Q ) , d ) .
P̂ (Q 0 ) Ψd (Q 0 )
(12)
28
Firms’ Problem
A firm, in any given period, tries to maximize the returns to the tree by choosing the right market, Q.
Note that, by choosing a market Q, the firm is effectively choosing the price of consumption, P̂ (Q).
Since there is nothing dynamic in the choice of market (note that, we are assuming firms can choose
a different market in each period), we may write the problem of a firm as:
π = max P̂ (Q) Ψt (Q) . (13)

Q
The first order condition for the optimal choice of Q is
P̂ 0 (Q) Ψt (Q) + P̂ (Q) Ψ0t (Q) = 0,
This condition determines the price as a function of Q as:
P̂ 0 (Q) Ψ0t (Q)

=− . (14)
P̂ (Q) Ψt (Q)
Equilibrium
Before defining the equilibrium concept for this economy, let’s simplify the optimality conditions we
have derived so far, a bit further; note that, by the choice of a Cobb-Douglas matching function, we
have:
Ψ0t (Q) = −ϕQ −ϕ−1 ,

Ψ0 (Q) ϕQ −ϕ−1 ϕ
∴− t = −ϕ
= , (15)
Ψt (Q) Q Q
and:
Ψ0d (Q) = (1 − ϕ) Q −ϕ ,
Ψ0 (Q) (1 − ϕ) Q −ϕ (1 − ϕ)
∴ d = 1−ϕ
= . (16)
Ψd (Q) Q Q
By Equations (14) and (15), we have:
P̂ 0 (Q) ϕ
= . (17)
P̂ (Q) Q
By substituting Equations (16) and (17) into (12), we get:
29
X
ϕ Q
θuc (θdΨd (Q) , d) =P̂ (Q) 1 + β Γθθ0
Q (1 − ϕ) θ 0
(1 + R 0 ) 0

0 0 0 0 1 0 0 0 0
P̂ (Q 0 ) Ψd (Q 0 )
(18)
Simplifying, we can write this equality as:
P̂ (Q) X
θuc (θdΨd (Q) , d) = β Γθθ0
1 − ϕ θ0
(1 + R 0 ) 0

0 0 0 0 1 0 0 0 0
P̂ (Q 0 ) Ψd (Q 0 )
(19)
If we compare Conditions (11) and (19), we observe they are very similar. One difference is the term
ud in (11). The reason that this terms does not appear in (19) is that households are the ones who
have to exert search effort to find the trees in this environment.
Note that, since all the firms are assumed to be ex-ante identical, the value of the firm in a given
period, π in Problem (13), would be the same for all the firms; if not, a firm has a profitable deviation
from his strategy, that he is not exhausting. This is what is paid to the households who are the owners
of the firms.
Now, we are ready to define an equilibrium of this economy.
Definition 9 An equilibrium with competitive search in this economy consists of a collection of func-
tions, c : Θ → R+ , d : Θ → R+ , s : Θ → [0, 1], P : R+ → R+ , Q : Θ → R+ , π : Θ → R+ , and
R : Θ → R+ , that satisfy:
1. individuals’ shopping and budget constraints, Constraints (5) and (6),
2. individual’s Euler equation, Equation (11),
3. firms’ first order condition, Equation (19),
4. firms’ value function, (13), and,
5. market clearing conditions; s (θ) = 1, Q (θ) = d (θ)−1 , and R (θ) = π (θ), for all θ ∈ Θ.
30
At the end, note that we assume dividends, R (θ), are paid out in units of the tree. So that, in
equilibrium, using the two constraints in the household problem, consumption is given by
R (θ)
C (θ) = = Q (θ)−ϕ .
P̂ (θ)
One way of thinking about competitive search is that, instead of having one attribute which is the
price, goods have two attributes; price and the difficulty of getting the good.
This definition is excessively cumbersome, and we can go to the core of the issue by writing the two
functional equations that characterize the equilibrium. Note that in equilibrium, we have c = Q −ϕ
and c = R/P̂. Putting together the household’s Euler ((11)) and the tree’s FOC ((19)), we have that
{Q, P̂(Q)} have to solve
ud (θQ −ϕ , Q −1 )
ϕθuc (θQ −ϕ , Q −1 ) = − , (20)
Q 1−ϕ
X 1 −ϕ −ϕ −1
θuc (θQ −ϕ , Q −1 ) = P̂β Γθθ0 [ + Q (θ0 ) ]θ0 uc (θ0 Q (θ0 ) , Q (θ0 ) ).
θ0
P̂(θ0 )
Pareto Optimality
One of the fascinating properties of competitive search is that the resulting equilibrium is optimal. To
see this, note that, from the point of view of optimality, there are no dynamic considerations in this
model; there are no capital accumulation decisions, etc.
Therefore, we may consider a social planner’s problem (in each possible state, θ) as:
max u (θC , D)
C ,D
s.t. C = D ϕ .
The first order condition for this problem implies
ϕD ϕ−1 θuc (θD ϕ , D) + ud (θD ϕ , D) = 0. (21)
On the other hand, if we combine Equations (11) and (19), we get:

ud (θdΨd (Q (θ)) , d)
θuc (θdΨd (Q (θ)) , d) + = (1 − ϕ) θuc (θdΨd (Q (θ)) , d) .
Ψd (Q (θ))
If we substitute from the equilibrium conditions for Q = 1/D and Ψd = Q 1−ϕ , we can write this
condition as:
ud (θD ϕ , D)
θuc (θD ϕ , D) + ϕ−1
= (1 − ϕ) θuc (θD ϕ , D) ,
D
which can be rearranged as Equation (21).
31
Measure Theory
This section will be a quick review of measure theory to be able to use it in the subsequent sections.
Definition 10 For a set S, S is a family of subsets of S, if B ∈ S implies B ⊆ S (but not the other
way around).
Definition 11 A family of subsets of S, S, is called a σ-algebra in S if
1. S, ∅ ∈ S;
2. A ∈ S ⇒ Ac ∈ S (i.e. S is closed with respect to complements); and,

T
3. for {Bi }i∈N , Bi ∈ S for all i implies i∈N Bi ∈ S (i.e. S is closed with respect to countable
intersections).
1. The power set of S (i.e. all the possible subsets of a set S), is a σ-algebra in S.
2. {∅, S} σ-algebra in S.

3. ∅, S, S1/2 , S2/2 , where S1/2 means the lower half of S (imagine S as an closed interval in R),
σ-algebra in S.
4. If S = [0, 1], then

