MIT LIBRARIES
3 9080 01
tf§.
:..-.
..."
Bi;.'
-'.
<
-
:
:
:
.
iimlkimlmmwMM
•
;
:l
ffiM/lMS/'W!')
ft'/iMi/'W/l
iff!
HB31
.M415
DEWEY
X)0 .00- I
Massachusetts Institute of Technology
Department of Economics
Working Paper Series
CREDIT MARKET IMPERFECTIONS AND PERSISTENT
UNEMPLOYMENT*
Daron Acemoglu,
MIT
Working Paper 00-20
SEPTEMBER 2000
Room E52-251
50 Memorial Drive
Cambridge, MA 02142
downloaded without charge from the
Social Science Research Network Paper Collection at
http://papers.ssrn. com/paper. taf?abstract_id=242228
This paper can be
INSTITUTE
MASS/
OF TECHNOLOGY
OCT
3
2000
LIBRARIES
Massachusetts Institute of Technology
Department of Economics
Working Paper Series
CREDIT MARKET IMPERFECTIONS AND PERSISTENT
UNEMPLOYMENT*
Daron Acemoglu, MIT
Working Paper 00-20
SEPTEMBER 2000
Room
E52-251
50 Memorial Drive
Cambridge, MA 02142
downloaded without charge from the
Social Science Research Network Paper Collection at
http: //papers, ssrn. com/paper. taf?abstract_id=242228
This paper can be
Digitized by the Internet Archive
in
2011 with funding from
Boston Library Consortium
Member
Libraries
http://www.archive.org/details/creditmarketimpe00acem2
Credit Market Imperfections and Persistent
Unemployment
*
Daron Acemoglu
Department of Economics
Massachusetts Institute of Technology
50 Memorial Drive
MA
Cambridge,
September
11,
02142
2000
Abstract: This paper develops the thesis that credit market frictions
contributor to high unemployment in Europe.
new
When
may be an important
a change in the technological regime ne-
can happen relatively rapidly
where credit
markets function efficiently. In contrast, in Europe, job creation is constrained by credit market
imperfections, so unemployment rises and remains high for an extended period. The data show
that there has not been slower growth in the' most credit dependent industries in Europe relative
cessitates the creation of
to the U.S., but the share of
firms, this
employment
in these industries
is
in the U.S.
lower than in the U.S..
This
suggests that although credit market imperfections are unlikely to have been the major cause of
the increase in European unemployment, they
may have
played some role in limiting European
employment growth.
'Paper prepared
European Economic Association 2000, Bolzano, Italy- I thank Pol Antras for
and Abhijit Banerjee, Olivier Blanchard, Jaurae Ventura and seminar participants
European Economic Association 2000 Conference for helpful comments.
for the
lent research assistance,
excelin
the
Introduction
l
The problems facing European labor markets are now well-known, and
to explain how unemployment could be so high for such a long time in
economies
(see, for
a
number
of
European
example, Layard, Nickell, and Jackman, 1991; Bean, 1994 or Nickell, 1997
Most economists believe that the
reviews).
a large literature tries
are, at least in part, responsible for the
rigid labor
for
market institutions of European economies
high rates of European unemployment. In a recent review
"Labor Market Rigidities: At the Root of Unemployment in
Siebert put this in a stark form:
Europe".
It
would not be an unfair statement, however,
a loss in understanding the European
market
economics profession
is
plausible, there have
been only a few major changes
the institutions of European economies over the 1970's to lead to the high rates of
of the 1980's
and 1990's
(see, for
is still
at
unemployment phenomenon. Although the idea that labor
cause unemployment
rigidities
to say that the
example, Nickell, 1997). For this reason,
in
unemployment
many economists have
emphasized not only institutions, but also other macroeconomic changes.
The famous Krugman
ment
hypothesis, for example, explains the increase in European unemploy-
as a result of the interaction between skill-biased technical change
stitutions
(Krugman, 1994, OEC'D,
international trade put
Adrian
Wood
in-
(1994) similarly argues that increased
of less skilled workers,
and unemployment
European labor markets prevented unskilled wages from
resulted because the rigid
recently, Ball (1997)
1994).
downward pressure on the wages
and labor market
falling.
More
and Blanchard and Wolfers (1999) have emphasized the interaction between,
macroeconomic shocks, such
as disinflation, the
slowdown
in
TFP
growth and the increase
in
the real interest rates, and the rigid labor market institutions as an explanation for European
unemployment.
Nevertheless,
the evidence
in
Europe
and Pischke, 1998).
(see, for
difficult to
generate large responses to any of these shocks. For example,
hard to reconcile with skill-biased technical change being the driving force of
is
unemployment
it is
(e.g.,
Nickell
and
Bell, 1996;
Card, Kramartz and Lemieux, 1998; Kruger
Similarly, the effect of increased international trade
example, Krugman, 1995; Lawrence and Slaughter, 1994).
seems
Macroeconomic shocks, on
the other hand, do not appear to be large enough to lead to such a persistent
disinflation during the early 1980's could
relatively small
have increased unemployment,
it
is
effect.
