econstor
A Service of
zbw
Make Your Publications Visible.
Leibniz-Informationszentrum
Wirtschaft
Leibniz Information Centre
for Economics
Holmlund, Bertil; Lang, Harald
Working Paper
Quit Behavior under Imperfect Information:
Searching, Moving, Learning
IUI Working Paper, No. 100
Provided in Cooperation with:
Research Institute of Industrial Economics (IFN), Stockholm
Suggested Citation: Holmlund, Bertil; Lang, Harald (1983) : Quit Behavior under Imperfect
Information: Searching, Moving, Learning, IUI Working Paper, No. 100
This Version is available at:
http://hdl.handle.net/10419/95057
Standard-Nutzungsbedingungen:
Terms of use:
Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen
Zwecken und zum Privatgebrauch gespeichert und kopiert werden.
Documents in EconStor may be saved and copied for your
personal and scholarly purposes.
Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle
Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich
machen, vertreiben oder anderweitig nutzen.
You are not to copy documents for public or commercial
purposes, to exhibit the documents publicly, to make them
publicly available on the internet, or to distribute or otherwise
use the documents in public.
Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen
(insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten,
gelten abweichend von diesen Nutzungsbedingungen die in der dort
genannten Lizenz gewährten Nutzungsrechte.
www.econstor.eu
If the documents have been made available under an Open
Content Licence (especially Creative Commons Licences), you
may exercise further usage rights as specified in the indicated
licence.
A list of Working Papers on
the last pages
No. 100
QUI'T BEBAVIOR UllDER IIIPERFEC'l"
IHPORMA'TIOH: SEARCHING, MOVING,
LEARflING
by
Bertil HolmIund and Harald Lang
This is a preliminary paper. It is
intended
for
pri vate
circulation,
and should not be quoted or referred
to in publications without permission of the authors. Comments are
vvelcome.
September, 1983
QUI"I' BEBAVIOR UllDER DlPERFEC"l" DIPORMA"I'IOI:
SEARCHIIIG. MOVIIIG. LEARNIIIG
by
Bertil Holmlund and Harald Lang
September 1983
-
1
2 -
Int~oction
The
application
issues has
ing
of
search theory to
labor market
yielded important new insights regard-
individual
behavior
under
unemployment.
A
model frequently cited portrays the behavior of an
unemployed worker in an environment characterized
by
(i)
offer
eons tant
search
distribution,
probability, and
costs,
(iii)
an
(ii)
a
known wage
exogenous
job offer
(iv) an infinite time horizon.
It
can be shown that a strategy that maximizes expected life-time income implies the selection of a reservation wage,
constant over time when the hori-
zon is infinite. The worker leaves the unemployment pool when he receives an offer above the
reservation wage and stays with the particular
firm
for
quite a
long timei
in
fact,
the employ-
ment contracts lasts for ever.
The basic search model has,
limitations. It is,
however,
for example,
some obvious
not without prob-
lems to use this model as a vehicle for understanding turnover
in
the
labor market.
ture of job mobility is
A typical
fea-
job-to-job changes without
~1attila
intervenient unemployment.
(1974) has esti-
mated that the major fraction of voluntary jOb
changes in the United States occur without interim
unemployment
qui ts
into
and
the
Holmlund
(1983)
unemployment
pool
indicates
that
are
more
even
rare in the Swedish labor market. An obvious inferene e
from
these observations
is
that
on-the-job-
search appears to be of some empirical importance.
The superior search efficiency sometimes attributed to the unemployment state is not very convincing
given
the
magnitude
intervenient unemployment.
of
job
mobility without
-
3 -
Another aspect of mobility behavior is its repetitive
naturej
of ten
workers
those
who
are
empirical
studies
ties
declining
are
suggested
with
most
job
likely
to
changes
qui t.
are
Various
demonstrate that quit-probabilias
job
explanation
firm-specific
recent
is
capital,
tenure
that
which
non-transferable across
increases.
workers
-
One
accumulate
by definition -
firms.
is
The amount of firm-
specific capital will increase as tenure increases
and
the
worker s
current
I
wage
will
grow in
rela-
tion to his market alternatives i hence, quit probabilities will decrease. (See,
for
example,
Parsons ,
1972.)
