LABOR ECONOMICS
LABOR ECONOMICS
Sponsored by a Grant TÁMOP-4.1.2-08/2/A/KMR-2009-0041 Course Material Developed by Department of Economics,
Faculty of Social Sciences, Eötvös Loránd University Budapest (ELTE) Department of Economics, Eötvös Loránd University Budapest
Institute of Economics, Hungarian Academy of Sciences
Balassi Kiadó, Budapest
LABOR ECONOMICS
Author: János Köllő
Supervised by: János Köllő January 2011
ELTE Faculty of Social Sciences, Department of Economics
LABOR ECONOMICS
Week 11
Equilibrium in a competitive labor market
János Köllő
• Marshallian equilibrium
• Delays of adjustment: cobweb cycle
• Frictions: job search and unemployment
• Frictions: equilibrium with job destruction and job creation
• Supply or demand?
Equlibrium in a perfectly competitive
labor market with no frictions
Excess demand
Shortages tend to push wages upward.
Demand falls and supply rises in
response to growing wages.
E is the only point, where the number of people willing to
work and the number of jobs offered are equal.
D,S
w S
D
Labor shortage
E
Marshallian equilibrium
Excess supply
Wages fall under the pressure of
unemployment
Demand rises and supply falls in response to falling wages
E is the only point, where the number of people willing to
work and the number of jobs offered are equal
D,S
w S
D
unemployment
E
Marshallian equilibrium
Cobweb cycle
Delays in adjustment
Cobweb cycle: conditions
Closed occupational labor market
Entry only from the school system. Skills are not convertible.
Naive expectations
Expectations concerning the future returns to education are based on current returns.
Cost-based planning in education
Quotas are determined on the basis of current social costs and returns.
Labor demand is more elastic than labor supply.
D0
D1 w
S
Let the demand for an occupation rise, shifting the
demand curve from D
0to D
1.
The market can not move immediately to the new equilibrium point because labor supply reacts with a delay.
D0
D1 w
S
?
D0
D1 w
D,S A
B
Step 1
Wages and enrolment rise but it takes time until students
graduate and enter the labor market.
Step 2
Students graduate supply rises
D0
D1 w
D,S A
B
Step 3
Growing supply puts a pressure on wages. Increased supply is absorbed only at lower wages.
D0
D1 w
D,S A
B
Step 4
Since wages are now lower, enrolment falls. After a while the number of graduates falls, too.
D0
D1 w
D,S A
B
Step 5
A smaller cohort of students graduate. Falling supply pushes wages upward and leads to an increase in enrolment.
And so on …
D0
D1 w
D,S A
B
D0
D1 w
D,S A
B
… and on
D0
D1 w
D,S A
B
… and on
D0
D1 w
D,S A
B
… and on
D0
D1 w
D,S A
B
Until the market finds the new equilibrium.
Note that if supply is more elastic than demand the process does not
converge.
Factors working against a cobweb cycle
• Skills are convertible to some extent.
Workers can leave and enter the market.
• Adult training can further increase mobility.
• Future changes are predictable to some
extent.
Frictions
Job search
• In practice, markets never reach the equlibrium point.
• Even in tight markets, many workers remain unemployed. In mature market economies
unemployment seldom fell below 3 per cent in the last few decades.
• A part of this problem arises because job
seekers may be interested in searching further rather than accepting the job offers they meet.
• Let us examine how job seekers make their
decisions.
The costs (and benefits) of searching
Shall we search further in a short period t after meeting a wage offer w*?
Consider an unemployed job seeker, who has information (or beliefs) on the distribution of wages but not on actual wage offers.
Unemployed workers can receive benefits, produce goods in the household and has to cover search costs.
b = unemployment benefit
h = value of household production c = direct costs of job search
F(w) = the distribution of wage offers
< 1 = the probability of an offer in period t
r = discount rate measuring risk aversion
The costs of searching further
) ( c b h w
I have to cover the costs of further
search (+)
I can work further in the household
(–)
I continue to receive unemployment benefit
(–) I lose wage w* by refusing
the offer (+)
w* = opportunity cost, c–b–h = direct costs
w r w
w w E w
F 1
( 1 1
Probability that the offer is higher than w*
Expected gain
Risk aversion
Probability of an offer
Recall that Pr(w>w*)=1–F(w*)
The benefit from further search
The asking wage* (w R )
c h
b r w
w w
w w E w
F R R R R
1 ( 1
1
w r w w w E w
F 1
( 1 1
c h b w
c h
b w*
Gains
wR
(gain if w* is accepted)
(gain if w* is rejected)
The wage, at which the job seeker is neutral between accepting and refusing an offer:
It is also called the reservation wage. We follow Borjas (2008) in calling it differently from the concept discussed in Labor supply
If the offer is lower than the asking wage further search Further search implies further (search) unemployment
w r w w w E w
F 1
( 1 1
c h b w
c h b
w*
Gains
wR
(gain if w* is accepted)
(gain if w* is rejected)
If the offer is higher than the asking wage stopping
w r w w w E w
F 1
( 1 1
c h b w
c h b
w*
Gains
wR
(gain if w* is accepted)
(gain if w* is rejected)
The probability that a random offer is accepted
(a is exogeneous):
• Benefits, household production decrease
• Search costs: increase
Lower costs make further search more attractive.
