Déclin et renouvellement de la main-d
17 Pages
Gain access to the library to view online
Learn more

Déclin et renouvellement de la main-d'oeuvre industrielle


Gain access to the library to view online
Learn more
17 Pages


De 1975 jusque dans les années 90, la réduction des effectifs ouvriers s'est portée d'abord sur les plus jeunes par la diminution des recrutements, puis sur les anciens par les départs en retraite ou préretraite de plus en plus précoce. Cette réduction a aussi eu pour effet de modeler la structure d'âge de la main d'oeuvre, la centrant sur les âges intermédiaires. Les structures d'âge, telles qu'elles se présentaient il y a dix ou quinze ans dans les différents secteurs, permettaient de faire façe à certaines contraintes de travail par leur répartition différenciée selon les âges. Ces formes de sélection sur l'âge sont désormais remises en question. L'analyse des structures d'âge et de leur évolution conforte donc la nécessité de s'attaquer à la conception et à l'aménagement des moyens de travail, aux conditions et à l'organisation du travail pour pallier les effets du viellissement en cours de la population ouvrière.



Published by
Reads 54
Language English


Industrial Workforce Decline
and Renewal
A Reading of Age Structures*
The reduction in the number of manual employees from 1975 to the 1990s hit firstAnne-Françoise
the youngest in the form of less recruitment and then the oldest in the form ofMolinié**
retirement and increasingly young early retirement. This reduction also changed
the workforce’s age structure, centring it on the intermediate ages.
The age structures found in the different sectors ten to fifteen years ago helped
overcome certain working constraints by means of their different age
breakdowns. These forms of age-based selection are now being challenged.
An analysis of age structures and their evolution upholds the need to address the
design and development of ways of working, working conditions and job
engineering to offset the effects of the current ageing of the manual employee
he French working population, like working policies are starting to be questioned in certain* Originally published as
“Déclin et renouvellement Tpopulations in most of the industrialised countries such as Germany and Italy. Moreover,
de la main-d’œuvre
countries, has seen two contradictory trends in the working-life characteristics of the post-warindustrielle : une lecture
des structures d’âge,” recent years. Firstly, the overall ageing of the generations (later entry into the labour market,
Economie et Statistique, population due to variations in birth rates and the high female participation, etc.) could create an
No. 316-317, 1998.
later age at which people are starting work has environment encouraging a change in the working
raised the proportion of oldest employees. patterns of the oldest workers.In any case, the
Secondly, retirement policies – retirement, early proportion of 40-55 year olds will probably
retirement and all other forms of distancing the continue to rise and the share of those aged 55 and
senior workforce – have brought about a decrease over remain stable over the next twenty years.
in participation rates among the oldest members of
the population. In coming years, the proportion of Underlying these general trends are substantial
young people will probably decrease further due to disparities by sector and company. This said, many
the smaller sizes of these generations and a companies, especially in industry, now have an age
probable continuation in prolonged education, pyramid predominated by 35-50 year olds (Le Minez,
** Anne-Françoise Molinié even if the rate slows (Brondel et al., 1996). 1994; Marchand and Salzberg, 1996). Yet the
works for CREAPT (Centre
However, given the uncertainties currentlyfor Research and Studies
1 A variant of the French working population forecasts was puton Working Ages and affecting the social security systems, the
Populations), 41, rue together to study the effects of later retirement, mainly by gradually
downward trend in retirement ages could well slowGay-Lussac, Paris, France. raising the retirement age (by five years from 2000 to 2020)
1down or even reverse. Early retirement incentive (Blanchet and Marioni, 1996).
INSEE Studies no. 43, April 2000 1existence and magnitude of any “problems” on car assembly lines, shifts and night shifts in the
associated with ageing depends to a large extent on chemical industry, etc. (Teiger, 1989). Hypothesis
working conditions and job engineering. It is with A based on the findings of national statistical
regard to these elements that the disparities should be surveys posits that this selection is made
assessed. essentially in line with two main models (Volkoff
and Molinié, 1995): selection-reassignment in
sectors such as motor vehicles where the gradual
Harder working conditions for old appearance of problems with “keeping up with the
workers line” with age could be solved by a sufficient
number of off-line jobs; and selection-exclusion in
The general change in working conditions in electronics and clothing where considerable speed
France, as described by the national surveys and precision requirements are satisfied by the
(Cézard and Vinck, 1996; see box 1), is believed to ongoing recruitment of very young female workers
have created more problems for ageing employees: and the departure of many of them before the age of
working hours are increasingly unstable 30. As regards shift work, shift workers often
(staggered, split, night shifts, etc.) and time request reassignment to the daytime around the age
pressures are heightening with the persistence of of 40 to 45 years old. Long-term individual paths
line work, growing “commercial” constraints in analysed using the Santé, travail et vieillissement
industry (growing dependency on demand) and, at (ESTEV) survey, confirm the selective nature of
the same time, growing “industrial” constraints in these constraints. Each age group comprises as
services (production norms) (Gollac and Volkoff, many individuals exposed in the past, but no longer
1996). exposed, to repetitive work with a strong time
constraint or shift work as those still exposed. Long
Adjustments of time constraints to age have long periods of exposure are rare. It is as if a large
been largely regulated by selection mechanisms.