It
is
questionable,
however,
if
this
story
can
explain the apparently rapid decline in quit rates
that
occurs
estimates
percent
as
that
in
tenure
separation
the
one year the
increases.
first
months
separation
and 5 percent.
rates
of
Hedberg
(1967)
run between
15-20
employmentj
af ter
frequencies
are between 2
It is unlikely that this pattern of
rapidly declining
(conditionaI )
quit-probabilities
can be attributed to accumulation of firm-specific
human capital.
Several empirical
job
attributes
studies have shown that non-wage
are
important
example,
rates.
(See,
for
Bartel,
1982.)
This,
ic
sense j
improving
workers
job
determinants
Viscusi,
of course,
who
changes
are
1979
and
makes good econom-
interested
will
of qui t
take
in
utili t y-
account
aspects of current and prospective
jobs,
of
all
including
non-pecuniary components.
Many
non-""age
covered
sions
by
are
job
attributes
job
searchers
made.
Workers
are
before
may
have
not
easily dis-
acceptance
to
learn
deciabout
- 4 them by
of the
sed
actually working on
the
job.
This
aspect
job matching process has of ten been stres-
in
literature
mobili ty
is
workers
on
labor
essential
.....
a
workers
mobility:
form
have
of
attractiveness of a
over
company' s
in
the
job-shopping
by
difficulty
in
great
judging the
"Voluntary
job by talking it
employment office.
The only
way to
judge it accurately is to work on it for a
while.
Af ter
can
tell
few
whether
explains
first
a
why
few
weeks
the
quits
months
months
is
worth
job
are
of
or
most
of work,
keeping.
frequent
service
and
one
This
during
diminish
the
rapidly
af ter that point." (Reynolds, 1951.)
In this paper we attempt to incorporate uncertainty
about
non-wage
search-theoretic
such
job
attributes
framework.
attempts.
Borjas
in
the
basic
We know of two earlier
and
Goldberg
(1978)
show
that the reservation wage rule holds even when the
non-wage
ance
attribute
decision
ployed)
is
worker' s
is
made.
how
reservation
wage
is
attempt
analysis
is by vVilde
process
where
the
the
show
about
study
af ter
They
uncertainty
no
revealed
the
non-pecuniary
of
quit
(1979).
worker
the
(unem-
affected
by
component but
behavior.
The
Re considers a
is
accept-
allml7ed
to
other
decision
qui t
when
the non-wage attribute of the
job is observed. On-
the-job-search
(by
the
worker
is
ruled
therefore
out
always
assumption)
qui ts
into
and
unemploy-
ment. vvilde shows that quit probabilities are positi ve but provides
no
information of how condi tio-
nal quit probabilities change as tenure increases.
Our
analysis
search
and
non-wage
focuses
quit
on
behavior
characteristic
is
the
in
a
employed
worker's
setting where
revealed
the
(imrnediately)
af ter the acceptance decision is made.
It is shown
-
5 -
that quit probabilities decline as tenure increases,
consistent
ties.
is
observed
We also prove that the
prepared
below
his
moving
full
with
to
current
costs).
wage
worker I s
accept
and
full
a
empirical
(risk-neutral) worker
wage
wage
regulari-
offer
(in
the
that
falls
absence
of
The di fference between the current
the
acceptance wage
willingness
to
the "job change lottery".
pay
for
represents
participating
the
in
2
The Basic Mode1
Consider a
worker who is
money wage \'l=w.
particular
in
6 -
The
employed at a
job is also associated with a
non-pecuniary
attribute,
as vJ.
the same uni ts
particular
Z=z,
The worker I s
measured
full wage is
accordingly y=w+z.
A
worker
wage
contempIating
offer
search
distribution,
distribution
of
f(w)
non-wage
assume that E(Z)=O
is
facing
and
for all w.
An
weIl
as
an
unobservable
characteristic of the
job. l
to accept an offer or to
case
the
worker
offer
accepts
expires
the
non-wage
attribute
realized
value
is
of the
We
the money wage,
part,
the
non-wage
The worker may choose
refuse itj
and
in the latter
case
the
In
to work,
the
revealed.
