Are labor market "rigidities" responsible for Europe's unemployment?

Yes, according to the OECD, IMF, and the conventional view among many academics. The simple intuition that if you make it more difficult to fire workers employers will be less willing to create new jobs, and that in a dynamic economy this will lead to higher unemployment, seems compelling enough. Plus, there is a large number of cross-country econometric studies that apparently show labor-market protections, particularly unemployment benefits, to be correlated with unemployment levels.

A new paper by Howell, Baker, Glyn, and Schmitt reviews this empirical literature and finds it badly wanting.

The main culprits are held to be protective institutions, namely unemployment benefit entitlements, employment protection laws, and trade unions. Our assessment of the evidence offers little support for this orthodox view. The most compelling finding of the cross-country regression literature is the generally significant and robust effect of the standard measure of unemployment benefit generosity, but there are reasons to doubt both the economic importance of this relationship and the direction of causation. The micro evidence on the effects of major changes in benefit generosity on the exit rate out of unemployment has been frequently cited as supportive evidence, but these individual level effects vary widely across studies and, in any case, have no direct implication for changes in the aggregate unemployment rate (due to “composition” and “entitlement” effects). Finally, we find little evidence to suggest that 1990s reforms of core protective labor market institutions can explain much of either the success of the “success stories” or the continued high
unemployment of the large continental European countries. We conclude that the evidence is consistent with a more complex reality in which a variety of labor market models can be consistent with good employment performance.

In his comment on this paper, Jim Heckman agrees that the cross-national evidence is weak and fragile. But:

In the absence of better data, and better measurement frameworks, prior beliefs will continue to dominate how one interprets the evidence. This is not as
much about dogmatism or conspiracy as it is about good science. In the absence of empirical evidence, logically consistent stories that accord with intuition have great appeal. At both an intuitive level and at the level of formal economic theory, incentives matter. If a person is paid not to work, the person will likely not work. If the costs of hiring a worker rise, fewer workers are likely to be hired. The microevidence supports these basic predictions of theory. Like the controversy over the effects of minimum wages, disagreements are not over qualitative predictions of the theory, but are about quantitative empirical responses.

HBGS are splendid critics. However, they do not offer a constructive empirical alternative to existing practices in the literature. They have not proved that institutions do not cause the pattern of European unemployment. They have, instead, shown that the current data base and models are too weak to decide the issue.

In at least one area of labor-market intervention, employment protection (firing restrictions), the orthodox expectation is not the only "logically consistent" story that accords with economic intuition. If you make it harder to fire an employee, you essentially give that employee some property rights over the job he occupies. Now, according to the Coase theorem, how property rights are allocated (to the employer versus the employee) has no effect on efficiency in the absence of bargaining and other transaction costs. If eliminating a job is the efficient thing to do, an employer can do it either by fiat or by paying the employee to leave. There are distributional implications (obviously the employer is worse off in the latter case), but nothing to stop the efficient thing from getting done. This is an idea I associate with my Harvard colleague Richard Freeman.

In the real world, there are of course transaction costs, and bargaining between worker and employer is not costless. But automatic dismissal has its own costs. Ultimately an evaluation of the "protective" regime vis-a-vis the liberal model is not that straightforward. We have to compare two regimes with different distributional incidence and different sets of transaction costs.

This way of thinking has another implication. Employment protection legislation should have the least adverse effects in systems that facilitate employer-employee bargaining. There is a complementarity between these two elements of labor-market institutions.

Comments

[robertdfeinman]

One of the sites I visit frequently is the European Tribune blog at http://eurotrib.com.

It's a European-based blog (in English) and there are several regular essayists who have discussed the issues mentioned in these papers frequently. Much of their complaint seems to be the rise of Reaganism/Thatcherism in many of the EU countries.

Sarkozy is see as the most recent example of this as is the attack on the unions in Germany.

Anyone interested in the topic might want to check the site from time to time and see how Europeans view the situation from the inside.

[hari]

Dani-

Labour market regidities are part and parcel of EU social market economy.