2 The “ESTEV” survey (health, work and ageing survey) was firstStudies of companies by ergonomists have long
carried out by nearly 400 company doctors in 1990. It provided
shown that work constraints in certain jobs set “age various elements on the current and past working lives of 21,000
employees of four different ages (37, 42, 47 and 52 years old inlimits” that vary depending on the case: combined
1990) as well as the results of a detailed medical examination.precision and speed requirements in certain
The same subjects were seen again in 1995 (Derriennic et al.,
clothing workshops, effort and posture constraints 1996).
Box 1
The Conditions de travail (working conditions) surveys term – rely more on the interviewee’s assessment (“are
(1978, 1984 and 1991) were made of employees and you exposed to …”). These areas concern physical
were based on their statements. These surveys were exertion, temperature, exposure to risks and noise
organised and processed by DARES at the French (although an attempt at objectiveness is made in this
Ministry of Labour. They provide complementary area), and even memory requirements.
information to the INSEE Emploi (employment) survey.
The samples comprise some 18,000 employees A comparison of the 1984 and 1991 surveys reveals a
interviewed at their homes. deterioration in working conditions as perceived by
employees in different fields. The survey authors
These surveys are not based on measurements or (Cezard et al., 1992) consequently endeavoured to
valuations of jobs in companies. What the employees estimate the extent to which this deterioration was due
say reflects their knowledge – or lack of knowledge – of to a change in attitudes to working constraints. The
their work situation. However, the questionnaire is changes measured from 1984 to 1991 refer to a number
designed to limit or curb “subjectivity” effects to a of different interpretation registers (Gollac, 1994). The
maximum by using as objectively orientated first concerns changes to the survey itself. These
formulations as possible and by looking at certain work changes probably had some effect, albeit limited.
aspects using cross-referenced questions. For Others concern the social processes that objectivise
example, pace constraints are studied using a question working conditions and render them verbalised and
on line work and an inventory of possible work rate declared in the surveys: mobilisation of occupations,
determinants (constraints associated with the raised awareness of risks, better standards of living,
automatic speed of a machine or the automatic moving etc. However, the deterioration in working conditions
of a product or part; constraints associated with felt by the employees is also due to other more objective
compliance with norms and deadlines; and constraints causes. The main cause is probably the increase in
associated with the more or less pressing nature of sources of pressure on the work rate. The hardness of a
client or public demand). task depends a great deal on the time available to
Questions regarding other areas – especially those accomplish it and a tightening of time constraints can
concerning the “work load” in the traditional sense of the influence the survey “declaration threshold”.
2 INSEE Studies no. 43, April 2000proportion (approximately half) of the employees the meat and dairy industries, the other food
exposed to such constraints decide to leave such industries and the rubber-plastics sector.
work or are removed from such work after a certain
number of years (Volkoff et al., 1992). Age structures were changed a great deal by
general demographic trends, decreases in worker
numbers and employment management methods
Age-based selection challenged from 1975 to 1990 (see table 2). The proportion of
young manual employees plummeted in virtually
Today, these selection processes can no longer all the industrial sectors. It only remained stable in
work overall in the same way due to the the food industries, which posted much more
contradiction between widespread ageing, keeping favourable and even positive growth in worker
women in the middle age bracket in work, and the numbers. Sectors such as textiles and clothing,
change in working conditions and job engineering. leather and footwear, and the electrical and
Formal and informal “human resources” electronic industry had the highest proportions
management methods based on forms of mobility ofyoungmanualemployees in 1975 (over 45%).