• Risk aversion: increases
• Abundance of job offers: decreases
This relates to accepting a given offer. In the same time, abundance
increases the job offer arrival rate and thereby decreases unemployment.
• Wage distribution: depends
Job seekers often notice changes in the wage distribution with a delay
w
ROffers accepted Density of wage
offers
Job seekers may not immediately notice changes in the wage distribution. Despite f
1f
2their asking wage may remain at w*
• Implications:
• Exit from unemployment falls since Prob(w>wR) decreases.
• Since the condition for entry to employment is still w>wR, accepted wages do not fall.
Wages are seemingly rigid.
Accepted offers
f
1f
2w
RFrictions
With job destruction and job creation
• Markets are in motion: jobs are destroyed
and created, and workers are hired and fired, on a massive scale even when aggregate
employment is constant.
• We can speak of equilibrium (steady state), when the number of exits from and entries to employment are equal.
• Steady states can be achieved at different
levels of unemployment. This is what we try
to understand in the forthcoming section.
The UV curve
Let us keep the rate of job destruction constant. How many vacant jobs should be created in order to keep unemployment (U) constant?
U V
When unemployment is high, it is easier to fill vacancies. Therefore, less vacancies should be created in order to keep unemployment constant.
Along the UV curve s(1–u) = x(u,v), where s is the separation rate and x(u,v) is a matching function describing the number of entries to employment. Unemployment is u while v denotes vacancies.
In other words, along the UV curve the number of new matches keep
unemployment constant for a given rate of job destruction.
Different UV curves
Let us keep the rate of job destruction constant. How many vacant jobs should be created in order to keep unemployment (U) constant?
If matching is inefficient (because workers and firms are poorly informed, mobility is constrained, the demanded and supplied skills differ to a large extent) more V is required at a given level of U.
U
V
The VS curve
VS=vacancy supply
How many jobs are created by firms? How job creation varies with U?
High U falling wages and recruitment costs it is now cheaper to create V.
U
V
Different VS curves
How many jobs are created by firms? How job creation varies with U?
If wages are rigid and recruitment costs do not fall with U (because the
unemployed live in remote areas, are unskilled, etc.) firms create less V at a given level of U.
U
V
Steady state
Firms are interested in creating as many vacancies as required to keep unemployment constant for a given level of job destruction.
U V
UV
VS UV
VS
Right of the steady state
Firms create more vacancies than required for the equality of flows U starts to fall.
u v
UV
VS UV
VS
Left of the steady state
Firms create less vacancies than required for the equality of flows U starts to grow.
u v
UV
VS UV
VS
Different steady states
A: good equilibrium. Flexible wages, low mobility costs, low information costs, demanded and supplied skills are similar.
B: bad equilibrium. Rigid wages, high mobility costs, high information costs, severe skills mismatch.
U V
B
A
Supply or demand?
• How can we ascertain whether market changes are supply or demand driven?
• We shall see that there is no reassuring answer to the question. However, we can arrive at a second best answer by studying how changes of wages and employment
are related to each other.
• The forthcoming slides draw from L. F. Katz–D. H. Autor (1999):
Changes in the Wage Structure and Earnings Inequality, in: Orley
Ashenfelter and David Card, eds., Handbook of Labor Economics
(Amsterdam, North-Holland).
Supply or demand? In practice, we only see how the market moved from one point to another. We do not know if the market moved along a highly elastic demand curve (D*) or the supply and demand curves S,D were shifted by some external shocks.
w
L D*
D
S
The Katz-Murphy (1992) cross-products provide some insight. Let us use the following notation:
Xt – K 1 vector of factor demands in year t.
wt = K 1 vector of wages (factor prices) in year t.
Zt = m 1 vector of demand shifters in year t.
Factor demands depend on prices and the shifters:
) , ( ]
1
[ X
tD w
tZ
tIn terms of differentials:
t Z t
t w t
Z t
w
t
D dw D dZ D dw dX D dZ
dX ]
2 [
where D
wis the matrix of cross price effects (Slutsky matrix)
3 2 1
3 2 2
2 1
2
3 1 2
1 1
1
2 1
2 2 1
2
2 1 1
1
2 1
dz dz dz
z x z
x z
x
z x z
x z
x dw
dw w
x w x
w x w
x dx
dx In case of two factors and three demand shifters, e.g.
Multiplying equation [2] with dw
tfrom the left yields:
0 ) (
] 3
[ dw
t,D
wdw
tdw
t,dX
tD
ZdZ
tIf demand is constant (dZ=0), the cross product of changes in wages and employment will be non-positive:
0 ]
4
[ dw
t,dX
tIn other words, positive cross products can not be reconciled with purely supply driven scenarios. This is less than a full answer but much more than nothing.
Why? Applying Shepard’s lemma we have:
j i j
i
w w
C w
x (w) 2 (w)
.
It is easy to see that the Slutsky matrix is the Hessian of the cost function. If the cost function is concave its Hessian is negative semi-definite. If a matrix is negative semi-definite then
0 0
'Ax x
x .