CHRONOBIOLOGY INTERNATIONAL

Vol. 21, No. 6, pp. 831–844, 2004

Flexible Working Hours, Health, and Well-Being in

Europe: Some Considerations from a SALTSA Project

Giovanni Costa,1,* Torbjorn Akerstedt,2 Friedhelm Nachreiner,3

Federica Baltieri,1 Jose Carvalhais,4 Simon Folkard,5

Monique Frings Dresen,6 Charles Gadbois,7 Johannes Gartner,8

Hiltraud Grzech Sukalo,9Mikko Harma,

10Irja Kandolin,

10

Samantha Sartori,1 and Jorge Silverio11

1Department of Medicine and Public Health, University of Verona, Italy2IPM and Karolinska Institutet, Stockholm, Sweden

3Institut fuer Psychologie, Carl von Ossietzky Universitaet, Oldenburg, Germany4Faculdade de Motricidade Humana, Universidade Tecnica de Lisboa, Portugal

5Body Rhythms and Shiftwork Center, University of Wales, Swansea, UK6Coronel Instituut voor Arbeid, Milieu en Gezondheid, Amsterdam, Netherlands

7Laboratoire d’Ergonomie Physiologique, EPHE, Paris, France8XIMES and University of Technology, Vienna, Austria

9AWiS-Consult, Oldenburg, Germany10Brain@Work Laboratory, FIOH, Helsinki, Finland

11Universidade do Minho, Porto, Portugal

ABSTRACT

The project brought together researchers from 9 EU-Countries and resulted in a

number of actions, in particular the following: (a) There is an urgent need of

defining the concept of flexible working hours, since it has been used in many

different and even counterintuitive ways; the most obvious distinction is where

*Correspondence: Giovanni Costa, Department of Medicine and Public Health, University of

Verona, Strada le Grazie 8, Verona 37134, Italy; Fax: +39 045 8027633; E-mail: giovanni.

costa@univr.it.

831

DOI: 10.1081/LCBI-200035935 0742-0528 (Print); 1525-6073 (Online)

Copyright & 2004 by Marcel Dekker, Inc. www.dekker.com

ORDER REPRINTS

the influence over the working hours lies, that is between the ‘‘company-based

flexibility’’ and the ‘‘individual-oriented flexibility’’; (b) The review of the

Legislation in force in the 15 European countries shows that the regulation of

working times is quite extensive and covers (Council Directive 93/104/EC) almost

all the various arrangements of working hours (i.e., part-time, overtime, shift, and

night work), but fails to provide for flexibility; (c) According to the data of the

Third EU Survey on Working Conditions, longer and ‘‘irregular’’ working hours

are in general linked to lower levels of health and well-being; moreover, low

(individual) flexibility and high variability of working hours (i.e., company-based

flexibility) were consistently associated with poor health and well-being, while low

variability combined with high autonomy showed positive effects; (d) Six

substudies from different countries demonstrated that flexible working hours

vary according to country, economic sector, social status, and gender; overtime is

the most frequent form of company-based flexibility but has negative effects on

stress, sleep, and social and mental health; individual flexibility alleviates the

negative effects of the company-based flexibility on subjective health, safety, and

social well-being; (e) The literature review was able to list more than 1,000

references, but it was striking that most of these documents were mainly

argumentative with very little empirical data. Thus, one may conclude that there

is a large-scale intervention ongoing in our society with almost completely

unknown and uncontrolled effects. Consequently, there is a strong need for

systematic research and well-controlled actions in order to examine in detail what

flexible working hours are considered, what and where are their positive effects, in

particular, as concerns autonomy, and what regulation seem most reasonable.

Key Words: Flexible working hours; Health; Well-being; Shift work; Work

flexibility.

THE CONTEXT

The borders between working and leisure times are no longer fixed and rigidlydetermined by the normal diurnal working day. Not only are working hoursextended into the evening and night hours, as well as to weekend days, but hours ofduty have become more variable (e.g., part-time, 6 and 12 h shifts, irregular shiftschedules, compressed work weeks, split shifts, etc.) in a society that never stops(‘‘24 h Society’’), and requires continuous adjustments to its evolution andorganization.