For example, Germany is par excellence the prime case of a labour market which is not only rigid but ideologically tilted against capital, as in France.

Yet, with some adjustments under Schroder and now under Merkel, Germany's export engine is at its peak now> can it be that social market regidities are not the real culprit?

In my opinion, EU ( with exception of UK and Scandinavians) does not regard its social market economy as a constraint to expansion and growth.

Space here is limited to go into details - but your fellow economists make a mistake by comparing EU internal market with US.

I remember a colleague who wrote a Ph.D. disseration at Berkely - during Eisenhower admin - on "How to Lie with Statistics".

In all fairness to the "soft" science of economics, I suggest you don't mix or confuse EU internal market developments (rigidities) without focusing on its
cost/benefit analysis.

[Slavisa]

The parallel with the Coase theorem does not work here, because it applies only on the workers-employers relationship. There will be an efficient solution between workers and employers, no matter who has property rights.
But there is no Coasian bargaining for the unemployed, they are out of the game. So your Coase parallel really discusses unemployment by excluding it.

[paine]

class struggle politics
dressed for high tea
in the front parlor

"There are distributional implications (obviously the employer is worse off in the latter case)"

[paine]

slav

"there is no Coasian bargaining for the unemployed "

unless there's
a universal citizens
right to a job
but of course that's beyond coase
the bargaining whole may indeed
be the sum of its individual units
but a diffrerent whole
bargained for as a
different whole ....

bargaining
for revision of
the social contract
led to the short week
and jobholders rights
but a right to be
a jobholder ...

might that not threaten the very system
of wages and profits itself ???

[Peter]

The intuition that Heckman draws on regards jobs as created by firms and offered, at a moment in time, to workers. A lower price for labor means a larger offer. But in workplaces in which worker performance has an effect on productivity, workers also, in a sense, create jobs. They do this by improving firm performance, which has employment consequences. It is not difficult to come up with a counterintuition about employment security, worker investment and performance. In fact, this is how lots of people think about this question in social market economies.

Empirically this is difficult to get at, since labor productivity has many determinants over which workers have no control. We are interested in the net effect of employment relations systems, but net of what? Of course, there is an enormous literature on this question, which Rick Freeman knows well (and has contributed mightily to).

I get the impression, however, that this large body of economic research is simply ignored by much of the profession, which can see the world only through the lens of instantaneous, impersonal supply and demand.

[paine]

hari

" ideologically tilted against capital "

the truth of that comrade may reside in the eye of the beholder

to the evil blood soaked
red (like me)
the social market
is capitalism
with a nurses face

contextual question :

which nurse face
is it
Ratchet or Nelly ??

[paine]

peter

"We are interested
in the net effect of
(the)employment relations systems "

vide
das crap-ital vol v :
the critique of
comparative employment systems

[Butter]

Who is the least cost avioder in this employee-employer Coasean model? I would argue that the employee is, since the employee knows how productive she is while the employer relies on signals. I'm thinking the property right should be assigned to the employer based on the fact the employer is at an asymetric information disadvantage.

[paine]

butter
with your cost avoider paradigm
you underline
the ever present danger
of the chinaski effect
ie
shirking
the rise of
the iww
the i won't work movement
implicit in any
job right system

where negative incentives are disolved away
after the sack....

you show up to the job sight and shirk
till they find out just how useless you are
and
you get laid off

but there's always another job for you out there
by right ...

its like
the old soviet joke
about useless rubles
creating useless
laborers

if at the margin
your job income
is collecting in the bank
because of
a chronic consumer product famine
you get this:

"they pretend to pay us
so we pretend to work "

analogy
under job rights system

"they pretend to fire us
so we pretend to work "

its like school compulsion
up to age 16

they can make you attend but they can't make you learn

so we're well trained for this wonderland reform

conclusion
some reforms might trigger
a collapse of the system

i'm all for that of course

but we all can read about
the battery of neasures used
to re impose " factory labor discipline"
during the soviet NEP ...