and reassignment associated with age or seniority Yet by 1990, they were among the relatively “old”
will probably be challenged. Some manual sectors with similar proportions to those found for
employees in the middle age bracket are already the oldest sectors such as iron and steel and
facing organisational forms that concerned building materials in 1975 (30% under 30 years
primarily the youngest a few years ago (Molinié old). The “natural” evolution in the age structure
and Volkoff, 1993). and the slowdown in recruitment brought about an
increase in the percentage of employees in the
However, companies take varying attitudes to “middle” age bracket in sectors where the initial
ageing and there are still relatively few initiatives age structure was relatively young and led to an
aiming explicitly to anticipate ageing by altering
working conditions and job engineering. The most Table 1
Nearly all the industrial sectors cut manualfrequent attitudes are still indifference and
employees from 1975 to 1990attempts to take action mainly by remodelling the
age structure, often based on early retirement
Growth in manual employees in the main industrial
measures (Volkoff, 1996). It is these management sectors from 1975 to 1982 and from 1982 to 1990
methods that will be analysed here from the point
%of view of how they are linked with demographic
developments and changes in employment. The
Growth instudy is based on data taken from various censuses
Sectors manual
since 1975 (see box 2). It concerns solely industrial employees
manual employees and uses a sector approach (see From From
1975 1982box 3).
to to
1982 1990
1Iron and steel -34 -50
Basic chemicals and man-made fibresStaff cuts have completely changed age -30 -28
Textiles and clothing -27 -28structures
Leather and footwear -27 -26
Shipbuilding, aircraft and armaments -16 -31Demographics and employment trends are two
Building materials -24 -19
major determinants in the development of age
Mechanical engineering -20 -7
structures (Molinié, 1993). From 1975 to 1990, the
Electrical and electronic manufacture -20 -15
number of manual employees in industry fell Land vehicles -10 -20
sharply (see table 1). The average annual rate of the Paper and cardboard -24 -5
drop was 6% from 1975 to 1982 and 8% from 1982 Building, civil engineering and -15 -10
agricultureto 1990 in the iron and steel industry (ores, ferrous
Smelting and metalworking -19 -3metals and primary steel processing) and at least
Wood, furniture and miscellaneous -6 -134% to 5% in sectors such as chemicals, textiles and industries
clothing, and leather and footwear. These rates are Printing, press and publishing -11 3
comparable to the rate of decline in agricultural Other food industries -5 -0.5
workers during the mass rural exodus from 1955 to Rubber and plastics -6 2
Meat and dairy industries1972 (-4% per year) (Marchand and Thélot, 1991). 19 -2
The only sectors to remain relatively stable or 1. Minerals and ferrous metals and primary steel processing
increase their numbers of manual employees were Source: population censuses.
INSEE Studies no. 43, April 2000 3increase in the percentage of employees aged 45 especially retirement ages and procedures. The
and over in sectors where it was older. However, potential impact of retirement age decisions can be
since 1982, the introduction and expansion of early estimated by comparing the observed trends with a
retirement measures have curbed the increase in scenario in which all the manual employees in
the proportion of the oldest employees. work at the beginning of the period remain in the
sector with the exception of those who have
reached the official retirement age and with no
The age structures have influenced other entries than any recruitments of young
management methods people.
Conversely, the age structures have probably This reference scenario, with its stability of manual
affected workforce management methods, employee numbers in the sector and no mobility
Box 2
Population censuses
This study covers the 1975-1990 period and draws on This source covers much larger numbers of people than
three censuses: 1975, 1982 and 1990. The classification the other sources on occupational mobility (the
of activities (Nomenclature des Activités et des Produits) Vocational Skills Training survey and the Careers
did not change over these fifteen years. However, the survey, complementing the 1989 Employment survey).
socio-economic classification was overhauled in 1982. This provides large enough numbers of manual
The 1982 and 1990 censuses used the PCS employees in most of the NAP40 classification’s activity
classification (occupations and socio-economic sectors. Moreover, recording occupational mobility
categories), whereas the 1975 census was based on the using the EDP means there is no need to call on
CSP (socio-economic category). To ensure memory. Nevertheless, the absence of an interviewer
comparability over time, occupations were also coded in makes it impossible to identify fictitious changes of
thline with the old classification for the 1/20 sampling of situation. For example, when a company is taken over
1982. by another company working in a different sector,
manual employees may change sector even though
The manual employee category was put together they still have the same job in the same company. The
differently in the two classifications (Seys, 1984). In indicators presented in this article should therefore be
order to work on fields with the maximum comparability, used more to compare sectors than as intrinsic data.
only non-agricultural manual employees were included
with the exception of supervisors (who were in The aim of this study was not to track individual paths,
the “manual employee” category in the CSP but to identify the main features of sector workforce
classification, but excluded from it in the PCS management methods and how they developed from
classification). The two sub-groups do not completely 1975 to 1990. The manual employees tracked therefore
tally, but the differences are fairly minor in most of the change between 1975 and 1990. The changes over
sectors. these two inter-census periods of 1975-1982 and
1982-1990 were studied separately.