Looking at the data of the Third EU Survey on Working Conditions carried outin the year 2000 in the 15 European countries, we can see that the so-called ‘‘normal’’or ‘‘standard’’ working hours represent more the exception than the rule. In fact, ifwe consider the 21,703 workers of the sample, and we progressively exclude thepersons (a) working more than 40 h/week, (b) working more than 10 h/day, (c)working on shifts, (d) working at night, (e) working on Sunday, (f) working part-time, and (g) working on Saturday, we discover that people working on ‘‘normal’’daytime work (i.e., between 07:00–08:00 h and 17:00–18:00 h) from Monday toFriday) now constitute only 24% of the entire population, in particular 27% ofemployed and only 8% of self-employed workers (Figure 1) entire population.Significant differences can be seen also according to gender, where overtime for men

832 Costa et al.

ORDER REPRINTS

and part-time work for women cause the largest differences, also among the 15 EUcountries, being widely influenced by different economic, cultural, political, andsocial backgrounds.

The increasing diversification of working time patterns reflects broader societalreasons, involving social development, economic efficiency, and individualpreferences. The modern society is changing quite rapidly not only in terms ofeconomic and productive strategies (e.g., new technologies, market globalization,information processes), but mainly in terms of social organization and individualbehaviors and careers.

There are several approaches and models to define ‘‘labor flexibility,’’ dependingon the different work, social, and personal parameters, criteria, or perspectives onemay consider. As pointed out by Goudswaard and De Nanteuil (2000) ‘‘flexibilityemerges as a heterogeneous concept, mixing two series of variables: quantitative/qualitative and external/internal, resulting in several possible combinations betweenthese variables;’’ e.g., numerical, geographical, functional, and temporal flexibility.What we considered in this project is the ‘‘temporal flexibility,’’ but we are awarethat the various forms are often strictly linked to one another; consequently, whenlooking at the impact of flexibility on working conditions, health, and well-being, it isnecessary to bear in mind that the positive or negative consequences that may derivefrom different arrangements are influenced more from their interactions ratherthan as a direct effect of one single factor. Moreover, working times are directlyrelated to nonworking times, both on the chronometric as well as chronologic

8%27%

15%34%

19%45%

20%50%

24%57%

26%64%

44%84%

100%100%

0% 10% 20% 30% 40% 50% 60% 70% 90%80% 100%

self-employed employed

≤40 h/w

≤10h/d + ≤40 h/w

no night + ≤10 h/d + ≤40h/w

no Sunday + no night + ≤10 h/d + ≤40h/w

no shift + no Sunday + no night + ≤10 h/d + ≤40 h/w

no part-time + no shift + no Sunday + no night + ≤10h/d + ≤40 h/w

no Saturday + no part-time + no shift + no Sunday + no night + ≤10h/d + ≤40h/w

TOTAL

Figure 1. Percentages of people involved in different patterns of working hours (21,505

workers examined in Third European Survey on Working Conditions, 2000).

Flexible Working Hours, Health, and Well-Being in Europe 833

ORDER REPRINTS

dimensions of time. Changes in the arrangements of working time must thereforelead (or be at least related) to changes in leisure, social, or family time. Changes intemporal flexibility or changes in the arrangements of working hours thus cannot beconsidered in isolation but have to be related to their effects on the individual, her/his social surrounding, and whole society. All these considerations brought us toplan a project aimed at analyzing how flexible working hours are considered andimplemented in the work organization, and what kind of effects they may have onworker’s health and well-being.

THE CONCEPTS

Over the last decades an increasing interest in ‘‘temporal flexibility’’ can beobserved. This occurred in relation to policies concerning working time reduction andemployment, on the one hand, and, on the other hand, to a progressive transfer ofattention from the quantitative to the qualitative aspects of work and social activities,among which working times represent the most important interrelated factor.

