Population (english edition)
I.N.E.D
I.S.B.N.sans
200 pages
p. 93 à 118
doi: en cours
Veille sur la revue
Veille sur l'auteur
I.N.E.D
I.S.B.N.sans
200 pages
p. 93 à 118
doi: en cours
Veille sur la revue
Veille sur l'auteur
Vous consultez
Volume 60 2005/1-2
Working Conditions: How are Older Workers Protected in France?
Ariane Pailhé [*] Ariane Pailhé, Institut national d’études démographiques, 133 boulevard Davout, 75980 Paris cedex 20, France, Tel: 33 (0)1 56 06 22 74, Fax: 33 (0)1 56 06 21 99
Since the 1980s, working conditions in France have been deteriorating, notably as a result of organizational changes in the industrial and business environment. Older employees make up a group that is particularly vulnerable to this sort of transformation. We will therefore examine to what extent protection designed for these workers in the 1970s still remains in place. Are older workers less exposed to occupational strain? If so, does this result from protection or selection? Cross-sectional surveys of working conditions conducted by DARES and INSEE in 1984, 1991 and 1998 have been used to answer these questions. Data analysis brings to light three groups of work constraints: physical strain, working time flexibility and commercial demands. All things being equal elsewhere, we will see that age protects workers against physical strain, particularly women. Men, for their part, become less exposed to commercial demands as they grow older. But although protection mechanisms designed for older people continue to exist, the relative protection they afford has decreased over time. Furthermore, beyond the age of 55, protection against physical strain often takes the form of exclusion from the labour market.
Depuis les années 1980, on assiste à une dégradation des conditions de travail en France, notamment en raison des changements organisationnels mis en place dans les entreprises. Face à ces transformations, les salariés âgés forment un groupe particulièrement vulnérable ; c’est pourquoi nous étudions en quoi les protections de ces salariés qui prévalaient dans les années 1970 perdurent. Les salariés vieillissants sont-ils moins exposés que les autres aux pénibilités professionnelles ? Si oui, est-ce le fruit d’une protection ou d’une sélection ? Pour répondre à ces questions, nous utilisons les enquêtes transversales sur les conditions de travail effectuées par la Dares et l’Insee en 1984, 1991 et 1998. Une analyse de données permet de distinguer trois groupes de contraintes de travail : les pénibilités physiques, la flexibilité horaire et les exigences marchandes. Toutes choses égales par ailleurs, on montre que l’âge protège les salariés des pénibilités physiques, surtout parmi les femmes ; pour les hommes, l’avancée en âge les rend moins exposés aux exigences marchandes. Mais si les mécanismes de protection des salariés âgés perdurent, leur protection relative a diminué au fil du temps. En outre, au-delà de 55 ans, la protection face aux exigences physiques s’exerce souvent par le biais d’évictions du marché du travail.
En Francia, las condiciones de trabajo se han deteriorado desde los años ochenta; tal deterioro es debido, principalmente, a los cambios organizativos llevados a cabo por las empresas. Frente a tales transformaciones, los trabajadores de más edad constituyen un grupo especialmente vulnerable. Por esta razón estudiamos qué sistemas de protección, entre los que estaban disponibles durante los años setenta, permanecen disponibles en la actualidad. ¿Es posible que los trabajadores mayores estén menos expuestos a las penalidades profesionales que los trabajadores más jóvenes? Y si lo están, ¿es ello debido a una mayor protección o a un efecto de selección? Para responder a estas preguntas, utilizamos las encuestas transversales sobre las condiciones de trabajo llevadas a cabo por la Dares y el INSEE en 1984, 1991 y 1998. Un análisis de los datos de estas encuestas permite distinguir tres tipos de penalidades en el trabajo: las penalidades físicas, la flexibilidad horaria y las exigencias del mercado. El artículo muestra que, en igualdad de condiciones, la edad protege a los trabajadores de las penalidades físicas, especialmente entre las mujeres; entre los hombres, la exposición a las exigencias del mercado disminuye con la edad. Sin embargo, aunque los mecanismos de protección de los trabajadores mayores siguen vigentes, su protección relativa ha disminuido. Por otra parte, para los mayores de 55 años, la protección frente a las exigencias físicas del empleo se ejerce a menudo a través de la evicción del mercado de trabajo.
Since the 1980s, changes in the organization of production and distribution have led to a general deterioration of working conditions, though employees sometimes enjoy greater autonomy in the management of their work. In this context, are older workers better protected than their younger counterparts, be it in terms of working time flexibility, physical strain or commercial constraints such as the pace of work dictated by demand, contact with the public or other sources of tension?
These questions are difficult to answer, since results may be distorted by selection bias, notably due to exclusion of older workers from the labour market (early retirement, unemployment etc.). Using survey data on working conditions in 1984, 1991 and 1998, Ariane Pailhé attempts to overcome this difficulty.
Following the recent reform of the pension system, and the lengthening of the contributory period, studies of this kind are of crucial importance.
In France, labour force participation rates for people over 55 are low: 62% for the 55-59 age group and 15% for the 60-64 age group according to the latest population census in 1999. The employment rate, which measures the proportion of persons in employment among the population of working age, is even lower (respectively 54% and 13%). At the same time, the proportion of under-25s is falling — the size of these birth cohorts has decreased and the time spent in schooling has lengthened — while the share of 40-55 year-olds is increasing, leading to a rise in the average age of persons in employment (Molinié, 1998).
Aiming to address the problems of pension system financing and increasing dependency rates, the pension reform of 21 August 2003
[1] is founded on a lengthening of the contributory period and hence the prolongation of working careers. This new approach, initiated by the act of 1993
[2], goes against practices set in place during previous decades, when institutional measures such as the lowering of normal retirement age, the development of early retirement, dispensation from job seeking and non-degressive benefits for unemployed people over 55 were introduced as a means to lower the number of surplus workers quickly without generating industrial unrest. This approach also runs counter to the practices of employers, who are often unwilling to hire workers over the age of 50 (Jolivet, 1999; Le Minez, 1995). Furthermore, it goes against the aspirations of many older workers who would like to retire earlier. The question of working conditions is often central to their demands, as illustrated by the development of labour disputes within certain professions (Belfer, 2004) and their role in the protests against pension reforms in the spring of 2003.
Poor working conditions damage health and lead to differential ageing (Teiger, 1989). Working history and, more specifically, working conditions influence life expectancy (Hayward et al., 1989, Moore et al., 1990, Desplanques et al., 1996), disability-free life expectancy (Cambois et al., 2001), and, more generally, health in old age (Cassou et al., 1994). Moreover, studies by epidemiologists and ergonomists indicate that certain work constraints are particularly detrimental for older workers (Krause et al., 2000; Laville, 1989).