The
may weIl be
such
starts
immediately
full
g(ZiW).
offer is associ-
immediately.
offer
known
conditionaI
attributes,
ated with an observable component,
as
a
a
wage
that continued search is considered.
The decision process is illustrated in Figure l.
Figure 1
Search
Accept offer,
sample Z
Employed
at y=w+:z
Stop search
Refuse offer,
continue to
search
Stop
search
Period 1
Period 2
- 7 The worker's objective is to maximize his discounted
life-time
costs.
Let
income,
V(y)
net
denote
of
the
search
and
moving
value of pursuing an
optimal policy for the employed worker,
00
=
V(y)
maxi y(l+r)
f
y-c+ l!r
r
max{V(y); - CM +
O
(l )
J
V(w+z)g(z;w)dz } f(w)dw }
-00
where c is the marginal search cost, CM is the
moving cost and r is the discount rate. For simplicity,
it
is
that qui ts
range of Z is
such
into search unemployment never will be
considered.
present
assumed that the
Note
value
that
of
the
keeping
term
the
y(l+r)/r
current
is
job
the
(and
thereby receiving y forever).
The
first
claim establishesthat
(l)
has a
solu-
tion.
Proposition l:
(a) There
exists
a
unique continuous
solution
to
the functional equation (l).
(b) V(y) is convex and increasing in y.
The proof is shown in the Appendix.
The optimal decision process involves two critical
wage rates. First, there is a critical "search reservation wage",
this
level,
search.
i. e.
k. Hhen the full wage falls below
y < k,
The cri tical
solution to
the worker
search wage
is
is induced to
given as the
-
8 -
00
k(l+r)
= k - c + T+r
l
r
J max t. V(k) ; - C'1L"•. +
r O
( 2)
j V(w+z)g(z;w)dz } f(w)dw
-00
Clearly,
ferent
at the full wage y=k the worker is indif-
between
searching and
keeping his
current
job. Eq.(2) expresses this equality condition.
The next problem is to show that k is unique.
Proposition 2:
There exists a unique solution to (2).
The proof is given in the Appendix.
The searching worker has
tion
wage
that
separates
to dec ide on a
acceptable
reserva-
offers
from
unacceptable ones. Hage offers for which w>w* hold
are acceptable, w* being the "acceptance reservation wage". The latter is given as the solution to
V(y) =
CM +
J
V(w*+z)g(z;w*)dz
( 3)
-00
where w*=w* (y). The next claim gives a sufficient
condition for a unique w*.
Proposition 3:
If W and
solution.
Z
are
independent,
(3)
has
a
unique
If g(z;w)=g(z) is independent of w, the RHS of (3)
is increasing in w* since V(·) is increasing. Of
course, uniqueness of w* holds true also under
less restrictive assumptions.
-
9 -
vve have now established the existence of an optimal search strategy for the decision process illustrated
in
Figure
unique
"search
l.
This
strategy
reservation
wage"
as
involves
a
weIl
a
as
unique "acceptance reservation wage".
Some properties of the problem may be illustrated
by means of Figure 2. The ray from the origin has
slope
(l+r)/r.
ray and
The
vertical
the horizontal
axis
of remaining in the current
distance between
represents
the
the value
job forever (for each
y). vVhen y<k it holds that V(y) > y(l+r)/r; hence,
the worker will engage in job search. The vertical
distance between the V(.) function and the ray
from the origin expresses the expected increase in
life-time income implied by an optimal search strategy.
For
hence,
y > k
the
there
is
no
ray and the V(·)
payoff
to
searchj
function coincide in
that interval.
Figure 2
V(y)
{
--------------- I
I
//
'
1/
I
I
I
I
I
:
I
I
I
I
I
I
"I
I
1
I
/
I
I"
I
)"
l
I
II
I
1
I
II
I
I
I
I
I
1/
I '
,
,
"/ I
(
,,' 1
I
,
,,"
/
,
,
,
,,"
1
1
, " II
/
I
"I
/
I
I
1
cc..
1
I
1
I
I
~--~--~~~----~-----------~y
k
w*
y
Recall that V(·)
10 -
is an increasing and convex func-
tion. Thus, by (3) and Jensen's inequality we have
(4 )
where
or
is a
€
Z had
size of
positive number.
zero
variance,
If V(·)
were affine
we would have
=
€
O.