Four series were therefore used to measure the
thchanges in staff numbers: the 1975 census’ 1/5 The EDP was used to answer two questions:
thsampling and the 1982 census’ 1/20 sampling for
changes from 1975 to 1982, and the 1982 and 1990 What had become of the different aged manual
censuses’ ¼ sampling for changes from 1982 to 1990. employees working in a given sector (in 1975 or in 1982)
However, the only source used to analyse the change in by the following census?
age structures for 1982 was the ¼ sampling. This was
done once the age structures obtained from this source -Where did the manual employees working in a given
in the chosen sectors had been checked to make sure sector come from (in 1982 and 1990): what was their
ththat they tallied with those calculated from the 1/20 situation in the previous census?
sampling using the old classification.
Their fate and origin were analysed and grouped into
To have age data covering a large enough population, one of the following situations: manual employee in the
only those activity sectors employing at least 100,000 same sector; other socio-economic category in the
manual employees in 1975 were chosen. same manual employee in another sector; other
socio-economic category in another sector;
The demographic panel unemployed; out of the labour force, student or
conscript; and retired.
INSEE’s Demographic Panel (Echantillon
démographique permanent or EDP) comprises all civil Despite its large size, the sample does not contain
status records and census forms since 1968 for one enough individuals by detailed age and sector.
group of people. This sample accounts for Consequently, age brackets were studied (15-24 years
approximately 1% of the population, covering individuals old, 25-34 years old, 35-49 years old and 50 years old
born on four days of the year. and over).
4 INSEE Studies no. 43, April 2000Box 3
The study covers solely manual employees. This is “Corporate and activity sector histories create certain
because work-related selection effects, i.e. those characteristic ways of integrating the workforce into
elements of the working conditions involved in work. Over the course of their history, a system of rules
excluding the oldest workers, are probably best is established (with regard to recruitment conditions,
identified for the manual employee category, even at redundancy conditions, wage setting, the working
macrostatistical level. week, etc.), which may or may not be made legally
formal (labour law and labour agreements). Moreover,
specific constraints result from the type of manpowerThe decision to restrict the study to industrial sectors is
employed by the company,” “Eymard-Duvernay, 1981);due to the fact that the frontiers between the manual
see also Blosseville, Clemenceau and Grando (1982),employee category and the other categories are
Du Tertre (1989), and Podevin (1989 and 1990).probably more clear-cut in industry than in other sectors
(such as market services, trade, etc.) (Chenu, 1993).
Note:Changes in category from one period to the next
therefore risk reflecting much more a similarity between
As with all mobility studies, the rate of the flowscategories than a real change of job (Merllié, 1990).
depends on the breakdown: the smaller the sector, theMoreover, detailed analyses of in-company situations
greater the probability of leaving it ;are available for this type of employment and can be
used in the interpretation of the findings. Other
-Movements within the sector are ignored (change ofcategories are probably affected by this phenomenon,
company);but highly detailed groups (e.g. occupations or
sub-sectors) would have to be studied to identify them
-Movements are measured by comparing situations onand certain problems would probably be incurred when
two dates, but no information is available on anyusing the statistical sources.
changes between these two dates;
The study is made at sector level. It would probably
-Exits due to a change of category within the samehave been more interesting to make it at company level,
sector are generally ignored. However, a figure hasbut no data are available to do so. Hence we more often
been put to this phenomenon and may be referred to inthan not had to ignore the internal heterogeneity of the
certain observations without being systematicallysectors when studying the variables. However, specific
studied.logics can sometimes be identified at sector level.
Table 2
The percentage of young manual employees then old manual employees falls
Sectors Age structure of the manual employee population in industry in
1975, 1982 and 1990
Under 30 years old 45 years old and over
1975 1982 1990 1975 1982 1990
Meat and dairy 36 37 36 30 29 22
Other food industries 43 43 43 30 26 20
Ferrous metals 32 23 19 32 35 26
Building materials 31 27 26 34 37 31
Basic chemicals and man-made fibres 27 21 20 36 39 36
Smelting 40 36 34 28 28 22
Mechanical engineering 43 36 32 26 27 23
Electrical manufacture 47 36 29 24 25 22
Land vehicles 40 30 24 25 27 26
Shipbuilding, aircraft and armaments 36 29 20 31 32 30
Textiles and clothing 49 42 29 27 25 22
Leather and footwear 46 40 30 29 29 23
Wood and furniture 43 41 38 29 28 21
Paper and cardboard 37 28 28 32 33 27
Printing, press and publishing 40 32 33 28 27 25
Rubber and plastics 39 33 33 28 28 23
Building, civil engineering and agriculture 41 37 34 23 26 25
Reading: For a given year, the percentage of 30-44 year olds in each sector corresponds to the remainder out of 100 of the sum of the two
age brackets given in the table.
Source: Population censuses.