For this reason, selection mechanisms traditionally provided a means to make adjustments between age and working conditions in the workplace through a process of worker reassignment, but also through exclusion from the workforce (Volkoff, Molinié, 1995; Molinié, 1998). Hence, during the post-war boom years, protection policies for older workers were prevalent, particularly in large companies, with personnel management based on tenure, protection against redundancy for the oldest workers, and allocation of less strenuous tasks to workers approaching retirement age. The labour market was segmented (Doeringer, Piore, 1971; Piore, 1978; Cases et al., 2001) and workers with long tenure were relatively well protected.
Starting in the 1980s, and even more so in the 1990s, faced with an increasingly unstable environment and exposure to international competition, companies embarked on a quest for productivity gains and a reduction of overall
[3] costs (Askenazy et al., 2002). With regard to employment, they reacted by raising productivity and searching for both quantitative and qualitative flexibility (diversification of working hours, versatility, development of atypical forms of employment, outsourcing of activities). At the same time, computers and new information technologies revolutionized work management (Gollac et al., 2000). Overall, these organizational changes made work more interesting in many cases, but at the cost of greater pressure and a deterioration of working conditions (Bué et al., 1999; Valeyre, 2001, Green et al., 2001). They also upset the vertical employment structure specific to Taylor’s model (Coutrot, 2000).
Against this background, our study will focus on age-related selection mechanisms. Does the deterioration of working conditions have a uniform effect on all age groups? Are older workers still protected? In other words, do older workers still belong to a protected segment of the labour market? To answer these questions, we use data from surveys on working conditions conducted by DARES and INSEE in 1984, 1991 and 1998. Although the descriptive analysis shows that older workers are less exposed to certain types of strain than younger ones, they are not necessarily better protected. Some of them have been excluded from the labour market, notably for health reasons linked to these working conditions. They are replaced by younger people and are therefore under-represented in the most physically strenuous positions. These selection effects need to be taken into account, as will be the case for our multivariate model estimations. To this end, we will use the most recent data, provided by the 1998 survey.
I. Changes in working conditions since the 1980s
1. The data
The data used are those from the second, third and fourth Working Conditions Surveys carried out in March 1984, 1991 and 1998 by DARES and INSEE in association with the Employment Survey
[4]. They were administered to all employed persons in one out of three households surveyed in the Employment Survey, i.e. about 22,000 people in 1998, who were each required to answer individually. The 1998 survey was the first to use the new computer-assisted data collection system (CAPI). The scope of these surveys is that of the employment surveys. It includes all ordinary households in metropolitan France, and a part of the non-household population insofar as the persons concerned have family ties with ordinary households. However, certain labourers working on temporary sites, young people and foreigners living in hostels were excluded, as were people living in residential establishments such as hospitals, schools or hotels. For this study, we limited the sample to wage earners aged between 16 and 64 (18,800 in 1998).
The themes studied were nuisance and risk (physical, biological, chemical) in the workplace, pace of work, working hours and organization, the degree of employee autonomy and initiative, employee communication networks and responsibilities, and finally, the use of modern technologies, particularly computers and their conditions of use. The questionnaire originally focused on the physical demands of work, but questions concerning psychological aspects or “mental demands” were added over time.
For the survey, the employees were asked to say whether they were subjected to a list of constraints by giving answers of the type yes/no, always/often/seldom/never, or by indicating the number of incidents or the intensity of the constraint. The data from the Working Conditions Survey are not therefore based on objective measurements. For this reason, other objective questions were added to the survey (exposure to noise
[5], working hours, working methods and organization, etc.) to supplement the subjective questions concerning employees’ perception of their working conditions (questions concerning physical effort, temperature, exposure to risk, use of memory, risk of error and cognitive, psychological or relational aspects of work). The questions concerned the employees’ main employment for a given period: a given day or the last week worked.
Due to the data collection method used, it is difficult to measure the changes in working conditions between the various surveys as they may result either from a change in perceptions or a real deterioration (or improvement). The appreciation of working conditions is closely linked to the fact that risks are “objectified”, i.e. made explicit, once they are defined as risks, for example, following industrial unrest
[6] or work accidents. Once individuals have learnt to assess their working conditions more precisely, certain difficulties which were previously considered “natural” are subsequently reported as risks
[7]. Faced with the deterioration in working conditions between 1984 and 1991, particularly regarding strain and risk, Cézard, Dussert, Gollac (1999) and Gollac (1994) analysed the various sources of this deterioration. They showed that changes in the survey protocol played a minor role compared with changes in perception. However, this “objectivation” of the situation was not enough to explain the measured changes: the deterioration in working conditions was, at least in part, real.
Surveys of working conditions thus make it possible to pinpoint tendencies. In our work, uncertainty associated with the data collection method can be limited by overlapping a few simple questions. The survey’s subjectivity should be taken into account when interpreting the results. Note also that the subjective experience of working conditions should be considered in its own right, insofar as the perceived strain of work exposes workers to increased risk of pathologies associated with psychological and physical fatigue (Lasfargues, 2004).
2. A general deterioration of working conditions
While exposure to physical strain stabilized during the 1990s, time pressures increased, marked by the expansion of commercial constraints in industry and of industrial constraints in the service sector (Gollac et al., 2001).
Towards a stabilization of physical strain
Following a sharp increase in nuisance and strain between 1984 and 1991, the number of workers doing physically strenuous jobs stabilized between 1991 and 1998 (Cézard and Hamon-Cholet, 1999a). For example, the proportion of workers who reported having to remain standing for long periods increased from 49% in 1984 to 53% in 1991 then to 54% in 1998, while for physical strain it rose from 68% to 82% then to 84%
[8]. The proportion of workers exposed to at least one risk factor (risk of burns, a serious fall, inhalation of fumes, dust, toxic chemicals, etc.) rose from 52% in 1984 to 68% in 1991 then to 70% in 1998.
Though exposure to physical strain stabilized during the 1990s, workers were exposed to a larger number of constraints: between 1991 and 1998, they were exposed to one additional physical strain (out of a total of 10) and 0.5 additional risk (out of a total of 14) on average. This increase can be partly explained by the continuing objectivation of working conditions, but also by work intensification resulting from the quest for greater productivity, shorter production lead times and growing competition (Aquain et al., 1994). Working under pressure reduces workers’ room for manœuvre, and this disrupts the strategies they use to manage physical effort and strain. Though a job may, intrinsically, be less physically strenuous, the faster pace of work makes it feel more tiring.
A faster pace of work
Compared with 1984 and 1991, constraints on the pace of work have increased. The pace of work is now more often imposed by industrial constraints (11% in 1991, 16% in 1998), automatic movement of a product or a piece (6% in 1984 and 1998, 4% in 1991), quantified production standards (18% in 1984, 16% in 1991, 29% in 1998), or controls and monitoring (23% in 1991, 29% in 1998). Deadlines are becoming increasingly short: 60% of industrial employees have task completion deadlines of less than one day. In 1998, 52% of employees were obliged to hurry, and a greater proportion of employees considered that the allotted time was not long enough “to do their job properly” (22 in 1991 and 24% in 1998).