1 1 1
S = ∅, 0, , , ,1 ,S
2 2 2
is not a σ-algebra in S.
Note that, a convention is to
1. use small letters for elements,
2. use capital letters for sets, and
3. use fancy letters for a set of subsets (or families of subsets).
Now, we are ready to define a measure.
Definition 12 Suppose S is a σ-algebra in S. A measure is a function x : S → R+ , that satisfies
32
1. x (∅) = 0;
2. B1 , B2 ∈ S and B1 ∩ B2 = ∅ implies x (B1 ∪ B2 ) = x (B1 ) + x (B2 ) (additivity); and,
3. {Bi }i∈N ∈ S and Bi ∩ Bj = ∅, for all i 6= j, implies x (∪i Bi ) = i x (Bi ) (countable additivity).8
P
A set S, a σ-algebra in it, S, and a measure on S, define a measure space, (S, S, x).
Definition 13 Borel σ-algebra is a σ-algebra generated by the family of all open sets (generated by a
topology).
Since a Borel σ-algebra contains all the subsets generated by the intervals, you can recognize any subset
of a set using Borel σ-algebra. In other words, Borel σ-algebra corresponds to complete information.
Definition 14 A probability (measure) is a measure with the property that x (S) = 1.
Definition 15 Given a measure space (S, S, x), a function f : S → R is measurable (with respect to
the measure space) if, for all a ∈ R, we have
{b ∈ S | f (b) ≤ a} ∈ S.
One way to interpret a σ-algebra is that it describes the information available based on observations; a
structure to organize information, and how fine are the information that we receive. Suppose that S is
comprised of possible outcomes of a dice throw. If you have no information regarding the outcome of
the dice, the only possible sets in your σ-algebra can be ∅ and S. If you know that the number is even,
then the smallest σ-algebra given that information is S = {∅, {2, 4, 6} , {1, 3, 5} , S}. Measurability has
a similar interpretation. A function is measurable with respect to a σ-algebra S, if it can be evaluated
under the current measure space (S, S, x).
We can also generalize Markov transition matrix to any measurable space. This is what we do next.
Definition 16 A function Q : S × S → [0, 1] is a transition probability if
1. Q (·, s) is a probability measure for all s ∈ S; and,
2. Q (B, ·) is a measurable function for all B ∈ S.
Intuitively, given B ∈ S and s ∈ S, Q (B, s) gives the probability of being in set B tomorrow, given
that the state is s today. Consider the following example: a Markov chain with transition matrix given
8
Countable additivity means that the measure of the union of countable disjoint sets is the sum of the measure of these
sets.
33
by
 
0.2 0.2 0.6
Γ =  0.1 0.1 0.8  ,
0.3 0.5 0.2
on the set S = {1, 2, 3}, with the σ-algebra S = P (S) (where P (S) is the power set of S). If Γij
denotes the probability of state j happening, given a present state i, then
Q ({1, 2} , 3) = Γ31 + Γ32 = 0.3 + 0.5 .
As another example, suppose we are given a measure x on S; xi gives us the fraction of type i, for
i ∈ S. Given the previous transition function, we can calculate the fraction of types tomorrow using
the following formulas:
x10 = x1 Γ11 + x2 Γ21 + x3 Γ31 ,
x20 = x1 Γ12 + x2 Γ22 + x3 Γ32 ,
x30 = x1 Γ13 + x2 Γ23 + x3 Γ33 .
In other words
x0 = ΓT x,
where xT = (x1 , x2 , x3 ).
To extend this idea to a general case with a general transition function, we define an updating operator
as T (x, Q), which is a measure on S with respect to the σ-algebra S, such that
x 0 (B) =T (x, Q) (B)
Z
= Q (B, s) x (ds) , ∀B ∈ S.
S
A stationary distribution is a fixed point of T , that is x ∗ so that
x ∗ (B) = T (x ∗ , Q) (B) , ∀B ∈ S.
We know that, if Q has nice properties,9 then a unique stationary distribution exists (for example, we
discard flipping from one state to another), and
x ∗ = lim T n (x0 , Q) ,
n→∞
for any x0 in the space of measures on S.
Consider unemployment in a very simple economy (we have an exogenous transition matrix). There
are two states; employed and unemployed. The transition matrix is given by

0.95 0.05
Γ= .
0.50 0.50
Compute the stationary distribution corresponding to this Markov transition matrix.
9
See Chapter 11 in [5].
34
Industry Equilibrium
Preliminaries
Now we are going to study a type of models initiated by [2]. We will abandon the general equilibrium
framework from the previous section to study the dynamics of distribution of firms in a partial equilibrium
environment.
To motivate things, let’s start with the problem of a single firm that produces a good using labor input
according to a technology described by the production function f . Let us assume that this function is
increasing, strictly concave, with f (0) = 0. A firm that hires n units of labor is able to produce sf (n),
where s is a productivity parameter. Markets are competitive, in the sense that a firm takes prices as
given and chooses n in order to solve
π (s, p) = max {psf (n) − wn} .

n≥0
The first order condition implies that in the optimum, n∗ ,
psfn (n∗ ) = w .
Let us denote the solution to this problem as a function n∗ (s, p).10 Given the above assumptions, n∗
is an increasing function of s (i.e. more productive firms have more workers), as well as p.
Suppose now there is a mass of firms in the industry, each associated with a productivity parameter
s ∈ S ⊂ R+ , where
S := [s, s̄] .
Let S denote a σ-algebra on S (Borel σ-algebra for instance). Let x be a measure defined over the
space (S, S) that describes the cross sectional distribution of productivity among firms. Then, for any
B ⊂ S with B ∈ S, x (B) is the mass of firms having productivities in S.
We will use x to define statistics of the industry. For example, at this point, it is convenient to define
the aggregate supply of the industry. Since individual supply is just sf (n∗ (s, p)), the aggregate supply
can be written as11
Z
Y (p) = sf (n∗ (s, p)) x (ds) .
S
S
Observe that Y S is a function of the price p; for any price, p, Y S (p) gives us the supply in this
economy.
10
As we declared in advance, this is a partial equilibrium analysis. Hence, we ignore the dependence of the solution on
w to focus on the determination of p.
11
S in Y S stands for supply.
35
Search Wikipedia for an index of concentration in an industry, and adopt it for our economy.
Suppose now that the demand of the market is described by some function Y D (p). Then the equilibrium
price, p ∗ , is determined by the market clearing condition
Y D (p ∗ ) = Y S (p ∗ ) . (22)
So far, everything is too simple to be interesting. The ultimate goal here is to understand how the
object x is determined by the fundamentals of the industry. Hence, we will be adding tweaks to this
basic environment in order to obtain a theory of firms’ distribution in a competitive environment. Let’s
start by allowing firms to die.
A Simple Dynamic Environment
Consider now a dynamic environment, in which the situation above repeats every period. Firms discount
profits at rate rt , which is exogenously given. In addition, assume that a single firm, in each period,
faces a probability 1 − δ of disappearing! We will focus on stationary equilibria; i.e. equilibria in which
the price of the final output p, the rate of return, r , and the productivity of firm, s, stay constant
through time.
Notice first that firm’s decision problem is still a static problem; we can easily write the value of an
incumbent firm as
∞ t
X δ
V (s, p) = π (s, p)
t=0
1 + r

1+r
= π (s, p)
1+r −δ
Note that we are considering that p is fixed (therefore we can omit it from the expressions above).
Observe that every period there is positive mass of firms that die. Therefore, how can this economy be
in a stationary equilibrium? To achieve that, we have to assume that there is a constant flow of firms
entering the economy in each period, as well.
As before, let x be the measure describing the distribution of firms within the industry. The mass of
firms that die is given by (1 − δ)x (S). We will allow these firms to be replaced by new entrants. These
entrants draw a productivity parameter s from a probability measure γ.
One might ask what keeps these firms out of the market in the first place? If
π (s, p) = psf (n∗ (s, p)) − wn∗ (s, p) > 0,
which is the case for the firms operating in the market, then all the (potential) firms with productivities
in S would want to enter the market!
36
We can fix this flaw by assuming that there is a fixed entry cost that each firm must pay in order to
operate in the market, denoted by c E . Moreover, we will assume that the entrant has to pay this cost
before learning s. Hence the value of a new entrant is given by the following function:
Z
V (p) = V (s, p) γ (ds) − c E .
E
(23)
S
Entrants will continue to enter if V E is greater than 0, and decide not to enter if this value is less
than zero. As a result, stationarity occurs when V E is exactly equal to zero (this is the free entry
assumption, and we are assuming that there is an infinite number (mass) of prospective firms).
Let’s analyze how this environment shapes the distribution of firms in the market. Let xt be the cross
sectional distribution of firms in period t. For any B ⊂ S, portion 1 − δ of the firms with productivity
s ∈ B will die, and that will attract some newcomers. Hence, next period’s measure of firms on set B
will be given by:
xt+1 (B) = δxt (B) + mγ (B) .
That is, mass m of firms would enter the market in t + 1, and only fraction γ (B) of them will have
productivities in the set B. As you might suspect, this relationship must hold for every B ∈ S.
Moreover, since we are interested in stationary equilibria, the previous expression tells us that the cross
sectional distribution of firms will be completely determined by γ.
If we let mapping T be defined by

