Digitized by the Internet Archive
in
2011 with funding from
Boston Library Consortium IVIember Libraries
http://www.archive.org/details/isthereviableeurOOblan
DEWEY
—
.-..-
HB31
.M415
Massachusetts Institute of Technology
Department of Economics
Working Paper Series
IS
THERE
A VIABLE EUROPEAN SOCIAL AND
ECONOMIC MODEL?
Olivier
Blanchard
Working Paper 06-21
July 11,2006
Room
E52-251
50 Memorial Drive
Cambridge,
This
MA 02142
paper can be downloaded without charge from the
Paper Collection at
Social Science Research Networl<
http://ssrn.com/abstract-916606
MASSACHUSETTS INSTITUTE
OF TECHNOLOGY
AUG
8 2005
LIBRARIES
Is
there a viable European social and
economic model?
Olivier Blanchard
*
July 2006
Abstract
In this lecture,
I
argue that the efficiency cost of generous but well designed social
insurance need not be very large, and that there
model, based on three
legs:
competition
in
is
indeed a viable European
goods markets, insurance
labor
in
markets, and the active use of macroeconomic policy.
Europe has performed poorly
more
observers, on both sides of the Atlantic, doubt that there
viable
European
definitive
social
More and
since the beginning of this century.
and economic
modc;l.
pronouncements on such large
I
While
disagree.
issues are unwise,
I
I
very
is
indeed a
realize that
much
believe
that the European model can work.
By "European
model,"
I
mean
a model that combines economic efficiency and
generous social insurance. So, put more precisely,
I
believe,
based on empirical
evidence, that the efficiency cost of generous but well designed insurance need
not be very large. This
in
I
two
is
the theme
I
shall develop in this lecture.
shall
do
it
steps.
shall first present
what
I
see as the architecture of a
(actual implementations, across countries
short,
I
good European model
and across time, have typically
but have informed us as to what works and does not work).
as relying
on three equally important
legs:
Competition
in the
I
see the
I
am
model
goods market,
insurance in labor maxkets, and the active use of macroeconomic policy.
that, in giving this lecture here in the Netherlands,
fallen
I
realize
in effect bringing coal
to Newcastle: In other countries, these three legs are seen as central to the
Van Lanschot
lecture, Tilburg University,
June 2006.
"Dutch model"... And
think these other countries are basically right: These
I
Dutch may see the flaws
days, the
do, but, seen from a distance, the
of their
model more
than foreigners
clearly
Dutch model seems indeed
remain a good,
to
incarnation of the ideal European model.
real flesh,
then want to confront three issues, which are often behind the growing doubts
I
about the
feasibility of the
European model. The
first is
the slowdown in Eu-
ropean productivity growth since the mid 1990s, and whether
it is
a sign that
the efficiency costs of the European model have bo;come larger than they used
to be.
The second
whether the labor market institutions which have proven
is
successful in one country can be imported with the
tries,
whether
for
example the much vaunted Danish
really solve the
problems of France or
faced by those
European countries that
architecture of the
me
that worries
The
1
I
Italy.
are
The
European model. To
the most, for
now and
"flex-security"
third
members
Euro imposes on the use of the third
straints the
same success by other coun-
is
system can
the specific challenges
of the Euro, and the con-
leg,
and thus on the
anticipate, the last issue
is
overall
the issue
for the future.
overall architecture
stated that a good model must be built on three legs, competition in the goods
me
market; insurance in labor markets; and active macroeconomic policy. Let
take each one in turn.
1.1.
Competition
The more we
in
goods markets
learn about the workings of the economy, the
more apparent
the role of competition in goods (and financial) markets. Perhaps,
more modest
in
my
claims:
of productivity, the
more
achieving efficiency.
I
my
A
I
The more
I
I
is
should be
have learned about the determinants
have been impressed by the role of competition in
shall just
mention two pieces of evidence, both related to
research interests:
while back,
MGI,
the research
ferences in productivity
ipated in
many
among
of these studies,
arm
of
McKinsey, decided to investigate
similar sectors across countries.
comparing
for
I
dif-
have partic-
example telecommunication
or
road transportation sectors across Germany and Prance, or banking sectors in
Turkey or
in
a relatively
at
how,
in
came
to the issue with
tighter. I
have been struck
Russia to benchmarks in richer countries.