Although
only an extreme
degree of hysteresis that can sustain very high levels of unemployment for over twenty years.
Similarly, in
most standard models, a permanent slowdown
in
TFP
or a
permanent increase
in
the interest rate only have a small effect on the unemployment rate.
This assessment of the state of literature on unemployment suggests that we need new ways
of looking at the
unemployment problem. This paper makes a preliminary attempt
an alternative hypothesis.
1
I
develop the thesis that differences
at developing
in the credit markets
may have
'Previously, Krueger and Pischke (1998) argued that differences in product market competition could be im-
portant
in
understanding European unemployment.
A
recent paper by
Wasmer and Weil
(2000) provides a
played an important role in the increase in European unemployment.
simply as possible,
I
will abstract
from labor market
unemployment
differences, the
model presented here
and labor market
joint behavior of access to credit
Credit markets differ in
this idea as
and focus only on differences
rigidities,
importance of credit market problems
In addition to illustrating the
credit markets.
To develop
in
in
causing
a simple starting point for analyzing the
is
prices.
many dimensions between
the U.S. and Europe. In the U.S., stock
venture capital finance, and funding of small businesses appear more important
market
activity,
than
Europe. In contrast, in Europe, lending by large banks appear to play a more prominent
role.
in
in
Most measures
Europe
in fact
Rajan and
(e.g.
show that
markets are more developed in the U.S. than
credit
Zingales, 1998).
will
I
make
complexities, and presume that the U.S. credit market
new
The main argument
firms.
paper
in the
markets respond to new opportunities
performances
in these
Even though,
librium in an
have large
will get
it
as
I
more
many
flexible in providing loans to
way
that this difference in the
in
which credit
an important contributor to the different employment
economies.
show below, coordination
economy with imperfect
effects
is
is
is
a heroic abstraction from the
failures
markets
credit markets, differences in credit
on steady state unemployment
eventually. However, in the
can lead to a high unemployment equi-
medium
rates,
not
will often
who need credit
money to the correct
because entrepreneurs
run, the failure to channel
To capture these
entrepreneurs can have a large effect on unemployment.
issues,
I
think of
the U.S. and Europe during the 1950's and 1960's as characterized by the steady state of two
economies, one with better credit markets than the other.
be low to start with.
new
the arrival of a
markets
will
I
will
both economies
common
will
shock:
The main result is that the economy with better credit
new technologies without an increase in unemployment. In
set of technologies.
contrast, in the
economy with worse
in technologies
can have a persistent adverse
markets, the agents
credit
markets (the model equivalent of Europe), the change
who need
effect
on unemployment because,
to have the cash to start
borrow the necessary funds. They have to accumulate
rely
in
then consider the response of these two economies to a
respond to the arrival of
efficient credit
Unemployment
this cash
in the
absence of
up new businesses cannot
through their own savings
(or
on more expensive alternative sources of finance, such as borrowing from their family), so the
economy goes through an extended period
entrepreneurs
unemployment
may be made even more
will
of depressed job creation.
difficult
by the
be high and wages low, so earnings
of self-reinforcing factors slow
down
Accumulation by potential
fact that in a depressed labor market,
for
the transition of the
many
individuals will be low. So a set
economy
to low
unemployment
for
an
extended period.
The
idea that there has been a change in technology (or the pattern of comparative advantage),
and that the U.S. has adapted to
this
change faster than Europe
think that the engines of growth of the U.S. economy today
is
is
Many commentators
companies like Ford or GM.
plausible.
not
simple model combining labor and credit market imperfections, and shows that credit market imperfections tend
unemployment. Another recent paper by Blanchard and Giavazzi (2000) emphasizes the interaction
between product markets and labor markets, while Bertrand and Kramartz (2000) investigate the impact of the
zoning restrictions in France on employment creation.
to increase
but new firms
like
Microsoft and Intel, which require different
of financial arrangements.
importance of these factors
in explaining
OECD
and look
(1998),
at
medium and low credit-dependent
This result
higher in the U.S., suggesting that differences in
constraining employment creation in Europe.
relevant
is
is
The data I have do not enable an
find
I
Rajan and Zingales
no evidence
for a
office
major
most credit-dependent industries
in the
credit,
may be
markets
Furthermore,
new
firms within
investigation of whether there has been less
is
playing some role in
might be argued that what
it
Europe than the U.S. over the past 25 years. This
in
manufacturing
classify
categories following
not growth of specific industries, but job creation by
by new firms
I
I
not encouraging for the credit market story.
employment
find that the fraction of
I
countries.