The
is determined by the degree of convexity
€
of V(') and the variance of Z.
Consider
(i. e.,
the
CM
=
case when
O).
It
there
is
are
obvious
no
from
rroving costs
(4)
that
the
acceptance wage then falls below the worker's current
full
tween
may
wage
the
be
worker
full
wage
interpreted
is
vl* < y).
( i. e.,
and
the
of
the
willing to pay
The di fference beacceptance
maximum
wage
"price"
(€)
the
for participation in the
"job change lottery" .
worker
A
who
accepts
an
~'V =
w*
expects
that
he
draws
offer
a
discounted life-time income of
J
V(w*+z)g(z;w*)dz
-00
before
Z
is
observed.
expected val ue of Z,
Suppose
i. e.
Z = O.
expected life-time income falls!
In
the
that case his
This follows
from
the inequality
00
f
V(w*+z)g(z;w*)dz
~
V(w*)
( 5)
-00
which holds by an application of Jensen's inequality.
the
(Recall
worker
realized
that
is
a
convex
function.) If
vv =
w*
draws
wage will be y
=
w* + €;
who
full
V(·)
accepts
Z
expected life-time income will not change.
=
€,
hence,
the
his
3
11 -
The Quit Rate
Consider now a
=
the wage \tv
worker who
just has been hired at
w but not yet observed Z. His probabi-
lit Y (ex ante) of quitting in period n + l, conditionaI on employment in n periods, is given by
k-w
=
Pr(Qln,w)
f [F(W*(W+z»)]n ll-F(w*(w+z»)]g(z,w)dz
_-_oo____________ ______________________-k-w
~
00
n
-l
klw g(z;w)dz +
g(z;w)dz
LF(w*(w+z»)]
(6)
where F(·)
require
has
is
some
the c.d. fe
explanations.
essentially
three
the probabili ty of
the
worker
of v'L
The
a
value
(6) may
numerator ,
components.
searching;
observes
Expression
first,
them
is
search continues
if
of
One of
Z
such
that
the
full wage falls below the search reservation wage,
Le.,
w + z < k.
The
probability
searcher in n periods is
lit Y
of
accepting
l - F(w*).
two
The
terms,
observing a
an
first
value of
cess
is
terminated
term
is
the
remaining
as
LF(W*)]n and the probabi-
offer
denominator,
the
of
being
Z such
(i. e.,
probability of
in
period
next,
is
the
that
n + l
sum
of
probability
of
the
w + z > k).
remaining
the
is
search proThe
as
second
searcher
in n periods.
Inspection of
claims:
(6)
allows us to make the
following
-
12 -
Proposition 4:
(a)
Z
=
If
there
are
no
non-wage
attributes
(i.e.,
O with probability one), then the quit-probabi-
lit Y is independent of n, Pr(aln,w)
=
l - F(w*).
(b)
If there is a distribution of non-wage attributes, then, in general, the quit-probability is
öPr(aln,w)
decreasing in n,
< O
on
(c)
The quit probability
wage attributes is
in
the presenee of non-
in general approaching zero as
n increases lim Pr(aln,w)
=
O
n+ co
Claims (a) and (c)
are obvious; the proof of claim
(b) is given in the Appendix.
The
negative
relationship between quit
rates
and
tenure has a straightforward intuitive interpretation. Workers who sample very low values of Z will
set
quit
low acceptance wages
rates.
~vorkers
and
therefore have high
with higher values of Z will
select higher acceptance wages and therefore quit
at a slower pace. Those observed as having long
tenure will be those who have drawn high values of
Z; hence their acceptance wages will be higher and
their quit rates lower.
The distribution of workers by tenure reflects an
underlying
distribution
of
full
wages.
Non-wage
attributes, however, are in general only partially
observed and tenure may be used as a proxy for the
unobserved characteristics.
-
13 -
Comparative Statics
4
Expression (6) provides information about the workerls "quit-tenure profile". In what follows we
will
foeus
on how quit-probabilities are affeeted
byehanges
marginal
the
in
the
seareh
diseount
eurrent
eost
rate
(e),
(r)
money
the
and
wage
moving
the
varianee
).