Tx (B) = δx (B) + mγ (B) , ∀B ∈ S, (24)
a stationary distribution of productivity is the fixed point of the mapping T ; i.e. x ∗ with Tx ∗ = x ∗ ,
implying:
m
x ∗ (B; m) = γ (B) , ∀B ∈ S.
1−δ
Now, note that the demand and supply relation in (22) takes the form:
Z
y (p (m)) = sf (n∗ (s, p)) dx ∗ (s; m) ,
d ∗
(25)
S
whose solution, p ∗ (m), is continuous function under regularity conditions stated in [5].
We have two equations, (23) and (25), and two unknowns, p and m. Thus, we can defined the
equilibrium as:
Definition 17 A stationary distribution for this environment consists of functions p ∗ , x ∗ , and m∗ , that
satisfy:
1. y d (p ∗ (m)) = S sf (n∗ (s, p)) dx ∗ (s; m);

R
R
2. s V (s, p) γ (ds) − c E = 0; and,
3. x ∗ (B) = δx ∗ (B) + m∗ γ (B) , ∀B ∈ S.
37
Introducing Exit Decisions
We want to introduce more (economic) content by making the exit of firms endogenous (a decision of
the firm). One way to do so is to assume that the productivity of the firms follow a Markov process
governed by a transition function, Γ. This would change the mapping T in Equation (24), as:
Z
Tx (B) = δ Γ (s, B) x (ds) + mγ (B) , ∀B ∈ S.
S
But, this wouldn’t add much economic content to our environment; firms still do not make any (inter-
esting) decision. To change this, let’s introduce cost of operation into the model; suppose firms have
to pay c v each period in order to stay in the market. In this case, when s is low, the firm’s profit
might not cover its cost of operation. So, the firm might decide to leave the market. However, firm
has already paid (a sunk cost of) c E , and, since s changes according to a Markov process, prospects
of future profits might deter the firm from quitting. Therefore, negative profit in one period does not
imply immediately that the firm’s optimal choice is to leave the market.
By adding such a minor change, the solution will have a reservation productivity property under some
conditions (to be discussed in the comment below). In words, there will be a minimum productivity,
s ∗ ∈ S, above which it is profitable for the firm to stay in the market.
To see this, note that the value of a firm with productivity s ∈ S in a period is given by
Z
1 0 0 v
V (s, p) = max 0, π (s, p) + Γ (s, ds ) V (s , p) − c .
(1 + r ) S
Show that the firm’s decision takes the form of a reservation productivity strategy, in which, for some
s ∗ ∈ S, s < s ∗ implies that the firm would leave the market.
In this case, the transition of the distribution of productivities on S will be:

Z s̄
0 ∗
x (B) = mγ (B ∩ [s , s̄]) + Γ (s, B ∩ [s ∗ , s̄]) x (ds) , ∀B ∈ S.
s∗
A stationary distribution of the firms in this economy, x ∗ , is the fixed point of this equation.
How productive does a firm have to be, to be in the top 10% largest firms in this economy? The
answer to this question is the solution to the following equation, ŝ:
R s̄ ∗
x (ds)
R ŝs̄ = 0.1 .
x ∗ (ds)
s ∗
Then, the fraction of the labor force in the top 10% largest firms in this economy, is
R s̄ ∗
n (s, p) x ∗ (ds)
R ŝs̄ .
s∗
n∗ (s, p) x ∗ (ds)
38
Compute the average growth rate of the smallest one third of the firms. What would be the fraction
of firms in the top 10% largest firms in the economy that remain in the top 10% in next period?
To see that this will be the case you should prove that the profit before variable cost function π (s, p) is
increasing in s. Hence the productivity threshold is given by the s ∗ that satisfies the following condition:
π (s ∗ , p) = cv
for an equilibrium price p. Now instead of considering γ as the probability measure describing the
distribution of productivities among entrants, you must consider γ
b defined as follows
γ (B ∩ [s ∗ , s])
γ
b (B) =
γ ([s ∗ , s])
for any B ∈ S.
One might suspect that this is an ad hoc way to introduce the exit decision. To make things more
concrete and easier to compute, we will assume that s is a Markov process. To facilitate the exposition,
let’s make S finite and assume s has transition matrix Γ. Assume further that Γ is regular enough so
that it has a stationary distribution γ ∗ . For the moment we will not put any additional structure on Γ.
The operation cost c v is such that the exit decision is meaningful. Let’s analyze the problem from the
perspective of the firm’s manager. He has now two things to decide. First, he asks himself the question
“Should I stay or should I go?”. Second, conditional on staying, he has to decide how much labor to
hire. Importantly, notice that this second decision is still a static decision. Later, we will introduce
adjustment cost that will make this decision a dynamic one.
Let φ (s, p) be the value of the firm before having decided whether to stay or to go. Let V (s, p) be
the value of the firm that has already decided to stay. V (s, p) satisfies
( )
1 X
V (s, p) = max spf (n) − n − c v + Γss 0 φ (s 0 , p)
n 1 + r s 0 ∈S
Each morning the firm chooses d in order to solve

φ (s, p) = max dV (s, p)
d∈{0,1}
Let d ∗ (s, p) be the optimal decision to this problem. Then d ∗ (s, p) = 1 means that the firm stays in
the market. One can alternatively write:
" #
1 X
φ (s, p) = max d π (s, p) − c v + Γss 0 φ (s 0 , p)
d∈{0,1} 1 + r s 0 ∈S
or even
" #
1 X
φ (s, p) = max π (s, p) − c v + Γss 0 φ (s 0 , p) , 0 (26)
1 + r s 0 ∈S
39
All these are valid. Additionally, one can easily add minor changes to make the exit decision more
interesting. For example, things like scrap value or liquidation costs will affect the second argument of
the max operator above, which so far is just zero.
What about d ∗ (s, p)? Given a price, this decision rule can take only finitely many values. Moreover,
if we could ensure that this decision is of the form “stay only if the productivity is high enough and go
otherwise” then the rule can be summarized by a unique number s ∗ ∈ S. Without doubt, that would
be very convenient, but we don’t have enough structure to ensure that such is the case. Because,
although the ordering of s (lower s are ordered before higher s) gives us that the value of s today
is bigger than value of smaller s 0 , depending on the Markov chain, on the other hand, the value of
productivity level s tomorrow may be lower than the value of s 0 (note s 0 < s) tomorrow. Therefore we
need some additional regularity conditions.
In order to get a cutoff rule for the exit decision, we need to add an assumption about the transition
matrix Γ. Let the notation Γ (s) indicate the probability distribution over next period state conditional
on being on state s today. You can think of it as being just a row of the transition matrix. Take s and
sP
b . We will say thatPthe matrix Γ displays first order stochastic dominance (FOSD) if s > b s implies
0 0
s 0 ≤b Γ (s | s) ≤ s 0 ≤b Γ (s | b
s ) for any b ∈ S. It turns out that FOSD is a sufficient condition for
having a cutoff rule. You can prove that by using the same kind of dynamic programming tricks that
have been used in a different course for obtaining the reservation wage property in search problems.
Try it as an exercise. Also note that this is just a sufficient condition.
Finally, we need to mention something about potential entrants. Since we will assume that they have
to pay the cost c E before learning their s, they can leave the industry even before producing anything.
That requires us to be careful when we describe industry dynamics.
Now the law of motion becomes:
Z s̄
0 ∗
x (B) = mγ (B ∩ [s , s̄]) + Γ (s, B ∩ [s ∗ , s̄]) x (ds) , ∀B ∈ S.
s∗
Stationary Equilibrium
Now that we have all the ingredients in the table, let’s define the equilibrium formally.
Definition 18 A stationary equilibrium for this environment consists of a list of functions (φ, n∗ , d ∗ ),
a price p ∗ and a measure x ∗ such that
1. Given p ∗ , the functions φ, n∗ , d ∗ solve the problem of the incumbent firm