Europe since 1970.
in
growth across these sectors.
Nevertheless,
a preliminary attempt
whether the most credit-dependent industries, such as electronics and
and computer equipment, grew slower
differential
different types
necessary to assess the
is
As
European unemployment.
look at the evolution of sectoral employment across
industries into high,
and perhaps
skills,
more empirical evidence
Nevertheless,
is
more
industries.
all
employment creation
an interesting research area
is
for future work.
The Model
2
Consider an economy that consists of a mass
by an
Each agent can become
offspring.
Production takes place
type of agent
type
«j.(,
i
in
of agents
1
t.
Then,
if
Agents
agent
one period, and are replaced
Let
differ in their skills.
becomes an entrepreneur and
i
an
be the
hires worker j with
they produce
3a ht + a jit
This form of the production function implies that
matter
live for
an entrepreneur, a worker, or remain unemployed.
worker-entrepreneur pairs.
of generation
who
for productivity,
distribution of
a
in
skills
but that of the entrepreneur
the population
is
given by
G
(1)
.
is
more important."
(a) with lower support
t
and the worker
of both the entrepreneur
I
assume that the
a mm and upper support
max
ft
All agents have preferences given
by
(l-sf-s'Clj'BZt-eit,
C
is
consumption,
B
is
function
is
convenient as
it
implies a constant saving rate,
where
utility function
is
bequest
linear, in particular
left
U
(y)
to their offspring,
=
y
—
e,
where y
and
s.
is
e
It
is
cost of effort.
This
utility
also implies that the indirect
income, so that
all
agents are risk
neutral.
The
simplify
fact that the productivity of the entrepreneur
some
of the expressions below.
is
three times
more important than that
of the
worker
is
to
I
also
assume that the cost of
effort is
{j
w
given as follows
if
if
worker and exert
7E
where j w <
*y
E For
.
A7 =
skill level
agents do not change across generations,
The
assume that
if
—
1
In contrast,
q.
contract of a worker of type
efficiency
I
effort choice of
common
is
an individual
utility to shirking (not
=
Oij-i for
a has
effort,
is
qw
knowledge.
all i.
effort
output
effort,
to encourage
(a)
through monetary incentives.
him
and
be high with probability
will
1.
to exert effort as in the standard
possible, so
will get a zero
whereas the
q,
w (a) >
wage.
Since
0.
all
Therefore, the
utility to exerting effort is
w (a) — e.
=
(l
-q)~
1
-
'
t
W
.
always in the interest of the entrepreneurs to encourage high
effort.
an incentive compatibility constraint (the equivalent of the no-shirking condition
Stiglitz,
of
the wage, w(a), needs to be greater than
v
Shapiro and
his
not observed by others introduces a
is
who produces no output
exerting effort)
This implies that to encourage
it is
is
and assume that the type
this here,
assume that negative wages are not
agents are risk neutral, a worker
is
a^t
i.e.,
he exerts
if
wage model.
assume that
of a worker
Workers need to be induced to exert
The wage
there
The
.
the worker does not exert effort, output will be high with probability
with probability
I
7 s~ 7w/
workers need to exert costly effort that
fact that
moral hazard problem.
I
entrepreneur
adopt a particularly simple form of
I
effort
important that the ability to perform entrepreneurial tasks are correlated
it is
across generations.
if
define
I
In contrast, the
private information.
For the analysis,
future use,
effort
worker and exert no
1984, model) which requires that wages for
all
Hence,
in the
possible types of workers
satisfy
w (a) >
Finally, to
The
to
(IC)
v.
become an entrepreneur, an individual needs an investment
exact timing of events
become an entrepreneur
is
as follows.
or a worker.
If
of 2 A'.
At the beginning of the period, each agent decides
there
is
necessary funds, otherwise they have to use their
a credit market, entrepreneurs borrow the
own
wealth.
Entrepreneurs hire workers,
production takes place, and they pay back the loans. Finally, consumption and bequest decisions
are made.
(see, e.g.,
The
is
Walrasian except for the presence of the moral hazard constraint
Acemoglu and Newman,
Since there
at the
labor market
is
1997).
no discounting within a period, an entrepreneur who borrows an amount
beginning of the period has to pay back
2 A".
So the
employing a worker of type a' as a function of the wage,
UE
(«, «')
= 3a +
a'
- w
(0')
utility of
w (a),
-
2
A
IK
an entrepreneur of type a
is
- 1e
,
(2)
+a
where 3a
capital,
the revenue that this employment relationship generates,
1
is
and j E
Since the labor market
the effort cost.
is
entrepreneur has to obtain the same
= Ue
Ue{oc,q')
(a, a"), for all a'
wage contract has
and a" that are workers.
be unemployed.
worker for an agent with a'
>
(3)
The
market.