There are different points of view about ‘‘labor flexibility’’ between employersand employees. The former are keener to view it in terms of prompt adaptation ofproduction/service systems to market demands and technological and organizationalinnovations; whereas, employees consider it an important tool to improve workingand social life by decreasing work constraints, increasing employment, and (re-)gaining autonomy over one’s own affairs.

As concerns the working times, that leads, on one hand, to a demand for anincrease in the number of hours during which the production is possible and, on theother hand, to a desire for a reduction of individual working hours and/or anincrease in the autonomy of their regulation (for both parties; employees andemployers). Also, society is interested in flexible working hours, as many advantagescan derive from a more flexible arrangement of social activities and services. One ofthe problems in negotiating flexible working hours is discretion or autonomy ofboth parties. Flexibility has to do with making decisions about working hours andadapting them to the demands of the company and/or worker.

Company-oriented flexibility meets the needs of employers, e.g., changingoperational times, varying customer and service times. The company-drivenflexibility is exemplified by the negotiated right to determine a moderate extension(or reduction) of the working week, without overtime pay or reduced wages,respectively. Other varieties include adding shifts, extending shifts, introducingovertime and weekend work, all after separate negotiations however, and thus theydo not really qualify as true flexibility of working hours.

Individual-oriented flexibility of working hours meets the changing needs ofemployees in different phases of life (e.g., studies, family, aging). Individualflexibility provides more autonomy regarding starting and ending times as well asbreaks, days off, and vacations. The possibility of choosing the amount of workinghours in different periods of one’s working life (e.g., part-time work, bank of hours)can also meet the needs of employees.

There are also many differences as concerns the ways of implementation: some areprescriptive (legislation, work contracts), others output-oriented, others procedural

834 Costa et al.

ORDER REPRINTS

and social security–related, and others ethical. Some include flexibility only on behalfof the employers, others only on behalf of the employees, others on both. Theimplication of this distinction for health, safety, and well-being is as yet unclear.

Shift work, for example, enables the company to extend operating hours, as wellas enables Society to extend public or private services. However, shift work mayinterfere with workers’ health and well-being, for which it is necessary to takeappropriate countermeasures. As concerns health, epidemiological evidence revealsnight and shift work schedules negatively affect family and social life, sleep andenergy level (increasing one’s risk to accidents), and they are associated withincreased risk of mood disorders, cardiovascular disease and altered eating patterns,and digestive ailments (see for example in these proceedings: Folkard and Lombardiet al., 2004; Giebel et al., 2004; Ingre et al., 2004; Knutsson, 2004; Nakata et al.,2004; Portela et al., 2004; Rajaratnam and Jones, 2004; Santos et al., 2004). But it isalso evident that such effects can be strongly influenced by many intervening factorsrelated to several aspects pertaining to personal characteristics (e.g., age, gender) andcoping strategies as well as by family, social, and working conditions (Costa, 2003).The result of their interactions depends not only on the specific load of each factor,but also on their temporal occurrence and duration in the worker’s life.

At present, there are large variations in the conceptualization and approaches toflexible working hours among countries, industries, and companies according todifferent cultures, history, socio-economic conditions, work sectors, unions’ power, andindustrial relations. As mentioned above, the most obvious distinction was where theinfluence over the working hours lies, that is, with the company, the individual, or both.

The simplest and most common type of flexible working hours is the ability toswitch shifts between workmates. A typical and traditional example of flexibleworking hours is flexitime with self-determined start- and end-times of work, withcore time when presence is required. The type with maximum individual flexibility issome variety of ‘‘trust hours,’’ that is, working hours are completely under thecontrol of the employee as long as performance objectives are met. A combinedversion is computer-aided self-selection of work schedules, permitting the individualto select a preferred work-shift pattern with different weight given to ‘‘uncomfor-table’’ shifts, thus reducing total working hours load.