INSEE Studies no. 43, April 2000 5between socio-economic groups, more or less company [and who] still have a long future with
corresponds to what J. M. Smith (1973) called the company.” In this case, “the first component of
“hypothesis H (as in history)”. This is the case in an employee reduction strategy is to restrict entry
which “the structure of an activity is simply a fuzzy flows”. “The second element of an employee
image of the history of this activity […]. Under reduction strategy can but focus on the age profile
Hypothesis H, the age structures, calculated over by lowering the retirement age and stepping up
successive periods, are expected to show that the retirements” (Favereau et al., 1991).
peaks and troughs move over time.”
[…]“Hypothesis H assumes the occupational
The use of retirement options does notstability of the workforce or, in the event of a
change of employer, mobility within the same always prevent ageing
activity.” Smith compared this model with
“hypothesis A (as in age)”, i.e. the case in which In 1975, manual employees likely to retire before
“the age structure of an activity reflects the extent the next census (in 1982) were over 58 years old,
of difficulty that this activity represents for each since the retirement age was set at 65 years old at
age group. Hypothesis A assumes the existence of the time. The proportion of these “old” manual
an ongoing adjustment process and an employees was low over this period, at 4% to 8% of
occupationally mobile workforce” (Smith, 1993). average numbers depending on the sector. For the
Later, we will present situations more or less in line purpose of comparison with the following period
with this scenario. and because practices often precede the law, it was
assumed that manual employees aged 53 and over
Smith’s hypothesis “H” is also similar to the way in in 1975 could have retired at 60 years old during
which other researchers define the effects of the 1975-1982 period. The proportion of
internal labour markets whose “operating logic potentially “departing” manual employees under
[…] favours workers in the middle age bracket this assumption thus topped 15% in the chemical,
[who] have tried and tested skills and loyalty to the building materials, paper-cardboard, and printing
Table 3
Potential departures of old manual employees and potential replacement by young people from
1975 to 1982 and from 1982 to 1990
Sectors Manual employees aged ...
Manual employees likely to retire Estimate of entries
26 and
58 and over 53 and over 52 and over 25 and under
under in1 2 3 4in 1975 in 1975 in 1982 in 1982 31990
Meat and dairy 5.0 11.6 14.4 22.8 19.3
Other foods 6.9 14.6 14.1 27.3 28.3
Iron and steel 4.8 14.0 18.0 7.1 6.4
Building materials 6.9 16.0 20.8 11.3 12.6
Basic chemicals and man-made fibres 6.9 17.4 21.4 7.2 8.3
Smelting 6.1 13.4 14.3 17.0 19.9
Mechanical engineering 5.5 12.4 15.5 16.5 16.9
Electrical manufacture 4.7 10.8 13.8 14.4 15.4
Land vehicles 4.3 10.5 13.1 12.0 12.9
Shipbuilding, aircraft and armaments 5.0 14.4 18.7 12.3 8.4
Textiles and clothing 8.1 15.9 15.1 20.0 13.2
Leather and footwear 7.3 15.7 18.4 19.6 13.0
Wood and furniture 7.1 14.2 16.0 23.3 9.5
Paper and cardboard 7.5 16.4 16.7 11.2 14.8
Printing, press and publishing 7.2 15.2 12.8 15.2 19.9
Rubber and plastics 5.6 12.5 14.4 15.8 19.4
Building, civil engineering and agriculture 4.6 9.7 12.8 17.0 19.9
1.Under the assumption of retirement at 65 years old by 1982; (% of half the 1975+1982 staff)
2. the of retirement before 60 years old in 1982; (% of half the
3.% of half the 1982+1990 sector staff
4.% of half the 1975+1982 staff
Reading: To evaluate the possibilities of old employees being replaced by young employees, the “school hypothesis” is made that departu-
res are due solely to retirement at the legal retirement age and that entries are estimated by the number of under-25s present at the end of
the period.
Source: Population censuses.
INSEE Studies no. 43, April 2000 6sectors and also in textiles and clothing and leather finding an organisation and ways of working
and footwear, which are in other respects very that reconcile the production efficiency,
young sectors (see table 3). continued health and skills development of
employees over 40 years old.