<a=
incentive compatibility con-
+
a'
<
with
v, i.e.,
— w*,
v
(4)
v
employed workers exert
all
— w*
eqs. (2)
and
(5),
the utility to becoming a
effort, so
is
UW
= w* +
(a')
we obtain that
a*
=
w*
a'
- lw
(5)
individuals with
all
skill level
greater than
+ K + Ao
become entrepreneurs. Whether they can do
(6)
so or not will
depend on the
availability
3
Equilibrium Without Credit
3
The equilibrium with
credit
market
frictions
is
want to become an entrepreneur can obtain
consistent with
am
+ a',
w*
to be paid to solve the moral hazard problem.
Along the equilibrium path,
of credit.
Therefore,
This implies that the equilibrium
Intuitively, their contribution to the revenues of the firm falls short of the
wage that they have
like to
=
(a/)
will adjust to clear the
a'
would
competitive, in equilibrium an
is
imposes a wage floor and implies that workers with w*
Combining
the cost of
to take the simple linear form
where the constant term, w*
will
is
from hiring two different workers.
level of profits
w
straint, (IC),
IK
full
employment, w m
(a)
relatively easy to characterize since all agents
a-
Let
finance.
= w m + a,
be the median of the distribution of
Market Frictions
(i.e.,
in the
me
define the
who
wage function that
absence of credit market frictions.
G (a m =
)
1/2).
For
full
agents with
indifferent
a
greater than
full
my
focus
is
is
m = am -
K
-
w (am = w m + a m
)
become entrepreneurs. I assume
So workers with low ability will be unemployed
that
.
satisfy
A-),
ability distribution,
also possible that workers with ability less than a,
constraint, might want to
out.
the equilibrium wage,
on the interaction between credit market frictions and unemployment,
impose that the lower support of the
'It
for
employment equilibrium wage function must
iu
Since
all
should become entrepreneurs. Hence, the median agent should be
between entrepreneurship and working
Therefore, the
Let
employment, half of
the agents need to become entrepreneurs. Without credit market frictions, this implies that
am
is
Qm
i
n
,
is
less
than
wm
I
:
who cannot become workers because of the wage
3a - u>* - IK - -y E < 0, which rules this possibility
Assumption
am
1:
which implies that
Equilibrium
is
\
n
full
now
< wm
,
employment
will
not be an equilibrium.
a
characterized by a cutoff level of ability
satisfying equation (4) to
become
a worker, a cutoff level of ability a* for entrepreneurship that satisfies equation (6), the
This market clearing condition requires that the
function (3), and a market clearing condition.
number
of entrepreneurs
equal to the
is
number
of workers, or
1-G(q*) = G(a*)-G(a),
where
1
— G
(a*)
is
number
the
become entrepreneurs, and
G (a*)
G (a)
be entrepreneurs, but have a sufficient
equals
in
demand
w* demand
,
which
The
condition.
falls as
number
the
is
left-hand side,
demand
the
is
equilibrium wage function,
i.e.,
w*
Figure
Once w*
.
who
for labor,
The right-hand
the price of labor increases.
to*.
of agents
is
a simple supply
and since a*
side
is
is
increasing
the supply of labor,
shows the determination of the unique
1
determined, the unemployment level
is
who
are not skilled enough to
to be employed. Equation (7)
skill level
a strictly increasing function of
is
(7)
of agents with ability greater than the cutoff level, a*,
—
wage
is
given by
G(v-w*).
Wage
intercept
.G(a*)-G(a)
/
/
/
/
/
/
X
-
\
/
f—
/
\
\
/
*-
/
/
V
X
N
*
^
%
"--^
,''
l-G(a')
Supply and demand
Figure
1:
The determination
of equilibrium
wage
level
to an increase in
Proposition
1
by equations
(4),
There
and
(6),
become entrepreneurs,
w
(a
1
)
= w* +
a'
and
The comparative
Figure
1.
exists a
all
all
(7).
A7
w*, and comparative statics in response
or
K.
unique steady state equilibrium characterized by [a,a*,w*) given
In this equilibrium,
agents with
agents with
skill level a'
skill level
€
a < a
all
agents with
[a, a*}
K
or
A7 make
greater than o*
become workers and
receive a
wage
are unemployed.
statics of this equilibrium are straightforward,
Increases in
skill level
and can be obtained using
entrepreneurship less attractive, shift both the supply
6
and demand curves down to the broken curves, and lead to a lower equilibrium wage intercept
As a
w**.
result,
both a* and a increase, so there are fewer entrepreneurs, and more unemployed
workers.
It is
useful to characterize the steady state wealth distribution, even
bution does not
affect the
a
who want
will leave bequest of
to
become an entrepreneur
=
Bf+l (a)
3a —
+
(Bf (a)
s
bequest of Bf (a) from his parent, invested 2A", paid a wage of w*
+
a'.