Since the introduction of shift and part-time work are discrete events of lowfrequency, resulting in very rigid working hours, and since flexibility has theconnotation of continuous choice, it was decided to exclude such systems from theconcept of flexible working hours and rather focus on such changes that may bemade relatively continuously. Thus, the following tentative definition of flexibleworking hours has been agreed upon: ‘‘Flexible Working Hours should involve acontinuous choice on behalf of employers, employees, or both, regarding the amount(chronometry) and temporal distribution (chronology) of working hours.’’

THE PROJECT

Although it can be easily speculated that, in general, individual-based‘‘flexibility’’ should improve health and well-being, and especially satisfaction,

Flexible Working Hours, Health, and Well-Being in Europe 835

ORDER REPRINTS

whereas company-based ‘‘flexibility’’ might interfere negatively, the effectiveconsequences on health and well-being have still to be analyzed properly, takinginto account both subjective and objective criteria and outcomes, with differentiationof the diverse forms of flexible working hours.

So, the aims of the project were:

To review the European legislation on working hours and the scientific literatureconcerning ‘‘flexible working hours’’ dealing with health and well-being.

To analyze the data of the last European Survey on Working Conditions (Paoliand Merllie, 2001) concerning the distribution of the different kinds ofworking hours organization in Europe, and to evaluate possible relations offlexible working hours with health outcomes.

To collect practical examples of innovative flexible working hours in some EUcountries and evaluate them according to ergonomic criteria.

To report some national case studies and surveys concerning interactionsbetween arrangements of more flexible working hours and effects on healthand well being of the workers, in particular: (a) Working hour flexibility andthe ability to sustain working life to retirement in Sweden (3,493 workersrepresenting the Swedish 25–75-year-old population); (b) The impact of ageand the flexibility of working hours on outcome measures in English aircraftmaintenance engineers (2,093 people working in 197 different companies and156 sites); (c) The use of irregular working schedules in France and some of itssocial and economic factors (19,000 people representing the entire populationof French wage earners); (d) Work demands, working times and sleep amongFinnish information technology professionals (2,355 representatives of theFinnish Information Processing Association); (e) Flexible working hours andwell-being in Finland (2,053 employees aged 25–64 yrs); (f ) The French 35 hwork week and its effects on the quality of life (1,618 representatives of peopleemployed in companies switched to the 35 h work week).

The project was funded by the Joint Programme for Working Life Research inEurope (SALTSA) and driven by the SwedishNational Institute forWorking Life andthe Swedish Confederation of TradeUnions (LO), Professional Associations (SACO),and Professional Employees (TCO). The results are reported on a compact diskpublished by the Swedish National Institute for Working Life (Costa et al., 2003).

THE MAIN OUTCOMES

Review of the legislation in force in the 15 European countries shows that theregulation of working times is quite extensive and covers (also with reference to theCouncil Directive 93/104/EC, 1993) almost all the various arrangements of workinghours from part-time to overtime, from shift and night work, to weekly and dailyworking hours, and from temporary work to daily and weekly rest periods. While thecontent of the review testifies to the great attention paid to this crucial aspect ofwork organization, it fails to provide for flexibility as defined in this report.

836 Costa et al.

ORDER REPRINTS

That was confirmed by the review of the literature that yielded more than1,000 references, but it is striking that most of these documents are mainlyargumentative with very little empirical evidence. In addition, the overall quality ofthe reports was in most cases unacceptably low. The 31 studies presenting at leastsome kind of empirical evidence on the effects of flexible work hours are listed in aseparate Table (Table B1) in Costa et al. (2003), which interested readers shouldconsult.

Thus, from a scientific perspective one may conclude that there is a large-scaleintervention on going in our societies with completely unknown and uncontrolledeffects, especially concerning long-term effects on the safety, health, and well-beingof workers. The additional sub-review of aging in relation to flexible working hoursyielded similar results.