My
flat prior.
many
posterior beliefs are
much
cases, differences in productivity
I
can be traced to differences in
the competitive environment rather than from access to technology, or capital,
or skilled labor.
technology more
More
intense competition truly forces firms to use the existing
efficiently,
and
forces inefficient firms out.
Turning to complementary and more quantitative evidence, very much the same
come out from the wealth
conclusions
of data on job flows
which the profession has accumulated over the
The
first
major conclusion
is
how much
last
reallocation
is
taking place in
economies: Each quarter in the United States, more than
end. and roughly the
The
same percentage
of jobs
and on reallocation
two decades:
5%
modern
of the existing jobs
created. (These are "job flows".
is
flows of workers are even higher). Perhaps surprisingly, the levels appear
to be very similar even
in "sclerotic''
Europe. This
is
dramatic evidence of the
process of job creation/destruction described more than sixty years ago
— but
— by Schumpeter.
without much empirical evidence at the time
The second major conclusion
is
that reallocation
is
indeed the main force behind
productivity improvements. Por example, a study by Foster et
concluded that more than
trade in the 1990s
retail
90%
al
[2002]
has
of the large increase in productivity in U.S.
came from
reallocation,
i.e.
the replacement of less
productive by more productive establishments, rather than from productivity
growth within a given establishment. Behind
this reallocation
is
the force of
competition.
These findings have two main implications.
First, productivity
implication, steady growth in the standard of living,
reallocation.
refer to
One just cannot have one without
There
lies
limiting
means painful job
also often
be
losses of jobs for
called "dislo-
some workers.
the challenge for policy: allowing for reallocation and growth, while
its
painful effects
on those who are displaced. This naturally takes me
to the second leg, social insurance.
1.2.
levels of
the other. But what economists
by the neutral name of "reallocation" can
cation": Reallocation often
growth, and, by
come with high
Insurance
in labor
markets
A popular slogan
is
in
FVance-at least
"Protect workers, not jobs".
among reformers,
makes
also
many European
from job
how
far
among all
politicians-
system should do and should not
do. It
from optimal actual labor market institutions are
countries: Faced with the
demand by workers
governments have too often tried to
loss,
not
indeed a good slogan, which goes to the
It is
heart of what a social insurance
clear
if
to protect
limit job destruction,
in
them
without
understanding the adverse implications of their policies for job creation.
easy to see the flaws of existing systems. And, based on both theory and
It is
a growing amoimt of empirical evidence from the diversity of experiences and
body
policies across countries (a
no small part
of empirical knowledge built in
by Jan Van Ours and others here at Tilburg),
also relatively easy to define
it is
the contours of what a good system should and can do:
A
good system should include some employment protection. But
take a rather different form from what the form
simply,
it
should be more financial,
imposes economic costs on
society,
less judicial.
it
takes in
A
it
should
Europe today. Put
firm that lays off workers
only the unemployment and other benefits
if
that the worker receives while unemployed. These costs should be internalized
by firms. The best way to do this
is
to have firms
pay contributions equal
to
the benefits paid to the workers they layoff, rather than through payroll taxes
as
the case in
is
European countries today. Experience rating systems, such
as
that used in the United States, show the ways in whicli this can be done, for
example by charging firms ex-post
have
laid-off,
This
shift
and allowing
for a
for
the benefits paid to the workers they
smoothing of
firms' contributions over time.
away from payroll taxation should come however with a reduction
the role of labor judges. Judges should be involved only
it
for age, sex, or
union
activities,
is
suspected. But
nalize the costs of layoffs, judges should have
justification
of firms, as
behind a
is
layoff.
Letting
the case in France,
is
no say
if
if
in
discrimination, be
firms are
made
in assessing
to inter-
the economic
them second guess the economic
decisions
the source of uncertainty and inefficiency,
which benefit neither firms nor workers.