Bf
(a)
and produced 3a
In steady state,
He consumes
= Bf+1
a fraction
—
1
steady-state wealth distribution for workers,
In particular,
Bf+l (a)
(a
1
Finally, those workers with ability
be
level will
u
B =
in the
a < a
will
—
+ a'
worker of
for a
and leaves a fraction
s as
bequest.
a €
those with ability
[a, a*], is
given
So
+ w*)
-s
be unemployed, therefore their steady state wealth
Equilibrium with Credit
economy without
market
credit
Market Frictions
frictions, equilibrium prices
economy with
In contrast, in an
block-recursive).
depends on the wealth
investment IK.
ability a',
have wealth
will
determine the
bution of wealth, but are themselves not influenced by the distribution of wealth
is
Therefore,
2A") since he received a
0.
4
Note that
i.e.
= s (Bf (a) + a + w*).
s
hire a worker.
xv*
-s
;
1
The
s of this,
an entrepreneur of type a
(a), so
v
similarly.
distri-
equilibrium allocation in this economy without credit market frictions.
First, notice that all agents
entrepreneur
though the wealth
level of individuals
with high
market
credit
skill,
who need
(i.e.,
distri-
the system
frictions, the equilibrium
to undertake the up-front
This feedback raises the possibility of multiple steady state equilibria and richer
dynamics.
Here
I
assume an extreme form of
individual could runaway with the
this case,
we need
distribution.
Let
to
me
credit
denote the steady-state wealth distribution by
p(a)dG(a)=
F (2K
i.e.
wc
|
a)
is
is
a who have wealth
I
(1
level less
the fraction of agents with
who have
skill level
than
economy with
no borrowing
(e.g.,
an
risk of being caught).
In
is
F (b
\
a),
i.e.
this is the
b.
credit
market
frictions
is
- p (a))dG (a) + G (K + A 7 + w c ) - G [v - w c )
the constant in the equilibrium wage function
have wealth greater than 2A".
agents
there
money he has borrowed without any
Then, the market clearing condition in this
where
frictions:
determine the equilibrium allocation jointly with the steady-state wealth
fraction of workers with ability
/
market
skill level
w c (a') = w c +
a who can
In the expression (8), J
K+ &
greater than A"
+ A7 +
iv
c
afford to
+w c P
a',
and p(a)
(8)
,
=
1
-
become an entrepreneur,
(«)
dG (a)
is
the fraction of
and have a wealth greater than
2 A'.
So
—
become entrepreneurs. The remainder become workers except
these agents like and can afford to
those with
skill level less
Assumption
2:
than v
^±fl
— w c who
cannot be employed profitably.
> IK.
This assumption ensures that in an allocation corresponding to the equilibrium of the economy
with no credit market
eventually
worker with
frictions, a
become an entrepreneur. So
skill level
neighborhood of the equilibrium of Proposition
in the
without credit market problems), those agents
(i.e.
can accumulate this money by saving.
a* can accumulate enough wealth to
who need money
to
1
become entrepreneurs
This implies
exists a steady state equilibrium identical to that described in Proposition
Proposition 2 There
1.
economy without
In contrast to the
credit
market
frictions, the steady-state equilibrium
no longer unique: there can now exist other steady state equilibria.
is
an example to
Consider the following situation in which only a fraction A of agents with
illustrate this point.
skill level
I
will give
greater than
K + A7 + w
c
have enough wealth to become an entrepreneur.
Then, the
equilibrium condition can be written as
2A
dG (a) + G
/
.lh'
where
w
the constant in the wage function.
b —
—
— max
is
and moreover as A
skills,
a max
,
is
»
0.
w
>
not very high
the implied wage level
w
a
v
(i.e.,
.
a max
A
- wb) =
lower level of A implies a lower equilibrium wage,
< (1 — s) 2K/s), we can
r—
< 2K. Then no
'-
enough wealth to become an entrepreneur, and the economy
entrepreneurs (equal to a fraction A
market
frictions, there
JK A +wb dG(a)
is
is
additional agent can accumulate
of the population). Therefore, with credit
can exist other steady state equilibria with higher unemployment and
state equilibrium exists because
is
always find a level of A such that
stuck with only a small fraction of
lower wages than the equilibrium characterized in Proposition
wealth, there
1.
'
Hence, as long as the upper support of the distribution of
—
satisfies -*
(v
^
+ A"/+w b
a limited
demand
when only
for labor,
2.
Intuitively,
another steady-
a few of the potential entrepreneurs have enough
and
this depresses wages.
With depressed wages,
it
harder for potential entrepreneurs to accumulate enough wealth to finance their investments,
and
in fact, as the
above example demonstrates, they
may
never be able to do so, leading to
another steady state equilibrium with greater unemployment and lower wages." This intuition
1
very similar to the reasoning for multiplicity of steady states in the Banerjee and
model
of occupational choice.