It is apparent there is very little systematic study of the effects of company-driven flexibility, but our impression is that there is a danger of overloadingemployees with work during long periods of time without providing sufficientrecuperation. There are also indications of increased irregularity of working hours indifferent ‘‘on demand’’ systems, making planning and organizing life difficult. Thus,there seems to be great need for evaluation of the effects of this type of flexibility andfor establishing reasonable limits of flexibility.

On the other hand, individual flexibility would appear almost by definition tobe good, but no systematic evaluations have been taking place. It is our impressionthat there are applications where, for example, trust hours may be associatedwith excessively high goals and thus yield an excessive workload. Similarly,computer-aided selection of work schedules needs to contain limits against excessiveamounts of self-inflicted long or irregular working hours that may impair safety andhealth.

The analysis of the data of the Third EU Survey on Working Conditions (whichinvolved 21,505 workers, 11,815 men and 9,690 women, aged between 16 and 65 yrs)confirmed how widely differentiated working hours are at present, both in terms offorms of arrangements of working hours and in relation to way of employment,gender, age, job activity, work sectors, and country.

In general, more than one-third of self-employed and more than one-fifth ofemployed workers, both men and women, declared that they did not work the samehours every day and/or the same number of days every week and/or at fixed startingand finishing times. Moreover, about 38% said they were able to influence in someway their working hours and about 50% said they could choose holidays or a dayoff. However, longer and ‘‘irregular’’ working hours appeared to be, in general,associated with lower levels of health and well-being. As concerns the perceivedstress, for example (Table 1), there is a significant trend with increasing workinghours, a higher prevalence in shift and night workers, and progressive increase inrelation to Saturday, evening, and Sunday work, respectively; it increases also withmore irregularity/variability of working timetables, but without significantdifferences between genders (except for night work, which appears worse forwomen). Similar patterns were also recorded for overall fatigue, stomach ache,sleeping problems, and anxiety.

Central elements of the above mentioned definition of ‘‘flexible working hours’’are the facts that employers and/or employees and/or both have a continuous option

Flexible Working Hours, Health, and Well-Being in Europe 837

ORDER REPRINTS

for choice to adapt the working hours to the needs/requirements at hand. Thatmeans that central elements of ‘‘flexibility’’ are ‘‘variability’’ and ‘‘control,’’ with the(degree of) discretion available to the parties involved being relevant for thedifferentiation of company- or employee-based flexibility. The main question iswhether such different kinds of flexibility can be differentiated and whether theywould show different effects on health and well-being.

Table 1. Prevalence (%) of perceived stress according to working hours patterns (21,505

workers examined in Third European Survey on Working Conditions, 2000).

Self-employed Employed

Men Women Men Women

Working hours per week

<30 23.8 24.4 20.4 20.0

30–39 38.1 36.4 26.0 30.9

40 20.9 28.5 25.4 31.2

>40 33.5 32.6 36.1 38.1

Part-time

Yes 26.9 21.1 20.9 21.8

No 31.8 32.9 28.2 31.9

Overtime

No>10 h/d 21.1 24.3 22.4 24.9

Yes 36.5 38.4 36.2 42.4

Shift work

Yes 35.8 39.4 32.2 37.5

No 31.1 30.6 26.5 26.5

Night work

No 28.0 28.4 25.3 26.9

Yes 41.1 46.0 35.5 40.3

Sunday work

No 26.2 25.6 24.6 25.6

Yes 37.2 39.5 36.1 36.9

Evening work

No 19.2 19.5 21.8 23.7

Yes 35.7 37.7 34.2 36.0

Saturday work

No 21.7 23.1 24.3 26.5

Yes 33.9 33.5 31.5 31.1

Same hours every day

Yes 28.2 30.0 23.2 24.4

No 34.0 31.9 35.3 35.1

Same no. of days every week

Yes 30.7 31.6 25.5 26.3

No 33.3 30.3 35.1 35.9

Fixed start and finish times

Yes 29.1 29.7 24.5 26.2

No 32.7 32.3 35.1 35.1

838 Costa et al.

ORDER REPRINTS

According to these statements, we tried to build up and compare two variables

describing different forms of ‘‘flexible working hours,’’ the one we called

‘‘variability,’’ related to the possibility of not having the same number of hours

