Gilbert Cette is Professor of Economics
at the University of Aix-Marseille II (postgraduate courses). He is
the author of several articles and books on reducing working hours,
including "Reduire la durée du travail: les 35 heures" in collaboration
with Dominique Taddei (published by Hachette, Le livre de poche, 1998).
The opinions expressed in this article are those of the author.
The main aim of the French authorities in the move
towards the 35-hour week started in 1997 has been to cut unemployment.
However, the reduction of working time (RWT) [in French RTT - réduction
du temps de travail] was never considered "the" solution
to the mass unemployment France was then experiencing. The best way
of fighting unemployment is obviously still growth, and RWT was at that
time considered a way of maximizing the growth in jobs so as to augment
the impact of resumed growth in the economy. The project seems for the
moment to be on the way to succeeding: in the four years from 1997 to
2000, about 1.6 million civilian jobs were created in France, including
500,000 in 2000 alone, which makes that the best year of the twentieth
century, and the unemployment rate has declined by about 1% per year
since it peaked in June 1997. The 35-hour working week will have accounted
for about 150,000 - 200,000 of these new jobs. But this job-creation
effect of the 35-hour week was made possible only by adhering to very
strict conditions of financing.
The analysis of the effects of RWT has given rise to
much debate and many publications. There is broad agreement in this
vast literature that the effects of RWT very much depend on the conditions
in which it is carried out and the ways in which it is financed. By
contrast, opinions differ sharply as to the economic and social realism
of expecting people to abide by the conditions which will eventually
make RWT effective, especially in a context where is it is pushed by
government (through legislation and regulation), as is currently the
case in France.
But before tackling this question of the conditions
for the success of RWT, we need to place the present French approach
to RWT in a more general historical context. We shall then explain the
details of this approach as driven by the two "Aubry Acts",
before concluding with a brief account of the present situation.
A universal but diversified process
RWT is not a recent development. Long-term studies
show that between 1870 and 1990, the annual average hours worked by
employees (including part-time workers) in the most highly industrialized
countries declined from about 3,000 to fewer than 2,000. This figure
still underestimates the real decline since, over the employees’ whole
working life, account must also be taken of the effect of later entry
into the labour market. This long-term trend in RWT has been paralleled
by a rise in living standards, as some of the productivity gains have
been traded off against higher wages and more leisure time.
Over recent decades, RWT has been variously reflected,
depending on the country, in the extension of part-time work, reduction
in the collective working time, longer paid holidays, and so on. In
the most recent period, this trend in RWT has been halted in many countries,
or even in some cases reversed.
Over recent years, a comparison of the figures for
average annual working time reveals three groups of countries: the United
States and Japan, where the average is apparently of the order of 1,900
to 2,000 hours, the Netherlands and Norway, where it seems to be about
1,400 hours, and the other European countries, where it ranges between
1,550 and 1,800 hours (1). France belongs to the third group, with an
"intermediate" number of hours worked.
In France, changes in working time, particularly when
it comes to reducing standard hours, require statutory action. In most
other European countries, such changes have usually come about through
the negotiation of agreements: national collective agreements concluded
on a multisectoral basis (Denmark, Netherlands, Belgium) or on a sectoral
basis (Germany, Britain).
These agreements point to a relative consensus between
employers and unions in these countries, or at least to an ability to
work out social compromises which take account both of the economic
and social constraints on the economy and on companies, and the aspirations
of wage-earners. It is this ability that explains why, despite conflicts
of interests between the two sides of industry as entrenched and powerful
as any observed in France, the move towards the reduction of working
time has continued in several of these countries.
Effects on jobs and unemployment
An analysis of the effects of RWT on jobs and unemployment
needs to distinguish between the short-to-medium term and the long term.
In the long term, the impact of RWT on the unemployment rate depends
on its impact on companies’ unit production costs. If the unit production
cost remains unchanged, RWT has no long-term impact on the unemployment
rate. RWT will reduce or increase the unemployment rate if it entails
a fall or rise, respectively, in the unit production cost. If it does
not affect the unit production cost one way or the other, there are
three sources from which RWT can be financed over the long term: the
gains in productivity per hours worked which it may induce; the gains
in capital productivity which it may also induce, through an extension
in the working life of equipment; and a wage component which may be
static (wages fall with RWT) or dynamic (wages rise less than without
RWT).
Over the short-to-medium term, and given cyclical unemployment,
the work-sharing effects of RWT are favourable to job creation and cutting
unemployment. In addition to the three forms of financing mentioned
above, there are also government subsidies, justified in the main by
a reduction in the cost of unemployment benefit and increased social
and tax revenues.
The French move towards the "35-hour week"
was launched by two Acts, those of 13 June 1998 (first Aubry Act) and
of 19 January 2000 (second Aubry Act). The question of the inflationary
dangers of such an approach has to be asked, especially in the present
context of rapidly growing employment, even excluding the positive effects
of RWT. While these risks cannot be ignored, there are a number of mitigating
factors to bear in mind:
- The process of introducing the 35-hour week for all
employees is very much spread out over time. Between the National Conference
on Employment, Wages and Working Time of 10 October 1997, which announced
the approach, and the end of the transitional regime whereby overtime
was dealt with in the context of a statutory working week reduced to
35 hours, companies with more than 20 employees will have had over four
years in which to adjust, and those with 20 or fewer employees over
six years.