Newman
is
(1993)
There, too, different wealth distributions can be self-sustaining
because the equilibrium wage rate depends on the distribution of wealth.
In terms of Figure 1, this corresponds to the demand for labor being an upward sloping curve because when
wages are low, potential entrepreneurs cannot accumulate the necessary funds to hire workers. As a result, there
can be multiple intersections of supply and demand.
Change
5
I
now
discuss
technology.
how economies with and without
in
specific form: a fraction
«2-
I
have ability
<f>
1
cti,
—
2<p
—^
skills,
of agents have a distribution
and another fraction
have
</>
F (a)
respond to a change in
modifying the pattern of
G (.)
takes the following
a4
with upper support
03, and the fraction k have
skill level
,
a
skill
assume that
Two
>
(%2
>
ct3
In particular,
2).
a\ and
>
C14,
>
and 1/2 > k
2c/>.
economies, one with and the other without credit market frictions, are in the minimum-
unemployment steady-state equilibrium
0.2
indifferent
I
equivalent to that characterized by Propositions
(i.e.,
assume that a\
also
become entrepreneurs.
will
>
0:2
> a* >
>
Q: 3
£*;
skill level Q'2
has to be
So the equilibrium wage function,
wage.
for a
is
w=
this, I also
a = v — w,
obtain that
so
a.2
—K—
A7.
(9)
unemployment
is
equal to
U = {1-2<P-k)F{v-Q).
Furthermore,
Assumption
1
so the agents with skill level
This implies that an agent with
between entrepreneurship and working
w (a) = w + a,
Using
frictions
In particular, suppose that
entrepreneurship.
a\
and
market
credit
consider a very simple shift in the distribution of
I
comparative advantage
fraction
Technology
in
3:
I
(10)
assume that
Bw
(a 3 )
=
^T
1
< 2K < B ™ n'i) =
(
s
~^^1
This assumption ensures that the steady state wealth
-
level of
a worker with
03
skill level
is
not enough to afford the up-front investment for entrepreneurship, while that of a worker with
skill level
Now
a\
is
sufficient for entrepreneurship.
suppose that there
agents with a\ and
0:3
economy without any
is
an unanticipated
are switched:
credit
market
not be any macroeconomic changes
in the
economy with
agents
who
credit
market
i.e.,
those
shift in the distribution of skills at
who had
frictions will
is still
previously had productivity 03, and
C13
and vice
given by Proposition
now have productivity
first,
to
become entrepreneurs. The evolution of unemployment
now depends on
now have
and
versa.
1.
5
in the
the transitory dynamics of the model.
all
The
will
In contrast,
Assumption 3 implies that unemployment
at
frictions
«i
t*,
immediately adjust to this change, and there
—the equilibrium
frictions
skill
time
increases:
oj, cannot afford
economy with
credit
market
Transitory dynamics are in
general quite complicated, since the equilibrium wage rate changes as the wealth distribution of
agents become less productive. It is conceptually
I am assuming that some
cumbersome to extend the model so that the productivity of all agents increases
and the changing technology can be modeled as some productivities growing more than others over a
'Perhaps this
is
too simple, since
straightforward, but the notationally
over time,
certain time period
I
Nevertheless, the specific assumptions that
agents evolves.
wage
particular, they ensure that the
equal to w. This
and equation
is
rate
(9) still
t*
unemployment
{l-2(j)-
level at
This implies that immediately after the
time
t*
+
K)F(v-w) +
who now have
r can be written as
skill level
At
r*, Bf. +T , (a\) exceeds 2A' the for the first time.
become entrepreneurs, and unemployment
Therefore, while
unemployment
is
aj start accumulating wealth.
so there will exist a threshold time r* such
productivity a\
falls
this point, the agents
Uf
from
in the
economy with
persistently between the date of technology change,
By changing parameters within this
arbitrarily late (e.g., if '^'^^ ~ 2 A"),
credit
simple example we can
so this
model
market
and the date
t*,
In general, with a depressed labor
and
their
own
make
it
businesses.
given by (10).
is
t*
It
it
stays high
.
decline in
unemployment
capable of generating an arbitrarily long
demand, the wage
useful to note that
It is also
a potentially important factor contributing to unemployed persistence
this will
U
frictions.
+
t*
make the
period of unemployment, or an arbitrary amount of persistence.
example.
to
with
always constant in the economy with perfect credit markets,
upon change of technology
increases
(t>> U.
5
> 2 A,
3 implies that lim r ^oo Bf. +T [a\)
+
t*
In
increases to
After this point, the agents (dynasties)
that at
analysis.
independent of the wealth distribution, and always
determines the equilibria wage.