More flexible working hours
During the 1990s — before the introduction of the 35-hour week — working time flexibility evolved less than in the 1980s (Bué and Rougerie, 1999b). Yet there were more employees working regularly or occasionally on Sundays (21% in 1991, 27% in 1998), on Saturdays (respectively 48% and 50%), or on night shifts (13% and 15%). The proportion of employees working under short-term (4% in 1991, 7% in 1998), temporary (respectively 1% and 2%) or part-time contracts (12% and 18%) also increased. A non negligible proportion of employees whose working hours were set by their employer had a high level of uncertainty concerning their work schedules: in 1998, 13% did not know their schedule for the coming month, 7% for the coming week and 3% for the following day. Employees were more often expected to switch to a new post to suit the employer’s needs (28% in 1991, 30% in 1998), and to stop what they were doing to work on a different, unscheduled task (respectively 48% and 56%). These interruptions were considered disruptive in 48% of cases. Even though job rotation was the source of errors in 21% of cases, it was judged beneficial more often than detrimental with regard to workload (29% in 1991, 32% in 1998) and especially with regard to job appeal (respectively 12% and 67%).
Greater job autonomy
Scope for initiative grew between 1991 and 1998
[9] and a smaller proportion of employees reported having to follow orders to the letter: it fell from 42% to 38% over the period. Similarly, employees have fewer deadlines (72% in 1991, 69% in 1998) and more often manage incidents themselves (49% in 1991, 56% in 1998). Tasks are also less strictly formalized: 89% in 1998 had a choice of work methods, compared with 82% in 1991.
Growing tensions
While this increased freedom gives work greater meaning, it is also a source of stress and tension (Cézard and Hamon-Cholet, 1999b)
[10]. In 1998, employees are often expected to work things out for themselves in difficult situations (often 24%, sometimes 56%) and tension is on the rise, particularly for those in contact with the public (22% in 1991, 29% in 1998), tension is also relatively high in relations with colleagues and with superiors (30% in 1998).
3. A decrease in constraints with age
Less frequent work constraints for the oldest employees
The ergonomic literature underlines the fact that physical demands are particularly penalizing for the oldest, as they affect workers who are already physically worn out. Furthermore it becomes increasingly difficult with age to withstand several types of constraint at the same time (Molinié, 2001). Preserving the health of the oldest workers thus depends to a great extent on protecting them from these strains.
Older workers are slightly less exposed to strains and risks than their younger counterparts. For example, the proportion of workers exposed to physical constraints decreases with age (Figure 1). Yet, the proportion of seniors affected by these constraints remains high, falling barely 2-3 points below average, so reassignment to other jobs appears to be uncommon. Postural constraints seem to be felt even more strongly by the oldest workers. For example, from 1991, workers over 55 more often reported working in difficult or painful postures for long periods (Figure 2).
Figure 1
Proportion of employees reporting exposure to physical constraints (%)
Source: DARES/INSEE, Working Conditions surveys.
Figure 2
Proportion of employees who reported working in difficult or painful postures (%)
Source: DARES/INSEE, Working Conditions surveys.
Overall, this relatively lower exposure to strain and risk combined with the benefit of experience makes them less vulnerable to workplace accidents. They concerned 6% of older workers, compared with 8% on average, in 1998.
Constraints associated with the pace of work (production line work, pace of work dictated by quantified production standards, automatic movement of a product or part, or automatic machine operating speed) is slightly lower for older workers: in 1998, 38% of 55-60 year-olds were exposed, compared with 45% on average
[11]. The time pressures of work lessen slightly with age: the proportion of workers “often obliged to hurry” is slightly lower for workers over 55 (49% in 1998 compared with 52% on average) and slightly fewer of them report not having enough time “to do their job properly” (22% compared with 24% on average in 1998).
With increasing age, time constraints become less strict and flexibility requirements are more limited. So the freedom to choose work schedules is greater: the over-50s, for example, are more often entitled to set their own work schedules. In 1998, 18% of 50-54 year-olds and 21% of over-55’s set their own work schedules compared with 14% on average. Likewise, uncertainty regarding work schedules decreases: imposed variations in work schedules or working days become slightly more limited with age. While in 1998, the over-50s had slightly longer working hours, the over-55s declared the least amount of over-time. So working hours seem to be slightly longer for the oldest workers, but more stable and less subject to activity peaks.
Similarly, older workers are less exposed to atypical working hours: the proportion of employees who work on Sundays or Saturdays falls with age. Furthermore, the proportion of employees with alternating hours is lower for the over-55s: 6% compared with 10% on average in 1998. Beyond the age of 50, employees are less likely to work night shifts, though “occasional” night work falls off faster than “regular” night work.
Lastly, seniors show greater autonomy in their work, particularly due to greater work experience: 29% of over-55s do not receive orders (20% on average) and their superiors are more likely to give them a target (89% of over-50s, compared with 86% on average) rather than impose a defined way of doing a job. Although they appear to enjoy greater freedom, older employees rely less on teamwork: they more often report that they are expected to work things out for themselves in difficult situations (28% compared with 24% on average) and deal with problems by themselves (63% of over-50s compared with 54% on average). Likewise, fewer of them claim to be able to receive help from a superior (46% vs. 59% on average), from a colleague (61% vs. 72% on average), from another person (34% vs. 41% on average). Lack of cooperation is often cited as a factor that prevents them from doing their job properly (18% of over-50s, compared with 14% on average).
This increased autonomy does not appear to be accompanied by increased mental strain. There is less fear of making mistakes than among younger workers: 56% of over-55’s believe that an error on their part could affect product quality, compared with 65% on average, 42% believe that it could lead to financial loss for the company (50% on average), 31% that it could cause safety risks (38% on average). Likewise, fewer of them fear sanctions if they make a mistake (49% vs. 60% on average). Lastly, tensions felt both in relations with the public (Figure 3) and with colleagues or superiors are less common among seniors. Among those in contact with the public, 43% experience tense situations compared with 48% on average.
Figure 3
Proportion of employees reporting exposure to tensions in their contacts with the public (%)
Source: DARES/INSEE, Working Conditions surveys.
An increase in constraints over time for all ages
Constraints have increased for all age groups since the 1980s. Figures 4 and 5 show the evolution of physical strain and occupational risks for different five-year cohorts. The average level of reported physical strain
[12] and occupational risk
[13] increases from one cohort to the next: at each of the three survey dates, the youngest cohorts slightly more frequently reported exposure to strain and risk than the older cohorts. Furthermore, for a given group of cohorts, strain and occupational risk increased over time. The only cohorts whose exposure decreased were those over 60 in 1998, though there may be strong selection bias due to retirement of workers in the most exposed occupations. Workers over 50 in 1998 are thus more exposed than their counterparts were ten years earlier.
Figure 4
Average number of physical strain constraints per cohort
Reading: The average number of physical strain constraints or occupational risks can be followed between 1984 and 1998 for each five-year cohort. To make the graph more legible we have only represented every other cohort. Some cohorts are represented by only one or two points because they are either too old or too young to have participated in the labour market during the unmarked years.