2. V E (p ∗ ) = 0
3. For any B ∈ S (assuming we have a cut-off rule with s ∗ is cut-off in stationary distribution)12
12
If we do not have such cut-off rule we have to define
40
Z s̄
0 ∗
x (B) = mγ (B ∩ [s , s̄]) + Γ (s, B ∩ [s ∗ , s̄]) x (ds) .
s∗
You can think of condition (2) as a “no money left over the table” condition, which ensures additional
entrants find unprofitable to participate in the industry.
We can use this model to compute interesting statistics. For example the average output of the firm
is given by
sf (n∗ (s))x ∗ (ds)

P
Y
= P ∗
N x (ds)
Next, suppose that we want to compute the share of output produced by the top 1% of firms. To do
this we first need to compute e
s such that
Ps̄ ∗
s x (ds)
e
= .01
N
where N is the total measure of firms. Then the share output produced by these firms is given by
Ps̄
sf (n∗ (s))x ∗ (ds)
Pes̄s ∗ ∗
s sf (n (s))x (ds)
Suppose now that we want to compute the fraction of firms who are in the top 1% two periods in a
row. This is given by
XX
Γss 0 x ∗ (ds)
s s 0 ≥e
s≥e s
We can use this model to compute a variety of other statistics including the Gini coefficient.
Z X Z
x ∗ (B) = Γss 0 1{s 0 ∈B} 1{d(s 0 ,p∗ )=1} x ∗ (ds) + µ∗ 1{s∈B} 1{d(s,p∗ )=1} γ (ds)
S s 0 ∈S S
where
Z X
∗
µ = Γss 0 1{d(s 0 ,p∗ )=0} x ∗ (ds)
S s 0 ∈S
41
Adjustment Costs
To end with this section it is useful to think about environments in which firm’s productive decisions
are no longer static. A simple way of introducing dynamics is by adding adjustment costs.
We will consider labor adjustment costs.13 Consider a firm that enters period t with nt−1 units of
labor, hired in the previous period. We can consider three specifications for the adjustment costs, due
to hiring nt units of labor in t, c (nt , nt−1 ):
• Convex Adjustment Costs: if the firm wants to vary the units of labor, it has to pay α (nt − nt−1 )2
units of the numeraire good. The cost here depends on the size of the adjustment.
• Training Costs or Hiring Costs: if the firm wants to increase labor, it has to pay α [nt − (1 − δ) nt−1 ]2
units of the numeraire good, only if nt > nt−1 ; we can write this as
1{nt >nt−1 } α [nt − (1 − δ) nt−1 ]2 ,
where 1 is the indicator function, and δ measures the exogenous attrition of workers in each
period.
• Firing Costs.
The recursive formulation of the firm’s problem would be:

( )
1 X
V (s, n− , p) = max 0, max sf (n) − wn − c v − c (n, n− ) + Γss 0 V (s 0 , n, p) , (27)
n≥0 (1 + r ) s 0 ∈S
where c gives the specified cost of adjusting n− to n. Note that due to limited liability of the firm, the
exit value of the firm is 0 and not −c (0, n− ).
Now, a firm is characterized by both its productivity s and labor n− in the previous period. Note that
since the production function f has decreasing returns to scale, there exists an amount of labor N̄ such
that none of the firms hire labor greater than N̄. So, n− ∈ N := [0, N̄]. Let N be a σ-algebra on N.
If the labor policy function is n = g (s, n− ), then the law of motion now becomes:
Z s̄ Z N̄
0 ∗
S N S
1{g (s,n− )∈B N } Γ s, B S ∩ [s ∗ , s̄] x (ds, dn− ) ,

x B ,B = mγ B ∩ [s , s̄] 1{0∈B N } +
s∗ 0
∀B S ∈ S, ∀B N ∈ N.
Write the first order conditions for the problem in (27).

13
These costs work pretty much like capital adjustment costs, as one might suspect.
42
Define the recursive competitive equilibrium for this economy.
Another example of labor adjustment costs is when the firm has to post vacancies to attract labor. As
an example of such case, suppose the firm faces a firing cost according to the function c. The firm
also pays a cost κ to post vacancies, and after posting vacancies, it takes one period for the workers
to be hired. How can we write the problem of firms in this environment?
Incomplete Market Models
A Farmer’s Problem
Consider the following problem of a farmer:

( )
X
V (s, a) = max 0
u (c) + β Γss 0 V (s 0 , a0 ) (28)
c,a
s0
s.t. c + qa0 = a + s
c ≥0
a0 ≥ 0,
where a is his holding of coconuts, which can only take positive values, c is his consumption, and s is
amount of coconuts that he produces. s follows a Markov chain, taking values in a finite set S. q is
the price of coconuts. Note that, the constraint on the holding of coconuts tomorrow, is a constraint
imposed by nature; nature allows the farmer to store coconuts at rate 1/q, but, it does not allow him
to transfer coconuts from tomorrow to today.
We are going to consider this problem in the context of a partial equilibrium, where q is given. The
first question is, what can be said about q?
Assume there are no shocks in the economy, so that s is a fixed number. Then, we could write the
problem of the farmer as:
V (a) = max
0
{u (a + s − qa0 ) + βV (a0 )} .
c,a ≥0
If u is assumed to be logarithmic, the first order condition for this problem implies:
c0 β
≥ ,
c q
with equality if a0 > 0. Therefore, if q > β (i.e. nature is more stingy than farmer’s patience), then
c 0 < c, and the farmer dis-saves (at least, as long as a0 > 0). But, when q < β, consumption grows
without bound. This is the reason we put this assumption on the model, in what follows.
A crucial assumption for generating a bounded asset space is β/q < 1, stating that agents are sufficiently
impatient so they tend to consume more and decumulate their asset when they get richer and far away
43
from the non-negativity constraint, a0 ≥ 0. However, this does not mean that, when faced with a
possibility of very low consumption, agents would not save, even though the rate of return, 1/q, is
smaller than the rate of impatience 1/β.
The first order condition for farmer’s problem (28) is given by:
βX
uc (c (s, a)) ≥ Γss 0 uc (c (s 0 , a0 (s, a))) ,
q s0
with equality when a0 (s, a) > 0, where c (·) and a0 (·) are policy functions from the farmer’s prob-
lem. Notice that a0 (s, a) = 0 is possible, if we assume appropriate shock structure in the economy.
Specifically, it depends on the value of smin := minsi ∈S si .
The solution to the problem of the farmer, for a given value of q, implies a distribution of coconut
holdings in each period. This distribution, together with the Markov chain describing the evolution of
s, can be summed together as a single probability measure for the distribution of shocks and assets
(coconut holdings) over the product space E = S × R+ , and its σ-algebra, B, which we denote by X .
The evolution of this probability measure is given by:
XZ
0
X (B) = Γss 0 1{a0 (s,a)∈Ba } dX (s, a) , ∀B ∈ B, (29)
s 0 ∈B s
where Bs and Ba are the S-section and R-section of B (projections of B on S and R+ ), respectively, and
e (Γ, a0 , ·) be the mapping associated with (29) (the adjoint operator),
1 is the indicator function. Let T
so that:
X 0 (B) = T
e (Γ, a0 , X ) (B) , ∀B ∈ B.
e n (Γ, a0 , ·) as:
Define T

e n (Γ, a0 , X ) = T
T e Γ, a0 , T
e n−1 (Γ, a0 , X ) .
Then, we have the following theorem.
e (Γ, a0 , ·), there is a unique probability measure X ∗ , so that:

Theorem 3 Under some conditions on T
X ∗ (B) = lim T
e n (Γ, a0 , X0 ) (B) , ∀B ∈ B,
n→∞
for all initial probability measures X0 on (E , B).
A condition that makes things considerably easier for this theorem to hold is that E is a compact set.
Then, we can use Theorem (12.12) in [5], to show this result holds. Given that S is finite, this is
equivalent to a compact support for the distribution of asset holdings. We discuss this in detail in
Appendix A.
44
Huggett’s Economy[3]
Now we modify the farmer’s problem in (28) a little bit:

( )
X
V (s, a) = max
0
u (c) + β Γss 0 V (s 0 , a0 ) (30)
c,a
s0
s.t. c + qa0 = a + s
c ≥0
a0 ≥ a,
where a < 0, so now farmers can borrow and lend with each other, but with a borrowing limit. We
continue to make the same assumption on q; i.e. β/q < 1. Solving this problem gives the policy
function a0 (s, a) . It is easy to extend the analysis in the last section to show that there is an upper
bound of the asset space, which we denote by ā, so that for any a ∈ A := [a, ā], a0 (s, a) ∈ A, for any
s ∈ S.
One possibility for a is what we call the natural borrowing constraint. This is the constraint that ensures
the agent can pay back his debt for sure, no matter what the nature reveals (whatever sequence of
idiosyncratic shocks is realized). This is given by
smin
an := − .
1
q
−1
If we impose this constraint on (30), even when the farmer receives an infinite sequence of bad shocks,
he can pay back his debt by setting his consumption equal to zero, forever.
But, what makes this problem more interesting is to tighten this borrowing constraint; the natural
borrowing constraint is very unlikely to be hit. One way to restrict borrowing is no borrowing at all, as
in the previous section. Another case is to choose 0 > a > an . This is the case that we consider in this
section.
Now suppose there is a (unit) mass of farmers with distribution function X (·) , where X (D, B) denotes
fraction of people with shock s ∈ D and a ∈ B, where D in an element of the power set of S , P (S)
(which, when S is finite, is the natural σ-algebra over S) and B is a Borel subset of A, B ∈ A. Then
the distribution of farmers tomorrow is given by:
Z X
0 0 0
X (S , B ) = 1{a0 (s,a)∈B 0 } Γss 0 dX (s, a) , (31)
A×S s 0 ∈S 0
for any S 0 ∈ P (S) and B 0 ∈ A.
Implicitly this defines an operator T such that T (X ) = X 0 . If T is sufficiently nice, then there exits
a unique X ∗ such that X ∗ = T (X ∗ ) and X ∗ = limn→∞ T n (X0 ), for any initial distribution over the
45
product space S ×A, X0 . Note that the decision rule is obtained given q. Hence, the resulting stationary
distribution X ∗ also depends on q. So, let us denote it by X ∗ (q).
To determine the equilibrium value of q, in a general equilibrium setting, consider the following variable
(as a function of q):
Z
adX ∗ (q) .
A×S
This expression give us the average asset holdings, given the price, q.
We want to determine an endogenous q, so that the asset market clears; we assume that there is no
storage technology so that asset supply is 0 in equilibrium. Hence, price q should be such that asset
demand equals asset supply; i.e.
Z
adX ∗ (q) = 0.
A×S
In this sense, equilibrium price q is the price that generates the stationary distribution that clears the
asset market.
We can show that a solution exists by invoking intermediate value theorem, by showing that the
following three conditions are satisfied (note that q ∈ [β, ∞]):
adX ∗ (q) is a continuous function of q;

R
1. A×S
2. limq→β A×S adX ∗ (q) → ∞; (Intuitively, as q → β, interest rate increases. Hence, agents would
R
like to save more. Together with precautionary savings motive, they accumulate unbounded
quantities of assets in the stationary equilibrium, in this case.) and,
3. limq→∞ A×S adX ∗ (q) < 0. (This is also intuitive; as q → ∞, interest rate converges to 0.
R
Hence, everyone would rather borrow.)
Aiyagari’s Economy[1]
In an Aiyagari Economy, there is physical capital. In this sense, the average asset holdings in the
economy must be equal to the average (physical) capital. So, if we keep denoting the stationary
distribution of assets by X ∗ , we must have:
Z
adX ∗ (q) = K ,
A×S
where A is the support of the distribution of wealth. (It is not difficult to see that this set is compact.)
46
On the other hand, the shocks affect the labor income; we can think of these shocks as fluctuations
in the employment status of individuals. Thus, the problem of an individual in this economy can be
written as:
( )
X
0 0
V (s, a) = max
0
u (c) + β Γss 0 V (s , a ) (32)
c,a
s0
0
s.t. c + a = (1 + r ) a + ws
c ≥0
a0 ≥ a,
where, here, r is the return to savings, and w is the wage rate. Therefore,
Z
sdX ∗ (q)
A×S
gives the average labor in this economy. If we decide to think of agents supplying one unit of labor,
then, we may think of the expression as determining the effective labor supply.
We assume the standard constant returns to scale production technology for the firm, of the form:
F (K , L) = AK 1−α Lα ,
with the rate of depreciation δ. Hence:
r =Fk (K , L) − δ
−α
K
= (1 − α) A −δ
L

K
=:r ,
L
and
w =Fl (K , L)
1−α
K
=αA
L

K
=:w .
L
In terms of Huggett [3], q, the price of assets, is given by
1 1
q= = .
(1 + r ) [1 + Fk (K , L) − δ]
Therefore, prices faced by agents are all functions of the capital-labor ratio. As a result, we may write
the stationary distribution of assets as a function of capital-labor ratio as well, X ∗ KL . Thus, the
47
equilibrium condition becomes:
∗ K
R
K adX L
= RA×S ∗ K
.
L A×S
sdX L
Using this condition, one can solve for the equilibrium capital-labor ratio.
Policy
In Aiyagari [1] [or Huggett [3]] economy, model parameters can be summarized by θ = {u, β, s, Γ, F }.
In stationary equilibrium, value function v (s, a; θ) as well as X ∗ (θ) can be obtained, where X ∗ (θ)
is a mapping from model parameters to stationary distribution of agent’s asset holding and shocks.
Suppose now there is a policy change that shifts
θ to θ̂ = {u, s, Γ̂, F }. Associated with this new
β,
environment there is a new value function v s, a; θ̂ and X ∗ θ̂ . Define η (s, a) to be the solution
of:

v s, a + η (s, a) , θ̂ = v (s, a, θ)
which is the transfer payment necessary to the households so that they are indifferent between living
in the old environment and in the new one. Hence total payment needed to compensate households
for this policy change is given by:
Z
η (s, a) dX ∗ (θ)
A×S
Notice that the changes do not take place when the government is trying to compensate the households.
Hence we use the original stationary distribution associated with θ to aggregate the households.
R
If A×S v (s, a) dX ∗ θ̂ > A×S v (s, a) dX ∗ (θ) , does this necessarily mean that households are willing
R
to accept this policy change? The answer is not necessarily because the economy may well spend a
long time in the transition path to the new steady state, during which there may be severe welfare loss.
Aggregate Shocks
In this section, we consider an economy that is subject to both aggregate and idiosyncratic shocks, at
the same time; consider Aiyagari’s [1] economy again, now, with a production function that is subject
to an aggregate shock; zF K , N̄ .
Let X be the distribution of types; then the aggregate capital is given by:
Z
K = adX (s, a) .
K 0 = G (z, K )
48
The question is what are the sufficient statistics for predicting the aggregate capital stock and, con-
sequently, prices tomorrow? Are z and K sufficient detemine capital tomorrow? The answer to these
questions is no, in general; this is true if, and only if, the decision rules are linear. Therefore, X , the
distribution of types becomes a state variable (even in the stationary equilibrium) for this economy.
Then, the problem of an individual becomes:

( )
0
X
V (z, X , s, a) = max
0
u (c) + β Πzz 0 Γzss 0 V (z 0 , X 0 , s 0 , a0 )
a
z 0 ,s 0
s.t. c + a0 = azfk K , N̄ + szfn K , N̄

Z
K = adX (s, a)
X 0 = G (z, X )
c, a0 ≥ 0.
Computationally, this problem is a beast! So, how can we solve it? To provide some idea, we will first
consider an economy with dumb agents!
Consider an economy in which people are stupid; people believe tomorrow’s capital depends only on
K , and not X . This, obviously, is not an economy in which expectations are rational. Nevertheless,
people’s problem in such settings becomes:
( )
0
X
V
e (z, X , s, a) = max u (c) + β Πzz 0 Γz 0 V (z 0 , X 0 , s 0 , a0 )
ss
a0
z 0 ,s 0
s.t. c + a0 = azfk K , N̄ + szfn K , N̄

Z
K = adX (s, a)
X0 = G e (z, K )
c, a0 ≥ 0.
Next step is to allow people become slightly smarter; they now can use extra information, like mean
and variance of X , to predict X 0 . Does this economy work better than our dumb benchmark? Compu-
tationally no! This answer, as stupid as it may sound, has an important message: people actually act
linearly in the economy; decision rules are approximately linear. Therefore, we may use Aiyagari’s [1]
results without fear; the approximations are quite reliable!
Monopolistic Competition, Endogenous Growth and R&D
So far, we have seen the neoclassical growth model as our benchmark model, and built on it for
the analysis of more interesting economic questions. One peculiar characteristic of our benchmark
model, unlike its name suggests, is the lack of growth (after reaching the steady state), whereas, many
49
interesting questions in economics are related to the cross-country differences of growth rates. To see
why this is the case, consider the standard neoclassical technology:
F (K , N) = AK θ1 Lθ2 ,
for some θ1 , θ2 ≥ 0. We already know that the only possible case that is consistent with the notion
of competitive equilibrium is that θ1 + θ2 = 1. However, this implies a decreasing marginal rate of
product for capital. Given a fixed quantity for labor supply, in the presence of depreciation, this implies
a maximum sustainable capital stock, and puts a limit on the sustainable growth; economy converges
to some steady-state, without exhibiting any balanced growth.
So if our economy is to experience sustainable growth for a long period of time, we either give up the
curvature assumption on our technology, or we have to be able to shift our production function upwards.
Given a fixed amount of labor, this shift is possible either by an increasing (total factor) productivity
parameter or increasing labor productivity. We will cover a model that will allow for growth, so that
we will be able to attempt to answer such questions.
Consider the following economy due to the highly cited model of endogenous growth[4]; there are three
sectors in the economy: a final good sector, an intermediate goods sector, and an R&D sector. Final
goods are produced using labor (as we will see there is only one wage, since there is only one type of
labor) and intermediate goods according to the production function
Z At
α
N1,t xt (i)1−α di
0
where x (i) denotes the utilization of intermediate good of variety i ∈ [0, At ].14 Note that marginal
contribution of each variety is decreasing (since α < 1), however, an increase in the number of varieties
would increase the output. We will assume that the final good producers operate in a competitive
market.
If the price of all varieties are the same, what is the optimal choice of input vector for a producer?
What if they do not have the same amount? Would a firm decide not to use a variety in the production?
Intermediate producers are monopolists that have access to a differentiated technology of the form:
k (i)
x (i) = .
η
Therefore, they can end up charging a mark up above the marginal cost for their product. This is the
main force behind research and development in this economy; developer of a new variety is the sole
proprietor of the blue print that allows himRearn profit. It is easy to observe that the aggregate demand
A
of capital from the intermediate sector is 0 t ηx (i) di.
14
The function that aggregates consumption of intermediate goods is often referred as Dixit-Stiglitz aggregator.
50
The R&D sector in the economy is characterized by a flow of intermediate goods in each period; a new
good is a new variety of the intermediate good. The flow of the new intermediate goods is created by
using labor, according to the following production technology:
At+1
= 1 + ξN2,t .
At
Notice that, after some manipulation, one can express growth in the stock of intermediate goods as
follows:
At+1 − At = At ξN2,t . (33)
Hence, the flow of new intermediate goods is determined by the current stock of them in the economy.
This type of externality in the model is the key propeller in the model. This assumption provides us
with a constant returns to scale technology in the R&D sector. In what follows, we will assume that
the inventors act as price takers in the economy.
The reason we see At on the right hand side of (33) as an externality is that it is indeed used as an
input in the process of R&D, while, it is not paid for. Thus, inventors, in a sense, do not pay the
investors of the previous varieties, while using their inventions. They only pay for the labor they hire.
Perhaps, the basic idea of this production function might be traced back to Isaac Newton’s quote: “If
I have seen further, it is only by standing on the shoulders of giants”.
The preferences of the consumers are represented by the following utility function:
∞
X
β t u (ct ) ,
t=0
and their budget constraint in period t is give by:

ct + kt+1 ≤ rt kt + wt + (1 − δ) kt .
In this economy, GDP, in terms of gross product, is given by:

Yt = Wt + rt Kt + πt ,
where πt is the net profits. On the other hand, in terms of expenditures, GDP is:
Yt = Ct + Kt+1 − (1 − δ) Kt + πt ,
where Kt+1 − (1 − δ) Kt is the investment in physical capital. At last, in terms of value added, it is
given by:
Z At
α
Yt = Nt xt (i)1−α di + pt (At+1 − At ) .
0
Certainly, this is not a model that one can map to the data. Instead it has been carefully crafted to
deliver what is desired and it provides an interesting insight in thinking about endogenous growth.
51
Solving the Model Let’s first consider the problem of a final good producer; in every period, he
chooses N1,t and xt (i), for every i ∈ [0, At ], in order to solve:
Z At Z At
α 1−α
max N1,t xt (i) di − wt N1,t − qt (i) xt (i) di,
0 0
where qt (i) is the price of variety i in period t. First order conditions for this problem are:
R At
α−1
1. N1,t : αN1,t 0
xt (i)1−α di = wt ; and,
α
2. xt (i): (1 − α) N1,t xt (i)−α = qt (i), for all i ∈ [0, At ].
From the second condition, one obtains:

α1
(1 − α)
xt (i) = N1,t , (34)
qt (i)
which, given N1t , is the demand function for variety i, by the final good producer.
Next, let’s consider the problem of an intermediate firm; these firms acts as price setters. The reason is
the ownership of a differentiated patent, whose sole owner is the intermediate good producer of variety
i. In addition, as long as α < 1, this variety does not have a perfect substitute, and always demanded
in the equilibrium. Therefore, their problem is to choose qt (i), in order to solve:
πt (i) = max qt (i) xt (qt (i)) − rt ηxt (qt (i))

1
(1 − α) α
s.t. xt (qt (i)) = N1,t ,
qt (i)
where xt (qt (i)) is the demand function, substituted from (34). Notice that we have substituted for
the technology of the monopolist, x (i) = k (i) /η. First order condition for this problem, with respect
to qt (i), is:
∂xt (qt (i))

xt (qt (i)) + (qt (i) − rt η) = 0,
∂qt (i)
which can be written as

1 1
(1 − α) α (qt (i) − rt η) (1 − α) α
1 N1,t = 1+α N1,t .
qt (i) α α qt (i) α
Rearranging yields:
1
qt (i) = rt η. (35)
(1 − α)
52
This is the familiar pricing function of a monopolist; price is marked-up above the marginal cost.
By substituting (35) into (34), we get:

" # α1
(1 − α)2
xt (i) = N1,t , (36)
rt η
and the demand for capital services is simply ηxt (i). In a symmetric equilibrium, where all the inter-
mediate good producers choose the same pricing rule, we have:
Z At
kt
xt (i) di = At xt = ,
0 η
where xt is the common supply of intermediate goods. Therefore:
kt
xt = .
ηAt
Moreover, if we let Yt be the production of the final good; by plugging (36) we get:
" # 1−α
(1 − α)2
α
Yt = N1,t At . (37)
rt η
Hence the model displays constant returns to scale in N1,t and At .
Let us study the problem of the R&D firms, next; a representative firm in this sector (recall that this
is a competitive sector) chooses N2,t , in order to solve the following problem:
max pt At ξN2,t − wt N2,t .

N2,t
The first order condition for this problem implies:

wt
pt = .
At ξ
In summary, there are two equations to be solved form; one relating the choice of consumption versus
saving (or capital accumulation), and one dividing labor demand for R&D, and that for final good
production. Consumption-investment decisions result from solving household’s problem in equilibrium,
and the corresponding Euler equation:
u 0 (ct ) = βu 0 (ct+1 ) [rt+1 + (1 − δ)] .
For determining the labor choices N1,t and N2,t , first note that the demand for patterns produced by
R&D sector, is from the prospect monopolists. As long as there is positive profit from buying demand,
the new monopolists would keep entering markets in a given period. This fact derives the profits of
53
prospect monopolists to zero. So, the lifetime profit of the monopolist, must be equal to the price he
pays for the blueprints;
∞ s
!
X Y 1
pt = πs .
s=t τ =t
1 + rτ − δ
This completes the solution to the model. Notice that output can grow at the same rate as At , from
Equation (37). In addition, Kt grows at the same rate. As a result, the rate of growth of At would
be the important aspect of equilibrium. For instance, if At grows at rate γ in the long-run, we have
a balanced growth path in equilibrium. This growth comes from the externality in the R&D sector.
Without that, we cannot get sustained growth in this model. The nice thing about this model is how
neat is is in delivering the balanced growth, with just enough structure imposed on the economy.
A Farmer’s Problem: Revisited
Consider the following problem of a farmer that we studied in class:

( )
X
V (s, a) = max 0
u (c) + β Γss 0 V (s 0 , a0 ) (38)
c,a
s0
s.t. c + qa0 = a + s
c ≥0
a0 ≥ 0.
As we discussed, we are in particular interested in the case where β/q < 1. In what follows, we are
going to show that, under monotonicity assumption on the Markov chain governing s, the optimal
policy associated with (38) implies a finite support for the distribution of asset holding of the farmer,
a.15
Before we start the formal proof, suppose smin = 0, and Γssmin > 0, for all s ∈ S. Then, the agent
will optimally always choose a0 > 0. Otherwise, there is a strictly positive probability that the agent
enters tomorrow into state smin , where he has no cash in hand (a0 + smin = 0) and is forced to consume
0, which is extremely painful to him (e.g. when Inada conditions hold for the instantaneous utility).
Hence he will raise his asset holding a0 to insure himself against such risk.
If smin > 0, then the above argument no longer holds, and it is indeed possible for the farmer to choose
zero assets for tomorrow.
Notice that the borrowing constraint a0 ≥ 0 is affecting agent’s asset accumulation decisions, even if
he is away from the zero bound, because he has an incentive to ensure against the risk of getting a
15
This section was prepared by Keyvan Eslami, at the University of Minnesota. This section is essentially a slight
variation on the proofs found in [3]. However, he accepts the responsibility for the errors.
54
Figure 1: Policy function associated with farmer’s problem.
series of bad shocks to s and is forced to 0 asset holdings. This is what we call precautionary savings
motive.
Next, we are going to prove that the policy function associated with (38), which we denote by a0 (·),
is similar to that in Figure 1. We are going to do so, under the following assumption.
Assumption 1 The Markov chain governing the state s is monotone; i.e. for any s1 , s2 ∈ S, s2 > s1
implies E (s|s2 ) ≥ E (s|s1 ).
It is straightforward to show that, the value function for Problem (38) is concave in a, and bounded.
Now, we can state our intended result as the following theorem.
Theorem 4 Under Assumption 1, when β/q < 1, there exists some â ≥ 0 so that, for any a ∈ [0, â],
a0 (s, a) ∈ [0, â], for any realization of s.
To prove this theorem, we proceed in the following steps. In all the following lemmas, we will assume
that the hypotheses of Theorem 4 hold.
Lemma 1 The policy function for consumption is increasing in a and s;
ca (a, s) ≥ 0 and cs (a, s) ≥ 0.
55
By the first order condition, we have:

0 βX 0 a + s − c (s, a)
u (c (s, a)) ≥ Γss 0 Va s , ,
q s0 q
with equality, when a + s − c (s, a) > 0.
For the first part of the lemma, suppose a increases, while c (s, a) decreases. Then, by concavity of
u, the left hand side of the above equation increases. By concavity of the value function, V , the right
hand side of this equation decreases, which is a contradiction.
For the the second part, we claim that Va (s, a) is a decreasing function of s. To show this is the case,
firs consider the mapping T as follows:
( )
X
Tv (s, a) = max
0
u (c) + β Γss 0 v (s 0 , a0 )
c,a
s0
0
s.t. c + qa = a + s
c ≥0
a0 ≥ 0.
Suppose van (s, a) is decreasing in its first argument; i.e. van (s2 , a) < van (s1 , a), for all s2 > s1 and
s1 , s2 ∈ S. We claim that, v n+1 = Tv n inherits the same property. To see why, note that for
an+1 (s, a) = a0 (where an+1 is the policy function associated with n’th iteration) we must have:
βX
u 0 (a + s − qa0 ) ≥ Γss 0 van (s 0 , a0 ) ,
q s0
with strict equality when a0 > 0. For a fixed value of a0 , an increase in s leads to a decrease in both
sides of this equality, due to the monotonicity assumption of Γ, and the assumption on van . As a result,
we must have
u 0 a + s2 − qan+1 (s2 , a) ≤ u 0 a + s1 − qan+1 (s1 , a) ,

for all s2 > s1 . By Envelope theorem, then:

van+1 (s2 , a) ≤ van+1 (s1 , a) .
It is straightforward to show that v n converges to the value function V point-wise. Therefore,