every day and/or the same number of days every week and/or fixed starting and

finishing times, and the other we called ‘‘flexibility,’’ that is the possibility of having

some control over the working hours, to take breaks when desired, and/or to be free

to decide when to take holidays or days off.In our opinion, ‘‘variability’’ can be viewed as an index of less rigid (or more

variable) working timetables, and synthesizes a form of flexible arrangement of

working hours generally more subjected to company control and decision; whereas,

‘‘flexibility’’ can be seen as an index of arrangement of working hours more

connected to individual discretion and control. The two variables are neither directly

nor inversely related to one other, but they appear to have a different impact on

workers’ health conditions.As can be seen in Table 2, showing the different prevalence of health problems

according to these two variables, both factors seem to influence significantly the

outcomes, but in opposite ways: ‘‘variability’’ in a negative and ‘‘flexibility’’ in

a positive way. In general, the more individually-based flexibility (‘‘flexibility’’)

appears to be associated with better health conditions; whereas, the more company-

based flexibility (‘‘variability’’) appears to be associated with decreased health

and well-being. This was confirmed by multiple logistic regression analysis (that

took into account several other factors dealing with personal, family, social,

and working conditions) on several health outcomes (e.g., overall fatigue, anxiety,

stomach ache, heart diseases) that showed also that the feeling of not being able to do

the same job at the age of 60 was mainly associated with high physical work load

(Odds Ratio or OR¼ 2.02) and lack of ‘‘flexibility’’ of working hours (OR¼ 1.67).That was confirmed also by a second approach using factor analysis and

analysis of variance to differentiate their independent and combined contribution.

Low ‘‘flexibility’’ (indicated by I� in Figure 2) and high ‘‘variability’’ (Vþ) of work

hours were both consistently associated with poor health and well-being, and

especially in combination, while low variability (V�) combined with high autonomy

(Iþ) was associated with the opposite, positive, effects (see Figure 2 for sleeping

problems as a prototypical result). Significance of differences between groups was

tested using a conservative one-way analysis of variance (due to the unequal cell

sizes) with a total n> 5,000 (depending on the specific health variables). Tests for

main effects for variability or flexibility were accomplished by t-test for independent

samples, with sample sizes> 8,000, resulting in highly significant main effects, all

p< 0.001) with remarkable effect sizes, as can be seen from Figure 2. This kind of

analysis revealed that these effects were not the result of shift work (Sþ, e.g.,

variability of working hours induced by shift work), because the factor analysis

provided orthogonal factor scores for shift work (S), Variability (V), and

Flexibility (I). As can be seen from Figure 2, the well-known effect of shift work

on sleep problems was reproduced by this analysis (see Sþ/S�), resulting in the

highest impairment under the combination of shift work, high variability

(independent of the variability induced by shift work), and lack of autonomy (Sþ,

Vþ, I�). Lowest impairment was associated with regular work, low variability,

Flexible Working Hours, Health, and Well-Being in Europe 839

ORDER REPRINTS

Table

2.

Prevalence

(%)ofhealthtroublesin

relationto

thepersonalconditionsof‘‘variability’’and‘‘flexibility’’ofworkinghours

(21,505

workersexamined

inThirdEuropeanSurvey

onWorkingConditions,2000).