For the 1982-1990 period, the “natural” ageing of
the in-service population created an increase in the
Diverse sector strategies to replace (or not)number of workers likely to retire at 60 years old
(the legal age for this period). Over 20% of manual old employees
employees were concerned in sectors such as
Retirement is at least partially offset bychemicals and building materials, and 17% to 20%
recruitment. This is measured by making thewere concerned in the shipbuilding, aircraft and
assumption that recruitment is entirely made up ofarmaments sectors, the leather and footwear
young people. To be more precise, it is estimatedindustry and the ferrous metals and primary steel
by the workforce of young people that could haveprocessing industries (see table 3).Even without
been recruited as of 17 years old (and have not leftearly retirement measures, retirement at 60 years
since) during the inter-census period studied, i.e.old alone therefore concerned a much higher
the number of under-25s in 1982 for the first periodpercentage of manual employees in this period than
and the number of under-26s in 1990 for thein the previous period. In certain sectors, early
1982-1990 period (the second period contains anretirement measures thus had an extremely large
additional year). This workforce is actually aquantitative impact and played a key role in
balance of entries-exits at these ages. Nevertheless,managing decreases in worker numbers, such as in
it is worth noting that the development ofthe iron and steel industry (see below). It is worth
temporary work means that a certain proportion ofnoting that massive exits by employees in the
employment, of young people in particular,“middle” age bracket could have disrupted the
escapes the sector analyses. The sector consideredrunning of work activities, with possible losses of
is the legal employer and temporary agencies areknow-how, the destructuring of work units and so
classed as “market services provided mainly toon.
businesses”. The actual user sector is not known.
When the previous exercise is extended past
In sectors with large workforce cuts, the level of1990, “in-service” ageing alone through to 1998
recruitment (or, more specifically, the balances ofof manual employees in service in 1990 results
entries and exits) of young manual employees wasin potential departures at the legal retirement
much lower than the level of potential retirement,age of 60 of 19% of manual employees in
even under the assumption of retirement at 60 yearsthechemicalssector, 17% of those in building
old as of 1975. This practice of closing off entrymaterials and 14% of manual employees in
was particularly marked in iron and steel, buildingshipbuilding, aircraft and armaments and the
materials, chemicals, and shipbuilding, aircraft andpaper-cardboard sector. Bringing the retirement
armaments during the entire 1975-1990 period.age down to 55 years old means that over 30%
This was also the case for paper-cardboard, theof the manual employees working in the
leather and footwear industries and the wood andchemicals sector in 1990 would be affected,
furniture industries from 1975 to 1982. Givenone-quarter or more of those in building
equivalent workforce variations, this practice wasmaterials and shipbuilding, aircraft and
even more marked from 1982 to 1990 than in thearmaments and 23% of the manual employees in
previous period.paper-cardboard. These sectors are made up of
boom generations starting to near the end of
Nevertheless, despite a drop in their workforce,their working lives. This opens up new
certain sectors maintained a larger proportion ofpossibilities for an infusion of youth and a
young people from 1975 to 1982 than would haverenewal of staff, even if the departures do not
been possible solely by replacements of “potential”take place before the retirement age. However,
retirements at 60 years old. Such was the case withwhen the age pyramid is more concentrated
the textiles and clothing industry, the leather andaround the 40-50 year bracket, which is
footwear industry (with decreases of over 25% infrequently the case, the oldest generations are
manual employees), and building and smeltingstill relatively small and potential departures
(with decreases of 15% to 20%). All of these(even as early retirement) provide few
sectors were relatively young in 1975. It is as if,opportunities for replacement recruitment. In
despite the drastic reductions in their manualaddition to the temptation to remodel the age
employees, these sectors tried to maintain a fairlypyramid, the companies in these sectors risk
young age structure or at least curb the effect offinding themselves faced with the problem of
7 INSEE Studies no. 43, April 2000workforce cuts on the entry of young people For example, the level of change in worker
(Smith’s “Hypothesis A”?). There were actually numbers cannot be compared with the level of the
more departures around the age of thirty, with the balance of entry and exit movements for the same
numbers of manual employees dropping less in the period. However, thedifferences between sources
youngest age brackets. One explanation of this are less marked as regards comparisons between
phenomenon could therefore be the number of industrial sectors.
business closures in these relatively
unconcentrated sectors. Businesses that close often Over 40% of the manual employees working in a
have older staff than business start-ups or given sector in 1975 were no longer manual
businesses in a position to recruit. However, no employees in the sector in 1982. Similarly, half to
trace of such strategies is found in the subsequent three-quarters of the manual employees in a given
period. sector in 1982 were no longer there in 1990.
Conversely, one-quarter or more of the manual
employees working in a given sector in 1982 were
Substantial renewal of the manual not there seven years previously, and 40% to
employee population two-thirds of those employed in 1990 had become
manual employees in the sector since 1982 (see
Our comparisons have hitherto focused solely on tables 4 and 5).
the general situations found by the censuses. The
sectors have been implicitly considered as These flows of entries and exits increased sharply
companies with a strong internal market, whose from 1982 to 1990. In terms of average annual
only movements comprise the entries of young rates, the percentage of manual employees leaving
people and the exits of old people (“Hypothesis a sector or a category varied from 8% to 11%
H”). The data contained in the Demographic Panel depending on the sector for the 1975-1982 period.