Unemployment insurance
the
way
of job search.
cal evidence
is
The
should be provided, but
lesson from the
it
should not stand
now considerable body
in
of empiri-
that insurance should be conditional, not only on reported job
search, a hard-to-monitor activity, but also on training
and job acceptance
if
jobs are available. Such conditional insurance
An
is fair:
older unskilled worker
in a region with httle job creation should receive insurance for the rest of her
working years
if
needed; but a younger, skilled worker, in a large
have the option of living on the dole
if
jobs
enough; what we have learned however
ing
way
the other
will
The
should not
principle
is
simple
the details. Defin-
is
always be fuzzy and somewhat
unemployment agencies may not have
unemployed back
get the
available.
that the devil
is
what constitutes an "acceptable job"
arbitrary. Public
ai'e
city,
sufficient incentives to
may be
to work; private placement agencies
— forcing workers into inappropriate jobs. Nevertheless,
biased
existing
systems can be substantially improved, and the direction of improvement
is
a
clear one.
A
specific
There
is
workers
to
problem here
is
the employment of the least skilled or able workers.
unfortunately no guarantee that the productivity of the least skilled
will
be
sufficient to give
be a decent standard of
way
the wrong
living.
A
to
i.e.
what
society considers
minimum wage however
high
more
to deal with the issue,
unemployment than
income
them a "hving wage,"
likely to lead to
any positive outcome. The right way
is
is
clearly
high unskilled
negative
to use a
tax, under which firms pay workers according to their productivity,
and the state then adds enough to allow these workers to achieve a decent
standard of
living.
There
be a role
will still
for a
—low— minimum wage, so as
to avoid the worst cases of exploitation; but such a
be used
for
accumulating body of evidence, coming
tax,
Were such
should not
purposes of income redistribution. Here again, the devil
details, in particular the details of the negative
income
minimum wage
is
is
in the
income tax schedule. But an
in particular
from the U.S. negative
helping design better systems.
institutions put in place,
Obviously, the correct answer
is:
We
how
large
would the
efficiency cost
be?
do not know. But looking around gives us
Europe have a
hints.
These days, economists and
affair
with Denmark: The Danish combination of low employment protection
and active labor market
policies,
politicians throughout
the so called "flex-security model"
,
with low unemployment and a solid economic performance. Love
fickle.
The
has lost
its
love
have come
affairs are
previous one was with the Netherlands and the Dutch model, but
flame as the Dutch labor market has softened.
My
that the two love affairs were and are justified. Not everything
country, and the flaws are surely
more
visible
from
close.
is
sense however
it
is
perfect in either
But both
countries.
as well as a few others in Europe, provide convincing evidence that high social
insurance
is
not inconsistent witli low unemployment and sustained growth. In
other words, they provide evidence that the efficiency cost of generous social
insurance need not be very high.
1.3.
Active macroeconomic policy
The measures
I
can operate as
have described aimed at the supply
efficiently as possible, that potential
be. But, to
take up an old Keynesian theme, there
output
always equal potential output.
will
Some have argued
nomic policy
side,
will
that,
if
is
making sure that firms
output
is
as high as can
no guarantee that actual
the measures sketched earlier are taken, macroeco-
no longer be needed. They argue that
"flexibility" will lead
the economy to remain close to potential output without macroeconomic policy
argument however
intervention. Their
ferent
meanings of the word
meaning
of the
word
is
is
"flexibility"
.
based on a confusion about the
In the context of labor markets, one
the ability of firms to adjust employment in response to
changes in the economic environment. In this sense,
the need to allow firms to be
flexible.
to remain close to potential output
is
What
is
I
have indeed argued
needed however
a different dimension of
a strong response of nominal wages in response to activity.
for the
for
economy
flexibility,
namely
The two dimensions
do not necessarily come together. Perhaps the best example of
is
dif-
this proposition
the United States: U.S. firms have nearly complete flexibility in adjusting
their labor force. Yet
nominal wage
flexibility
appears very limited: Nominal
wages adjust slowly to labor market conditions, more so indeed
European
tlian in
most
countries.
This has a simple implication, the need to use macroeconomic policy to maintain output close to potential.