‘‘Variability’’

‘‘Flexibility’’

Variable

Partiallyvariable

Fixed

Flexible

Partiallyflexible

Rigid

Backache

35.1

33.7

28.6

a25.9

a32.6

36.6

Headache

15.8

17.1

12.7

a12.1

a15.0

18.4

Stomach

ache

6.3

5.4

3.1

a3.6

a4.4

5.8

Respiratory

difficulties

4.1

4.0

3.9

3.1

a4.2

4.7

Heart

disease

2.0

1.2

0.8

a1.1

1.1

1.3

Injury

7.8

7.0

6.0

a4.7

a7.1

8.0

Stress

34.8

33.1

22.9

a26.5

a28.3

31.8

Overallfatigue

25.8

23.1

20.3

a19.5

a21.0

28.1

Sleepingproblems

11.5

9.6

5.3

a6.0

a8.1

9.6

Allergies

4.8

4.8

3.8

a3.3

a4.6

5.1

Anxiety

9.5

8.1

5.4

a6.6

a6.5

8.7

Irritability

11.9

11.3

7.0

a7.5

a9.2

12.5

Skin

problems

7.5

7.0

6.0

a5.2

a7.0

7.7

Healthandsafety

atrisk

32.1

28.0

25.2

a23.1

a27.3

32.6

WH

bdonotfitwellwithfamily/socialcommitments

32.4

21.8

12.5

a17.2

a17.7

24.1

Notable

todothesamejobwhen

at60yrs

old

46.1

43.1

42.8

a33.2

a44.2

54.5

Jobdissatisfaction

15.6

15.2

14.1

a8.9

a14.2

23.5

ap<0.01at�2test

foreach

factorandcondition.

bWH¼work

hours.

840 Costa et al.

ORDER REPRINTS

and high autonomy. This resulted as a consistent pattern with a number of health

indicators (for details see Costa et al., 2003).These findings were also illustrated by the six case studies conducted in different

countries that showed that:

The majority of the workers in the European Union have some form of

individual or company-based flexibility in working hours that varies with

country, economic sector, social status, and gender.

The economic sectors with highest flexible working hours are agriculture,

transportation, information technology, education, and health.

Overtime is still the most frequent form of company-based flexibility; about

40–60% of women and 60% of men work overtime rather regularly.

Overtime and weekend work have negative effects on the stress, sleep, and social

and mental well-being of the workers; total overtime and variable working

hours are associated with one’s opinion of being able to remain in the work

force until the age of retirement.

Individual flexibility alleviates (but does not compensate) the negative effects of

the company-based flexibility in working hours on subjective health, safety,

and interference of the working hours with social well-being; however,

individual flexibility is not associated with subjectively reported sickness

0 0,05 0,1 0,15 0,2< low high >

Figure 2. Sleeping problems, as related to shift work, variability, and influence (21,505

workers examined in Third European Survey on Working Conditions, 2000). (Legend:

V¼Variability of working hours; S¼ Shift and night work; I¼Personal Influence;þ¼

High;�¼Low).

Flexible Working Hours, Health, and Well-Being in Europe 841

ORDER REPRINTS

absence or one’s opinion of being able to remain in the work force untilretirement age. It should also be noted that even self-induced variability ofworking hours was detrimental to well being.

SOME SUGGESTIONS

The collected data point out the need of further systematic research studies toexamine in detail what flexible working hours are considered to be, what and wheretheir positive (as well as negative) effects are, and what regulation (orientation,means, actors) seems most reasonable.

In our opinion, to improve our knowledge it is now important:

To promote a general consensus on a valid (and possibly specific) definition offlexible working hours at the international level, and to acquire more detailedinformation and statistics on the prevalence of its different forms throughnational and international surveys.

To develop methods for the evaluation of the impact of flexible working hourson workers’ conditions, focusing more on human-centered outcome variableson health and well-being (long-term effects, more objective assessments),preferably by longitudinal studies with adequate time intervals accordingto the different effects monitored.

To analyze better the interactions between work load, flexible working hours,and work organization, and to support intervention studies with carefulcontrol of the effects.

To verify the appropriate reference periods for determining maximum workinghours (weekly and daily), rest pauses, etc.; at present the data are insufficientto support both what is stated in the EU Directive and other recommenda-tions (e.g., ILO) or alternatives, e.g., at the national levels.

To focus better on age/gender and flexibility (or other relevant individualdifferences, e.g., domestic situation) taking into account the increasingfeminization and aging of the work force.