Consider,
to
the
quit-response
the
(C~1)'
eost
distribution of non-wage attributes (0 2
z
first,
(w) ,
of
a
the
higher
money wage.The derivative oPr(Qln,w)/ow in (6) is
not easily signed for arbi trary val ues of n. However, for n=O we have
k-w
Pr(QIO,w)
= J
( 7)
Ll-F(w*(w+z»)]g(z:w)dz
-00
whieh
is
deereasing
in
w.
Henee,
an
inerease
in
the eurrent money wage unambiguously deereases the
first-period quit probability.
Next,
eonsider
the
effeet of an
marginal seareh eost. We
in
ök + ~ • öw*
oe
ow*
oe
oP
be
=
where P
It
inerease
is
(8 )
Pr(Q In,w).
obvious
ereases,
the
obtain 2
from
(6)
V( ) deereases
that oP/ok > O.
If e
in-
(see Appendix) and inspec-
tion of Figure 2 reveals that k deereases as weIl,
i.e. ok/öe < O. Henee, an inerease in the marginal
seareh
eost
makes
the
worker
less
inelined
seareh,
as refleeted in the negati ve first term of
(8): this resul t does hardly offend intuition.
to
- 14 Expression
case
(8)
(since
cannot
o p /ow*
that a higher c
uniess
2).
to
be
is
signed
ambiguous).
the
Note,
general
however,
will have negligible impact on w*
CM
is
very
Thus,
we
may conclude
an
in
increase
large
in
the
relative
that
to V( -)
the quit
marginal
search
(Figure
response
cost
most
likely will be negative.
Consider now the
quit
response
to
an
increase in
moving costs. We have
op
öp
ök
=
ok
öC
oC M
op
ow*
(9 )
+ ow* - oC~~
M
An increase in CM implies a
Appendix).
By
inspection
reduction in V( -)
of
Figure
that a downward shift of the V(-)
a
lower search reservation wage;
Since aP/ok >
(9)
o
2
we
(see
conclude
function implies
hence ok/oC~1 < O.
it is clear that the first term in
is negative,
i.e.,
an increase in moving costs
implies a lower search probability.
From Figure
when
the
revises
we can also infer that aw*/aC
2
cost
his
the
uously
moving
acceptance
second term in
(7»)i
of
wage
upwards.
the
quit
> Oj
worker
Hence,
is negative for n = O (cf.
(9)
first-period
falls.
increases,
M
the
Eq.
probability
unambig-
For arbi trary val ues of n,
however,
the quit-effect of higher moving costs is indeterminate.
Suppose
that
the
variance of
the
distribution of
non-wage attributes increases. We have
uP
oo~
=
aP
ok
oP
ow*
ow*
00 2
+ -- -
z
(10)
- 15 where
2
z
0
is
implies
a
the
variance of
higher
Z.
of
value
2
z
Appendix),
increase in
An
V(.)
(see
0
which in turn implies ök/oo2 > O. A. larger disperz
sion in the values of prospective job offers increases
the
returns
to
search;
hence
the
search
probability increases.
An
2 will also produce a larger c:: in
z
since the value of the "job lottery" has
increase in
Figure 2;
0
increased,
the worker
price
participating
for
åW*/ÖO~ < O unIess
crease
will
in
the
most
is
C~'1
is
willing
to
pay a
in
It
follows
it.
very large.
higher
that
Hence,an in-
variance of the non-wage attributes
likely
increase
the
first-period
quit
probability.
Finally,
it
should
be
noted
the discount rate implies a
income
associated
forever
ward
with
(y(l+r)/r).
shi f t
of
V(.)
an
increase
in
reduction in life-time
keeping
However,
the
that
the
current
there is also a
function
(Figure
job
downThe
2).
net effect on k and w* appears impossible to sign.
5
Conc1uding Remarks
Several different explanations have been suggested
for observed empirical
and quit
rates.
regulari ties between tenure
One of them is the specific human
capital
hypothesis,
between
the
according
worker I s
to
current wage
which
the
and his
gap
market
alternative increases with the length of the job.
Another
story
relies
among workers.
mobile
tenure,
are
the
on
unobserved
heterogenei ty
Since those who are inherently most
most
likely
"duration
to
show
up
with
short
dependence"
of
quit
rates
- 16 -
may reflect heterogeneity in unobserved individual
characteristics.