Adopting a passive policy stance, and
letting
nominal wages adjust slowly to labor market conditions, would lead to too slow
an adjustment process. In the United States, the lesson has been weU understood. Faced with sharp
movements
authorities have not hesitated to step
aggressively. This
The same
was very much
consumer spending or
in
in,
and
to use
in asset valuation,
monetary and
lesson applies to Europe: Increasing potential output
sure actual output
is
equal to potential
macroeconomic pohcy.
fiscal
policy
in evidence during the 2000-2001 recession.
is
just as important,
is
good; making
and requires
active
me
Let
take stock.
I
have described the basic architecture of what
viable European model.
have argued that real
I
see as a
I
flesh incarnations, in countries
such as the Netherlands or Denmark, as well as specific experiments in other
be very high.
countries, suggest that the efficiency cost of insurance need not
At
this point,
that simple?
you are
In the second part of
make a
longer
Issues
2
2.1.
likely to ask:
The answer
list,
my
is,
Are things
talk,
I
want
really that obvious,
One
to turn to three such reasons.
but these three are at the top of
really
my
could
hst.
and worries
The European productivity growth slump
European productivity growth has been very low
1
is it
There are plenty of reasons to worry.
as always, no.
since the mid-1990s. Table
(based on the work of the researchers at the University of Groningen) gives
total factor productivity
members
of the
growth
for the
United States,
for the
European Union before the recent extension),
Netherlands, since 1980.
Until the mid-1990s,
It
TFP
EU-15
(the 15
as well as for the
leads to two conclusions.
Europe than
in the
United
growth has increased
in the
United
growth was higher
States. Since then however, productivity
in
And European TFP growth
States, decreased in Europe.
is
now
substantially
lower than in the United States.
1990-1995
2000-2004
1995-2000
U.S.
0.6
0.5
1.1
1.7
EU-15
1.1
1,2
0.9
0.4
Netherlands
1.1
0.6
0.6
0.4
Source: Timmer,
Some have
1980s
Ypma and Van Ark
[2003],
and
GGDC
data base.
read tliese numbers as showing that the European model
The arguments
are
many:
"Rigidities"
to adjust, killing reallocation
make
are obviously off the mark. First,
it
doomed.
impossible for European firms
and growth. Low hours of work (the "lazy Euro-
peans") translate into low productivity.
technological frontier,
it
is
it is
And
so on.
Most
of these arguments
normal that, as Europe has gotten
has experienced a decrease in
growth. Second, "rigidities" were very
much
its
to the
rate of productivity
present before 1995, and did not
prevent nearly
of
productivity convergence to the U.S.
full
work mean a lower income per worker
have no obvious implication
Still,
one must look at the facts at
appear to have done
well,
Third, low hours
at a given level of productivity; they
for productivity
the facts are worrisome.
level.
growth.
To make progress
in
a more disaggregated
understanding their cause,
level, to find
which sectors have done poorly
in
which sectors
Europe
relative to
the United States. Here, again, Herculean work has been done by the Groningen
team.
read
I
There
is
its
conclusions as follows:
no evidence of a productivity growth problem across the board. In
manufacturing
for
example, while productivity growth has decreased over time
(presumably as a result of catch up with the United States),
than
in the
remains higher
it
United States. Instead, the difference in performance between the
two sides of the ocean can, from an arithmetic viewpoint, be traced mostly to
what has happened
in three sectors, retail
and wholesale trade, and
in financial
services.
Given the
difficulty of
measuring productivity
be skeptical of what productivity
trade
U.S. and
European
almost
As
reflects
retail
trade
in that sector. Productivity
trade productivity growth numbers
retail
some true
is
mean
one should
nearly as hard to measure, but the difference between the
in retail
is
statistics
in financial services,
difference
so large that
is
it
between the two continents.
a heavy user of IT (information technologies),
many
observers
have identified the source of the difference as coming from the use of IT. Firms
in the
their
United States, the argument goes, have been better at using IT than
European counterparts.
limits to competition he
al I
I
am
skeptical. I
am more
inclined to think that
behind much of the difference. The paper by Foster
talked about earlier has
shown that nearly
all
et
the productivity growth in
U.S. retail was due to the replacement of less efficient stores by
more
efficient
ones (the so called "Wal-Mart" revolution). There exists no equivalent study
for
Europe, but there
is
plenty of evidence that zoning and other regulations
have led to lower turnover of stores; this appears to be a good candidate for
explaining lower productivity growth over the period.