Source: DARES/INSEE, Working Conditions surveys.
Figure 5
Average number of occupational risks per cohort
Reading: The average number of physical strain constraints or occupational risks can be followed between 1984 and 1998 for each five-year cohort. To make the graph more legible we have only represented every other cohort. Some cohorts are represented by only one or two points because they are either too old or too young to have participated in the labour market during the unmarked years.
Source: DARES/INSEE, Working Conditions surveys.
Although, for each period, the proportion of people affected by the various pace of work constraints decreases from the youngest cohorts to the oldest (Figures 6 to 8), work intensification occurs at all ages. Only for the 1929-33 and 1924-28 cohorts, who were over 57 in 1991, did the pace of work decrease in the 1980s. For all other cohorts, work intensity increased during the 1984-1998 period.
Figure 6
Pace of work imposed by technical constraints per cohort (%)
Source: DARES/INSEE, Working Conditions surveys.
Figure 7
Pace of work imposed by commercial constraints per cohort (%)
Source: DARES/INSEE, Working Conditions surveys.
Figure 8
Pace of work imposed by organizational constraints per cohort (%)
Source: DARES/INSEE, Working Conditions surveys.
The youngest are the most exposed
Whatever the type of strain or risk studied, young people are systematically the most exposed; they also accumulate a larger number of constraints. This accumulation, exacerbated by a lack of experience, makes them particularly vulnerable to work accidents: 13% of 20-25 year-olds were victims of work accidents in 1998 compared with 8% on average. Furthermore, young people under 25, particularly men, are exposed to substantial working time flexibility and to the highest pace of work constraints (62% vs. 52% on average in 1998).
The descriptive study of the evolution of working conditions reveals a regular increase in exposure to poor working conditions between 1984 and 1998 among the older workers but even more so among the youngest. In this respect, France has followed the European tendency, despite already being among the countries with the worst working conditions (Volkoff et al., 2000).
From age 50 onward, employees are relatively less exposed to strain and risks. However, structural effects may moderate or, on the contrary, exaggerate this reduced exposure, as older people do not hold the same positions or do not work in the same sectors as the youngest workers. We will isolate the specific effects of age on exposure to constraints using a multivariate regression.
However, our sub-population, comprising persons in wage employment, is not representative of the total population. In particular, pensioners, those in early retirement and unemployed people are missing. Yet the probability of belonging to the sample may not be independent of past exposure to work constraints: certain seniors previously exposed to strenuous work may have been excluded from the labour market for health reasons. Worn-out workers may be declared unfit by their doctors to release them from an untenable situation
[14]. If these workers are replaced by younger people, fewer seniors are present in demanding positions. This selection bias needs to be corrected.
II. The effect of age on exposure to constraints
1. Typology of working conditions on the French labour market
Given that a single post combines several different job requirements, we performed a multiple correspondence analysis (MCA) of the qualitative variables. This makes it possible to synthesize fragmented survey information and to establish a typology of working conditions. There are three main types of constraint
[15]:
- Strain and risks: the variables that contribute strongly to the overall variance are, for example, the risk of projection or of falling materials, the risk of injury by a machine, the risk of a serious fall, physical exertion, working in a noisy environment, etc.
- Flexible and atypical work schedules: the active variables determining this axis are work on Saturdays, Sundays, nights, number of days worked during the week variable and imposed by the company, absence of mealtime breaks, more than five days worked in a week, less than 48 consecutive hours’ rest, alternating work schedules, several work periods in one day, etc.
- Commercial demands: the most discriminating active variables are the absence of regular hours, pace of work imposed by demand, the experience of tense situations, being obliged to manage on one’s own, receiving unclear or incomplete instructions, coping with an outside demand that requires an immediate response, being in contact with the public, using computers, not having a fixed work schedule, etc.
Commercial demands are ambivalent. On the one hand, autonomy and taking responsibility increase the appeal of a job and the positive image it projects. On the other, they may be a source of stress and strong tension, particularly when work intensity increases or when insufficient means are available to produce good quality work, in which case they may be felt as a constraint.
This type of constraint became apparent from 1991 due to the development of new forms of work organization, new skills and the transformation of skills requirements. Since the early 1980s, the extension of the commercial model
[16] to industrial structures has modified the nature of work and of required skills (Eymard-Duvernay, 1987). Work has become less formalized and more adaptable, so that commercial demands and customers’ needs can be addressed more quickly. It is increasingly oriented towards customer satisfaction, to the extent that employee autonomy and responsibility are now highly regarded qualities. New questions pertaining to psychological and organizational constraints were introduced into the Working Conditions survey from 1991 onward. They concern the degree of freedom to interpret instructions given, room for manœuvre in case of problems, the possibility of conferring with other people, work requirements and use of personal initiative.
2. The estimated model
Using multiple correspondence analysis, we build three synthetic variables that reflect each individual’s exposure respectively to physical strain, flexible or atypical hours and commercial demands
[17]. These three variables are continuous
[18]. To measure the effect of age on exposure to different constraints independently from structural effects, we use the ordinary least squares regression method for each of these variables. The estimates were carried out for 1998. The variable of interest is the age group, i.e., 15-24, 25-34, 35-49, 50-54, 55-60 and 61-65. The reference age group is 35-49. The other independent variables are sex, level of education, years of tenure with current employer, socio-occupational category, sector of activity, size of the establishment, employer’s status (public or private, type of work contract (temporary, fixed-term, open-ended contract) and category of municipality. To correct for selection bias, we use the method proposed by J. Heckman [1979] (Box).
Correcting for selection bias
Most studies that correct for selection bias use heckman’s (1979) two-step procedure. It consists in first estimating the probability of participating in the labour market as an employee (probit model). It is assumed that being an employee, D, is determined by a latent variable D* and that D is defined by the following equation:
Zi is a vector of independent variables for participation. ei corresponds to an error term whose distribution function, Φ is assumed to be that of average.
The estimated parameters can be used to calculate the inverse Mill’s ratio, λi.
φ is the density function of the normal law.
Next, we estimate the equation for exposure to the various risks by integrating the λi variable. A Student’s test is performed to determine whether the coefficient of the estimate corresponding to this ratio is significant. If so, the same is true for selection bias. By introducing this new term, the coefficients of the other variables are modified and become unbiased.
This method does not, however, correct fully for selection bias since individual histories cannot be taken into account. Histories differ from one industry to another, in the public and private sector, etc. The Heckman method does not allow us to correct for selection due to working history or state of health, insofar as we have access neither to retrospective information, nor to health indicators to estimate the probability of participation. Only a longitudinal survey, covering working histories and reasons for non-participation, for health reasons in particular, would make it possible to control fully for this bias. Nonetheless, this method can be used to approach working histories indirectly, via independent variables of the probability of participation. Women with little education, for example, are more likely to be employed as manual workers. Likewise, the level of education has a significant impact on health.