- The application of the 35-hour week to all full-time
employees represents an average RWT of well below 10%. This is because
companies applying for incentive payments under the Act of 13 June 1998
are required (among other conditions) to apply a RWT of 10%, reducing
the average working week to 35 hours or less. But those applying for
payments under the Act of 19 January 2000 have no obligation to reduce
working time in percentage terms, only to bring the average working
week down to 35 hours or less (or the working year to 1,600 hours or
less). Also, many companies are adjusting the way they calculate working
time. While respecting the new definition of effective working time
(2), a more scrupulous calculation of the length of time worked often
results in a lower RWT, thereby achieving an average 35 hours a week
or 1,600 hours a year. This new method of calculation can, for example,
mean excluding certain break periods from the weekly figure, or certain
days off (in addition to official holidays) from the annual figure.
According to recent estimates, the transition to the 35-hour week thus
represents an average RWT of 4 % to 6%. On the other hand, the overtime
quota, maintained at a final rate of 130 hours a year, represents an
increasing percentage (and hence "flexibility") of the new
working time. Finally, as concerns small enterprises (10 or fewer employees),
it should be noted that there are specific provisions (pre-dating the
Aubry Acts) permitting great flexibility in the application of certain
constraints on the number of hours (concerning compensatory rest periods).
- Finally, the two Aubry Acts permit a variety of flexible
arrangements whereby companies may derogate (within limits) from certain
provisions of the working time legislation, provided such arrangements
are negotiated and organized through collective bargaining. This applies,
for example, to the annual calculation of overtime if the agreement
provides for an annual adjustment of working hours, or in the calculation
of executives’ working time by days worked in the course of the year.
Among these flexible arrangements, we may also note the flexibility
which allows observance of the 35 hours (again within certain limits)
to be spread over five years [the hours worked over 35 can be added
together and compensated for by holiday - Compte épargne-temps or
CET in French]. Moreover, the Act affords increased security for the
35-hour agreements by ruling out, in the event of "majority agreement"
(3), the risk previously faced by companies of having to go down the
expensive road of collective layoffs and the accompanying redundancy
programme if ten employees reject certain clauses of an agreement on
the grounds that this represents a substantial change in their work
contract. In fact, so far as the working time legislation is concerned,
the two "35-hour Acts" have considerably widened the scope
of contract law at the expense of statutory law. In this respect, France
has come closer to the practice of other industrialized countries where
collective bargaining plays an important role.
The Act of 19 January 2000 also provides a guarantee
that there will be no decrease in the monthly wage for workers being
paid the SMIC (minimum wage). The risk of an increase in wage costs
because of this provision appears small. The lower level of social contributions
provided for in the Act of 19 January 2000 accounts for only 6% of the
wage cost of such workers. Assuming productivity gains equivalent to
one third of the RWT, the cost of going from 39 to 35 hours is almost
completely covered.
The process of RWT involving the universal introduction
of the 35-hour week is, for the moment, progressing very well. In September
2000, out of 12.5 million full-time employees falling within the scope
of the two Aubry Acts (i.e. not including civil servants, local government
officers and employees of public corporations), about five million,
i.e. more than 40%, are reported as already on 35 hours or covered by
a 35-hour agreement. The figure is said to be over 50% for companies
with more than 20 employees, but about 5% for companies with 20 or fewer,
for which the deadlines are later. The collective bargaining process
thus stimulated is unprecedented in France, both for companies and for
sectors of the economy.
In terms of the financing of these RWT, the assumption
that companies’ unit production costs would remain neutral seems on
average to be proving correct, with about a third of the cost incurred
being offset by productivity gains, another third by the lower level
of social contributions, and the rest by wages in the form of wage restraint
or even a wage freeze over a period of one to three years.
Lastly, RWT seems on average to have spurred the creation
of two thirds of the number of jobs which could theoretically have been
expected as a result of RWT. If the process continues in the same favourable
financing conditions, the total number of jobs created, when the 35-hour
week has become universal, i.e. in a few years’ time, could be 400,000
to 500,000 in companies with more than 20 employees and 150,000 to 200,000
in those with 20 or fewer.
These favourable results show that the introduction
of the 35-hour week is confirming a trade-off temporarily oriented towards
more leisure time rather than (in part) higher wages. This certainly
does not mean there is no danger that the favourable effects of this
process may not be to a greater or lesser degree diluted over the next
few years by wage inflation adversely affecting its financing. The risks
are principally of two kinds:
- there could be a selective bias in the negotiating
process now under way, companies in which negotiations have been completed
enjoying a doubtless more favourable social climate than the others.
On this assumption, which looks pretty realistic, the difficulty of
financing the transition to the 35-hour week without an increase in
unit production costs would be greater in companies where negotiations
have not yet started or been concluded than in companies where they
have already resulted in an agreement;
- growing tensions in the labour market, sharpened
by the employment effects of the spread of the 35-hour week, could culminate
in wage acceleration. In this context, competitive hiring could spell
the end of the wage-restraint clauses in some 35-hour agreements.
As can be seen, the risks exist and call for the closest
attention to developments in the labour market. For all that, no ambitious
policy is entirely risk-free, and it should be stressed that, so far,
the 35-hour week has led to the creation of many jobs without wage inflation.
(1) These figures are taken from the OECD’s Employment
Outlook, July 2000 edition.
(2) Article 5 of the Act of 13 June 1998 reads as
follows: "The effective duration of work is the time during which
the employee is at the disposal of the employer and must comply with
his/her instructions without being free to go about his/her private
business."
(3) I.e., in the case of agreements signed by trade
unions representing a majority of members following the most recent
elections, or validated by a majority of employees when the work force
is consulted on the specific issue.
(January 2001)./.