In short, there
to the
is
no strong evidence that a deep structural problem,
intrinsic
European model, hes behind the poor productivity growth performance
of the last decade in Europe.
At the same time, the most hkely explanation
for
the slowdown in retail trade (namely the protection of existing establishments)
provides a good example of the efficiency costs of good intentioned policies.
reasonable for European governments to protect the quality of their city
It is
But the
centers.
cost
may
not be negligible, and Europe
may have
to accept
permanently lower productivity (although not permanently lower productivity
growth) as a
Can
2.2.
I
result.
labor market institutions be transplanted?
argued earlier that, while no country
is
perfect, the
Dutch
models embody many of the characteristics of the economic and
had
in
mind. But, to put
Danish
One may
well
The
summarized
first,
bluntly,
it
if
doubt
it.
me
Let
in
give
the graph above, gives the
Denmark and
lower. It
is
conclusion
relation
may be
World Value Survey) "Can
piece of evidence
my
is
Put another way, the
more
in
Denmark
indirect,
papers, with
much
trade-off
will
between
may
or in the Netherlands.
Phillipon [2005].
in the 1990s),
clear:
OECD
It
shows the
and the unemployment
Again the conclu-
countries.
Countries with higher trust appear to have experienced smaller
unemployment
This correlation remains when controlling
rates.
existing measures of labor market institutions.
again that there
is
more
to success than just a
tions. Especially in the labor
tant.
be
between a measure of trust between business and labor (based on the
increases in
modify
it
but also potentially more general.
Thomas
rate in various decades since 1965, across
is
OECD
steeper, the efficiency cost of insurance
answers to a World Forum survey of firms
sion
across
There are large differences across countries:
in the four countries.
from one of
it
clear:
R-ance and Greece than
in
The second
take
mean answers
by no means obvious that the system of unemployment benefits
insurance and search effort
I
country of origin?
in their
the Netherlands score very high, France and Greece score
work the same way
be higher
is
I
which you have no right? (never=l, other an-
justified to claim state benefits to
The
model
you two pieces of empirical evidence:
countries to the following question (from the
swers=0").
social
France and Italy were to import Dutch or
would they work just as well as
institutions,
Danish
or the
—elements such as
market,
"trust"
,
less
I
for
read this evidence as saying
good design of formal
—and
easy to measure
less
institu-
easy to
or the quality of labor relations, are impor-
Importing institutions from Denmark or the Netherlands may not deliver
Dutch
or
Danish unemployment
rates. Or,
put in a positive
light,
designing a
9
Nw
Dk Swd Nth
Source:
Ajs
Cahuc and Algan,
good European
good labor
Usa
Uk Cd Che Jp
Ire
The Euro
is
Sp Pd
R
Mx
Bg Chi
Fra
a* Gre
2005.
This naturally takes
The Euro, macroeconomic
3.
Aul Ger
policy,
me
to
my
just importing
and
requires developing trust between labor
It
relations.
O
Ita
and economic model implies more than
social
formal institutions.
hfe
capital,
and
last topic.
and wage adjustments
an exciting project. But there
is
no question that participation
in
the Euro imposes sharp constraints on national macroeconomic policy. Namely,
the loss of monetary and exchange rate policies, with only
it
worth
it?
Many
fiscal
years ago, Mundell gave us the basic answer.
countries to join in a
common
pohcy
The
left. Is
costs for
currency area will be smaller, the more limited
the size of country-specific shocks, or the higher the labor mobility between
countries, or the higher the
To
wage
flexibility
within each country.
state the obvious, these conditions are far
The
result has
been a
Reunification in
series of rotating slumps, as countries
Germany
slump from which
scenario:
satisfied
today
in
Europe.
have experienced
and then gone through long and painful adjustment processes.
specific shocks,
In Portugal, a
from
it
led to a
a real appreciation, followed by a
has taken the best of ten years
demand boom
A boom
boom and
in anticipation of the
and an appreciation
loss of competitiveness,
for
Euro has
in the late 1990s,
a large current account
deficit
Germany
to reemerge.
led to a similar
which has turned into a
and anemic growth
in
the
10
Unemployment and Trust
1965-74
1975-84
1985-94
1995-2002
6
3
labor trust
Source: Blanchard and Philippon. 2004.