3. Main characteristics of exposed employees
Exposure to the various constraints depends strongly on individual characteristics such as age, sex, level of education or tenure (Table 1). All things being equal elsewhere, women are less exposed than men to all constraints, particularly physical strain and occupational risks. Education protects against physical strain and working time flexibility while commercial demands rise with the level of education, most noticeably with the level of higher education. Tenure, for its part, significantly increases exposure to physical and commercial constraints.
Table 1
Factors of exposure to work constraints, 1998 (ols with correction for selection bias)
Exposure to different constraints depends strongly on types of occupation and sectors of activity. Employees with manual jobs (manual workers, skilled workers and supervisors in particular) are much more exposed to physical constraints than white-collar workers, intermediate and higher-level occuptions, while working time flexibility is more likely to concern sales employees and service employees working directly for private individuals. People in management and intermediate professions are more affected by commercial demands. Sector of activity is an important explanatory factor, close behind occupation, sex and level of education. The sectors with the highest levels of strain and physical risks are intermediate goods manufacturing, construction, education and health, the automotive industry, agriculture, the agri-food industry and transportation. Employees in education and health, services to individuals and transportation are the most exposed to working time flexibility. Employees in public and government services, transportation, energy, services to individuals and to business are, all things being trade and equal elsewhere, more Contrary to certain preconceived ideas, working in the private sector reduces exposure to physical constraints and working time flexibility, but increases exposure to commercial demands. While the weight of constraints increases with the size of the work entity (except for the very largest ones), atypical working hours are also frequent in structures employing fewer than ten people.
Lastly, note that occupational insecurity is accompanied by increased exposure to work constraints, all things being equal elsewhere: temporary employees are particularly exposed to physical strain while employees on short-term contracts are more likely to have atypical working hours.
4. Protecting the oldest employees
If selection bias is not taken into account, age becomes a determining factor in exposure to the various types of work constraints. Exposure to physical strain and commercial demands decreases significantly as age progresses (Table 2). The youngest employees, particularly those under 25, are the most exposed. Starting at age 50, a significant decrease in exposure to physical demands is observed. For employees between the ages of 50-54 and 55-59, the strain indicator is 3% lower on average than that of 35-49 year-olds
[19]. It is mainly from the age of 60 that exposure falls (–9%). Employees between the ages of 50 and 59 are also less affected by commercial demands, perhaps because they do not possess the right skills, or are thought not to possess them by employers. On the other hand, employees aged 55 and over are more often exposed to working time flexibility.
Table 2
Effect of age on exposure to work constraints, 1998 (OLS, estimate without correction for selection bias)
When a distinction is made between the sexes, protection against physical demands beyond age 55 is no longer significant for men, though it remains at the 10% threshold for women (Table 3). So for men, reassignment away from the most strenuous positions would appear to play a role until age 55. From age 50 onward, women, and men in particular, are less affected by commercial demands.
Table 3
Effect of age on exposure to work constraints by sex, 1998 (OLS with correction for selection bias)
These initial results indicate that older employees are protected against the most strenuous work. Nonetheless, to reach this conclusion it is necessary to correct for selection bias. If, after doing so, the coefficients still have the same negative sign, protection for older employees does indeed exist. If, on the contrary, age becomes a non-significant variable, this indicates that older employees are neither more nor less exposed than others.
The independent variables in the equation of labour force participation (equation [2]) are age group, level of education and category of municipality. Family situation variables (living in a couple or not), nationality and area of residence serve to identify the equation. Age is an important selection variable that significantly influences the probability of participation: compared with 35-49 year-olds, participation is lower for the youngest (especially for the under-25s) and the oldest (especially for the 55-59 age group, and even more so for the 60-64 age group)
[20]. Likewise, the probability of participation is lower for women. On the other hand, level of education and living in a couple increase the probability of employment.
The coefficient of the lambda parameter (equation [4]) is significantly different from zero in each of the three regressions for exposure to work constraints, so there is selection bias linked to being in employment. Correcting this bias primarily affects the coefficients of age-related variables. So, once bias has been corrected (Table 1), it is revealed that people over 55 are no longer significantly less exposed to physical strain. The coefficient remains negative for 50-54 year-olds however
[21], indicating that protection of older employees applies to the 50-54 age group only.
For exposure to working time flexibility, the sign of age coefficients goes from positive without the correction to negative with the correction, indicating that the oldest workers are protected. Lastly, for commercial constraints, 50-54 year-olds are likewise less exposed. However, as is the case for physical strain, the coefficient is no longer significant for 55-59 year-olds.
We suggest that beyond age 55, protection against physical strain often operates through exclusion from the workforce, particularly by means of early retirement, rather than reassignment to other positions within the company. The low employment rates of the over 55s, due largely to the weight of early retirement, bear witness to this. The employment rate for 55-59 year-olds is 61.7% for men and 45.7% for women (resp. 84.1% and 66.6% for 50-54 year-olds). Note however that the selection bias cannot be perfectly controlled by the technique used because we do not have the necessary tools to model early departures. Such a model could be developed using panel data. For instance, the two-wave ESTEV
[22] survey provides longitudinal information on work and health, as well as information on exposure to certain working conditions. Due to its data collection method — the works doctors had to be volunteers — and its scope, the survey is not representative of employees. A longitudinal survey project concerning health and working histories is currently being planned and will enable us to analyse these results in greater depth (Cristofari, 2003).
5. Working conditions and health
The Working Conditions survey does not allow us to directly measure the consequences of working conditions on health. Nonetheless, whether epidemiological or ergonomic, studies analysing longitudinal data on working conditions and state of health, clearly show a negative relationship between exposure to risks and damage to health. Ergonomists have established that musculoskeletal disorders (hand, wrist, elbow and shoulder problems) are caused by movements at work, generally repetitive in nature, when carried out in uncomfortable positions (Laville, 1989). Work based on the ESTEV survey has shown that physical constraints and unnatural posture, vibrations or repetitive tasks increase the prevalence of certain musculoskeletal pains for example (Monfort et al., 2001). Certain forms of work organization may also have irreversible effects on health: night work or alternating hours increase long-term cardiovascular and coronary risk (Lasfargues, 2004). The link between exposure to commercial demands and impact on health is less direct. These demands create their own type of strain and wear however, most noticeably through the development of stress and mental strain, which can be just as selective as physical strain (Bertin et al., 2001; Vezina et al., 2001; Carayon et al., 1999, Siegrist et al., 2004).
The intensification of work has increased strain but has also affected the conditions in which work is carried out: operators have less room for manœuvre and carry out their work in increasingly uncomfortable positions, and they have less time to develop compensation mechanisms. Ergonomists have shown that operators develop strategies to preserve their health: changing posture, fixing up their work space, taking breaks to recuperate, choosing adequate tools, anticipating tasks to come, cooperation, collecting information (Gaudart, 1996). These strategies are especially important as they provide a means to offset the increase in functional impairments with age (Ilmarinen, 1999). Faced with the intensification of work, workers have less time to establish compensatory mechanisms, which in turn creates more tension.