Uv =
Assumption
imposed simplify the
because agents with productivity «2 continue to be the marginal entrepreneurs,
technology change at
Their wealth
is
I
is
missing in this simple
rate will also
fall
during transition,
harder for potential entrepreneurs to accumulate funds sufficient to finance
This
is
in fact exactly the
same
force that led to the multiplicity of steady
state equilibria in the previous section.
6
So
far
I
vs.
have presented an abstract model of unemployment
imperfections.
they
Empirical Evidence: U.S.
may
Europe
in the
presence of credit market
Credit market problems do not necessarily lead to higher unemployment, though
introduce additional equilibria with high unemployment levels.
The major
result
is
that economies with and without credit market frictions respond to changes in technology quite
differently.
In response to a shock that changes
in entrepreneurship,
which agents have a comparative advantage
an economy without a well-functioning credit market
may
suffer a lengthy
period of unemployment, as productive agents lack the necessary funds to create jobs.
situation
is
This
reversed only slowly as these agents gradually accumulate enough wealth to finance
their investments.
It is possible to tell a
story of European
unemployment based on
this
model.
Europe and
the U.S. had similar unemployment levels during the 19G0s because they were both in a "steady
6
To obtain this expression note that
(q 3 + s(w + Qi), etc.
sB w
at
time
£*,
they start with wealth
)
10
Bw
(as)
=
iiSjtSa^
amj
then B^i + 1 (oi)
=
A
state situation".
variety of changes during the early 1970s
and early 80s
and firms to assume a more important
sectors of the economy, requiring different entrepreneurs
This happened relatively quickly the U.S. thanks to a more
role.
with institutions
In contrast, in
like
growth
shifted the
fluid credit
market, especially
venture capitalists channeling funds to the rapidly growing high-tech sectors.
Europe, lack of funds in the hands of the people with the right
slowed
skills
down
job creation.
there any evidence supporting this story?
Is
The second maintains
Although
constraint.
story,
I
can look
With
I
it is
this purpose,
firms in
some
I
follow
some
The
sectoral
first
views
sector.
particularly credit-dependent sectors that are most
do not have a way of investigating the validity of the
Rajan and Zingales (1998)
in the U.S..
at
Europe to be those using new technologies within each
support for the second version using the
for
dependents
credits
that
end the paper with a quick look
There are two interpretations of the model.
evidence to answer this question.
the firms that are constrained in
I
Using the results
erages and tobacco, textile, apparel and leather,
in
Table
OECD
first
version of the
sectoral database.
in classifying sectors
according to their
of these authors,
classify food, bev-
1
wood and
furniture, paper
I
and paper products,
basic metals, and metal products as low credit-dependent industries; chemical, coal, rubber and
plastic products, metal products,
machinery and equipment, and transport equipment
credit-dependent industries, and electrical goods and
are high credit-dependent industries.
By
office
as
medium
and data processing machine industries
focusing on manufacturing, this classification avoids
other potential differences in the U.S. and European economies in the growth of service sectors,
though
service industries are
if
more credit-dependent,
it
might understate the importance of
credit problems.
Figure 2 plots the share of employment in a number of European countries in the low, medium,
and high categories
unemployment
is
relative to the share of the
same categories
fall
7
If
European
fall
behind the U.S.
in
the high
In other words, in this case, the curve with the squares (the high category) should
relative to the other curves, especially relative to the circles,
For most countries,
all
Germany, employment
sectors.
the U.S..
mainly due to the failure of the European economies to expand into new sectors
because of credit constraints, we would expect these economies to
credit sectors.
in
three curves
in
Therefore, there
move
in a
more
which denote the low industries. 8
or less parallel fashion.
In fact, for France and
high credit-dependent sectors seem to be growing more than another
is
no evidence that European economies have been
falling
behind the
U.S. in the credit-dependent industries over the period of rising European unemployment.
are from the OECD ISDB data set, and refer to the total number of employees in that sector relative
employment. More formally, let e c be total employment in the low group in country c at time t, and Ect
be total employment- I am plotting (e c ,/E c t) / (e v st / En st) Data for the high category in Sweden and Finland
1
The data
to total
<
are missing.
An
in
increase in
all
West Germany
employment compared to the
three curves, as in the case of
the relative share of manufacturing
11
or
Denmark, would correspond
U.S..
to
an increase
Aempl
medium
In
ind rel to
Aempl
US
--
I
2000
1970
medium
in
ind
rel to
A empl
US
in
mednjm md t?
lo
U
...,
1990
980
7
year
Belgium
Bempl
empl
In
tow ind
in
high ind
Aempl
rel !o U.S.
rel lo U.S.
medium
in
ind
ret lo
US
Bempl
empl
in
low ind
in
high ind
rel to U.S.
rel to U.S.
Aempl
in
medium
ind ret lo
US
Aem.pt m
1980
1970
year
year
Denmark
Western Germany
Italy
Aemp!
in
medium
ind
rel lo
US
I
empl
empl
in
in
tow ind
rel to U.S.
rel to U.S.
