SOLUTION PROBLEM SET 3

LABOR ECONOMICS
Question 1:
Answers should recognize that this result does not hold when there are search frictions in the labour
market. The proof should follow a simple matching model in which the worker bargaining power parameter
is set at zero, and the firm appropriates the whole rent from job matches.
In this set-up equilibrium unemployment is given by
u=

+ q ()

(1)

Hence, the only way to eliminate unemployment when > 0 is to have , which makes it
infinitely easy to find a job. Labour market tightness and wages are jointly determined by the following
job creation and wage setting conditions:

c
(2)
w = P (r + )
q ()
w = P + c + (1 )z
(3)
= z when = 0
Therefore, the equilibrium tightness parameter is such that

r+
c
q () =
P z
Hence is finite and, although very low, unemployment is not eliminated.
The opposite case, where = 1 implies that
w = P + c > P
Then the Job Creation condition reads
c = (r + )

c
q ()

However, it can never be satisfied since the term on the left-hand side is positive and the one on the
right-hand side is negative. The intuition is quite simple: since the worker appropriates all the rent, firms
cannot recover their searching costs. Consequently no firm will enter the market and we have full
unemployment. In other words, the equilibrium conditions for an interior solution cannot be satisfied
when = 1.
Question 2:
The reservation wage of a worker facing a non-degenerate wage oer distribution is given by

Z +

(w w ) dF (w)
w =z+
r+
w

(4)

where is the rate at which workers contact firms. The term z is the flow income of unemployed workers.
It aggregates unemployment income, alternative revenues and searching costs.
Hence, the first obvious answer to the question is that governments do not have full control over the
worker outside option and so cannot directly determine the reservation wage.
1

The second and more interesting point is that, even when z is equal to zero, workers might turn
down some job oers (as seen in PS.5). The rationale behind such a behavior is that, by accepting a
given job oer, the worker abandons the possibility to search for better opportunities. This is why he
will sometimes refuse job oers although they provides him with a higher income than what he gets
while being unemployed. In other words, an acceptable job oer must also compensate the worker for
abandoning his option to search.
The discussion of the provisions of reasonableness could led to dierent answers. It is related to the
first part of the question, since we have seen that unemployment income do not allow direct control over
the reservation wage. This is why governments have turned to dierent policies in order to modify the
behavior of job seekers. By excluding workers from the services provided to job seekers, they can reduce
the rate of contact and to some extent stigmatize the unemployed workers for not complying with the
rule. Therefore, the provisions of reasonableness is likely to increase the rate at which workers leave
the unemployment pool. However, such a policy also has important drawbacks. The principal one is
that workers choose the reservation wage which maximizes their utility. By interfering, the government is
likely to generate ineciencies from the workers point of view. It is also dicult to imagine that workers
will be motivated for jobs that they would have refused in the first place. Of course, many other answers
are possible and were given points.
Question 3:
The common mistake was to discuss shocks that increase whereas the question was about shocks
that shift the Beveridge curve. The expression of the Beveridge curve is derived from
u=

+ q ()

(5)

1. The first type of shocks directly aect the matching function q (). Such shocks reflect an increase
in the degree of mismatch in the economy. They might be due, for example, to a higher proportion of
long term unemployed workers whose skills have become obsolete. Consider for example a Cobb-Douglas
matching function, so that
q () = A with (0, 1)
A decrease in the scale factor A leads to an increase in unemployment for a given value of . Since
the matching function does not enter the JC and WC, the equilibrium value of and w do not change.
Hence the new equilibrium can be illustrated considering solely the (u, v) plane.

U*

2. The second type of shocks increase the separation rate . They might be due, for example, to a
reduction in employment protection. A change in also aects the JC curve, so the new equilibrium
looks as follows.
V

W*

U*

Exercise:
1. This question asks to redo the calculations explained during the lectures. We briefly repeat them
here.
Job Creation Curve: The asset equations for Vacancies and Filled Jobs are
rV
rJ

= C + q () (J V )
= P w + (V J)

Free-Entry implies that in steady-state V = 0, therefore

J=

P w
C
=
r+
q ()

(JC)

Wage Curve: The asset equations for Unemployed and Employed workers are
rU
rE

= Z + q () (E U )
= w + (U E)

The second asset equation can be rewritten as follows

w = (r + )(E U ) + rU

(6)

To obtain the value of (E U ) in terms of the models parameters, we use the Nash-bargaining
solution

P w
C
(1 )(E U ) = (J V ) =
=
r+
q ()
where the last equality follows from the Free-Entry condition.
Now replace into the asset equation for U the value of (E U ) to obtain

rU = Z + q () (E U ) = Z +
C
1
Finally reinsert this solution into (6) to obtain
w

= (r + )(E U ) rU

=
(P w) + Z +
C
1
1
3

Simplifying this expression yields

w = P + C + (1 ) Z

(W)

2. If the minimum wage is below the market wage, then the regulation is irrelevant. Thus the lower
bound is w = w = P + C + (1 ) Z.
If the minimum wage is above the jobs output, firms have no incentives to post vacancies since they
cannot recover their searching costs. Thus the upper bound is w = P.
For any minimum wage between these two bounds, the equilibrium is well defined and diers from
the free-market equilibrium. Since the wage is fixed exogenously, we do not need to consider the wage
curve. The equilibrium tightness is given by the JC condition, so that
P wm = (r + )

C
q ()

When wm is above the market wage, it must be the case that q () increases. Since q 0 () < 0, this
implies that decreases. In other words, there is less vacancies per unemployed worker.
To obtain the impact on the equilibrium rate of unemployment, we use the Beveridge curve
u=

+ q ()

We know that decreases and since (q())

> 0, we can deduce that the unemployment rate increases.

3. The hazard rate of leaving unemployment is equal to q (). Thus the aggregate flow of workers is
given by
q ()
u (q ()) =
+ q ()
Dierentiating this expression with respect to yields

(q())
(q())

(
+
q
())

q
()

(uq ())
=
2

( + q ())

(q())
2

=
>0
( + q ())2
Since is a decreasing function of the minimum wage, the size of the flows out of the unemployment
pool also decreases. In other words, the unemployment rate increases proportionally less than the hazard
rate decreases.
4. We replace the new expressions in the WC to obtain
w

= P + cP + (1 ) zP
= P ( + c + (1 ) z)

Therefore the wage is given by a constant factor times the jobs output.
5. Replace the new expression into the JC condition
P w
r+

P (1 + c + (1 ) z)
cP
=
r+
q()
1 + c + (1 ) z
c
=
r+
q()

The equilibrium tightness and so unemployment are independent of the jobs output. This is because
the wage and the cost of posting a vacancy increase proportionally to output.
4

6. P is a measure of jobs productivity. Thus one can think of changes in P as measuring technological
progress.
If the minimum wage grows at a slower rate than technological progress, it is obvious that after
a sucient length of time w (t) = P (t) ( + c + (1 ) z) > wm (t) so that the market wage will
eventually exceeds the minimum wage.
On the contrary, if the minimum wage grows at a faster rate than technological progress, it will
always be true that limt wm (t) > limt P (t). Then the equilibrium will collapse and there will be
full-unemployment.
To conclude, for the minimum wage to be consistent with steady state and growth, it must increase
at the same rate than technological progress.