2000s.
that of
and
to
growth
The way
its
out for Portugal
competitors, or productivity growth above that of
do so
for
many
years. Neither
other Euro members, the
in
nominal wage growth below
to achieve low
is
growth, something
difficult to
is
competitors,
its
easy to achieve: Given low nominal wage
first
may well require negative nominal wage
achieve in the best of circumstances. And, while
Portugal has plenty of room to improve productivity, the required reforms are
politically
hard to
sell
and
to
Elsewhere (Blanchard 2006),
Portugal
I
is
for
many more
could go on. Italy
tiveness,
and
little
is
I
implement when the economy
have argued that the most
same problems
in a similar
policy,
I. see
and
for
a slump.
predicament, with a steady loss of competi-
wage adjustment. Spain, where
in the not too distant future.
blame the Euro, or suggest
in
years of low growth and current account deficits.
internal
steady growth but also to very large current account
the
is
likely forecast for
its
abandonment, but
deficits, will
My
to
demand has
led to
probably face
point however
is
not to
draw the implications
for
the implementation of the "European model" in Euro countries.
two such implications, neither of them being much emphasized these days.
11
,
Absent monetary policy and the use of the nominal exchange
tries
must adjust using two
Both must be used
actively;
tools. Fiscal policy,
which one, or
Euro coun-
rate,
and nominal wage adjustment.
what combination, depends on the
in
need to adjust internal or external demand. But the bottom hne
is
that both
are essential.
This means returning to an active use of
tool.
And, more importantly
fiscal
in the context of
policy as a niacroeconomic policy
my
argument, countries must be
able to adjust nominal wages so as to maintain competitiveness. This
adjustments up,
the case of Portugal and Italy,
or, as in
wage. If Portugal could, overnight, reduce
would imply a much smaller decrease
goods would adjust downwards as
its
of the nominal
nominal wages by 10 or 20% (which
wages as prices of domestic
in its real
well),
down
may imply
it
would substantiahy decrease
its
current account deficit and restart growth without having to endure years of
high unemployment.
Can
these adjustments take place?
—
they can only take place
I
believe that, because of coordination issues,
at least take place quickly
collective bargaining at a centralized level
and perhaps the
and Social
state.
Council...).
But
I
they come out of
between business and labor unions,
this in turn requires representative unions,
beheve however, that
fiscal
policy
lead to a rejection of the
first
two.
It is
it
continuous dialogue
—go very much against the current
in their absence, the
third leg. In the absence of this third leg,
who
— a centralized bargaining structure
in case of need, representative unions, a
between unions and firms, active
grain.
if
(Think Wassenaar here; think of the Dutch Economic
can negotiate on behalf of workers. All this
ready to be used
—
will
European model
will lack its
work poorly, and quite possibly
my main
worry
for
the future.
12
References
Blanchard, Olivier, 2006, "Adjustment within the Euro; the
Portugal",
difficult case of
mimeo MIT.
Blanchard, Olivier and
and unemployment",
Thomas
NBER
Philippon, 2004,
"The
quality of labor relations
working paper 10590.
Cahuc, Pierre, and Yann Algan, 2005, "Civic attitudes a.nd the design of labor
market institutions: Which countries can implement the Danish Flexisecurity
model?
",
mimeo
Foster, Lucia,
Paris
I,
September.
John Haltiwanger, and
C..I.
Krizan, 2002, "The hnk between
aggregate and micro productivity growth; Evidence from retail trade"
,
NBER
working paper 9120.
Timmer, Marcel. Gerard Ypma. and Bart van Ark, 2003, "IT
pean Union: Driving Productivity Divergence?", Research
67,
in the
Euro-
Memorandum GD-
Groningen Growth and Development Centre, October.
13
o35
21
Date Due
Lib-26-67