Aempl
in
medium
ind
rel lo
Aempl
U S
high ind
1980
year
UK
Finland
Figure
2:
The
evolution of the fraction of total
employment
in low,
medium, and high
credit-dependent industries in European countries relative to the U.S.. Data from
data
employment
in the
OECD ISDB
set.
Nevertheless, except for the Netherlands and the
substantially less
UK, European economies appear
most credit-dependent
gium, Italy and Denmark, relative employment
industries.
in the least
to have
For example, in Bel-
credit-dependent industries
is
on
average about 50 percent more than in the U.S., while the relative employment in the most
credit-dependent sectors
is
about 30 percent
less
than the U.S..
This suggests that credit market
problems may have played some role in constraining employment
UK
commonly thought
among
in
Europe. Interestingly, the
the European economies, so
it is
suggestive that the most credit-dependent industries have relatively high employment in the
UK
is
to have the best credit markets
as well as in the U.S. (though their share in the
1980s).
These observations give some support
been an important factor
in
UK
employment seems
for the idea that credit
to have fallen over the
market problems may have
the recent European employment experience, and encourage further
research in this area.
12
me
in
medium
<nd
re< to 11
i
References
Acemoglu, Daron and Andrew
ture
"
CEPR
Newman
(1997) "The Labor Market and
Struc-
working paper.
NAIRU" NBER
Ball Lawrence (1997) "Disinflation and the
Banerjee, A. and A.
of Political
The Corporate
Economy,
Bean Charles
Newman
WP
#5520.
'Occupational Choice and the Process of Development.' Journal
101, 274-298 (1993).
(1994);
"European Unemployment a Survey" Journal of Economic Literature
Vol 32, pp 573-619.
Bertrand, Marianne and Francis Kramartz (2000), paper presented in the
NBER labor studies.
Blanchard Olivier and Justin Wolfers (1999) "Shocks vs Institutions" forthcoming Economic
Journal.
Blanchard Olivier and Francesco Giavazzi (2000) "The Macroeconomic Effects of Regulation
and Deregulation In Goods and Labor Markets"
MIT
mimeo.
Card David, Francis Kramartz and Thomas Lemieux
ture of
Wages and Employment:
A
Comparison
(1995);
"Changes
of the United States,
in the Relative Struc-
Canada and France" Mimeo.
Krueger, Alan B. and Jorn-Steffen Pischke (1997) "Observations and conjectures on the U.S.
employment
miracle,"
NBER
Working Paper No.
6146,.
Krugman, Paul (1994) "Part and Prospective Causes
of High
Unemployment"
in Federal
Reserve Bank of Kansas City Reducing Unemployment: Current Issues and Policy Options.
Krugman, Paul, "Growing World Trade: Causes and Consequences" Brookings Papers
Economic
pp 327-377.
Lawrence, Robert and Matthew Slaughter "Trade and U.S. Wages: Giant Sucking Sound
Activity
on.
(1995).
or
Small Hiccup?" Brookings Papers on Economic Activity, (1993) pp. 161-226.
Layard, Richard, Steve Nickell and Richard Jackman (1991) Unemployment: Macroeconomic
Performance and the Labor Market Oxford University Press.
Legros, Patrick and
Andrew
of Organization," Journal of
Nickell, Steven (1997),
F.
Newman
(1996); "Wealth Effects, Distribution
Economic Theory.
"Unemployment and Labor Market
America," Joum,al of Economic Perspectives.
Nickell Stephen
and Brian
employment Across the
OECD
Bell (1996)
OECD"
(1994) Jobs Study
and the Theory
Rigidities:
11, 55-74.
"The Collapse
in
Demand
For the Unskilled and Un-
Oxford Review of Economic Policy Vol
Volumes
1
and
Europe versus North
11,
pp
40-62.
2.
Rajan Raghuram and Luigi Zingales (1998) "Financial Dependence and Growth" American
Economic Review,
88, 559-586.
Wasmer, Etienne and Philippe Weil (2000) "The Macroeconomics Of Credit and Labor Market
Imperfections"
UBL
Wood, Adrian
Working Paper.
(1994), North-South Trade,
Employment and
a Skill Driven World, Oxford, Clarendon Press.
13
Inequality:
Changing Fortunes in
5236 Okk
MIT LIBRARIES
3 9080 01917 6046
;:;.:
I
1
illll
m
Wmm
tiipifcpfl:
I II lllil
illli
mil %
'
:
'
.
<-H
'»
.
(;
m$i
:
IIISIffiitSW'''
WmMmf
MM
fflfii
«i
II
'
B;^^ Will
1
iiiiii
:.. ...
.",.....'..
m
Mffiiiiiiiiii
H
immsm
ii»
llSi
I9H1
H