(EDP), which tracks a panel of individuals over a This percentage topped 10% in only four sectors:
number of censuses, can be used to evaluate the smelting and metalworking, mechanical
entry and exit flows and hence assess the extent of engineering, textiles and clothing, and the leather
manual employee population renewal in the and footwear industry. For the 1982-1990 period,
different sectors. However, differences between this percentage was less than 10% in only two
the two sources (see box 2) mean that sectors: motor vehicle manufacture and
EDP-estimated “movements” cannot be directly paper-cardboard. An annual average of 14% to
related to the “stocks” evaluated by the censuses. 15% of manual employees left the sector or manual
Table 4
Exits from the sector (or category)
For manual employees present at the beginning of the period, from 1975 to 1982 and from 1982 to
Sectors 15-49 year old “exit” rates (age“Exit” rates all ages together
at the beginning of the period)
1975-1982 1982-1990 1975-1982 1982-1990
Meat and dairy 44 57 41 50
Other food industries 61 47 5450
Iron and steel 4451 72 66
Building materials 48 64 42 56
Basic chemicals and man-made fibres 46 4073 67
Smelting and metal working 67 48 6253
Mechanical engineering 69 6352 49
Electrical and electronics manufacture 60 48 5451
Land vehicles 42 52 36 44
Shipbuilding, aircraft and armaments 46 64 39 55
Textiles and clothing 59 5255 52
Leather and footwear 59 5154 50
Wood, furniture and miscellaneous industries 52 70 49 65
Paper and cardboard 47 51 42 42
Printing, press and publishing 45 54 41 48
Rubber and plastics 47 57 42 48
Building, civil engineering and agriculture 49 55 47 50
Source: Demographic panel.
INSEE Studies no. 43, April 2000 8employee category in iron and steel, chemicals, the extent of sector entry and exit movements.
mechanical engineering, and the wood and Exits should be more frequent where the decrease
furniture industry. From 1975 to 1982, the annual in employee numbers is sharper and more entries
entry flows represented 4% to 6% of the numbers should be found where growth in numbers is more
of manual employees counted at the end of the positive. In fact, the sectors with a particularly high
period. From 1982 to 1990, these flows varied from proportion of manual employee departures were
7% to 13%. generally sectors that drastically cut their numbers
of manual employees. Conversely, the lowest exit
The volume of exits could be thought to depend in rates concerned more sectors whose numbers of
part on the number of age brackets affected by manual employees decreased little.
retirement. However, an analysis restricted to
manual employees aged 15 to 49 (age at the Yet the number of exits was always very high, even
beginning of the period) reveals relatively high in sectors with relatively favourable growth in
proportions of exits: over 40% in most of the numbers. In sectors where numbers dropped only
sectors from 1975-1982 and 50% from 1982 to slightly (less than 10%) or where they rose, the
1990. This increase is partially due to the percentage of manual employee exits varied from
inter-census period’s being extended by one year 44% to 52% from 1975 to 1982 and was around
(see tables 4 and 5). The effect of early retirement 50% to 70% for the 1982-1990 period. Conversely,
on the oldest workers was probably greater during the weight of entries was far from negligible in
this second period than from 1975 to 1982 due to certain sectors where numbers fell a great deal,
the spread of early retirement measures. The such as in mechanical engineering.
sectors with the sharpest increases in average
annual exit rates for manual employees aged 15 to Lastly, highly disperse exit rates were found in
49 widely applied such measures to relatively sectors with the same extents of growth in
young workers. In iron and steel (where employees employee numbers. This dispersion was even
were retired as early as 50 years old) and in more marked from 1982 to 1990 than in the
chemicals, the proportion of manual employees previous inter-census period (see charts I and II).
leaving each year rose from 7% to 8% in the This observation ties in with Podevin’s finding
1975-1982 period to 13% in the 1982-1990 period. (1990) that, “no significant liaison is found
between these variations in employee numbers
More importantly, a close relationship could be and the intensity of labour turnover” (p. 48),
expected to be found between the growth in the confirming the importance of employment
number of manual employees over the period and management methods.
Table 5
Entries into the sector (or category)
For manual employees present at the end of the period, from 1975 to 1982 and from 1982 to 1990
Sectors Entry rate for 35 year olds and over (age
Entry rate All ages together
at end of the period)
1975-1982 1982-1990 1975-1982 1982-1990
Meat and dairy 38 59 23 44
Other food industries 35 55 27 40
Iron and steel 24 50 14 40
Building materials 33 59 20 47
Basic chemicals and man-made fibres 30 53 17 40
Smelting and metalworking 34 2266 56
Mechanical engineering 37 68 25 58
Electrical and electronics manufacture 34 54 18 41
Land vehicles 32 43 17 28
Shipbuilding, aircraft and armaments 31 53 18 47
Textiles and clothing 29 41 21 30
Leather and footwear 28 45 19 35
Wood, furniture and miscellaneous industries 38 60 26 49
Paper and cardboard 27 53 14 40
Printing, press and publishing 30 51 19 35
Rubber and plastics 36 67 20 53
Building, civil engineering and agriculture 29 48 18 35
Source: Demographic panel.