[Kimmitt]

Is the employee's right to a job transferable -- that is, can a contract with a severance be drawn up which is enforceable under the various European laws?

[corvad]

i can't help it, I love it so much: "the chinaski effect."

I just want to make sure everyone who sees it knows to read charles bukowski (please don't ever see a film about him, unless it's footage of the man himself). not sure there has ever been such a disastrous horrifying sick drunk pig of self-reflexive unitedstatesian cultural critique, and oh how erudite and beautiful.

reading his ouvre also worth a degree in labor studies, if you ask me.

thanks paine.

[wjd123]

Thinking intuitively if I own a house most people would say that I'll take better care of it than if I rent. Wouldn't this be the same for an employee who has a greater proprietary right to his or her job, particularly if getting another one will be hard to do. Job holders will want to see that the job lasts. The employer will also want to see that employees are well trained since it will be harder to replace them.

A few years back I read that France mandates employers to spend a certain percentage of their yearly profits to train their employees--I believe it was 2%--or pay an equivalent tax to the state. It would seem that the transaction cost of losing an employee would be a lot higher for the French employer than for a U.S employer. Yet, thinking intuitively, even if France's employment transaction cost are higher there should be less of them.

[paine]

hey
job "markets "

sound nature ??

once slave markets did too

[Peter]

I forgot to mention: the Coase template has been shown not to apply in the case of employment protection, at least in some of the articles collected in Behavioral Law and Economics (Sunstein ed., Cambridge UP, 2001). One interesting tidbit: most US workers erroneously think they have employment protection. They don't realize they can be legally dismissed without any justification, so of course they won't bargain for protection.

A second point will be familiar to academics who follow this blog. Tenure for professors is justified, among other reasons, by the role faculty play in university governance: employment protection enables them to be freer agents when it comes to exercising their governance powers. A parallel argument can be made in European countries with a degree of co-management, with Germany the most obvious example. I will grant, however, that protection is also strong in many countries with no provisions for worker participation.

[corvad]

i've always wondered and perhaps this is a good opportunity while on the subject of job loss: what economic incentive theories justify golden parachutes for bad CEOs (or college football coaches, for that matter)?

[Rupert]

There is no single labour market in many European countries, there are two. The insiders have protected jobs and generous benefits if they are unemployed or retire early. But most people under 25, immigrants and (sometimes) women are not covered by the job protection or eligible for the benefits.

So the reason that HBGS can't find evidence that generous benefits raise unemployment might be that France, Germany, etc have kept their generous benefits for insiders, but made it harder to get them. So unemployment falls as students and immigrants are forced into poorly paid, insecure jobs . . . while the insiders get benefits worth 80% their previous salary.

This could be a lasting adaptation: after all, lots of young Europeans (and grad students) are willing to work as unpaid interns if it helps them break into the inside market!

[Jeremy McKibben]

Dani,
I do not understand how your example contradicts Heckman's comment, because it would seem that the distributional effects do matter when employers are making hiring decisions. Even if workers will still be fired when it is efficient to do so, the costs of firing will be higher, which will discourage employers from hiring in the first place. I agree that the effects will be softened when transactions costs are lowered, perhaps by institutions that facilitate bargaining, but our best model still predicts a negative employment effect of employment protection.

I don't think we can conclude from the above reasoning that employment protection is a bad idea, but it was not Heckman's intention to make a normative judgment.

Do you have any other logically consistent models that suggest there will be no negative employment effect?

[piglet]

"what economic incentive theories justify golden parachutes for bad CEOs"

Indeed one of the most amazing aspects of this discussion (which frequently makes me want to puke) is the fact that in most companies, the employees who enjoys the highest degree of job protection is, by far, the CEO. Yet the same economists who fulminate against the little of job protection that ordinary workers have in countries like Germany (it is little, and it certainly doesn't prevent employers from frequently making mass layoffs), those are usually the same who will cite the most absurd arguments in support of wildly exaggerated CEO compensation.