We have
tion,
in
this paper suggested another explana-
based
on
a
search
panded to account
non-wage
job
theoretic
framework
ex-
for imperfect information about
attributes.
Those
attributes
are
of ten difficult to observe prior to acc
de-
hence jobs will be accepted (or refused)
without full knowledge of the value of offers.
cisionsi
The introduction of non-wage attributes in a framework
of
derive
optimal
on-the-job
the worker's
Conditional
quit
wage
are
rates,
who
~Jorkers
are
less
search
decreasing
likely
as
favorable
to
disproportionately
us
to
expected quit-tenure profile.
probabilities,
sample
allows
quit
and
at
given
money
tenure
increases.
non-wage
attributes
will
represented
among
therefore
those
be
with
long tenure.
A number of issues have not been dealt with in the
paper.
A
fairly
straightforward
extension of
the
model would be to allow quits into search unemployment. Another,
and more difficult,
extension would
be to model
learning behavior among employees. An
environment
in
which
a
worker
makes
a
once-and-
for-all discovery of the true qualities of his job
is clearly not very realistic. Although the importance of learning behavior should be recognized,
we suspect that an appropriate analys is of it may
require contributions from other disciplines than
economics.
- 17 APPEllDIX
Let
B be the Banach space of continuous
LO,AJ,
on
will
never
V € B,
n
=
one
Vo =
~(V
n-
chosen
A.
so
"contraction
R.H.S.
is a
of
(l),
contraction
=
constant"
° and inductively
l)
=
~
that
that ~J + Z
large
~(V)
If
verifies
with
mapping
V
exceed
then
Define
A is
where
functions
l/(l+r).
1-
, n;;.
By the "contraction mapping theorem"
lim V = V
n
n+ co
exists
also
and
easy
is
the
to
verify
increasing,
which
is
convex
less
qualities
the
which
R.H.S.
(l+r)/r,
is
it
will
of
not
the
(2)
a
that
rate
equal
(2)
to
has
exceed
of
It
is
V
is
n
increase
(l+r)/r.
These
and this proves
l
a
rate
of
increase
+[l/(l+r)][(l+r)/r]
rate of increase of the L.H.S.
satis fied
also
be
of k.
This
unique.
It
follows
smaller
c
for
proves
also
and/or
=
Thus,
two distinct val ues of k,
satisfied
values
that
(l).
induction)
has
or
to
and also proposition 2 if we observe
does
if
(by
and
than
solution
will be inherited by V,
proposition l
that
unique
CM
for
that
all
k
is
readily
will
intermediate
(essentiaIly)
(by
give
induction)
larger
V I s,
n
i.e., larger V.
An
the
increase of the variance 0~ of Z will increase
value
of
the
integral
in
(l),
since
"Jensen effect" will be more pronounced.
same argument
as
by an increase of
above shows
0
2•
z
that V will
Thus,
the
the
increase
- 18 It
remains
rals
of
to prove proposition 4b.
(6)
we
make
the
x
=
F(W*(w+z») and introduce
c
=
j
g(z;w)dz > O
k-w
In the integ-
change
of
variables
Now
Pr(Oln,w)
=J
xn(l-x)~(dx)
xn~(dx)
f
c +
where ~ is some posi ti ve measure on LO, l J. The
condition Pr(Oln+l,w) ~ Pr(Oln,w) is now seen to
be equivalent to
J x n + l (l-x}~(dx) • [c +
~
f
x
n
Since
(l-x)~(dx) • le +
n
~(dx)J ~
f
x
f
n
x + l ~(dx)].
J xn + l (l-x)~(dx) ~ J x n (l-x)~(dx)
it
re-
mains to prove that
j
x(l-x)~(dx)
where
~(dx)
=
•
J
~(dx)
~
x n ddx).
f
~,
generali ty,
~
ure, Le.
f
~(dx)
=
that
• j x~(dx)
Sine both sides are homo-
geneous of degree two in
assume
(l-x)~(dx)
we may without loss of
is a
probabili ty meas-
1. The relation is now
or
(f
x~(dx)
)
2.
~
J
x 2 ~(dx ) .
But this is a special case of Jensen1s inequality.