9 INSEE Studies no. 43, April 2000signify stability on the internal markets, but ratherHigh labour turnover sectors and stable
high intrasectoral mobility in a particularly largesectors
Certain sectors have a high manual employee
Manual employee turnover is especially highturnover. Movements, in terms of both entries and
among the youngest manual employees andexits, are much more frequent in these sectors than
subsequently decreases with age.in others with the same level of growth in
employee numbers. This relatively high manual
Exit rates in the 35-to-49-year-old age bracketlabour turnover is observed as much where
were particularly marked for the youngest ageemployee numbers dropped considerably (at least
brackets in the meat and dairy industry, in rubberover one of the two periods studied) such as in
and plastics, and even in the leather and footwearmechanical engineering, smelting and
industry (solely from 1975 to 1982 for this lattermetalworking, and the wood and furniture
sector). They were nearly twice as high among 15industries, as in sectors with a more favourable
to 24 year olds. However, relatively littlegrowth in employee numbers, albeit accompanied
difference was found in iron and steel, buildingby a high turnover: food industries, and the rubber
and, to a lesser extent mechanical engineering.and plastics industries.
Other sectors appear to be relatively more stable
Age management and labour constraintsthan the above: motor vehicle manufacture,
in certain sectorsshipbuilding, aircraft and armaments,
paper-cardboard, printing-press-publishing, and
A more detailed analysis of some contrastingeven building-civil engineering, and agriculture.
sectors can illustrate how these age managementThere are probably various explanations for this
methods tie in with changes in working conditionsrelative stability. In the first cases, the sectors (or
and job engineering. The age structures in the lategroups of sub-sectors) are fairly concentrated with
1970s helped address certain labour constraints byrelatively large internal corporate markets. This
sharing them out differently by age. These forms ofmay also be the case for part of
“age selection” are now in question today.printing-press-publishing. However, the internal
heterogeneity of this activity is such that other
The numbers of manual employees in iron andhypotheses could be envisaged, such as a fairly
steel dwindled over the period, falling one-thirdhigh internal sector mobility on the professional
from 1975 to 1982 and then 50% from 1982 tomarkets (Eyraud et al., 1990). In the case of
1990. The age structure in 1975 was bimodal with abuilding-civil engineering and agriculture, with
“peak” around 24-25 years old and a “bump” fromtheir large weight of small enterprises, the low rate
40 to 47 years old (see chart III).of sector entry and exit flows most likely does not
Chart I Chart II
Proportion of manual workers “exiting” the sector or The proportion of manual employees “exiting” the
the category and proportion of manual employees sector or category and the proportion of manual
“entering” from 1975 to 1982 employees “entering” from 1982 to 1990
% “exits” % “exits”
60% 80%
Growth in manual employee numbers from 1982 to 1990:
. Stability or increase (from 0% to 4%): sector name in italics
. from -2 % to -10 %: sector name in normal type
. from -13 % to -20 %: sector name in bold
Textiles and clothing
55% 75% . from -25 % to -50 %: bold underlined
Leather and footwear
Smelting Mechanical engineering ChemicalsFerrous metals
Wood anf furnitureFerrous metals
Electrical manufacture
Other food industries 70%50% Wood and furnitureBuilding and public works
Mechanical engineering
Building materials
Rubber and plasticsPaper and cardboard Smelting
Chemicals Shipbuilding, aircraft and armaments
Printing, press and publishing45% 65%
Meat and dairy Shipbuilding, aircraft and armaments Building materials
Motor vehicle manufacture
Other food industries
Electrical manufacture
40% 60%
Leather and footwear
Textiles and clothing Rubber and plastics
Growth in manual employee numbers from 1975 to 1982: Meat and dairy
. Increase: sector name in italics
Building and public works. from - 5% to -11%: sector name in normal type 55%35% Printing, press and publishing. from -15% to -24%: sector name in bold
. from -27% to -35%: bold underlined
Motor vehicle manufacture
Paper and cardboard
30% 50%
20% 25% 30% 35% 40% 45% 50% 40% 45% 50% 55% 60% 65% 70%
Source: Demographic panel. % “entries” Source: Demographic panel. % “entries”
10 INSEE Studies no. 43, April 2000