To develop guidelines on nonimpairing or positive arrangements of flexibleworking hours according to ergonomic principles through participatoryplanning and publishing ‘‘best practice’’ examples showing that this ispossible.

To establish more links and exchange of information among people in charge ofwork organization (employers, managers, work councils, labor inspectorates)and social partners about the use of time and related effects to achieve goodintegration of working and nonworking hours.

To support for extending individual flexibility as a tool for (a) for companies toimprove their economic efficiency (optimize manning), (b) improvingworker’s autonomy and control of her/his working conditions, (c) thus

842 Costa et al.

ORDER REPRINTS

improving her/his working and living conditions, and (d) adding social value

to the quality of life in general.

To improve and increase research on the international level; since the market is

becoming more and more global, we have to improve our knowledge of the

effects of flexible working hours with respect to legal, economic, cultural, and

social differences.

ACKNOWLEDGMENT

The authors wish to thank Pascal Paoli of the European Foundation for the

Improvement of Living and Working Conditions (Dublin), for his kind support

concerning the dataset of the EU Survey on working conditions.

REFERENCES

Costa, G. (2003). Factors influencing health and tolerance to shift work. Theor.

Issues in Ergon. Sci. 4:263–288.Costa, G., Akerstedt, T., Nachreiner, F., Carvalhais, J., Folkard, S., Frings

Dresen, M., Gadbois, C., Gartner, J., Grzech Sukalo, H., Harma, M.,

Kandolin, I., Silverio, J. (2003). As Time Goes by—Flexible Work Hours,

Health and Wellbeing, Working Life Research in Europe Report No. 8;

The National Institute for Working Life: Stockholm.European Council Directive 93/104/EC. (1993). Concerning certain aspects of

working time. Off. J. Eur. Commun. L307:18–24.Folkard, S., Lombardi, D. A. (2004). Toward a ‘‘risk index’’ to assess work

schedules. Chronobiol. Int. 21(6):1063–1072.Giebel, O., Janßen, Schomann, C., Nachreiner, F. (2004). A new approach for

evaluating flexible working hours. Chronobiol. Int. 21(6):1015–1024.Goudswaard, A., de Nanteuil, M. (2000). Flexibility and Working Conditions: a

Qualitative and Comparative Study in EU Countries. Dublin: European

Foundation for the Improvement of Living and Working Conditions.Ingre, M., Kecklund, G., Akerstedt, T., Kecklund, L. (2004). Variation in sleepiness

during early morning shifts: a mixed model approach to an experimental field

study of train drivers. Chronobiol. Int. 21(6):973–990.Knutsson, A. (2004). Methodological aspects of shift-work research. Chronobiol. Int.

21(6):1037–1047.Nakata, A., Haratani, T., Takahashi, M., Kawakami, N., Arito, H., Kobayashi, F.,

Fujioka, Y., Fukui, S., Araki, S. (2004). Association of sickness absence

with poor sleep and depressive symptoms in shift workers. Chronobiol. Int.

21(6):899–912.Paoli, P., Merllie, D. (2001). Third European Survey on Working Conditions 2000.

Dublin, Ireland: European Foundation for the Improvement of Living and

Working Conditions, Loughlinstown, Co.

Flexible Working Hours, Health, and Well-Being in Europe 843

ORDER REPRINTS

Portela, L. F., Rotenberg, L., Waissmann, W. (2004). Self-reported health and sleepcomplaints among nursing personnel working under 12 h night and day shifts.Chronobiol. Int. 21(6):859–870.

Rajaratnam, S. M. W., Jones, C. B. (2004). Lessons about sleepiness and drivingfrom the Selby rail disaster case: R v Gary Neil Hart. Chronobiol. Int.21(6):1073–1077.

Santos, E. H. R., deMello, M. T., Pradella-Hallinan, M., Luchesi, L., Pires, M. L. N.,Tufik, S. (2004). Sleep and sleepiness among Brazilian shift-working busdrivers. Chronobiol. Int. 21(6):881–888.

844 Costa et al.