O.E.D.
- 19 -
l The money wage offer is interpreted as involving
all observable aspects of the offer. To the extent
that compensating wage differentials are present,
those are captured by ~J.
2 op/ow* denotes the rate of increase of P due to
variation of the function w*{y).
-
20 -
REFERERCES
Bartel, A. P. ,
and
"Wages ,
Labor
Nonwage
Mobility",
Job
Characteristics
Industrial
and
Labor
Relations Review, April 1978.
Borjas , G. J.
Job
and
M. S. Goldberg,
Search:
A
Comment",
"The
Economics
Economic
of
Inquiry,
January 1978.
Hedberg, t1.
The
Process of Labor Turnover,
Stock-
holm 1967.
Holmlund, B.,
Labor
"The
Mobility
Turnover
and
of
Labor.
Migration
in
Studies
the
of
Swedish
Labor Market", IUI, Stockholm 1983.
Mattila, J .P.,
"Job
ployment",
Quitting and
American
Frictional
Economic
Review,
UnemMarch
1974.
Parsons, D.O.,
tion
nal
to
of
"Specific Human Capital: An ApplicaQuit
Rates
Political
and
Layoff Rates",
Economy,
Jour-
November/December
1972.
Reynolds, L.,
The
Structure of Labor Markets,
Ne\<v
York 1951.
Viscusi,
~v.
K.,
Employment Hazards, Cambridge, Mass-
achusetts 1979.
~lilde,
L.L.,
Job
"An
Quits",
(Eds.)
Information-Theoretic Approach to
In
Studies
S.A. Lippman
in
North-Holland, 1979.
the
and
Economics
J.J. McCall
of
Search,
WORKING PAPERS (Missing numbers indicate publication elsewhere)
1976
1.
Corporate and Personal Taxation and the Growing Firm
by Ulf Jakobsson
7.
A Micro Macro Interactive Simulation Model of the Swedish
Economy.
Preliminary model specification
by Gunnar Eliasson in collaboration with Gösta Olavi
8.
Estimation and Analysis with a WDI Production Function
by Göran Eriksson, Ulf Jakobsson and Leif Jansson
1977
12.
The Linear Expenditure System and Demand for Housing
under Rent Controi
by Per H ögberg and N Anders Klevmarken
14.
Rates of Depreciation of Human Capital Due to Nonuse
by Siv Gustafsson
15.
Pay Differentials between Government and Private Sector
Employees in Sweden
by Siv Gustafsson
1979
20.
A Putty-Clay Model of Demand Uncertainty and Investment
by James W. Albrecht and Albert G. Hart
1980
25.
On Unexplained Price Differences
by Bo Axell
26.
The West European Steel Industry - Structure and
Competitiveness in Historical Perspective
by Bo Carlsson
33.
The Demand for Energy in Swedish Manufacturing
by Joyce M. Dargay
34.
Imperfect Information Equilibrium, Existence, Configuration
and Stability
by Bo Axell
- 2 1981
35.
Value Added Tax: Experience in Sweden
by Göran Normann
36.
Energi, stabilitet och tillväxt i svensk ekonomi (Energy,
Stability and Growth in the Swedish Economy)
by Bengt-Christer Ysander
37.
P icking Winners or Bailing out Losers? A study of the
Swedish state holding company and its role in the new
Swedish industrial policy
by Gunnar Eliasson and Bengt-Christer Ysander
40.
Wage Earners Funds and Rational Expectations
by Bo Axell
41.
AVintage Model for the Swedish Iron and Steel Industry
by Leif Jansson
43.
An Econometric Model of Local Government and Budgeting
by Bengt-Christer Ysander
44.
Local Authorities, Economic Stabil ity and the Efficiency of
F iscal Policy
by Tomas Nordström and Bengt-Christer Ysander
45.
Growth, Exit and Entry of Firms
by Göran Eriksson
52.
Swedish Export Performance 1963-1979. A Constant Market
Shares Analysis
by Eva Christina Horwitz
56.
Central Control of the Local Government Sector in Sweden
by Richard Murray
58.
Industrial Subsidies in Sweden: Macro-economic Effects and
an International Comparison
by Bo Carlsson
59.