But in addition to work-related ageing, working conditions give rise to processes that exclude people from the labour market. Saurel-Cubizolles et al. (2001) have shown that a decline in health caused by exposure to physical demands contributes to exclusion from the labour market.
Conclusion
The organizational changes in the 1980s and 1990s caused a deterioration of working conditions in France. Far from disappearing, physical strain and occupational risks are ever-present in the workplace. Likewise, the pace of work has intensified while working time flexibility and demand constraints have developed considerably. This degradation is felt by workers of all ages and the exposure of the over-50s to work constraints has increased with respect to previous birth cohorts at the same age.
Against this backdrop, mechanisms for protecting older people continue to exist. Compared with the youngest employees, the over-50s are protected, with women being protected from physical strain while men are protected from commercial demands. Older employees still appear to belong to a protected segment of the labour market. Young people, for their part, claim to be more exposed to all types of risk. It is possible that their lower status (which sometimes helps to counterbalance strain) makes them more sensitive to working conditions; but young people are also subjected to constraints that can no longer be imposed upon older employees (Davezies, 2001). Even though some older employees are protected from strain and risks by being reassigned to new tasks within the company, selection takes places through exclusion from the labour market of individuals who would traditionally have been protected but who are now exposed to redundancy, forced early retirement or self-exclusion due to poor health.
The 2003 pension reform encourages continued employment beyond the age of 60, so the elderly employee population is likely to grow. To restrict self-exclusion by employees, companies should be encouraged to pay greater attention to occupational strain. Moreover, from the viewpoint of social justice, strain should be taken into account when calculating retirement age in order to avoid penalizing workers who are the most exposed to risk. Negotiations between social partners will need to be initiated; though the definition of criteria remains open to broad debate (Volkoff, 2004).
BIBLIOGRAPHIE
·
Aquain Valérie, Michel Cézard, Alain Charraud, Lydie Vinck, 1994, “Vingt ans d’évolution des conditions de travail”, Travail et emploi, no. 61, pp. 81-91.
·
Askenazy Philippe, Ève Caroli, Vincent Marcus, 2002, “New organizational practices and working conditions: Evidence from France in the 1990s”, Couverture orange, CEPREMAP, no. 01-06.
·
Bertin Françoise, Francis Derriennic et le groupe Estev-Psychopathologie des Pays de Loire, 2001, “Sentiment de lassitude, âge et conditions de travail”, in Travail, santé, vieillissement. Relations et évolutions, Proceedings of the conference of 18 and 19 November 1999, Toulouse, Octares Editions.
·
Belfer Richard, 2004, “Les conditions de travail au cœur du malaise hospitalier”, Santé et travail, no. 46, pp. 10-13.
·
Bué Jenifer, Catherine Rougerie, 1999a, “L’organisation du travail : entre contrainte et initiative”, Premières synthèses, DARES, no. 32.1.
·
Bué Jenifer, Catherine Rougerie, 1999b, “L’organisation des horaires : un état des lieux en mars 1998”, Premières synthèses, DARES, no. 30.1.
·
Cambois Emmanuelle, Jean-Marie Robine, Mark D. Hayward, 2001, “Social inequalities in disability-free life expectancy in the French male population: 1980-1991”, Demography, no. 4, pp. 513-524.
·
Carayon Pascale, Michael J. Smith, Marla C. Haims, 1999, “Work organization, job stress, and work-related musculoskeletal disorders”, Human Factors and Ergonomics in Manufactoring, 41(4), pp. 644-663.
·
Cases Chantal, Nathalie Missègue, 2001, “Une forte segmentation des emplois dans les activités de services”, Économie et statistique, no. 344, pp. 81-108.
·
Cassou Bernard, Francis Derriennic, Yuriko Iwatsubo, Maurice Amphoux, 1994, “Incapacité physique après la retraite et conditions de travail”, Gérontologie et société, no. 70, pp. 179-192.
·
Cézard Michel, Françoise Dussert, Michel Gollac, 1999, “Méthodologie et évolution des réponses dans les trois enquêtes”, Les conditions de travail en enquêtes, Documents d’études de la DARES, no. 29.
·
Cézard Michel, Sylvie Hamon-Cholet, 1999a, “Efforts et risques au travail”, Premières synthèses, DARES, no. 16.1.
·
Cézard Michel, Sylvie Hamon-Cholet, 1999b, “Travail et charge mentale”, Premières synthèses, DARES, no. 27.1.
·
Coutrot Thomas, 2000, “Innovations dans le travail : la pression de la concurrence internationale, l’atout des qualifications”, Premières Synthèses, DARES, no. 09.2.
·
Cristofari Marie-France, 2003, Bilan des sources quantitatives dans le champ de la santé et de l’itinéraire professionnel, Report of the Centre d’études de l’emploi 03/11.
·
Cru Damien, 1987, “Les règles des métiers”, in Christophe Dejours (ed.) Plaisir et souffrance dans le travail, CNRS.
·
Davezies Philippe, 2001, “Vieillissement et processus de sélection : de l’observation à la recherche de modalités d’action”, in Travail, santé, vieillissement, relations et évolutions, Octarès, pp. 43-52.
·
Desplanques Guy, Andrée Mizrahi, Arié Mizrahi, 1996, “Mortalité et morbidité par catégories sociales”, Solidarité santé, études statistiques, no. 4, pp. 75-85.
·
Doeringer Peter B., Michael Piore, 1971, International Labour Market and Manpower Analysis, Heath Lexington Books.
·
Eymard-Duvernay François, 1987, “Introduction : les entreprises et leurs modèles”, Cahiers du Centre d’études de l’emploi, no. 30, V-XXII.
·
Gaudart Corinne, 1996, “Vieillir, mais ‘tenir’ la cadence”, Gérontologie et société, no. 77.
·
Gollac Michel, 1994, “Donner un sens aux données, l’exemple des enquêtes statistiques sur les conditions de travail”, Dossiers de recherche du Centre d’études de l’emploi, no. 3.
·
Gollac Michel, Nathalie Greenan, Sylvie Hamon-Cholet, 2000, “L’informatisation de l’ancienne économie, nouvelles machines, nouvelles organisations et nouveaux travailleurs”, Économie et statistique, no. 339-340, pp. 171-201.
·
Gollac Michel, Serge Volkoff, 2001, Les conditions de travail, La Découverte (coll. Repères, no. 301).
·
Green Francis, Steven McIntosh, 2001, “The intensification of work in Europe”, Labour Economics, 8(2), pp. 291-308.
·
Hayward Mark D., William R. Grady, Melissa A. Hardy, David Sommers, 1989, “Occupational influences on retirement, disability, and death”, Demography, 26(3), pp. 393-409.
·
Heckman James, 1979, “Sample selection biais as a specification error”, Econometrica, pp. 153-162.