[Justin Rietz]

I haven't had a chance to carefully read the entire paper, but my initial reads revealed glaring problems with the authors' claims / methodologies:

1. The chart on page 17 plots net replacement rate against the unemployment rate, and the authors observe that that there are multiple countries with higher NRRs yet lower unemployment rates than the U.S. However, the authors fail to mention that the U.S. was in the middle of a recession - a recession that hit U.S. employment particularly hard because of the dot-com bust. In addition, the correlation is poor, so the data neither proves nor disproves anything.

2. Page 18, graph of duration of benefits vs. unemployment: ditto comments from #1.

3. The data does not take into account government provided jobs. Government jobs that are relatively easy to come by could be considered an entitlement. It would be interesting to look at the % of the population employed by the government.

4. Country size: eyeballing the data, I noticed that there are a disproportional number of small countries that are inconsistent with the consensus theory. I won’t try to theorize why this may be the case, but certainly there are many reasons why small countries are not good “laboratories” for economic phenomena.

In general, given the authors' use of admittedly so-so data and their inattention to the elements I mentioned above, I don't think their inability to find correlations that match the consensus view proves anything.

[derrida derider]

There's some disturbing methodology questions raised by Heckman's response. IMO theory is no substitute at all for empirics simply because its quite easy to invent a plausible theory to fit almost any conceivable result. If we don't have the empirics then the right answer is "we don't know", not "theory suggests ...". And a lot of trouble in economics has flowed from people's failure to grasp that.

It's easy to invent theoretic reasons why, for example, labour supply curves slope down (eg, crowding out of intrinsic motivations) or demand curves slope up (eg positional goods). Or where minimum wages raise employment, or where employment protection reduces unemployment. As Heckman above all should know theory is the way to develop testable predictions for empiric research, but is no guide to policy absent those empirics.

As an aside, I hate it when US economists refer to "Europe" as if it's one country. "Europe" is very diverse in its national institutions, its economic structure, its cultural propensities, and accordingly in its labour market outcomes.

[derrida derider]

And justin, that they haven't proved anything and neither have those who claim "flexibility" is the answer is precisely the paper's point - something that Heckman acknowledged.

Even if your strong priors lead you to discount this paper, you still have to explain the "disappointing" results of the OECD's much more extensive studies. Simply put, there is no smoking gun linking adverse unemployment outcomes to employment protection legislation, minimum wages or union involvement in management. Absent this, the gun itself is a purely theoretic construct which we could easily imagine pointing in the other direction.

[Justin Rietz]

Derrida derider -

While I haven't read all of the studies mentioned and thus will take at face value what has been said about them, I agree that there is no empirical smoking gun.

However, the authors of the study in question go beyond saying "no one knows." They suggest that the consensus theory is wrong. Given their data and methodology, I don't think they have a strong foundation to make this case.

I do agree with you that empirical evidence is important. However, we often run into situations where there is not, and may likely never will be, good, clean, consistent data. This is especially true for "big" topics (as we have here) that involve so many relevant variables. In addition, as you say yourself, countries have a wide range of cultural propensities, further complicating the picture regardless of data quality.

In these cases, we are left with reason, logical thinking, and common sense.

[Patrick R. Sullivan]

'Now, according to the Coase theorem, how property rights are allocated (to the employer versus the employee) has no effect on efficiency in the absence of bargaining and other transaction costs. If eliminating a job is the efficient thing to do, an employer can do it either by fiat or by paying the employee to leave. There are distributional implications (obviously the employer is worse off in the latter case), but nothing to stop the efficient thing from getting done.'

Interesting implications here, to say the least.

Repeal the XIII Amendment and give employers a property right to their employees. Would that have much effect on how many people would ask to be hired?

[robertdfeinman]

You might be interested in this essay from "Le Monde" on the issues in France.

This version on eruotrib has a parallel English translation and lots of graphs to support the author's case.

Even if you don't agree with his argument, the charts are a valuable resource in themselves:

http://www.eurotrib.com/story/2007/9/10/9417/13559