Longitudinal Lessons from the Panel Study of Income
Dynamics
by Greg J. Duncan and James N. Morgan
- 3 -
1982
61.
Var står den nationalekonomiska centralteorin idag?
av Bo Axell
63.
General Search Market Equilibrium
by James W. Albrecht and Bo Axell
64.
The Structure and Working of the Isac Model
by Leif Jansson, Thomas Nordström and Bengt-Christer
Ysander
65.
Comparative Advantage and Development Policy Twenty
Years Later
by Anne O. Krueger
67.
Computable Multi-Country Models of Production
and Trade
by James M. Henderson
68.
Payroll Taxes and Wage Inflation: The Swedish Experiences
by Bertil Holmlund (Revised, September 1982).
69.
Relative Competitiveness of Foreign Subsidiary Operations
of a Multinational Company 1962-77
by Anders Grufman
70.
Optimization under nonlinear constraints
by Leif Jansson and Erik Mellander
71.
Technology, Pricing and Investment in Telecommunications
by Tomas Pousette
72.
The Micro Initialization of MOSES
by James W Albrecht and Thomas Lindberg
73.
Measuring the Duration of Unemployment: A Note
by Anders Björklund
74.
On the Optimal Rate of Structural Adjustment
by Gunnar Eliasson
75.
The MOSES Manual
by Fredrik Bergholm
76.
Differential patterns of Unemployment in Sweden
by Linda Leighton and Siv Gustafsson
- 4 77.
Household Market and a Nonmarket Activities (HUS)
- A Pilot Study
by Anders Klevmarken
78.
Arbetslöshetsersättningen i Sverige
- motiv, regler och effekter
av Anders Björklund och Bertil Holmiund
1983
79.
Energy Prices, Industrial Structure and Choice of
T echnology; An International Compar ison with Special
Emphasis on the Cement Industry
by Bo Car lsson
80.
Energy Usage and Energy Prices in Swedish Manufacturing
by Joyce Dargay
81.
ELIAS - A Model of Multisectoral Economic Growth
in a Small Open Economy
by Lars Bergman
82.
Oil Prices and Economic Stability - Simulation
Experiments with a Macroeconomic Model
by Tomas Nordström and Bengt-Christer Ysander
83.
Statlig kontroll av kommunerna
En översikt av svenska erfarenheter under efterkrigstiden
av Richard Murray och Bengt-Christer Ysander
84.
Monopolyand Allocative Efficiency with Stochastic Demand
by Tomas Pousette
85.
Export Performance of the Nordic Countries 1965-80
A Constant-Market-Shares Analysis
by Eva Christina Horwitz.
86.
The Micro (Firm) Foundations of Industrial Policy
by Gunnar Eliasson
87.
Excessive Government Spending in the U.S.: Facts and
Theories
by Edward M. Gramlich.
88.
Controi of Local Authority Expenditure - The Use of Cash
Limits
by Noel Hepworth
89.
Fiscal Containment and Local Government Finance in
The U.K.
by Peter Jackson
- 5 90.
Fiscal Limitations: An Assessment of the U.S.
Experience
by Wallace E. Oates
91.
Pr icing and Privatization of Public Services
by George E. Peterson
92.
Export Performance of the Nordic Countries
1965-80
by Eva Christina Horwitz
93.
Was Adam Smith Right, Af ter All?
Another Test of the Theory of Compensating
Wage Differentials
by Greg J. Duncan and Bertil Holmiund
94.
Job Mobility and Wage Growth:
A Study of Selection Rules and Rewards
by Bertil Holmiund
95.
Sweden's Laffer Curve:
Taxation and the Unobserved Economy
by Edgar L. Feige and Robert T. McGee
96.
The Machine Tool Industry - Problems and Prospect s
in an International Perspective
by Bo Carlsson
97.
The Development and Use of Machine Tools in Historical
Perspective
by Bo Car lsson
98.
Technical Progress and Structural Change in the Swedish
Cement Industry 1955-1979
by Finn R. Fl,lSrsund and Lennart Hjalmarsson
99.
An Equilibrium Model of Search Unemp10yment
by James W. Albrecht and Bo Axell
100. Quit Behavior under Imperfect Information: Searching,
Moving, Learning
by Bertil Holmiund and Harald Lang