·
Ilmarinen Juhani, 1999, Ageing Workers in the European Union : Status and Promotion of Work Ability, Employability and Employment, Helsinki, FIOH, 274 p.
·
Jolivet Annie, 1999, Entreprise et gestion de la main-d’œuvre vieillissante : organisation, discrimination, doctoral thesis, Université Paris 1.
·
Krause Niklas, John W. Lynch, George A. Kaplan, Jukka T. Salonen, Richard D. Cohen, 2000, “Standing at work and progression of carotid atherosclerosis”, Scandinavian Journal of Work, Environment and Health, 26(3), pp. 227-236.
·
Lasfargues Gérard, 2004, “Pénibilité du travail et santé : des effets multiples et parfois irréversibles”, Travail et emploi, no. 46 (dossier spécial Pénibilités et retraite : repères pour négocier).
·
Laville Antoine, 1989, “Vieillissement et travail”, Le travail humain, 52(1), pp. 3-20.
·
Le Minez Sylvie, 1995, “Les entreprises et le vieillissement de leur personnel : faits et opinions”, Travail et emploi, no. 63, pp. 23-39.
·
Molinié Anne-Françoise, 1998, “Déclin et renouvellement de la main-d’œuvre industrielle. Une lecture des structures d’âge”, Économie et statistique, no. 316-317, pp. 109-126.
·
Molinié Anne-Françoise, 2001, Age and Working Conditions in the European Union, Report for the European Foundation for the Improvement of Living and Working Conditions.
·
Monfort Claude, Bernard Cassou, Joanna Norton, Francis Derriennic, 2001, “Conditions de travail et évolution des douleurs musculo-squelettiques chroniques du cou et des épaules – ESTEV, analyse longitudinale 1990-95”, in Travail, santé, vieillissement, relations et évolutions, Octarès, pp. 81-89.
·
Moore David E., Mark D. Hayard, 1990,” Occupational careers and mortality of elderly men”, Demography, 27(1), pp. 31-53.
·
Pailhé Ariane, 2005 (forthcoming), “L’intensité du travail s’allège-t-elle avec l’âge?”, in P. Askenazy, D. Cartron, F. de Coninck, M. Gollac (dir.), Organisation et intensité du travail, Octarès (Entreprise, travail, emploi).
·
Piore Michael. J., 1978, “Dualism in the Labor Market”, Revue économique, 29(1), pp. 26-48.
·
Saurel-Cubizolles Marie-Josèphe, Fabienne Bardot, Bernadette Berneron, Michèle Fromet, Gérard Lasfargues, Marie-Christine Minois, Christine Monfort, Catherine Robida, Catherine Rondeau du Noyer, Francis Derrienic, 2001, “État de santé perçu et perte d’emploi”, in Travail, santé, vieillissement. relations et évolutions, Proceedings of the conference of 18 and 19 November 1999, Toulouse, Octarès.
·
Siegrist Johannes, Michael Marmot, 2004, “Health inequalities and the psychosocial environment: two scientific challenges”, Social Science and Medicine, 58(8), pp. 1463-1473.
·
Teiger Catherine, 1989, “Le vieillissement différentiel dans et par le travail : un vieux problème dans un contexte récent”, Le travail humain, 52(1), pp. 21-56.
·
Valeyre Antoine, 2001, “Le travail industriel sous la pression du temps”, Travail et emploi, no. 86, pp. 127-149.
·
Vezina Michel, Francis Derriennic, Claude Monfort, 2001, “Tensions au travail et atteintes à la santé mentale : l’éclairage de l’enquête ESTEV”, in Travail, santé, vieillissement, relations et évolutions, Octarès, pp. 177-187.
·
Volkoff Serge, 2004, “Expert en pénibilité, un métier d’avenir? », Travail et emploi, no. 46 (dossier spécial Pénibilités et retraite : repères pour négocier).
·
Volkoff Serge, Anne-Françoise Molinie, 1995, “Éléments pour une démographie du travail”, in Le travail au fil de l’âge, Toulouse, Octarès, pp. 99-119.
·
Volkoff Serge, Anne-Françoise Molinie, Annie Jolivet, 2000, “Efficaces à tout âge? Vieillissement démographique et activités de travail”, Dossier du CEE, no. 16.
NOTES
[*]
National d’Études Démographiques, Paris.Translated from the French by Jane Muir Dickinson.
National d’Études Démographiques, Paris.Translated from the French by Jane Muir Dickinson.
[1]The contributory period will increase from 40 to 41 years in the private sector over the period 2009-2012 and from 37.5 to 41 years in the public sector over the period 2004-2012.
The contributory period will increase from 40 to 41 years in the private sector over the period 2009-2012 and from 37.5 to 41 years in the public sector over the period 2004-2012.
[2]Law that enacted a lengthening of the contributory period from 37.5 to 40 years for private sector employees.
Law that enacted a lengthening of the contributory period from 37.5 to 40 years for private sector employees.
[3]They transformed the organization of production, notably through production line standardization, just-in-time production, total quality procedures, decentralized production and sub-contracting.
They transformed the organization of production, notably through production line standardization, just-in-time production, total quality procedures, decentralized production and sub-contracting.
[4]The Employment Survey, or labour force survey, is used to establish data on employment and unemployment using International Labour Organization definitions.
The Employment Survey, or labour force survey, is used to establish data on employment and unemployment using International Labour Organization definitions.
[5]The question “When you are working, if a person standing two or three meters away from you speaks to you: (1) you can hear the person if he/she speaks normally; (2) you can hear the person if he/she raises his/her voice; (3) you cannot hear the person” aims, for example, to indirectly pinpoint the 85-decibel threshold, generally considered by ergonomists as the noise level above which irreparable damage to hearing is liable to occur.
The question “When you are working, if a person standing two or three meters away from you speaks to you: (1) you can hear the person if he/she speaks normally; (2) you can hear the person if he/she raises his/her voice; (3) you cannot hear the person” aims, for example, to indirectly pinpoint the 85-decibel threshold, generally considered by ergonomists as the noise level above which irreparable damage to hearing is liable to occur.
[6]Strikes reveal latent difficulties concerning working conditions and encourage workers to declare them more often. This was the case for nurses, police officers, and SNCF rail employees in 1991 (Gollac, 1994).
Strikes reveal latent difficulties concerning working conditions and encourage workers to declare them more often. This was the case for nurses, police officers, and SNCF rail employees in 1991 (Gollac, 1994).
[7]The degree of allegiance to what is known in the field of labour psycho-dynamics as a ”defensive strategy of the trade” also has an impact on these declarations. For example, not all drivers mention the risk of having a road accident. For some of them, admitting such a risk would weaken their status (Cru, 1987). Since 1984, the proportion of drivers mentioning the risk of a road accident has increased noticeably however, indicating a decline in the use of this type of strategy.