Va (s2 , a) ≤ Va (s1 , a) ,
for all s2 > s1 .
Now, note that, by envelope theorem:

Va (s, a) = u 0 (c (s, a)) .
As s increases, Va (s, a) decreases. This implies c (s, a) must increase.
56
Lemma 2 There exists some â ∈ R+ , such that ∀a ∈ [0, â], a0 (a, smin ) = 0.
It is easy to see that, for a = 0, a0 (a, smin ) = 0. First of all, note the first order condition:
βX
uc (c (s, a)) ≥ Γss 0 uc (c (s 0 , a0 (s, a))) ,
q s0
with equality when a0 (s, a) > 0. Under the assumption that β/q < 1, we have:
βX
uc (c (smin , 0)) = Γs s 0 uc (c (s 0 , a0 (smin , 0)))
q s 0 min
X
< Γss 0 uc (c (s 0 , a0 (smin , 0))) .
s0
By Lemma 1, if a0 = a0 (0, smin ) > a = 0, then c (s 0 , a0 ) > c (smin , 0) for all s 0 ∈ S, which leads to a
contradiction.
Lemma 3 a0 (smin , a) < a, for all a > 0.
Suppose not; then a0 (smin , a) ≥ a > 0 and as we showed in Lemma 2:

X
uc (c (smin , a)) < Γss 0 uc (c (s 0 , a0 (smin , a))) .
s0
Contradiction, since a0 (smin , a) ≥ a, and s 0 ≥ smin , and the policy function in monotone.
Lemma 4 There exits an upper bound for the agent’s asset holding.
Suppose not; we have already shown that a0 (smin , a) lies below the 45 degree line. Suppose this is not
true for a0 (smax , a); i.e. for all a ≥ 0, a0 (smax , a) > a. Consider two cases.
In the first case, suppose the policy functions for a0 (smax , a) and a0 (smin , a) diverge as a → ∞, so that,
for all A ∈ R+ , there exist some a ∈ R+ , such that:
a0 (smax , a) − a0 (smin , a) ≥ A.
Since S is finite, this implies, for all C ∈ R+ , there exist some a ∈ R+ , so that
c (smin , a) − c (smax , a) ≥ C ,
which is a contradiction, since c is monotone in s.
Next, assume a0 (smax , a) and a0 (smin , a) do not diverge as a → ∞. We claim that, as a → ∞, c must
grow without bound. This is quite easy to see; note that, by envelope condition:
Va (s, a) = u 0 (c (s, a)) .
57
The fact that V is bounded, then, implies that Va must converge to zero as a → ∞, implying that
c (s, a) must diverge to infinity for all values of s, as a → ∞. But, this implies, if a0 (smax , a) > a,
X
uc (c (smax , a0 (smax , a))) → Γsmax s 0 uc (c (s 0 , a0 (smax , a))) .
s0
As a result, for large enough values of a, we may write:

βX
uc (c (smax , a)) = Γs s 0 uc (c (s 0 , a0 (smax , a)))
q s 0 max
X
< Γsmax s 0 uc (c (s 0 , a0 (smax , a)))
s0
≈uc (c (smax , a0 (smax , a))) .
But, this implies:
c (smax , a) > c (smax , a0 (smax , a)) ,
which, by monotonicity of policy function, means a > a0 (smax , a), and this is a contradiction.
58
References
[1] Aiyagari, S. R. (1994): “Uninsured Idiosyncratic Risk and Aggregate Saving,” The Quarterly
Journal of Economics, 659–684.
[2] Hopenhayn, H. A. (1992): “Exit, Selection, and the Value of Firms,” Journal of Economic
Dynamics and Control, 16, 621–653.
[3] Huggett, M. (1993): “The Risk-Free Rate in Heterogeneous-Agent Incomplete-Insurance

Economies,” Journal of Economic Dynamics and Control, 17, 953–969.
[4] Romer, P. M. (1990): “Endogenous Technological Change,” The Journal of Political Economy,
98, 2.
[5] Stokey, Nancy L., R. E. L. J. and E. C. Prescott (1989): Recursive Methods in

Economic Dynamics, Harvard University Press.
59

Econ 702 Macroeconomic Theory Spring 2016

Uploaded by

Copyright:

Available Formats

Econ 702 Macroeconomic Theory Spring 2016

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

Econ 702 Macroeconomic Theory Spring 2016

Uploaded by

Copyright:

Available Formats

Econ 702 Macroeconomic Theory Spring 2016∗

April 26, 2016

2 Review: Neoclassical Growth Model 5

2.1 The Neoclassical Growth Model Without Uncertainty . . . . . . . . . . . . . . . . . . 5

2.2 A Comment on the Welfare Theorems . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3 Recursive Competitive Equilibrium 8

3.1 A Simple Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3.2 Adding Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3.2.1 Markov Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3.2.2 Problem of the Social Planner . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3.2.3 Recursive Competitive Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . 12

3.3 Economy with Government Expenditures . . . . . . . . . . . . . . . . . . . . . . . . 13

3.3.1 Lump Sum Tax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3.3.2 Labor Income Tax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.3.3 Capital Income Tax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.3.4 Taxes and Debt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

4.1 Heterogeneity in Wealth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

4.2 Heterogeneity in Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

4.3 An International Economy Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

5 Some Other Examples 20

5.1 A Few Popular Utility Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

5.2 An Economy with Capital and Land . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

6 Asset Pricing: Lucas Tree Model 22

6.1 The Lucas Tree with Random Endowments . . . . . . . . . . . . . . . . . . . . . . . 22

6.2 Asset Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

6.3 Taste Shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

7 Endogenous Productivity in a Product Search Model 25

7.1 Competitive Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

7.1.1 Problem of Individuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

7.1.2 Firms’ Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

7.1.4 Pareto Optimality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

9.2 A Simple Dynamic Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

9.3 Introducing Exit Decisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

9.4 Stationary Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

10 Incomplete Market Models 43

10.1 A Farmer’s Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

10.2 Huggett’s Economy[3] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

10.3 Aiyagari’s Economy[1] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

10.3.2 Aggregate Shocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

11 Monopolistic Competition, Endogenous Growth and R&D 49

A A Farmer’s Problem: Revisited 54

1. Output per capita has grown at a roughly constant rate (2%).

6. Labor income as a share of output has remained roughly constant (0.66).

7. Hours worked per capita have been roughly constant.

Review: Neoclassical Growth Model

We review briefly the basic neoclassical growth model.

The Neoclassical Growth Model Without Uncertainty

The commodity space is

ct , kt+1 ≥ 0, x1t + (1 − δ)kt = ct + kt+1 , −kt ≤ x2t ≤ 0, −1 ≤ x3t ≤ 0, k0 = k 0 }.

Definition 1 An Arrow-Debreu equilibrium is (x ∗ , y ∗ ) ∈ X × Y , and a continuous linear functional ν ∗

A Comment on the Welfare Theorems

Recursive Competitive Equilibrium

Now we can define the Recursive Competitive Equilibrium.

Definition 2 A Recursive Competitive Equilibrium with arbitrary expectations G is a set of functions2

1. given G ; Ω, g solves the household problem in (RCE),

Definition 3 A Rational Expectations (Recursive) Equilibrium is a set of functions Ω, g , R, w , G ∗ ,

1. Ω(K , a; G ∗ ), g (K , a; G ∗ ) solves HH problem in (RCE),

Pr(zt+1 = z j |ht ) = Γij , zt (ht ) = z i