The degree of allegiance to what is known in the field of labour psycho-dynamics as a ”defensive strategy of the trade” also has an impact on these declarations. For example, not all drivers mention the risk of having a road accident. For some of them, admitting such a risk would weaken their status (Cru, 1987). Since 1984, the proportion of drivers mentioning the risk of a road accident has increased noticeably however, indicating a decline in the use of this type of strategy.
[8]These figures correspond to the proportion of people who claimed to be exposed to at least one of the following forms of physical effort: remaining standing or working in painful or tiring postures for long periods, carrying or moving heavy loads, performing other physically strenuous tasks, being exposed to vibrations, not being allowed to take one’s eyes off one’s work, reading small letters or numbers, examining very small objects or fine detail, watching out for short visual signals that are unpredictable or hard to detect, or, lastly, listening out for brief audio signals.
These figures correspond to the proportion of people who claimed to be exposed to at least one of the following forms of physical effort: remaining standing or working in painful or tiring postures for long periods, carrying or moving heavy loads, performing other physically strenuous tasks, being exposed to vibrations, not being allowed to take one’s eyes off one’s work, reading small letters or numbers, examining very small objects or fine detail, watching out for short visual signals that are unpredictable or hard to detect, or, lastly, listening out for brief audio signals.
[9]Job autonomy is assessed by the degree of freedom to interpret orders received, room for manœuvre in case of problems, the possibility of conferring with others, work requirements and the use of initiative.
Job autonomy is assessed by the degree of freedom to interpret orders received, room for manœuvre in case of problems, the possibility of conferring with others, work requirements and the use of initiative.
[10]Tension in the work place is measured by direct questions relating to the perception of tension in relations with superiors or with the public, but also fear of sanctions, unexpected switching of jobs, having to work quickly, not being allowed to take one’s eyes off one’s work, working in a noisy environment, lack of time, or pressure of responsibility.
Tension in the work place is measured by direct questions relating to the perception of tension in relations with superiors or with the public, but also fear of sanctions, unexpected switching of jobs, having to work quickly, not being allowed to take one’s eyes off one’s work, working in a noisy environment, lack of time, or pressure of responsibility.
[11]For a detailed analysis of work intensity by age, see Pailhé, 2005.
For a detailed analysis of work intensity by age, see Pailhé, 2005.
[12]A strain indicator is calculated for each individual, using the sum of constraints taken from the following list: remaining standing for long periods, working for long periods in a difficult posture, exposure to jolts or vibrations, carrying or moving heavy loads, walking for long periods or frequently, other forms of strenuous physical exertion, not taking ones eyes off one’s work, reading small letters, examining small objects, being exposed to brief, unpredictable visual or audio signals.
A strain indicator is calculated for each individual, using the sum of constraints taken from the following list: remaining standing for long periods, working for long periods in a difficult posture, exposure to jolts or vibrations, carrying or moving heavy loads, walking for long periods or frequently, other forms of strenuous physical exertion, not taking ones eyes off one’s work, reading small letters, examining small objects, being exposed to brief, unpredictable visual or audio signals.
[13]The total number of risks incurred is counted from among the following: risk of burning, electrocution, serious fall, falling materials, injury, traffic accident, contact with explosives, inhalation of fumes, dust or toxic products, handling of toxic products.
The total number of risks incurred is counted from among the following: risk of burning, electrocution, serious fall, falling materials, injury, traffic accident, contact with explosives, inhalation of fumes, dust or toxic products, handling of toxic products.
[14]Workers over 57 declared unfit for work are entitled to unemployment benefits up to retirement age.
Workers over 57 declared unfit for work are entitled to unemployment benefits up to retirement age.
[15]These three constraints correspond to the first three axes of the MCA. Each axis compares employees exposed to the various constraints with those who are not exposed. The first three axes explain 20% of the scatter diagram variance (10% for the first, 5.3% for the second and 4.7% for the third).
These three constraints correspond to the first three axes of the MCA. Each axis compares employees exposed to the various constraints with those who are not exposed. The first three axes explain 20% of the scatter diagram variance (10% for the first, 5.3% for the second and 4.7% for the third).
[16]This model is characterized by an adaptation of products and product distribution to the market and to customers.
This model is characterized by an adaptation of products and product distribution to the market and to customers.
[17]In practice, we use the coordinates of each individual on each of the factorial axes of the MCA.
In practice, we use the coordinates of each individual on each of the factorial axes of the MCA.
[18]The variable synthesizing the information relative to physical constraints has a minimum of –0.592 and a maximum of 1.423; the one concerning working hour flexibility has a minimum of –0.763 and a maximum of 0.808; for the variable relating to commercial demands, the figures are respectively –0.655 and 0.881. Their respective means are 0.001, –0.001 and 0.000 and their standard deviations are 0.367, 0.248 and 0.237.
The variable synthesizing the information relative to physical constraints has a minimum of –0.592 and a maximum of 1.423; the one concerning working hour flexibility has a minimum of –0.763 and a maximum of 0.808; for the variable relating to commercial demands, the figures are respectively –0.655 and 0.881. Their respective means are 0.001, –0.001 and 0.000 and their standard deviations are 0.367, 0.248 and 0.237.
[19]In an ordinary least squares regression, the estimated coefficient directly represents the impact of a change in the exogenous variable on the dependent variable. The coefficient is thus interpreted in this case as the variation in the constraint indicator in relation to reference age category. We checked that coefficients were significantly different, other than those for the 50-54 and 55-59 age groups.
In an ordinary least squares regression, the estimated coefficient directly represents the impact of a change in the exogenous variable on the dependent variable. The coefficient is thus interpreted in this case as the variation in the constraint indicator in relation to reference age category. We checked that coefficients were significantly different, other than those for the 50-54 and 55-59 age groups.
[20]The results are not given here. They are available from the author.
The results are not given here. They are available from the author.
[21]The coefficients of the age variables are statistically different.
The coefficients of the age variables are statistically different.
[22]The “health, work and ageing” (ESTEV) survey was carried out in 1990 and 1995 in seven French regions on employees monitored by their company doctors. It concerns four employee cohorts born in 1938, 1943, 1948 and 1953, forming a panel of more than 20,000 people altogether in 1990 and 1995. It includes information on working history, exposure to certain working conditions, biometric data and psycho technical tests concerning cognitive functions.
The “health, work and ageing” (ESTEV) survey was carried out in 1990 and 1995 in seven French regions on employees monitored by their company doctors. It concerns four employee cohorts born in 1938, 1943, 1948 and 1953, forming a panel of more than 20,000 people altogether in 1990 and 1995. It includes information on working history, exposure to certain working conditions, biometric data and psycho technical tests concerning cognitive functions.
© Cairn 2007
Vie privée |
Conditions d’utilisation |
Conditions générales de vente
À propos | Éditeurs | Bibliothèques | Aide à la navigation | Plan du site | Raccourcis
À propos | Éditeurs | Bibliothèques | Aide à la navigation | Plan du site | Raccourcis
Retour en haut de page
Ici s'achève le contenu de la page, la suite ne concerne que les attributs d'affichage
